Cover Image
close this bookAgroforestry in the Pacific Islands: Systems for Sustainability (UNU, 1993, 297 pages)
View the documentPreface
Open this folder and view contents1 Introduction
Open this folder and view contents2 Pacific Island agroforestry: Functional and utilitarian diversity
Open this folder and view contents3 Agroforestry in Melanesia: Case-studies from Papua New Guinea and the Solomon Islands
Open this folder and view contents4 Agroforestry in Melanesia: Case-studies from Vanuatu and Fiji
Open this folder and view contents5 Agroforestry in Polynesia
Open this folder and view contents6 Agroforestry in Micronesia
Open this folder and view contents7 Pacific Island urban agroforestry
View the document8 Agroforestry on smallholder sugar-cane farms in Fiji
Open this folder and view contents9 Institutional agroforestry in the Pacific Islands
Open this folder and view contents10 Agroforestry in the Pacific Islands: Systems for sustainability
View the documentAppendix One hundred Pacific Island agroforestry trees (1)
View the documentAppendix One hundred Pacific Island agroforestry trees (2)
View the documentAppendix One hundred Pacific Island agroforestry trees (3)
View the documentAppendix One hundred Pacific Island agroforestry trees (4)
View the documentAppendix One hundred Pacific Island agroforestry trees (5)
View the documentAppendix One hundred Pacific Island agroforestry trees (6)
View the documentReferences (A-E)
View the documentReferences (F-R)
View the documentReferences (S-Z)
View the documentContributors

Preface

This study examines the diverse traditional systems of agroforestry that have evolved over millennia in the Pacific Islands. Attention is also given to present-day urban agroforestry, to agroforestry as practiced in conjunction with monocultural cash-cropping, and to the modern agroforestry projects promoted by governments and international funding agencies. The study is based on an aggregate of several decades of research on Pacific Island agro-ecosystems by the five contributors. From 1982 through 1989, field research for the study was funded by the United Nations University.

We intend the study to be a contribution to an increased understanding of the diversity and utility of traditional agroforestry systems in the Pacific. Our further purpose is to provide an inventory of Pacific trees and a description of Pacific Island agroforestry systems that might serve as bases for further development or re-development of Pacific agroforestry in the face of pressures for agrodeforestation.

We would like to thank the United Nations University for funding the research project "Pacific Island Agrosilviculture Systems: A Basis for Sustained Development and Environmental Protection" and also for providing funds for final bibliographic work and consultation between the editors. We are also grateful to Lee MacDonald, UNU Programme Officer at the time of the initiation of the project, to Kathleen Landauer, who served in the same position later, and to Professors Walter Manshard and Roland Fuchs of the UNU for their patience, understanding, and support over the long life of the project. We would also like to thank the many practicing agroforesters (professionals and Pacific Island villagers and farmers) and other per sons throughout the Pacific who have so freely shared their knowledge. Thanks are also due to the University of the South Pacific for its continued assistance and institutional support. Particular thanks to Sheila Singh, Sharon Bing, Iliana Seru, Marica Bolabola, and Sharon Smith-McGowan, who typed the numerous drafts with their daunting lists of scientific and vernacular plant names.

W.C. Clarke and R.R. Thaman, The University of the South Pacific, Suva, Fiji

Context of the study

In much of the tropical world, agriculture and forests compete for the use of land, with agriculture usually the winner. The defeat of forests in this competition is the continuation, at a faster rate than ever before, of one of the most ancient relationships between humanity and the environment - the replacement of wildlands with human domesticates, including humans themselves (Janzen 1990, xi). Now, because a near total loss of tropical forest is foreseen and because newly cleared as well as older agricultural lands suffer increasing degradation, governments, funding agencies, and scientists seek ways to rehabilitate the damaged interface of agriculture and forests and to prevent further deterioration.

To achieve these goals, attention has turned strongly to "agroforestry," a term and concept that became widely known after the publication in 1977 of the seminal work Trees, Food and People: Land Management in the Tropics by Bene, Beall, and Cote, who recommended the establishment of what became the International Council for Research in Agroforestry (ICRAF). Since then, interest in agroforestry has greatly increased, and the concept has now been brought into prominence under the banner of "sustainable development." Researchers now specialize in the discipline of agroforestry; many conferences have focused on agroforestry; and a rapidly growing literature has been published - e.g., the ICRAF journal Agroforestry Systems, many articles in other journals, and at least 16 books, manuals, or conference proceedings in the past 6 years (Baumer 1990;

Beets 1989; Buck 1989; Budd et al. 1990; Gholz 1987; Gregersen et al. 1989; Kartasubrata et al. 1990; MacDicken and Vergara 1990; Nair 1989b; Raintree 1987a, 1987b; Rocheleau et al. 1989; Steppler and Nair 1987; Vergara and Briones 1987; Withington et al. 1988; Wood and Burley 1991).

Examination of this literature reveals that there are two distinct, although not always entirely separated, approaches to agroforestry: the modern, institutional; and the traditional, or indigenous. Institutional agroforestry relies on modern agronomic science and field experimentation for such purposes as assessing yields of crops, trees, and animals in varying combinations and under varying conditions; determining competition or complementarily of system components; seeking maximization of aggregate production; or determining which leguminous tree species produce the green manure supportive of the greatest nitrogen mineralization. The approach that focuses on traditional or indigenous agroforestry (Denevan et al. 1984) arises from cultural geography and ecological anthropology; it seeks to record the attributes of traditional or non-institutionalized agroforestry systems that are in use now and that may have been practiced by tropical peoples for centuries or millennia. Some cultural-ecological studies derive purely from ethnographic interest but most are now also motivated by the belief that the traditional or non-institutionalized systems are worthy of preservation and dispersal. The present study, written by geographers, mainly follows the cultural-ecological approach rather than the quantitatively analytic agronomic approach, but our interest extends beyond the ethnographic to the applied, for we believe that the past and present agroforestry systems that have been empirically developed by Pacific peoples have much to offer in the search for sustainable development.

A merit of the indigenous systems is that they are based on indigenous trees, which are familiar and useful to local people and already adapted to local climatic and soil conditions (Montagnini 1990, 5051). Further, though any particular local agroforestry system probably yields less than its potential maximum, its technologies are adapted to local conditions by virtue of empirical experimentation and have already been adopted by local people. In many cases, the system could be improved by tinkering based on analytic agronomic knowledge but it does not have to be invented anew. It is in place and it works.

The value of the indigenous systems is acknowledged by most agroforestry researchers, and the antiquity of agroforestry among tribal and peasant peoples is widely recognized (e.g., King 1989). One of ICRAF's projects, launched in the early 1980s, was a systematic inventory of agroforestry systems (Nair 1987): an effort that showed "there was a bewildering array of agroforestry systems worldwide" with more than 2,000 species of multi-purpose trees in use (Steppler 1987, 15). None the less, most government-supported, aid-funded projects in agroforestry follow the institutional approach, which usually requires the introduction of unfamiliar, non-indigenous trees and associated slow, complex experimentation. In discussing the difficulty of gaining scientific understanding of the large number of agroforestry systems, Steppler (1987, 17) noted:

Any one system undergoing experimentation would include, at a minimum, a tree species and a crop species. Each of these could have variation in genotype and management such as spatial arrangement, maturity type for crop and harvesting methods (e.g., lopping and coppicing timing for the tree). It quickly becomes apparent that we are dealing with a multifactor design with many combinations. As we add species of trees and/or crops or introduce animals, the experiment grows in size logarithmically.... The other dimension to the problem is the fact that we have combined long-lived woody perennials with annuals, short-lived perennials and/or animals. Ideally, experiments should continue for the life of the longest-lived component; this could be upwards of 40 years and we cannot wait that long. Thus, we must also devise tests and methods of prediction that will have acceptable levels of confidence in predicting long-term effects.

These sorts of difficulties suggest to some planners and students of agroforestry that, while there is a need to build institutionalized analytic capacity, it should be recognized that "many of the problems which are addressed by agroforestry do not have clear and explicit objectives associated with them. That is, the task is not necessarily one of solving a problem, but rather one of establishing a context in which social learning can take place" (Budd et al. 1990, 332-333).

The existence of so many site-specific indigenous agroforestry systems and the information that is available about them makes it less necessary to seek "valid" scientific findings by means of long experimentation with a fixed set of variables according to a strict problem-solving paradigm. As Nair (1984, 73) says, such luxuries cannot be afforded in agroforestry research at this stage. Existing indigenous systems already satisfy, to varying degrees, the three basic criteria that Raintree (1990, 58) argues should be built into the design of a good agroforestry system: productivity, sustainability, and adoptability. That the systems have remained in use shows that local farmers are convinced of the adequacy of their production in relation to inputs. The presence of trees in the system lengthens the time-span of production and protection and provides an inherent basis for sustainability. That the local people have already adopted the system meets the requirement - increasingly recognized as basic to the success of any project to ameliorate living conditions - that the intended users accept the project and be active participants in its design, trial, evaluation, and redesign. The indigenous systems have already been well tested by local farmers from season to season and generation to generation and so provide a strong locally based framework for sound management and incremental agro-ecological innovation (Gliessman 1990, 36; Richards 1985).

In their review of institutional agroforestry in the South Pacific region, Vergara and Nair (1985, 377) comment that, "in general, trees may be considered suitable to agroforestry if they complement and support rather than compete with the interplanted food crops.... Unfortunately, out of the over 2000 species that satisfy these characteristics, only a handful have been tested and used in agroforestry, such as Leucaena, Albizia, Gliricidia, and Calliandra. The rest remain untried and therefore their potentials unrealized."

In contrast, our point is that in the Pacific- and elsewhere, as in the Peruvian Amazon (Denevan and Padoch 1988; Denevan et al. 1984) - many or most of the indigenous trees suitable to agroforestry have already been tried and their potentials at least partially realized in traditional systems, even if they have not yet been the subject of institutional agronomic experiments.

Our purpose in this book is to tap this already-existing knowledge by offering details on the wide range of non-institutional, informal agroforestry systems practiced in the Pacific Islands, now or previously. We hope this information will be of applied use in the further adaptation and diffusion of agroforestry practices, whether institutional or developed informally. We do include some details about institutional agroforestry developments in the Pacific (chapter 9), but emphasize the traditional or indigenous systems because we believe that Pacific landscapes themselves and the agroforestry systems they already contain are the most useful drawing-board from which to move agroforestry into the landscapes of the future.

Geographical background

The Pacific Islands are, for the purposes of this study, defined as the islands of Melanesia, Micronesia, and Polynesia (excluding Hawaii and New Zealand) (see map p. 5). The study area includes the large continental island of New Guinea in the west and extends to the small atolls and recent volcanic islands of the central and eastern Pacific, where traditional agroforestry systems remain common in unchanged or only slightly modified forms. Some of the political units are single islands rather than island groups (e.g., Nine and Nauru), or territories rather than independent countries (e.g., New Caledonia, French Polynesia, American Samoa, and Guam). For simplicity's sake, they will be included in discussion of "groups" or "countries."



The Pacific islands

The study area includes a diversity of island types:

  1. Continental islands, such as New Guinea and New Caledonia, which are composed of geologically-ancient sedimentary, metamorphic, and igneous rocks of continental origin.
  2. Andesitic-arc islands, such as most of the islands of the Solomon Islands, Vanuatu, Fiji, Guam, and the Northern Marianas, as well asthe eastern islands of Papua New Guinea (e.g., Manus, New Britain, and Bougainville) and the more recent volcanic islands of Tonga (e.g., Niuatoputapu, Niuafo'ou, Kao, and Tofua). Andesitic-arc islands have been formed by recent andesitic volcanic activity in proximity to the subduction zone (previously, but now inappropriately, referred to as the Andesite Line), where the newer crust of the Pacific Plate and the older crust of the Indo-Australian Plate come together, with one being forced (subducted) under the other.
  3. Basaltic volcanic ("hot-spot") islands, which are high oceanic islands such as the Samoas, the Cook Islands, Tahiti in French Polynesia, the Hawaiian Islands, and Pohnpei (Ponape) and Kosrae in the Federated States of Micronesia. Hot-spot islands have been formed as the result of the extrusion of magma through cracks or rifts in the Pacific Plate as it moves over "hot spots" in the Earth's mantle.
  4. Raised limestone islands such as Nauru and Nine, the islands of Vatulele in Fiji, Aniwa in Vanuatu, Ouvea in the Loyalty Islands of New Caledonia, most of the Tongan group, and portions of Mangaia and Atiu in the southern Cook Islands and Vanuabalavu in Fiji. These islands, formed of old-reef and foraminiferous limestone, have experienced considerable uplift relative to sea level.
  5. Coral atolls is a term that technically refers to the roughly circu lar coral-reef structures at, or just below, or slightly above, sea-level that usually support scattered reef islets (motu) and that surround a central lagoon. Most of the islands in the Marshall Islands, Kiribati, Tokelau, and the Tuamotus in French Polynesia are atolls. Often included in the category are reef islets or "table reefs," such as Arorae and Tamana in Kiribati and Nukulaelae in Tuvalu, which have no lagoons and should probably be categorized as raised coral-limestone islands. Reef islets would also include islets on barrier reefs surrounding larger islands, but clearly separated from the main island by a lagoon. Similarly, some barrier reefs and associated reef islets, such as those around the central volcanic peaks and lagoons of the islands of Aitutaki in the Cook Islands and Bora Bora in French Polynesia, take the form of atolls. These islands are commonly referred to as "almost atolls" because remnants of the volcanic peak that forms the foundation of all atolls still remain emergent above sealevel in the central lagoon.

It must be stressed that, just as there are continental or andesitic islands very far from continental shores in the Pacific (e.g., Easter Island), there is also basaltic volcanic activity close to subduction zones that cannot be reconciled with the hot-spot model and remains largely unexplained (e.g., Taveuni in Fiji). Similarly, raised limestone islands, atolls, and reef islets can be found on both sides of the subduction zone, thus adding considerable ecosystemic and environmental diversity, with most island groups including more than one island type (table 1).

There is also great geographical and demographic diversity among the islands. Easter Island, Guam, Kosrae, Nauru, and Niue consist of a single small island; Fiji, Tonga, French Polynesia, and Hawaii consist of hundreds of large and small, widely dispersed islands; Papua New Guinea and Irian Jaya share the very large, high continental island of New Guinea, and both include many smaller offshore islands. Total land areas vary from 10 to 26 sq km for groups of low-lying, coral-limestone islands like Tokelau and Tuvalu to over 400,000 sq km for the continental island areas of Irian Jaya and Papua New Guinea (Thaman 1988a).

Population densities for entire groups range from just over 1 person per sq km for the Galapagos and Pitcairn Island and 2.5 for Irian Jaya, to almost 300 or more for Nauru, Truk, and Tuvalu. If the "most populous islands" are considered, the figures jump to over 100 persons per sq km for four islands, and over 200 for three islands;

Table I Types of islands up selected island nations and territories of the tropical Pacific Ocean. (Some individual islands may be composite, combining more than OK island type) and are 421 for Koror in Palau, 757 for Funafuti in Tuvalu, 1,179 for Majuro in the Marshall Islands, and 2,190 for Tarawa in Kiribati. The estimated population for Betio Islet of Tarawa atoll is expected to reach 34,066 by 1993, which will give it a population density of 4,705 per sq km, thus rivalling the population densities of Hong Kong (Carter 1984, 231). If we consider Ebeye, one of some 90 islets of Kwajalein Atoll in the Marshall Islands, to which people have been relocated by the US military to free the atoll's lagoon for intercontinental ballistic-missile testing, the population density sky-rockets to 25,000 per sq km (Keju and Johnson 1982)!

 

Nation or territory Continental Andesitic arc High basaltic Raised limestone Coral Atoll
Irian Jaya + +   +  
Papua New Guinea + +   + +
Torres Strait + +   +  
New Caledonia + +   + +
Solomon Islands   +   + +
Vanuatu   +   + +
Fiji   + + + +
Tonga   +   +  
Wallis and Futuna     + +  
Western Samoa     +    
American Samoa     +   +
Nine       +  
Tuvalu       + +
Tokelau         +
Cook Islands     + + +
French Polynesia     + + +
Pitcairn Island     +   +
Hawaii     +   +
Easter Island     +    
Galapagos Islands     +    
Palau       +  
Guam   +   +  
Northern Marianas   +      
Yap     +   +
Truk     +   +
Pohnpei     +   +
Kosrae     +    
Nauru       +  
Marshall Islands         +
Kiribati       + +

Sources: Personal observation; Carter 1981, 1984; Dahl 1980; Klee 1980b; Thaman 1988a.

This range of diversity in island types and population densities when combined with differences in climate, geological resources, topographical features, soil types, water availability, flora and fauna, and culture - goes a long way to explain the diversity of agroforestry systems found in the Pacific Islands. The more specific nature of the physical and biological resources of individual island ecosystems, and the extent to which they are currently protected or endangered by exploitation, have been comprehensively analysed by Arthur Dahl (1980) in his Regional Ecosystems Survey of the South Pacific Area, which contains descriptions and the conservation status of all marine and terrestrial ecosystems and physical and biological features or resources of particular ecological and cultural importance; as well as lists of rare, endemic, or endangered species; existing and proposed conservation legislation; and existing, proposed, and recommended reserves for each island group.

It is beyond the scope of this study to present such detailed information, although some of it will be considered in the case-studies of individual agroforestry systems. For detailed data on various aspects of island groups see Bakker (1977a, 1977b), Brookfield with Hart (1971), Carter (1981, 1984), Dahl (1980), Douglas and Douglas (1989), McArthur (1967), Thaman (1988e), Ward and Proctor (1980), and Winslow (1977).

Definition of terms

As noted earlier in this chapter, agroforestry is a new name for an old practice. As the word and concept became widely accepted in international land-use circles, many definitions of the term were put forward, as described in detail by Nair (1989a). The definition of agroforestry that ICRAF has used since the early 1980s is as follows (Lundgren 1987, 48):

Agroforestry is a collective name for all land-use systems and practices in which woody perennials are deliberately grown on the same land management unit as crops and/or animals. This can be either in some form of spatial arrangement or in a time sequence. To qualify as agroforestry, a given land use system or practice must permit significant economic and ecological interactions between the woody and non-woody components.

Of the many other definitions of "agroforestry," one of the most comprehensive is that of King and Chandler (1978) in an early ICRAF publication The Wasted Lands. Recently reproduced by Nair (1989a, 13), the definition reads: "Agroforestry is a sustainable land-management system which increases the overall yield of the land, combines the production of crops (including tree crops) and forest plants and/or animals simultaneously or sequentially, on the same unit of land, and applies management practices that are compatible with the cultural practices of the local population."

Along similar lines, for the purposes of this report, "agroforestry" is defined as: "The deliberate incorporation of trees into, or the protection of trees within, an agro-ecosystem in an effort to enhance its short- and longterm productiveness, its economic and cultural utility, and its ecological stability."

In this context, an "agroforestry system" is defined as: "Any agricultural system (agro-ecosystem) in which planted or protected trees are seen as economically, socially, or ecologically integral to the system."

These non-restrictive and functional definitions have been selected because they can cover the great diversity and functional utility of existing Pacific Island agroforestry/agricultural systems, which range from home-garden or household and squatter-garden agroforestry in both urban and rural areas to deliberate intercropping and the protection of trees and tree-like perennials in gardens and pastures and the planting of woodlots and protection of inland and coastal forest stands (which are seen as part of integrated agroecosystems) in sparsely populated rural areas.

Finally, the new term, "agrodeforestation," is introduced, defined by Thaman (1988b, 1988c, and 1989a) as: "The removal of trees or the deemphasis on the planting and/or protection of trees in agroecosystems."

Deforestation and agrodeforestation in the Pacific

Pacific Islanders, like people everywhere, "prospered by disturbing the natural order," as Carl Sauer (1952, 3-4) put it. The pioneering Pacific mariners found islands almost entirely covered by a mosaic of natural forest types. As the islands were occupied, the newly arrived settlers cleared forest for gardens and established orchards or agroforests that provided many valuable foods and materials. People also opened forest land to provide materials and space for houses. The ubiquitous use of fire, often for hunting, was a major tool in the change from forest to more open landscapes. These activities modified the natural landscape, creating a human habitat that was more congenial to occupation and much more productive of food than were the closed native forests. But as Sauer went on to say, human beings often overreach themselves, and the new order they introduce may contain the seeds of disaster. Or, as Oedekoven (1962, 55) suggested with regard to forests, humankind may ultimately cut off the branch it is sitting on.

In the Pacific before European contact, human activities caused many kinds of degradation. Deforestation has been prevalent in Pacific history; subsequent repeated burning has been responsible for the evolution of fire-climax forests, grassland savannas, and degraded fern and scrub lands (Clarke 1965; Farrell 1972; Manner 1981; Thaman and Clarke 1983). Such a process has undoubtedly been the main cause of the extensive anthropogenic grasslands of highland New Guinea; the xerophytic niaouli (Melaleuca leucadendra) savanna lands of New Caledonia; the highly degraded "sunburnt lands," or talasiga, found throughout Fiji; and the rapidly expanding saafa (Panicum maximum) grasslands of Tongatapu in Tonga.

Deforestation has led to severe erosion in Wallis and Futuna, the Cook Islands, French Polynesia, and Hawaii, where most of the indigenous forest has been removed, leaving degraded fern lands and grasslands no longer suitable for agriculture (Kirch 1982, 4). Flenley and King (1984) go as far as to suggest that deforestation was responsible for the collapse of the pre-European megalithic culture on Easter Island, a view supported by McCoy (1976, in Kirch 1982, 4), who argues that the "radical reduction of forest, shrub, and grassland communities, following over-exploitation and misuse by man," was responsible for a change from open-field cultivation to protected stone garden enclosures (manavai). Similarly, drastic deforestation of the central plateau on the Hawaiian island of Kahoolawe because of shifting cultivation and increasing population pressure between A.D. 1375 and 1600 reportedly led to a "dramatic population crash" and the total abandonment of the interior of the island by 1700 (Hammon 1980; Kirch 1982, 4).

It is clear that the Pacific Islands' early inhabitants did not avidly practice a conservation ethic that preserved their habitat as an unchanging paradise until Europeans brought major disturbances and degradation; instead, the early settlers caused many extinctions (notably of birds), reduced forest cover, initiated massive soil erosion, and created or extended degraded grasslands. In short, they did what all peoples, especially pioneers, do in their efforts to make a living: they actively manipulated, modified, and at times degraded the ecosystems in which they lived (Clarke 1991; Kirch 1984, 123-151).

But in their transformation of natural landscapes into cultural landscapes, the early inhabitants of the Pacific also developed - partly as an adjustment to the degradation they had caused - sustained-yield systems of agriculture, agroforestry, hunting, gathering, and fishing that still operate productively today but that are in danger of disappearing in the face of changing technological, social, demographic, and economic conditions.

The natural forests of the Pacific Islands fall into the general categories of lowland tropical rain forest, montane forest, swamp forest, mangrove forest, or coastal-strand communities. Locally, many Pacific forests are unique; their high endemism and fragility have attracted the attention of scientists, beginning with the first comprehensive biological studies on Cook's voyages during the eighteenth century and Darwin's observations on his voyage on H.M.S. Beagle in the 1830s. The forests' vulnerability to human-induced change has been stressed by Darwin (1895), Fosberg (1965), Mangenot (1965), Carlquist (1965, 1980), Mueller-Dombois (1975), Dorst (1972), and Dasmann et al. (1973). All categories of remaining natural forest are increasingly endangered as part of the worldwide process of deforestation resulting from urbanization, industrialization, commercial logging, agricultural development projects, and increasing population. As mentioned above, Dahl (1980) provides descriptions and conservation status for all terrestrial ecosystems in the South Pacific as well as lists of rare, endemic, or endangered species. Dahl also provides details for each island group or biogeographical province of the current and proposed conservation legislation, and of the existing, proposed, and recommended reserves.

All parts of the Pacific have ecologically and culturally important forest types or individual species that are in danger of depletion by human action. Some countries and territories have conservation legislation and forestry ordinances (Pulea 1984); Papua New Guinea and Hawaii have increasingly effective systems of forest reserves and conservation areas; and other places, such as New Caledonia, the Solomon Islands, Vanuatu, Norfolk Island, Fiji, French Polynesia, Kiribati, and some of the American territories, have had similar developments recently on a more limited scale. However, forest products continue to be shipped off for an inadequate return, while Japan, South Korea, China, and other countries continue to protect their forest resources and to implement major reforestation efforts (Richardson 1981). In New Caledonia, for example, where nearly all exploited timber species are endemic, most of the surprisingly rich native gymnosperm flora of 44 endemic species are now limited to a few restricted habitats. Most of the 13 species of Araucaria are restricted to active mining areas, and the local kauri species, Agathais lanceolata, has been exploited to near-extinction in southern New Caledonia (Dahl 1980, 37). Similarly, Agathis macrophylla, formerly abundant on Aneityum in Vanuatu, has been almost logged out. Selective unrestricted cutting, sometimes for shipping as saw logs, also threatens the Fijian form of this stately species, mature individuals of which may be centuries old.

Deforestation is proceeding rapidly in most of the Pacific. Forests, both primary and secondary, continue to be transformed into degraded savannas and fern-grasslands, mangroves into housing and industrial estates or other lifeless land-sea interfaces; and polycultural, treestudded, traditional agroforested gardens into monocultural plantations. Urban areas lose trees to make way for industrial, commercial, and residential areas or to fuel the cooking fires or to erect the squatter housing of low-income families. The trends are the same from the high continental islands of Melanesia to the smallest atoll islets of Polynesia and Micronesia.

Although deforestation, seen as the loss of forest as such, has received much more attention, "agrodeforestation" is probably of tantamount importance culturally and ecologically. Fewer trees are planted, and the great variety of useful tree species in gardens, villages, and towns is suffering depletion. The situation is particularly serious on smaller islands with little or no remaining native forest, where agricultural areas and home gardens serve as the few reserves where endangered plant varieties or cultivars can be protected. In Tonga, for example, during the height of the banana boom, so many trees were cut to provide shooks for banana boxes, and to extend banana plantings, that sawmillers had to move from Tongatapu to the nearby island of 'Eua. Thus, the search for meagre export earnings diminished valuable native species as well as food-bearing trees such as mango and citrus cultivars (Thaman 1976).

Most of the trees that now provide food, timber, firewood, and medicines, or that serve other cultural and ecological functions in Pacific agro-ecosystems, were deliberately planted or protected in the past. But few of them are being replaced or protected by the present generation. Opening a tin of imported peaches for a feast, going to the local dispensary or pharmacy for medicines, or purchasing imported plastic flowers, perfumes, and deodorants, have replaced the products that came from trees. Of particular concern is the ubiquitous senility of Pacific Island coconut palms, the only source of export income on many of the smaller, more isolated islands, as well as a very important source of food, drink, and materials. Despite limited replanting, the declining yield of palms, often planted before the turn of the century, augers poorly for the economic future of these areas. On the other hand, areal expansion of coconuts, as currently promoted, is commonly at the expense of other land use-soften arboreal. A singular case was that described by Spoehr (1949) in the Marshall Islands, where the Japanese ordered the removal of breadfruit trees so that copra production could expand, thus lessening arboreal diversity and eliminating a tree that produced food, medicine, canoe hulls, and caulking. Similarly, in Kiribati, only recently has the Government acknowledged that some 20 years of institutionally-sponsored coconut replanting and rehabilitation have led to the gradual elimination of a wide range of ecologically and culturally important tree species, all traditionally components of the Kiribati agricultural system (Thaman 1989b).

Although some countries have increasingly effective systems of forestry reserves, conservation areas, or national parks, few, if any, have legislation or programmes prohibiting the cutting - or promoting the replanting - of endangered tree species as part of agricultural development. Thus, agrodeforestation continues with little or no official recognition and, therefore, few attempts to reverse the trend (Thaman 1989a).

Aside from the loss of materials and ecological services, which will be discussed in detail in the next chapter, agrodeforestation also means a cultural loss because a significant part of Pacific intellectual heritage is an intimate feeling for the social and spiritual meaning of trees, together with an immense knowledge of their habits and products. As the trees disappear, traditional knowledge is eroded, and landscapes lose the depth of meaning imbued by protected or planted trees. Although commonly useful trees are in no immediate danger of becoming extinct because of agrodeforestation, biodiversity is diminished because many agroforest species contain a great number of varieties or cultivars, each with its own characteristics. If these varieties, which are the result of generations of careful selection, are lost, the local food-production system will be degraded. For instance, in a recent study, Fownes and Raynor (in press) report that as many as 130 cultivars of breadfruit (Artocarpus altilis) are recognized by farmers on the volcanic Micronesian island of Pohnpei, where breadfruit collected from a traditional agroforestry system is a major staple food for people during several months of the year and is also a major food source for pigs, which roam the understorey consuming fallen fruit during the peak season. Fownes and Raynor investigated five cultivars (a small fraction of the recognized number) and found that they varied in seasonality, growth form, and yield. Complementary seasonality among the cultivars led to an extended fruiting season in the aggregate.

The Pacific remains fortunate because many such traditional agroforestry strategies still exist, if only in relict form. None the less, increasing agrodeforestation and the gradual disappearance of time-tested agroforestry systems and their component species and varieties in the face of the expansion of monocultures and commercial livestock, population growth, increasing demands for fuel, continued urbanization, and the "commercial imperative" (fudge 1977) are the dominant trends, which will only be reversed by deliberate planning and action. In an attempt to help facilitate such planning and action, this study emphasizes throughout the roles that particular traditional and existing agroforestry systems and their component trees play, and could continue to play, in the provision of useful materials, the enhancement of the environment, the maintenance of the stability of agroecosystems, and the reversal of deforestation and agrodeforestation.

Organization of the study

Following this Introduction, chapter 2 examines agroforestry in the Pacific generally, and with particular regard to its functional and utilitarian diversity. Chapters 3, 4, 5, and 6 present case-studies of specific agroforestry systems grouped according to the long-standing geographical and ethnographic division of the Pacific Islands into Melanesia, Polynesia, and Micronesia. This division is not meant to suggest that each of these three regions has a distinct "agroforestry environment." Although only Melanesia contains continental islands, all three regions contain all the other four kinds of islands: andesiticarc islands, high volcanic islands, raised limestone islands, and coral atolls (table 1). Nor does the division into Melanesia, Polynesia, and Micronesia reflect any rigidly distinct contrast in flora, crops, or agri culture in general. The division is used because it is widely familiar and it provides some convenience in discussion and research. Moreover, certain distinctions can be made in agroforestry practices from region to region, as will be discussed in appropriate chapters.

Following the case-studies of agroforestry in the rural Pacific, most of which remains at least partially subsistence-based, attention is turned to urban agroforestry and to agroforestry practiced in conjunction with the intensive cash monoculture of sugar cane in Fiji. The penultimate chapter examines institutional agroforestry in the Pacific - that is, the more formal agroforestry activities that are promoted by governments, companies, and various agencies, and that involve external funding, training, agronomic research, and extension services. Also briefly described in that chapter is the status of education about agroforestry in the Pacific's universities and the work undertaken by some scientific research organizations. The final chapter offers general conclusions and recommendations having to do with agroforestry in the Pacific. In the Appendix, information about the characteristics of 100 important Pacific Island agroforest species is drawn together. Although the total number of tree or tree-like species found in use in agroforestry systems in the Pacific is more than 400, the more modest annotated listing of 100 species is certainly sufficient to give a clear indication of the remarkable richness of the agroforestry resource already available in the Pacific.

Integration and sustainability

In traditional Pacific Island societies, aspects of living such as for estry, agriculture, housing, medicine, and tool making were not com partmentalized into economic sectors. Instead, they were parts of an integrated system of production tailored to the environmental condi tions and material needs of each island society. As most of the sys tems that evolved in the Pacific contained both annual crops and trees, they were true agroforestry systems. While not unchanging, these tree-rich systems had a high degree of stability and would fit into Janzen's category of sustained-yield tropical agro-ecosystems (SYTA). The "resilient permanence" of these traditional Pacific agro-ecosystems rested on seven "principles of permanence" that made possible their continuing operation for centuries or millennia.

The systems (Clarke 1977)

  1. did not depend on external energy subsidies or extra-system nutrient sources - i.e., no imported fuel, fertilizers, or other imports were required;
  2. did not receive applications of poisonous agricultural chemicals or eutrophic materials to pollute the environment;
  3. had strongly positive net energy yields - i.e., for every joule of energy invested, 18-20 joules of food energy were returned;
  4. used only renewable resources as inputs - e.g., trees for fencing, ash as fertilizer- rather than imported, often non-renewable, inputs such as inorganic fertilizers derived from phosphate deposits or fossil fuels that took millions of years to form;
  5. were structured so that the resources supporting agriculture (energy, land, vegetation) were equitably spread throughout the community rather than being concentrated in the hands of a few or in urban areas;
  6. contained resources that were looked upon as productive capital to be preserved - i.e., attempts were made to preserve for future generations a habitat and set of resources only slightly modified from what parents had themselves inherited; and
  7. were based on polyculture and a diversity of tree and non-tree crops, wild plants, and animals rather than on monoculture or on specialized animal production.

Diversity of function

In terms of the more specific attributes of individual Pacific agroforestry systems, table 2 shows the many functions of these systems as well as the value of the individual arboreal components. Although modern resource developers see economic value and, possibly, even ecological, recreational, or nutritional value in native forests, in silvicultural tree plantings, in plantations of coconut, oil-palm, cocoa, coffee, or banana, and in orchards of orange, avocado, or macadamia, it is clear that Pacific Island agroforesters perceived arboreal resources to be far more pervasive in the landscape and still more multi-purposeful .

To emphasize the variety of arboreal functions in Pacific Island agrosilvicultural systems, we provide the following examples from the immense list possible.

Shade

Shade provides valuable protection to humans, plants, and animals, especially in savannas and in highly reflective low-lying coral island and lagoonal environments. Sunburn can be very severe in the tropical Pacific; solar-induced skin cancer is common. Trees provide not only a protective habitat that is open to cooling breezes but also materials for the production of almost all locally-produced headgear and structures used to protect Pacific fisherfolk and agroforesters and their animals and shade-loving plants against the sun's rays.

Protection from natural calamities

Damage from wind, salt spray, erosion, and flood are increased when forests are removed. This was evidenced in 1982 in Tonga by the comparatively minor damage to crops by Isaac in areas with even small groves of trees. This worst hurricane in Tonga's recorded history caused damage estimated at T$8 milion (aprox. US$9 milion) to crops and livestock alone (Thaman 1982a)

Table 2 Ecological and cultural functions and uses of trees in agroforestry systems in the Pacific Islands, based on fieldwork in Papua New Guinea, the Solomon Islands, Vanuatu, Fiji, Rotuma, Tonga, Western Samoa, Kiribati, and Nauru comparatively minor damage to crops by hurricane Isaac in areas with even small groves of trees. This worst hurricane in Tonga's recorded history caused damage estimated at T$8 million (approximately US$9 million) to crops and livestock alone (Thaman 1982a).

Ecological    
Shade Soil improvement Animal/plant habitats
Erosion control Frost protection Flood/runoff control
Wind protection Wild animal food Weed/disease control
Cultural/economic    
Timber (commercial) Broom Prop or nurse plants
Timber (subsistence) Wrapping materials Staple foods
Fuel wood Abrasive Supplementary foods
Boat building (canoes) Illumination/torches Wild/snack/emergency
Sails Insulation foods
Tools Decoration Spices/sauces
Weapons/hunting Body ornamention Teas/coffee
Containers Cordage/lashing Non-alcoholic beverages
Wood carving Glues/adhesives Alcoholic beverages
Handicrafts Caulking Stimulants
Fishing equipment Fibre/fabric Narcotics
Floats Dyes Masticants
Toys Plaited ware Meat tenderizer
Switch for children/ Hats Preservatives
discipline Mats Medicines
Brush/paint brush Baskets Aphrodisiacs
Musical instruments Commercial/export Fertility control
Cages/roosts products Abortificants
Tannin Ritual exchange Scents/perfumes
Rubber Poisons Recreation
Oils Insect repellents Magico-religious
Toothbrush Deodorants Totems
Toilet paper Embalming corpses Subjects of mythology
Fire making Lovemaking sites Secret meeting sites

Deforestation commonly leads to accelerated erosion, and such degradation is ubiquitous in all high-island groups. Mangrove forests stabilize tidal-zone soil and reduce the impact of storm surge and salt spray. Trees also reflect terrestrial radiation and may provide some protection from frost to garden-edge crops at higher elevations in the Papua New Guinean highlands. It has even been suggested that pro grammes of coastal reforestation and agroforestry-based coastal reclamation, based on indigenous, salt- and wind-resistant trees and plants, could be one of the most effective strategies for addressing both the predicted short- and long-term effects of global warming such as increasing storminess, coastal erosion, and soil deterioration; declining fisheries and aquacultural yields; and decreasing soil fertility (Thaman 1989b).

Soil improvement

Natural soil improvement is another benefit provided by trees, especially given the high cost of fossil-fuel-dependent inorganic fertilizers and concern as to the detrimental impact on soil of long-term use of such fertilizers (Commoner 1971). Alioizia and Casuarina spp. are known by scientists and Papua New Guinean tribal farmers alike to enrich the soil for shifting gardens; as the Papua New Guineans say, such trees serve as a "garden mother" (Clarke 1965). Koka (Bischofia javanica), which is also believed to enhance soil fertility, is one of the most commonly protected trees in fallow areas in both Tonga and the wet and intermediate zones in Fiji. The extensive spread of the nitrogenfixing Middle American tree Leucaena leucocephala in many Pacific islands provides a recent example of this function (National Academy of Sciences 1977).

Forests and trees as habitats

Bird extinctions have been common on Pacific islands, and now many of the world's rarest or most threatened birds (the world's two rarest birds are the Mariana's mallard and the Kauai o'o of Hawaii) are found on Pacific islands. The past and threatened extinctions seem to be primarily the result of habitat elimination through deforestation (Dahl 1980; King 1981). Forest removal is also responsible, along with predation by profiteering collectors, for the endangerment of the worldrenowned birds of paradise and giant birdwing butterflies of Papua New Guinea. Without the recent introduction of protective legislation and butterfly-farming schemes, including the planting of host tree species, these valuable cultural, economic, and scientific resources would be lost (Pyle 1981).

Because deforestation may destroy the habitats of insects, birds, and other vertebrates that prey on crop pests, the process of forest loss may limit the potential for implementing integrated pest management programmes designed to minimize reliance on dangerous herbicides and pesticides (Bottrell 1978).

Timber, construction, and handicraft materials

Commercial timber operations supply local construction needs throughout the Pacific as well as generating significant amounts of foreign exchange in Papua New Guinea, the Solomon Islands, Fiji, and Western Samoa. Trees are also important in the informal sector of most countries for house construction, fencing, boat building, tool making, weaponry, making containers, and fishing gear (table 2). The coconut palm provides a whole storehouse of materials for house building, mat making, containers, fish traps (roots), and an array of handicrafts. The breadfruit for many islands is another tree of life, used for medicine, food parcelling, mulching, and canoe making. Bamboo contributes fishing poles, fencing, housing containers, and rafts. Many native species- e.g., Intsia bijuga, Cordia subcordata, and Thespesia populnea - are favoured for wood carving, but are now in short supply in Fiji and Tonga because of overexploitation. With the decline in these species, the introduced raintree (Samanea saman) has taken on increasing importance for making the traditionally-important kava bowl (tanoa in Fiji or kumete in Tonga).

Food resources

The nutritional importance throughout the Pacific of staple foods from trees such as coconut, breadfruit, bananas, sago palm, and Pandanus spp., along with a wide range of supplementary foods, snacks, or famine foods from other trees, has been widely stressed elsewhere and needs no further mention (Coyne 1984; Parkinson 1982; Rody 1982; Thaman 1979, 1982b, 1982c, 1983a, 1985b; Yen 1980a). Supplementary foods, snacks, and wild foods are described by Thaman (1975, 1976/77, 1982b) for Tonga and other Pacific islands and by Clarke (1965, 1971) for a highland Papua New Guinean community. Powell (1976a) provides a comprehensive coverage of wild foods and other important aspects of ethnobotany for the island of New Guinea.

Although many tree foods are energy-rich in carbohydrates and/or vegetable fats, it is in other nutritional essentials that they often excel compared with the ubiquitous root-crop staples and other annual nonarboreal plants. For example: mango, papaya, and some Panda nus spp. are excellent sources of provitamin A; Canarium spp., Inocarpus fagifer, and avocado (Persea americana) provide Bcomplex vitamins; and guava, mango, papaya, and Citrus spp. are rich in vitamin C. Most seeds or green leaves (for instance, from Ficus spp., Gnetum gnemon, which also provides edible seeds, and Moringa oleifera) are good sources of plant protein and a range of other micronutrients necessary for optimum health (Leung et al. 1972; Miller 1927; Miller et al. 1965; Murai et al. 1958; Thaman 1982c, 1983a). Moreover, all foods from trees and associated agroforestry-system ground crops are rich in fibre, which is essential to good health, but which is noticeably lacking in highly processed foods of urban diets (Coyne 1984; Thaman 1983a).

Spices and sauces from tree products can also be of great nutritional importance. Coconut cream or milk is used very widely in cooking, and local variants of the Rotuman taroro or Samoan sami lolo (concoctions of sea water aged or fermented with coconut flesh, often with chili peppers, in a coconut nut) enhance local cuisines. The sauce from Pandanus conoideus syncarps provides vitamin-A precursors and vegetable oil, nutritionally important additions to highland New Guinean diets. Indian cooking in Fiji utilizes tamarinds for chutney, the "curry leaf" (Murraya koenigii), a wide range of pickled fruits (achar), and many other tree products.

Trees are also important sources of food and fodder for domesticated animals. Pisonia grandis leaves, for example, are used as pig feed in Tonga; avocados are fare for pigs in the Cook Islands; Leucaena leaves and pods are used widely for goats, pigs, and cattle; and coconuts and papaya are abundant and important animal foods throughout most of the Pacific.

Wild food and other valuable products are lost to subsistence communities when the diverse plants and animals that supplied them disappear along with the forest that served as their habitats (Clarke 1965, 1971; Thaman 1982b). Destruction of the Calophyllum inophyllum forest and stands of Pisonia grandis on Nauru's central plateau, as a result of opencast phosphate mining, has eliminated roosting areas for the noddy bird, which is of considerable social and dietary importance to the Nauruan people (Manner et al. 1984, 1985). Deforestation has also restricted the habitats for wallabies and the valued cassowary bird of Papua New Guinea, and a great number of vertebrate and non-vertebrate wild foods that contribute significantly to the dietary well-being of many Pacific Islanders, particularly in the interior of large continental islands.

Mangrove ecosystems, which can be considered as a kind of outrider to agroforestry systems, contribute, either directly or indirectly through primary and secondary productivity, to the nutritional requirements of many marine species used as food by humans (Watling 1985). Research in Fiji has shown that over 60 per cent of commercially-important species are associated with mangroves at some stage in their life cycle (Lal et al. 1983), whereas more rigorous research gives figures of 67 and 80 per cent for eastern Australia and Florida (Walling 1985). Destruction and "reclamation" of such resources undoubtedly have deleterious effects on fisheries yields, with studies in the Malacca Straits indicating that mangrove clearing for industrial expansion led to a substantial drop in catches per effort (Khoo 1976). Baines (1979) argues that mangrove removal can lead to yield declines in offshore fisheries of 50-80 per cent.

Medicinal value

The arboreal pharmacopoeia is widely known and valued by modern science and industry as well as by local inhabitants. As everywhere in the tropics, all parts of the Pacific possess medicine-producing trees. For example, Nauru, which has an impoverished flora (some 49 indigenous species including non-trees), has 24 tree species with medicinal uses. Similarly, of 93 medicinal plant species found in urban gardens in Fiji, Tonga, Kiribati, and Nauru, 55 per cent (51) were trees and another 10 were woody shrubs (Thaman 1987).

Love potions and perfumes

More for pleasure than curing are love potions or aphrodisiacs and perfumes from trees. Guettarda speciosa, a coastal tree common in home gardens in Nauru and Kiribati, has flowers that when boiled in water produce a liquid that a woman can drink to "make a man go crazy" when she sweats. Perfumes or scents such as sandalwood are well known outside the Pacific and have drawn foreign exploiters from early European times to the present. In the late 1970s the profligate shipping of sandalwood from Tonga to Singapore and Hong Kong was stopped because of public outcry. Less cosmopolitan fragrances are derived from Cananga odorata and other scenting agents that are put into coconut oil from trees such as Pimenta, Plumeria, Gardenia spp., Parinari glaberrima, Aglaia saltatorum, Fagraea berteriana, and Calophyllum inophyllum. In Tonga, for example, there are over 50 species of sacred or fragrant plants, known as 'akau kakala, that are central to the spiritual and economic fabric of Tongan society and that are planted or protected as integral components of Tongan agroforestry.

Other uses

Wrapping materials include coconut leaves, leaves of Artocarpus altilis, Musa cultivars, Hibiscus tiliaceus, and Macaranga spp. Other leaves, notably Ficus spp., serve as effective abrasives. Dyes are derived from many sources, e.g., Bischofia javanica (a major brown dye for tape), Bruguiera spp. and Aleurites moluccana (black), Morinda citrifolia (yellow and red), and Bixa orellana (red).

These few examples from the list in table 2 show the utilitarian diversity and the economic and cultural value derived from trees and agroforestry in the Pacific - values that are rarely acknowledged in planning or project documents but that would be costly, difficult, or impossible to replace with imported substitutes. The elimination of such utilitarian and cultural diversity can only serve to lock Pacific societies more tightly into economic and cultural dependency.

Bases for innovation and sustainability

Instead of seeing trees as a basis for stability in agro-ecosystems, modern developers often focus on their disadvantages - e.g., trees take up space, may compete with annual crops, require years to reach maturity, and inhibit the use of some agricultural equipment such as the plough and mechanical harvester. These characteristics have often led to the domination or replacement of trees by more immediately productive annuals. However, in a world where biological stability is increasingly precarious, many characteristics of trees become advantageous. The "frozen" quality of trees - once established they are awkward to replace with other species - and the related lack of a quick turnover of product or land use provide a permanence in ecosystems that slows misuse and provides a wide range of ecological benefits: diversity of habitat, diversity of species, prevention of accelerated erosion, maintenance of soil fertility and arable soil structure, flood retardation, weed suppression, increased slope stability, and wind protection.

Trees for maintenance, insurance, and intensification

Agroforestry is now widely seen as a way to combine production with sustainability. In addition to the cultural and economic contributions already described or listed in table 2, trees also

  1. require less labour for maintenance than do annuals;
  2. provide the insurance of a diversified reserve of foods should annual crops fail; and
  3. produce, in combination with annuals, an aggregate yield greater than many monocultures of annuals.

As populations continue to grow and economic demands on land escalate, the use of land as a resource intensifies - which raises the trenchant question asked by Rambo and Hamilton (1991, 121), with regard to the upland areas of Asia and the Pacific: "How can we devise strategies for intensified resource use that will meet the broadest range of needs of the greatest number of people in the most sustainable way?" Part of their answer is that particular attention needs to be paid to agroforestry as a way to increase the productivity of existing agricultural land while reducing pressures to clear or otherwise exploit remaining forest land. Similarly, Raintree and Warner (1986) discuss the pathways by which agroforestry can aid in the intensification of shifting cultivation.

Antidote to nutritional degradation

The culinary and nutritional values of tree foods were discussed in a previous section. Further mention should be made of their potential to counter the rapid increase in nutrition-related maladies among Pacific peoples; an increase that has been widely documented and includes very high incidences of nutritional disorders such as iron-deficiency anaemia, vitamin-A-deficiency-induced night blindness, obesity, and general micronutrient deficiency, and of nutritionrelated non-communicable diseases such as cardiovascular disease, hypertension, diabetes, various forms of cancer, hyperuricaemia and gout, dental disease, and alcoholism (Cornell 1984; Coyne 1984; Fitzroy 1981; Jansen and Wilmott 1971; Johnson and Lambert 1982; Keith-Reid 1982; Pargeter et al. 1984; Parkinson 1982; Rody 1982; Thaman 1979, 1982c, 1983a, 1985b, 1987).

Although Pacific peoples seem to be genetically predisposed to obesity, diabetes, cardiovascular disease, and gout (Baker 1979), the main cause of the increase of these diseases today seems to be the shift to a diet of imported, highly refined foods from a diet of fresh foods high in fibre, vitamins, and minerals and low in sugar, salt, animal fats, and refined carbohydrates. A diet based on imported, highly refined foods is the reverse - as well as containing carcinogenic food additives. Table 3 is an attempt to assess the degree of correlation between these dietary changes and the major nutritional and nutritionrelated disorders. Cigarette smoking, increasing alcohol consumption, and decreasing physical activity are also contributing factors to the rising incidence of such diseases (Coyne 1984; Thaman 1983a).

As can be seen from table 3, the fresh fruits, nuts, vitamin-rich green leaves, derived juices, and complex-carbobydrate-rich and fibre-rich staple foods such as bananas, breadfruit, and even coconut (which has no animal fat or cholesterol, and the high Pacific-islander consumption of which does not seem to be correlated with the increase in any of these diseases, except possibly gout and hyperuricaemia) are exactly the types of foods needed to stem the Pacific's dangerous nutritional transformation. These foods also constitute the traditional snacks, drinks, and supplementary foods that are now being replaced by soft drinks, candy, and other modern but nutritionally-poor processed foods.

The nutritious local staples, fruits, and vegetables are increasingly scarce (in actual amount and because of their high price) owing to population growth, urbanization, and the emphasis on monocultural production for export. All-but-forgotten by today's youth and in danger of disappearing are ingenious and time-tested strategies of acquiring wild food, of practicing polycultural agriculture and multi-storey agroforestry, and of processing, storing, and preserving traditional foods (Barrau 1958, 1961; Massal and Barrau 1956; Parkinson 1984a; Thaman 1982c, 1985b; Yen 1980a).

Food dependency and decreasing self-reliance

Many countries have become dangerously "food-dependent," with food imports constituting far more, in terms of value, than returns from all export earnings (Carter 1984; Coyne 1984; Heywood 1991; McGee 1975; Parkinson 1982, 1984b; Rody 1978; Thaman 1982c, 1985b). As early as 1968, food imports in the Cooks, Kiribati, Tuvalu, Nauru, Niue, Western Samoa, and Tonga made up between 25 and 35 per cent of import expenditure (Fairbairn 1971; McGee 1975).

Table 3 Degree of correlation between the increasing incidence of major nutritional and nutrition-related disorders and dietary changes (increasing and decreasing consumption of specified nutrients or substances)

  Increasing consumption Decreasing consumption
  Refined
carbo-hydrate
Sugar Saturated
fat
Salt Alcohol Fibre Micronutrients Breast milk
Marasmus - - - - + - + + + + +
Kwashiorkor + + + + - - + + + + + + + +
Obesity (adult) +++ +++ +++ ++ +++ ++ + +++
Obesity (infant) + + + + + + + + + + - + + + + + +
Anaemia +++ +++ - ? + - +++ ++
Vitamin-A deficiency +++ +++ - ? ++ + +++ ++
Vitamin-B deficiency + + + + - ? + + + + + + + +
Micronutrient deficiency +++ +++ - - ++ + +++ +++
Infant mortality + + - - + - + + +  
Cardiovascular disease ++ ++ +++ +++ ++ ++ + ++
Hypertension + + + + + + + + + + + + + + + +
Diabetes +++ +++ ++ + +++ + ++ ++
Cancer ++ ++ ++ ++ +++ ++ +  
Gout/arthritis + + + + + + + + + + + + + +
Dental disease + + + + + - - + + + + + +
Alcoholism + + - ? + + + ? + +
General morbidity + + + + + + + + + + + + + + + +++ +++ +++

Source: Adapted from Thaman 1983a.

+ + + Very high positive correlation (i.e., a major determinant). + + High positive correlation (i.e., a significant determinant). + Some positive correlation (i.e., a complicating factor). - Not significant, no correlation, or a negative correlation. ? No data.

Although the percentages of total imports of foodstuffs have not changed significantly, the total values of food imports, and their value compared with the value of exports, has increased drastically. In Kiribati, for example, where the dominant locally-grown staple foods are coconut, breadfruit, and pandanus (together with the giant swamp taro Cyrtosperma chamissonis), the 1989 value of food imports was $8,580,000 (Australian), which was 134 per cent of the value of exports. The import percentage increases still more if beverages and tobacco are included. The situation is similar in Tuvalu, Niue, and the Cook Islands, where food-import bills are larger than the total value of locallyproduced exports. The decrease in self-reliance is even greater in American Micronesia, but less in Melanesia, although food dependency in urban areas is increasing rapidly there (Cornell 1984; Forum Secretariat 1991).

A study in Papua New Guinea comparing rural eastern highlanders with those living in urban Lae showed that, whereas sweet potato and other starchy foods provided 85 per cent of the dietary energy and 65 per cent of the protein in the rural area, rice had become the main staple - ahead of sweet potato and taro, with animal protein being eaten almost daily - in Lae (Jeffries 1979). The country's rice imports increased 40 per cent between 1975/76 and 1979, of which approximately one-half was consumed in rural areas (Bourke et al. 1982). Although energy, vitamin, and mineral intakes seem to be sufficient in rural areas where root crops, bananas, and tree foods predominate, the energy intakes in some low-income urban areas in the capital of Port Moresby, where 72 per cent of the energy was provided by flour, rice, sugar, bread, and biscuits, were the lowest ever recorded for the country: 1,200 kilocalories (5.0 MJ) for men and 1,035 kilocalories (4.3 MJ) for nonlactating women (Jeffries 1979). Sugar provided 11 per cent of the energy, whereas root crops provided only 9 per cent (Coyne 1984).

In Fiji, food-balance studies in the late 1970s showed that the total population received only 29 per cent of total energy from root crops and fruits (NFNC 1979). Of particular concern in Fiji is the increasing emphasis on the consumption of cassava; in some areas, its contribution to dietary energy rose from 36 per cent in 1953 to 59 per cent in 1963 (Parkinson 1984; Thaman and Thomas 1982' 1985). The increasing monoculture of cassava, which can be cropped almost continuously on even the poorest soils, has also been a factor in deforestation. Previously, yams, taro, and other traditional staples were more commonly cultivated along with trees as components in integrated agroforestry systems.

In Tonga and Samoa, similar trends are evident. Diets in urban Nuku'alofa contained more bread, sweet potato, cassava, mutton, pork, tinned fish, beef, butter, and tea than diets in rural areas, where more taro, plantain, coconut, ripe bananas, fresh fish, shellfish, green vegetables, and fruits were consumed (Jensen 1973). In Western and American Samoa, taro, green bananas, breadfruit, and coconut contributed 50-64 per cent of the dietary energy in rural areas but only 31 per cent on the more urbanized island of Tutu'ila, and only 20.5 per cent among highly urbanized Samoans in Hawaii (Bindon 1982; Parkinson 1984b).

In the Cook Islands, studies comparing rural Mitiaro with urbanized Rarotonga show that less than half the amount of root crops and one-eighth the amount of coconut, but over thirty times the amount of cereals (flour, bread, and rice) and nine times the amount of sugar were consumed in Rarotonga (Coyne 1984). In French Polynesia, where the trend is well advanced, by 1973 90 per cent of food intake was imported (Jacober 1977).

The changes are, however, most dramatic on some of the atolls, such as in the Tuamotu atolls of French Polynesia, where the traditional foods, such as taro and Polynesian arrowroot, have been all but forgotten and breadfruit and bananas are becoming rare. As early as 1956, Barrau (1961) found that flour, sugar, rice, biscuits, vegetables, fats, and tinned foods accounted for more than 2,000 kilocalories of individual daily food intakes.

In Tuvalu, where the traditional diet consisted of fish, coconut, breadfruit, bananas, and Cyrtosperma taro, by 1976 the island of Funafuti was 80 per cent dependent on imports for its food needs (Zimmet et al. 1981). In Kiribati, the per capita consumption of rice and sugar increased, respectively, three-fold (from 15 to 52 kg per year) and five-fold (from 8 to 40 kg) between 1950 and 1979. Even on rural Maiana, 60 per cent of the population regularly consumed sugar, 90 per cent flour, and 95 per cent rice (Pargeter et al. 1984). Studies on Namu atoll in the Marshall Islands in the late 1960s showed that 93 per cent of all copra income was spent on food, and that sugar, tea, and rice had replaced breadfruit, coconut, and Cyrtosperma taro as the most commonly consumed foods (Pollock 1970, 1974).

In parts of the Pacific, such as Tonga, where the traditional agro forestry system is relatively intact, the per capita bill for imported food has remained fairly modest - US$75 in 1984 (Heywood 1991, 75). Elsewhere, food dependency has reached far more vulnerable levels, as in French Polynesia with per capita annual food imports of US$554 (Hamnett et al. 1981).

Agroforestry and national development goals

Despite the relevance of agroforestry to modern development initiatives, few if any national development plans have formally included it in their lists of priorities or strategies, perhaps because trees and tree planting as components of agricultural systems "fall into the gaps" between the institutionalized sectoral responsibilities of "agriculture" and "forestry" (Chambers 1983).

An analysis of commonly stated national development objectives as well as of agricultural and forestry sector objectives from the most recent national development plans for Fiji (1985), Kiribati (1983), Papua New Guinea (1983), Tonga (1981), Vanuatu (1982), and Western Samoa (1984) shows considerable overlap, with the major objectives of national development being to

  1. improve productivity and the economic well-being of the country and its people;
  2. maximize national self-sufficiency and self-reliance;
  3. enhance the quality of life for all people, with particular em phasis on women, in both urban and rural areas - i.e., to distribute the benefits of development more equitably and to maximize the realization of the national human resource potential;
  4. preserve the culture and traditions of the nation;
  5. transfer or promote appropriate technologies; and
  6. conserve natural resources and protect the environment.

Sevele (1981) has argued that the wide-ranging development objectives "can be reduced to a few basic ones." These include:

  1. improved quality of life or standard of living;
  2. more equitable distribution of development benefits;
  3. increased employment opportunities;
  4. improved social and cultural facilities and services;
  5. protection of the environment;
  6. promotion of economic independence; and
  7. fostering regional cooperation.

Similar themes are seen in the more narrowly focused agricultural and forestry development objectives, with common objectives being:

  1. agricultural diversification;
  2. promotion of appropriate agricultural technologies and farming systems, both modern and traditional;
  3. strengthening of agricultural infrastructure, including extension,
  4. credit, transport, storage, processing, and marketing; improvement of the nutritional status of the people;
  5. increased selfsufficiency in timber and fuel wood; and
  6. promotion of social forestry or village-level agroforestry.

Almost all of these objectives could be furthered, either directly or indirectly, through the promotion of polycultural agroforestry systems. On the national level, in terms of encouraging import substitution, improving the balance of payments, and maximizing national selfsufficiency, agroforestry could have a significant effect in reversing dependency on imports and on ever-increasing foreign exchange problems. An expansion of agroforestry would also increase returns on underutilized natural and cultural resources and increase long-term productivity in both urban and rural areas.

Because almost all households can benefit from agroforestry, its systematic promotion on a national level could bring about more equitable and balanced development. It could also lead to improved use of scarce capital and aid and could minimize public expenditure through the maximization of self-help on the part of the communities involved.

For individual families, the economic importance seems to be great indeed, especially for unskilled workers and poor urban immigrants, who would benefit from the fruit, medicines, firewood, and other products provided by trees and associated plants and animals. For example, in Suva, Fiji, many residents of government housing would not have been able to pay their rents if it were not for the estimated $812 (Fijian) per week per family (approximately US$9.514) they saved by growing their own cassava, taro, tree crops, other foods, medicines, and firewood on idle urban land and in home gardens (Thaman 1984a). The value of coconuts, toddy, firewood, medicines, and the wide range of products of the coconut palm would likewise be of critical economic importance, particularly in densely-settled-atoll urban areas such as Tarawa in Kiribati or Majuro in the Marshall Islands.

The money saved can be used for other purchases - for example, of fish, meat, eggs, or dairy products - to supplement proteindeficient urban diets. The significance of subsistence provision in urban areas was recognized at least 20 years ago in the British Solo mon Islands Protectorate by the Committee on Food Supplies (1974), which noted in its report that ". . . non-cash or so-called 'subsistence incomes' are more important in towns, and cash incomes more important in rural areas than has been generally understood."

Agroforestry development could contribute to tourism development by ensuring that workers in the tourism industry maximize their real incomes and maintain subsistence production as insurance against downturns in the fragile international tourism economy. Moreover, coastal reforestation and coastal agroforestry would enhance the beauty and stability of beaches, which are so important to Pacific Island tourism.

The social benefits of agroforestry are manifold and must be seen as contributing significantly to the quality of life and the protection of cultural values. For instance, social ties are maintained through the distribution of produce and provision of food for feasts; trees are provided for the recreational activities of children; and urban dwellers and their children, who often have limited knowledge of or appreciation of plants, are linked with their past through an understanding of the traditional uses of trees.

There is considerable scope for involving traditional or existing leadership, increasing local participation, and increasing the involvement of women in agroforestry development. On Aitutaki in the Cook Islands, for example, successful schemes for replanting the endangered coastal species Cordia subcordata and Thespesia populnea have met with considerable success and have created greater environmental awareness.

Given access to land, agroforestry and tree planting are technically within the means of even the poorest families. They depend on inexpensive and readily available time-tested local technologies, plants, and cultural practices rather than on unfamiliar, often expensive, and ecologically-suspect imported technologies such as hybrid seeds, inorganic fertilizers, pesticides, mechanized equipment, and imported food and fossil fuels. Inexpensive and often locally produced hand tools are the only implements required.

Agroforestry inherently increases agricultural diversification in predominantly monocultural rural areas, as well as helping with most other development objectives such as achieving self-sufficiency in food, livestock, timber, and fuel; promoting appropriate agricultural technology; improving nutrition; and taking pressure off existing forestry resources while bestowing to agricultural landscapes the environmental benefits of trees.

Existing models and the need for appropriate innovation

How can agroforestry be encouraged so that the many advantages of tree-rich agro-ecosystems are more widely realized? How can an arborocultural strategy be maintained or intensified? We suggested earlier that functioning models have long existed within the Pacific, and that these should provide the region with the basis for modern interventions and the further evolution of agroforestry systems. Modified forms have developed with urbanization and some forms of commercial agriculture. A few of the possible models are described in the case-studies in the following chapters.

Although there will remain a continuing need for innovation and for the modification of existing agroforestry systems, it should be remembered that the existing systems have never been static; they are the product of constant evolution and selection over centuries or millennia and are evidence of the continued willingness of Pacific Island agroforesters to make rational decisions to adapt their systems to changing conditions and technologies. Because the existing systems are largely endogenous, rather than imposed in accord with the beliefs of urban-based governments or aid-funding agencies, the incremental improvement of existing systems - as by the introduction of appropriate new species, varieties, and technologies - offers a strong, locallybased means of promoting sustainable development in the small-island states of the Pacific Ocean.

A note on Melanesia

On most of the larger islands of Melanesia, sufficient land and relatively low population densities allow for the practice of extensive agricultural systems within largely forested landscapes. Human settlement and use of these lands have caused a humanization, taming, or "agriculturalization" of the forest. Although there is variation from place to place, the basic agricultural strategy consists of felling or ringbarking some trees and clearing the underbrush while at the same time protecting selected tree species that will remain or be allowed to regenerate as part of the garden of deliberately planted short-term crops and domesticated trees.

In most cases, the debris from forest clearing is allowed to dry and is then burned before or during planting, although in some areas such as on the Great Papuan Plateau in Papua New Guinea (Schieffelin 1975), on the Weather Coast of Guadalcanal, and in the upper Wainimala River area of Tailevu Province, Fiji - burning is discouraged. The debris is allowed to decay around emergent crops, thus retarding soil erosion and enhancing the development of the soil structure and the accumulation of organic matter.

The trees that have been preserved usually have some utilitarian value, such as provision of fruits, nuts, edible leaves, medicines, or wood for special purposes; or trees may be left intact or not weeded out because they improve the soil (e.g., the leguminous Albizia falcataria) or serve as habitats for desired prey such as birds of paradise, pigeons, or flying foxes (fruit bats). When such favoured trees are left unfelled on new garden sites, they are often pruned or pollarded to open up the ground to sunlight, to add additional organic material to the soil, or to provide support for climbing or sprawling crops. Thus, the practice of the classic system of shifting cultivation of gardens in forest results not only in the maintenance of soil fertility on garden sites but also in the development of a humanized forest fallow that itself contains many trees of economic and cultural significance.

The case-studies of Melanesian agroforestry systems presented in this chapter and chapter 4 are drawn from research carried out in various parts of Papua New Guinea, the Solomon Islands, Vanuatu, and Fiji (see map p. 36); the studies illustrate the range of variation possible within this general pattern. Some of the examples described are from areas of relatively low population density (Nduimba Basin in the highland fringe of Papua New Guinea and, in chapter 4, Namosi and Matainasau in Fiji); others are from areas of fairly high density (Buma on Malaita in the Solomon Islands and, in chapter 4, Tanna in Vanuatu). Not included in the Melanesian case-studies, except by inference in the discussion of the highland fringe of Papua New Guinea, is a description of the intensive, quasi-permanent sweet-potato cultivation of the densely settled Papua New Guinea Highlands, where labour-demanding tillage in grasslands is an important aspect of cultivation. But even here, where the forest has largely been replaced by anthropogenic grasslands, trees remain significant. Groves of Casuarina provide wood for fuel and fencing and serve as a planted, soilenriching fallow; Ficus dammaropsis (the "highland breadfruit") provides edible leaves; planted Pandanus conoideus is important nutritionally at midelevations and also in forested areas; and Pandanus julianettii and P. brosimos, which grow spontaneously or as quasi-domesticates in the remaining high-elevation forests above the zone of cultivation, provide oil-rich nuts that are an important supplementary food.

Highland fringe, Papua New Guinea

Clarke's (1965) comparative study of the intensification of shifting agriculture in highland Papua New Guinea provides information on variations between the agroforestry systems practiced by people of the highland fringe of Papua New Guinea.



Melanesia

Nduimba Basin

When studied in the mid-1960s (Clarke 1971), the Maring people of the Nduimba Basin numbered about 150 and occupied several square kilometres of mountainous land adjacent to an immense stretch of unoccupied rain forest on the northern slopes of the Bismarck Range. The population density was low- approximately 12 persons per sq km. Within their occupied territory, with elevations centred around 1,000 metres, no more than two per cent of the forest cover had been converted to grassland. In these few areas, the dominant species were Imperata cylindrica and Ischaemum digitatum, both of which are common components in the early stages of the Nduimba Basin plant succession that leads from cleared garden to well-developed secondary forest. Furthermore, the presence in the grasslands of pioneering trees such as Dodonaea and Alphitonia and tree ferns also points to the transient dominance of the grasses.

The comments of the people of the Nduimba Basin support the botanical evidence with respect to lack of pressure on the land. When asked why they had given garden land or usufructuary rights to outsiders, a frequent reply was: "We are a small group, and we have abundant land; therefore we give land away to friends and kin" (Clarke 1965, 348).

Agriculture
Even when compared with other types of simple shifting cultivation, the system in the Nduimba Basin is remarkably casual. Gardens are started sporadically throughout the year; part of the cut debris is usually burned, but the burning phase may be omitted if the weather remains wet. Most men clear their plots in secondary forest; a few of the "very strong men" sometimes make a clearing in the primary forest, a process that is more work than cutting the slenderer trees of the secondary forest. Low bush and the limited grasslands are never used for gardens. No tillage - in the sense of turning or working the soil - is practiced. Neither irrigation nor ditching is present or necessary. The planting tool is a simple wooden dibble. The polycultural mix of many species of tuber and fruit- or leaf-bearing crops that are planted in the gardens provides a fairly good diet, which is complemented by wild plant foods (including wild yams, ferns, nuts, fruits, and leaves) and fish, eels, domestic pigs and fowls, insect larvae, birds, and other wild game from the surrounding forest, fallow areas, and streams.

Agroforestry practices
Maring agroforestry practices come into play particularly when the gardens are left to fallow after the harvest of short-term crops. No fallow cover is planted but, because gardeners deliberately leave tree seedlings in the ground while weeding, the plot is usually well colonized by saplings when cultivation ceases. Additionally, as the harvesting of short-term crops comes to an end, the garden sites are transformed into orchards with major plantings of Gnetum gnemon (fibre, edible leaves and fruit) or Pandanus conoideus (oil-rich mesocarp providing very important supplementary food and condiment) and minor plantings of Artocarpus altilis (breadfruit) and Ficus wassa (bark cloth, edible leaves and fruit). The orchards thus come to be scattered widely throughout Maring territory and provide a valuable supply of materials and foods throughout the year. As they age over several decades, the orchards merge back into secondary forest. Their sites again become available for swidden clearing so that the orchards extend production from garden lands for several decades, while at the same time the land is receiving benefits similar to those gained from fallow under spontaneous secondary forest.

The Maring also recognize unplanted secondary-forest communities as a resource - one that is more valuable than primary forest. Because the secondary forest provides the reinvigorating functions of fallow vegetation that make shifting cultivation possible, it is referred to in the vernacular as the "garden mother." Beyond this important function, the whole secondary zone - which derives from clearing for gardens combined with the protection of tree seedlings during weeding - serves as a foraging zone for domestic pigs and as a hunting ground for the men who pass through it daily on their rounds. The secondary communities also serve as sources of plant foods and materials. The common Alphitonia-Cyathea woodlands provide easily-cut wood for fencing and fuel. In addition, the Cyathea provides edible leaves; and the understorey plants, growing in the light shade, provide cordage, edible leaves, and medicines. The secondary Albizia falcataria community is recognized as most valuable for increasing soil fertility for future gardens, presumably through the usual nitrogen-fixing processes of leguminous plants; Albizia's wood is also valued for carving. Jungle regrowth provides edible leaves as well as wild Musa leaves for wrapping materials and for clean surfaces for food preparation. In other words, not only does the temporal alternation of forest fallow with gardens make the gardens produc five, the successional stages of the forest are themselves valuable for other purposes (Clarke 1971, 60-64).

The people of the Nduimba Basin are not oppressed by garden work. Even the women, who do most of the steady work of weeding and harvesting, frequently take a day's vacation and eat food gathered on the previous day or collect snacks of sugar cane from their nearest garden. Food shortages and crop failures are unknown, and when the gardens are abandoned they contain considerable amounts of unharvested produce.

Kompiai

Kompiai is the name given now to a territory on the southern slopes of the Bismarck Range (Jimi River valley) occupied by the Kauwatyi clan cluster, a group that numbers about 850. The average population density at Kompiai is 29 people per sq km, more than double that in the Nduimba Basin. At Kompiai, retrogressive succession of the vegetation is more advanced than in the Nduimba Basin, which is about 16 km distant. Perhaps as much as 20 per cent of the Kauwatyi land is grassland containing genera such as Themeda, Arundinella, Ophiurus, and Eulalia, which are thought to be indicative of degraded sites. Kompiai has a smaller percentage of primary forest than the Nduimba Basin. On average, the secondary forest at Kompiai is younger, and much of it is floristically less complex; large areas are covered by almost pure stands of the weed tree Dodonaea viscosa - a type of simple regrowth that is absent from the Nduimba Basin.

This circumstantial evidence of an intensity of land use greater than that of the Nduimba Basin is compatible with what the Kauwatyi say about their land and with their quite unusual action in the mid-1950s when they attempted to make gardens in the territory of a defeated neighbour before certain prescribed ritual procedures had taken place. Such an attempt almost certainly indicated considerable pressure on the land because, in this part of Papua New Guinea, the extension of territory is scarcely ever an immediate result of warfare.

Agriculture
The gardens at Kompiai look almost like those in the Nduimba Basin: there is neither tillage, irrigation, nor any attempt to improve soil drainage; nor are the grasslands used for gardens except on rare occasions. The methods of planting, the agricultural tools, and the "messy" intermixture of garden plants are the same in both places. The crops grown are the same, too, but the proportions are different. At Kompiai, sweet potatoes become relatively more important among the starchy staples grown - an indication of the beginning steps toward the very heavy dependence on that crop that characterizes the areas of highest population density in the central highlands of Papua New Guinea, where fields are used almost continuously and forest fallow is wholly absent. The orchards of Pandanus conoideus, Gnetum, and breadfruit are rarer at Kompiai than in the Nduimba Basin, and game is scarcer.

The intensity of harvest and the age and type of fallow cover also vary between the two places. Native informants all agreed that the Kauwatyi harvest their gardens more thoroughly than the inhabitants of the Nduimba Basin. Moreover, when the Kauwatyi decide that an old garden is ready for fallow, they bring their pigs to the plot to root for the small tubers that still remain in the ground. In the same situation, the people of the Nduimba Basin simply abandoned the plot; if the pigs find it, well and good; if not, it does not matter. Doubtless, if the pig herd of the Nduimba Basin were to increase, the Kauwatyi custom would be adopted.

Agroforestry practices
Although fallow periods are shorter at Kompiai than in the Nduimba Basin and consequently there is less well-developed secondary forest, the Kauwatyi habit of planting or encouraging quick-growing Casuarina oligodon trees in old gardens may slow the decline in soil fertility associated with decreasing lengths of fallow between periods of gardening. Interestingly, although the people say that the soil under Casuarina groves is better than that under other types of fallow, the main motive for planting the trees seems to be to obtain wood for fences, fuel, and house construction. Despite the value placed on the trees, the people's efforts to establish groves of Casuarina are haphazard; if the bare ground of a new garden develops Casuarina seedlings, the gardener cherishes them; but if seedlings do not develop, he does not always transplant seedlings into the garden (Clarke 1965, 349-352).

In contrast, not far away in the upper Kaironk Valley - where population density is even higher, and spontaneous forest almost absent except on the cloudy mountain crests - planting of Casuarina as a fallow cover for gardens is much more systematic. Groves of these graceful trees dot the grassy landscape, and a high percentage of the gardens are planted beneath pollarded Casuarina. Despite the soil improvement brought by these nitrogen-fixing trees, land degradation has proceeded in the Kaironk to the point where labour-intensive tillage is required to maintain soil fertility, and sweet potatoes (which require a less fertile soil than Colocasia taro or yams) are far more important than at Kompiai or in the Nduimba Basin.

Implications of increasing agricultural intensification

The variations between the agro-ecosystems of the Nduimba Basin, Kompiai, and the Kaironk Valley can be interpreted as a sequence of increasing agricultural intensification. The consequences of a decline in the amount of well-developed natural forest because of extended cropping cycles and shortened fallow periods (caused in part by a recent expansion of smallholder commercial coffee planting and beef cattle production) include an increase in the intensity of garden harvesting, an increasing use of agronomic techniques to maintain soil fertility, a decline in yield per unit labour, and a concentration of effort towards production of the highest-yielding crops (sweet potatoes in the highland New Guinean case). The implications for agroforestry are that, as primary forest and well-developed secondary forest are converted to grassland or impoverished scrub under regimes of increasing population density and shortened fallow, the resources available from forest disappear, and the regenerating "work" done by the forest must be taken over by the human cultivators if an acceptable level of soil fertility is to be maintained. In parts of New Guinea a traditional solution to the loss of spontaneous forests was the creation of artificial forests of Casuarina (or Dodonaea viscosa in some places) to aid soil regeneration and to provide wood for fuel and fencing. Although of great value, such single-species stands of trees did not provide the variety of wildlife habitats or the food, medicine, or other materials found in the more species rich natural forests.

Kologhona village, Weather Coast, Guadalcanal, the Solomon Islands

In 1972, population densities on the Weather Coast of southern Guadalcanal, the Solomon Islands, ranged from 1.5 to 26.4 persons per sq km (Chapman and Pirie 1974). Pressure on land was low in the vicinity of Kologhona village, some 10 km inland from Babanakira in the upper Tina River Basin of the Wanderer Bay area. The village, which had a population of approximately 60 in early 1975, had access to extensive areas of alluvial and colluvial soils, steeply sloping garden land covered with secondary forest, and considerable areas of slightly disturbed primary forest along the crests of the ridges and on the steeper slopes.

Shifting agriculture
The two most common subtypes of agroforestry practiced in Kologhona are ago-puka and ago-male. The ago-paka method is practiced in old forest at some distance from the village. All but the biggest trees are felled, after which the debris is piled around the bases of bigger trees and burned. No attempt is made to clear all stumps or level the ground before planting commences. Traditionally, planting was almost exclusively women's work, using digging sticks and hoes; men are increasingly doing more of this work. Yams and Colocasia taro are the most common crops planted in these gardens, although other ground and tree crops are also planted. Fallen trees are allowed to lie in the garden, often placed along contours to retard erosion and to provide for trellising yams and other climbing plants.

In the more heavily cropped alluvial and colluvial soils and gardens closer to the village, the ago-male method is more common. Its practice means the extension of existing gardens into surrounding secondary vegetation by clearing, moving the debris to the side, and planting without burning (Rainbow and Teteha 1983). Sweet potato is the major crop in these gardens, intercropped with yams, taro (Colocasia and Xanthosoma), banana cultivars, and a diversity of other crops, including sugar cane, Saccharum edule, Hibiscus manihot, pumpkin, pineapple, maize, chill) peppers, and tobacco.

Agroforestry practices
Trees with edible fruits and nuts are commonly protected when secondary vegetation is cleared, or they are planted amongst crops. Trees so treated include breadfruit, coconut, betel-nut, papaya, Citrus spp., Canarium spp., Inocarpus fagifer, Barringtonia edulis, Syzygium malaccense, and Ficus copiosa. Other useful tree or tree-like species maintained in gardens include Pandanus tectorius, kapok (Ceiba pentandra), and sago palm (Metroxylon salomonense) and

Heliconia indica, both of which are found mostly in poorly drained areas close to the river. In fallow areas, the commonest pioneer species are Kleinhovia hospita, A lstonia spp., Ficus spp., and Macaranga aleuritoides. Wild foods - including wild yams (Dioscorea spp.), a range of ferns, and other wild greens - and animal foods are either preserved through selective weeding or occur in fallow and secondary vegetation.

Despite increasing population pressure, the preservation of trees as part of an integral agroforestry system has continued. However, increasing pressure by the government to expand monocultures of copra or cocoa and the smallholder production of beef cattle, with no emphasis on the maintenance of arboreal diversity, is accelerating agrodeforestation on the Weather Coast and will play a major role in the decline of arboreal diversity and self-sufficiency and the loss of knowledge of traditional agroforestry systems among young agroentrepreneurs.

Buma village, West Kwara'ae, Malaita, the Solomon Islands

With a population density of 14.2 people per sq km in 1976, the island Malaita was (and remains) one of the more densely populated provinces of the Solomon Islands (Solomon Islands Government 1979). While more than 68 per cent of all rural Solomon Islands adults were fully engaged in subsistence or semi-subsistence agriculture and another 23 per cent were partly so engaged, comparable figures for Malaita were 77 per cent and 17 per cent respectively (Eele 1978). On Malaita, only 1,467 people were employed wageearners; most Malaitans gained their livelihood through subsistence fishing, shore and reef gathering, and shifting agroforestry. The pressures, however, for increased cash income through smallholder cattle ranching and cocoa and copra production were having a significant impact on the sustainability of shifting agroforestry on Malaita. This was readily apparent at Buma, a coastal village of approximately 361 Kwara'ae-speakers, located approximately 9 km north of Auki, the provincial capital.

Agriculture

The systems of shifting agroforestry at Buma, and throughout much of Malaita, are not unlike the ago-paka and ago-male systems of Kologhona, except that burning is a characteristic feature of all Buma systems. Almost all shifting agroforestry is practiced on well-drained lowland rain forest or secondary forest fallows, on both steep and level terrain. Selection of garden sites is determined by distance from the village and older gardens, availability of adjacent or nearby fallows for future plantings, soil and fallow characteristics, and previous productivity of the site (Manner 1980). New gardens are usually located adjacent to producing gardens, which serve as sources of planting materials for the new gardens. The old gardens are thus extended in a style resembling the linear arrangement of gardens described by Oliver (1955) and Connell (1978) on Bougainville.

Both men and women participate in gardening activities such as slashing the undergrowth, transporting planting materials, burning, planting, and weeding. Men are usually responsible for the more arduous tasks such as tree felling, log removal, and their emplacement as boundaries. All trees except Areca catechu, Artocarpus altilis, and Canarium indicum are cut down by axe. After a drying period of 1-2 weeks, the litter is gathered into small piles and burned. If young (5-12 years) secondary forest is cut for gardening, the Buma Kwara'ae ensure that all parts of the garden receive ash. On the other hand, if mature secondary forest or rain forest is felled, the regrowth is slashed and the litter burned - although little is done to ensure that all parts of the garden receive ash.

The cleared plot is marked off into sections, then a dibble stick is used to plant a wide variety of root-crop cultivars, fruit and nut trees, and other crops such as pineapple, water melon, Hibiscus manihot, tomatoes, and tobacco (table 4). Newly cleared gardens (2-4 months old) are dominated by taro (Colocasia esculenta), giant taro (Alocasia macrorrhiza), and the sweet or lesser yam (Dioscorea esculenta). As taro and other cultigens mature, replanting is carried out. By way of contrast, older gardens are dominated by Alocasia macrorrhiza, pineapple (Ananas comosus), sweet potato (Ipomoea batatas), and a heavier weed cover. The gardens are abandoned to fallow two years after initial planting.

This human-directed succession of cultivars and the release of gardens to fallow can be seen as a reflection of Kwara'ae understanding of crop requirements and ecological processes. The presence of sweet potatoes and cassava in older gardens indicates poorer soil fertility a condition to which these species are adapted (Haynes 1977; Thaman 1976; Thaman and Thomas 1982, 1985). Bananas and other tree and shrub crops, because of their height and longer growth cycles, may be better adapted to resist the increased weed and pest in festations of older gardens. Their more extensive rooting systems may also be more efficient in nutrient uptake than shorter-lived plants. Finally, as the fallow period continues, there is the increasing regeneration of pioneering forest trees, which brings renewed fertility and filth and increases habitats for other forest-dwelling organisms. Wild food products multiply and noxious garden pests disappear. In addition, the mixed forest is the source of building materials, medicines, dyes, and food; and its regeneration may help to maintain the cultural and economic stability of the Buma Kwara'ae. Major tree or tree-like species include a range of bananas and plantains (Musa cultivars), the culturally important betel-nut (Areca catechu), the ngali nut (Canarium indicum), papaya (Carica papaya), and the nutritious hibiscus spinach (Hibiscus manihot) (table 4).

Monetization and agrodeforestation

For the past two decades, the sustainability of the Kwara'ae shifting agroforestry system has been threatened by government-sponsored smallholder cattle, cocoa, and copra projects, and by sweet potato cashcropping for the Auki market. The resultant agrodeforestation brought about by the increasing emphasis on cash cropping has greatly increased pressures on Buma land resources and caused the spatial displacement of traditional agroforestry, a process noted as occurring elsewhere by Chapman and Pirie (1974) and Quartermain (1980). At Buma, cocoa is intercropped with coconuts, and cattle are grazed under coconuts and other valuable tree species. Prior to these introductions, the Buma Kwara'ae practiced traditional agroforestry on lands classified as "highly suitable agricultural opportunity" areas (Wall and Hansell 1974), and which were located within 15-30 minutes walking time from the village. Today, however, most traditional agroforestry is practiced on more distant and steeper-sloping lands. Walking times to these areas range from 30 to 60 minutes. Unfortunately, the burden of transporting garden produce, which may weigh as much as 15 kg per trip, often falls on the women. The Buma Kwara'ae have potential gardening sites located 2 hours farther inland, but few gardens have been cleared in these relatively distant areas. It is ironic that a major reason for this decline in traditional agroforestry is narrowly focused, institutional smallholder agroforestry focused primarily on the tree crops, cocoa and coconuts.

Similarly, cash cropping of sweet potatoes for the Auki market has also displaced traditional agroforestry farther inland. Moreover, this cash cropping is conducted in forest or bush fallows often less than 3 years old. This shortening of the gardening cycle has led to significant soil deterioration and a retrogression of the fallow vegetation from the useful arboreal species that characterized traditional forest fallow to a community dominated by Acalypha grandis, ferns, and Imperata cylindrica.

Table 4 Species composition of Buma village gardens, West Kwara'ae, Malaita, the Solomon Islands. (Data expressed as number of plantings per 5m x 5m quadrat; C =cormels; D = dominant crop)

Garden age 2-4 months 6-12 months 12-18 months 18+ months
Quadrat number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Scientific name (Kwara'ae or common name)                                
Alocasia macrorrhiza(giant taro) 2 5 5 5 19 19 2   4 3 12         3
Ananas comosus (pineapple) 4 6           9 5 7         3 2
Areca catcheu (betel-nut)                             1  
Canariurn indicum (ngali nut)       1                        
Carica papaya (pawpaw) 11         1                    
Colocasia esculenta (taro) 38 42 2   2 15 13C 2C 2C 7C   15C        
Citrullus lanatus (water melon)     2 1                        
Dioscorea esculenta (sweet yam)     16 17                        
Hibiscus manihot (hibiscus spinach)             5   1     1     2  
Ipomaea batatas (sweet potato)           10 5 D D D D 9D D D D D
Lycopersicon esculentum (tomato) 1           1                  
Manihot esculenta (cassava)                 4       2 9    
Musa cultivars (banana/plantain)                       2     1  
Nicotiana tabacum (tobacco)       1   2       1   2        
Total plants 56 53 25 25 21 47 43 5 15 18 11 41 3 10 7 6
Number of species 5 3 4 5 2 5 5 3 4 5 3 6 2 2 5 3

Source: Adapted from Manner 1980.

The south-eastern Solomon Islands

Among the most thorough accounts of traditional Pacific agroforestry are those of the south-eastern Solomon Islands by Yen (1974, 1976a, 1976b) and of the Polynesian outlier Tikopia by Kirch and Yen (1982). These works, coupled with Powell's (1976b) study of plant communities and Hendren's (1976) study of Ulawan settlement patterns, provide an excellent basis for understanding agroforestry in the south-eastern Solomon Islands.

The areas studied include the island of San Cristobal (3,500 sq km); the smaller islands of Ulawa and Santa Ana; the Santa Cruz group, including the main island of Nendo (660 sq km); and the islands of Utupua and Vanikolo, the Duff Islands, the main and outer Reef Islands, and the Polynesian outliers of Anuta and Tikopia in the far south-east. The environments of these islands range from the heavily forested interiors of the larger high islands of San Cristobal, Nendo, Utupua, and Vanikoro (which is over 900 metres in elevation) with extensive coastal plains and swampy areas, to smaller volcanic islands, and the coral atolls of Nupani and Nukapu in the outer Reef Islands with very restricted floras.

Agriculture

The dominant agricultural system in most of the south-eastern Solomon Islands is shifting cultivation, with yams (Dioscorea alata and D. esculenta) and taro (Colocasia esculenta) the dominant staples in newly cleared plots. Sweet potatoes (Ipomoea batatas) are commonly planted next in succession. Primary forest is rarely cleared; rather, secondary forests on coastal plains and lower slopes are the most common sites for new gardens. Some inland forest sites are occasionally cleared for yam gardening, including the cultivation of minor species such as Dioscorea pentaphylla, D. nummularia, and D. bulbifera. Yams often occupy drier sites, whereas taro, which is often planted twice before letting the land revert to fallow, occupies wetter areas. Giant taro (Alocasia macrorrhiza) is occasionally planted as a supplementary staple (Hendren 1976; Kirch and Yen 1982; Powell 1976b; Yen 1973, 1976a, 1976b).

Gardens are usually cropped for no longer than three years, with fallows ranging from 6 to 12 years. More permanent gardens of perennials such as Cyrtosperma chamissonis and trees are also found in some areas, such as Nendo. Wetland cultivation of taro and sago palm (Metroxylon salomonense) is practiced in the Duff group. On Anuta and Tikopia, more permanent or intensive systems incorporating cassava into taro-cassava or sweet potato-cassava rotations have displaced yam cultivation.

The greater intensity of land and resource utilization on the small isolated islands of Anuta and Tikopia is evidenced by the great diversity of seasonal surpluses of foods stored using pit-preservation or ensiling by semi-anaerobic fermentation. Herbaceous and tree species thus utilized included breadfruit, Burkella fruit, Ananas and plantains, sago, taro, cassava, Polynesian arrowroot (Tacca leontopetaloides), giant taro (Alocasia macrorrhiza), and giant swamp taro (Cyrtosperma chamissonis) (Yen 1976a). Further evidence of intensification is provided by the existence of walled garden complexes (hatarau or pwainua) and non-irrigated terraces on Ulawa (Hendren 1976) and stone-walled terraces on Anuta (Yen 1976a).

Agroforestry and arboreal components

All agricultural and land-use zones have major arboreal components. The main species found in villages and home gardens, in permanent village tree groves, and as protected or deliberately planted intercrops in coastal and inland food gardens include coconut palms (the most common intercrop), a wide range of banana and plantain cultivars (Musa cultivars), breadfruit (Artocarpus altilis), sago palm (Metroxylon salomonense), the nut trees Canarium spp., Barringtonia edulis, Inocarpus fagifer, and Terminalia catappa, Burkella obovata, edible pandanus (Pandanus dubius), Malay or mountain apple (Syzygium malaccense), oceanic Iychee (Pometia pinnata), the hogplum or Polynesian vi-apple (Spondias dulcis), Gnetum gnemon, edible figs (Ficus spp.), and the betel-nut palm (Areca catechu). More recent introductions include mango (Mangifera indica), citrus trees (Citrus spp.), and papaya (Carica papaya). Of the non-food species, Ceiba pentandra is very common.

Other less common or locally important species found planted or protected in gardens and tree groves include Antiaris toxicaria, Sterculia sp., Semecarpus sp., and Corynocarpus cribbeanus, with Sterculia and Semecarpus being planted in breadfruit groves as living ladders. Species found around former inland settlement sites include Areca catechu, Burkella obovata, Terminalia catappa, Gnetum gnemon, Corynocarpus cribbeanus, and Cordyline fruticosa and Cyathea sp.

Shrubby or herbaceous supplementary food species intercroppped in food gardens and planted in villages and home gardens include hibiscus spinach (Hibiscus manihot), sugar cane (Saccharum officinarum), Cordyline fruticosa (also represented by cultivars of ornamental and magical or ceremonial importance), edible panax (Polyscias fruticosa and P. scutellaria), and maize (Zea mays) and non-food species such as Pandanus spp., Euadia sp., Alpinia sp., Codiaeum variegatum, Gardenia sp., Halfordia sp., and tree ferns (Cyathea sp.). Species commonly planted along garden borders or as hedges include vegetatively planted Barringtonia edulis cultivars, Ananas and plantains, sugar cane, Cordyline fruticosa, and Polyscias spp.

For many of these species, both wild (or feral) and cultivated forms are recognized, with numerous cultivars or varieties specified for the more common species. For example, there are 13 varietal forms of coconut, 33 banana or plantain cultivars, 18 breadfruit cultivars, 3-5 cultivars each for Canarium spp., Pometia pinnata, Barringtonia edulis, and Burkella sp., and at least 2 cultivars each for Terminalia catappa, Metroxylon salomonense, and Gnetum gnemon. The wide range of arboreal husbandry strategies includes deliberate planting of seeds, shoots, branches, and other vegetative parts, the transplanting of self-sown seedlings, and their protection in fallows. Of particular interest is the hypothesis that many trees that grow wild elsewhere (e.g. Burkella obovata, Pometia pinnata, and Calophyllum inophyllum) have been domesticated or semi-domesticated in the southeastern Solomon Islands because of the "accent on trees there" resulting from "an intensification of attention on the gathering aspect of subsistence economy." This focus on arboriculture seems to be Melanesian (that is, nonAustronesian) rather than Austronesian or Polynesian (Yen 1974).

None the less, arboriculture in the definitely Austronesian areas, such as on the Polynesian outlier of Tikopia, also reflects great agroforestry sophistication, possibly as a result of continued contact with Melanesia. Because of the unusual richness and significance of the Tikopian orchard tracts, the description of them by Kirch and Yen (1982, 38-39) is provided at some lenght:

The orchard tracts are not exclusively devoted to productive trees. Rather, they mimic the mixed nature of the low-altitude forest associations typical of the Solomon Islands flora, not only in tree species, but also in the subcanopy. These artificial associations are exemplary of the multi-storey forms of cropping sometimes recommended as innovations for modern agricultural development in the Pacific, but which actually already exist as indigenous adaptations (Ward and Proctor 1980), albeit sometimes erased by the adoption of commercial cropping modes. A characteristic Tikopia orchardholding has a lower planted storey of Cyrtosperma, which, it should be noted, is adapted well beyond its "swamp taro" environment. Bananas generally form the second storey upward, but in scattered distribution, their situation being favored by more open portions of the canopy. The vines of Piper belle may climb to quite prodigious heights on trees, but are generally kept in control by constant picking for betel chewing, while the yams . . . climb to the upper reaches, depending on season and time of planting. The density and diversity of the middle storey is compounded by the presence of Antiaris and Gnetum trees, the occasional cycad, and the saga palm at different stages of growth before flowering or maturity, when its height may reach 20 meters. The bulk of the canopy, generally below that height, consists of the orchard trees proper- breadfruit, Eugenia, lnocarpus, Spondias, and the occasional Calophyllum. On the flatland and at lower altitudes (below 50 meters above sea level), this list is often enriched by Canarium, Burkella, Terminalia catappa, Barringtonia, and Pometia.... The coconut, important not only as a food source but also for its multiple industrial uses, does not form the familiar belts of Pacific island strands; rather it is distributed on Tikopia throughout the orchard gardens, in all but the highest sections of the island to protrude above the canopy of economic trees.

More monocultural smallholder coconut plantations, the main source of cash income, extend from coastal plains up to an elevation of 22 metres. Breadfruit and bananas or plantains are planted in almost monospecific groves or patches on the coastal plains and near active garden areas.

When clearing secondary vegetation for new gardens, most of these species are preserved, although some - such as Pometia pinnate and Ficus spp. - may be severely pruned or pollarded to open up new garden areas to sunlight and to provide additional organic material or ash. Self-sown seedlings of these species are also protected through selective weeding and are allowed to regenerate as part of the fallow vegetation.

Pioneer species, which are abundant in secondary and open disturbed forest and common in young regrowth in still-productive gardens on most of the larger islands, include Macaranga, Trema, Pipturus, Ficus, Glochidion, Acalypha, Piper and Euodia spp., Morinda citrifolia, Kleinhovia hospita, and Hibiscus tiliaceus. Species more prominent in mature secondary or older regrowth forests, ready for reworking into gardens, include many of the commonly cultivated species such as Pometia pinnata (often as a dominant), Canarium spp., Artocarpus altilis, Syzygium spp., Burkella obovata, Antiaris toxicaria, Spondias dulcis, wild varieties of Areca catechu, stands of bamboo (possibly including both Schizostachyam glaucifolium and Bambusa spp.) along with the indigenous species, Alphitonia incana, Securinega flexuosa, Pterocarpus indicus, Semecarpus sp., and Dysoxylum, Ficus and Elaeocarpus spp. Less common secondary forest species include Stemonurus sp., Boerlagiodendron sp., Horsfieldia spicata, Nauclea orientalis, Garcinia sessilis, Cryptocarya medicinalis, Pimleodendron amboinicum, with common understorey species in mature fallow forest including Cordyline fruticosa, Fagraea racemosa, Euadia, Psychotria, Calophyllum, Codiaeum, Dracaena, and Pseuderanthemum spp. (Powell 1976b).

The climax or only partially disturbed inland forests of the larger islands provide the main hunting and gardening reserves; the dominant trees are Agathis macrophylla, Terminalia calamansanai, Calophyllum vitiense, Dacrydium elatum, and Albizia falcataria with Campnosperma brevipetiolata being common in more disturbed areas.

Protected areas of coastal strand, mangrove, or coastal plain and lake-shore vegetation can also be seen as integral to traditional agroforestry strategies in the south-eastern Solomon Islands. Species commonly found in these areas include Calophyllum inophyllum, Pandanus tectorius, Casuarina equisetifolia, Hibiscus tiliaceus, Terminalia catappa, Inocarpus fagifer, Cerbera manghas, Barringtonia asiatica, Barringtonia racemosa, Heritiera littoralis, and Intsia bijuga. Less common species include Thespesia populnea, Cordia subcordata, Hernandia nymphaeifolia, Tournefortia argentea, and Pisonia grandis.

The "extension of strand vegetation" inland was encouraged through the tending of coastal tree species of value for consumption and industrial use. There were also plantings of feral genera such as Pandanus, Barringtonia, and Terminalia, and species such as Calophyllum inophyllum (Yen 1976a). C. inophyllum, in particular, was commonly planted in the main Reef Islands and domesticated and seed-sown in large-scale plantings (aururakau o ngatai) for the purpose of shore-line stabilization and in garden areas up to 300 metres on Tikopia, with Casuarina equisetifolia, H. tiliaccus, Ochrosia sp. (Neisosperma oppositifolia?), and Pipturus sp. reportedly being semi-domesticated and planted inland and for coastal stabilization.

Agroforestry on Aneityum and Tanna, Vanuatu

Aneityum and Tanna, two of the southernmost islands of Vanuatu, have agroforestry systems that are similar in many ways, but very different in others. Aneityum is the smaller island and is geologically older than Tanna. It is less than 200 sq km in area, with a degraded and deforested mountainous interior rising to 850 metros. The highly eroded and laterized soils of the interior, which support only an impoverished vegetation, have been interpreted to be the result of the activities of the large populations (between 3,000 and 4,000 people) that inhabited the island prior to the severe post-missionization depopulation that resulted in a low of 516 people in 1979. The initial decline resulted from a measles epidemic in 1860, followed by an outbreak of dysentery (Carter 1984, 495), and perhaps by other epidemics two or three decades earlier (Spriggs 1981, 73-77). Tanna is a larger (560 sq km), geologically younger, island with an active volcano. Its highly fertile young volcanic soils support a population of over 15,000 (Bonnemaison 1986).

Although excellent studies have been conducted on traditional agriculture on Aneityum (Spriggs 1981) and Tanna (Bonnemaison 1984, 1985, 1986), on the vegetation and flora (Guillaumin 1931, 1932; Schmid 1973), and on the trees of Vanuatu (Gowers 1976), little information has been published on the nature of trees found in active garden areas and in villages and home gardens. The information given here on these trees is derived from field research by R. Thaman and W. Clarke on Tanna and Aneityum in 1988.

The traditional significance of trees in Vanuatu

A particularly interesting aspect of agroforestry in Vanuatu, compared with many other areas? is the relative unimportance of recently introduced trees in villages and agricultural areas, and the predominance of indigenous and aboriginally introduced species. Bonnemaison (1985, 56-57) has remarked on the failure of colonial planters and missionaries to eliminate kastom (tradition) on Tanna; he further observes that the Melanesian society there has "retained in its heart of hearts a memory of its identity and an ability to reconstruct itself according to its own standards." More specifically, just before the Second World War, most people of Tanna rejected Christianity and returned to the use of "magic and to their old values and beliefs" and invented their own Messianic religion, the John Frum Movement, and the return to "the truth of their traditional culture - "Kastom'" (Bonnemaison 1986).

In his study of Melanesian identity, Bonnemaison (1985, 32) argues that the tree serves as a "metaphor" or "symbol of rootedness and stability," and that, in Tannese cosmogony, men wandered without shelter or protection from the sun, rain, or the cold of night. The growth of vegetation, first in the form of lichen and then real trees, provided the first protection. At the foot and in the shade of banyan trees, men hollowed out the first dancing grounds (yimwayim) and built their houses. They then spread gradually throughout the whole island, following wherever banyans sprang up and scattering their houses (Bonnemaison, 1985, 34).

He goes on to say (Bonnemaison, 1985, 37) that "in this 'geographical society' man is compared to a tree whose roots thrust deep into the sacred earth. The banyans around the most prestigious dancing grounds are symbolic of men and bear the names of ancestors who founded the clans. If the banyan leaves the soil it dies, its land and political rights are extinguished, its magical and life-giving powers fade away."

Bonnemaison (1985, 39) also writes of the importance of trees in defining "social space," which is structured by networks of central places bearing symbolic and ritual significance: dancing grounds in the shade of the great banyan trees; sacred or "tabu" places connected with magic stones; dwelling sites and garden areas. Around this living heart, the peaceful dwelling place of followers of custom, there is usually a rather extensive and encircling belt of forest, punctuated by places of identity and security. If the territory's heart is an uninhabited homeland, its periphery is a forest given up to wandering evil spirits. Only by day and with precaution does one venture into these fringe areas, hunting or gathering reserves where from time to time men may make a few temporary gardens to cultivate food crops. The forest, although a place of uneasiness, is also a protection and, because of this, a controlled space.

Plant use on Tanna remains highly traditional and secretive, with all trees perceived to be useful and to possess spiritual significance or power. The traditional perception of trees in Tanna is well reflected in the response to questions about the cultural utility of almost any species: "some man uses it for medicine," which often implies magic.

As Bonnemaison (1986) remarked in a seminar on "Magic Gardens in Tanna," if the island's traditional gardens are still, today, filled with beauty and abundance, it is because they have gone back to their magical foundations, thanks in part to the John Frum movement. The gardens are beautiful because they are traditional and traditional because they are beautiful.

All forces of garden magic are localized in territorial networks of stones and places, all dedicated to the god of food and the master of all fertility magic, Mwatikitiki (a Polynesian name). Garden magic is practiced by garden magicians, or naotupunas, in sacred gardens, or nemai assim, which are "made" by the magician. Magic is practiced by "awakening power" in sacred stones by rubbing each with special assortments of leaves and tree barks from a precise set of plants that the magician alone knows. In the case of magic for a yam garden, for example, the leaves of Cordyline fruticosa, Mwatikitiki's personal emblem, are used. Each of the tasks a magician carries out in the sacred garden is then repeated simultaneously by all other members of the community in their gardens. In addition to his supernatural function, the garden magician is also a master of agricultural technology who guides traditional gardening. The sacred garden is then abandoned after the magician takes the first fruits, distributes them to members of the residence group in exchange for banana laplap (traditional pudding made from root crops or bananas mixed with coconut cream and cooked in an earthen oven), a pig, a fowl, or a kava root.

Such magic practices exist or existed for all plants traditionally grown in gardens: yams, taros, bananas, kava "native cabbages" (Hibiscus manihot), sugar cane, and so forth. Thus, our informants' insistence that virtually all native or aboriginally introduced plants on

Tanna were "used by someone for 'marasin' [medicine]," probably referred more to their traditional magical and spiritual importance and utility in the context of kastom than to "medicine" in the Western context. Thus, "medicine" encompasses many concepts including drugs, magic, fertilizers, pesticides, repellents, and others. Anywhere on Tanna or Aneityum, the great importance of a wide range of indigenous or aboriginally introduced plants, rarely found in agroforestry systems to the east of Vanuatu, probably reflects the resilience and deep significance of the traditional system as well as the richer indigenous flora of Vanuatu compared with the more oceanic islands to the east.

Agroforestry in household gardens and village groves

As in most areas of the Pacific, home gardens and groves surrounding villages in Vanuatu include many tree and shrub species, and seem to be the main avenue through which exotic species are introduced. The dominant tree species of home gardens and groves surrounding coastal villages on Aneityum include mangos, coconut, breadfruit, and Tahitian chestnut (Inocarpus fagifer), with other significant fruit-bearing species such as banana and plantain cultivars, papaya, citrus trees - including orange (Citrus sinensis), mandarin orange (C. reticulata), lime (C. aurantiifolia), and grapefruit (C. paradisi) - oceanic Iychee (Pometia pinnata), Malay apple (Syzygium malaccense), and several edible fig (Ficus) species.

Also common today are the culturally important banyans (Ficus spp.), Pandanus spp., and Casuarina equisetifolia. Species also seen in villages included Cerbera manghas, sandalwood (Santalum austrocalidonicum), and Securinega flexnuoa. Recently introduced species, most of which had no local names, included sweetsop and soursop (Annona squamosa and A. muricata), avocado, guava, strawberry guava (Psidium cattleianum), tamarind (Tamarindus indica), frangipani (Plumeria rubra), the poinciana (Delonix regia), Cassia fistula, Bauhinia sp., and Leucaena leucocephala. Conspicuous by their absence are exotic trees widely established in other areas of the Pacific, species such as rain tree (Samanea saman), African tulip tree (Spathodea campanulata), kapok (Ceiba pentandra), and Cananga odorata.

Shrubby species included the hedge panaxes (Polyscias guilfoylei, P. fruticosa, and P. balfouriana), the leaves of which are commonly cooked as green vegetables, hibiscus (Hibiscus rosa-sinensis), the copperleaf or beefsteak plant (Acalypha amentacea), Cordyline fruticosa, Euodia hortensis, and the important food plants, sugar cane and bush hibiscus spinach (Hibiscus manihot), known in Bislama (Vanuatu pidgin) as aelan kapis (island cabbage), of which there are numerous recognized cultivars.

In Middle Bush, Tanna, at an elevation of about 300 metres, home gardens in villages contain fewer trees than are found in villages along the coast because the people of Middle Bush can easily acquire many important plant products from the surrounding forest and their richly varied bush gardens. Trees that are commonly found in these home gardens include coconut palms, bananas and plantains, edible figs (Ficus spp.), citrus trees, Barringtonia edulis, Terminalia catappa, Pometia pinnata, guava, soursop and bullock's heart (Annona muricata and A. squamosa), and more temperate species, such as mulberry (Mows alba) and the common peach (Prunus persica), both of which bear fruit in Middle Bush. Other trees found around villages and in ceremonial dancing grounds, or nakamal, in or surrounding villages include large sacred banyans (Ficus spp.), Casuarina equisetifolia, and red-leaved varieties of Pipturus argenteus. Other less common species found in or surrounding villages include Finschia chloroxantha, with an edible fruit, and Solanum aviculare, plus a number of other unidentified species. Also common in Middle Bush home gardens are shrubby species with colourful foliage such as Cordyline fruticosa, Codiaeum variegatum, Breynia disticha, Coleus blumei, and Pseuderanthemum sp., all of which seem to be used in either traditional (kastom) medicine and magic or for body ornamentation. Conspicuously absent in Middle Bush village areas were mango and breadfruit trees, so common on the coastal strip at lower, warmer elevations.

Also important in the areas close to villages are species commonly planted as living fences or pig pens. These species include Erythrina variegata, Hibiscus tiliaceus, Pterocarpus indicus, Polyscias spp., and Jatropha curcas. Other food or shade plants allowed to grow within pig pens adjacent to villages included coconut palms, Morinda citrifolia, Glochidion ramiflorum, Terminalia catappa, and Ervatamia orientalist.

Garden and fallow-forest agroforestry

The agricultural landscapes and associated agroforestry systems in coastal areas and inland along river flood plains on Aneityum and Tanna - as well as at higher elevations on the volcanic soils of Middle Bush, Tanna - illustrate the great importance and diversity of trees and treelike plants. In coastal areas, agroforestry constitutes a mosaic of smallholder coconut plantations, active food gardens, and areas in various stages of fallow vegetation, all within a matrix of scattered useful trees. Small groves, containing both cultivated and wild species, are also common, as are stands of mature secondary forest and coastal strand forest. In Middle Bush, smallholder coffee plantations are found in some areas, but formal smallholder copra plantings are absent.

In coastal areas, within gardens and areas under short-term fallow, the coconut palm is ubiquitous. In more mature fallow areas, the coconut is also the most common tree, although a wide range of other trees is present. Other useful trees are sometimes scattered amongst coconut palms in copra plantations. Whereas in copra plantations, coconut palms tend to be evenly spaced and of a uniform age because of the active promotion of commercial plantation culture in the colonial past, the palms in food gardens and fallow areas tend to be randomly scattered and of different ages. Spriggs (1981, 83) notes that in pre-colonial times, coconuts would have been few in number compared with today's coconut overlay, although none the less of great subsistence importance.

Among the most common cultivated or protected tree species in plantations, gardens, and fallow areas are fruit-trees such as breadfruit, mango, papaya, citrus species, Tahitian chestnut (Inocarpus fagifer), Pacific Iychee (Pometia pinnata), beach almond (Terminalia catappa), and Malay or mountain apple (Syzygium malaccense). Also important in Vanuatu are other less widespread fruit-trees such as pilinut (Barringtonia edulis), canarium almond (Canarium indicum), Burkella obovata, Corynocarpus similis, dragon plum (Dracontomelon vitiense), and Adenanthera pavonina, together with a wide range of edible fig or banyan species (Ficus spp.).

Among tree crops in the pre-colonial economy, Spriggs (1981b, 84) believes lnocarpus fagifer to have been particularly important in times of failure of other crops. Breadfruit was important in some areas at certain seasons, as it is today. Of particular interest is the range of cultivated and selfsown edible figs, which include Ficus aspera, F. copiosa, and F. wassa, each of which has a number of distinguishable red- or green-leaved cultivars. In addition to their edible fruit, many Ficus spp. also have edible leaves, which are cooked as green vegetables and/or are of medicinal importance, as in the case of F. septica.

A wide variety of bananas and plantains (Musca cultivars) is also present in both lowland and upland gardens. Of lesser importance, but found occasionally in garden areas, are the soursop and bullock's heart (Annona muricata and A. reticulata), avocado (Persea americana), and jambolan (Syzygium cumin)). Cocoa, which is an important cash crop in the northern islands of Vanuatu, is also occasionally found in lowland gardens.

In Middle Bush gardens, coconuts are fewer, and mango and breadfruit absent, replaced by more temperate trees such as the mulberry (Mows alba) and the peach (Prunus persica). Finschia chloroxantha is also grown for its edible fruit.

Non-fruit trees commonly cultivated, often as living fence posts or boundary markers, include Erythrina variegata, Pterocarpus indicus, Hibiscus tiliaceus, Garuga floribunda, and Jatropha curcas. Casuarina equisetifolia is also common. Occasional in some plantations and garden areas are the exotic shade trees Delonix regia and Albizia lebbeck and the indigenous kauri Agathis obtusa. Stands of bamboo - including the presumably aboriginal introduction Schizostachyam glaucifolium, the recent introduction Bambusa vulgaris, and other locally recognized bamboo species or cultivars are also scattered throughout garden lands.

Other tree-like or shrubby food and non-food plants commonly cultivated in lowland or higher-elevation garden areas include sugar cane (Saccharum officinarum), bush hibiscus spinach (Hibiscus manihot), kava (Piper methysticum), panax (Polyscias spp.), tobacco (Nicotiana tabacum), Cordyline fruticosa, Ricinus communis, Pseuderanthemum sp., Codiaeum variegatum, Heliconia indica, and Crinum asiaticum, all of which have considerable cultural utility or subsistence importance. Important components are Hibiscus manihot, which is the main green vegetable and an exceptionally good source of vitamins and minerals, Saccharum officinarum, a major snack food and plant that features prominently in traditional legends, and Piper methysticum, the ubiquitous ceremonial beverage that is accorded such cultural, mystical, and magical significance, especially on Tanna. On Aneityum, the leaves of Heliconia indica, Crinum asiaticum, Scaevola cylindrica, Inocarpus fagifer, Acacia spirobis, and Epiprenum pinnatum are among those used by different gardeners for preparing fertilizer or mulch for swamp taro cultivation. On Tanna, the leaves of Crinum asiaticum are particularly important and used as "fertilizer" for yams, taro, sugar cane, and coconuts.

In both lowland and Middle Bush gardens, dominant and minor nontree crops are almost always intercropped with the tree or shrubby species already listed. Particularly on Tanna, planting of almost all these crops is accompanied with customary ritual and garden magic, with the leaves and bark of many trees and other plants commonly found in the gardens being used either in rituals (Bonnemaison 1986) or as fertilizers, mulch, pesticides, or taste enhancers (applied at the time of planting).

Other common species protected or encouraged in active garden areas or fallow regrowth, often through selective weeding, include widespread pioneer species such as Ficus spp., Hibiscus tiliaceus, Macaranga spp., Glochidion ramiflorum, Alphitonia spp., Morinda citrifolia, Commersonia bartramia, Pipturus argenteus, Trema orientalis, Cyathea spp., Grewia crenata, Ervatamia orientalis, Maesa nemoralis, Melochia odorata, Pittosporum spp., Tarenna sambucina, and Securinega flexuosa. Other common-tooccasional species include Leucaena leucocephala, Cordia myxa, and Melia azedarach.

A number of common coastal strand species, including Barringtonia asiatica, Calophyllum inophyllum, Cerbera manghas, Gyrocarpus americanus, Hernandia nymphaeifolia, Neissosperma oppositifolia, Premna serratifolia, and Scaevola sericea, were also commonly found planted or protected in lowland garden sites, while other more exclusively strand species, including Acacia simplicifolia, Caesalpinia bonduc, Colubrina asiatica, Cordia subcordata, Guettarda speciosa, Leucaena insularum, Ochrosia elliptica, Pemphis acidula, Pisonia grandis, Vitex negundo, and mangrove species (such as Bruguiera eriopetala, Excoecaria agallocha, and Heritiera littoralis), border coastal gardens, sometimes extending into the garden areas, and provide the very important service of protecting gardens from salt spray and coastal erosion. These species were restricted almost exclusively to coastal garden areas, although Cerbera manghas and Scaevola sericea were seen planted or protected in upland gardens above 400 metres.

Species that seem to be common in, or confined to, stands of mature fallow forest include Acalypha insulana, Adenanthera pavonina, Aleurites moluccana, Bischofia javanica, Canthium odoratum, Codiaeum variegatum, Croton insularis, Cryptocarya turbinata, Dysoxylum spp., Elattostachys falcata, Euodia spp., Ficus spp., Garcinia pancheri, Gardenia tannaensis, Garuga floribunda, Geniostoma rupestre, Litsia tannaensis, Manilkara dissecta, Mimusops elengi, Nauclea spp., Ochrosia odollam, Phyllanthus sp., Pisonia umbellifera, Polyathia nitidissima, Polyscias sp., Pterocarpus indicus, Semecarpus vitiensis, Syzygium richii, and Tarenna sambucina. Also common-to-occasional in lowland forests or mature secondary or fal low forest were mature fruit-tree species also found in active garden areas including Mangifera indica, Inocarpus edulis, Pometia pinnata, Terminalia catappa, Barringtonia edulis, Syzygium malaccense, Burkella obovata, Canarium indicum, Dracontomelon vitiense, and a wide range of edible fig or banyan species (Ficus spp.) and the palm, Veitchia sp.

Agroforestry in undisturbed forest

As noted by Bonnemaison (1985, 39), the relatively undisturbed forest areas that remain on Aneityum and Tanna constitute "controlled space" and "identity and security" to the local people and in Tanna are the "heart of their territory" and an "uninhabited homeland," that also serve as hunting and gathering reserves, where temporary gardens are established. As such, forest areas must be seen as integral to the total agroforestry resource.

The species found in forests undoubtedly include many of those inventoried in remnant stands of mature fallow forest described above, plus a wide range of other indigenous species.

Other species, many of which were not seen during the present study but were reported by Guillaumin (1931,1932), Schmid (1973), and Gowers (1976) to be present in lower-elevation and mature secondary forests on Aneityom and/or Tanna, include Aglaia eleagnoides, Alstonia villosa, Astronia aneityensis, Beilschmiedia sp., Bischofia javanica, Breynia sp., Cleidion spp., Cyrtandra sp., Dolicholobium aneityense, Endiandra aneityensi, Leucosyke spp., Maba buxifolia, Maesa efatensis, Melastoma denticulatum, Micromelon minutum, Mimusops parvifolia, Mussaenda frondosa, Myrishca inutilis Palaquium neo-ebudicum, Planchonella aneityense, Psychotria necdado, Randia sezitat, Serianthes spp., Sideroxylon aneityense, Sterculia tannaensis, and Tapeinosperma kajewskii.

In addition to species common in lower elevation forests, species seen or reported in mature secondary forest and forests surrounding garden areas at elevations between 200 and 500 metres in the environs of Middle Bush include Breynia sp., Croton sp., Cryptocarya tannaensis, Dillenia crenata, Elaeocarpus hortensis, Pittosporum campbelli, and Semecarpus tannaensis.

In more highly degraded areas and in higher elevations, common species include Acacia spirobis, Geissois denhamii, Metrosideros villosa, Weinmannia sp., together with shrubby species such as Vaccinium macgillivrayi.

Species found in swampy areas adjacent to garden lands include the sago palm (Metroxylon sp.) and Barringtonia racemosa.

Fijian agroforestry at Namosi and Matainasau

Namosi and Matainasau villages are located in the interior wet zone of Viti Levu, the main island of Fiji, some 56 and 95 km respectively from the capital of Suva. The village lands extend from elevations of about 150 to 1,000 metres, with much of the land in slopes steeper than 50 per cent. Rainfall ranges from 250 to 500 cm per year, although it can be as high as 900 cm. Both villages are situated on major rivers, Namosi on the upper Waidina River and Matainasau on the Wainimala River. Both have access to sizeable areas of alluvial land and extensive areas of more mountainous agricultural and forest land. The humic latosols that predominate in the upland areas are characterized by stony clay, stony sandy clay, and stony silt to 60 cm in depth. Although prone to erosion, these soils have relatively high fertility and are particularly suited for short-term root cropping, yaqona (kava) production, or long-term tree cropping. The alluvial sandy loams and sandy soils of the alluvial flats and colluvial areas have relatively high fertility when drained (Groom and Associates 1981).

In the early 1980s, it was estimated that there were some 45 households in Namosi, with a total population of approximately 250 (Rizer et al. 1982). The population of Matainasau is slightly less, with some 30 households in the mid1980s.

Cultural importance of trees and forests

The Fijian term for land, vanua, "has physical, social and cultural dimensions which are interrelated" (Ravuvu 1983, 70). These include the vegetation and animal life as well as the social and cultural system. It follows that all trees on a community's land are seen to be integral to the whole agricultural system and to human welfare. In this context, the major agroforestry land-use zones would include:

  1. the village site (koro);
  2. the areas surrounding the village (bill ni koro);
  3. the agricultural lands (qele ni teitei), including active gardens (vei were) and fallow areas (raki) on the alluvial flats (buca);
  4. agricultural lands (qele ni teitei), including active gardens (vei were) and associated fallow areas (raki) on rolling colluvial and mountain soils of the uplands (vei delana);
  5. secondary forest areas (veikau) on both alluvial flat and upland areas;
  6. dense primary forest (lekutu or veikau loa);
  7. scattered fenced beef-cattle pastures (loma ni ba or ba ni bulumakau); and
  8. river bank or riparian areas (bati ni wad).

During R. Thaman's in-depth field studies of the agroforestry systems of Namosi and Matainasau, conducted from 1979 to 1988, over 100 trees or tree-like species or cultivars were encountered. Most of these were in existing agricultural areas, rather than in surrounding primary forest stands, although some of the forest species are occasionally found as protected individuals in recently cleared upland garden sites. In keeping with the vanua concept, veikau, or forest, areas are seen by Fijians as integral components of the wider land-use system, and were generally used by the entire community, regardless of the more restrictive clan (mataqali) affiliations required to obtain access to agricultural lands. Forest areas supply materials for construction and firewood and provide a domain for hunting and foraging. Although the pressure for commercial logging in such areas has made communal use of forest lands more restrictive, access to more distant forest areas is still very much open, as long as a member of the community is extracting resources for personal or communal use, rather than for commercial purposes (Ravuvu 1983,74).

Important cultigens

The most common cultigens are bananas and plantains (Musa cultivars), kava, or yaqona (Piper methysticum), hibiscus (Hibiscus manihot), Citrus species, breadfruit (Artocarpus altilis), duruka (Saccharum edule), sugar cane, papaya, coconut, Malay or mountain apple (Syzygium malaccense), Polynesian viapple (Spondias dulcis), soursop (Annona muricata), vutu kana (Barringtonia edulis), guava (Psidium gunjava), cocoa (Theabroma cacao), jakfruit (Artocarpus heterophyllus), two palms (Veitchia joannis and Pritchardia pacifica), now found more often growing wild, and waciwaci (Sterculia vitiensis). All are commonly planted or protected in active garden areas, fallow areas, and in tree groves around villages, although, as Ravuvu (1983, 73) reports, they are "usually grown in small patches away from the land used for root crops." The mango and the avocado are found infrequently because they seem to bear few fruit in wet inland areas, but are common, especially the mango, in drier coastal agri cultural areas. Although still present, Veitchia joannis, Pritchardia pacifica, and Sterculia vitiensis were all more common in the past.

In active garden areas, these species are generally found interspersed with the dominant staple food crops in both upland and alluvial lowland and river terrace gardens, as well as remaining there throughout fallow periods, which traditionally ranged from 5 to 15 years, with cropping periods of 2-7 years. Correspondingly longer cropping periods and shorter fallow periods are characteristic on the richer alluvial and colluvial soils nearer the villages. Although burning of debris cleared from new garden patches is practiced widely in Fiji, including Namosi, the practice has been traditionally discouraged at Matainasau because it was believed to have deleterious effects on soil and arboreal regeneration.

In the upland garden areas, taro and cassava are the dominant ground crops. Taro is generally planted after clearing and is intercropped with kava as a co-dominant crop. Cassava is planted next, sometimes up to three or more times in succession. Less common crops or intercrops in these gardens include yams (Dioscorea alata), wild yams (D. nummularia) (which are both cultivated and grow wild in fallow and secondary forest areas), tannia (Xanthosoma sagittifolium), giant taro (Alocasia macrorrhiza), and giant swamp taro (Cyrtosperma chamissonis), which is occasionally found growing wild, although probably originally planted, along small streams and poorlydrained areas bordering the garden areas. One factor responsible for decreasing fallow periods and increasing cropping periods and associated agrodeforestation has been the propensity of Fijians to abandon more labourintensive traditional crops, such as yams and taro, in favour of cassava, which is less often intercropped and which requires little or no fallow between successive plantings (Thaman and Thomas 1982, 1985).

The same crops are also found in alluvial soils, although kava is less commonly planted there; sweet potatoes (Ipomoea batatas) grow particularly well in sandy alluvial soils. Some non-traditional vegetables and fruits, such as Chinese cabbage (Brassica chinensis), tomatoes, long beans (Vigna sequipedalis), corn (Zea mays), and eggplant (Solanum melongena), are increasingly intercropped between staple root crops during the early stages of gardens, particularly on the alluvial flats near villages, where water for shortterm seed crops is readily available. Pineapple (Ananas comosus) is also increasingly common.

Of the tree crops, banana and plantain cultivars are probably of greatest economic and subsistence importance, with a range of plantain (vudi) cultivars (Musa AAB Group) found in most gardens as an important staple. Vudi cultivars scattered throughout garden areas may cover as much as 5 per cent of total food-crop area. As perennials, plantains can continue bearing without replanting for 10 years or more if kept free of diseases and weeds.

The common banana of commerce, the Musa AAA Group, which was formerly an important export crop in both the Matainasau and Namosi areas, is still common in gardens, especially on alluvial and colluvial soils. It remains an important local cash crop in Matainasau village. Although bananas are susceptible to black-leaf-streak fungal and bunchy-top viral diseases, the wide dispersal of banana plants and their intercropping with other species controls damage from these diseases, which proved to be the death knell of banana monocropping for export to New Zealand and Japan in the mid-1960s. Other banana cultivars include the liga ni marama, or lady's finger, banana (Musa AAB Group), the bata, or blue Java plantain (Musa ABB Group), and the vinvialevu or qamure (an uncommon plantain cultivar). Musa fehi, the wild banana, which is occasionally cultivated now, was more widely cultivated in the past; it remains as a relict here and there in old secondary vegetation.

Piper methysticum, a tree-like shrub known widely in the Pacific as kava (yaqana in Fijian), is currently the most important cash crop in both Namosi and Matainasau. It is planted as a monocrop but more commonly intercropped with taro (Colocasia esculenta) and other crops, which are harvested first, leaving the kava to mature over its 4-7year optimum yield cycle. Because kava cultivation has been newly extended into more-distant upland areas with the spread of road transport, the new gardens often have fewer deliberately planted cultivated trees than is the case in the more traditional garden areas closer to the villages. As with bananas, there is considerable intraspecies diversity, with at least six recognized kava cultivars in the Namosi area. Being relatively shade resistant, small plantings of kava are commonly found in tree-dominated gardens near settlements.

Although more a shrub than a tree, vauvau, bele, or hibiscus spinach (Hibiscus manihot) is a perennial that, under good conditions, can grow to over 3 metres in height. Along with taro-leaf spinach and many wild ferns, which are found throughout garden and fallow areas, bele is one of the main green vegetables of Namosi and Matainasau villages. It is reportedly one of the most nutritious green vegetables, being very high in vitamins A and C and in iron, and having 12 per cent protein by dry weight (Standal et al. 1974), which makes it a valuable food in interior villages where animal protein is scarce. It is easily propagated from cuttings, requires little cultivation, is relatively disease-resistant, and people like it as a vegetable and also consider it to have medicinal value. Planted along borders of gardens or as an intercrop throughout gardens, it yields a valuable food for a long time.

The wide range of Citrus species cultivated and protected as volunteers throughout garden and fallow areas and around villages are a major economic resource in Namosi, which is renowned for its mold (known in the local dialect as soco), or oranges. The fruits produced include the sweet orange, or mold Taiti, literally Tahitian orange (C. sinensis), and the mandarin orange, or mold madarini (C. reticulata), which are sold in thousands at the Suva Municipal Market (Thaman 1976177) and provide a major seasonal source of cash as well as a very nutritious snack food. The rough lemon, or moli karokaro (C. hystrix), is also common, especially in fallow areas, and is widely used to marinade raw fish or squeezed on a wide range of foods, as well as constituting one of the main ingredients, along with coconut cream, chillies, onions, and salt, of the sauce known as mitt. The young leaves are also used to make tea (drau ni molt). The sour orange, or mold hula (C. aurantium), the kalamantsi, or mold witiwiti (C. micro-arpa), and the pomelo, shaddock, or mold kana (C. grandis), are also commonly cultivated or protected for use in drinks, for squeezing on food, and, in the case of the sour orange, for eating and occasional selling at the market.

The coconut and the breadfruit, although common in garden areas and around villages, especially on the alluvial flats, are not as dominant as in coastal areas. Nevertheless, the coconut remains a very important supplementary staple. The cream expressed from the flesh of the endosperm of mature nuts is widely used in cooking, the green nuts for drinking, and the mature nuts are commonly fed to poultry and pigs. Similarly, breadfruit constitutes an important seasonal staple. Both trees, but especially the coconut, have many non-food uses, and the coconut, the Pacific's "tree of life," is an excellent arboreal intercrop because of its small crown structure. At least five cultivars of coconut and three of breadfruit are recognized in Namosi. Two other palms with edible seeds and useful fronds, Veitchia joannis and Pritchardia pacifica, are found occasionally, often growing wild, but reportedly were planted more widely in the past.

Unlike commercial sugar cane cultivation, where the entire crop is taken at each harvest, the many traditional, aboriginally-introduced cultivars of cane are kept in tree-like clumps and continually harvested to chew as an energy-rich snack food. Functionally, sugar cane, which is widespread in most gardens, constitutes an important "agroforestry" species. The closely related duraka (Saccharum edule), known locally as "Fiji asparagus," provides an edible inflorescence and is found in a cultivated or almost wild state throughout garden areas, especially in poorly drained alluvial sites. In Namosi, where it is an important seasonal food and source of cash income, some nine cultivars are recognized.

Also of considerable economic and subsistence importance is a range of other traditional and more recently introduced fruit-trees, found planted or protected. Traditional - possibly aboriginally introduced- trees that provide seasonal flushes of fruit for consumption and sale include the kovika, or Malay or mountain apple (Syzygium malaccense), the wi, Polynesian vi-apple or hog plum (Spondias dulcis), and the vutu kana (Barringtonia edulis). Of the recent introductions, papaya (Carica papaya) is particularly common and an excellent non-seasonal vitamin- and mineral-rich fruit; seremaia, or soursop (Annona muricata), is common in gardens and around villages; and the uto ni Idia, or jakfruit (Artocarpus heterophyllus), is increasingly common. Guava is an important seasonal fruit; it is generally found growing wild, especially where livestock have been grazed, but is occasionally planted or protected in gardens or village areas. The mango, which is believed to be a post-European introduction to Fiji, and the avocado are only occasionally found and bear little fruit in the cold, wet conditions of Namosi and Matainasau.

Senile experimental plantings of cocoa (Theobroma cacoo) and coffee are found at Namosi, with scattered trees found at Matainasau. No cocoa or coffee is produced, although the ripe cocoa fruit pulp is consumed as a snack food and occasionally sold. Cocoa is, however, an increasingly important commercial crop in some areas, such as the nearby Wainibuka River valley, and does offer some prospect for agroforestry cash cropping.

Of cultural importance are a number of cultivated non-food plants. Voivoi, or pandanus (Pandanus spp.), of which there is a range of cultivars, is used in the production of plaited ware such as ceremonial mats, rough mats, baskets, and hats; the many cultivars of the ti plant (Cordyline fruticosa) are used as ornamentals in villages and to provide the most commonly used leaves in traditional dance costumes.

The bright red cultivars are planted as `'protective" plants in gardens to ward off evil spirits and to ensure good yields, while the larger darkgreen variety, vasili ni Toga, was formerly an important famine food and source of sugar from the root, which was baked in an earthen oven for four days. Other non-food plants include panax (Polyscias spp.), a common ornamental hedge, living fence, boundary marker, and ornamental plant of medicinal value, which is commonly planted in villages and occasionally in gardens; kalabuci (Acalypha wilkesiana), also a common ornamental and hedge plant; bua ni Vavalagi, or frangipani (Plumeria spp.), widely planted in and around villages as an ornamental and for the use of its fragrant flowers in garlands; banidaki, wiriwiri, or physic nut (Jatropha curcas), has medicinal value and is planted as living fencing; vauvau, vauvau ni Vavalagi, or kapok (Ceiba pentandra), is occasionally planted around villages as a source of fibre to fill mattresses and pillows; uci (Euodia hortensis), a shrub of medicinal value with pungent flowers and leaves used in garlands and to scent coconut oil, is commonly planted in villages and occasionally in garden areas; vasa damu (Euphorbia fidjiana), an attractive rust-redcoloured tree-like shrub, is planted as a protective plant to ward off evil spirits and to ensure good yields; and sago palm (Metroxylon vitiense) provides a favoured thatching for roofs. Other shrubby ornamentals commonly planted in and around villages are Dracaena fragrans, Bougainvillea spp., Pseuderanthemum curruthersii, P. bicolor, and Groptophyllum pictum.

Non-cultigens

Non-cultivated or self-sown food trees such as the oceanic Iychee (Pometia pinnata), Tahitian chestnut (Inocarpus fagifer), Ficus vitiensis, the beach or Indian almond (Terminalia catappa), Adenanthera pavonina, Elaeocarpus chelonimorphus, and sukau (Gnetum gnemon) are also commonly protected in garden areas, around villages, and in forest stands. Other less commonly used wild food sources, which were reportedly more widely used in the past, include vava (Heliconia indica), the flower bracts of which are baked or roasted; losilosi (Ficus barclayana), the leaves and fruit of which are eaten and used medicinally; and waciwaci (Sterculia vitiensis), the edible seeds of which were roasted or fried in the past over open fires.

Of almost tantamount economic, ecological, and cultural importance to the many cultivated and wild food trees and the cultivated non-food trees is the host of other useful non-cultivated species that are commonly found, often as dominants, in agroforestry areas. These trees provide timber, fuel wood, medicines, fibre, perfumes, and dyes; they also give shade, control erosion, improve soil, offer habitats to wildlife, and may have considerable spiritual importance. These and other uses are specified in the final section of this chapter. In terms of medicinal importance, for example, field surveys and Weiner's (1984) study of Fijian medicinal plants indicate that almost 50 per cent (47 of 101) of the plants listed in this chapter's final section have some medicinal use.

Most of the non-cultivated agroforestry species are indigenous and some are endemic. Others such as Bischofia javanica, Hibiscus tiliaceus, Canangaa odorata, and Aleurites moluccana may be aboriginal introductions. Although most are generally found growing naturally, some - such as Hibiscus tiliaceus, Erythrina variegata, Premna obtusifolia, and Dillenia biflora- are commonly planted as living fencing, and others - such as Parinari glaberrima and Cananga odorata, which are particularly valued- are commonly planted around villages. Trees that are sacred totems, or i cavuti, of the various descent groups (mataqali) of Namosi include mako (Cyathocalyx vitiensis), bua (Fagraea berteriana), bitu (Schizastachyam glaucifolium), and niu, the coconut (Cocos nucifera).

Exotic trees of widespread agroforestry importance include yaqona ni Onolulu (Honolulu), Onolulu, or qonaqona (Piper aduncum); bitu ni Vavalagi, or common bamboo (Bambusa vulgaris); and two species referred to as vaivai ni Vavalagi (literally, foreign vaivai), leucaena (Leucaena leucocephala) and the rain tree, or monkey-pod (Samanea saman). Piper aduncum is a weedy large shrub to small tree abundant throughout garden and fallow areas, often being coppiced as a readily available source of fuel wood from active garden areas. Bambusa vulgaris is occasionally cultivated, but is usually found wild and protected in large stands on the alluvial flats and occasionally in upland garden areas to provide a ready source for construction materials of wide utility, including bamboo rafts, or bilibili, which were a main means of transporting agricultural produce before road access to urban markets became available in the past 10 years or so, and to provide a ready source of fuel, especially on rainy days, when other sources of fuel are wet. Leucaena leucocephala is locally important, especially in grazing areas, and is commonly used for fencing and construction of pig pens and as a source of firewood and fodder. Samanea saman is commonly planted or protected around villages or in grazing areas as a shade tree, and is particularly common on the alluvial garden flats near Matainasau, where it is an important source of timber, fuel wood, and wood for carving.

Importance of wild foodstuffs

Also of importance to inland Fijian villages is the great diversity of nonarboreal wild plant and animal foodstuffs, medicines, and other useful products found within the matrix of the agroforestry system in active garden areas, fallow vegetation, and secondary forests, in primary forests bordering active agricultural areas, and in or along streams that border or flow through agroforestry areas and that, if agrodeforestation continues, will be considerably impoverished (Thaman 1982b).

Most notable among the non-arboreal wild foodstuffs is an almost baffling diversity of wild yams, the most important species being Dioscorea nummularia, D. pentaphylla, and D. bulbifera, and ferns, mostly referred to as wata or ota, the most commonly consumed species being Athyrium spp., Diplazium spp., Tectaria latifolia, Stenochlaena palustris, and Marattia smithii. Black nightshade (Solanum nigrum); sou (Solanum repandum); moca, tubua, or spleen amaranth (Amaranthus viridis); wild chill) peppers (Capsicum frutescens); kosopeli, or cape gooseberry (Physalis angulata); inoka, or kudzu root (Pueraria lobata), and a diversity of daliga, or fungi, that are commonly found on dead trees, are all common wild foods from agroforestry areas. It is estimated that these more common species, together with less important species and the previously mentioned tree or tree-like sources of wild food products, number over 60 in the Namosi area alone. When the wide range of edible birds, frogs, snakes, grubs, insects, fishes, eels, freshwater prawns, and other foods that are found within agroforestry zones is included, the significance of wild food resources to mountain villages becomes obvious. Moreover, apart from being nutritionally important - particularly in the cases of some seasonally abundant fruits, nuts, wild yams, and wild greens - these wild products also constitute important low-capitalinput, low-risk cash "crops" for seasonal sale at the Suva Municipal Market (Thaman 1976/77).

Importance of village tree groves

Also of agroforestry significance are the relatively dense groves of mature trees found around the boundaries of villages. These groves, which are present at Namosi and some other villages in Fiji, include a wide range of cultivated trees of high utility and easy accessibility. Important species are sweet oranges, mandarin oranges, rough lemon, breadfruit, jakfruit, coconut, bananas and plantains, Tahitian chestnut (Inocarpus fagifer), oceanic lychee (Pometia pinnate), coral tree (Erythrina variegate), and Cordyline terminalis, along with other trees of medicinal or spiritual importance, such as those used in ceremonial garlands or for scenting coconut oil.

Pressures toward agrodeforestation

Over 100 trees or tree-like species are found in the agroforestry systems of Namosi and Matainasau villages; collectively they represent a resource of enormous economic, cultural, and ecological importance. These trees, along with the many other species found in surrounding forest stands, have been preserved as part of an integral agroforestry system for generations but are almost totally neglected by most present-day agricultural developers and researchers. Consequently, although the agroforestry systems of both villages remain relatively intact, recent pressures to encourage cash cropping of bananas, cocoa, Lava, and root crops, and to develop commercial livestock grazing, have led to deforestation and agrodeforestation. The new generation of farmers, which has not been educated to see the long-term utility of integrated agroforestry, neglects the trees. As an example of the richness and diversity of this agroforestry resource, table 5 in the final section of this chapter lists the important agroforest trees and some tree-like plants recorded in the landscapes of Namosi and Matainasau villages and provides a brief description of their uses and significance. Lack of space prohibits the presentation in this book of the similar sort of information available for all the Pacific Island regions discussed here, but an aggregate listing of 100 important agroforest species is offered in the Appendix.

A listing of agroforestry components in the landscapes of Namosi and Matainasau

The following table contains a list of important tree and tree-like species of agroforestry systems of the Namosi and Matainasau village landscapes in Fiji.

It should be noted that under "vernacular names," B refers to the common Bauan dialect, used as the Fijian lingua franca; N refers to the Namosi dialect, and M to the Matainasau dialect. The common

English or other widely used vernacular names are given with no further designation. In Fijian orthography, the "c" is pronounced like the "th" in the word "the," the "d" like the "nd" in "candy," "g" as the "ng" in the word "sing," and "q" like "ngg."

Table 5 Tree and tree-like species in the agroforestry systems of the villages of Namosi and Matainasau, Fiji

Latin name Vernacular names Notes
Acalypha wilkesiana kalabuci, kalabuci damu (N,M,B); Joseph's coat, beefsteak plant, copperleaf Common in villages, and occasional in garden areas; important planted ornamental and hedge plant; used medicinally.
Adenanthera pavonina vini (N); vaivai (M); lera (B); red-bead tree eaten. Occasional in garden areas and around villages in tree groves; timber used in house building and for firewood; seeds
Aleurites moluccana lauci (N,M,B); waiwai (N); candlenut Occasional in garden and fallow areas, in secondary forests, and around villages; seeds used to produce oil beneficial to the skin; used medicinally.
Alphitonia zizyphoides doi (N,M,B) secondary forests; timber useful; good Occasional in garden areas and in firewood; used medicinally.
Alstonia vitiensis sorua (N,M); bulei (B) Occasional in garden areas and in secondary forests; good fuel wood; sap dried for chewing gum.
Annona muricata seremaia (N,M,B); soursop Common around villages, and occasional in garden areas; ripe fruit eaten and made into drinks.
Antidesma classophylum molau (N,M,B); molau vuloa (N) Occasional in fallow areas; leaves used medicinally; good fuel wood when thoroughly dried.
Artocarpus altilis uto sori (N); uto (N,M,B); breadfruit Common around villages, in garden areas, in secondary forests, and in village tree groves; fruit cooked as an important seasonal staple; leaves used for wrapping food for cooking in the earthen oven or boiling; occasionally sold; used medicinally; various cultivars recognized.
Artocarpus heterophyllus uto ni Idia (N,M,B); jakfruit Occasional around villages in garden lands; fruit eaten and occasionally sold.
Astronidium confertiflorum sakelo (N); dava (M) Occasional in secondary forest in garden areas; used as fuel wood.
Bambusa vulgaris bitu ni Vavalagi (N,M,B); common bamboo Common in garden areas, fallow and secondary forest areas, and around villages; large stems used for housing and other construction purposes, for construction of bamboo rafts (bilibili), for ladders, and smaller sections for steaming food in the earthen oven or over open fires.
Barringtonia edulis vatu kana, vatu (N,M,B) Occasional in garden areas on alluvial flats, in tree groves, and near or in villages; seeds eaten raw as a snack food; occasionally sold.
Bischofia javanica koka (N,M,B) Abundant in garden and fallow areas, and common on alluvial flats; protected in the past when clearing garden lands, but increasingly felled by younger generations; favoured tree for house posts, good firewood, used medicinally, leaves boiled with pandanus leaves to dye them black; bark formerly used

with other ingredients to dye hair red; its presence believed to be a sign of good soil.

Bougainvillea spp. bougainvillea Common ornamental in villages.
Buchanania atrenuata kaukaro (N) Infrequent in secondary forest areas; used medicinally.
Cananga odorata makosoi (N,M,B); perfume flower, ylangylang Occasional in garden areas and around villages; leaves used medicinally; fragrant flowers used in garlands and for scenting coconut oil.
Carica papaya uto, uto maoli (N); weleti (M,B) Common in garden and fallow areas and around villages, both planted and as a volunteer; ripe fruit eaten raw; green fruit occasionally cooked or mixed with meat as a tenderizer; used medicinally.
Ceiba pentandra vauvau, vauvau ni Vavalagi Occasionally planted in garden areas and around villages; occasionally planted as living fences; fibres surrounding seeds used to stuff pillows and mattresses.
Citrus aurantium mold kula (M,B); kula (N); sour orange Occasional in garden and fallow areas and around villages; fruit used to make drinks, to squeeze on food, and occasionally sold.
Citrus grandis mold kana (N,M,B); soco vi kana (N); pomelo, shaddock Occasional in garden and fallow areas, in secondary forests, and around villages; excellent firewood; ripe fruit eaten and occasionally sold; used medicinally.
Citrus hystrix mold karokaro (N,M,B); soco ni Vavalagi (N) Common in garden and fallow areas, and occasional in secondary forests and village tree groves; juice of fruit used to marinate raw fish and to flavour foods, young leaves used to make tea; used medicinally; occasionally sold; good fuel wood.
Citrus microcarpa mold witiwiti (N); kalamantsi (Philippines) Occasional in garden areas, tree groves, and around villages; ripe fruit used to make drinks and to squeeze on food.
Citrus reticulata soco madirini (N); mold madirini (M,B); mandarin orange, tangerine Common in garden areas, fallow vegetation, around most villages, and in community tree groves, and occasional in secondary forests; ripe fruit eaten and made into drinks; sold as a major seasonal source of cash income.
Citrus sinensis soco ni Taiti (N); mold lecau (M); mold Taiti (B); orange, sweet orange Common in garden areas, fallow vegetation around villages, and in community tree groves, and occasional in secondary forests; ripe fruit eaten and made into drinks; sold as a major Seasonal source of cash income; used medicinally.
Cocos nucifera niu (N,M,B); coconut palm Occasional in garden areas, and more common on alluvial flats and surrounding villages; a minor food plant in the interior, unlike in coastal areas, where it is a major source of vegetable fat and energy in cooking and a poultry and pig feed, and is used in a great variety of ways; still an important intercrop, but more important in the past; the totem of the Natuvora descent group (mataqali) of Namosi village; used medicinally.
Commersonia bartramia sama (M) Common in garden and fallow areas, in secondary forests, and along road frontages; good fuel wood; inner bark used for string and lashings; used medicinally.
Cordyline fruticosa vasili (N,M,B); ti plant Abundant in villages, common in garden areas, and occasional as an escape in fallow areas; a wide variety of cultivars exists; bright red varieties commonly planted as 'protector" plants to ward off evil spirits and to bring good luck to gardens and as ornamentals in villages; large roots of green varieties (vasili ni

Toga) cooked in earthen oven in the past as a food, now only as a famine food; leaves of many varieties used in ceremonial skirts and other ornamental attire; used medicinally.

Cyathea spp. balabala (N,M,B); tree fern Common in fallow areas and secondary forest, and occasional in garden areas; trunk used in house construction and occasionally sold to and used by ornamental gardeners as orchid

supports; when dry, used to carry fire long distances before matches were available.

Cyathocalyx vitiensis mako (N,M,B) Occasional in garden and fallow areas, and in secondary and primary forest; important totem (i cavuta) of the high ranking Nabukebuke yavusa and mataqali descent groups of Namosi village; used for firewood; inner bark Occasionally used for string or lashing.
Cyrtandra jugalis micra (N,M) Occasional in garden and fallow areas and in secondary forests; timber used for house frames.
Dendrocnide harveyl salato (N,M,B) Infrequent in fallow areas and in secondary forests; used medicinally.
Dillenia biflora kulava (N,M,B) Occasional in garden lands and around villages, common in secondary forests; planted as living fencing, pig pens, and bathhouse walls.
Dracaenafragrans dracaena Common ornamental in villages.
Dysoxylum richii vesida (N); sasawira (M); sasauira (B) Common in secondary forests, and occasional in garden areas; timber used in construction and for firewood.
Elaeocarpus chelonimorphus kabi (N) Occasional in garden and fallow areas and in surrounding secondary forest; seeds eaten as a snack food and favoured food of fruit-bats.
Elattostachys

falcata

isusu (N) Occasional in fallow and secondary forest areas; timber used in construction and for firewood.
Endospermum macrophyllum lekutu (N); lekuti (M); kauvula (B) Occasional in secondary forests, protected in garden lands, and common in surrounding forests; wood used in house construction, for dug-out canoes, and for fire wood; an important timber species for local milling and export.
Erythrina variegata drala (N,M,B); coral tree, dadap Common in garden areas and around villages; commonly planted as property markers, living fences, or pig pens; flowering indicates the onset of the yam planting season; used medicinally.
Euedia hortensis uci (N,M,B) Occasionally planted in garden areas and common in and around villages; flowers and leaves used to scent coconut oil, to make garlands (salusalu), and commonly worn behind the ear; leaves used medicinally.
Euodia vitiensis tokatolu (N) Infrequent in secondary forests; leaves used medicinally and taken by women after giving birth.
Euphorbia fidjiana vasa damu (N,M,B) Shrub occasionally planted as a "protec tion" or spiritual plant to protect gardens from evil spirits and to ensure good yields; used medicinally.
Fagraea berteriana bua, bua ni Viti (N,M,B); bua tokaikau (N) Occasional in garden and fallow areas, in secondary forests, and around villages; fragrant flowers used in garlands and for scenting coconut oil; the totem (i cavati) of the Nasilime mataqali, or descent group, of Namosi village; used medicinally.
Ficus barclayana losilosi (N,M,B) Occasional in garden and fallow areas, in secondary forest, and around and in villages; leaves and fruit eaten; leaves use medicinally.
Ficus fulvo-pilosa al mast (N) Occasional in garden and fallow areas and in secondary forest; leaves used as sandpaper in the past.
Ficus obliqua bake, baka ni Viti (N,M,B); strangler fig; banyan Occasional in garden areas and common in secondary forest; fruit eaten by birds and fruit-bats; used medicinally.
Ficus vitiensis lolo (N,M) Occasional in garden and fallow areas and in secondary forests; sweet fruits eaten as a snack food; used medicinally.
Gironniera

celtidifolia

sisisi (N) Infrequent in secondary forests and in fallow areas; wood good for making digging sticks.
Glochidion spp. molau (N,M,B) Common in fallow vegetation, and occasional in garden areas and secondary forests; leaves used medicinally; one of most widely used medicinal plants.
Gnetum gnemon sukau, bele sukau (N,M,B) Occasional in secondary forest and in forest areas surrounding garden areas; seeds edible, young leaves cooked as a spinach.
Graptophyllum pictum caricature plant Occasional ornamental in villages.
Heliconia indica vava (N,M,B) Occasional in garden areas, fallow areas, and secondary forests; flower bracts cooked as a famine food; leaves used to parcel food.
Hibiscus manihot vauvau (N,M); bele (B) Abundant in gardens and around villages; major leafy green vegetable, very rich in vitamins A and C, iron, and plant protein; commonly sold in markets; used medicinally.
Hibiscus tiliaceus vau (N,M,B); beach hibiscus tree Common in fallow and garden areas, and around villages; commonly planted as living fences and pig pens; wood used in walling for houses, inner bark as fibre for skirts, for straining yaqana, and for house and canoe lashings; leaves used medicinally.
Homolanthus nutans tadau (N); tadanu (B) Occasional in fallow areas and secondary vegetation; used for firewood.
Inocarpus fagifer ivi (N,M,B); Tahitian chestnut Common in poorly drained areas, along rivers, and in tree groves, and occasional in alluvial and some upland garden areas and around villages; seed cooked as a supplementary staple; leaves used to cover food in earthen oven; good fuelwood; used medicinally; sold occasionally.
Jatropha curcas banidaki (N,M); wiriwiri (B) Occasional in garden areas and around villages; planted as a living fence; used medicinally.
Leucaena leucocephala vaivai, vaivai ni Vavalagi (N,M,B); leucaena Common locally as a long-introduced inventive in garden areas, around villages, and in some areas where cattleare grazed; good source of firewood, posts for fencing and wood for general construction purposes, such as animal

pens; foliage and young pods used as fodder and occasionally eaten.

Macaranga graeffeana mavu (N,M); davo (B) Common in garden and fallow areas, and as a pioneer species in secondary forests; wood used for construction and for firewood; used medicinally.
Macaranga harveyana gadoa (N,M) Common in fallow areas and secondary forest, and occasional in garden areas; wood used in construction and for firewood; used medicinally.
Mangifera indica maqo (N,M,B); mango Occasional in garden areas and surrounding and in villages; ripe fruit eaten; fruit rarely sets at wet higher elevations.
Melochia mollipila samaloa (N) Occasional in fallow areas and secondary forests; good firewood; used medicinally.
Metroxylon

vitiense

soya (N,M,B); sage palm Occasionally planted along rivers or around villages and gardens; fronds used for thatching houses; maristem sometimes sold to Indians for use in curries; leaf ribs used for brooms.
Morinda citrifolia kura (N,M,B) Occasional in garden and fallow areas, in secondary forest, and around villages; used medicinally; pungent fruit eaten in the past.
Musa AAA Group jaina(N,M,B); banana, Cavendish banana Common intercrop in garden areas especially on alluvial flats; important cash crop for local sale and eaten green as a supplementary staple and ripe as a fruit; leaves used to wrap food for cooking and for general food parcelization.
Musa AAB Group liga ni marama (N,M,B); lady's-finger banana Occasional in garden areas and around villages; ripe fruit eaten, very minor cash crop for local sale; leaves used to wrap food.
Musa AAB triploid vadi (N,M,B); plantain Abundant in garden areas, in both alluvial flats and on hill slopes; common intercrop in almost all gardens; green fruit an important staple, ripe fruit occasionally cooked as a dessert (vakasoso); leaves used to wrap food for cooking in the earthen oven and for general food parcelization; important supplemental cash crop for local sale.
Musa ABB Group bata (N,M,B) ; blue plantain Occasional in garden areas; green fruit cooked as a Java supplementary staple and eaten ripe occasionally; leaves favoured for parcelization of foods for earth oven.
Mussaenda

raiateensis

Hobo (N,M); bovu (B) Common in fallow areas and in secondary forest; bark and leaves used medicinally.
Musa cultivar viavia levu (M); qamure (B); plantain Uncommon banana or plantain cultivar in garden areas; green fruit cooked as a staple.
Musa fehi sowaqa (N,M,B); fe'i banana, mountain banana Occasional in secondary forest as a reict

of former cultivation, occasional in garden areas; fruit cooked as a staple and occasionally sold at the SuvaMarket.

Neonauclea

forsteri

bo (N,M) Occasional in garden areas and secondary forests; wood used in construction.
Neubergia

corynocarpa

bo nokonoko (N); boloa (N,M) Occasional in garden and fallow areas and in secondary forests; timber used in house construction.
Pandanus spp vadra, voivoi (N,M,B); pandanus, screw pine Commonly planted in garden areas and in and around villages, often along fence lines or garden boundaries; leaves treated and dyed for use in a wide range of plaited ware including ceremonial and rough mats, baskets, hats, and lashings; used as a fuel; prop roots usedmedicinally; fruit and fibre used in garlands.
Parinari

glaberrima

makita (N,M,B) Occasional in garden and fallow areas, in secondary forest, and around villages; timber used for construction and small straight rods used for spears; seed used for dyeing hair and scenting coconut oil; leaves used for thatching houses.
Parinari insularum sa (N,M,B) Occasional in garden and fallow areas and

in open forest stands; excellent firewood.

Persea americana pea (N,M,B); avocado Occasionally planted in garden areas and around villages; ripe fruit eaten, commonly as a butter substitute; does not bear fruit as well at wet higher elevations.
Piper aduncum onolulu (N), qaneqona (M), yaqona ni Onolulu (B) Common in fallow areas and protected in garden areas; coppiced as a very accessible and renewable firewood source; used as a shade tree in active garden areas.
Piper methysticum yaqona (N,M,B); kava Very abundant; most important cash crop in both Namosi and Matainasau; important alkaloid-rich social beverage made from the roots and lower stems; commonly intercropped with taro on rich slope soils; used medicinally.
Plumeria spp. bua ni Vavalagi (N,M,B); frangipani, plumeria, coconut oil Common in villages, and occasional in nearby gardens; fragrant flowers used in garlands and to scent.
Polyscias spp. danidani (N,M,B) Common in villages, and occasional in garden areas; planted as live fencing, hedges, pig pens, and ornamentally in villages; leaves used in preparation of attire for traditional dances; used medicinally.
Pometia pinnata dawa (N,M,B); Pacific Iychee Common in garden areas, fallow areas, and secondary forest; usually Protected when clearing garden lands; ripefru. eaten; leaves and bark used medicinally; good firewood and good timber.
Premna serratifolia yaro (N,M,B) Common in garden and fallow areas, in Secondary forest, and around villages;commonly planted as living fencing; leaves and bark used medicinally.
Pritchardia pacifica sakiki (N,M); masei (B); Fiji fan palm In&equently planted around villages and garden lands; seeds eaten; leaves once used for fans.
Pseuderanthemum

curruthersii

false eranthemum Occasional ornamental in villages.
Psidium guajava quwawa (N,M,B); guava Common in cattle paddocks, and occasional in garden areas and around villages; leaves used as a treatment for diarrhoea, fruit a favoured, vitamin-C rich snack food; good firewood.
Saccharum eduk duraka (N,M,B); Fiji asparagus Common crop in garden and fallow areas,

especially in moist alluvial soils; plantedand protected in a somewhat naturalized state; internal inflorescence an important seasonal food and cash crop.

Saccharum officinarum dovu (N,M,B); sugar cane Common intercrop in garden areas and around villages; stems an importantsnack food; used medicinally, leaves of some varieties occasionally used for thatching.
Samanea saman vaivai, vavai ni Vavalagi (N,M,B); vu ni kau ni Vavalagi (M); rain tree, monkey-pod Commonly planted or protected in garden and grazing areas on the alluvial flats and near villages; excellent shade trees, wood favoured for wood carving and firewood.
Santalum yasi yasi (M) Uncommon in secondary forests; harvested commercially in some areas of Fiji.
Schizostachyam glaucifolium bitu, dina (N,M,B); bamboo Common in garden areas and in open secondary forest in hilly areas; largestems used for general construction purposes, in the construction of light rafts (bilibili), for piping, and for cooking food, especially freshwater prawns in the earthen oven or over an open fire; the totem (i cavati) of the Soloira mataqali, or descent group, of Namosi village; used medicinally.
Spathodea campanulata tulip (M); African tulip tree Occasional ornamental and living fence around villages, occasional as a naturalized escape in garden areas.
Spondias dulcis wi (N,M,B); Polynesian vi- apple, hog plum Commonly planted or protected in garden areas, around villages, and in village tree groves; ripe and green fruit eaten and occasionally sold at markets; leaves boiled with fatty food, particularly pork.
Sterculia vitiensis waciwaci (N) Infrequent in garden areas and in secondary forests, and occasionally planted around villages; seeds eaten roasted or fried in the past.
Syzygium malaccense kavika (N,M,B); Malay apple, mountain apple Commonly planted or protected in garden areas, around villages, and in village tree groves; leaves and bark used medicinally; fruit eaten green and ripe, leaves cooked in the past.
Syzygium seemannianum yasi ni wai (N) Common along streams and river banks; leaves used medicinally.
Terminalia catappa tavola (N,M,B); beach Indian almond Occasional in garden areas and around villages; ripe seeds eaten; young leaves and bark used medicinally; the totem (i cavuti) of the Navatusila mataqali, or descent group, in Namosi village.
Theabroma cacao koko (N,M,B); cocoa, cacao Occasionally planted in old garden areas; fruit eaten and occasionally harvested to make cocoa; increasingly important cash crop in some nearby areas in the wet zone.
Timonius affinis tirivanua (N) Occasional in fallow areas; used medicinally by women.
Trema orientalis drou (N,M,B) Common pioneer species in fallow areas, secondary vegetation, and disturbed ruderal sites, particularly along road cuts.
Veitchia joannis saqiwa (N,M); niusawa (B) Infrequent in secondary forests and around villages; planted more in the past; immature seeds eaten as a snack food; the totem (i cavuti) of the Naqelekautia mataqali of Namosi

village.

Sources: Field surveys by R.R. Thaman from 1979 to 1988; Weiner 1982 for some medicinal usages.

A note on Polynesia

Although the agroforestry systems of the smaller, more geologicallyrecent islands of Polynesia are in many ways similar to those of the larger, more heavily forested islands of Melanesia, there are significant differences. Little truly native forest remains in Polynesia, agricultural cycles have shortened, there are generally fewer tree species, and agrodeforestation is more widespread, both in terms of the elimination of trees and forest stands and the failure to protect or replant trees in active agricultural areas. The main causes of these processes are the comparative scarcity of land, rapid population growth on some islands, habitual burning, the increasing use of land for monocultural commercial agricultural production or non-agricultural purposes, and a loss of appreciation of the importance of trees among the younger generation. However, as the four Polynesian case-studies illustrate, great variation remains in Polynesia, with very rich agroforestry traditions present in areas such as Tonga and Rotuma, whereas at the other extreme - as in Aitutaki, in the Cook Islands agrodeforestation has proceeded to a dangerous stage.

Although similar agroforestry zones are found in Melanesia and Polynesia, very little primary or relatively undisturbed forest remains in Polynesia, and what natural forest there is has always been floristically less diverse than that in Melanesia. In Polynesia, the dominant vegetation communities consist of coastal strand and, in some cases, mangroves, isolated relict stands of tropical lowland or upland forest, mature fallow forest, highly degraded fire-climax communities rarely used for agriculture, and the much modified communities and fallow vegetation found in active agricultural areas, villages, and towns.

Shifting agroforestry remains a widespread practice in well-drained areas, with semi-permanent, sometimes irrigated, Colocasia taro gardens in poorly-drained alluvial lowlands and river valleys, and home gardens in settlements. In the complex systems of Tonga and Rotuma the dominant root-crop staples and a wide range of banana and plantain cultivars are grown in a shifting pattern on land so heavily planted with coconuts, breadfruit, citrus trees, and other trees that most garden areas resemble polycultural orchards. Interspersed throughout garden areas are small groves of secondary fallow forest or bushfallow containing both indigenous and exotic species of great cultural utility. As in the case of Melanesia, coastal strand forest, swamp forest, and mangrove forest are considered integral and useful components of the productive landscape.

Where staple and supplementary food crops continue to be partly derived from and integrated with a matrix of trees, the landscape also contains many important non-food trees or tree-like plants, for example, pandanus cultivars for handicrafts (Pandanus spp.), paper mulberry (Broussonetia papyrifera) for bark-cloth manufacture, and kava (Piper methysticum), the traditional ritual beverage of Polynesia. Most of these plants, which will be further described in the casestudies, are also important in, and probably introduced through or from, Melanesia.

An aspect of Polynesian agriculture that is not so well developed in the areas studied in Melanesia is the increasing, often intensive, monocultural commercial cultivation of non-tree crops, both for export and for the rapidly expanding local urban markets. These crops, most of which also have subsistence value, include water melons, pineapples, peanuts, tomatoes, Chinese cabbage, capsicum, cucumber, and other short-term vegetables, and traditional staples, such as yams, Colocasia and Xanthosoma taros, sweet potato, and cassava. In Tonga, vanilla (Vanilla fragrans) and butter squash (Curcurbita sp.), the latter destined for the Japanese market, have become monocultural components of the system. Vanilla is also produced for export in French Polynesia.

Bananas are grown on an intensive monocultural basis for export in Tonga, Western Samoa, and the Cook Islands, and there are attempts to develop monocultural citrus production for export and local processing in Nine and the Cook Islands. Both bananas and citrus have suffered from continuing problems with pest and disease control, damage from tropical cyclones and high winds, poor management, inadequate quality control, marketing difficulties, and import and quarantine restrictions. Consequently, these crops have been largely uneconomic, and their further development has been highly aid-dependent. The coconut remains the main cash crop for the majority of smallholder agroforesters in Polynesia, whereas the more diversified commercial arboricultures found on the large islands of Melanesia, based on the inclusion of cocoa, coffee, rubber, and oil palm, are largely absent in Polynesia, except in Samoa, where cocoa has been a major cash crop since the time of German occupation. Coffee has been widely tried, with limited production for local consumption in Tonga, Western Samoa, the Cook Islands, and French Polynesia.



Polynesia

Exotic species of post-European introduction that have become integral components of Polynesian agroforestry systems include kapok (Ceiba pentandra), frangipani (Plumeria rubra and P. obtusa), allspice and bay rum (Pimenta doica and P. racemosa), and jambolan (Syzygium cumin)), with leucaena (Leucaena leucocephala) and guava (Psidium gunjava) having become widely naturalized components of fallow vegetation that grow in almost impenetrable monospecific stands on many islands. There are also a number of exotic timber species introduced as part of reforestation schemes; these include Albizia falcataria, silky oak (Grevillea robusta), eucalyptus (Eucalyptus spp.), cigar box cedar (Cedrela odorata), West Indian mahogany (Swietenia macrophylla), and Caribbean pine (Pinus caribaea), which have been planted in small woodlots, as border plantings, or as scattered individual trees.

Home gardens are an important component of Polynesian agroforestry strategies, particularly in areas like Aitutaki and Rarotonga, where agrodeforestation has been extensive. Home gardens serve as refugia for endangered species, as well as avenues for the testing and introduction of new species. Home-garden agroforestry is characterized by a wide range of fruittrees, ornamental trees, and shrubs with fragrant flowers or leaves, and a number of other useful trees that serve as a matrix for lawns, structures, and scattered flower gardens.

Throughout most of Polynesia, continuing official bias toward monocultural export cultivation of citrus, banana, papaya, water melon, tomatoes, and other crops, and on monocultural root cropping, has led to widespread deforestation and agrodeforestation. Concomitantly, polycultural root and tree cropping has declined, existing tree groves have been neglected, and large areas of previously important fallow forests and associated tree species have been cleared. Agrodeforestation and the cultural abandonment of traditionally important trees on Rarotonga and Aitutaki have perhaps been worse than in most areas of the Pacific, although similar trends are evident elsewhere - as in Tonga, where 90 farmers, in a survey of 100 farmers, reported declining numbers of trees on their own fields (Kunzel 1989, 27). If not reversed, such trends could lead to the abandonment of agroforestry traditions that have sustained the highly sophisticated hierarchical societies of Polynesia for many centuries.

Tongatapu island, Tonga

Tongatapu, the largest and most populous island of the Tongan group, is an upraised limestone island, approximately 257 sq km in area, lying just within the tropics. Its mild and moist climate, with a mean annual temperature of 23°C, a seasonal range of only 4°C, and a mean annual rainfall of 189 cm, provides a favourable agricultural environment. The soils are fertile, deep friable clays of prehistoric volcanic origin that overlay the upraised coral-limestone base of the island. Little soil erosion occurs on the mostly flat surface, the highest elevation being only 83 metres above sealevel. Major constraints to agriculture include the absence of surface streams or surface water, undependability of rainfall, and frequent tropical cyclones and galeforce winds (Thaman 1976).

Owing to population pressure (over 251 per sq km in 1986) and over 3,000 years of human occupance (Groube 1971), very little primary vegetation remains. Vegetation communities currently that exist include: coastal littoral forest, coastal savanna, swamp forest, marshland, mangrove swamps, and isolated stands of tropical lowland forest, plus the highly modified vegetation communities and fallow and ruderal vegetation found in active agricultural areas and villages and towns.

Agricultural land use

A bush-fallow agroforestry system is practiced on some 5,652 bush allotments ('apt 'uta) averaging about 3.2 ha (8.25 acres by law) in area, to which individual farmers or farm families have individual tenure rights. The characteristic vegetation on Tongatapu bush allotments can be best described as a dynamic mosaic of individualized allotments, each containing a diversity of subsistence and cash ground crops and secondary vegetation in various stages of regrowth, spread throughout a matrix of coconut palms and a wide variety of useful tree species. Also found on many allotments, and considered integral and useful components, are small protected areas of old fallow forest, coastal strand forest, or mangrove forest. Most families also have an additional town allotment ('apt kolo), ranging from 0.1 to 0.16 ha (0.250.4 acre), located in town reserves, where they maintain their permanent residences, a home garden, and associated structures. Some families, however, have permanent residences on their larger bush allotments, with many families maintaining temporary thatched shelters or small houses there for use when needed (Thaman 1975, 1976).

Cropping patterns

The cropping cycle begins with the clearing of secondary vegetation, including many pioneer tree species, while protecting (although sometimes pruning or pollarding) useful species, the most common being coconut, breadfruit, mango, citrus, Tahitian chestnut (Inocarpus fagifer), oceanic Iychee (Pometia pinnatae), candlenut (Aleurites moluccana), and koka (Bischofia javanica). The dried plant remains are then usually burned, and the first crop, often yams (Dioscorea alata), is planted, and usually intercropped with plantains (Musa AAB Group), giant taro (Alocasia macrorrhiza), or other crops. After the yams are harvested, true taro (Colocasia esculenta) and tannia (Xanthosoma sagittifolium) or cassava are generally the next crops in the succession. Other crops, including sweet potatoes, bananas, peanuts, paper mulberry (Broussonetia papyrifera), or Lava (Piper methysticum), are also integrated into the cropping sequence and may be planted at any time during the sequence. If taro has been planted, cassava often follows and is sometimes planted up to four or five times consecutively on a given plot. Then, after the cassava is harvested, the land will be allowed to revert to fallow, although some cassava is commonly left as a component in the tumble-down fallow vegetation (Thaman 1975, 1976).

In a sample survey of 101 bush allotments comprising 326 ha, conducted in 1971-1972, approximately 111 ha, or 34 per cent of the area, was under cultivation, with the balance under various stages of fallow regrowth (Thaman 1976). There were 87 allotments with some cultivation and 14 that were not cropped at all. With respect to cultivated land, it was often difficult to differentiate between subsistence and commercial cropping, because crops such as coconuts, bananas, water melons, pineapples, garden vegetables, and peanuts, which are commonly grown for export or local sale, are also used for home consumption. Conversely, many important subsistence crops such as cassava, yams, sweet potato, tannia, and plantains are also produced commercially. On this basis, 97 of the 111 ha under cultivation, or 86.5 per cent, constituted subsistence cropping; 12 ha, or 10.7 per cent, commercial cropping; and 3 ha, or 2.7 per cent, a combination of the two. Of the 87 allotments under some form of cultivation, all had subsistence crops, whereas only 31 had "commercial" crops (Thaman 1975, 1976).

Of the eight species of staple subsistence ground crops, cassava covered 37 per cent of the area, tannia (Xanthosoma taro) 24 per cent, yams 21 per cent, sweet potatoes and taro 3 per cent each, a combination of tannia and Colocasia taro and plantains 2 per cent each, and sweet yam (Dioscorea esculenta) and giant taro (Alocasia macrorrhiza) 1 per cent each. These figures represent areas for these crops as dominants on a given plot and, in the cases of plantains and giant taro (which are commonly intercropped with yams and infrequently planted alone), their relative importance was greater than these figures indicate. Similarly, the relative importance of sweet potato is probably understated because of the time-period selected for the inventory.

A range of supplementary food crops also plays an important role in the Tongan diet. Because most supplementary food plants are either intercropped with staple crops or protected in the fallow vegetation, their frequency of occurrence was surveyed rather than the area they occupied. Of these plants, the papaya was the most widespread, occurring in either a cultivated or semiwild state on 88 per cent of the sample allotments. In only three cases was it planted in rows resembling a plantation. Supplementary food plants present on over 50 per cent of the sample allotments included Cordyline fruticosa, chill) peppers (Capsicum frutescens), and bananas (Musa AAA Group) (planted as isolated or small groups of plants). Plantains (Musa AAB Group), hibiscus spinach (Hibiscus manihot), pineapple, sugar cane, and pata variety bananas (Musa ABB Group) were found present on at least 25 per cent; wild or cherry tomatoes (Lycopersicon esculentum var. cerasiforme), pumpkin, kava (Piper methysticum), and corn on at least 10 per cent; with the remaining supplementary food plants found on the bush allotments (which bring the total to 18) including green or bunching onions (Allium spp.), coffee, lemon grass (Cymbopogon citratus), Amorphophallus campanulatus, and granadilla (Passiflora quadrangularis).

Throughout the matrix of staple and supplementary food crops and secondary vegetation are a number of important non-food plants. The most common non-food plants, which are both cultivated and protected in a natural state, are a number of species or cultivars of pandanus (Pandanus spp). These provide fibre for baskets, mats, and other handicraft items and were present on 76 per cent of all allotments. On four of these allotments, solid plantings averaging about 0.1 ha were found. Second in frequency to pandanus, but occupying considerably more area, was paper mulberry (Broussonetia papyrifera) (used for making tape cloth), which was planted on 38 per cent of the sample allotments and comprised about 3 per cent of the subsistence area. Other plants useful for construction, wrapping food, decorative purposes, handicrafts, scenting oil, adhesives, and medicinal purposes included bamboo (Schizostachyum glaucifolium and Bambusa vulgaris), Heliconia indica, cycad (Cycas circinalis), Job's tears (Coix lachryma-jobi), members of the ginger family (Zingiber zerumbet and Amomum cevanga), sword grass (Miscanthus foridulus), Polynesian arrowroot (Tacca leontopetaloides), maile (Alyxia stellata), cotton (Gossypium barbadense), tobacco, tumeric (Curcuma longa), and 'uhf (Euodia hortensis).

Of the non-tree crops classified as commercial - although many of them are also of subsistence value - bananas (Musa AAA Group) covered 42 per cent of the area, water melons 17 per cent, plantains (Musa AAB Group) 13 per cent, pineapples 11 per cent, tomatoes and sweet capsicum, or bell pepper (Capsicum annuum var. grossum), both 6 per cent, and peanuts 4 per cent. The balance of the area was made up by rock melons (Cucumis melo var.), corn, and vanilla (Vanilla fragrans), all in contiguous plots. Subsequent to the survey, there has been an expansion of monocropping of butter squash (Cucurbita sp.) for export to Japan. All of these crops, with the possible exception of vanilla, also have subsistence value that, in some cases, may be greater than their commercial value. Bananas in particular, which are cooked in a green state, were, in 1971, one of the major staple foods, together with cassava, American taro, and yams (Thaman 1976).

Trees and agroforestry practices

Although the major staple and commercial crops dominate the cultivated area of bush allotments, the most prominent plants are trees. The coconut palm, which was present on over 98 of the 101 surveyed bush allotments, was by far the most common and most important cash crop on Tongatapu, as well as being the most frequently consumed subsistence crop. Over half of the allotments had over 200 mature palms, whereas some had 500 or more, the median coverage being about 250 palms. Although 56 allotments also had immature replanted coconuts, in only a few instances were they wellmaintained (Thaman 1976, 257-258).

In addition to coconut palms, there were at least 25 other fruit trees among the plant associations. Trees, primarily important because of their edible fruit, present on over half of all sample allotments, included mango (Mangifera indica), citrus (including Citrus sinensis, C. reticulate, C. hystrix, C. aurantium, and C. aurantiifolia), oceanic Iychee (Pometia pinnata), breadfruit (Artocarpus altilis), Tahitian chestnut (Inocarpus fagifer), and guava (Psidium gunjava). Other common species included avocado (Persea americana), beach almond (Terminalia catappa), soursop (Annona muricata), Polynesian vi-apple (Spondias dulcis), and Malay apple (Syzygium malaccense). Other species such as sugar apple, or sweetsop (Annona squamosa), rose-apple (Syzygium jumbos), cocoa (Theobroma cacao), cashew nut (Anacardium occidentals), the common fig (Ficus carica), apple (Malus pumila), and peach (Prunus persica) were also found as isolated individuals on allotments on some occasions.

Of even greater importance in terms of numbers and frequency of occurrence are at least 67 other non-food tree species, all useful in some way to Tongan society. They are generally found in a protected state, although sometimes deliberately planted. Apart from the value of their timber, their fruit, seeds, leaves, bark, wood, and roots are used in the preparation of medicines, dyes, handicrafts, cordage, scent for coconut oil, soap substitutes, handicrafts (including wood carving), canoe construction, and other material goods. The most common trees in this category include Bischofia javanica, Aleurites moluccana, Hibiscus tiliaceus, Morinda citrifolia, Macaranga harveyana, Rhus taitensis, Grewia crenata, Erythrina variegate, Glochidion spp., Pandanus tectorius, Ceiba pentandra, Santalum yasi; Cananga odorata, Pittosporum arborescens, Pisonia grandis, Elatto stachysfalcata, Garciniasessilis, Adenantherapavonina, Mertyamacrophylla, Ficus scabra, Micromelum minutum, and bamboos (Schizostachyam glaucifolium and Bambusa vulgaris), all of which were found present on over 10 per cent of all allotments.

Other species of widespread agroforestry importance throughout the Pacific, found on over S per cent of the sample allotments and which, in some cases, constitute endangered species, include Cerbera manghas, Diospyros laterifolia, Jatropha curcas, Ervatamia orientalis, Grevillea robusta, Syzygium corynocarpum, Xylosma orbiculatum, Plumeria rubra, Syzygium clusiaefolium, Neisosperma oppositifolium, Trema orientalis, and Dysoxylum forsteri. Other species of particular cultural significance or utility, found on less than 5 per cent of allotments, some of which must be considered endangered species, include Garuga floribunda, Parinari glaberrima, Fagraea berteriana, Diospyros ferra, Ficus obliqua, Calophyllum inophyllum, Myristica hypargyraea, Pimenta racemosa, Burkella richii, Casuarina equisetifolia, Tarenna sambucina, Homolanthus nutans, Aglaia saltatorum, Syzygium neurocalyx, and Premna serratifolia.

Most of the tree species found on bush allotments are native to Tonga or are aboriginal introductions, although recent introductions such as Leucaena leucocephala, kapok (Ceiba pentandra), Jatropha curcas, and frangipani (Plumeria rubra) were also common. There are also a number of species introduced by the government as part of reforestation schemes; these include the silky oak (Grevillea robusta), blue gum (Eucalyptus spp.), and cigar box cedar (Cedrela odorata), which have been planted, either as small woodlots, border plantings, or scattered individual trees, on some allotments.

Fallow and natural vegetation

Another dominant vegetational feature on bush allotments was the fallow or secondary vegetation, which has a very important role in the cropping succession. Approximately two-thirds of the combined area of all sample bush allotments were under some stage of bushfallow or natural vegetation, with some allotments (14 per cent of the sample) completely under various stages of fallow vegetation. In no case was an individual allotment under 100 per cent cultivation. There were, however, very few areas under fallow for sufficient time for forest to regenerate. Only 9 per cent of the uncultivated area had been under fallow over 10 years, the majority of which comprised uncultivable stands of coastal strand or swamp forest. Another 21 per cent had been under fallow for approximately 5-10 years, with the remaining 70 per cent having been under fallow for less than 5 years and being characterized by bushes, grasses, and ephemeral pioneer species, most of postEuropean-contact introduction, and few trees. Extensive areas of fallow vegetation, especially in the eastern, or Hakake, district, were under almost monospecific stands of Leucaena leucocephala, which spread so fast and is so difficult to clear that people considered it a weed (Kunzel 1989, 41), while recognizing the tree's value as an enhancer of soil fertility and a major source of saleable firewood.

Similarly, many of the other fallow species (which number more than 60) are useful as sources of food, medicines, firewood, fodder, fish poisons, and other valuable products; but some, such as Cyperus rotundas, Commelina spp., and Psidium gunjava, spread so rapidly as noxious weeds that they have become the focus of control programmes.

As suggested above, in addition to the cultivated and protected plant species and fallow vegetation, a number of allotments also had uncultivable areas under "natural" vegetation. These allotments were usually found along the coast and included areas of littoral or coastal strand forest (on the higher windward southern portions of Tongatapu) and swamp or mangrove forest (on the low-lying leeward northern coast). Although of limited agricultural potential, such areas constitute sources of many useful products, and embody gene pools of species rapidly disappearing from a majority of Tongatapu bush allotments, as well as serving as buffer zones protecting crops and soils from sea spray and salt-water incursion.

Rotuma island, Fiji

Rotuma is an isolated volcanic island to the north-north-west of the Fiji group, with fertile, well-drained basaltic soils, an area of some 45 sq km, and a population of around 3,000. The majority of its indigenous Polynesian people (approximately 5,000) have emigrated to the main Fiji group, thus relieving pressure on Rotuman land.

Ecologically, Rotuma illustrates the human ability to design with Nature to create productive and sustainable agroforestry. The dominant staples - yams, Colocasia, Xanthosoma, and Alocasia taros, cassava, and a wide range of banana cultivars - are grown in a shifting system, with old gardens having been so commonly planted with coconuts, breadfruit, oranges, and other trees that most of the island

Table 6 Sampling frequency of mature tree species from 40 sites along a 2-km transect on Rotuma island, F'iji (bananas or plantains [Musa spp.] and papaya [Carica papaya], which were very abundant, have been excluded) appears to be part of a giant plantation - but decidedly a polycultural plantation, with many other useful trees interspersed with the coconuts, and small groves of secondary forest containing both indigenous and exotic species of great cultural utility. In order to convey an image of the orchard-like character of Rotuma's agricultural landscape, data collected on a 2-km transect are summarized in table 6. The transect was run from the coast inland along a line close to the trans island road. The survey technique used enabled 120 trees to be identified at 40 sampling points. Aside from the very common Musa cultivars and papaya, which were excluded from the sample, only 19 species of trees were identified. This uniformity, when contrasted with the greater diversity recorded in more comprehensive botanical studies of the entire island and of secondary forest groves, demonstrates how what would have been a considerably greater natural diversity has been transformed into a quasi-orchard, a sort of tamed forest of mostly useful trees. Only about a quarter of the trees recorded had grown spontaneously; the rest had been planted.

Botanical name (Rofuman; common names) Frequency Comments
Cocos nucifera (niu; coconut) 58 The coconut trees provide Rotuma's major export (copra) as well as a variety of materials and several nutritious additions to the Rotuman diet; planted and volunteer.
Artocarpus altilis ('ulu, ulu; breadfruit) 43 Major seasonal staple prepared and eaten in a variety of ways; made into traditional pudding or dessert, fekei; trunk favoured for canoes; medicinal;

planted.

Morinda citrifolia (ura; beach mulberry) 38 Small fruits pounded and mixed with water for children's cough syrup; leaves soaked in oil to put on boils; spontaneous volunteer; fruit eaten by older persons.
Citrus sinesis (mori; orange) 15 Plantings of the famous Rotuman orange - highly esteemed in Suva - have increased in recent years as demand for their export has grown; also used in renowned Rotuman "orange wine"; planted occasionally, usually spontaneous.
Inocarpus fagifer (ifi, Tahitian chestnut) 13 Frequently used as a boundary marker; edible seed; planted and spontaneous.
Macaranga spp. (sa'a) 10 Sometimes planted by seedling near houses; light wood used for canoes and house flooring; good firewood; medicinal use; spontaneous in fallow lands.
Hibiscus tiliaceus(hau; beach hibiscus) 8 The pan-Pacific beach hibiscus; traditionally an important fibre plant; important medicinally; spontaneous, occasionally planted.
Flacourfia rukum (firmoto; governor's plum) 5 A very abundant spontaneous naturalized tree in the forest, along roads, and in coconut plantations; wood used forhouse and fence posts and fuel; fruit eaten.
Mangifera indica(mago; mango) 5 Common tree; nutritious fruit; good firewood.
Pometia pinnata (fao, fava; oceanic Iychee) 5 Fruit eaten; excellent firewood and good timber; medicinal value.
Ceiba pentandra (sipi; kapok) 5 Silk cotton used for stuffing pillows and mattresses; occasionally used as a living fence; planted.
Spondias dulcis (vi; Polynesian plum) 5 Planted around houses for its delicious fruit; said to be spontaneous in the interior and in plantations.
Citrus hystrix (rough lemon, kaffir lime) 5 Fruit important for making drinks, spicing food, and marinading sea food; leaves used medicinally and in herbal teas; planted and spontaneous.
Pandanus sp. (sa'aga) 3 Planted for leaves, which are used to make fine white mats.
Psidium gunjava (guava) 3 Fruit a nutritious seasonal snack food; leaves medicinal; excellent firewood; spontaneous.
Theabroma cacoa (cocoa) 3 Cocoa, tried as an export crop, is not presently harvested; pulp surrounding seed eaten as a snack food; planted.
Pipturus argenteus (armea) 3 Wood, which is strong but light, is used for canoe paddles.
Swietenia macrophylla (ai hapa; mahogany) 3 A recent introduction; timber tree.
Erythrina orientalis (ratu'a; coral tree) 3 Planted as a boundary marker and living fence; used medicinally.
Metroxylon vitiense (ota; sago) 3 Planted for the thatch; starch from trunk used occasionally in puddings.

Other important tree species found in reconnaissance surveys of active garden areas, various stages of fallow vegetation, and protected secondary forest stands included Dracontomelon vitiense, Terminalia catappa, Elaeocarpus grandis, Syzygium spp., Sterculia fanaiho, Gardenia fijiensis, Diospyros spp., Dysoxylum spp., Intsia bijuga, and Neonauclea forsteri.

Closely connected with Rotuman agroforestry is the pig-rearing system, wherein the animals are kept in large communal stone-walled enclosures located separately from garden lands in protected secondary forest stands, often between villages and inland gardens. Trees provide much of the pig food, notably coconuts and the abundant papaya. Also of great abundance are a wide range of banana and plantain (Musa) cultivars, including the mountain banana (Musa troglodytarum), which are eaten both green as a starch, and also ripe as a fruit and in a variety of puddings. It has been suggested that banana cultivars may, along with Colocasia taro and cassava, be among the most universally important staple food crops in the Pacific (Thaman 1987). The diversity of banana cultivars (about eight) and their many uses on Rotuma bear this out.

Although it is difficult for an outsider - particularly if a whistle-stop central planner, let alone a trained botanical researcher - to inventory or fully to comprehend the Rotuman agroforestry system, it clearly contains much diversity in species and cultivars and contributes much that is of cultural, economic, and ecological utility. Although much of Rotuma's landscape is forested, it is a much modified landscape, laced with planted or preserved trees and designed to meet human needs. After their fieldwork on the island, Thaman and Clarke (1987) commented that walking through Rotuman forests is- appropriate changes being made - not unlike strolling in the gardens of Versailles, except that the Rotuman land use is so much more productive of foods and useful materials.

Rarotonga and Aitutaki, the Cook Islands

Rarotonga and Aitutaki, two small islands in the southern Cook Islands, have agroforestry systems that have been subjected to relatively high population densities as well as to strong pressures from urban growth and the commercialization of agricultural production.

Rarotonga, with a total land area of 67 sq km, is a high basaltic island with a rugged forested interior rising to 652 metres, areas of highly degraded bush and fern-clad hills, deep well-watered valleys, and a fertile low-lying, somewhat swampy coastal plain fringed by a narrow, slightly-raised sandy coastal strip. With Mangaia, the second-highest island in the Cook Islands, reaching only 168 metres, the native upland vegetation of Rarotonga is not found elsewhere in the Cook Islands (Merlin 1985,84).

Agricultural activities in Rarotonga have traditionally focused on taro cultivation in the rich soils of the swampy lowlands and the alluvial and colluvial soils of stream valleys surrounding the mountainous interior, with diversified agroforestry and other root cropping being practiced throughout these areas, as well as on extensive areas of colluvial soils on the lower slopes and interfluves of the interior and in home gardens in villages located on the lower slopes of the interior and on the well-drained coastal strip. The current population of Rarotonga is estimated to be over 10,000, which constitutes over half of the 15-island country's population.

Aitutaki, a smaller island lying some 225 km north of Rarotonga, has a total area of 19.9 sq km and an estimated 1981 population of 2,335 (Douglas and Douglas 1989). The land area consists of a somewhat lowlying volcanic island, rising to a maximum elevation of 137 metres at Maunga Pu (Sykes 1976), with limited areas of alluvial valleys and coastal plain surrounding upland agricultural areas and highly degraded scrub lands and fern lands at higher elevations. There are also a number of low-lying coralline and volcanic reef islets, or motu, surrounding an extensive lagoon, "generally considered to be one of the most beautiful in the Pacific" (Douglas and Douglas 1989, 53).

Agroforestry activities on both islands have been seriously affected by emigration, heavy emphasis on export cash cropping, and increasing monetization, as well as other urban-biased developments, particularly tourism, and increasing educational, economic, social, and nutritional dependence on New Zealand.

Because of the unusual political status of the Cook Islands, first as a New Zealand colony since the turn of the century and then, since 1964, as a self-governing state in free association with New Zealand (with Cook Islanders remaining New Zealand citizens with free access to New Zealand), there has been steady emigration of Cook Islanders to New Zealand since World War II, with well over half of all Cook Islanders now residing there (Douglas and Douglas 1989). Remittances from overseas Cook Islanders in New Zealand, as well as from those in wage employment on Rarotonga, to outer islands such as Aitutaki constitute the major source of income for many families.

The main agroforestry zones on Rarotonga and Aitutaki are:

  1. home and village gardens;
  2. coastal littoral strand areas encircling the islands;
  3. poorly-drained alluvial agricultural lowlands, or pa'i;
  4. upland and interior garden lands on rolling or undulating hill-lands and up-valley bottoms up to 200 metres in elevation; and
  5. steep forested upland areas.

Aitutaki, being a smaller, older, lower island than Rarotonga, has very limited areas of poorly drained lowland and virtually no indigenous upland forest.

Home and village gardens

Because of the emphasis on commercial cropping and the highly derived nature of the vegetation at lower elevations, especially on Aitutaki, it is in home and village gardens that one finds agroforestry at its most diverse. On both Rarotonga and Aitutaki, home agroforestry is characterized by a wide range of fruit-trees, ornamental trees and shrubs with fragrant flowers or leaves, and a number of other useful trees, which serve as a matrix for lawns, structures, and scattered flower gardens.

Most common among the fruit or edible-seed trees are coconut palms, breadfruit, mango, citrus trees, papaya, avocado, guava (Psidium guajava and P. cattleianum), Tahitian chestnut (Inocarpus fagifer), the red-bead tree (Adenanthera pavonina), sweetsop and soursop (Annona squamosa and A. muricata), Polynesian vi-apple (Spondias dulcis), Malay apple (Syzygium malaccense) and beach almond (Terminalia catappa). Among the citrus trees, species such as limes and lemons, which are commonly required for marinading raw fish (ika mata) and which are not normally planted in the larger commercial citrus plantations, are usually found near homes. Mandarin orange, grapefruit, and the orange are also common in home gardens.

Other fruit-trees occasionally found in home gardens include jakfruit (Artocarpus heterophyllus), carambola (Averrhoa carambola), the star apple (Chrysophyllum cainito), koko (Inga edulis), tamarind (Tamarindus indica), the Pacific fan palm (Prichardia pacifica), and a range of fruits, all referred to locally as vinevine, including Surinam cherry (Eugenia uniflora), white mulberry (Mows alba), governor's plum (Flacourtia ramontchi), acerola (Malpigia glabra), and the sea grape (Coccoloba uvifera).

Common non-fruit trees found in home gardens include Norfolk Island pine (Araucaria heterophylla), banhinia (Bauhinia spp.), kapok (Ceiba pentandra), flame tree, or poincianna (Delonix regia), coral, or dadap, tree (Erythrina variegata), beach hibiscus (Hibiscus tiliaceus), pandanus (Pandanus spp.), bay rum tree (Pimenta racemosa), frangipani, or plumeria (Plumeria rubra and P. obtusa), African tulip tree (Spathodea campanulata), and yellow bells (Tecoma stans).

Among tree-like shrubby species, the Tahitian gardenia (Gardenia taitensis) is particularly common and very important, along with the frangipani, in the production of the ubiquitous flower garlands, or lets, that are still of considerable cultural and economic importance, both for everyday use and for the expanding tourist industry. Other common shrubby species, most of which are referred to as kapaie (literally hedge), commonly planted as living hedges or fences or as ornamentals, include a number of panax species (Polyscias guilfoylei, P. scutellaria, P. balfouriana, and P. tricochleata), the copper leaf, or beefsteak, plant (Acalypha amentacea), croton (Codiaeum variegatum), cordyline (Cordyline fruticosa), the caricature plant (Graptophyllum pictum), and the common hibiscus (Hibiscus rosasinensis).

Coastal and strand agroforestry

Because the sandy low-lying coastal zone and smaller offshore islands are commonly owned and used by the same families who actively cultivate inland gardens, these areas can be seen as integral parts or extensions of the agroforestry system. Moreover, such areas border or extend into agricultural areas as well as being integral components of coastal home gardens.

The dominant trees or tree-like shrubs in the outermost zones, which are most exposed to salt spray, include Messerschmidia argentea, Leucaena insularum, Sophora tomentosa, and the often dominant saltbush, or halfflower, Scaevola sericea. Two very hardy plants known locally as ngangie, Pemphis acidula and Suriana maritima, which seem to be absent on Rarotonga- possibly because of human clearance, although they may never have been present (Sykes, 1976) are common on the reef islets, or motu, of Aitutaki.

In more sheltered areas on Rarotonga, beach hibiscus (Hibiscus tiliaceus), Indian mulberry (Morinda citrifolia), and Pipturus argenteus are common, with Hibiscus tiliaceus forming dense stands. Emergent coconut palms commonly dominate this somewhat shrubby vegetation formation, which merges on its inner margin with semi-open, semi-natural or planted coconut groves (Sykes 1976). Similarly, on Aitutaki, Hibiscus tiliaceus, often along with Pandanus tectorius, dominates large areas along the coast, especially between Tautu and Vaipae, where lateritic soils overlying the volcanic bedrock reach the sea (Sykes 1976).

Where the indigenous vegetation has not been cleared for coconut monoculture, or in coastal areas where indigenous strand species have not been felled or replaced by home gardens, dominant large trees include Barringtonia asiatica, Calophyllum inophyllum, Hernandia nymphaeifolia, and Pisonia grand is, with Casuarina equisetifolia being abundant locally in coastal areas on both Rarotonga and Aitutaki. Guettarda speciosa and Pandanus tectorius are also quite common locally in coastal areas of Aitutaki, especially near the Ootu airstrip, and on some of the reef islets. Thespesia populnea is reportedly found on some reef islets of Aitutaki (Sykes 1976), and, along with Cordia subcordata (both endangered coastal strand species highly valued for wood carving in the Cook Islands), it is currently being planted along roads on the coastal strip of Aitutaki as part of current nationwide conservation activities.

On Aitutaki, coconuts are similarly ubiquitous emergents along sheltered lagoons, as well as the dominant tree on the more productive coastal fringe areas, with large numbers of regularly-spaced palms having been replanted after World War II deforestation to make way for an airstrip. Most palms, however, except some along the north coast of Aitutaki, which have been planted as part of a coconut rehabilitation scheme, are irregularly spaced, often in dense stands (Johnston 1967).

Agroforestry in poorly drained lowlands

On Rarotonga there are extensive poorly-drained agricultural lowlands, or pa'i, which constitute the most important taro-growing areas on the island. They are dominated almost entirely by excavated and mulched swamp taro (Colocasia esculenta) cultivation and fallowed taro beds in various stages of abandonment. Scattered throughout these areas are a wide range of planted or protected trees, usually along the bunds and pathways between plots or on scattered areas of high ground. Although only limited studies of these areas were conducted, the main trees in these pa'i areas include coconut palms, Tahitian chestnut, mango, Polynesian vi-apple, guava, Hibiscus tiliaceus, the jambolan (Syzygium cumin)), and banana and plantain cultivars. Although not the dominant vegetation, these scattered trees add important diversity and welcome shade and snack foods for gardeners in these highly productive agricultural areas, which have provided Cook Islanders with their traditionally important staple, taro, for over a thousand years.

On Aitutaki, areas under swamp taro cultivation were very limited in extent and found only in small patches inland and along the east coast. Where such swamp taro gardens exist, similar tree species to those on Rarotonga were often associated with them, particularly Hibiscus tiliaceus, which bordered some gardens.

Agroforestry in the uplands and inland gardens

Well-drained upland and interior garden areas extend inland from the coastal strip and the poorly drained taro swamps up to heights of 50-200 metres on Rarotonga and up to some 100 metres on Aitutaki. In these areas, traditional diversified agroforestry once predominated, but mostly has been gradually replaced by monocultural root cropping and institutional plantation culture of citrus, pawpaw, and tomatoes on Rarotonga and citrus, bananas, and tomatoes on Aitutaki. Johnston's (1967) study of agriculture on Aitutaki and Wilder's (1931) Flora of Rarotonga provide some indication of what the system was like prior to World War 11.

Whereas on Rarotonga there are still scattered groves of trees in some areas, particularly in well-watered valleys and beyond the limits of current cultivation, these are now virtually absent on Aitutaki, where there was an absence of primary forest by the early 1960s as a result of the "rotational bush fallow" agricultural system in which areas of active garden production for a given family were rotated "through a fixed area of fallow grass or woody plants where woodland is not allowed to regenerate" (Johnston 1967,28).

In both areas, permanent cultivation or protection of useful trees such as coconuts, citrus, breadfruit, mango, Tahitian chestnut, avocado, papaya, oceanic Iychee (Pometia pinnata), matakoviriviri (Adenanthera pavonina), kapok, and candlenut (Aleurites moluccana), and a wide range of banana and plantain cultivars - was also an integral part of the system.

Even the export citrus industry, which flourished from the late nineteenth century until the 1930s, was based on the "haphazard" (in Johnston's view) production of dispersed "wild" orange trees, which had been incorporated into the subsistence economy and were found scattered among other tree crops (Johnston 1967, 50-51). As late as the 1950s, these "wild oranges," which included fruit from trees scattered in the forest beyond the limits of cultivation, still constituted the overwhelming proportion of citrus production and exports (Johnston 1951, 130).

Evidence indicates that a fallow period of 10-20 years, followed by a cropping period of 2-4 years, was the traditional fallow-crop cycle. By 1960, however, fallow periods had dropped to only 5 years, and to as low as 1-2 years in more accessible areas closer to villages, with the cropping period remaining the same (Johnston 1967, 34-35), a process that explains the absence of primary forest and a scarcity of secondary forest in the gardening areas.

Although the traditional criterion for selecting garden sites was sufficient regrowth of mature secondary forest, land that had lain fallow for 10-20 years was extremely limited (Johnston 1967, 29). What forest did exist usually consisted, as it does today, of limited areas of relatively mature, almost homogeneous, stands of au (Hibiscus tiliaceus) and, in some cases, guava. Priority in use of such areas was given to bananas, which bear for 5-10 years on one plot, although existing banana areas and individual family needs were taken into account. Yams were almost always planted on newly cleared land, where uncleared trees and stumps provided trellising or support for the vines. Other non-tree and root crops were usually planted on land cleared from younger secondary forest, areas of bush-fallow, or in succession after the harvest of the previous crop, although tomatoes were also commonly planted on recently cleared fallow land, but generally in less mature areas that were relatively easy to clear (Johnston 1967, 15, 21-30).

Plot histories indicated that bananas were always planted for a minimum of 5 years, with some varieties for at least 10 successive years. Cassava was generally planted for 2 years in succession, although sometimes for 4 years, whereas yams, sweet potatoes, tan nia (Xanthosoma taro), and tomatoes were generally planted only once in succession on a given plot, although tannia was occasionally planted for 2 years (Johnston 1967, 33).

Of the major food crops, coconuts excepted, cassava was most common, covering 39 per cent of the area under crops. Reportedly introduced by the missionaries in the last century to supplement mission funds, cassava spread rapidly, possibly because of the prestige growers received by contributing to the church and the ability of cassava to grow in the conditions of relative water scarcity and declining soil fertility on Aitutaki, conditions that restrict the growth of wetland tarot Bananas or plantains cover 29 per cent, and tannia (Xanthosoma sagittifolium) 12 per cent. Other crops such as sweet potato, taro, yams, and minor crops such as melons, pumpkins, and onions, covered 7.8 per cent, and the two major commercial crops at the time, oranges and tomatoes, constituted 7.6 per cent and 4.6 per cent of the area respectively, making up the balance of the area (Johnston 1967, 14-21).

Although second in importance in terms of area, banana or plantain plots were the most common, with over two-thirds of all gardens containing them as one of several crops or as the sole crop, with some plots bearing up to 15 years. Moreover, such plots were the most localized and often at a greater distance from the village than root crops (Johnston 1967, 13-14).

Cassava and sweet potato were almost always planted as monocultures, and bananas and tannia usually as monocultures, although 18 per cent of tannia plots had been intercropped with bananas for 23 years. Intercropping was characteristic only of yam plots, which were often interplanted with bananas or tannia and occasionally with cassava.

Tree crops were scattered throughout garden areas, as were mixed groves of valuable trees. Coconut palms, the dominant crop on the coastal fringe, were found as isolated individuals or irregular clusters. Tahitian chestnut, mango, breadfruit, papayas, rough lemon, Malay apple, Polynesian vi-apple, and avocados, along with the widespread wild "native" orange and a range of other useful trees, including kapok, the banyan (Ficus prolixa), mata'oi (Cananga odorata), and candlenut, were the main trees, although breadfruit was more commonly found around villages, particularly on Aitutaki (Johnston 1967, 16).

On Rarotonga, where extensive areas of low-lying swamp taro beds and systems of irrigated taro terraces existed historically to take pressure off interior and upland garden areas, trees were more numerous than on Aitutaki, with a much greater range of trees being found either planted or naturalized in interior and upland agricultural areas.

Hibiscus tiliaccus was, and still is, the commonest and often dominant tree, with Tahitian chestnut (Inocarpus fagifer) and the candlenut (Aleurites moluccana) being particularly common in valley entrances and on lower slopes on Rarotonga. Formerly cultivated trees that have become extensively naturalized on Rarotonga include Cecropia palmata, Inga edulis, mango, and jambolan (Syzygium cumin)). Commonly found around garden areas, especially on the lower slopes of the mountainous interior, are Bischofia javanica, Homalium acuminatum, and Xylosma gracile, which Sykes (1976) believes to be "almost certainly . . . principal components of the original forest." Homalium acuminatum (moto) - the durable wood of which was formerly important in canoe construction and which was formerly "common from sea-level to 1000ft" - is one of the few native trees still occasionally found around plantations where it "was most likely the 'most common dominant' before human disturbance" (Cheeseman 1903; Sykes 1976, quoted in Merlin 1985, 90).

Other species found in the more disturbed areas of the lower montane slope forest include Elaeocarpus rarotongensis, Weinmannia rarotongensis, Macaranga harveyana, Planchonella grayana, and Terminalia glabrata, with Fitchia speciosa found in the more recently disturbed sites (Merlin 1985, 89). In the case of Bischofia javanica, some authorities such as Whistler (1980a) and Thaman (1988g) believe that B. javanica could have been a deliberately cultivated aboriginal introduction that, because of its widespread cultural importance as a source of dye for tape cloth, its use as a medicinal plant, its value as a sign of soil fertility, and its ability to withstand human abuse, remains as a relict or a possible natural dispersal from ancient garden areas. It was particularly common in most valleys and hillsides at lower elevations in the 1920s (Wilder 1931, 70), is common in low-elevation disturbed forest (Merlin 1985), and is still quite common on lower mountain slopes and surrounding old irrigated taro terraces, such as in the Takuvaine Valley.

The pattern of shifting cultivation amongst scattered fruit-trees and other useful trees and permanent tree groves has, however, been significantly altered because of increasing emphasis on commercial and subsistence monocropping, which began as early as last century and which has accelerated since World War II with the systematir institutional promotion of comercial export cropping coupled with the increasing monetization of the economy. The earliest changes were probably the introduction of cassava by the missionaries to raise funds for the church and tannia (Xanthosoma sagittifolium) - both of which grow well on drier upland soils. Both are commonly grown as monocrops and are rarely intercropped because of their ability to grow under conditions of declining soil fertility. Increased cassava cultivation, in particular - cassava is commonly planted up to four times in succession on the same plot of land - is a major factor in deforestation and agrodeforestation and the movement towards shorter fallow periods. On Aitutaki, cassava, in addition to being a prestige crop because of its association with "Christians contributing to the church," has also been an important supplementary cash crop for export production of starch.

The trend towards agrodeforestation continued in the 1930s, when intensive clearing of wild oranges to obtain land for seasonal planting of tomatoes took place, and when increasing dependence on cash incomes from expanding employment opportunities in areas such as teaching, the medical profession, and government service, coupled with contract work in the phosphate mines on Makatea in French Polynesia and emigration and associated cash remittances, began "drawing people away from the soil" (Johnson 1951, 125).

About the same time, due to falling production from wild orange trees, many of which were dying of old age and uncontrolled disease, a citrus replanting scheme was instituted by New Zealand authorities to shift the production from "haphazard" production of dispersed wild oranges to "scientific plantation agriculture." The scheme was organized so that, once the prospective owner had established secure tenure to a suitable piece of land, the administration took over and "established and matured the young plantation." The "considerable initial and recurrent costs for labour, fertilizers and machine hireage" were charged against the owner's account, to be paid when the trees began to bear, with the owner's sole responsibility being to arrange for the picking of the fruit and, in some cases, weeding and pruning, although few owners in Aitutaki exercised this option (Johnston 1967, 5).

Initially, 100 citrus plots were established on Rarotonga, each of 1.5 acres and containing 90 trees. By 1950 some 150 plots had been established on Rarotonga on the gently sloping garden lands inland from the low-lying taro swamps, or pa'i. The scheme also provided for the establishment of 50 acres on Aitutaki. Johnston (1967, 130) relates that, during this time, many existing tree groves and orange plantings were "devastated" for tomato cultivation, for "new" orange orchards, and commercial cropping of water melon, sweet potato, and tannia for the expanding local urban market. Although orange production was the main focus of the citrus scheme, limited quantities of mandarin orange or tangerine and grapefruit were also processed at the local fruit-processing factory on Rarotonga or exported by sea to New Zealand during the 1960s and 1970s (Department of Agriculture 1979).

During this period, the only government-supported activities related to agroforestry were the planting of dadap (Erythrina variegata) between alternate orange trees, with the foliage being trimmed to provide green manure; and the planting of jambolan (Syzygium cumini), Albizia falcataria, hibiscus (Hibiscus rosa-sinensis), and hedge panax (Polyscias guilfoylei), along with dadap, as wind-breaks or shelterbelts, or with the matured Albizia to provide timber for fruit-cases (Johnson 1951, 129-130).

Unfortunately, production fluctuated, even in the 1960s and 1970s when production was at its peak, with returns to growers being small, and with resultant widespread indebtedness due to charges against the owners, which increased at a faster rate than their repayment (Johnston 1967, 52). Moreover, by 1978, many of the citrus plantings were in a rundown state or beyond their "economic life" of approximately 25 years, thus requiring replanting programmes to rejuvenate the industry (APU 1978). Ward and Proctor (1980, 375) suggest that serious consideration should be given to the possible "winding up" of the citrus industry and the use of land for other purposes, because in only one year during the 1970s did real gross receipts exceed real purchased input costs in the citrus industry, with the net income of producers during other years being "wholly financed by subsidies."

The economic failure of the citrus industry, the low prices received for tomato exports, and a continuing focus on export diversification have led the government and local farmers to look for other commercial agricultural alternatives. Most prominent have been the implementation of banana replanting and extension schemes on both Rarotonga and Aituaki in the 1960s to increase banana exports to New Zealand and the official promotion of plantation culture of papaya for air-freight export to New Zealand. There was also an unsuccessful attempt to establish a commercial pineapple project on Rarotonga in 1966 (Tudor 1972, 143). Also significant have been the increasingly capital-intensive and monocultural production of taro, sweet potato, tannia, and cassava, both for the rapidly expanding local market and for export to New Zealand as well as for subsistence consumption.

As in the case of citriculture, subsidization of production of such activities has been considerable. In the case of the Aitutaki Banana Scheme of 1964, which established a 50-acre nursery and encouraged replanting with free planting material and loans for the purchase of fertilizer, the estimated cost was NZ$106,000, of which NZ$70,000 represented advances to growers. By 1970, 500 acres had been planted under banana replanting schemes on both Aitutaki and Rarotonga (Tudor 1972, 143). Unfortunately, monocultural banana production has always been beset with disease problems, most notably black-leaf-streak fungus and bunchy-top virus, and widespread devastation due to periodic tropical cyclones, not to mention marketing problems related to unreliability of ship transport and competition from Ecuadorean bananas. These problems, plus an increasing focus on the commercial production of papaya and, to a lesser extent, beans and capsicum for air freight to New Zealand, and on rapid tourism development since the opening of the new international airport in 1973, brought about the cessation of banana exports from Rarotonga in the late 1970s, leaving Aitutaki as the sole producing area for export production of bananas in the Cook Islands. As of July 1988, despite a devastating tropical cyclone in December 1987, 150 growers on Aitutaki had 350 acres (141 ha) under banana monoculture, with the individual holdings ranging from 2 to 7 acres. Disease problems and irregularity of transport, however, continue to hinder development of the industry.

Upland forest agroforestry

Although technically not part of the agricultural system, the forest of the hill slopes and mountainous interior of Rarotonga, which in most areas seems to be in nearly its original state (Sykes 1976), contains numerous species of widespread cultural utility and ecological importance and, like the coastal strand forest, must be seen as an extension of the agroforestry system. Such areas take on greater importance as refugia for endangered species as agrodeforestation proceeds in more accessible garden areas. As argued by Merlin (1985, 81), the preservation of the upland forest of Rarotonga, where over 92 per cent of all woody plants are either indigenous or endemic, is particularly important because "the native coastal and lowland vegetation of this high volcanic, tropical island has either been completely removed or heavily disturbed."

The major species in the upland forest include Fitchia speciosa, in more open sites, and Homalium acuminatum, as the dominant in the lowerslope forest areas. Other common species include Hibiscus tiliaceus, which is very common up to 250 metres, Aleurites moluccana, Elaeocarpus rarotongensis, Xylosma gracile, Pittosporum rarotongensis, Bischofia javanica, Macaranga harveyana, and Weinmannia rarotongensis, with Planchonella grayana and Terminalia glabrata being more localized, and the recent introduction, Cecropia palmata, scattered throughout some areas of forest. Common understorey shrubs include Canthium barbatum, Ixora bracteata, Macropiper latifolium, Mertya pauciflora, and, sometimes, Alstonia costata. Numerous lianas of considerable cultural importance, such as Freycinetia wilder) (kiekie), Alyxia elliptica (maire rakau), and Jasminum didymum, are also present (Merlin 1985; Sykes 1976).

On the wetter, higher slopes, Metrosideros collina is dominant, with Fitchia speciosa, Elaeocarpus rarotongensis, Pittosporum rarotongensis, Weinmannia rarotongensis, and, to a lesser degree, Coprosoma laevigata, Xylosma gracile, Geniostoma rarotongensis, Morinda forester), and a small shrub, Vaccinium cereum, being present. On ridge crests, Fagraea berteriana is most dominant, with other species including Fitchia speciosa, Homalium acuminatum, and, to a lesser extent, Canthium barbatum, Alyxia elliptica, Metrosideros collina, Hibiscus tiliaceus, Coprosoma laevigata, and Mertya pauciflora.

Increasing monoculture and agrodeforestation

The impact of export-oriented monocultural agricultural development and increased monocropping of root crops on both Rarotonga and Aitutaki have led to the widespread recent removal of even common useful trees, such as beach hibiscus (Hibiscus tiliaceus), Tahitian chestnut, mango, kapok, Polynesian vi-apple, candlenut, coral tree (Erythrina variegata), matakoviriviri (Adenanthera pavonina), and even coconut palms, as well as recently introduced species such as jambolan (Syzygium cumin)) and Albizia falcataria on both islands. Even the cultivation of a range of banana cultivars, which have long constituted dominant staples, has declined dramatically. Whereas on Aitutaki, there remain few, if any, forest stands or tree groves, except for stands of Hibiscus tiliaceus, on Rarotonga there are large areas of native upland forest, and tree groves can still be found in some areas, with traditionally important species still scattered throughout low-lying swamp and and in upland garden areas.

Some trees that were formerly abundant, such as mata'oi (Cananga odorata), the flowers of which were highly valued for garlands and scenting coconut oil; Malay apple, or ka'ika (Syzygium malaccense), and oceanic lychee, or tava (Pometia pinnate), both with edible fruit; mati (Ficus tinctoria) and ava (Ficus prolixa), the best fibre of which was so important in the manufacture of bark cloth; tou (Cordia subcordata) and miro (Thespesia populnea), two of the most valued woods for wood carving; and other species of widespread medicinal importance throughout Polynesia, such as Glochidion ramiflorum, Alphitonia zizyphoides, and Grewia crenata, to mention only a few, are increasingly rare, and, in some cases, endangered or possibly extinct.

The development processes that have been described in this section have led on Rarotonga and Aitutaki to agrodeforestation and to a cultural abandonment of, and failure to replant, traditionally important trees. The antiarboreal pressures are worse in these islands than almost anywhere else in the Pacific, with the possible exceptions of Nauru and Hawaii. One could go so far as to suggest that a combination of factors has brought about an "agrodeforestation of the mind" in a generation of young Cook Islanders, who neither know the names of their trees nor are able to find one!

The Marquesas Islands, French Polynesia The Marguesas Islands, French Polynesia

Decker's (1971) study of "Plants, Man and Landscape in Marquesan Valleys," although not exclusively focused on agroforestry, provides a thorough description of agroforestry in the inhabited valleys and middle uplands of the Marquesas Islands almost a quarter of a century ago.

The islands of the Marquesas are highly eroded, steep, cliffed remnants of ancient submarine volcanoes. The islands lack coastal plains, except at heads of embayments, but possess large habitable, well-watered amphitheatreheaded or elongate canyon valleys, with boulders in tributary beds upstream and rich colluvium in the back-valleys. Recognizable flood plains occur only in the lower reaches of some valleys. In the absence of coastal plains, backvalley colluvial slopes are the most naturally productive areas and the mainstay of Marquesan agroforestry activities. Because of the steeply sloping nature of the islands, severe erosion threatens unvegetated areas. The two valleys studied in detail by Decker, the relatively wet Puama'u valley on Hivaoa Island and the relatively arid Vaipaete valley on Uahuka, illustrate the range of Marquesan lowland environmental conditions (Decker 1971, 9-31).

Population densities were, until recently, relatively low because of widespread post-European-contact depopulation last century owing to disease and outright slaughter. Population has increased, particularly since World War II, although continued out-migration to Tahiti, New Caledonia, and France has kept populations relatively small. None the less, long human occupation has led to the extinction of many species and the substitution of a largely alien flora for the indigenous flora (Decker 1971, 31), with habitual burning being responsible for Miscanthus tall grasslands and Cleichenia fern lands in upland areas (Decker 1971, 95).

Agroforestry zones

Marquesan agroforestry as a land use can be divided into five zones:

  1. strand-side and wetlands;
  2. village areas, including home gardens;
  3. subsistence gardens and grove lands, usually found up-valley, from which the bulk of vegetable food, marketable produce, and raw materials of everyday utility originates;
  4. tracts of productive fallow forest land in the back-valleys or in the valleys proper; and
  5. extensive tracts of non-arable uplands, used mainly as fattening grounds for horses, for hunting feral animals, and gathering natural products (Decker 1971, 107-108).

Strand-side and wetlands

Bayshore areas are dominated by vast stands of coconuts with other useful large trees such as breadfruit, mango, and kapok. Relict large trees such as Barringtonia asiatica, Calophyllum inophyllum, and Terminalia catappa, formerly useful to the Marquesans, but rarely planted today, are also found in foreshore areas (Decker 1971, 110). Other ubiquitous strand species, so common in Kiribati and coastal areas elsewhere in the Pacific, are conspicuously rare here, although many can be found in inland forests. Stream bed or riparian and marshy areas are dominated by thickets of Hibiscus tiliaceus, with Tahitian chestnut (Inocarpusfagifer) also common.

Village areas and home gardens

Marquesan villages are located in valleys at a distance from the shore to reduce risk from tsunami damage with dwellings located along principal paths and occupying discretely bounded yards of various sizes. Village or home-garden agroforestry is characterized by a plethora of exotic plants accumulated during some 1,500 years of human arrivals from over the sea, with the accumulation of exotics having increased considerably since European contact (Decker 1971, 124).

Dominant species include the coconut and breadfruit, with mango and kapok (Ceiba pentandra) being very common, along with the ubiquitous and usually spontaneous and protected beach hibiscus (Hibiscus tiliaceus), which forms a part of the backdrop of every village. Common spreading ornamental shade trees include the rain tree, or monkey-pod (Samanea saman) and the flame tree, or poinciana (Delonix regia), with Albizia lebbeck also present in dwelling areas.

Immediately surrounding dwelling areas in uncared-for places, on the islands of Nukuhiva and Uapou, are extensive monospecific stands of Leucaena leucocephala, serving as the main lowland fodder for horses, which are highly useful and abundant draught animals in the Marquesas and which are rotationally fed on Leucaena in the lowlands and grazed on limited areas of upland pasture and shrub land (Decker 1971, 129).

The village home gardens are a very directly useful kind of agroforestry serving as a source of:

  1. breadfruit and other starchy foods to supplement produce from back-valley gardens;
  2. esteemed seasonal fruits and other rare plant products for every day use, safe from pilfering under the eyes of the household;
  3. flowers for ornamentation and perfumery of home and person; and
  4. a source of pride, sentimental satisfaction, and ostentation, based on special plants from distant places or from special persons.

Because such plants must be accommodated along with stray animals and children, they are usually hardy, thus favouring trees and tree-like shrubs. Less hardy shrubs and colourful herbs are commonly planted against the house beneath the eaves, with hardier shrubs and trees scattered elsewhere, particularly along borders delineating home-garden boundaries.

The most common staple plants in villages and home gardens are coconuts, breadfruit, and banana cultivars, with important fruit-trees including mango, papaya, lime, avocado, soursop and sweetsop (Annona muricata and A. squamosa), guava, and tamarind. Sugar cane is also common in home gardens.

Tree-like shrubs most commonly planted alone borders include Polyscias guilfoylei, Hibiscus rosa-sinesis, and Gardenia taitensis. All are important ornamentals, with the latter highly favoured for its fragrant flowers, which are used in garlands and for scenting coconut oil. Other trees and shrubs with fragrant flowers or of ornamental importance include the perfume tree or ylangylang (Cananga odorata), bougainvillea (Bougainvillea spectabilis), croton (Codiaeum variegatum), cordyline (Cordyline fruticosa), and the coral hibiscus (Hibiscus schizopetalus). The cultivated pandanus (Pandanus tectorius var. Iaevis), so important in the production of plaited ware, and kapok (Ceiba pentandra) are also common in home gardens (Decker 1971, 141-142).

Where households border streams, hardy useful species, both spontaneous and planted, include Hibiscus tiliaceus, Tahitian chestnut (Inocarpus fagifer), coconuts, kapok, and guava (Psidium spp.). Also found in home gardens are a wide range of non-tree, mostly ornamental, species, but also including staple ground crops, such as taro (Colocasia esculenta), tannia (Xanthosoma sagittifolium, known locally as tarua), and the ubiquitous pineapple (Ananas comosus). Also present is giant taro (Alocasia macrorrhiza), or kape, a staple in western Polynesia and a famine food elsewhere, but serving only as an ornamental and never eaten today in the Marquesas.

Back-valley or up-valley gardens and grove lands

Back-valley or up-valley agroforestry gardens and grove lands are the main sources of Marquesan subsistence and commercial production. With rapid population increase since World War II and recent coconut-planting subsidies, gardens and grove lands, particularly coconut plantings, have expanded into extensive back-valley areas of old fallow forest, much of which remains from agroforestry activities prior to widespread depopulation in the Marquesas during the nineteenth century.

Back-valley agroforestry activities provide:

  1. staple breadfruit, coconut, banana, and root crop production;
  2. food for household animals, including forage for family horses;
  3. cash crops, predominantly coconuts; and
  4. firewood, construction materials, cordage, medicines, perfumery, handicrafts, and other utilitarian products (Decker 1971, 160).

Marquesan agriculture or horticulture has remained remarkably Polynesian in character, being based on intercropping of long-lived root crops with trees that are mostly vegetatively propagated. The system favours the survival of trees and perennials because of its emphasis on diversity and variety rather than maximum yield, the extensive application of labour in terms of weeding and crop care, and the propensity to intercrop old gardens or clear new land when yields decline. Throughout all Marquesan operations having to do with plants there is an emphasis on the care, protection, and utility of individual plants, rather than on the crop or yield of an entire field or planted stand of trees. Consequently, the word "horticulture" is used advisedly to highlight that emphasis.

For the clearing of new gardens, wooded localities are always preferred because ground preparation is less laborious in the shade where weedy undercover is little developed and soil structure is generally loose. Old fallow areas dominated by Hibiscus tiliaceus are seen as the best sites for gardens, although Xylosma-Sapindus forest and transitional thickets are commonly cleared in drier areas. Edaphic requirements influence site selection, with bananas being planted in wetter alluvial sites, often in sheltered ravines, and cassava higher upslope in lighter soils up to the practical limit of cultivation on slopes up to 30° (Decker 1971, 164).

More permanent difficult-to-clear groves of trees such as Ficus prolixa, Mangifera indica, and Inocarpus fagifer are seldom cleared. In garden clearance, small additions to existing gardens are generally hacked out of the fallow vegetation at almost any time of the year, with Hibiscus tiliaceus being cleared and Ficus prolixa and mangoes generally burned at the base and pruned or pollarded, but left remaining to live in garden sites. After firewood has been saved and temporary shelters of Hibiscus tiliaceus poles and coconut thatch prepared, most areas are burned, although sometimes debris is left to rot in place.

The fallow forest is dominated by many species once of great subsistence utility to Marquesan society. The trees have been spared because they retain some use or because people see them as possessing traditional interest. Marquesan horticulture accepts their presence in gardens as readily as it does large old trees and bulky boulders, thus enhancing the variety and bounty of the new garden. The most common species in this category include the ubiquitous Hibiscus tiliaceus, coconuts, breadfruit, and mangoes, with other useful species spared during garden clearance including papaya, coffee, Tahitian chestnut (Inocarpus fagifer), banyans (Ficus prolixa), candlenut (Aleurites moluccana), Malay apple (Syzygium malaccense), guava (Psidium spp.), beach almond (Terminalia catappa), and Fagraea berteriana, plus a range of coastal strand species.

The persistent dominance of ocean-dispersed coastal strand species in Marquesan secondary forests is remarkable in light of Egler's (1939, 45-46) findings that such plants are poorly adapted for terrestrial migration. The major trees or tree-like plants that fall into this category are Hibiscus tiliaceus, Pandanus spp., Erythrina variegata, Cordia subcordata, Guettarda speciosa, Morinda citrifolia, Premna serratifolia, Sapindus saponaria, Cerbera manghas, and Thespesia populnea, all of which seem to regenerate well and are widely naturalized in inland localities. Other species that occur rather sporadically in old fallow forest or adjacent to suspected former habitation sites included Barringtonia asiatica, Calophyllum inophyllum, Terminalia catappa, and Cocos nucifera, most of which are encountered in fallow forests below 500 metres, with the exception of Hibiscus tiliaceus and Pandanus tectorius, which extend up to the cloud forest. Although illadapted to terrestrial dispersal, all these species possessed utility in the aboriginal economy, which argues strongly for their human carriage inland, a dispersal that would have been reinforced by profound human disturbance of the indigenous vegetation below the cloud zone.

The presence of large inland tracts of old forest fallow can be attributed to internecine warfare before and after European contact. These conflicts led to the retreat of gardens to steeper, secure backvalley slopes. Their persistence, today, in fertile inland valleys reflects the widespread abandonment of pre-1800 subsistence groves after drastic depopulation and a changing settlement pattern, with the abandonment of back-valley dwellings in favour of new coastal villages around foreigners' buildings near bay anchorages and mission dwellings (Decker 1971, 90, 72).

The great plant diversity within garden successions in garden and grove areas results from diverse planting and clearance strategies, all of which minimize toil in the long run. Once a new clearing is planted, soil and plant cover are only minimally disturbed, with minimum tillage with digging sticks or mattocks being used to plant new crops, and weeds being cleared by machete, thus maximizing soil protection against erosion, leaching, oxidation, and the depletion of organic material.

In terms of new garden development, perennial arborescent crops such as coconuts and bananas are sometimes planted in the fallow forest before clearance. When the fallow is finally cleared, if these trees have set little foliage, the debris is burned; otherwise, it is allowed to rot and cleared periodically around the bases of the trees. This practice seems to be followed primarily by individuals planting large areas of coconuts in response to the drive to expand copra production. As a result, it is common to see plantings of immature coconuts only a few years old in areas well along with fallow regrowth. Although perhaps detrimental to optimum growth of coconuts' such a multi-purpose agroforestry strategy is a productive compromise between the aims and resources of both administration and individual Marquesans, providing an investment in future production and being conservative of soil fertility. The fallow vegetation can be recleared slowly, a practice that allows labour input to be spread over time.

After the fallow has been cleared and burned, the staple root crops and various short-lived crops are established, together with a range of banana and plantain cultivars (Musa cultivars) and the less common fe'i banana (Musa troglodytarum), breadfruit, coffee, oranges, avocado (Persea americana), and sugar cane. The garden lasts for three or four years, after which it becomes a grove dominated by trees, aborescent herbs (mainly banana cultivars), and palms, with breadfruit, citrus, and coffee beginning to bear. Bananas and fe'i (huetu) bananas form discrete dense groves. Cassava still persists in the fallow as a source of planting material, with Xanthosoma and giant taro also prevalent, their leaves being used to wrap food. Regularly spaced coconut palms, which have not yet begun to bear, are the only evidence of geometric order. Among the thickening canopy, there remains some spontaneous growth of Commelina diffusa and Oplismenus composites as horse fodder, with a few papayas, chillis, guavas, and Morinda citrifolia, which have been spared, still persisting in some areas. In the few open areas, some grasses, herbs, and shrubs persist (Decker 1971, 188-190).

Of note is the abundance of papaya seedlings in areas recently cleared of fallow, especially Hibiscus tiliaceus fallow. Because of papaya's abundance, it is rarely deliberately planted and is not considered a primary human food, but is fed mostly to animals such as pigs, chickens, and dogs. Other pioneering weeds, such as Commelina diffusa and Oplismenus composites (this latter species occurring in abundance only on the island of Uahoka, where very few alien weeds had become established by the 1960s), are highly valued as horse fodder, as new gardens are among the few places where tethered animals can graze in lowland areas while farmers work in the field. Other weedy species, such as the perennial chill) (Capsicum frutescens), are spared, although shoots of Hibiscus tiliaceus, Morinda citrifolia, and Psidium spp. are usually suppressed by slashing in the early stages of gardens.

Arboreal species, either planted or protected, that persist in backvalley gardens and grove lands include fruit-trees, such as pomelo (Citrus grandis), Malay apple (Syzygium malaccense), hog plum, or Polynesian vi-apple (Spondias dulcis), Inga edulis, mango, Psidium species, and Iychee (Litchi chinensis); other utilitarian species present are pandanus, kapok (Ceiba pentandra), perfume tree (Cananga odorata), Coffea arabica, Tahitian gardenia (Gardenia taitensis), and a range of hibiscus cultivars (Hibiscus rosa-sinensis, H. schizopetalus, and ornamental hybrids). Aleurites moluccana, Canthium barbatum, and Thespesia populnea are found occasionally.

Productive fallow areas and mature groves

As garden groves begin to mature and coconut palms start to bear, some seven to ten years after original clearing, production shifts to copra from food, fibre, leaf, and fodder production, and groves take on a character that may persist for decades. Discrete groves of bananas and fe'i bananas (Musa troglodytarum) and breadfruit still produce in favoured sites, and family groves of Tahitian chestnut (Inocarpus fagifer) remain along with occasional remnant Xanthosoma plants. Sometimes, a new tree is added to an ageing grove. The area around individual coconut palms is generally kept free of high growth by slashing with machetes during the search for fallen mature nuts. Further clearing is brought about by the burning of husks, fallen fronds, and other debris during the extraction of the coconut meat, and the tethering of horses to graze the marginal pasture around each tree (Decker 1971, 195-196).

As coconut groves expand at the expense of old fallow forest and more diversified agroforestry activities, the burden of weeding and brush cutting increases and is made more difficult by the invasion of harder-to-clear shrubby weeds such as Indigafera suffruticosa and Ocimum gratissimum, which make collection of nuts more difficult. The mature grove will eventually consist of the remnants of the original fallow forest, planted trees, and pioneering trees such as several varieties of Hibiscus tiliaceus, Morinda citrifolia, guava (Psidium guajava and P. guineense), mango (Mangifera indica), and kapok (Ceiba pentandra), which exhibits phenomenal reproductive ability in some areas, with some "fallow" forests consisting entirely of tall coconut and kapok trees (Decker 1971, 200).

It is difficult to distinguish advanced stages of matured garden groves from old fallow forest, some areas of which have probably re mained out of cultivation for 100 years or more. Although Hibiscus tiliaceus and Inocarpus fagifer dominate old fallow forests in most back-valley ravines and canyons, harvestable coffee is found naturalized in many, and large banyans (Ficus prolixa), Aleurites moluccana, Terminalia catappa, and Spondias dulcis also remain in these areas. In drier areas of old fallow forest, I. fagifer becomes uncommon, with H. tiliaceus becoming the dominant along with scattered individuals of Canthium barbatum, Glochidion sp., Wikstroemia foetida, Morinda citrifolia, Piper latifolium, and Pipturus argenteus, with coffee flourishing and vanilla occurring occasionally as escapes. Discrete groves of bamboo (Schizostachyam glaucifolium) and mango are also found in areas of old fallow forest. Of particular importance are the widespread fallow groves of a vegetatively propagated sterile form of Hibiscus tiliaceus (H. tiliaceus var. sterilis), which was possibly an aboriginal cultigen because of its wide utility in construction and for fruitpicking poles (Decker 1971, 258-262).

In transitional or drier areas, where vegetative growth is slower, back- or up-valley gardens are maintained for longer periods of time, interspersed between uncleared areas of mid-slope forest dominated by Pandanus tectorius, Sapindus saponaria, and Xylosma suaveolens. The crops are essentially the same, although woody perennials including Canthium odoratum, Capsicum frutescens, Carica papaya, Ficus prolixa, Hibiscus tiliaceus, Morinda citrifolia, Cerbera manghas, Premna serratifolia, Psidium spp., Sapindus saponaria, and Xylosma suaveolens, are generally spared in the clearance for new gardens or protected in garden regrowth (Decker 1971, 206). On exposed uplands and ridge tops Casuarina and Pandanus groves, Miscanthus floridulus and Tricholena rosea grasslands, fern scrub lands, and scattered forest stands and thickets exist, along with occasional thickets of Hibiscus tiliaceus in more humid areas (Decker 1971, 231-232).

Thickets in transitional areas include, most commonly, Psidium guineense, P. guajava, Celastrus crenatus, Colubrina asiatica, Premna serratifolia, and Morinda citrifolia, all of these, even C. asiatica, approaching tree stature. Scattered in the widespread Miscanthus (kokaho) grasslands, which have been maintained by periodic burning since pre-European-contact times, are a number of tree species, including, most prominently, Psidium spp., as well as Morinda citrifolia, Celastrus crenatus, Pandanus tectorius, Hibiscus tiliaceus, Xylosma suaveolens, and Casuarina equisetifolia. The forest on lower canyon slopes is dominated in most areas by Xylosma suaveolens and Sapindus saponaria, which give way to Hibiscus tiliaccus in the wetter areas or to coconut groves on the lower slopes. Forests in the valley ravines and bottoms in the transitional zones include the common Sapindus saponaria and Xylosma suaveolens, as well as Hibiscus tiliaceus, Thespesia populnea, Erythrina variegata, Pisonia grandis, Pandanus tectorius, Morinda citrifolia, Canthium odoratum, Premna serratifolia, Aleurites moluccana, and occasionally Celtis pacifica, and Cordia subcordata, with large banyans (Ficus prolixa) as common large emergents. Also appearing sporadically in ravines and valleys are planted or spontaneous economic plants such as chillies (Capsicum frutescens), and kapok, orange, and coconut trees (Decker 1971, 244 245).

Most of the same species, with the exception of Cordia subcordata, Celtis pacifica, and Pisonia grandis, which are absent or uncommon away from the coast and valley bottoms, are also found in transitional forest on upland slopes and inland ridge crests, with Xylosma suaveolens and Sapindus saponaria being the co-dominants, except where the latter has been exploited for firewood for bakers' ovens. Species also found in ridge-crest areas include Glochidion sp., Canthium barbatum, Coffea arabica, occasionally Cocos nucifera, and Mangifera indica; the more frequent occurrence of Hibiscus tiliaceus; and discrete groves of Pandanus tectorius and Casuarina equisetifolia (Decker 1971, 246-252).

In the more arid zones, arborescent plants are more restricted to relict forest stands in the interior and in well-watered and sheltered ravines, with shrubby species (such as Indigofera suffruticosa) and Leucaena leucocephala invading extensive inland areas and coastal areas. Acacia farnesiana, although not invasive, persists in some areas. Other shrub and tree species found in non-garden areas include Gossypium barbadense, Jatropha gossypifolia, Cordia lutea, Tamarindus indica, and Albizia lebbeck, the latter two being occasionally planted (Decker 1971, 214-215).

Open woodland stands of Sapindus saponaria are common, with other tree species in arid areas including Celastrus crenatus, Celtis pacifica, Syzygium spp., Guettarda speciosa, Morinda citrifolia, Psidium spp., and Thespesia populnea. In areas denuded by goats, solitary banyans (Ficus prolixa) are occasionally found.

The most aggressive of the exotic plants and one of the most difficult to clear for new gardens is Leucaena leucocephala, which is controlled in some banana, cassava, and tannia gardens by deliberately planting cuttings of another aggressive weed, Commelina diffusa. After original Leucaena clearance, this weed retards further regen eration, at the same time providing valuable easily-cleared fodder for the ubiquitous horse. This is, however, not a practice in gardens with spreading ground crops such as sweet potatoes and melons, which thrive on clear cultivation and would suffer from competition from Commelina diffusa. Where Leucaena has become established, the tenacious stands seem to exclude all groundcover of other species, with little or no tendency toward succession by native or indigenous trees.

Non-arable uplands

The non-arable uplands, which are dominated by Gleichenia linearis fern lands, with scattered Casuarina equisetifolia and Pandanus tectorius, are found on the deep, degraded latosols of the upland interfluves. Along with the Miscanthus floridulus grasslands, these apparently pyrogenic and fire-maintained areas serve as important seasonal grazing areas covering most of the interfluvial terrain.

The humanized forest of the Marquesas

The close examination of Marquesan plant cover offered here reveals yet again the intense humanization of many Pacific Island landscapes. Looking at the Islands' often extensive forest cover, the uninitiated would be likely to assume that the forest was largely natural, punctuated here and there by gardens or grassy ridges - a land waiting for development. Instead, the forest - much of which is a human artifact is intricately linked to the inhabitants' production of food, materials, and cash crops. The individual trees are the wheels of the forest factory, turning out a great variety of products while also working slowly together to enhance the future productivity of garden soils and to help maintain the present stability of slopes and the clarity of streams. The efficiency of this "factory" could in many instances be improved; its human managers may make mistakes; hard-to-manage components, like Leucaena, may get introduced; none the less, a forest "factory" is already operating in place. It would be useful in planning for the future welfare of land and people to recognize this factory's existence and its significance, before dismantling it under the false impression that it serves little purpose and has no connection with human activities.

A note on Micronesia

Micronesia extends across the western Pacific Ocean from 15°N to 3°S latitude and from 132°E to 177°E longitude. Although covering an ocean area of over 7 million sq km, it contains a land area of only 2,706 sq km. Except for Kiribati and Nauru - which at the time of their independence were, respectively, a British colony and a UN trust territory administered by Australia - Micronesia's most recent colonial administrator was the United States of America. Today, what was the American Trust Territory is divided into the Territory of Guam, the Commonwealth of the Northern Marianas, the Republic of Palau, the Republic of the Marshall Islands, and the four Federated States of Micronesia - Yap, Chunk (Truk), Kosrae, and Pohnpei.

Micronesian agroforestry can be separated into two distinct types. The first is practiced on the higher and larger islands of central and western Micronesia. Although similar to many of Polynesia's agroforestry systems, certain differences of emphasis can be discerned. The giant swamp taro. Cyrtosperma chamissonis, is more important in Micronesia than in Polynesia; there is a greater emphasis on mulching, artificial soil improvement, and water management in Micronesia; and the mixed tree gardens, or "agroforests" - the equivalent of Polynesia's permanent tree groves - are perhaps even more important in Micronesia than in Polynesia.

The second type of Micronesian agroforestry is practiced on atolls, where environmental constraints are extreme and population densities are high compared with most of Melanesia and Polynesia. The agroforestry system developed in response to these severe conditions is unique in the relative dominance of trees over ground crops and in the sophisticated, intensive systems of mulching and soil improvement employed in the face of the impoverished atoll soils.



Micronesia

Traditional agroforestry in the high islands of Micronesia

Traditional subsistence agriculturalists of the higher and larger islands of Micronesia developed a wide range of agricultural technologies and systems for the production of food under different environmental conditions and in different locations. These included permanent systems of wetland taro cultivation, mixed tree gardening, intermittent tree gardening (shifting cultivation), home or kitchen gardening, and open grassland cultivation (Falaoruw 1985; OTA 1987). The use of any one of these systems did not preclude the use of others. Most islanders produced food using all five systems and, in addition, exploited the reefs and ocean for marine products. Most attention will be given here to mixed tree gardening, an elaborate and refined indigenous agroforestry that is the predominant rural land use on many Micronesian high islands (table 7; Raynor and Fownes 1991a, 139). The other four systems will be briefly discussed, particularly with regard to their relationship to trees.

Table 7 Areas in hectares of land classes in Micronesian high islands. (Total agroforest includes agroforest/tree gardens and coconut plantations; Truk data are for the high islands of Moen, Dublon, Fefan, and Eten only)

  Belau Kosrae Pohnpei Truk Yap
Forest 28,093 7,066 19,683 986 3,882
Secondary forest and vegetation 594 1,272 1,843 252 553
Agroforest/tree gardens 187 2,585 11,741 2,378 2,379
Coconut plantation 743 124      
Total agroforest 930 2,585 11,865 2,378 2,379
Non-forest 8,285 263 2,102 554 2,743
Total area 38,832 11,186 35,493 4,170 9,557
(Percentage agroforest) (0.48) (23) (33) (57) (25)

Sources: Belau: Cole et al. 1987; Kosrae: Whitesell et al. 1986; Pohnpei: McLean et al. 1986; Truk: Falanruw et al. 1987a; Yap: Falaaruw et al. 1987b.

Wetland taro cultivation

The production of Colocasia esculenta and Cyrtosperma chamissonis taros in essentially permanent patches has long been practiced on all the high islands of Micronesia, often by means of sophisticated systems of water management that minimized soil erosion and stagnation of water. Favoured areas for the wetland cultivation of taro are alluvial bottom lands and the lowland swamps and marshes located inland of the mangroves, generally within agroforests (Falanruw 1990). The taro patch may act as a filter to minimize sedimentation of the lagoons and ocean (Falanruw 1990; OTA 1987). When a patch under fallow is selected for replanting with Colocasia esculenta, it is cleared of vegetation and drained, the soil is dug up, and various leaves, twigs, and sea grasses are added as a mulch or green manure (OTA 1987), with Wedelia biflora, Carica papaya, and Macaranga sp. recognized as good manuring species (Sugiura 1942). After mulching, the patch is worked to produce a fertile muck of desired consistency and planted with cormels or corm tops. Harvesting occurs six months to a year after planting, and the patch may be replanted to provide a continuous supply (OTA 1987). If the yield or quality declines, the patch is allowed to lie fallow for a number of years.

The cultivation of Cyrtosperma chamissonis requires less labour and attention than is needed for Colocasia (Falanruw 1990; McCutcheon 1981). Green manures are not added to the soil to improve its fertility for cultivation of Cyrtosperma chamissionis, and cultivation methods include the periodic removal of fallen vegetation and debris in order to maintain the flow of water through the system. Cyrtosperma is more shade-tolerant than Colocasia and thus more compatible with tree culture, and an integral part of the agroforestry system of Yap (Falanruw 1990, 97-98). Cyrtosperma is the preferred aroid in Yap (Hunter-Anderson 1984) and Truk (Alex 1965); Colocasia remains the preferred food in Pohnpei (Hunter-Anderson 1984), where not all of the trees were cut back during garden clearance, with some, such as Hibiscus tiliaceus, being left standing to provide shade for the young plants.

Mixed tree gardening

A very well developed form of traditional agroforestry in Micronesia is the tree garden or agroforest (Falanruw et al. 1987a; 1987b; OTA 1987). Except for the initial planting and care while the trees are young, this agroforestry system requires little energy input, but provides Micronesians with an abundant supply of different tree crops and products from marginal lands over a long time period, while maintaining a permanent canopy cover. Except in Palau, such agroforests cover considerable areas of high-island Micronesia (table 7).

The composition and structure of these agroforest gardens varies from place to place. In coastal areas, they tend to be relatively simple, consisting of few species, dominated by coconuts. In Truk, breadfruit is a dominant species of mixed tree gardening (Goodenough 1951). In Guam, breadfruit, coconuts, and Cycas circinalis were harvested from mixed tree gardens. In Palau, these mixed forests were called chereomel and consisted of a wide range of food, fruit, and timber trees, including coconuts, mango, breadfruit, Terminalia catappa, and Inocarpus fagifer (McCutcheon 1981). The forests serve as a habitat for feral and domesticated animals, provide traditional medicines and other culturally valued products, and are or were a source for canoe hulls, building materials, and firewood.

As Falanruw (1990, 98) explains for Yap, the likely development process of tree gardens was the planting of trees for food and other uses around homesteads and in the drained areas created by the excavation of taro patches and the construction of drained paths between homes and villages. These home and path-side tree gardens became confluent and today make up a significant proportion of the island's terrestrial vegetation and contain about 50 native and introduced tree species, such as coconut, breadfruit, betel-nut (Areca catechu), Inocarpus fagifer, many varieties of banana and citrus species, Pangium edule, papaya, cacao, and guava. Of species that have long been cultivated on Yap, there are numerous named varieties-for example, 21 named coconut varieties and 28 named breadfruit varieties. Tree gardens are reported to be largely self-fertilizing. Owners maintain their tree gardens by selective pruning and removal of undesirable trees; occasionally desired trees are planted. Once established, such tree gardens require little maintenance.

The most detailed recent description of indigenous agroforestry in Micronesian high islands is the account by Raynor and Fownes (1991a; 1991b) of Pohopei, where the traditional agroforestry system still supports most of the island's 28,000 people despite increasing emphasis on the cash economy.

The general characteristics of the Pohnpeian agroforestry system are an extensive, permanent overstory of breadfruit, coconut, and forest trees above fruit and multipurpose trees, and an understory of shrubs, root crops, and herbaceous plants. Although sharing many characteristics of homegardens, the system is best classified as a ``multistory,, tree garden because it is not limited to the area immediately surrounding the house compound but extends throughout the landscape. Breadfruit and yams (Dioscorea spp.) are the major staples, and complement each other in seasonality.... Hibiscus tiliaceus is the premier "multipurpose" tree, because it is used for firewood and light construction, poles or whole trees are used for yam trellises, its leaves are used as green manure in soil pits dug for yams, its bark for rope, and its inner bark for best fibers for straining mashed Piper methysticum roots for drinking (Raynor and Fownes 1991a, 140).

Raynor and Fownes (1991a) carried out intensive sampling in 54 landholdings on Pohnpei and enumerated 102 different species, 26 of which were upper canopy species, 39 were sub-canopy, and the remainder were understorey. A few species found on nearly every holding constituted the typical Pohnpeian agroforest: coconut, breadfruit, Cananga odorata, mango, Musa spp., Hibiscus tiliaceus, Morinda citrifolia, Alocasia macrorriza, Dioscorea alata, and Piper methysticum. As on Yap and elsewhere, several of the important species have many recognized cultivars, a diversity that is an important component of the aggregate biological diversity of Micronesian agroforestry systems (Raynor and Fownes 1991a, 151).

Because they could detect no patterns in species composition dependent on elevation, soil type, or particular region, Raynor and Fownes (1991a, 155) concluded that " . . . the Pohnpeian agroforest is a managed landscape, despite its superficial resemblance to forest. Species presence or absence was apparently more strongly controlled by farmers' decisions than ecophysiological constraints."

Intermittent tree gardening (shifting cultivation)

Intermittent tree gardening (i.e., shifting cultivation, or swidden) is practiced in secondary forest fallows on all high islands of Micronesia. Structurally and functionally, this system is little different from shifting cultivation systems described for the other parts of the Pacific, except that in Kosrae burning was not used in garden clearing (Wilson 1968). In contrast to mixed tree gardening, intermittent tree gardening is not a permanently productive form of land use. When the gardens, cleared from forest, are no longer harvested after one to two years of production, the site returns to fallow, succeeding through stages to a forest of spontaneous secondary species, here and there enriched with useful species such as breadfruit planted during the garden phase. The major crop in such gardens is yams of the genus Dios corea - six species of which and 34 locally recognized varieties have been recorded from Yap (Falauruw 1990, 99).

Home gardens and lanchos

Throughout high-island Micronesia, home gardens are a common feature of most households. In Guam and the Northern Marianas a variation of the home garden is known as the lancho. These provide villagers with a ready source of food, fruit, spices, herbs, and, in some cases, medicinal plants. In urban areas, these gardens are, in the main, supplementary to a wage income.

Out of an extensive pool of fruit-trees, the most commonly found are varieties of citrus, coconuts, breadfruit, and bananas. Ornamental trees and shrubs, some of which have ritual or ceremonial significance, are other components of kitchen gardens. Hibiscus hybrids, Cordyline fruticosa, and Codiaeum varigatum are as significant in the Micronesian high islands as they are in Melanesian and Polynesian societies. The latter two species were sometimes not planted in Palauan house gardens because of their association with death and the supernatural (McCutcheon 1981). In Guam, the "pickle" tree, or bilimbi (Averrhoa bilimbi), carambola (Averrhoa carambola), mango, coconuts, soursop (Annona muricata), perennial chill) (Capsicum frutescens), annatto (Bixa orellana), Citrus spp., Jatropha integerrima, Cycas circinalis, Plumeria rubra, P. serratifolia, Araucaria excelsa, and Dracaena marginata are found in many home gardens. In Palau, the betel pepper vine (Piper betle) was zealously guarded against theft (McCutcheon 1981). Other useful plants were Areca catechu, Citrus mitts, and Muntingia calabura (McCutcheon 1981). Sproat (1968) notes that Crateava speciosa has special importance in the central Caroline Islands.

For Guam and the Northern Marianas, there is little documentation of traditional subsistence cultivation. Prior to European contact, the indigenous Chamorros were mainly dependent on the ocean; root crop agriculture was rudimentary and supplemented by hunting for fruit bats, birds, and land crabs (Underwood 1987). However, under Spanish rule, and by the end of the nineteenth century, subsistence agriculture on the ranch (lancho) became accepted as the Chamorro way of life (Underwood 1987). Today, most ranchos are located in southern Guam and consist of a simply built cooking and sleeping house surrounded by food trees, chickens, pigs, and gardens (OTA 1987). Relatively few ranchos are cultivated without the use of ferti lizers or pesticides, and not all of the production is consumed at home. In the Northern Marianas, ranchos are more difficult to find because of the impacts of economic development, division of family lands, foodstamp programmes, and population increases (OTA 1987; Sproat 1968). During the Japanese administration of the Northern Marianas, traditional subsistence agriculture was replaced by the development of sugar-cane plantation agriculture.

Traditional open-canopy (ked) agriculture

The ked area lying in the interior region of Babeldaob, Palau, is characterized by exposed and eroded oxisols and ultisols, and a degraded vegetation (McCutcheon 1981). Ked also refers to the fertile grassland areas in Palau, Yap, and Pohnpei, which are used for traditional subsistence agriculture (Hunter-Anderson 1984).

Ked agriculture involves burning the grass, turning the friable soil, and hoeing ridges along the contour to reduce erosion (McCutcheon 1981). Mulching (with sea grass and Cymbopogon citratus) and ditching are other features of this open-canopy agriculture (OTA 1987). Sweet potatoes, Colocasia taro, Manihot esculenta, and pineapple are the most commonly planted crops. Cymbopogon citratus (lemon grass) is planted as a border and to prevent soil erosion (McCutcheon 1981). With crop rotation to reduce species-specific insect predation, the ked garden can be cultivated for many years without fallowing (McCutcheon 1981).

On Yap, crops grown in open areas include sweet potatoes, cassava, and vegetables. The making of sweet-potato gardens involves the construction or reconditioning of drained garden beds by cutting or burning the vegetation on the site, piling the debris on the garden area, and sometimes adding additional mulch, including washed up sea grass. Ditches are dug or deepened around the garden bed, and the soil piled on top of the mulch, a process that drains the garden bed, suppresses weeds, and provides a fertile soil-mulch-soil sandwich. Open-canopy gardening on Yap today is reported to be causing a retreat of forest and an exhaustion of soils under secondary vegetation because of careless and too-frequent burning and too-short fallow periods between gardening (Falanruw 1990, 101-102).

The ditching, mulching, and other garden preparation activities, and the relatively long period of cultivation of ked agriculture, are very similar to the intensive, semi-permanent forms of cultivation found in Papua New Guinea grassland ecosystems. On the other

Micronesian high islands, agricultural mounds and terraces, with or without stone facing, attest to the intensive and long-term cultivation of food crops. In Pohnpei, for example, bananas, coconuts, Piper methysticum, and Alocasia macrorrhiza are grown in earthen mounds and hillside terraces (Hunter-Anderson 1987).

Sustainability of Micronesian high-island agroforestry

It is clear that in the high islands of Micronesia, subsistence agroforestry has a long history and has successfully incorporated many recently introduced cultigens into the continuing evolution of traditional systems. High-island peoples developed both permanent and impermanent forms of agroforestry that provided food and materials from a large number of cultivated species and varieties. Raynor and Fownes (1991b, 164) specifically address the question of the sustainability of the Pohupeian agroforestry system, noting that Pohnpei is thought to have been settled for at least 2,500 years, and supported a pre-European-contact population as large as 50,000, compared with 30,000 in 1990. Twenty-two of the 54 farms surveyed by Raynor and Fownes were reported to be over 100 years old, but farmers claimed that many of them had been farmed as long as people had been there. The presence of domesticated trees and abandoned garden beds in today's forests on Pohnpei and Yap, and probably on other Micronesian high islands, demonstrates that the conversion of forest to garden and back to forest has a long history (Falanruw 1990, 102; Raynor and Fownes 1991b, 164). The future sustainability of high island agroforestry probably depends less on ecological than on human factors, including not only knowledge of the systems and attention to their careful management but also the desire to maintain them in the face of food imports and the tendency of young people to seek wage employment in town or to emigrate (Raynor and Fownes 1991b, 164). For the present, traditional agroforestry and subsistence agriculture remain important land uses in many high islands of Micronesia but are hardly practiced at all any longer in Guam and the Northern Marianas.

Atoll agroforestry on Tarawa and Abemama, Kiribati

Of all non-urban, Pacific Island agroforestry systems, those on atolls operate under the most severe environmental constraints and greatest population pressures. Further, in response to these constraints and pressures, atoll-dwellers have created the most intensive agroforestry in all the Pacific, with the greatest relative dominance of trees over non-trees.

Tarawa and Abemama, two of the islands of the Gilberts group of Kiribati, provide excellent examples of atoll agroforestry and will serve here as casestudies of village-level agroforestry wherein a wide range of cultivated and protected wild trees and a more limited number of non-tree plants and livestock are raised within a relatively dense matrix of coconut palms. The descriptions are based on a 10 day reconnaissance survey of agroforestry on Tarawa and Abemama in 1984, subsequent visits in 1989 and 1991, and a survey of the available literature.

Population pressure on land in Kiribati is high, most particularly on the main atoll of Tarawa, where it is projected that by 1993 (with a high annual growth rate of 4.3 per cent because of migration from the outer islands) there will be a population of over 34,000 people living on only 920 ha (Douglas and Douglas 1989, 282). Population densities on Abemama and other outer islands are significantly lower. The only agricultural export from Kiribati and the mainstay of the export trade is copra, which was worth Australian $3.1 million in 1989 (Forum Secretariat 1991).

The atolls and table-reef islands of Kiribati rarely rise more than 3 metres above high-tide level, with the true atolls surrounding large central lagoons. The highly alkaline, calcareous, and rocky soils are among the most infertile on earth, with very low water-holding capacity, little organic material, few available soil macro- and micro nutrients, apart from calcium, sodium, and magnesium, and restricted availability of iron and other micronutrients because of the high pH. Rainfall is extremely variable, with extended periods of drought being common. Groundwater is brackish to slightly salty and subject to salt-water incursion. The islands, where one is never more than 0.5 km from the sea, are susceptible to inundation by storm surge and tsunamis and the constant effect of humid salt-spray-laden winds. As Small (1972, 5) said in his book on atoll agriculture: "all this adds up to a very difficult environment for plants, and produces problems for animals and man."

Agroforestry resources

Because of the small size of the islands of Kiribati and their isolation, geological youth, and harsh environment, indigenous plant species number only 66, none of them endemic. There are just under 300 species, including exotics, mostly ornamentals and weeds, that have ever been reported to grow there (Fosberg and Sachet 1987; Fosberg et al. 1979; 1982; Thaman 1989b).

It is under these harsh conditions and with a paucity of flora to choose from that the I-Kiribati (people of Kiribati) have evolved their distinctive agroforestry system, which incorporates into a matrix of the superdominant coconut palm:

  1. indigenous species (almost exclusively pan-Pacific or pan-tropical, ocean-dispersed species);
  2. selected aboriginally-introduced food plants, such as the giant swamp taro, or babai (Cyrtosperma chamissonis), and Pandanus tectorius;
  3. some recently-introduced exotics; and
  4. settlements or villages and other urban features.

Coconut palms

Almost all coconut palms seem to have been planted, either deliberately or accidentally, by the I-Kiribati. The resultant agroforested landscape takes the form of a natural forest, rather than an orderly plantation, because a great proportion of the trees are spontaneous occurrences of different heights and ages, rather than equally-spaced trees of the same age. On both the seaward and lagoon sides, coconuts lean outward interspersed with pan-tropical strand species, whereas in the higher central portions of the islands they generally form thick stands, with young coconut seedlings and other plants in various stages of growth often forming an almost impenetrable jungle. In many areas, plants suffer from excessive density, although towards the lagoon side, where most of the settlements and giant taro (babai) pits are found, the trees begin to thin out (Catala 1957, 22; Watters and Banibati 1977, 33). Moul (1957, 1), however, found concentrations to be denser along lagoon shores on Onotoa atoll in southern Kiribati.

Sixteen locally recognized coconut cultivars are divided into two main categories according to whether the mesosperm is edible (te bunia) or non-edible (te ni), the latter term also applying to coconuts in general. Some are favoured for their juicy flesh, the quality and sweetness of their toddy, and some for the quality of their fronds, coir from the husk, or wood for use as handicrafts and building materials (Catala 1957, 25-27).

Catala (1957, 30-34) stressed the "extraordinary resistance of the palm" in Kiribati to prolonged drought and its ability to continue to produce inflorescences, which although incapable of producing commercial-grade copra, still produced the nutritionally essential toddy. The ability to withstand prolonged drought depends on the nature of soils, the degree of salinity of groundwater, the nature of tides during droughts, and the sporadic occurrence of fire during drought periods. Despite this resistance to drought and increasing salinity, the production of most palms, particularly of copra, is severely affected by drought, although palms around village sites, beside babai pits or in abandoned babai pits, and around inland ponds seem to be affected only minimally by drought because of proximity to the freshwater lens or the presence of greater domestic and organic waste near villages and babai pits. Watters and Banibati (1977, 33) reported that, after a prolonged drought in the early 1970s, only 44.2 per cent of mature coconut palms surveyed on Abemama were bearing in 1972.

A transect across Bikenibeu islet, Tarawa, contained 138 irregularly spaced palms in an area of 5,950 sq m, a very high density of 231 per ha - compared with a density of only 157 per ha on a fully stocked regularly spaced copra plantation in Tonga. In this same area, 11 pandanus trees, most of them concentrated in the mid-island portion or toward the lagoon and village end of the transect, were also inventoried. Nearer to village sites, the density was considerably lower, with 14 surveys giving an average density of only 155 palms per ha, not counting other important trees. For example, one village, covering an area of some 10,750 sq m, had only 100 coconuts, a density of 93 per ha, as well as 36 breadfruit trees. For village areas, the average density ranged from 80 to 150, whereas densities were from 200 to 350 in bush garden areas.

Watters and Banibati (1977, 35) suggested that the density of coconut palms on rural Abemama was 321 palms per ha, with densities of bearing palms being 151.8 (given a figure of 47.3 per cent bearing palms). They estimated annual nut production per bearing tree at 17.8, a figure somewhat lower than Catala's (1957) earlier estimate of 23.1 nuts per year on Tarawa, possibly because the Abemama survey was conducted after an extended drought. Coconut palms provide not only the single agricultural export possibility of copra but also oil, drink, and food for both people and pigs.

For toddy (karewe) production, which perhaps nowhere has such fundamental importance as in the harsh environment of Kiribati, the flower spathes of selected trees are cut and bound and tapped twice a day, once in the morning and once in the afternoon, yielding approximately two coconut shells of liquid per day. A dietary staple for most I-Kiribati households, especially in times of severe drought when palms produce few fruit, fresh toddy is drunk daily by most I-Kiribati. Toddy is also fermented to make a vitamin B-rich (one-third the level found in brewer's yeast) drink of varying alcoholic content, a boiled-down syrup, which can be kept without fermenting, and a solid caramelized form (Catala 1957, 44-46).

Particularly in rural areas, coconut flesh is the major source of dietary fat and energy, as well as contributing some iron, fibre, and other nutrients, and is prepared and consumed in countless ways. Toddy is particularly rich in energy and vitamin C and has significant amounts of vitamin B and iron (Pargeter et al. 1984, 10-15). Bayliss-Smith's (1982, 62) study of Ontong Java atoll in the Solomon Islands stresses the dietary importance of coconut, which contributed 21 per cent of all food energy directly, as well as the copra, which provided the cash to purchase another 25 per cent of the total food energy consumed. In addition to its critical dietary importance, the coconut palm is used in a myriad of other ways to produce products of economic and cultural importance, the imported substitutes for which would either be too costly or unobtainable for most I-Kiribati.

Pandanus

After the coconut, the pandanus, or te kaina (Pandanus tectorius), is the most important tree of Kiribati agroforestry systems, with almost 200 different recognized cultivars, many of which may be exclusive to a given village or family (Luomala 1953; Overy et al. 1982). Catala (1957, 51) reports, however, that only 16 names were widely recognized on Tarawa.

Because pandanus will grow in very poor or thin soils, it can be found growing almost anywhere on atoll islets. Catala (1957, 52) found that for Tarawa atoll there was an equal density of pandanus, whether it was on the ocean or lagoon side or in the interior, although it grew more successfully where coconut density was lower, particularly in marshy areas or along the lagoon edge, where pandanus seems to have a definite advantage over the coconut. Moul (1957) also found pandanus present in most vegetation associations on Onotoa atoll.

Although natural stands commonly occur in swampy areas, in coastal littoral forests, and in long-neglected bush plantations, the majority of pandanus in garden lands or around villages or residences are planted and owned by individuals (Luomala 1953, 83). Because pandanus propagated from seeds will rarely reproduce desired characteristics, almost all planted pandanus are started from cuttings, ideally cuttings that already bear the beginnings of adventitious roots. At times, new trees will be mulched with leaves of Guettarda speciosa or other plants, and covered with black topsoil. Frequent tamping around young plants, even after they are fully established, is carried out to obtain low, easy-to-harvest high-yielding trees. Given optimum light availability and care, trees near villages can bear as soon as 10 months after planting, whereas they may take up to more than a year in bush gardens (Catala 1957, 53-54).

The fruit of pandanus is a very important part of the I-Kiribati diet; the tree also provides raw material for a wide range of plaited ware, construction materials, medicines, decorations, parcelization, perfumes, and other uses, as well as being the I-Kiribati ancestral tree, from which, according to mythology, the progenitors of the I-Kiribati came (Luomala 1953, 83).

The fleshy parts or drupes of the ripe fruits are consumed raw, as well as being prepared or included in other dishes in a variety of ways. Some of the commonest preparations are te tangauri, te tune, and te karababa. Te tangauri, a paste made from a mixture of a puree of the fresh fruit and grated coconut, can be eaten fresh or dried in the sun. Te tuae is prepared by cooking the fruit, removing most of the fibre, and making a paste, which is then spread on leaves and dried in the sun. The dried paste, which is then cut into pieces for further desiccation, will keep for years, constituting a food reserve that can be used on long voyages or prepared at a later time by softening in and/or prepared with coconut milk or grated coconut. Te karababa is prepared by cooking the drupes, then mashing and mixing them with grated coconut. This mixture is then dried and eaten or is further processed, by toasting and grinding into flour, into te kabubu, which keeps for long periods and which may be eaten straight or as an ingredient in a range of dishes, for example, toddy molasses (Catala 1957, 56-58).

Pandanus leaves are used in the production of thatching, roofing, a range of fine and everyday mats, hats, sails (in the past), cigarette wrappings, food wrappers, caulking material, and baskets for babai compost. The trunk and adventitious roots are used in house and in general construction, with particular cultivars being best for different uses (Catala 1957; Luomala 1953; Overy et al. 1982).

Breadfruit

The next most important cultivated plant is the breadfruit, of which there are two distinct species, the common breadfruit (Artocarpus altilis) and the Mariannas breadfruit (A. mariannensis), plus a hybrid of the two (Fosberg et al. 1979; Fosberg and Sachet 1987; Thaman 1989b). Although well-adapted to the atoll environment, its distribution seems to be directly related to the salinity of groundwater, being planted primarily in villages or their immediate vicinity, and occasionally along roadsides, particularly on the more protected lagoon side of the islands. Moul (1957, 11) reported that it was very common along village streets on Onotoa, and the canopy of one of the most extensive breadfruit groves on Tarawa almost covers the main road through the chiefly village of Eita.

Although pandanus is much more common, breadfruit rivals it in subsistence importance in some areas. Whereas pandanus is an important component of the bush flora, often forming pure stands, breadfruit is rarely found in the heart of the bush (Catala 1957, 61), but forms a major, often dominant, component of the vegetation around villages such as Eita and Betio on Tarawa. In one village on Tarawa, with a population of 115 (22 families), Catala (1957, 64) counted 93 trees, all of which belonged to the person or household that planted them, even if the planter moved to another village. The number of trees per household varied from 0 to 11, with the mode being 4 (Catala 1957, table 10).

Like pandanus, breadfruit are almost always deliberately planted in holes filled with waste, including the dead leaves of coconut and breadfruit and the leaves of te mao (Scaevola sericea), te ren (Tournefortia argentea), te uri (Guettarda speciosa), and te non (Morinda citrifolia), and often topped with black topsoil found under Guettarda speciosa (Catala 1957, 64; Moul 1957; Small 1972). Under favourable conditions, trees may grow over 20 metres high, with trunks almost 2 metres in circumference, although under less favourable conditions, trees may only reach 8-10 metres and 1 metre in circumference (Catala 1957, 64).

Of the major economic plants, the breadfruit seems to have the least resistance to prolonged drought. Sabatier (1939, in Catala 1957, 61) says that breadfruit trees survive with difficulty in the drier southern islands and "are practically exterminated every ten years." Moul (1957, 11) reports that a significant proportion died during the prolonged drought of 1949-50.

The trees generally bear for the nine months from May to January, during which time fruit is often very abundant, being eaten ripe, both raw and cooked, depending on the variety, as well as being preserved by cooking, crushing, and drying as te kabuibui ni ma, or te tune n-te mad. Young leaves of breadfruit are used medicinally to cure ear infections, and buds are used for conjunctivitis; mature leaves provide wrapping for food and serve as fertilizer or compost for babai and other plants; the wood can be made into outrigger canoe hulls and fishing floats (Catala 1957, 65-66).

Other cultivated fruit-trees

Other commonly cultivated fruit-trees are papaya, banana and plantain cultivars, the native fig, or te bero (Ficus tinctoria), the common fig, or te biku (Ficus carica), and the lime (Citrus aurantiifolia). Occasionally lemon trees (Citrus limon) are found, and guava and mango have been introduced but are rare, and, in the case of mango, survive with difficulty.

Papayas are particularly common in villages and, where well cared for and mulched, are healthy and produce good fruit that is eaten raw when ripe, especially by children, and cooked green with coconut milk. In a village of 23 households, with a population of 115, Catala (1957, table 10) found 111 papayas, amounting to just under one tree per person.

Banana and plantain cultivars are occasional to common on Tarawa and Abemama and much more common on wetter islands in north Kiribati, such as Makin and Butaritari, and in southern Tuvalu, where the main island of Funafuti (meaning the "place of the banana") is renowned for its extensive banana plantings. Bananas are commonly grown around houses in villages and occasionally planted in abandoned banai pits or in specially dug banana pits, a common practice at mission settlements and boarding schools. When grown in pits, bananas and plantains are normally not planted in flooded soil, as is babai, but in slightly higher parts of pits or in pits that have been partially filled.

Where well looked after, bananas grow well and are a favoured staple. If grown as a "pit plantation," it is usually necessary to dig a rather deep trench around the pit to keep coconut roots out. A layer of dark soil collected from under Guettarda and Scaevola is added along with composting and rusted tin cans to provide iron (Catala 1957; Small 1972). Although there seems to be considerable scope for an expansion of banana and plantain pit cultivation, the taro beetle (Papuana sp.) may have caused widespread damage to plants grown in pit plantations. When grown in villages, close to the lagoon side, mulching with organic material, coconut husks in particular, results in good yields.

The native fig, or te bero (Ficus tinctoria), is commonly cultivated around villages and occasionally in plantation areas. Moul (1957, 12) reported it as common around abandoned babai pits and present in small thickets in rich soils around Pisonia grandis groves on Onotoa atoll in southern Kiribati. It is propagated vegetatively by planting branch cuttings, and its fruits are an important staple in the drier southern islands. The fruits are picked when ripe and sometimes when green, cooked, crushed in a mortar into a puree that can be eaten after being sweetened with toddy molasses (kamaimai) or sugar and grated coconut or preserved by drying in the sun on Guettarda speciosa leaves. Fig trees reportedly bear many times throughout the year, and have wide cultural utility (Catala 1957; Luomala 1953; Small 1972).

The common fig, reportedly introduced by missionaries, seems to be very well adapted to the atoll environment and is occasionally found propagated by cuttings in village home gardens.

The lime (Citrus aurantiifolia) is by far the most common citrus fruit grown in Kiribati, but is found only occasionally in villages. The fruit is highly sought after for squeezing on fish and other foods and for making drinks. Lemon trees are present on the agricultural experiment farm at Bikenibeu, but are rare elsewhere.

Other cultivated food plants

Other cultivated but minor tree-like food plants include sugar cane (Saccharum officinarum) and hibiscus spinach (Hibiscus manihot), both of which are found occasionally in villages around homes. Although sugar cane grows poorly in some localities, it shows promise where well-mulched. Hibiscus spinach, a very nutritious green vegetable, reportedly introduced by contract workers returning from the phosphate mines on Banaba (Ocean Island), grows particularly well and shows little or no evidence of the insect or disease damage so characteristic in Fiji.

Giant swamp taro

The major understorey food plant in Kiribati is the ceremonially important staple, giant swamp taro, or te babai (Cyrtosperma chamissonis), which is cultivated in pits excavated to the freshwater lens, and is heavily mulched. About 20 named cultivars are recognized (Small 1982). Te babai is not a woody species, but because of the very sophisticated system of mulching and fertilization with leaves from numerous tree species, and because the babai pits are found within a matrix of coconut palms and other trees, te babai is an integral component of the Kiribati agroforestry system.

Te babai was probably more extensively cultivated in the past, as evidenced by the numerous abandoned pits, some of which have been overrun with coconut seedlings and weeds. Although pits are often abandoned as the water becomes more brackish, many were obviously abandoned long ago, with most pits being ancient, the inhabitants having no recollection of their origin (Catala 1957, 68). On Abemama, for example, Watters and Banibati (1977,37) found, in a survey of 16 households, that whereas the mean number of pits in use was only 4.2, the mean number of empty pits per household was 23.4, with only 7.7 still containing the water necessary to produce te babai. Moreover, few of the productive pits were fully stocked, thus "reflecting more basically the changing food preferences and habits and growing reliance on the cash component of a household's total income" (Watters and Banibati 1977, 38). On Onotoa, Moul (1957, 5) found that as many as 10 individuals had separate plots in single pits ranging from 25 to 30 feet long and from 10 to 20 feet wide.

As a result, te babai has become almost a luxury in many areas, reserved almost exclusively for ceremonial purposes, rather than constituting a staple food (Catala 1957, 67). Nevertheless, te babai cultivation continues to be surrounded with tradition, and there has been some recent rehabilitation of abandoned pits on both Tarawa and Abemama. As Catala (1957, 67) relates: "pulling up a babai in order to offer the tuber to a distinguished guest is considered the greatest honour that can be paid to him."

Te babai pits must be excavated through as much as 1.5 metres of hard conglomerate and limestone to reach the freshwater lens, with Moul (1957, 5) reporting pits up to 15 feet deep. The young shoots are planted in holes about 0.3 metres (2 feet) deep in the bottom of the pit and mulched and fertilized with black topsoil from stands of Guettarda speciosa, Scaevola sericea, and other plants and a variety of leaves, some of which are specially prepared for the purpose, using techniques generally not divulged. Baskets of pandanus or coconut leaves are commonly made, into which the shoot is planted or in which the fertilizer or mulch is administered to the plants in the pit.

Leaves used for fertilization and mulching, in order of importance, are te kaura (Sida fallax), te uri (Guettarda speciosa), te ren (Tournefortia argentea), te mad, or breadfruit (Artocarpus spp.), te woo (Boerhavia repens) and, to a lesser extent, species such as te kaura ni Banaba (Wollastonia biflora), te kanawa (Cordia subcordata), te kiaou (Triumfetta procumbens), and te kiaiai or te rao (Hibiscus tiliaceus). These leaves, with the exception of Sida fallax, are mixed with other plant waste, particularly old pandanus leaves and coconut refuse, black topsoil, and occasionally ground pumice (te uuan), and applied green or dried to the basket surrounding the plant or placed in the pit near the plant. Because the leaves of a variety of trees and the black topsoil found under trees are very important in te babai cultivation, increasing agrodeforestation may be, at least in part, responsible for the decline in its cultivation in Kiribati.

Cultivated exotic timber trees

Two trees deliberately introduced for reforestation purposes are Casuarina equisetifolia and Leucaena leucocephala. Casuarina, in particular, which was rare in the 1950s, has been widely planted on Tarawa as part of government-sponsored reforestation programmes to provide wind-breaks for recently planted coconut palms on the ocean sides of atoll islets (Overy et al. 1982, 14) and to provide firewood. Leucaena was also introduced for reforestation purposes, because of its nitrogen-fixing ability, but is not widely planted.

Cultivated ornamentals

Commonly cultivated ornamentals, most of which are found in home gardens, mission settlements, school grounds, or in major settlements, include plumeria, or frangipani (Plumeria rubra and P. obtusa), hedge panax (Polyscias guilfoylei and P. fruticosa), copperleaf, Jacob's coat, or the beefsteak plant (Acalypha amentacea), false eranthemum (Pseuderanthemum carruthersii), golden bells (Tecoma stans), bougainvillea (Bougainvillea spp.), Lantana camara, Hibiscus rosa-sinensis, dracaena (Dracaena fragrans), ixora (Ixora casei), and the poinciana, or flame tree (Delonix regia). Also present in home gardens, but not common, are the Tahitian gardenia (Gardenia taitensis), Acacia farnesiana, Cordyline fruticosa, and the Pacific fan palm, Pritchardia pacifica. With the possible exceptions of Dracaena fragrans and Pritchardia pacifica, all of these plants constitute important sources of flowers and leaves, which are used - along with flowers from native species such as Guettarda speciosa, Sida fallax, and Scaevola sericea - in the ubiquitous leis and head garlands so important for all social and ceremonial occasions.

Important indigenous species

Important indigenous trees or tree-like species, which are integral and widespread components of the Kiribati agroforestry system, include Scaevola sericea, Guettarda speciosa, Tournefortia argentea, Sida fallax, Morinda citrifolia, Clerodendrum inerme, Premna serratifolia, Pemphis acidula, and Dodonaea viscosa. Other indigenous trees, which are uncommon to rare in agricultural areas, but are sometimes found in coastal strand forest, home gardens, and villages, and as street trees in the main settlements, include Calophyllum inophyllum, Cordia subcordata, Terminalia catappa, Pisonia grandis, Hibiscus tiliaceus, Terminalia samoensis, Barringtonia asiatica, Hernand ia nymphaeifolia, Macaranga carolinensis, and Thespesia populnea. Also of localized importance are the mangrove species Rhizophora mucronata, Bruguiera gymnorhiza, and Lumnitzera littorea. All of these species have important cultural uses, some of which are described here or in the Appendix. The information is based on field surveys and on Luomala (1953), Catala (1957), Moul (1957), and Overy et al. (1982).

Scaevola sericea is the commonest understorey shrub in Kiribati (Moul 1957, 22). It is found everywhere in coastal littoral forest and forms dense thickets. It is common in plantations, especially where coconut density is low, and occasional in home gardens and in villages and other habitats. It is an important component of the coastal strand vegetation, which provides protection from salt spray to inland plantations and gardens, is an important producer of humus and organic material because of its abundance, and has wide-ranging cultural utility.

Guettarda speciosa, Tournefortia argentea, and Sida fallax are the most common sources of leaf compost for the cultivation of babai (Cyrtosperma chamissonis). Guettarda speciosa, one of the main components of the atoll vegetation, is occasionally cultivated in village gardens and is particularly common in the centre of islets, where it is important in the formation of the black topsoil mixed with leaf compost used in planting babai, pandanus trees, and other crops. Guettarda's wood is used in general construction, its leaves in the production of garlands and head wreaths and when spreading pastes or preserves for sun-drying. The plant is prominent in I-Kiribati legends, mythology, and is associated with phases of the moon and stations of the sun - all of which uses together make it one of the most culturally important plants in Kiribati.

Tournefortia argentea is commonly found scattered in groups in plantation areas, occasionally in strips of ocean or lagoon strand forest, and was reported by Moul (1957, 20) to be very common on the edges of babai pits on Onotoa. Like Guettarda speciosa, it has wide cultural utility. Its wood was occasionally used as a substitute for Calophyllum inophyllum for canoe bows and Y-shaped pieces as spar supports on outrigger canoes. It also provides a favoured fuel, and was used as the bottom piece in making fire by friction in the past. The leaves are reportedly eaten in salads by boat crews, used medicinally to reduce fever, and as a female deodorant, and also in magic and for scenting coconut oil, as well as being an important ingredient in compost or fertilizer for babai and other plants. Te ren also features in many IKiribati legends.

Sida fallax, a small shrub found scattered throughout plantations, is occasional in villages, and common on lagoon sides and on the inner margins of coastal ramparts of islands. It is a favoured species for personal ornamentation and magic, particularly love magic, and is used medicinally. Its flowers and leaves are shredded and dried to produce the "strongest" compost or fertilizer for babai. Also occasional in plantation areas and cultivated as a living hedge or ornamental in home gardens is Clerodendrum inerme. It is reportedly used medicinally and its flowers are used in garlands.

Pemphis acidula is very common on sandy areas inland from mangroves and in clusters in garden areas bordering the ocean coast and on beach ramparts, where it often forms almost pure stands and serves as protection against sea spray. It is important medicinally, and the dense, extremely hard, wood has wide utility because of its resistance to sea water, and is a favoured firewood. Dodonaea viscosa, indigenous to many Pacific islands, but possibly a recent introduction to Kiribati, locally common near existing villages and in sites of former dwellings and occasional in garden areas, also has a variety of uses.

Morinda citrifolia and Premna serratifolia, two of the most important medicinal and magical plants in Kiribati, are occasional in coastal areas and relatively common in bush gardens and home gardens in villages. The pungent ripe fruits of M. citrifolia are occasionally eaten after boiling by old people, as a famine food, and as a stimulant on long fishing trips or ocean voyages, and the consumption of the young leaves has been actively promoted recently as a rich source of vitamin-A to combat outbreaks of vitamin A-deficiency night blindness among children.

Other indigenous species of wide cultural utility occasionally present in the coastal strand forest bordering garden areas, in home gardens, or in settlement areas include Calophyllum inophyllum, Cordia subcordata, Terminalia catappa, Pisonia grandis, Hibiscus tiliaceus, Terminalia samoensis, Barringtonia asiatica, Hernandia nymphaeifolia, Macaranga carolinensis, and Thespesia populnea. All of these culturally useful species were more widespread in the past before official government emphasis was placed on clearing indigenous species to extend and rehabilitate coconut plantations and before current high population densities placed such pressure on limited arboreal resources.

Cordia subcordata is occasional in coastal forests and in villages, its attractive wood being highly valued for woodwork, and the inner bark, leaves, and attractive orange flowers highly valued for medicine, magic, composts, and garlands. Terminalia catappa and the related species, Terminalia samoensis, both useful trees, are occasional in villages and in tree groves in plantations and inland from coastal littoral forest, almost always as individual trees, and sometimes planted as ornamentals.

Calophyllum inophyllum, so important medicinally and for general construction, canoe building, and woodworking, is occasional around villages and towns, and was a sacred tree in the past on Tabiteuea.

Pisonia grandis, the favoured nesting tree for the black noddy, an important food resource, is uncommon to occasional as isolated individuals or small groups, and has been recently planted in villages and at the hospital in Bikenibeu for its edible leaves, which are rich in vitamin A. It was probably more common in the past as a dominant in the indigenous forest. There reportedly remains a large traditional

Pisonia reserve on the island of Onotoa in south Kiribati, which is surrounded by extensive guano deposits and the most luxuriant vegetation seen on the atoll (Moul 1957, 4).

Mangroves provide a habitat and serve as an important food supply for many of the important edible fish species. They also have an important role in coastal stability, land reclamation, and the protection of gardens from salt-water spray at the interface between the lagoon and agricultural areas. On Onotoa, they reportedly encircle fish-ponds (Moul 1957, 5). Mangroves are also used in construction and in the production of medicines, dyes, and garlands. They can, consequently, be considered integral components of agroforestry systems, particularly in land-scarce areas such as Kiribati. Rhizophora mucronata is the most common species, forming very dense stands on swampy lagoon shores as well as being found on the windward ocean coast at Bairiki, Tarawa. Bruguiera gymnorhiza is common to occasional, and Lumnitzera littorea, although rare on Tarawa and possibly absent on Abemama, is reportedly more common on Butaritari.

(introduction...)

Home-garden urban agroforestry
Urban agroforestry on undeveloped land
Problems of urban agroforestry
Integrating agroforestry into urban planning and policy

With the growth of towns and cities in the Pacific and the resultant increasingly large and dense concentrations of people isolated to varying degrees from rural production, urban agroforestry has taken on ever greater significance, although until recently it has received little official recognition or encouragement. Two main types of Pacific Island urban agroforestry are categorized here as: home-garden agroforestry adjacent to residences; and agroforestry on idle or undeveloped land within urban areas, but usually at a distance from the residence.

Home-garden urban agroforestry

Urban agroforestry in home gardens is today a ubiquitous feature of urban landscapes in the Pacific Islands. Even in areas not renowned for agricultural diversity, such as Kiribati and Nauru, urban gardens contain a wide range of food trees, non-tree staple and supplementary food plants, and non-food plants (table 8). Cultivation or protection of trees and non-tree plants on idle or undeveloped land in urban areas is also very widespread and provides an important source of other produce, including limited commercial production (Thaman 1977a; 1977b; 1984a; 1985a; 1987a; 1988e).

As suggested by table 8, more than 20 species of trees are common in the "food trees" category in urban agroforestry gardens. Also common are all the important staple root crops and a great range of supplementary non-tree food plants such as onions, amaranths, pineapple, peanuts, cabbages, a wide variety of legumes and "spinaches," cocurbits, okra, tomatoes, passion-fruit, sugar cane, eggplant, and corn. There are also condiments and spices such as chillies, ginger, coriander, and mint as well as plants producing beverages, stimulants, and depressants (betel-nut, betel pepper, kava [Piper methysticum], tobacco, and lemon grass). Many of the home gardens surveyed contained a large proportion of these plants.

Table 8 Number of species and distinct varieties found in surveys of urban agroforestry systems in Port Moresby in Papua New Guinea; Suva in Fiji; Nuku'alofa in Tonga; South Tarawa in Kiribati; Nauru; and Location, the contract-worker settlement in Nauru

Crop types Papua New Guinea Fiji Tonga Kiribati Nauru Location
Food trees 30* 39 27 20 14* 16
Non-tree staples 7 10 8 6 5 8
Non-tree supplementary foods 48 65 44 35 14 41
Total 85 114 79 61 33 65

*The totals for Papua New Guinea and Nauru would have been slightly higher for food trees if Musa clones and Citrus spp., respectively, had been differentiated.

As stressed by Soemarwoto et al. (1985, 44), in their study of Javanese home gardens, true plant diversity is far greater than indicated by the presence of many different species since many species are represented by numerous cultivars. In Tonga, for example, there are numerous distinct, named breadfruit cultivars. And in tree gardens in settlements in Yap, as already mentioned in the section on the Micronesian high islands, there are 21 named coconut cultivars, 28 breadfruit cultivars, and 37 banana cultivars (Falanruw 1985, 16). There is similarly great cultivar diversity among other crops such as mangoes, domesticated Pandanus, papayas, and especially among the traditional staple ground crops such as yams, taros, and sweet potatoes. This intra-species diversity adds economic, ecological, and nutritional stability to urban agroforestry systems.

"Weed" species are also myriad in urban agroforestry gardens, but as Soemarwoto et al. (1985,44) caution, "the term 'weed' should be used with extreme care" because of the many uses home gardeners have for weeds - as medicines, fodder, mulch, roofing, fish poisons, toothbrushes, and vegetables. "Weeds" such as Amaranthus spp., black nightshade (Solanum nigrum), purslane (Portulaca oleracea), and water spinach (Ipomoea aquatica), for example, are important pot herbs in Fiji and are often sold at the Suva Municipal Market (Thaman 1976/77), and almost all grass species are used for fodder.

Food trees

Although staple ground crops are most numerous, food trees such as coconut, breadfruit, papaya, Citrus spp., mango, Musa clones, guava, Annona and Syzygium spp., avocado, Tahitian chestnut (Inocarpus fagifer), hog plum, or vi-apple (Spondias dulcis), oceanic lychee (Pometia pinnata), and Terminalia spp. are the dominant plants of most urban landscapes, especially in long-settled areas. An exception would be Location, the indentured worker settlement in Nauru, where little or no space for gardening exists, let alone tree cropping, and where a high proportion of plants are grown in artificial boxed beds or containers. The only mature trees are bananas and scattered coconut palms.

Trees constitute a particularly important economic and nutritional resource on low-lying islands such as the atolls of Kiribati, where apart from giant swamp taro (Cyrtosperma chamissonis), which is generally reserved for special occasions, the main staples are all tree crops: coconut, breadfruit, Musa clones, pandanus, and the native fig (Ficus tinctoria). Trees are, nevertheless, very important in home gardens on high islands as well. For instance, as described in chapter 6, the tree gardens around homesteads in Yap contain some 50 species of introduced or indigenous food trees (Falanruw 1985, 15-16). Trees of particular importance to the Indian population of Fiji (who currently make up close to half of Fiji's population) include jakfruit (Artocarpus heterophyllus), horseradish, or drumstick tree (Moringa oleifera), curry leaf, or Indian bay (Murraya koenigii), and tamarind (Tamarindus indicus).

Staple crops

The most common staples include the important root crops, such as cassava, taro, sweet potato, and tannia (Xanthosoma spp.), with giant taro (Alocasia macrorrhiza) important in Tonga, and giant swamp taro (Cyrtosperma chamissonis) particularly important in the harsh environment of the lowlying atolls of Kiribati, as discussed in the previous chapter (Thaman 1984b).

True taro (Colocasia esculenta) is particularly well suited to urban conditions because it can be grown on small plots, either as a staple for its corms or for its leaves, which constitute the most common leafy vegetable, or "spinach," in many areas. It is often found planted in wet conditions along drains, near water taps, or washing areas (Thaman 1977a; 1977b).

Cassava is commonly planted in back and front home gardens in Suva and Port Moresby and along road frontages in Suva, and sweet potatoes commonly occupy large proportions of back and front home gardens in Port Moresby, as well as being found in small plots around homes in densely-settled urban Tarawa. Tannia (Xanthosoma) is also of increasing importance as it seems to be disease-resistant, relatively drought-resistant, and grows well in the shady conditions commonly encountered in older urban areas where mature trees dominate the environment.

Supplementary food crops

Supplementary crops, such as hibiscus spinach (Hibiscus manihot), Amaranthus spp., pineapple, cabbages, chillies, taro (grown especially for its leaves), cucurbits, tomatoes, sugar cane, and a wide variety of edible legumes, are also very common in urban gardens throughout the Pacific. These, plus a wide range of other supplementary, often shortterm crops, constitute a valuable nutritional and economic resource.

Non-food plants

In addition to plants for food, drink, or supplementary consumption, many useful non-food plants are also found in home gardens. These include handicraft plants such as Pandanus spp., used in plaited ware, paper mulberry (Broussonetia papyrifera), used for bark or tape cloth, and annatto (Bixa orellana) and Bischofia javanica (both sources of dyes); Leucaena leucocephala, an important renewable fuel-wood resource; a great range of medicinal plants, many of which are trees; plus many other plants with a variety of uses.

Medicinal plants have long been of significance and remain a resource, given the high cost of imported medicines, not to mention their frequent unavailability, misuse, and, in some cases, doubtful efficacy. In Fiji, where extensive areas of tropical rain forest, coastal strand, and mangrove forest still exist, 40 per cent (73) of 183 plant species reportedly used medicinally by the indigenous Fijians (Weiner 1984) are found in home gardens in a cultivated, protected, or weedy state. On the smaller, more densely-populated islands of Tonga and Kiribati, where less natural vegetation remains, approximately 75 per cent of all reported medicinal plants (56 of 77 and 33 of 44 respectively) are found cultivated or protected in home gardens (Luomala 1953; Thaman 1976; Weiner 1971). In Nauru, where over 70 years of open-cast phosphate mining and widespread bombing during World War II have devastated most of the natural and much of the traditional cultural vegetation (Manner et al. 1984; 1985), 28 (85 per cent) of 33 reported medicinal plants are now found in Nauruan home gardens.

Of the 93 medicinal plant species found in urban gardens in these four countries, 51 (55 per cent) were trees and another 10 were woody shrubs. It should be noted that the totals for medicinal plants in Fijian urban gardens would undoubtedly be much higher if data on Indian medicinal plants were also available.

The importance of sacred or perfumed plants to urban agroforestry is also considerable. Of some 49 plant species considered by Tongans to be sacred ('akau kakala), 36 were found present in a survey of home gardens in the capital of Nuku'alofa. Twenty-three of the 36 are trees and five others are woody shrubs. In addition to their sacredness, such plants constitute a very significant economic resource. Their flowers, leaves, fruits, and bark are used in leis and ornamentation for the expanding tourist industry, as well as being the main scents used in body oil (coconut oil), perfumes, and deodorants, the imported substitutes for which are extremely expensive and often not as culturally acceptable. Many of these plants, such as heilala (Garcinia sessilis), langakali (Aglaia saltatorum), feta'u (Calophyllum inophyllum), and sandalwood (Santalum yasi), are chiefly kakala, kakala hingoa, which are reserved for nobility or very special occasions and constitute a very important and sacred cultural resource (Thaman 1987a).

Evidence from Fiji, Kiribati, and Nauru indicates that sacred and perfumed plants are of similar importance there. In Kiribati, for example, where cash incomes are very low, headbands and garlands (primarily from flowers grown in home gardens) are widely sold (often by school children to help their families) to people to wear during work or for dances and other festivities.

Similar analyses of urban agroforestry systems for other plants that yield firewood, dyes, livestock feed, insect repellents, handicrafts, fish poison, etc., would yield equally impressive lists of plants and uses.

Characteristics of urban agroforestry gardens

The commonest plants in the inventory of Pacific urban agroforestry tend to be natives or pre-European introductions, except where the gardeners are from immigrant populations. For example, the Indian population of Fiji prefers species such as eggplant, okra, Amaranthus spp., a wide range of pulses and cucurbits, and tree crops, such as jakfruit, tamarind, mango, Citrus spp., curry leaf (Murraya koenigii), Sebesten plum (Cordia dichotoma), horseradish, or drumstick, tree (Moringa oleifera), and the spiritually and medicinally important neem (Azadirachta indica).

Also of importance, however, is a range of more recently introduced crops, including temperate vegetables, pineapple, papaya, avocado, guava, and improved citrus varieties and banana clones, as well as cassava, which is a ubiquitous staple in most Pacific Island towns (Thaman and Thomas 1985). In fact, Pacific home gardens seem to have been, and will probably continue to be, one of the most effective avenues for the introduction and acceptance of new plant species.

A great diversity exists in the actual area under food crops and in their spatial distribution. Whereas some households have only a few scattered fruittrees and vegetables, many cultivate food crops on over 50 per cent of their allotments. In Port Moresby, for example, in low-income areas settled by migrants, such as Morata and Gerehu, an average of approximately 40 per cent of 450 sq m allotments were under food crops. Similarly, in some cases in Nuku'alofa, up to 75 per cent of 500-1,000 sq m allotments were under food cultivation, mainly root crops, such as taro, tannia, and cassava, amongst Musa spp. and scattered trees. Trees become increasingly dominant in longsettled areas, as cash incomes increase, soils decline in fertility, and tree seedlings mature and increasingly shade garden areas.

Ornamentals are commonly planted closest to the home, often in the front, whereas medicinal plants, sacred or fragrant plants, and other culturally valuable, commonly multi-purpose plants, are scattered amongst the food plants.

In the atoll environment of South Tarawa, Kiribati, where the cal careous soils are poor and thin and population densities are very high, traditional staple tree crops, such as coconut, breadfruit, pandanus, the native fig (Ficus tinctoria), and, in some cases, Musa spp., are predominant. In the gardens of the indigenous Nauruans (who, as a result of phosphate mining royalties, have among the highest per capita incomes in the world), ornamental, fragrant, and medicinal plants dominate, along with the ubiquitous coconut, edible pandanus, some Musa spp., and breadfruit.

At the Location contract-worker settlement, where people live in multistorey tenements, and where family gardening is limited to no more than 1530 sq m, most families have only a few plants. In the case of the gardens of Tuvaluans (nationals of Tuvalu, formerly known as the Ellice Islands) and IKiribati, they often constitute juvenile tree seedlings, staple root crops, or a single coconut or stand of bananas.

In the case of gardens made by Chinese (mostly recruited from Hong Kong) and Filipinos, the emphasis is on intensive vegetable gardening, often in containers, reflecting a more intensive system than that practiced by most indigenous Pacific peoples. In Kiribati and Tonga, however, recent emphasis has been placed by the governments and non-governmental organizations on more intensive types of gardening, such as hydroponic techniques in Kiribati because of the highly calcareous and sandy soils there. In Kiribati, where vitamin-A deficiency-induced night blindness and xerophthalmia have become problems, the planting and consumption of the vitamin-rich leaves of two native tree species, Morinda citrifolia and Pisonia grandis, have been encouraged in urban areas.

Urban agroforestry on undeveloped land

Undeveloped or idle lands in urban and pert-urban areas are important sources of food and tree products such as timber, fence posts, fuel wood, medicines, leaves, flowers, fruits, and nuts. Such areas include road frontages, empty allotments, river banks and valleys, right-of-ways for proposed or existing paths and roads, and open land such as hillsides and swamp land.

In Port Moresby, over one-third of all households had gardens on idle land in addition to their home gardens. Kilakila villagers, as original inhabitants of the area, had particularly large tracts of undeveloped urban savanna land, and all households had, in addition to their home gardens, from one to four "bush" gardens averaging 1,135 sq m located on urban land within two miles of the urban village of Kilakila.

In Suva, about 20 per cent of all households cultivate "unused" open land, and it has been estimated that on the Suva Peninsula, approximately 5 sq km (over 70 per cent) of the "undeveloped" area (which does not include swamp or mangrove) was under this type of cultivation. The practice was most common in areas where there is a high proportion of Crown and leasehold land (as opposed to freehold land) and a high proportion of Fijian residents. Some 20 per cent of all households also planted along road frontages, despite Suva City Council regulations forbidding the practice.

In Tonga, Kiribati, and Nauru, there is little undeveloped "urban" land, although, in a number of cases, Tongans planted entire adjacent unoccupied "town allotments" ('apt kolo) in sweet potato, taro, tannia, and a mixture of trees, or in traditional mixed yam gardens (ma'ala 'up), where yams, giant taro, plantains, and taro are intercropped, usually under coconuts and other trees. There is virtually no open land in urban Kiribati, but in Nauru, some Chinese, Tuvaluan, and I-Kiribati contract labourers plant food gardens near the Nauru Phosphate Corporation's "topside" workshops, on the phosphate-rich central plateau, and in the swampy area surrounding land-locked Buada Lagoon. In Tuvaluan and I-Kiribati gardens, coconuts and banana clones were dominant.

Along road frontages, fruit-trees such as mangoes and coconuts are common, but ornamental and shade trees such as Plumeria spp., flamboyant (Delonix regia), Cassia spp., monkey-pod (Samanea saman), banyans (Ficus spp.), variegated coral tree (Erythrina variegata var. variegata), and the pride of India (Lagerstroemia speciosa), many of which are systematically planted by city councils or the government, as well as by individual households, are dominant. Living fences of fruittrees and other useful species, such as Polyscias spp., Leucaena leucocephala, Erythrina variegata, Hibiscus tiliaceus, Hibiscus rosasinensis, guava, and the recently introduced madre de cacoo (Gliricidia septum), are harvested, pruned, pollarded, or "grazed" and constitute important sources of food, fodder, firewood, medicines, and flowers, as well as being of considerable ecological importance. The balance of agroforested landscapes in urban areas includes the rare botanical garden or urban forest reserve, public parks, and institutional tree planting throughout cities, often as part of landscaping schemes.

Despite the current importance of agroforestry on undeveloped urban and pert-urban land, it is these areas that are most severely affected by wanton deforestation because of insecure tenure and undefined ownership. It is a classic example of Hardin's "Tragedy of the Commons" (1968), where the urban poor and entrepreneurs clear trees in order to plant crops and to glean the land of scarce fuelwood resources. As Eckholm argues (1976, 101), the "real energy crisis" is the daily scramble by the world's poor to find the wood they need to cook their dinner. This scramble for wood and associated deforestation is clearly visible and rapidly increasing in Pacific Island urban areas (Thaman and Ba 1979).

Animal husbandry and urban agrosilvipastoralism

Small-scale animal husbandry, although playing a minor role compared with plants, is also an important activity. Surveys in Port Moresby found animal-keeping to be minimal, with 11 of 79 households keeping pigs, chickens, or ducks, and a few households keeping tethered cows or goats. There were no pigs kept in Suva. In Tonga over half of all sample households kept tethered or penned pigs, and almost two-thirds kept chickens or ducks. In most cases, poultry were penned or tethered at night and allowed to forage during the day, and pigs and other larger animals were generally tethered or penned at all times. In Kiribati and Nauru, pigs and chickens are also kept on home allotments. In Nauru, there was a large communal pig-rearing area along the beach in Denigomodu District, and, in Betio, the most heavilypopulated area of South Tarawa, there was a large communal pigrearing area with individualized pens, established by the local town council, under coconuts, breadfruit, and other trees.

In terms of agrosilvipastoralism within the wider context of urban areas, livestock depend on trees to a great extent for shade, sustenance, and tethering. Apart from kitchen waste, the main feed for pigs and chickens in most areas is coconut. In Tonga, goats and pigs are commonly fed the leaves of Leucaena leucocephala, Pisonia grandis, and Erythrina variegate, while "living edible pens" for poultry and pigs are made of these same species, plus others such as Hibiscus tiliaceus and Polyscias spp., all of which are easily pruned or pollarded to provide fodder. On open land, horses, cattle, and goats are commonly tethered to trees, which also give them shade. Small animal pens, which are commonly constructed of coconut logs, bamboo, Leucaena, or other local timber, are found occasionally.

On the detrimental side, grazing animals and pigs seem to accelerate deforestation in urban areas through the consumption or destruction of tree seedlings and saplings. Once established, however, trees and animals co-exist well, except where goats eat the bark of trees. Cattle seem to enhance the establishment and spread of guava, which although an important fruit, medicinal, and fuel-wood source - has become a noxious pasture weed in many areas.

Problems of urban agroforestry

Urban agroforesters in the Pacific face a number of problems. Unfavourable climate, poor soils, high cost or unavailability of land and water, insufficient time and labour, theft, and lack of government assistance were most commonly mentioned.

Problems of drought are severe in Port Moresby, and include high cost of water, distance of community taps, and water cancellations in Morata, and fear of City Council regulations against the use of water for gardening purposes between 8.00 a.m. and 6.00 p.m. Restrictions on the use of water in gardens are also imposed during periods of extended drought in Fiji.

Urban gardeners commonly have to contend with poor, infertile soils, such as the very poorly developed rocky or stony lithosols of Port Moresby, the shallow soils that overlay a marl substrate in Suva, hydromorphic soils in low-lying areas, and the notoriously infertile calcimorphic soils of Kiribati. Continual cropping on small urban plots also leads to declining fertility and loss of soil structure, unless ameliorative measures are taken. Both water shortage and poor soils, however, often make trees a more attractive proposition than shortterm ground crops, which require water and higher soil fertility.

Insufficient land and insecurity of tenure were problems in most areas, with over half of all households in Suva mentioning land shortage as a problem. Insecurity of tenure, especially in Suva, where a number of people had short-term leases or were squatters, is a major problem and a strong disincentive to urban agroforestry. City Council regulations, although not strictly upheld, were also considered a disincentive that discouraged cultivation of ground crops and trees along road frontages and the keeping of pigs, goats, cows, and horses within the city limits. Other problems included disease, insects, birds, rats, dogs, mongooses, and noxious weeds; theft of produce, especially of banana bunches and tree fruit (approximately one-third of all households had experienced theft); insufficient time; high costs of poultry feed and fertilizer; predation of firewood and deforestation on undeveloped urban and pert-urban lands, where most low-income families still depend on firewood to cook their meals (Thaman and Ba 1979); boundary problems with respect to ownership of crops; and neighbours' unfavourable response to gardening or livestock rearing.

In Kiribati and Nauru, where constraints to expanded home gardening are the greatest, the most significant problems are extremely poor soils, limited water availability, and extremely high population densities, especially in South Tarawa and at Location, the contract-worker settlement. Among the indigenous Nauruans, who are considered to be wholly urbanized, extremely high per capita incomes from phosphate royalties and a resulting overdependence on imported foods act as disincentives to expanded urban agroforestry.

Integrating agroforestry into urban planning and policy

The significance of urban agroforestry is not clearly understood by most planners and policy makers in the Pacific Islands because of a lack of quantitative data on its nature, extent, and cultural and economic value. However, interest has been shown by some city planners and administrators. In Port Moresby, for example, the Housing Commission conducted a survey of urban gardening in the early 1970s. The Committee on Food Supplies of the Solomon Islands (1974) conducted studies of the production of major staple crops (primarily sweet potato) in Honiara and stressed the need to increase production per head in both rural and urban areas, and Fitzroy (1981) pointed out the correlation between vitamin deficiency in '`urbanized" people and the absence of garden plots in Honiara. Further studies stressing the importance of urban home gardens in the Pacific have been conducted since the mid-1970s (All 1976; Basha et al. 1974; Harris 1977; Kesavan 1979; Thaman 1977a; 1977b; 1984a; 1985a; 1987a; 1988e; Vasey 1985; von Fleckenstein 1978).

There have been campaigns encouraging the cultivation of food crops in Port Moresby, and, in Fiji, the National Food and Nutrition Committee (NFNC) and The Fiji Times, through their "Feed Fiji First" campaign, have placed major emphasis on home food production and have sponsored competitions in H.A.R.T. (Housing Assistance and Relief Trust) destitute areas, schools, government housing areas, and agricultural resettlement schemes. Major emphasis was placed on the planting and maintenance of food trees in these competitions.

Similar interest in urban agroforestry has recently been shown in Vanuatu, Tonga, Kiribati, the Federated States of Micronesia, and the Marshall Islands, where urban food dependency and increasing incidences of nutritional disorders have become serious. These countries, along with Papua New Guinea, the Solomon Islands, Fiji, and Western Samoa, are now part of a Unicef-sponsored "Pacific Island Regional Family Food Projection and Nutrition Project" (Sommers 1990). In a course entitled "Agriculture, Food and Nutrition in the Developing World" offered at the University of the South Pacific (a regional university located in Suva but serving 11 Pacific Island countries), a major component of the course involves students in the development and maintenance of mixed home gardens; Tonga and Fiji have both promoted tree planting in towns as integral parts of their World Environment Week and Arbor Day programmes, respectively; and, most recently, the Honiara Town Council has actively and successfully promoted home food gardening through its Sup Sup Gaden (soup soup garden) Club.

8 Agroforestry on smallholder sugar-cane farms in Fiji

Official neglect of traditional polycultural agroforestry systems can be seen as the opposite side of the coin of official emphasis on and encouragement of commercial monocropping, commercial production of livestock, and industrial forestry. And yet, as surveys of smallholder sugar-cane farms in Fiji demonstrate, small-scale commercial operations can maintain a diversity of useful trees in a landscape primarily dedicated to monocropping.

As a legacy of over a century of sugar-cane cultivation and emphasis on cash cropping for export, much of the drier western and northern sides of Fiji's two main islands, Viti Levu and Vanua Levu, are landscapes of smallholder cane farms (typically 4 ha in size), farmed mostly by the descendants of indentured workers from India. Homes are usually located on farms so that settlement is dispersed in contrast to the nucleated villages of the indigenous Fijians.

Although production has long been strongly focused on sugar cane (and to a considerably weaker degree on annual subsistence crops such as rice, pulses, maize, and a variety of vegetables), Indian farmers have traditionally planted or encouraged a wide variety of trees around their houses, as well as on grazing lands, river banks, and along roads and boundaries. Unfortunately, because of high world sugar market prices and a deliberate policy of the Fiji Government to increase sugar-cane production (Fiji's number one source of foreign exchange) in the mid-1970s and early 1980s, farmers extended their cane planting onto grazing lands, to the edges of rivers and drainage ditches, secondary forest stands, and areas formerly reserved for rice and other crops. The process resulted in widespread agro deforestation and resultant overexploitation of existing firewood and timber reserves and grazing land, thus tightening the causal circle of agrodeforestation, overgrazing, and plundering of scarce fuel-wood resources.

Ali's (1986) in-depth study of 26 smallholder sugar-cane farms near Tavua in northern Viti Levu showed that in the study area of 146.3 ha the land under sugar cane increased from 78.7 ha (54 per cent) in the early 1970s to 107.6 ha (74 per cent) in 1982. Associated with, and facilitating, the increase in sugar-cane monoculture were: uncontrolled felling of trees, often with bulldozers, on grazing lands, and along farm edges and river courses, to make room for more sugar cane; a decline in the cultivation of rice and vegetables on separate plots and as intercrops with sugar cane; and a decline in livestock husbandry.

Whereas in 1971, all 26 households planted vegetables and spices on separate plots, or as intercrops with sugar cane, away from the home site, by 1982, only 17 farmers (65 per cent) planted vegetables (3 on separate plots, 5 as intercrops with cane, and 9 with both). In the past, most farmers had set aside pieces of land for vegetable gardens, normally on the alluvial soils along river flood plains and near wells, but extension of sugar-cane plantings into these areas forced most farmers to move their vegetable plots to their home allotments (24 of the 26 farmers had allotments for a house on raised ground at a distance from their farm land).

The loss of the trees scattered in the landscape meant a loss of ornamentation, timber, shade for livestock, fruit, edible leaves, living fences, green manure, wood for handicrafts, food for livestock, medicines, and fuel. The ecological stability of the sugar-cane landscapes was also lessened as the trees had served to enrich soil, to control erosion, and as wind-breaks. The only beneficial effect was that, in the face of increasing agrodeforestation and pressure on remaining tree resources, farmers were forced and, in some cases, encouraged to increase tree planting around their home compounds. A survey of trees on sugar-cane farms throughout Viti Levu, Fiji, in 1985 provided an inventory of several score of tree species protected or cultivated around residences, the most common being coconut, mango, papaya, the drumstick, or horseradish, tree (Moringa oleifera), curry leaf (Murraya koenigii), Citrus spp., tamarind, monkey-pod (Samanea saman), soursop, and Albizia lebbeck (table 9).

As pressure increases for monocultural crop production, and as large tracts of rural land become increasingly scarce or the population of the rural landless increases, the relative importance of rural home gardens in the provision of food, culturally useful products, and ecological benefits will increase. Greater official recognition of, and support for, rural home-garden agroforestry alongside monocultural agriculture could help to stem the processes of rural agrodeforestation.

Table 9 Names and relative importance of tree species found on sugar-cane farms, western Viti Levu, Fiji, between Sigatoka and Tavua, 1985-1986. (Under "local names," I = Indian, F = Fijian; under "importance," 5 = found on 75-100% of farms, 4 = 50-74%, 3 = 25-49%, 2 = 10-24%, and 1 = <10%)

Scientific name Common names Local names Importance
Acacia farnesiana Ellington's curse ban baburi (I); vaivaivakovotona (F) 3
Aegle marmelos bel apple bel, bael (I) 1
Albizia lebbeck siris tree, white monkey-pod, woman's tongue siris (I); vaivai (F) 3
Anacardaum Occidentale Cashew supardi (I) 1
Annona muricata Soursop salifa (I); seremaia (F) 4
Annona squamosa sweetsop, sugar apple sitafal (I); sermaia (F) 2
Artocarpus altilis Breadfruit uto (I and F) 3
Artocarpus heterophyllus Jakfruit katthar (I) 2
Averrhoa carambola carambola kamrakh (I); wind Idia (F) 2
Azadirachta indica margosa tree nim, neem (I) 2
Bambusa vulgaris Bamboo baas (I); bitu ni vavalagi (F) 1
Bauhinia monandra pink butterfly tree, pink orchid tree   1
Bischofia javanica Java cedar koka, togotogo (F) 1
Brassaia actinophylla Queensland umbrella Tree   1
Carica papaya Pawpaw pipita (I); weleti, maoli (F) 5
Cassia fistula golden shower tree, Indian laburnum, pudding-pipe tree   2
Cassia glauca scrambled egg tree   1
Cassia grandis pink shower, horse cassia sirsa (I); vaivai (F) 3
Cassia javanica pink and white shower tree vaivai (F) 3
Casuarina equisetifolia Casuarina, ironwood jhau (I); nokonoko (F) 3
Ceiba pentandra Kapok rui (I); vauvau (F) 3
Chrysophyllum Cainito star apple   1
Citrus aurantiifolia lime nabbu kaghdi (I); laimi, mold laimi (F) 5
Citrus grandis pommelo; shaddock chakotra (I); moli kana (F) 1
Citrus hystrix rough lemon khatta nabbu (I); moli karokaro (F) 2
Citrus limon lemon khatta nabbu (I); moli karokaro (F) 2
Citrus reticulata mandarin narangi (I); mold madirini (F) 3
Citrus sinensis orange mitha nabbu (I); moli, moli Tahiti (F) 2
Citrus xx hybrid suncrest (hybrid) nabbu (I); moli karokaro (F) 1
Cocos nucifera coconut narial (I); niu (F) 5
Cordia dichotoma sebesten plum lasora, lasoda (I) 3
Delonix regia poinciana, flame tree sekoula (F) 2
Dracontomelon

Vitiense

dragon plum tarawau (F) 1
Erythrina variegata coral tree, dadap drala (F) 1
Eucalyptus citriodora lemon-scented gum   2
Eucalyptus deglupta gum tree   2
Eucalyptus sp. eucalyptus   1
Eugenia brasiliense Brazil cherry sinaili, oula, amla, aula (I) 1
Ficus benjamina Benjamin tree pakar (I); baka (F) 1
Ficus obliqua native banyan pakar (I); baka (F) 1
Gliricidia septum madre de cacao sirsa (I); ba ni cagi (F) 3
Hibiscus tiliaceus hibiscus tree vau (F) 1
Intsia bijuga ipil vesi (F) 1
Jatropha curcas physic nut bakrera (I); wiriwiri (F) 1
Leucaena leucocephala leucaena vaivai (F) 4
Mangifera indica mango aam (I); ma-to (F) 5
Manilkara achras sapodilla   1
Morinda citrifolia beach mulberry achi (I); kura (F) 2
Moringa oleifera horseradish tree, drumstick tree seijan, saijan (I) boro ni Idia (F) 5
Murraya koenigii curry leaf, Indian bay leaf tej patti (I) 5
Musa AAA diploid banana kera (I); jaina (F) 2
Musa AAB triploid lady finger banana liga ni marama (F) 3
Musa ABB triploid bluggoe, blue Java bata (F) 1
Pandanustectorius pandanus, screwpine balawa, vadra (F) 1
Peltophorum pterocarpum golden poinciana sirsa (I); vaivai (F) 1
Pinus caribaea Caribbean pine paint (F) 2
Pithecellobium dulce Madras thorn kukafalli, kataiya (I) 3
Plumeria obtusa frangipani, plumeria bna (F) 2
Plumeria rubra frangipani, plumeria bua (F) 3
Psidium guajava guava amrood (I); quwnwa (F) 3
Puncia granatum pomegranate anar (I) 2
Samanea saman pod rain tree, monkey- sirsa (I); vaivai (F) 5
Spathodea campanulata African tulip tree   1
Spondias dulcis Polynesian vi-apple, Polynesian plum amra (I); wi (F) 2
Syzygium cumini Jambolan jamun (I); kovika ni Idia 2
Tamarindus indica Tamarind imli (I); tamarind (F) 5
Thevetia peraviana yellow oleander kandel (I) 2
Ziziphus jujuba Chinese jujube ber (I) 2
Ziziphus mauritiana Indian jujube baher, bair (I) 1

Source: Field surveys by the authors, 1985-1986.

(introduction...)

Intercropping of tree crops/woody perennials with commercial or subsistence ground or tree crops
Planting of timber, fuel wood, and general-purpose trees in relation to agroforestry and agriculture
Grazing with commercial tree cropping and silviculture
The future of institutional agroforestry in the Pacific

The agroforestry systems so far described in this volume have been informal, often traditional systems that were developed on the basis of empirical, non-quantitative experimentation by local practitioners. We turn now to institutional agroforestry in the Pacific Islands, by which we mean those agroforestry activities that are promoted by governments, quasi-government organizations, private agencies, companies, and aid donors and that involve external funding, formal training, agronomic research, and extension services. Because the activities that fall under the rubric "institutional agroforestry" are becoming so many and varied, we provide here only a summary review rather than attempt a comprehensive survey. Other reviews of the topic are the paper by Vergara and Nair (1985) and the reports given by country representatives at a meeting on agroforestry held in 1987 at the University of the South Pacific's School of Agriculture in Western Samoa (Clements 1988).

Inasmuch as plantation agriculture is the core of much institutional agroforestry in the modern Pacific, it can be said that institutionalized activities were first promoted by colonial governments when they encouraged the plantation production of products for export in order to finance government expenditure and to provide cash income to local populations. Agroforestry projects are encouraged today for similar reasons as well as to meet local demand for forest products, to bring grass-fern "wastelands" to greater productive use, and to reduce the adverse environmental effects of rapid deforestation. A major emphasis of recent agroforestry programmes has been the promotion of increased productivity through multiple use, although multiple use is often defined solely with regard to the commodities derived from some mix of commercial crops, livestock, or timber, rather than in terms of the traditional context wherein an agroforested landscape showed increased multi-purpose utility and stability.

Table 10 Categories and examples of institutional agroforestry in the Pacific Islands

Intercropping of tree crops/woody perennials with commercial or subsistence ground or tree crops - Coconuts intercropped with ground and tree crops

  • Cocoa intercropped with ground crops under shade trees and/or coconuts
  • Coffee with shade/nurse trees or shelter-belts
  • Citrus with shelter-belts and multi-purpose species
  • Bananas intercropped with ground crops under coconuts or other trees
  • Vanilla with host plants under coconuts
  • Oil-palm with subsistence and commercial crops
  • Kava with ground crops and/or other trees
  • Other commercial agroforestry intercrops

Planting timber, fuel wood, and general-purpose trees in relation to agroforestry and agriculture

  • Pine plantations in relation to agroforestry
  • Non-pine forestry in relation to agroforestry

Grazing with commercial tree cropping and silviculture (silvipastoral or agrosilvipastoral systems)

  • Livestock under coconuts
  • Cattle under timber species
  • Other silvipastoral activities

Most modern agroforestry projects fall into one of three categories:

  1. intercropping of tree crops or woody perennials with commercial or subsistence ground or tree crops;
  2. planting timber and/or fuel-wood species (mostly exotic) within existing agricultural systems, as intercrops, rotational crops, or as small-scale monocultures, or woodlots; and
  3. 3. grazing under commercial tree cropping.

Table 10 provides examples of the kinds of activities that are practised under each category.

Intercropping of tree crops/woody perennials with commercial or subsistence ground or tree crops

In the first category, the focus has been on monocultural production, with limited intercropping, of major export crops such as coconuts (for copra and oil), cocoa, bananas, coffee, rubber, and oil from oil palms, and, to a lesser extent, Citrus fruit, papaya, mango, kava (Piper methysticum), and vanilla (which requires support, or mother, trees). Although most of the species in this category have been promoted on the basis of large-scale, often foreign-owned, plantations or estates, they have also been actively promoted for smallholder production by colonial and independent governments and aid agencies.

Since the early nineteenth century and until recently, most of the islands in the Pacific basin have been characterized as plantation economies dependent on the production of coconuts, cocoa, and to a lesser extent coffee or bananas. On the atolls and smaller isolated islands, the coconut, which was an integral component of subsistence agroforestry systems, became the main source of terrestrial-based cash income. For some of the atolls of Micronesia, the introduction of the "tin can copra economy" led to abandonment of Cyrtosperma chamissonis taro pits and the erosion of self-sufficiency (Farrell 1972). The cultivation of subsistence crops was replaced by the cultivation of coconuts for copra and coconut oil to earn a cash income. On the larger islands, the establishment of coconut plantations was soon followed by the development of intercropping systems that included cattle and later cocoa under the coconut palms.

Because of the economic significance of plantation crops, especially coconuts, to many of the islands, systems of raising cocoa, cattle, or coffee under coconuts and other trees have been the subject of concern by island departments of agriculture and regional and global development organizations. Most of the plantation crops have been characterized by fluctuating demand and prices, determined largely by metropolitan countries and larger producers, an uncertainty that has motivated many countries in the Pacific to investigate alternative export crops such as ginger and vanilla.

Coconut intercropping

Coconut agroforestry is the most widespread and important form of commercial and government-promoted agroforestry in the Pacific Islands. Although most smallholders have intercropped coconuts with ground crops and other tree crops or protected trees as part of the traditional polycultural agroforestry systems, many governments in the region continue to encourage the intercropping of ground crops and some tree crops with coconuts. In many cases, this effort has taken the form of the systematic planting or replanting in existing garden areas of regularly-spaced palms, which can then be under cropped. Alternatively, undercropping in existing commercial groves of coconut palms has been systematically encouraged.

Very early on in Papua New Guinea, "it was advocated that plantations could be virtually self-supporting with regard to the provision of food for labour" by intercropping subsistence foods between coconuts; "if grown in correct rotation, the fertility of the soil will not only be maintained, but improved" (Gallasch 1976). However, apart from the widespread intercropping of coconuts with cocoa and pasture, and to a lesser extent, with Robusta coffee, interplanting with food crops was actively discouraged by the Department of Agriculture, Stock and Fisheries. In the early 1970s, however, experiments conducted on the intercropping of coconuts with selected forage grasses and legumes, taro, sweet potatoes, corn, and peanuts indicated that a variety of food crops could be grown successfully between coconuts, and that the effect of intercropping tended to be beneficial in terms of reduced weeding and maintenance of all crops, and that such intensive use of scarce land yielded substantially greater returns than coconut monoculture (Gallasch 1976).

Similar experiments have been conducted in Vanuatu, Fiji, Tonga, and Western Samoa. For instance, in Tonga, a succession of coconut replanting schemes over the past two decades has encouraged the replanting of coconuts on smallholder agricultural allotments, and has actively encouraged intercropping with ground crops, a practice that would probably continue regardless of official support. Maude (1965) found that on some of the smaller islands of Ha'apai, where land is scarce, there were few allotments without full coconut coverage, under which ground-cropping was practiced. Thaman (1976) reported that the median coconut coverage on the main island of Tongatapu was 50 per cent, with over 50 per cent of all allotments having recently planted coconuts, most of which were undercropped at least partially with staple ground crops and supplementary crops. A wide range of cultivated and protected tree species also grow throughout the coconut groves. The normal spacing was 30 feet between rows and 20-30 feet within rows. Because of the increase in plough cultivation and the interest in intercropping, there have been moves to increase the between-row distance and to decrease the within-row distance to as little as 10 feet in order to maintain copra yields, while at the same time making cultivation easier and maximizing both subsistence and commercial intercropping between rows.

As Watt (1980, 303) argues, for most island groups and many small islands within groups, institutionalized coconut intercropping and re habilitation is of utmost importance since coconuts constitute the only sizeable wood resource: "only in the case of Papua New Guinea is the size of the coconut resource dwarfed by the commercial timber resource." The development of a successful coconut-milling industry for overmature stems in Tonga and other areas has not only made use of a resource that had been wasted or destroyed, but also has facilitated the improvement of coconut production by removing overmature trees so that effective replanting could take place. This process is facilitated in Tonga by coupling the well developed and ongoing coconut replanting scheme with a chain-saw service to remove overmature palms for milling (MAFF 1985, 41).

Cocoa intercropping

There are numerous examples of the intercropping of cocoa, or cacao, with food crops or with other tree species, many of which serve as shade plants. In the Tolai areas of the Gazelle Peninsula of New Britain in Papua New Guinea, large-scale planting of cocoa, which began in the 1950s, has either displaced food gardens or has resulted in significant changes to the traditional food systems, including the development of cocoa-intercropping systems. The system that operated in the mid-1970s consisted of clearing forest fallow and then planting root crops and bananas. After the root crops were harvested, cocoa was planted under the shade of the bananas, together with Leucaena leucocephala and/or coconuts, so that the food garden became a regularlyspaced cocoa plantation under Leucaena or a cocoa-coconut-Leucaena mixture (Bourke 1976).

In other places in the Pacific, several other intercropping combinations with cocoa have been developed. For example, in the Solomon Islands, cocoa has been intercropped with chillies, ginger, and tumeric; and Leucaena leucocephala and Gliricidia septum have been used as shade trees (Walton 1988, personal communication). In Vanuatu, the joint-venture Metenesel Estates Cocoa Project, covering 1,700 ha, has officially encouraged intercropping and the preservation of buffer zones of alternate 10-ha blocks of uncleared indigenous forest between cocoa planting, with Gliricidia septum as temporary shade. Most of the cocoa in Vanuatu is cultivated under the shade of coconuts or thinned natural bush or forest, although some areas have been formally planted under Gliricidia septum, Leucaena leucocephala, and Erythrina variegate (Jacovelli and Neil 1984, 6). In Fiji, cocoa is commonly intercropped with tannia (Xanthosoma sagitti folium)under Erythrina variegate as a shade tree. The Ministry of Primary Industries has recommended the planting of cocoa under coconuts, taro, bananas, cassava, and kava (Piper methysticum) (MAF 1974).

Coffee intercropping and agroforestry

Like cocoa, commercial coffee plantations are planted under shade trees and/or with wind-breaks. The major species used in Papua New Guinea for such purposes are Leucaena leucocephala, Casuarina spp., Albizia stipulate, and Erythrina variegate. Coffee has also been interplanted under coconuts in coastal Papua New Guinea, Vanuatu, and New Caledonia (Brookfield with Hart 1971, 157). In the Sepik district of lowland Papua New Guinea, Allen (1985, 227) reports that Robusta coffee, grown in permanent blocks, has been integrated into the traditional diversified shifting agricultural system, with food crops being planted during the establishment stage and bananas and Xanthosoma taro being retained in mature gardens, with Leucaena leucocephala serving as the shade tree.

On Tanna, in Vanuatu, intercropping and agroforestry strategies have been actively promoted on a 450-ha Arabica coffee nucleus estate and on a projected 250 ha of associated smallholder plots (of which 50 ha had been established at the end of 1987). The intercrops included peanuts and maize (root cropping is discouraged). Various models have been tried to protect the coffee from wind damage; these included leaving strips of indigenous forest between coffee blocks and planting casuarina shelter-belts. There is also a major French-sponsored research project on Santo, where experiments are being conducted on the intercropping of cocoa and coffee with coconuts and other species such as kava, black pepper (Piper nigrum), and vanilla, using Erythrina and Cyathea spp. (tree fern) trunks as nurse plants/climbing poles for pepper and Gliricidia for vanilla (Barrance 1988).

Apart from the almost universal planting of shade or wind-break trees, very few countries have actively and systematically promoted the intercropping of coffee with food crops. Many have actually discouraged it, even though, as noted above, studies by Carrad (1982) in the Papua New Guinea highlands showed that intercropping of root crops, bananas, and other crops with coffee, in a 15-year cycle, gave smallholder farmers almost twice the real income obtained from coffee monocropping.

Citrus with shelter-belts and intercropping

The Fiji Sugar Corporation (FSC) and its predecessor, the Colonial Sugar Refining Company (CSR), in conjunction the Agricultural Department, have, at various times, officially promoted the establishment of small mixed citrus orchards on smallholder sugarcane farms throughout Fiji. Such orchards are now common, and do provide some cane-farm families with an additional source of income through local sales, often at fresh produce markets in the main towns. In most cases, however, such efforts have met with only limited success beyond satisfying on-farm subsistence needs.

Perhaps most notable was the proposal for the eventual development of a system of 90 ha of smallholder citrus holdings, each of 2.5 ha, to supplement production from 162 ha of monocultural production at the Batiri Citrus Scheme, established on the island of Vanua Levu in 1978, funded by foreign and local investors and the Fiji Development Bank (FDB). Additionally, 200-500 ha of West Indian limes were to be established in nearby Seaqaqa and Bua to generate additional income and employment on Vanua Levu and to extend the processing season at the factory (C.P.O. 1980, 110). As early as 1984, it was suggested that the Batiri Citrus Scheme would not be feasible as a "stand-alone" orange-processing project, given the high costs of operations, inadequate financing, low long-term production levels, delayed investments in other aspects of the project, and the underutilization of land (McGregor 1984).

To achieve fuller utilization of the project, which has been hampered by the very short processing season of the oranges, diversification into pineapples, mangoes, guavas, and passion-fruit was recommended. Various intercropping trials were carried out, but, unfortunately, this highly export-oriented and capital-intensive project ran into major financial difficulties in 1988 and has subsequently been taken over by the National Marketing Authority (NMA). The smallholder "outfarm" Stage II component was never developed, although a few smallholder farmers did plant improved citrus cultivars, and surrounding Vanua Levu out-growers supplied some 10 tonnes of fruit, with another 109 tonnes coming from Rotuma in 1984 (McGregor 1984, 10). A major problem of this monocultural project has been the destruction of 25 per cent of the annual production by a fruit-sucking moth (Fiji Times 26 July 1988, 3).

In 1945 the Cook Islands began a large-scale Citrus Replanting Scheme, with the initial establishment by 1950 of 150 acres on Rarotonga and 50 acres on Aitutaki. which did include deliberate inter cropping with cover crops and multi-purpose trees. Crotalaria and tarapi grass were planted as green manure between rows to be ploughed under thrice yearly during inter-row cultivation, with leguminous dadap trees (Erythrina variegate) planted between alternate citrus trees to be trimmed to provide green manure. Shelterbelts of jambolan (Syzygium cumin)), Albizia falcataria, hibiscus (Hibiscus rosa-sinensis or H. tiliaceus - it was not made clear what species was planted), and hedge panax (Polyscias guilfoylei) were planted, with A. falcataria being grown to provide timber for citrus shipping-cases (Johnson 1951; Johnston 1967). Unfortunately, the scheme did not encourage self-reliance and has been less than successful. Many landowners have fallen into debt and have suffered widely fluctuating yields and returns.

In Nine, which has actively promoted the planting of extensive areas of limes (Citrus aurantiifolia), often as intercrops with coconuts and other crops, the export of limes and lime juice has been the second most important agricultural export after passion-fruit (Carter 1984, 298). Other countries that have tried to encourage limited intercropping of citrus include Tonga, Western Samoa, and French Polynesia.

Banana intercropping

Intercropping of bananas for export under coconuts, with cocoa, and to a lesser extent with subsistence ground crops and other trees, has been promoted from time to time in Fiji, Tonga, Western Samoa, and the Cook Islands, although recently there has been a trend towards more monocultural production.

In Fiji, the export banana industry was based on smallholder production, often under coconuts and other trees with root crops and supplementary crops cultivated in between, at least in the early stages of plantations. In the early 1960s, however, greater official emphasis was placed on monocultural production for expanding export markets in New Zealand and Japan. The major effort in this direction was the Lomaivuna Resettlement Scheme in south-east-central Viti Levu, which was designed to resettle some 200 landless families on 4ha (10-acre) allotments. The objective of the scheme was the monocultural production of bananas on 2 ha, with the balance being used for homestead and subsistence crops. As a banana scheme it failed because of serious outbreaks of black-leaf-streak fungal (Mycosphaerella musicola) and bunchy-top viral diseases, heavy nematode (Radolus similis) infestations, loss of trees rule to a series of tropical cyclones, the unsuitability of the soils and topography for banana cultivation, and the resultant poor quality of the fruit (Eaton 1988b). Because of these problems associated with monocultural banana production, export banana production in Fiji ceased in the early 1970s.

Intercropping of bananas for export with coconuts and staple root crops has also long been promoted in Tonga, Western Samoa, and the Cook Islands, and supported by considerable financial and technical support from the New Zealand marketing agent, Fruit Distributors Limited, and the New Zealand Government. As was the case in Fiji, banana production increased dramatically in the mid- and late 1960s in response to increasing market demands in New Zealand (partly due to the removal of an earlier quota system, after Fiji ceased export production), but then declined just as dramatically due to the same diseases as experienced in Fiji. Banana production in Tonga, for example, dropped 60 per cent in 1969, with further decreases in 1970 and 1971 (Thaman 1976).

In an attempt to combat declining production and improve export quality, the Departments of Agriculture in Tonga, Western Samoa, and the Cook Islands, with subsidization by Fruit Distributors Ltd and New Zealand AID, have all implemented planting or rehabilitation schemes. In Tonga, a Banana Rehabilitation Scheme, which required growers to plant one or more acres of at least 500 plants, with no intercropping, to control weeds to a reasonable standard, to practice desuckering and leaf-trimming, and to destroy bunchytopaffected plants, was instituted in the early 1970s (Thaman 1976). Despite continued efforts to rehabilitate banana production in Tonga, Western Samoa, and the Cook Islands, yields and fruit quality continued to deteriorate in all island groups during the 1960s and 1970s, with the major share of the New Zealand market being increasingly supplied by Ecuadorean bananas.

Efforts in all three countries during the 1980s to rehabilitate production of bananas for export have continued to face problems of disease, cyclone damage, and marketing. Many of these often heavily capitalized efforts have been strongly monocultural; several have failed entirely; none has met with real success; and most banana schemes have contributed to agrodeforestation.

Vanilla intercropping

Smallholder vanilla production, with appropriate support plants, and often under coconuts or other trees, has been increasingly promoted by governments throughout the Pacific. The only areas where vanilla has become a significant export, however, are French Polynesia and Tonga. In French Polynesia, where it has been intercropped under coconuts and other trees using Tecoma stans as a support plant, exports dropped from 28 tonnes in 1970 to a low of 0.6 tonnes in 1981, but renewed interest in the crop had increased production to 14.7 tonnes by 1984 (Douglas and Douglas 1989, 149).

In Tonga, where vanilla is becoming an increasingly important export crop, there is widespread official advocacy for its planting under coconuts and other trees, using physic nut (Jatropha curcas) as the support plant. Vanilla, given its high returns per acre and negligible demands on soil fertility as an epiphytic member of the orchid family, seems to be the perfect plant for Tonga, where land scarcity is increasing. Vanilla has been a major export crop in the northern Vava'u group since the 1960s, and the Department of Agriculture has been seriously promoting its cultivation in all island groups within the Kingdom for the past 10 years. By 1984, there were reportedly 105 vanilla growers with 21.2 ha of vanilla in production, 37.2 ha planted but not yet in production, and another 3.7 ha planted with support plants in preparation for vanilla planting.

Two different models have been developed to determine how best to incorporate vanilla into the traditional mixed cropping system in the belief that Tongan farmers will manage their plantations better if vanilla is introduced into a system that is already familiar (Fa'anunu n.d.). Other countries, such as Fiji, Vanuatu, and Papua New Guinea, have small vanilla development programmes, and a small number of growers in Fiji are producing marketable yields.

Oil-palm intercropping

Papua New Guinea and the Solomon Islands are the only two countries growing oil-palm (Elaeis guineensis) on a commercial scale. Papua New Guinea has three schemes, the first and largest being the Hoskins Oil Palm Development Scheme started in West New Britain in 1967, on the basis of a balance between a company plantation and adjacent smallholders; the Sangara Scheme started in the Northern Province in 1977; and the Bialla Scheme started in West New Britain in 1980. On the Hoskins oil-palm resettlement scheme, smallholders were encouraged to intercrop subsistence crops between oil-palms in the early stages of plantation development. Consequently, in the late 1970s, 31 per cent of the settlers were able to practice market garden ing of traditional foods, and, given access to land, the farming systems of almost all settler groups had remained virtually intact with people growing their own food (Benjamin and Wapi 1982). In the Solomon Islands, however, the large-scale Guadalcanal Plains joint venture oilpalm development scheme is entirely monocultural, with little encouragement of intercropping or agroforestry practices (Carter 1984).

Kava intercropping

Kava (Piper methysticum) is an important crop in Vanuatu, Fiji, Tonga, Pohnpei, and to a lesser extent in Western Samoa. Given its shade tolerance and that it takes 3-6 years to mature, it is particularly amenable to intercropping both with short-term food crops and with coconuts and other trees.

In Vanuatu, where kava cultivation and consumption is, in many rural areas, highly ceremonial and very significant in garden magic, the plant is almost always intercropped amongst trees in polycultural gardens. Kava is also one of the main cash crops promoted as an intercrop with coconuts and in timber plantations of Cordia alliodora.

In Fiji, kava (known locally as yaqona) is a major cash crop for local sale and the major cash crop for villagers in some of the smaller outer islands and the interiors of the larger islands. Programmes to rehabilitate the copra industry have encouraged the establishment of other economic crops such as cocoa and kava underneath coconut palms. Some Fijian kava farmers are beginning to intercrop kava with Calliandra calothyrsus, a nitrogen-fixing species being promoted for agroforestry in Fiji by the Fiji-German Forestry Project. Throughout Fiji, kava is commonly intercropped with staple root crops, particularly taro (Colocasia esculenta). The kava is then allowed to mature amongst existing cultivated and protected trees after the root crops have been harvested.

Whereas copra prices are dictated by outside market forces beyond the control of Fiji and its farmers, most of the kava produced is sold internally and has shown much greater price stability than copra. Of further significance is that almost all kava farms are Fijian-run. Sofer (1985) argues that the success of kava as a commercial crop in Fiji results from the familiarity of Fijian farmers with its husbandry. Also, kava provides higher returns to the grower, per unit of land, labour, and capital investment, than most alternative cash crops; it requires few outside inputs; its planting material is readily available and consists of parts of the plant not normally sold; it can be harvested or processed at any time; it is non-perishable and can be easily stored; and it is well-adapted to the environmental conditions and to the semisubsistence polycultural shifting agroforestry system in Fiji.

In Tonga, kavais also seen as an important local cash crop and is almost always planted under coconut palms and other useful trees.

Other commercial agroforestry intercrops

In Fiji, a number of successful smallholder papaya plantations, sometimes under coconuts or as part of diversified agricultural developments, supply the growing demand for the fruit in the country's tourist resorts. Unfortunately, a major 30-ha capital-intensive papaya plantation developed by the Native Land Development Corporation to air freight fruit to Japan has been a failure. Most recently, however, the Southern Development Company Ltd. the company responsible for directed smallholder tobacco production for Fiji's local cigarette manufacture, has attempted to diversify the production of its directed smallholder farmers into tomatoes, maize, chillies, mangoes, and papayas. Papayas seem to offer the greatest potential, with 18 farmers already growing the crop under the auspices of the company (Eaton 1988, 194-195).

Areas of improved mango varieties for export have also been planted by the Native Land Development Corporation and on the Government-run Yaqara Beef Cattle holding on Viti Levu as part of a diversification programme. Papua New Guinea, Vanuatu, Tonga, Western Samoa, and the Cook Islands have also experimented with improved mango cultivars, as well as avocados in efforts to diversify tree crop production.

Smallholder passion-fruit (Passiflora edulis) production, often as an intercrop with coconuts or other trees or on small plots within existing plantation areas, has been strongly promoted in Papua New Guinea, Fiji, Tonga, Western Samoa, and Nine, as part of agricultural diversification programmes. Results have been mixed. Hurricanes have damaged vines, market availability has been inconsistent, as has ability to supply, given low production because of poor management and insufficient labour to carry out daily the hand pollination of the flowers needed to obtain high yields. The industry has showed promise only in Fiji, where the production is run as a directed smallholder operation by a fruit processing and exporting company that provides most of the inputs, planting material, and research, extension, and managerial services, and in Nine, where passion-fruit is the most important agricultural export (Carter 1984).

In Fiji, the aid-funded Fiji-German Forestry Project, working together with Fiji's Forestry Department and the Extension Service of the Ministry of Primary Industries, is promoting a kind of alleycropping system through the introduction of hedgerows of Calliandra calothyrsus along the contours of sloping land used for various crops, particularly ginger. In less than two decades, ginger has grown to be Fiji's third-largest agricultural export earner, surpassed only by sugar and coconut products. A combination of the environmental requirements of ginger (high rainfall, fertile soils, good drainage, freedom from nematodes) and the low capital resources of the smallholders planting the crop has led to a "ginger frontier," where ginger is grown on plots newly cleared of forest for a few seasons and then moved on to still more recently cleared land, leaving behind a cleared, less fertile area (Overtop 1989, 83). The Calliandra hedgerows, which are recurrently pruned, are intended to slow erosion, which has been serious at times on the steep ginger-growing lands, to provide nutrients and firewood, and to serve as wind-breaks. Attempts to analyse the economic benefits of Calliandra hedgerows suggest that the labour costs of establishing the hedgerows are repaid in two years by savings on fertilizer (Künzel 1992, personal communication).

Planting of timber, fuel wood, and general-purpose trees in relation to agroforestry and agriculture

Because the imminent demise or depletion of commercially usable natural forests can be so readily foreseen in many Pacific Island countries (Watt 1980, 297), governments and development agencies have in several places promoted either some form of restocking or enrichment of commercially logged areas or the establishment of forest plantations on degraded grassland sites. Not all these efforts can be classified as agroforestry, strictly speaking; but in the Pacific context, as in most of the tropical world, the traditional, if transient, shift of land use back and forth between forest and agriculture on any particular site makes it relevant to consider what at first glance appear to be purely forestry projects.

Many of the timber species institutionally promoted have been exotics such as Caribbean pine (Pinus caribaea), West Indian mahogany (Swietenia macrophylla), cordia (Cordia alliodora), and Eucalytus spp., although some indigenous Pacific species such as Albizia falcataria, Agathis spp., Araucaria spp., and Endospermum spp. have been successfully established, often as exotics in areas beyond their natural range. Many other species - including West Indian cedar (Cedrela odorata), the silky oak (Grevillea robusta), teak (Tectona grandis), mahogany (Swietenia mahogoni), toon tree ( Toona australis), cadamba (Anthocephalus chinensis), and Albizia lebbeck along with several indigenous trees - have also been the subject of trials, and planted to various degrees throughout the islands.

Firewood and multi-purpose species that have been successfully introduced include Leucaena leucocephala, Erythrina spp., Casuarina spp., and Gliricidia septum, and, to a lesser extent, Securinega samoana and Adenanthera pavonina. Other species, all of which have been planted experimentally and which seem to grow successfully, but which have not yet become so well established, include Cassia, Acacia, and Calliandra spp. Apart from timber and fuel wood, the major multi-purpose objectives of such plantings are site reclamation and amelioration, erosion control, wind protection, shade, multipurpose construction and handicrafts, nurse cropping, fodder, green manure, and food.

The indigenous casuarinas, particularly Casuarina equisetifolia, have also shown considerable promise for reforestation programmes, and have been planted in Tonga in land reclamation projects, in the Cook Islands for the rehabilitation of degraded lands, and on atolls as sources of fuel wood and to protect coconut plantations from saltwater damage. C. oligodon and C. papuana are traditionally used for reforestation and to enrich fallow land in Papua New Guinea, and are now promoted in some areas for land rehabilitation and as shade plants for coffee.

Pine planting in relation to agroforestry

Of the total area of timber plantations in the Pacific, well over 50 per cent is accounted for by Caribbean pine (Pinus caribaea). The largest area of pine planting is in Fiji, where that country's Pine Commission together with the Forestry Department has established over 50,000 ha of plantation since 1960, mostly on degraded anthropogenic grasslands (Drysdale 1988a, 110; Watt 1980, 301). Some pine timber is used locally, but the wood was intended mainly for export, and a wood-chipping mill is now in operation. In the mid1960s, under a programme now discontinued, woodlots of Pinus caribaea on smallholder sugar-cane farms were promoted by the colonial government.

Sized from 0.4 to 2 ha, these woodlots were planted on steeper non-cane areas of farms to control erosion, provide on-farm supplies of timber and fuel wood, and for undergrazing by farm animals (Eaton 1988b, personal communication). Apart from this woodlot grazing and grazing of cattle in association with larger pine plantations (described below), there has been no institutional support for any form of intercropping or other agroforestry activities in pine plantations (Drysdale 1988b).

Similarly, in the limited areas of pine planting in New Caledonia, Western Samoa, Tonga, and the Cook Islands, there has been little or no link to agroforestry in such programmes, with the main focus being on creating a timber resource, land improvement, erosion control, and employment creation in rural areas.

In highland Papua New Guinea large areas of degraded grassland have been planted with pines (Pinus spp.) and Araucaria spp. Intercropping activities are few and consist of the intercropping of coffee and cardamon on a trial and demonstration basis (Howcroft 1983).

In Vanuatu, P. caribaea var. hondurensis is the main species planted in forest plantations in seasonally dry and highly degraded sites on the southern islands of Aneityum and Erromango, where some 550 ha had been established up to April 1985. The commercial viability of such plantings is still uncertain, however, due to poor access to markets and high transport costs. On Erromango, high costs of clearing land of the indigenous pioneering species Acacia spirobis has stopped the development of pine plantations. Benefits in the form of erosion control and aiding the local economy through wages were the main motives behind these programmes (Neil 1986a).

Non-pine forestry in relation to agroforestry

To judge from programmes in Papua New Guinea, the Solomon Islands, Vanuatu, Tonga, and Western Samoa, there seems to be greater promise and greater institutionalized promotion of intercropping with other, primarily broadleaved evergreen, species than has been the case with pines.

In Papua New Guinea, where extensive areas of Eucalyptus deglupta have been planted, cocoa and coffee have been successfully grown at 4 m x 4 m and 3 m x 3 m spacing, respectively, in conjunction with E. deglupta planted at 10 m x 10 m (Jacovelli and Neil 1984, 10).

Also in Papua New Guinea, severe environmental degradation resulting from rapid urban expansion and associated subsistence gardening and "fuel-wood mining" prompted the cities of Lae and Port Moresby to institute fuel-wood-planting programmes. In Lae, in 1978, it was decided to plant 200 ha of sloping land (20°-30°) in Leucaena leucocephala for firewood and to intercrop fuel-wood species with annual food crops in zones designated for subsistence food gardening. The project, which was allocated K250,000 (US$275,000) over six years, had a management component coupled with a public education programme and a team of local government rangers to control gardening and to police the area (King 1987). Follow-on projects were planned but not carried out because of lack of funding. By 1988 the project had ceased to operate, and the original plantings of some 100 ha of L. Ieucocephala, Acacia auriculformis, and Eucalyptus spp. and 5 ha of "agroforestry plantings" of fuel-wood species with food crops had been cut down or removed completely (King 1987).

In Vanuatu, Cordia alliodora, a hardwood native to Central America, has been the main commercial silvicultural species since the mid-1970s, with over 1,000 ha planted on 12 islands as of 1984 (Neil 1984). Cordia was first planted on various islands in 5-10-ha blocks called Local Supply Plantations (LSP). As the potential contribution of forestry to rural and national development became evident, larger, export-oriented Industrial Forest Plantations (IFP) were established on the islands of Pentecost, Erromango, and Aneityum (Jacovelli and Neil 1984). The rapid expansion of IFPs, sometimes with plantings of up to 200 ha per year on single sites, led to unprecedented demands for land and aroused fears among landowners, especially on Pentecost, that these silvicultural activities would make land unavailable for planting subsistence and commercial crops. This prompted the Vanuatu Forest Service to establish, on Pentecost in 1984, demonstration plots growing a wider range of subsistence and cash crops within forestry plantations of Cordia alliodora (Jacovelli and Neil 1984).

Crops established between line plantings of Cordia alliodora included 8 sweet potato cultivars, 6 cassava cultivars, 13 aroid cultivars from Colocasia esculenta, Xanthosoma sagittifolium, and Alocasia macrorrhiza, 12 yam cultivars, kava (Piper methysticum), and trials with coffee (Arabica and Robusta), cocoa, and cardamon. In addition to these trials, subsistence gardens have also been established under Cordia alliodora by both local landowners and forest workers alike (Jacovelli and Neil 1984, 8).

Because C. alliodora may be severely attacked by root rot (Phelli nus noxius) in some conditions, and does not perform well on some sites, other species currently being tried in Vanuatu include Terminalia brassii, T. calamansanai, Eucalyptus deglupta, Swietenia macrophylla, Toona australis, and Cedrela odorata. However, the barks of both T. brassii and E. deglupta are palatable to cattle (Jacovelli and Neil 1984, 10; MacFarlane 1980). The species showing greatest potential as an alternative species to C. alliodora may be S. macrophylla, and if grown with nurse species to reduce pest problems, intercropping should be possible during the early years of rotation (Neil 1986b).

Several other systematic experiments on tree species, both exotic and indigenous, have been carried out in Vanuatu in a search for species especially suitable for fuel wood, timber, or pulpwood, but none of this research was connected with agroforestry. Research on agroforestry has focused almost exclusively on "cash crops which appear to have great potential, particularly coffee and cocoa, and possibly kava and cocoa" (Jacovelli and Neil 1984, 11).

In Fiji, some 22,953 ha of tropical hardwood forests have been planted as of mid-1986. Of these, 14,987 ha are West Indian mahogany (Swietenia macrophylla), 3,058 ha are Cordia alliodora, 2,963 ha are cadamba (Anthocephalus chinensis), 928 ha are Maesopsis eminii, 438 ha are Eucalyptus deglupta, and 202 ha are the indigenous species Endospermum macrophyllum (ADAB 1986). Despite such considerable silvicultural activity, in terms of both hardwood and pines, it is essentially monocultural, and, as the General Manager of the Fiji Pine Commission has stated: "Institutionalized agrosilviculture is non-existent in Fiji at present" (Drysdale 1988b, personal communication).

Tonga's silvicultural activities are more diverse, some being significantly agrosilvicultural. More purely silvicultural activities include a major reforestation programme begun on the island of Eua in the mid-1960s. Over 40 ha of mixed exotic species including Toona australis, Cedrela odorata, Cordia alliodora, Grevillea robusta, Agathis robusta, Pinus caribaea, and Eucalyptus spp., as well as suitable indigenous species, such as Casuarina equisetifolia, Terminalia catappa, and Dysoxylum tongense, were planted on the Eua Forest Farm. Tests of seed stock from throughout the world were also carried out on the farm. Larger areas were subsequently planted, with 104 ha alone being planted in 1979 (Thaman 1984e, 3).

The species most commonly planted in 1984 were Eucalyptus saligna, E. tereticornis, Toona australis, and Pinus caribaea. Seedling pro auction for these species and other timber species, such as Cupressus lusitanica, amounted to 77,491 seedlings (42,427 of which were planted) in 1979 (MAFF 1985, 100-102). Reforestation continues, as the small areas of remaining indigenous forest on Eua are exploited, with the local mill "approaching the end of its productive life as the local hardwood timber supply is cut out and cannot be replaced from the Forest Farm for at least another 10 years" (MAFF 1985, 99). The only truly agroforestry aspect of the Eua silvicultural activities, a taungya system of combined tree-planting and temporary gardens, was phased out because "it has greatly increased pressures for settlement of unsuitable land, and is thus clearly not in the national interest" (MAFF 1985, 100).

A second and continuing agroforestry activity has been the Forestry Extension Programme, which began in the 1960s to produce seedlings for distribution to smallholder farmers for planting in small woodlots or as windbreaks around their agricultural allotments (see chapter 5 on Tongan agroforestry). The major species distributed included Casuarina equisetifolia, Grevillea robusta, Cedrela odorata, Eucalyptus spp., Agathis spp., and Gmelina arborea (Thaman 1984e, 3).

With the establishment of the Extension Nursery at Mataliku on the main island of Tongatapu in 1978, the programme was expanded to include the propagation and distribution of a wide range of timber trees, "cultural" species, and species providing food, medicine, and ornamentation. The considerable interest shown by the people for planting on both rural and town allotments led to a "blossoming of forest extension work" to the point that, in 1978, the nursery could not cope with the demand, which exceeded 8,000 trees per month (MAFF 1979, 99).

According to programme records, as of 1984, at least 155 species had been tested and/or propagated for distribution on Eua and Tongatapu. Of these, 66 were timber species, 45 ornamentals, 32 "cultural" plants of particular importance to the Tongan society, 11 food plants, 6 plants used for coastal protection or land reclamation, 4 for living fences or hedgerows, 3 medicinal plants, and 2 each for windbreaks and firewood. Among the most popular nontimber species were Casuarina equisetifolia (planted as an ornamental, living fence, or wind-break); culturally important sacred or fragrant plants, known locally as akau kakala, such as heilala (Garcinia sessilis), langakali (Aglaia saltatorum), sandalwood, or ahi (Santalum yasi), pua (Fagraea berteriana), pipi (Parinari glaberrima), huni (Phalaria disperma), perfume tree, or mohokoi (Cananga odorata), allspice (Pimenta doica), and Pandanus cultivars; fruit-trees, such as mango, Malay apple (Syzygium malaccense), and macadamia nut (Macadamia integrifolia); and ornamental or shade plants, such as flamboyant, or poinciana (Delonix regia), hibiscus (Hibiscus rosa-sinensis), Cordyline fruticosa, copperleaf, or beefsteak, plant (Acalypha amentacea), bougainvillea (Bougainvillea spp.), poinsettia (Euphorbia pulcherrima), gardenia (Gardenia spp.), and the hedge panaxes (Polyscias spp.) (Thaman 1984e).

The final major area of activity has been the testing and establishment of trees for land reclamation, such as the project to rehabilitate low-lying areas at Sopu to the west of the capital of Nuku'alofa on Tongatapu. Reclamation work at Sopu began in the 1960s, with the planting of Casuarina equisetifolia to stabilize the area, and has continued to the present with extensive plantings of Lumnitzera littorea, Rhizophora mangle, Bruguiera gymnorhiza, Xylocarpus granatum, and other selected species. As recently as 1980, 6 acres of Lumnitzera littorea, 4 acres of Terminalia catappa, and 3 acres of Queensland kauri (Agathis robusta) were planted. The vegetation has reportedly been well-established, with the operation becoming more maintenance than reclamation.

Grazing with commercial tree cropping and silviculture

Grazing, usually of cattle, with commercial tree cropping and silviculture consists mainly of the widespread practice of grazing cattle under coconuts or commercial timber species, and the limited grazing of cattle under Leucaena leucocephala or other fuel-wood or multipurpose species.

Livestock under coconuts

The grazing of cattle (primarily beef, but also dairy cattle) under coconuts (in some cases with pasture improvement) is by far the most widespread practice. It has been encouraged throughout the Islands since colonial times, particularly on large coconut estates. In addition to providing meat and dairy products, cattle are seen as effective weed control and fertilization agents, thus facilitating plantation management and the collection of fallen nuts.

Although primarily promoted on large, often foreign or state controlled estates or plantations, some governments, such as those in the Solomon Islands, Tonga, and Niue, have encouraged smallholder grazing of cattle under coconuts and other trees. In the case of Tonga, smallholder agriculturalists have been encouraged to fence limited portions of their 3.3 ha bush allotments to graze cattle, and sometimes horses, under coconuts and other tree crops and protected trees, or, alternatively, to tether animals to trees and graze on a rotational basis.

The practice has been particularly important in Vanuatu (both before and after independence in 1980) and New Caledonia, where beef cattle production is a major activity. Beef cattle production became so important in Vanuatu, prior to independence, that some plantations were turned into cattle properties. The importance of cattle grew in the 1950s, when steeply rising labour costs made planters increasingly dependent on cattle to keep their plantations clean. At one period in the 1950s, herds became larger than the plantations could support, especially during dry spells, and by the end of the decade, town butcheries had opened in both Port Vila and Luganville, the two main towns. By the end of the 1960s, copra production had become no more than a sideline on a number of plantations (Brookfield with Hart 1971, 164165).

In Fiji, in 1973, 10.5 per cent of the local beef requirements were supplied by the 9.9 per cent of the cattle population grazed under coconuts (MAF 1973; Manner 1983). This is particularly significant given the large proportion of range-fed cattle raised on extensive large-scale developments in the dry zones of Fiji. Papua New Guinea, the Solomon Islands, Vanuatu, and New Caledonia in Melanesia, and Western Samoa and French Polynesia have also actively encouraged cattle under coconuts with trials having been conducted on optimum stocking rates and pasture improvement. Much of the Western Samoa Trust Estates (WSTEC) Mulifanua Copra Plantation, reportedly one of the largest copra plantations in the world (Carter 1984), is undergrazed by cattle.

The potential for the formal promotion of large-scale grazing of cattle under coconuts is greatest on the larger islands of Melanesia and Polynesia. On smaller islands, such as those in Tonga and the Cook Islands, where high population densities and land scarcity make more extensive agrosilvipastoral developments less relevant, small-scale rotational undergrazing of tethered animals is more appropriate. In Nine, where population density is low because of emigration to New Zealand, there have been problems of overgrazing and lack of fodder during times of drought- for example, during the severe drought of 1977-1978, when hay had to be imported from New Zealand.

Richardson (1983, 59) cautions that grazing under coconuts can create problems of soil compaction and, especially in the case of free grazing, preclude intercropping, which should take precedence in areas with limited land resources. As shown by studies in Papua New Guinea and elsewhere, smallholder beef cattle production can have harmful impacts on subsistence cropping (Grossman 1981). Where cash cropping or subsistence production is feasible, Richardson (1983, 59) argues that intercropping should take precedence over grazing under coconuts.

Cattle under timber species

The grazing of cattle under commercial timber species has been actively promoted in Papua New Guinea, the Solomon Islands, Vanuatu, and Fiji. In Papua New Guinea, reforestation projects in both the highlands and lowlands offer opportunities for beef production, and cattle have been actively promoted to control weeds and reduce fire danger by consuming the fuel. Pinus caribaea planting has also been encouraged in order to provide shade for cattle in open grasslands (Watt 1980, 308). The introduction of pasture legumes into timber plantations and surrounding areas has also been actively encouraged, and the development of pastures, followed by grazing, has been more or less standard practice in a number of forest plantations in Papua New Guinea, where klinki and hoop pine (Araucaria spp.), Pinus caribaea, and Eucalyptus spp. are grown. Government forest plantations are made available to local Braziers who establish adequate fencing and pastures and follow acceptable range management and stocking practices (Howcroft 1974; 1983).

In the Solomon Islands, where there is a "Cattle Under Trees" (CUT) project, cattle have been grazed under Eucalyptus deglupta in forest plantations established by the government in logged forest (Macfarlane and Whiteman 1983; Schirmer 1983, 101; Watt 1980, 308) and in Vanuatu under both "Local Supply Plantations" and "Industrial Supply Plantations" of Cordia alliodora, as well as under Pinus caribaea on Aneityum, Erromango, Pentecost, and Santo (Jacovelli and Neil 1984, 8). Grazing under pines in Vanuatu is seen as a means of reducing the significant fire threat in plantations (Neil 1986a).

It is in Fiji that the practice has probably been tried most exten sively, owing to research undertaken by the Fiji Pine Commission (FPC), a statutory body with the objective of facilitating and developing "an industry based on the growing, harvesting, preserving and marketing of pine and other species of trees grown in Fiji" (CPO 1980, 141). The FPC is responsible for managing over 45,000 ha of Pinus caribaea out of an envisioned gross estate of 80,000 ha on the highly degraded talasiga (sunburnt) soils of the drier leeward grasslands of the two largest islands of Fiji. The relatively infertile and eroded areas are vegetated with a grassland sub-climax of presumed anthropogenic origin, including species such as Pennisetum polystachyon, Pteridium esculentum, Gleichenia liners, Psidium guajava, Dodonaea viscose, and Casuarina equisetifolia. On moister slopes, Miscanthus floridulus forms almost impenetrable thickets. These grasslands are subject to frequent and unauthorized burning.

The FPC undertook research into cattle grazing for two reasons: to examine the effects of cattle grazing on reducing fuel in high fire-risk zones; and to test the use of cattle as a site-preparation tool for clearing the land of Miscanthus floridulus, which proved difficult to eradicate by more conventional means such as slashing and burning (Drysdale 1982). Research has yielded variable results. Vincent (1971) concluded that grazing of cattle under 5- and 6-year-old pine plantations in poor soils had a detrimental effect on the incremental growth of pines, whereas grazing trials in the Nausori Highlands to determine the effect on fire hazard reduction resulted in a reduction in fuel from 2,500 kg per hectare to 800 kg per hectare, an average cattle weight gain of 0.24 kg per day, and no pasture deterioration despite heavy stocking rates (Gregor 1972). At Nawaicoba, Partridge (1977) reported weight gains twice this, when trees were planted at 2 m x 3 m spacing, with two rows in every five missing. In variable spacing trials, Bell (1981) found slight bark damage to trees less than one year old because of trampling, when the trees were spaced 3 m apart within rows and 2.5, 3, 3.5, and 4 m apart between rows, the cattle being introduced into the plantation when the pines were 54 cm high.

In 1982, the FPC reviewed various research projects on cattle under pines and concluded that given "the high overhead and general costs of FPC operations, commercial cattle grazing of unimproved pasture under pines, is an unlikely prospect" (Drysdale 1982, 4). Although fuel loadings were considerably reduced, the cost of using cattle for fuel reduction was "considered unacceptably high compared with alternatives such as burning" (Drysdale 1982, 3). In contrast, the use of cattle as a site-preparation tool where Miscanthus predominates was termed an "outstanding success" (Drysdale 1982, 8) because other methods of clearing the giant grass gave incomplete results, were impractical, or cost too much.

Because of the high cost of fencing, the long-term and extensive grazing of cattle under pines has been found to be an uneconomic proposition for the Fiji Pine Commission, although some 480 cattle are allowed to graze under pines free of charge at Drasa and Tavaka-bo, and some cattle owners unofficially graze their cattle in Fiji Pine Commission forests. Native landowners are also allowed to graze cattle under their own pine plantings, subject to certain restrictions. But cattle owners also are unlikely to find fencing a profitable venture. Open-range grazing with night-time penning may be a possibility. In addition, the economics of cattle grazing on improved pastures under trees in Fiji still needs to be ascertained.

Other silvipastoral activities

Trees such as Leucaena leucocephala are used as fodder in Tonga and Papua New Guinea, where they are browsed by cattle as a dietary supplement (Watt 1980, 308). There is perhaps some scope for the grazing of other animals such as pigs, goats, and chickens on improved legume pastures or fallows under coconuts, commercial timber species, or other trees (Quartermain 1980; Richardson 1983).

The future of institutional agroforestry in the Pacific

In the Pacific, as elsewhere, interest in agroforestry has recently grown rapidly among scientists, land-use experts, conservationists, and the development professionals of national governments and international agencies. As already noted, systems of commercial production that would now be classified as agroforestry were initiated early in the Pacific's colonial past, particularly in the form of multistorey arrangements of coconut palms with other crops or with cattle. With regard to agroforestry systems in the subsistence sphere, this book has sought to demonstrate their prevalence and antiquity in the Pacific Islands. As Yen (1980b, 91) comprehensively expressed it in his discussion of "Pacific Production Systems," there is nothing new about multi-storey cropping even though it has often been suggested to smallholders as an innovative technique they might adopt.

In fact native systems have always involved such techniques in village gardens with descending storeys of palms, trees, productive vines, shrubs, herbaceous root crops, and vegetable plants and ornamentals. Similarly, in swiddens, mixed species and variety plantings are themselves multi-storey. In this case such plantings also take on a successional aspect, for following the root crops, some cultigens such as banana and longer-term plants such as breadfruit and other fruit and nut trees, industrial shrubs, and vines, prolong the production of these gardens.

Geographers and anthropologists who have studied these sorts of indigenous systems find ironic some of the attempts made to introduce institutional agroforestry into the Pacific context. On the other hand, in a time of deforestation and agrodeforestation, it is apt to encourage both of the approaches to agroforestry described in chapter 1- the institutional approach, which generally seeks to introduce commodity-focused systems devised on the basis of modern forms of analysis, and the cultural-ecological approach, which is concerned more with long-standing indigenous systems, empirically devised and deeply embedded in the cultural landscape. Whether or not the two approaches can be usefully meshed remains open to question, although some forms of "progressing with the past" do seem possible (Clarke 1978).

When attention is turned to the future of institutional agroforestry in the Pacific, it can be clearly forecast that if individual smallholders are to benefit over the long term from the introduction of an unfamiliar institutionalized agroforestry system, they will need to receive an ongoing package of inputs and information, which suggests the need for some sort of extension service. Unfortunately, it is acknowledged that extension work in many Pacific countries is generally poor, and extension services often have only secondary ranking within ministries or departments (Hau'ofa et al. 1980, 188-189). How to remedy this deficiency raises several complex but pervasive issues, which have been dealt with at length in a large literature and which can only be superficially treated here.

With regard to the initial introduction of a new agroforestry system, it is easy - given the current popularity of agroforestry in the development world to find funding for workshops and projects, but these by their nature lack continuity, and they are often administered by staff unfamiliar with local agroforestry traditions. The Pacific is littered with projects advanced in support of all sorts of good causes their collapsed remnants remain, like the military paraphernalia rust ing on beaches after World War II. One way to incorporate continuity into projects and to move beyond reliance on inadequate extension services is to form a centralized management system for smallholders (sometimes referred to as a plantation mode of management). Such a system has been successful in several instances, notably the efficient smallholder production of sugar so important in Fiji's economy and also in tobacco production in that same country (Eaton 1988a). Some other attempts have been less successful. The pros and cons of the approach have been cogently summed up by Hardaker et al. (1984a; 1984b) and Ward (1984).

Aside from problems common to any project-based introduction, a specific constraint to the full realization of the potential of agroforestry by institutional means relates to the disciplinary compart-mentalization that characterizes institutions concerned with land use, whereby - as the Director of ICRAF commented - "agriculture and forestry normally fall under different ministries or, if they are under the same ministry, under separate departments,' (Lundgren 1987, 44). Writing specifically of the forestry sector in the South Pacific, Watt (1980, 302-303) noted that "the separation of agricultural and forestry extension services encourages the impression that agriculture and forestry are mutually exclusive alternatives rather than complementary land uses." Following on from and related to this sectoral compartmentalization is each institution's imperative to maximize the individual component that is the focus of that institution. In contrast, as has often been observed:

The subsistence land user's strategy and aims are to use his labour and land resources to optimize, with minimum risk, the production of various products and services required to satisfy all his basic needs. The fundamental inadequacy of conventional-discipline-oriented institutions lies in the failure to acknowledge and understand these basic facts, strategies and aims, and in the inability to adapt to them. The aims, infrastructure, rationale and philosophy of these institutions, as well as the training of their experts, are geared to the maximization of individual components, be they food crops, cash crops, animals or trees. There is little understanding that the land user needs to share out his resources for the production of other commodities or services (Lundgren 1987, 46).

When maximization is aimed at commercial products, as it most frequently is in the Pacific, a set of sometimes contradictory processes comes into play. For example, attempts to produce cash crops while continuing to meet subsistence needs may bring agricultural involution if land is limited, or it may result in an extension of cropping onto marginal sloping lands as cash crops or cattle take over better lands. A specialization in commercial products may not be accompanied by any concomitant increase in labour availability or extension advice (often restricted to larger producers) on how to increase subsistence production (Ward 1986; Yen 1980b).

Even the Fiji-German Forestry Project, which commenced in the mid1980s, appears mainly focused toward facilitating export cash cropping, although its terms of reference suggest a broader approach that includes "providing ecologically sound advisory assistance in the fields of forestry and agroforestry in line with the social, cultural and economic requirements of target groups" (Tuyll 1988, 3). Consultants to the Fiji-German Forestry Project have also made holistic and wide-ranging recommendations, but the Project's current activities, as described earlier in this chapter, are concentrated on improving the production of ginger as a cash crop by introducing exotic trees to prevent erosion and replace artificial fertilizer.

This accomplishment is not to be decried, but the approach, distinguished by its introduction of and experimentation with exotic trees alley-cropped with a cash crop, does little to preserve existing agroforestry systems or to maintain a balance between commercial agroforestry activities and activities that could protect the existing subsistence base. One consultant recommended to the Project that "agroforestry and forestry extension should not attempt to remain with or return to pure forms of subsistence economy but focus on including profitable cash crops at low risks" (von Maydell 1987, 35). This recommendation does indicate an appreciation of the need to minimize risk, but both it and all the other consultants' recommendations to the Project fail to support strongly the maintenance of a viable subsistence base. Another consultant, who had been selected to identify suitable sites for demonstration plots for the Project, was asked to comment on the idea of putting greater emphasis on the subsistence aspects of agroforestry and of analysing existing local agroforestry systems as demonstration plots into which selected improvements could be introduced. He responded that it was quite unrealistic to expect either the Fiji Government or the German funding agency to support such an emphasis in place of an emphasis on using agroforestry as a way to improve monocultural cash cropping.

In summary, export crops, timber trees, and grazing under coconuts have been the continuing focus of almost all official agroforestry activities for the past century. Regardless of whether it has been the colonial or post-colonial agricultural and forestry departments or, re cently, international aid agencies, the focus has been almost exclusively on monocultural, often large-scale production for export or, in the case of timber and fuel-wood production, for import substitution. Even the intercrops are usually cash crops for export or local sale. Consequently, most indigenous wild species and the wide range of traditional cultivars have received little official promotion and have been the focus of only limited research. Few technical experts or development entrepreneurs know enough about traditional mixed agricultural systems and their component plants to be willing or able to promote their expansion or maintenance. It is not only projects intended to develop commercial agriculture and forestry that may displace or degrade traditional agroforestry systems; modern institutional agroforestry projects may themselves play the same role.

Agencies and educational institutions promoting agroforestry

However, there are also movements in support of traditional systems. The growing popularization and recognition worldwide of the value of the "wisdom of the elders" (Knudtson and Suzuki 1992) may motivate increased institutional attention to indigenous polycultural systems of agroforestry in the Pacific. This section provides information on several examples of such attention and on the institutions involved; mention has been made earlier of some of these, but they will be referred to here briefly again to provide a coherent single account.

All the major universities within the Pacific region (University of Guam, both of Papua New Guinea's universities, the University of the South Pacific in Fiji and its School of Agriculture in Western Samoa, University of Hawaii, and the developing francophone institutions in New Caledonia and Tahiti) support staff with interests in traditional matters, including agriculture, agroforestry, and the management of soil and vegetation. Rather than attempt a full listing of course offerings relevant to agroforestry to at least some degree, we note here only that, on the basis of current information at hand, the courses most directly focused on agroforestry are found within the Geography Department at the University of the South Pacific in Suva, Fiji, and the Department of Agronomy and Soil Science at the University of Hawaii in Honolulu. To the best of our knowledge, the University of Hawaii is distinguished by being the only university in the region to have a named Professor of Agroforestry, who is located in the Department of Agronomy and Soil Science. The Col lege of Micronesia in Pohnpei also has staff with active and direct interests in indigenous agroforestry.

Agroforestry promotion by the Fiji-German Forestry Project, a bilateral agency, has been described in the previous section. A different approach is followed by the South Pacific Forestry Development Programme, which is a multilateral 5-year project funded by UNDP, executed by FAO, and now based in Suva, Fiji. The Programme is concerned with forests and trees in 15 countries, so far particularly with forests in the larger countries, but atoll countries are making enquiries about coconuts and other multi-purpose trees. The role of the Programme is to stimulate activities and provide technical advice, not to operate activities itself. For instance, it facilitated the import of seeds of superior rattan from Malaysia for planting in Pacific forests in order to increase their non-timber production capability. Aside from technical advice, the Programme acts as a focal point for information about forests and trees and publishes the quarterly South Pacific Forestry Newsletter. It is also trying to organize the documentation of local knowledge on indigenous agroforestry, with studies planned or under way in Pohnpei, Fiji, Kiribati, Tuvalu, Tonga, and other island countries.

The Programme has worked cooperatively with the international NGO The Foundation of the Peoples of the South Pacific (FSP) on a project intended to develop sustainable forestry in local areas while slowing down or stopping rapid conversion of forests by large-scale industrial logging. This objective is based in part on selling small mobile sawmills to rural entrepreneurs and community groups so that they may develop small-scale but profitable and locally utilitarian logging, carried out in ways that avoid major environmental damage and that maintain the essential structure of the forest for traditional uses and ecological services.

A US Government project based in Hawaii is carrying out work related to several aspects of agroforestry in Hawaii, American Micronesia, and American Samoa. Called Agricultural Development in the American Pacific (ADAP), the project has provided agroforestry educational materials to all the public (land grant) colleges and universities in the American-affiliated Pacific. In association with the US Department of Agriculture and the US Forest Service, ADAP is also developing training programmes in agroforestry.

The Environment and Policy Institute of the East-West Center in Hawaii maintained a strong programme of research, seminars, and publication on agroforestry for several years during the 1980s (e.g., Djogo 1992; Nair 1984). Although agroforestry is no longer a principal focus of its work, the Institute remains a repository of a large volume of published and unpublished material on the topic.

Mentioned at the beginning of this chapter was the report (Clements 1988) of a technical meeting on agroforestry in tropical islands held at the Institute for Research, Extension and Training in Agriculture (IRETA), which is part of the University of the South Pacific's School of Agriculture in Western Samoa. IRETA is also involved in research projects to improve or strengthen atoll agroforestry in Kiribati.

In the Melanesian countries, with their comparatively larger natural forests, forest-resource inventories are under way or planned, generally as a cooperative, aid-funded project between the local Forestry Department and overseas technical personnel. The inventories are intended to provide the information base necessary for effective land-use planning and management, but now, unlike some past forest assessments, the inventory process includes collection of data on watershed vulnerability and on the indigenous ethnobotanical value of forest plants, as in the forest-resource inventory now being completed by the Vanuatu Forestry Department with technical assistance from the Queensland (Australia) Forest Service and the Division of Tropical Crops and Pastures of the (Australian) Commonwealth Scientific and Industrial Research Organization (CSIRO).

Finally, mention should be made of the work of ORSTOM, the French organization that promotes French scientific research in the third world, mainly in the tropics. With centres in the Pacific in Nouméa and Tahiti, ORSTOM has sponsored work not only related to many aspects of modern development but also to traditional cultural-ecological matters, for example, with specific relevance to agroforestry, the work on the cultivars of kava (Piper methysticum) in Vanuatu (Lebot and Cabalion 1986).

Smallholder farmers and the larger community, individual land holdings and the landscape: The agroforestry predicament

It will be clear to any reader who has come this far that the authors of this book believe that the maintenance and further dissemination of agroforestry offer great benefits to Pacific Island peoples. We also recognize, however, that there can be costs, that there can be disadvantages, connected with agroforestry, at least as perceived by individuals or by some parts of the community. The views opposing agroforestry, as well as the views supporting it, need to be included in the discussion. The advantages and disadvantages of agroforestry have been subject to considerable analysis in the literature, for instance by Arnold (1984) with particular regard to economic constraints and incentives, or by Budowski (1982) with regard to biological as well as to socioeconomic aspects. Greatly simplifying their detailed considerations, it can be said, on the beneficial side, that agroforestry is "ameliorative and protective" that is, it can increase, diversify, and sustain crop production while improving environmental stability (de la Cruz and Vergara 1987). On the detrimental side, trees can be or be seen to be competitive - either economically or biologically or both - with annual-crop production rather than complementary or supplementary. Consequently, when farm size falls below a certain level, farmers may forego tree products and services in favour of staple food-crop production. Or if land tenure is not secure, the time-lag in realizing benefits gained by planting trees may become a severe disadvantage, and trees will not be planted or pro tected. Sustainability is not common sense if there is no future return to the individual on today's investment in conservation or long-term production. Also, trees hinder mechanization, and their establishment or maintenance may require more manual labour than is easily available.

All this is no more than to say, as has been observed by many authorities, that agroforestry is not a panacea. Policies promoting agroforestry may demand costs from those who will not receive the benefits. Planning for agroforestry becomes very complex if it takes into consideration the incongruencies that may exist between ecological and economic accounting or if it seeks to resolve the issues of equity that promotion of agroforestry may cause across time and between social sectors or between the individual and the community.

Looking specifically at the Pacific, Künzel (1989, 24-27) has pithily expressed the doubts raised by Tongan farmers about agroforestry. While recognizing the antiquity of their agroforestry tradition and appreciating the benefits of trees, the farmers also made the following points, which Künzel relates as though they were in a letter from a Tongan farmer:

I have farmed my land for all my life, and it has been a constant struggle against trees. Before I can plant any crops, I have to remove the trees. The more trees I remove, the more crops I can plant. It is our saying that good farmers have neat fields, which means neatly cleared. Trees are for the lazy. True, there have been less trees in recent years than when I was young, but this does not change the wisdom of our forefathers.

Of course trees make the soil more fertile, but this takes many years. During the few years that I need to cultivate a certain area, trees only shade out my food crops. I learnt to live with that while I had to clear my field by hand. Ecologists always marvel at the systems I developed. But now that I can hire a tractor easily, I clear my trees with enthusiasm. Believe me, it improves the yield.

The building materials, perfumes, medicines and dyes I get from trees are old-fashioned and ineffective.... I like those products from overseas.... Nobody is boiling tree bark to fight a stomach ache in New York. Pills work better, and I want to use them too.

If I cut a tree, I reap the benefits completely and immediately. If I plant a tree, who knows the person who will benefit? Probably not me, and maybe not my children.

Young trees are more vulnerable than young onions! One week when I have no time to water, or a runaway cow while I sleep, and all is gone. Not to mention fire. I do not like risks like that.

Finally, all the arguments about the ecological benefits of trees are de finitely true. But when the day starts and I have to decide whether or not to plant cassava or trees for my family, a day planting cassava is worth more.

These views of a fictitious Tongan farmer clearly provide reasons for current agrodeforestation. They also remind us that the first purpose of agricultural or agroforestry systems is the production of goods that have value for the farmer, not the maintenance of agro-ecosystems in a healthy and sustainable state. In today's Pacific, the production of goods with value is affected not only by rapid population growth in many countries but also by a whole range of nontraditional factors such as mechanized technology, monetization, a never-ending search for new export products, individualization, wage-paying employment, and new forms of communication in combination with advertising. Agroforestry often fits uneasily with these changes. Although everyone agrees that agroforestry has great contributions to make to biological and physical sustainability, many people have doubts about its social and economic benefits. As Künzel (1989, 27) says: "Clear and convincing reasons why agroforestry will benefit the individual farmer need to be found."

The relationship between the individual farmer and the larger community is in many ways analogous to the relationship between an individual landholding and the landscape in which it is situated. To maintain the landscape in good health, it is not necessary that every landholding, every stretch of land, contain trees, just as every farmer need not be an agroforester- but it is necessary that there be sufficient trees in the right places, at the least on sloping land and along streams. However, the predicament in the Pacific, as in much of the world, is that the landscape is not a unit of management even though it best expresses the integration of environmental processes and relations at the regional level. Governments everywhere are only now starting to grope toward the concepts of landscape ecology, a groping that may be motivated by demands to make development sustainable but that is - to be more optimistic- facilitated by the modern tools of computer databases and remote sensing of the environment. None the less, bringing the science of landscape ecology into land utilization and management is a difficult and slow process, for landscape ecology cuts across economic sectors and encompasses traditionally distinct agencies or ministries, such as agriculture, forestry, and urban planning (Naveh and Lieberman 1984). When this inertia is combined with population growth, with economic pressures, and with the complex issues of land tenure and of multiple control over land use, it is no wonder that implementing conservation measures is often difficult or that agrodeforestation is taking place. Even within this context, however, there are ways to plan for agroforestry, to encourage it, and to slow down agrodeforestation. These will be discussed later in this chapter. First, we will turn to a brief consideration of the basic component of Pacific Island, or any other, agroforestry systems - the trees themselves.

The component trees

The case-studies in this book illustrate the richness of the existing heritage of polycultural agroforestry possessed by all Pacific societies, both rural and urban, both predominantly subsistence and strongly focused on cash-crop production. By definition, it is trees and tree-like species that are fundamental to these almost endlessly variable systems. If the existing systems are to be preserved and utilized in future agroforestry development, it is of course necessary to preserve the trees themselves as well as to have an inventory of information about them. Analysis of available information from the literature and field surveys has led to the identification of 419 cultivated and/or wild tree or tree-like species or groups of closely related species that are of widespread or localized economic, cultural, and ecological importance in Pacific Island rural and urban agroforestry systems. As these plants have already been shown to be useful components of local agroforestry systems, they merit consideration for future use, but to discuss here, or even to list, all 419 species or groups of species would be impractical as well as daunting to all but the most botanically-minded of readers. In place of that overload of information, we discuss here only a few arbitrarily selected details. In the Appendix, the characteristics and uses of 100 Pacific Island agroforestry trees are described more fully so as to provide the interested reader with a picture of the rich variety of species available.

Selection of agroforestry resources

Whether on high islands, with their often rich soils, varied habitats, and low population densities; on the harsh atolls almost without soil and short of water; or in home gardens in densely settled urban areas and monocultural rural agricultural areas, Pacific Islanders have selected for incorporation into their agroforestry systems a wide range of tree and tree-like species that meet their particular en vironmental and cultural needs. The use of these species is the cumulative result of a selection process that has occurred over thousands of years, beginning in the ancestral homelands of today's Pacific Islanders in SouthEast Asia and the archipelagic areas of Indonesia, the Philippines, and Papua New Guinea, whence valuable cultigens and accumulated knowledge of wild species were transferred to the smaller Pacific oceanic islands.

After they arrived, early settlers domesticated previously unknown indigenous species or else incorporated them as wild species into agroforestry systems prior to European contact (Yen 1990). Similarly, post-Europeancontact introductions, including food plants, timber trees, and ornamentals have been tested, selected, and incorporated into today's systems to such an extent that the undiscerning visitor or agricultural "expert," and many of the current generation of islanders, believe the introduced trees to be traditional or even indigenous. The integration has been carried to such an extent that the status of many species remains unclear as to whether they are local domesticates or indigenous plants, aboriginal introductions brought to the islands by successive waves of Pacific settlers, or early postEuropean-contact introductions.post-European-contact

Growth form

Of the 419 agroforestry plants, approximately 329 are classified as large or small tree or tree-like species, whereas approximately 90 are smaller, more shrub-like perennials, which constitute common fixtures in Pacific Island agroforestry systems. The division into shrubs and trees is, of course, somewhat arbitrary because some species, depending on the environment, the variety, or cultivar, can be either shrubby or tree-like. For example, the species Hibiscus tiliaceus, Pipturus argenteus, and Vitex trifolia are all found as both shrubs and trees. Similarly, groups of highly variable, closely related species of the genera Leucosyke, Pandanus, Pittosporum, Psychotria, Solanum, and Timonius are all represented by both shrubs and trees.

A very limited sampling of shrubs and shrub-like species would include the indigenous, often wild, plants such as Acalypha insulana, Dodonaea viscosa, and Pemphis acidula. Important food or beverage plants in the "shrub" category include the bush hibiscus spinach (Hibiscus manihot); sugar cane, which, along with the related species with the edible inflorescence (Saccharum edule), forms tree-like stands in Pacific gardens; kava (Piper methysticum), the important social bev erage of Vanuatu, Fiji, Polynesia, and Pohnpei; betel pepper (Piper belle), which is cultivated in western Melanesia, high-island Micronesia, and by the Indian community of Fiji. Shrubby handicraft plants of a widespread importance include paper mulberry (Broussonetia papyrifera), which extends from the highlands of Papua New Guinea, where it is used in the production of string bags (bilum) and a wide array of loin cloths, to Polynesia, where it is used in the production of the ceremonially important tape cloth, an increasingly valuable source of cash income from the tourist industry; and a wide variety of Pandanus species or cultivars, whose leaves are used in plaited ware such as fine mats, thatching, hats, baskets, and other items of cultural and economic importance. The rattan palms (Calamus spp.) are important in Papua New Guinea for handicrafts and general construction. The introduced annatto (Bixa orellana) is widely planted for its seeds, which yield a red dye, and has been a minor smallholder export crop in Western Samoa.

"Protective" or magical plants, established in or around active garden areas to ward off evil spirits, include Cordyline fruticosa, Coleus scutellarioides, and Euphorbia fidjiana, which may be present only in Fiji and Tonga.

Origin or antiquity status

Of the 419 species, approximately 172 (41 per cent) are probably indigenous to most islands where they are found; 13 are probably aboriginal introductions; 40 either indigenous or aboriginal introductions; 147 (35 per cent) are recent post-European-contact introductions; 17 are indigenous in some areas but recent introductions in other areas; 17 are both aboriginal or recent depending on the area; and 13 are indigenous, aboriginal, or recent depending on the area. It is, however, extremely difficult to determine whether some species were indigenous or introduced by early settlers, and their status may be different in different areas. For example, Terminalia catappa may be both indigenous and/or an aboriginal introduction to some areas, whereas it is probably a recent introduction into Hawaii.

The 172 possibly indigenous species include plants from Pacific Island coastal strand forest, mangrove forest, marsh or riparian and indigenous lowland, and montane forest, as well as pioneer species common in fallow vegetation. All are important components of plots of relatively undisturbed vegetation bordering or included in agroforestry systems and are often deliberately protected or planted as in tegral components of rural gardens or plantations and urban gardens because of their cultural and ecological utility.

Of the more than 80 species that may be aboriginal introductions to at least some islands, the most widespread and culturally important species are food plants. These include the ubiquitous "tree-of-life" coconut palm, breadfruit, the many traditional banana and plantain cultivars, sago palm, edible pandanus cultivars, sugar cane, bush hibiscus spinach (Hibiscus manihot), and a variety of fruit and nut trees and other supplementary food crops.

Also, almost certainly aboriginal introductions to most areas were the important social masticant and beverage plants, betel-nut palm (Areca catechu) and betel pepper (Piper belle), kava (Piper methysticum); the important fibre and handicraft plants, paper mulberry (Broussonetia papyrifera) and a wide range of Pandanus species or cultivars; the candlenut (Aleurites moluccana) and the perfume tree (Cananga odorata), so highly valued for scenting coconut oil; and a wide range of other plants of cultural value for general construction and boat building, medicines, dyes, poisons, fibre, perfumes, decorations, magic, and other purposes.

The recent post-European-contact introductions include important longestablished food trees such as the papaya, mango, avocado, a range of Citrus species, guava, soursop, sweet sop, and three important commercial banana clones; the ubiquitous perennial chill) pepper; food trees of more localized importance, such as the jakfruit, tamarind, horseradish, or drumstick, tree (Moringa oleifera), and the Indian bay, or curry leaf (Murraya koenigii), all of particular importance to the Indian community of Fiji; and a wide range of fruit and nut trees of minor importance. (See the Appendix for further details on some of these plants.)

Recent introductions of commercial importance include tea, coffee, oilpalm, Para rubber, cocoa, and new, often high-yielding, cultivars or clones of plants such as the coconut palm, bananas, and citrus trees; the exotic timber species including Cordia alliodora, Eucalyptus deglupta, Caribbean pine (Pinus caribaea), and West Indian mahogany (Swietenia macrophylla); and a number of other species that are, or will be, important sources of income to some countries in the region.

Other recent introductions include the common bamboo, cotton, kapok, the neem tree (Azadirachta indica), allspice and bay rum tree (Pimenta doica and P. racemosa), and castor bean (Ricinus communis); a number of fastgrowing multi-purpose trees, most notably

Leucaena leucocephala, madre de cacao (Gliricidia septum), Pithecellobium dulce, Sesbania grandiflora, Erythrina spp., and Jatropha curcas; a wide range of exotic ornamental trees and shrubs, the most common including the orchid tree (Bauhinia spp.), bougainvillea, flowering cassias, poinciana, or flamboyant, banyans (Ficus spp.), gardenias, lantana (Lantana camara), Persian lilac, or China berry (Melia azedarach), mock orange (Murraya paniculata), the ubiquitous plumerias, or frangipanis, monkey-pod, or rain tree, and the African tulip tree; and a number of widely cultivated ornamentals, such as Acalypha amentacea, Codiaeum variegatum, Gardenia taitensis, Graptophyllum pictum, hedge panaxes (Polyscias spp.), and Pseuderanthemum spp. A number of palms, although recent introductions into the eastern parts of their range, are possibly indigenous to or even domesticates of Melanesia or Polynesia.

Range

All 419 plants or groups of plants are found in the diverse agroforestry environments of either high islands or the larger limestone islands, with no species of widespread or major local importance confined to the atoll environment. Only 83 species (20 per cent) are commonly found on Pacific atolls, which reflects both the poverty of the indigenous and exotic floras on atolls as well as the critical importance of all agroforestry species that do succeed in growing on atolls. Of these 83 species, 30 are ubiquitous Pacific strand or mangrove species, 26 are introduced ornamentals, with the balance consisting of aboriginally introduced plants, recently introduced food plants, and a small number of recently introduced timber, fuel-wood, or other useful plants.

Husbandry

Of the 419 agroforestry plants, some 113 are found almost exclusively in a wild state, 127 are cultivated, and about 179 are both cultivated and wild. The 113 exclusively wild or uncultivated species are made up of indigenous coastal strand, mangrove, and inland-forest species commonly protected in forest stands. Also included are several pioneering species that, although not generally protected when clearing new garden plots, are encouraged through selective weeding and become dominant components of the young fallow forest.

The strictly cultivated species, on the other hand, are almost entirely exotic cultigens. These include major and minor fruit or food species, ornamentals, exotic timber or fuel-wood species, major and minor export crops, and a number of handicraft, medicinal, spiritual, and other multi-purpose species.

Species found both cultivated and in a wild state include a wide range of indigenous species, which are also deliberately cultivated, as well as cultigens, most of which are exotic, that have escaped or become naturalized and are seen to have significant cultural or ecological value. Among the most commonly cultivated indigenous species are a large number of ubiquitous coastal strand plants valued for their multi-purpose utility. Some trees indigenous to Melanesia in the genera Acacia, Agathis, Albiza, Araucaria, Endospermum, and Terminalia have been deliberately planted or experimented with as timber species for reforestation programmes.

Species that are both cultivated and either commonly naturalized or indigenous include food species such as coconut, breadfruit, sago palm, sugar cane, Saccharum edule, and Cordyline fruticosa. Other food trees in this category are Adenanthera pavonina, Barringtonia spp., Burkella obovata, Canarium spp., Castanospermum australe, Corynocarpus similis, Dracontomelon vitiense, edible figs (Ficus spp.), Finschia chloroxantha, Gnetum gnemon, Musa troglodytarum, Pangium edule, Pometia pinnata, and Syzygium malaccense. Other cultigens found wild are candlenut (Aleurites moluccana), betel-nut and betel pepper, perfume tree, or ylang ylang (Cananga odorata), Schizostachyam glaucifolium, Sida fallax, and Solanum uporo. Most of these species are commonly found in mature fallow forests, where they are either remnants of former cultivation or naturalized escapes or volunteers that have either been protected by agriculturalists or, in some cases, may have been components of the indigenous flora.

Recent post-European-contact introductions that have become most widely naturalized, commonly as escapees from cultivation, include the perennial chill) pepper, papaya, lantana, leucaena, and guava. Other commonly naturalized species include Acacia farnesiana, Brassaia actinophylla, Cassia alata, citrus trees, derris root, mango, Melia azedarach, and castor bean. There are others of more localized importance, many of which are either selfsown or have grown from seeds discarded by humans. Although growing wild, most of these species cannot be considered fully naturalized as they grow mostly in disturbed sites and are generally displaced by indigenous vegetation in mature fallow forests.

Relative importance

In terms of relative importance, 56 of the 419 species are considered to be of major agroforestry importance in many island groups in terms of abundance and/or ecological importance and cultural utility. These include:

  1. the staple food species breadfruit, coconut, four banana or plantain clones, and sago palm;
  2. the fruit-trees soursop, papaya, three citrus species (orange, mandarin orange, or tangerine, and rough lemon, or kaffir lime, C. hystrix), Tahitian chestnut, mango, oceanic Iychee (Pometia pinnata), guava, Polynesian plum, or hog plum (Spondias dulcis), and Malay apple (Syzygium malaccense);
  3. supplementary food species including the perennial chill) (Capsicum frutescens), bush hibiscus spinach (Hibiscus manihot), and sugar cane;
  4. the important social beverage kava and the alkaloid masticant betel-nut;
  5. the important handicraft plants paper mulberry (Broussonetia papyrifera) and pandanus cultivars (Pandanus spp.);
  6. multi-purpose aboriginally-introduced or indigenous species including candlenut (Aleurites moluccana), Cananga odorata, Cordyline fruticosa, and the common hibiscus (Hibiscus rosa-sinensis);
  7. the primarily indigenous species Bischofia javanica, Calophyllum inophyllum, Casuarina equisetifolia, Cordia subcordata, Erythrina variegata var. orientalis, Fagraea berteriana, Ficus prolixa, Glochidion spp., Hibiscus tiliaceus, Macaranga spp., Morinda citrifolia, Pandanus tectorius, Pipturus argenteus, Premna serratifolia, Terminalia catappa, and Thespesia populnea;
  8. the mangrove species Bruguiera gymnorhiza and Rhizophora spp.;
  9. recently-introduced multi-purpose species including common bamboo (Bambusa vulgaris), kapok (Ceiba pentandra), and Leucaena leucocephala;
  10. the common ornamental and shade trees including poinciana, or flamboyant (Delonix regia), frangipani, or plumeria (Plumeria rubra), and the rain tree, or monkey-pod (Samanea saman);
  11. the living fence or hedge plant, panax (Polyscias guilfoylei); and
  12. the important export crop of cocoa.

Table 11 Tree or tree like species or groups of closeb related species of mqior agroforestry importance in certain localities and/or widespread supplementary ecological or cultural importance (See Appendix for detailed information on some of the species listed here)

Acalypha amentacea vars. Diospyros samoensis Musa troglodytarum
Adenanthera pavonina Diospyros spp. Musa cultivars
Aglaia saltatorum Dracontomelon vitiense Myristica spp.
Aglaia spp. Dysoxylum forsteri Neisosperma
Albiziafalcataria Dysoxylum spp. oppositifolium
Albizia lebbeck Elaeis guineensis Nerium oleander
Alphitonia spp. Flaeocarpus spp. Pandanus conoideus
Alyxia spp. Endospermum spp. Pandanus dubius
Annona squamosa Eucalyptus deglupta Pandanus julianettii
Antiaris toxicaria Eucalyptus spp. Pangium edule
Araucaria cunninghamii Euadia hortensis Parinari glaberrima
Araucaria heterophylla Euodia spp. Pemphis acidula
Artocarpus heterophyllus Ficus benjamina Persea americana
Artocarpus mariannensis Ficus copiosa Phaleria spp.
Barringtonia asiatica Ficus dammaropsis Pinus caribaea
Barringtonia edulis Ficus obliqua Pisonia grandis
Bauhinia monandra Ficus tinctoria Planchonella spp.
Bougainvillea spp. Ficus wassa Plumeria obtusa
Burkella obovata Ficus spp. Polysciasfruticosa
Cajanus cajan Garcinia sessilis Polyscias scutellaria
Canarium indicum Garcinia spp. Prichardia pacifica
Canarium spp. Cardenia jasminoides Pterocarpus indicus
Cassia alata Cardenia taitensis Rhus taitensis
Cassiafistula Carugafloribunda Saccharum edule
Cassiagrandis Cliricidia sepium Santalum spp.
Casuarina oligodon Gnetum gnemon Scaevola sericea
Cerbera manghas Grewia crenata Schizostachyum
Citrus aurantiifolia Heliconia indica glaucifolium
Citrus aurantium Hernandia Securinega flexuosa
Citrus grandis nymphaeifolia Sida fallax
Citrus limon Homalanthus spp. Solanum melongena
Codiaeum variegatum lntsia bijuga Spathodea campanulata
Coffea arabica Ixora spp. Sterculia spp.
Coffea canephora Jatropha curcas Swietenia macrophylla
Coleus scutellarioides Kleinhovia hospita Syzygium cumini
Commersonia bartramia Lantana camara Syzygium richii
Cordia alliodora Lumnitzera littorea Syzygium spp.
Cryptocarya spp. Melaleuca leucadendra Tamarindus indica
Cyathea spp. Melastoma Tarennasambucina
Cycas circinalis malabathricum Terminalia spp.
Decaspermum fruticosum Metrosideros collina Tournefortia argentea
Denrocnide spp. Micromelum minutum Trema spp.
Derris spp. Moringa oleifera Vitex spp.

Also of major agroforestry importance, although not quite as widespread or as abundant in indigenous agroforestry systems, are another 125 species or groups of closely related species. These include:

  1. important food species;
  2. multi-purpose aboriginal introductions;
  3. indigenous species that are protected in tree groves and fallow areas and/or planted in agricultural areas;
  4. recently introduced timber, fuel-wood, and multi-purpose species;
  5. important export crops; and
  6. a wide range of commonly cultivated ornamentals. These species are listed in table 11 and details of some of them are provided in the Appendix.

If all the individual plants belonging to the over 400 species of agroforestry trees and shrubs should suddenly vanish, Pacific Island landscapes and life would be dramatically altered, and a near irreplaceable resource would have been lost. The manifold utility of the agroforestry species has been stressed throughout this book. They fulfil cultural, economic, and ecological functions, with the majority of species serving multiple purposes, as shown by a study of coastal species commonly found in Pacific Island agroforestry systems. On average, each species had 9.2 purposes or uses, ranging from two reported uses per species to as many as 121 for the coconut. These uses did not include the strictly ecological functions such as shading, protection from wind and sand and salt spray, erosion and flood mitigation, coastal reclamation, provision of animal and plant habitats, or soil improvement (Thaman 1989c).

Encouraging agroforestry

We have already suggested the paradox to be seen in the concurrent existence, on the one hand, of a rich indigenous agroforestry resource containing both a great variety of component trees and an extensive and sophisticated body of knowledge and, on the other hand, institutional proposals for the introduction and development of agroforestry. This incongruous situation is partly the result of the tendency of development specialists to promote the currently popular "solution', before the problem has been carefully investigated. It is also, as we have argued earlier, simply a lack of knowledge in institutional circles with regard to indigenous agroforestry. Even at the recent technical meeting on Pacific Island agroforestry held in Western

Samoa (Clements 1988), few of the speakers (mostly Pacific Islanders) appeared to recognize that in areas of shifting (swidden) cultivation the forest had any function other than fallow. The intricate, usually intentionally created, multi-purpose character of secondary forests described in this book received little attention - except for the discussion of Pohnpei's "agroforests" and an occasional mention of fruit or nut trees elsewhere.

As the combined pressures of population growth and cash cropping bring shortened fallows, declining soil fertility, and increasing cultivation on marginal lands to many of the Pacific's high islands, institutionally promoted systems such as contour hedging with Leucaena, Gliricidia, or Calliandra can help maintain soil fertility and lessen erosion, as has been clearly demonstrated elsewhere in the world. It should be recognized, however, that the process of replacing the indigenous agroforest with a few, usually exotic, species, useful though they are, is part of the pattern of simplification of subsistence systems in the Pacific (Ward 1986, 218) and loss of general biodiversity that is often seen as one of the great ecological threats of the present world. While the forces driving toward simplification and loss of natural forest and indigenous agroforest cannot be easily slowed, some interventions may help.

Appreciation of indigenous agroforestry

If indigenous agroforestry is to be encouraged, the character and value of the many existing agroforestry systems need to be more widely and deeply appreciated. As it is to that purpose that the material in this book is largely dedicated, we will only recapitulate a few of the worthwhile and significant technologies and characteristics that need to be kept in mind with regard to Pacific Island agroforestry.

The composite extended nature of agroforestry systems
In the land-use strategy of Pacific Islanders almost all types of agricultural or wild lands are viewed as integral components of a larger agroforestry system. This view should be taken into consideration during development planning. Types of agricultural and wild lands found within the matrix of agroforestry systems include: native or secondary forest stands; sacred groves; monocultural woodlots, orchards, or coconut, cocoa, or oil-palm plantations; mixed tree- and ground-crop shifting agricultural plots; home gardens; and even small patches of grassland or adjacent mangrove or coastal forest.

Just as an entrepreneur wisely diversifies his assets, the Pacific agroforester diversifies his or her land use, both within individual plots or areas, or within the whole set of landholdings to which the agroforester or his or her extended family have usufruct, individualized, or even freehold (fee-simple) rights. Thus, any suggested changes or innovations to a given part of the system will affect the system's aggregate utility, stability, and sustainability.

Home gardens in agroforestry systems
Throughout the Pacific, home gardens are integral components of agroforestry strategies. They may exist as adjuncts to wider rural or village agroforestry systems, or stand on their own as urban home gardens. They are also among the most diversified and most dynamic agroforestry systems, often serving as both refugia for endangered or culturally-important traditional species, as well as avenues for the introduction and testing of new exotics. As suggested by Thaman (1988g), the promotion of urban and home gardening may be among the most cost-effective and culturally and ecologically appropriate means of satisfying many of the commonly stated objectives of national development plans.

High interspecies diversity
All traditional agroforestry systems, from the highlands of Papua New Guinea to the smallest atoll countries, are truly polycultural, exhibiting a high degree of interspecies diversity incorporating a wide range of cultivated, protected, and wild indigenous and exotic species. The numbers of species range from some 75 commonly encountered on atolls, which have among the poorest floras on earth, to over 300 widespread species found in the agroforestry systems of the larger islands of Fiji, Vanuatu, the Solomon Islands, and Papua New Guinea.

High intraspecies diversity
There is, additionally, a high degree of intraspecies diversity for most traditional tree and non-tree cultigens and for many recently-introduced cultigens. A wide range of named, locally-differentiable cultivars, varieties, or "land races" (Harlan 1975) exists for a given species. Within a given species, these cultivars have variable yield characteristics and seasonality, thus spreading yield distribution and seasonal surpluses more evenly. Different cultivars also have differential resistance to pests and diseases, tropical cyclone damage, salt water incursion and salt spray, and drought; differential ecological tolerance ranges in terms of adaptability to different soil types, shade, and hydrological regimes; and differential utility (for example, some coconut cultivars are used purely as drinking nuts, some for the flesh, and some for the large shells or the coir, which can be used for vessels or for cordage, respectively). As indicated in the case-studies, there is a particularly great range of cultivars of the traditional staple tree crops of coconut, breadfruit, pandanus, and banana and plantains; traditional perennial supplementary crops like sugar cane, sugar cane inflorescence (Saccharum edule), and hibiscus, or bush, spinach (Hibiscus manihot); traditional non-food plants, such as kava (Piper methysticum) and betel-nut (Areca catechu); recent introductions such as papaya, avocado, and a range of Citrus species; and many indigenous or, in some cases, possible aboriginally-introduced food and non-food species, such as Pandanus tectorius, Pometia pinnata, Hibiscus tiliaceus, Erythrina variegate, and a number of species of Ficus, which seem to have undergone systematic selection by the indigenous inhabitants of various islands and are now represented by a range of recognized cultivars or semi-domesticated genotypes.

Importance of wild food resources
Wild food resources are among the most important products of agroforestry systems, particularly in Melanesia, where extensive stands of primary and secondary forest remain. Most notable among these are a great diversity of wild yams, ferns, fungi, fruits, nuts and leaves and birds, frogs, snakes, grubs, insects, fin fish, eels, freshwater prawns, and other foods that are found within agroforestry zones. Even in grassland areas and on atolls, wild yams, ferns, wallabies, kangaroos, birds, edible plants, and crabs constitute important wild or emergency food resources. Apart from being nutritionally important, these wild products also constitute important low-capital-input, lowrisk seasonally abundant cash crops. When forests and trees are removed, however, in the process of agricultural intensification or replanted with exotic silvicultures, species diversity of wild plant and animal foods decreases significantly (Clarke 1965; Watling 1988), thus constituting a loss of food resources and cash and non-cash income.

Live fencing, hedges, animal pens, and boundary markers
Of almost universal importance is the use of live fencing, hedges, and boundary markers, which serve as productive and ecologically valuable components of agroforestry systems. The use of trees or shrubs for such purposes includes: living fence posts for modern barbed-wire fencing on livestock development schemes or around crops susceptible to damage by livestock; boundary markers demarcating boundaries of land parcels belonging to different landholding groups, between the active gardens of individual gardeners, or between individual plots within a given gardener's active garden area; wind-breaks or barriers to protect coastal gardens from salt-laden air and resultant physiological drought and excessive salinity; and living animal pens, either with or without wire.

Integration of livestock and agroforestry
A common characteristic of most traditional agroforestry systems is an integration with, rather than a segregation or dissociation of, agroforestry from livestock husbandry. Domesticated pigs foraged in secondary forests and woodlands during the day and were also fed forest products to supplement their root-crop diet. Moreover, traditional systems often provided a means of storing surplus crop production for times of scarcity or feasting, such as in highland Papua New Guinea, where pig production served this purpose (Bayliss-Smith 1982). The increasing institutional emphasis on smallholder cattle projects, modern piggeries, and battery-raised poultry seems to be riddled with failure, compared to traditional foraging, waste-disposal, or tethering systems. Notable exceptions include cattle under coconuts, which seem to be far less productive than crops under coconuts (Schirmer 1983, 104), and cattle under timber trees. Evidence from throughout the Pacific shows that the modern intensive systems are extremely difficult to maintain, and in the case of smallholder commercial grazing systems, have caused widespread destruction of food crops and agrodeforestation (Grossman 1981).

Value of trees as staple crops
Throughout the Pacific, trees or tree-like crops constitute major or supplementary staples. The coconut, for example, is the major source of dietary energy on atolls, and in some coastal areas. Bananas and plantains, as a group, are perhaps the most important staples, after taro, cassava, and sweet potato, in terms of total calories provided to Pacific societies, with breadfruit being a particularly important seasonal staple in most areas. Other, lesswidespread or major seasonal staples include edible pandanus (Pandanus spp.), Tahitian chestnut (Inocarpus fagifer), and Barringtonia and Canarium spp. Such renewable (without replanting) energy resources are of particular impor tance, given that in parts of Melanesia, energy deficiency and not protein deficiency is the most serious nutritional problem and that increasing food dependency on imported white flour and rice, in both urban and rural areas, is a serious economic and nutritional problem (Coyne 1984; Hayward and Nakikus 1982; Jeffries 1979).

Trees as major sources of micronutrients and fibre
Along similar lines, fruit-trees, nut trees, leaf-providing trees such as Hibiscus manihot, Ficus spp., Gnetum gnemon, and Moringa oleifera, and sugar cane are some of the best sources known of vitamins, minerals, fibre, and other micronutrients, so vital to optimum health, but often lacking in modern urban diets. The increased consumption of such foods could significantly address increasingly widespread incidences of micronutrient deficiencies, such as anaemia and vitamin-A and vitamin B-complex deficiencies.

Medicinal importance of trees
The strategic importance of agroforestry systems as sources of the medicines used by many rural and urban people, most of whom have no access to or could not afford imported medicines, cannot be stressed too strongly. The importance of medicinal plants in agroforestry systems is underlined by the very high numbers of medicinal plants found in home gardens, in garden areas, and in protected forests or tree groves. Some two-thirds of medicinal plants are trees or shrubs (Thaman 1988g; 1989c).

Tree products of subsistence and commercial value
In addition to foods and medicines, trees are the sources of a very wide range of construction and handicraft materials, cordage, dyes, poisons, scents, oils, decorations, and body ornamentation, and an almost endless range of utilitarian products of considerable subsistence and commercial value. Such products account for a significant proportion of real rural and poor-family urban incomes, incomes that are little affected by inflation or the vagaries of an unstable world economy.

Magico-religious and spiritual importance of trees
The magico-religious or spiritual importance of trees, both in terms of sacred groves or individual living trees and the use of plant parts in magic or ceremonies, particularly in Melanesia, is an important aspect of Pacific Island agroforestry, and a reason for preserving trees. Ceremonial uses of plants include ceremonies or rituals associated with death, war and peace, human sacrifice and cannibalism, circumcision or coming of age, canoe making and launching, fishing, planting cycles, prayer sessions, as well as species serving as symbols or totems or planted in sacred groves or burial grounds and mediums for communicating with spirits or gods. Others are associated with times of revelry or are used in the production of baskets, mats, and other articles reserved for ceremonial exchange or dress. Many plants also feature prominently in Pacific Island mythology, legends, songs, riddles, proverbs, and cosmogony. As recorded in Ethnobotany of the Samoans (Setchell 1924), plant names were given to gods or vice versa, and songs and legends have developed around them and the "heroes, families, or villages, etc. they represent." One particular Samoan text of the "battle of trees and stones" enumerates between 70 and 80 tree names. Also very closely related to ceremony and ritual are magic and sorcery, which are still very strong in the Pacific, especially in Melanesia and parts of Micronesia. Plants are integral to such practices, which include magic related to love, exorcism of evil spirits, gardening, and death (Thaman 1989c).

Protection, pruning, and pollarding of trees during garden preparation
Protection (not killing) of trees, although often with severe pruning or pollarding, in the preparation of new gardens, or when replanting old gardens, is a widespread practice that ensures the stability of agroforested landscapes. The ringbarking or removal of trees, often to allow for total tillage or ploughing, so commonly promoted to maximize the production of a single target crop, is a major factor in agrodeforestation. Not only does severe pruning or pollarding of valuable species such as Bischofia javanica, Hibiscus tiliaceus, and Macaranga spp., and fruit-trees such as mango, citrus, and Pometia pinnata, open up prospective garden areas to sunlight and provide additional organic material, and ash if gardens are burned, it also ensures renewed vigour of the regrowth and provides firewood and trellising for crops. Even ringbarking, if done in a particular way, will stimulate shooting from the base, thus not permanently eliminating a given tree from the garden area.

Minimum tillage
The practice of minimum or restricted tillage and the use of appropriate hand tools, which minimize soil degradation and favour the survival of trees, are universal and beneficial features of shifting agro forestry systems. It is especially important in the humid tropics and in steep terrain where heavy rainfall and high temperatures promote rapid leaching and oxidation and/or accelerated erosion of loosened soils. The practice of minimum tillage, where holes or mounds are prepared or the soil loosened only where plants are to be planted, is, thus, a positive, but labour-intensive, alternative to complete tillage or ploughing. Although appropriate in laboursurplus areas throughout the Pacific, the practice is being abandoned in areas such as Tonga and the Cook Islands, where ploughing and disc harrowing is encouraged in commercial monocultural production of root crops, vegetables, and fruits such as water melon and pineapple.

Mulching, composting, and fertilization
The use of leaves and other plant parts for mulching, composting, and fertilization in the cultivation of major staples and tree crops, particularly on atolls, is a well-developed practice that depends on the maintenance of a range of often minor non-food plant species found in agroforestry systems. In Melanesia, intensive mulching systems using the leaves of Pometia pinnata are employed in the East Sepik area of Papua New Guinea (Allen 1985), and a wide range of leaves are placed in planting holes of yams and taro or dug into the mud in irrigated taro gardens in Vanuatu, often to aid in insect or pest control, to enhance taste, or as part of diverse, but possibly functional, magicoreligious rites. Throughout the Pacific, various societies, such as the Koita in the arid savanna areas near Port Moresby, Papua New Guinea, deliberately mulch yam, cassava, taro, and sweet potato gardens with dried grass and other plant remains to reduce moisture loss and erosion and to improve soil fertility and yields. It is in Micronesia, in the harsh, almost soil-less atoll environment, that mulching systems are the most intensive. The most sophisticated systems are those used in the production of the ceremonially important giant swamp taro (Cyrtosperma chamissonis), with a wide range of leaves, some specially processed and dried, being mixed with scarce topsoil from under deliberately protected plants, and applied in woven pandanus or coconut frond baskets. Pulverized, rooted coconut logs, seaweed, pumice, and tin cans or rusting iron are also occasionally added to provide additional benefits, trace elements, or to alter the soil pH to free trace elements. The use of kitchen waste, animal manure, and ash is increasingly widespread in urban and home gardens, with mill-mud (waste from the sugar refining process) being used in some areas of Fiji (both on sugar-cane farms and in home gardens). Commercial inorganic fertilizers are only rarely, if ever, used in most existing agroforestry systems.

Selective weeding
The practice of selective weeding to encourage culturally useful and important pioneer fallow species, particularly in the later stages of fallow, is a widespread characteristic of most agroforestry systems. Through such practices, fallow species and non-planted volunteer seedlings of other useful plants, such as mangos, papaya, citrus, and nut trees, can be selectively encouraged during the latter stages of the gardening cycle, thus hastening and controlling or managing the formation of productive tumbledown fallows and fallow forests or orchards, often including still productive stands of cassava, yams, taros, and bananas and plantains.

Permanent polycultural orchards
The establishment of permanent polycultural orchards, which often become mixed fallow forests, is a widespread practice. Such orchards, which may be almost entirely planted or may be mostly mature fallow forests in old gardening areas where useful trees had been planted, are protected as permanent productive features in the agroforestry landscape. The orchards can take the form of communal groves adjacent to villages or they can be individually-controlled groves in distant garden areas. The species composition varies from area to area, with a wide range of fruit, nut, and other useful trees represented, e.g. the Pandanus-breadfruit-Gnetum-Ficus wassa orchards described for the highland fringe of Papua New Guinea.

Planting, transplanting, or protection of indigenous species
It is a widespread practice to plant, transplant, or protect indigenous species in agricultural areas and in urban and home gardens. Coastal strand species such as Barringtonia asiatica, Calophyllum inophyllum, Casuarina equisetifolia, and several others are often thus cared for, as are important food trees such as Canarium spp., Ficus spp., lnocarpus fagifer, and Pometia pinnata.

Domestication or semi-domestication of indigenous species
Related to the deliberate planting of indigenous species was the domestication or semi-domestication of many of these same indigenous species. This was particularly the case in non-Austronesian areas of Melanesia (Kirch and Yen 1982; Yen 1976a; 1974). Included in this Pacific Island domestication were Burkella spp., Calophyllum inophyllum, Casuarina equisetifolia, Hibiscus tiliaceus, Nersosperma oppositifolium, Pipturus argenteus, and Pometia pinnata, which came to contain a range of cultivars or genotypes and were deliberately planted in gardens and for coastal stabilization.

Improved fallows
Improved fallow systems based on deliberate planting of soil enhancing species was practiced in some areas, particularly in places in Papua New Guinea where agricultural intensification was far advanced. There, indigenous Casuarina spp. are deliberately planted as improved fallows and to provide fuel wood and fencing when cleared for the next garden. Albizia falcataria and Dodonaea viscosa are also planted to improve fallow in highland New Guinea. Guettarda speciosa and Tournefortia argentea are occasionally planted or encouraged to improve soils on atolls. A similar strategy has been adopted institutionally on Rarotonga in the Cook Islands, where Leucaena leucocephala, Albizia falcataria, Casuarina equisetifolia, and Eucalyptus spp. have been planted to rehabilitate degraded upland areas.

The productivity, sustainability, and adoptability of indigenous systems

The summary descriptions above of some technologies and characteristics of indigenous agroforestry systems serve to illustrate again the depth of knowledge and sophistication associated with the use of trees in the Pacific. The experts who understand, developed, and manage the systems - i.e. the local people - are already in place, "incountry," in the jargon of expatriate consultants. Formal experimentation and training might well improve the systems, but clearly, as noted in chapter 1, they possess already the criterion of adoptability, which, along with productivity and sustainability, is listed by Raintree (1990, 58) as one of the basic criteria possessed by any well designed agroforestry system. Similarly, that the systems are already local means that they meet the normative condition established by Bene et al. (1977); namely, that an agroforestry system be "compatible with the cultural patterns of the local population." As Raintree (1990, 92) says with regard to the introduction of agroforestry technologies, "there is little value in originality for its own sake" - espe cially when the local knowledge base is as immense as it is in the Pacific Islands. All of this suggests that encouraging local agroforestry systems would directly satisfy one of the currently most popular imperatives of development: community participation and concern with location-specific needs and opportunities (e . g., Chambers 1983).

Turning to Raintree's (1990) other two criteria, clearly the local systems possess adequate productivity or they would not have been maintained for so long. Indeed, archaeological and other evidence now supports the view that arboriculture has been practiced in the western Pacific for millennia (Yen 1990). Similarly, the systems possess sustainability and stability - even though they are dynamic - both in the sense that production has been maintained and that the environment has been protected if not enhanced. But being successful in the test of time does not mean that agroforestry systems achieve a high level of economic efficiency by modern measurements. The truth is of course that no matter how valuable the systems are believed to be, indisputable measurements are difficult to obtain. If we examine the anti-arboreal view of Künzel's composite Tongan farmer discussed at the beginning of this chapter, we can only agree that on a single field, planting or maintaining trees may lessen monocultural crop production. Yet the Tongan situation is unusual in the Pacific in that farmers hold individual tenure to single clearly demarcated fields, situated, on the main island of Tongatapu, in a favourable agricultural environment of mostly flat land with good volcanic soils. In more vulnerable areas of sloping land or poor soil, the protective and enhancing qualities of trees become more significant. Also, Pacific Island agroforestry systems have developed in a milieu where the unit of management was not a single field but a block of communal land wherein individuals held temporary tenure to various bits of garden land and enjoyed usufruct rights of varying sorts to trees, groves, and forest. This evershifting pattern of use amid a great diversity of arboreal and annual plants is antithetical to agronomic assessments of yield and economic efficiency. As Budowski (1982, 14) noted, comparisons of monocultures with agroforestry systems "may by no means be easy," especially as such evaluation is complicated by various short- and long-term economic projections concerning, for instance, the value of wood or the present and future estimations of environmental damage (for instance, erosion, use of pesticides in monocultures) and, even more so, by the appraisal of social and cultural factors, themselves complicated by a dynamic evolution in time that is difficult to foresee.

Encouraging agroforestry

Because institutional agroforesters, agronomists, and development agents are generally preoccupied with commodity production and maximization of yields, they seem at times almost to suffer from a sort of paralysis in the face of the difficulties or impossibility of producing quantitative assessments of agroforestry systems, especially ones as complex and unbounded as those developed indigenously in the Pacific Islands. What is called for as interest shifts to "sustainable development" is new ways of evaluation that integrate protective and productive functions, and new land-use institutions that focus on optimization of the whole landscape rather than component maximization aims. Lundgren (1987) has put forth a similar line of argument, but, no more than he could, can we suggest an "ideal" structure and set of objectives for tomorrow's land-use-related institutions. We will, however, end with a few recommendations on how further to encourage agroforestry in the Pacific Islands.

Existing systems as prototypes for development
Pre-packaged agroforestry technologies are by their nature alien, and their introduction is usually part of a short-term project, often set off from the general pattern of life and economy. When the project staff depart, the project often falls apart. Yet within the population already practicing indigenous agroforestry, experimentalists are legion. Pacific Islanders, like people all over the world, are fascinated by plants, like to experiment with them, and will happily carry a young individual plant from island to island or from valley bottom to mountain ridge or vice versa to establish it in a new home. If it works there, it will be reproduced and further disseminated. It makes sense, therefore, that at least some of the attempts at agroforestry intervention concentrate simply on supplying trees as raw materials for distribution. Some of these trees may be traditional, some newly introduced. The experimental work, the hybridization of new species into older systems, will be carried on in the place where it counts for most - among the local population. This does not of course mean that more formal agronomic experimentation is not equally or more valuable in terms of accurate information, but simply that experimenta tion carried on locally in the forests, woodland fallows, gardens, and grasslands is likely to be much more effective in achieving the goal of encouraging agroforestry.

Nurseries
Nurseries already successfully distribute seedlings and cuttings of fruit-trees, timber species, and legume trees in some Pacific Island countries (e.g., in Tonga, as described by Künzel 1989, 47-48), but they could be more comprehensively stocked and could incorporate extension work on agroforestry more strongly. Such nurseries would provide a central, permanent focus for agroforestry development.

Extension
Few Pacific Island countries would be able to establish an "agroforestry unit" on a permanent basis. If they were, the unit might better be placed in the Department of Agriculture rather than the Department of Forestry. In place of such an intersectoral unit, extension workers could at least be better educated in agroforestry and in ideas about the broad integration of protection and production. They could, for instance, stress the values of durable living fences or hedges; promote traditional techniques of mulching, composting, or improved fallow; and encourage agroforested home gardens and polycultural orchards.

Education
As with extension education, so too could agroforestry be promoted and publicized by other forms of education. Multimedia public programmes could be developed, in the vernacular, to stress the long-term value of existing agroforestry systems and the dangers of deforestation and agrodeforestation. Radio programmes along these lines already exist, as in Vanuatu, where the Environment Section has a weekly broadcast. Agroforestry can be better introduced, with appropriate field activities, into school curricula in both urban and rural areas. The nutritional value of tree foods can be incorporated into formal education or, for instance, into child-care information, as has been done by Fiji's National Food and Nutrition Committee.

Environmental impact assessment
When forests and woodlands are to be cleared, EIA generally includes a consideration of consequences with regard to erosion or hydrology, but the lost value of food and other products from trees also needs to be part of the formal assessment procedure. That this value often exceeds the value of the timber has been clearly demonstrated elsewhere in the world.

Agricultural intensification
Although long-term intensification can be seen as destructive of forests, intensification of annual production, often through the use of agroforestry techniques (Raintree and Warner 1986), lessens the extension of agriculture in the short term. Experience has shown, in Java, for instance, that severe agricultural intensification leads to a highly intensive, though spatially restricted, agroforestry. Inducing farmers to intensify is another matter, for it generally requires greater inputs, in some form such as capital, labour, fertilizers, management skills, and so forth. Appropriate consideration of the effects on the whole landscape of the intensification of certain kinds of production from land again requires that sectoral compartmentalization be lessened.

Further research and documentation of indigenous systems
Individual countries need to establish intersectoral committees or working groups to compile and publish lists and descriptions of useful tree species that should be protected or promoted locally. Further research is required into the existing systems, their component trees, the associated technologies, the methods of establishment, and the quantity and value of the products.

This book's descriptions of indigenous Pacific Island agroforestry systems may at times have seemed excessively detailed, but the book's coverage of the topic remains far from complete. What has been described, however, is sufficient to establish that traditional and existing agroforestry systems and their component trees have played and could continue to play a significant role in the provision of useful materials, in the enhancement of productivity and diversity, and in the stability of agro-ecosystems in the Pacific Islands. Recognition of agroforestry's multiple values and further documentation of indigenous systems can contribute to the effort to stem agrodeforestation and to stimulate the future flowering of Pacific Island agroforestry on the basis of the rich indigenous resource that has already for so long successfully blended production with protection.

Appendix One hundred Pacific Island agroforestry trees (1)

By definition, trees and tree-like species are essential components of all the agroforestry systems described in the body of this book. These arboreal species now in use need to be preserved and considered first as the basic components of future agroforestry development and in the reversal of agrodeforestation.

Whether on high islands, with their rich soils, varied habitats, and low population densities, on the harsh atolls, almost without soil and short of water, or in home gardens in densely settled urban areas or monocultural rural agricultural areas, all Pacific societies have selected and incorporated into their agroforestry systems a wide range of tree and tree-like species that suited local environmental conditions and met particular cultural needs. Out of the more than 400 such species or groups of closely related species (chapter 10), 100 are listed and described here (table A.2). All were and remain of economic, cultural, and ecological importance in Pacific Island rural and urban agroforestry systems.

The current use of these species, and many others not listed here, in agroforestry is the cumulative result of a selection process occurring over thousands of years. This process began in the ancestral homelands of today's Pacific islanders - in mainland South-East Asia and the archipelagic areas of Indonesia, the Philippines, and Papua New Guinea (known collectively as Malesia). From these areas, valuable cultigens and accumulated knowledge about uses of widespread indigenous species and genera were transferred to the smaller Pacific oceanic islands.

Subsequently, indigenous species encountered in the islands by the settlers were incorporated into Pacific agroforestry systems prior to European contact (Yen 1990). Some of those indigenous species of particular value were, in turn, carried farther east as new islands were settled. Finally, post-Europeancontact introductions have been tested and selected by Pacific islanders themselves and incorporated into indigenous systems to such an extent that the undiscerning visitor or agricultural "expert," and many of the current generation of islanders, believe them to be traditional or even indigenous plants. The integration of species (indigenous and aboriginal and post-European-contact introductions) has been carried out to such an extent that the status of many species remains unclear. Of the 100 tree or tree-like species or distinct cultivars listed here,

74 are usually classified as trees, whereas approximately 26 are better described as shrub-like perennials, which constitute relatively permanent fixtures in Pacific Island agroforestry systems. The classification system is, of course, arbitrary because some species, such as Hibiscus tiliaceus, Pandanus spp., Pipturus argenteus, and Vitex trifolia, are all found as both "shrubs and trees," depending on the environment, the variety, or cultivar. The species classified as shrubs or tree-like perennials (those not normally classified as trees) include:

  1. the coastal littoral species Pemphis acidula and Scaevola sericea;
  2. food species, including five banana or plantain cultivars (Musa cultivars), sugar cane (Saccharum officinarum), the closely related tall cane with an edible inflorescence (S. edule), bush hibiscus spinach (Hibiscus manihot), pigeon pea (Cajanus cajan), papaya (Carica papaya), and Cordyline fruticosa;
  3. coffee (Coffea spp.); the important handicraft plant, paper mulberry (Broussonetia papyrifera);
  4. the important social beverage plant kava (Piper methysticum);
  5. two widespread bamboos (Bambusa vulgaris and Schizostachyum spp.);
  6. the ubiquitous legume shrub or small tree Leucaena leucocephala;
  7. the widespread hedge or ornamental species Acalypha amentacea, Hibiscus rosa-sinensis, and Polyscias spp.; and
  8. Cycas spp., Euadia hortensis, Gardenia taitensis, and Heliconia spp., all culturally important (table A.2).

Table A.1 Origin or antiquity status of the 100 tree or tree-like perennial species listed in table A.2

Status Number (x/100)
Indigenous 19
Indigenous and aboriginal introductions 37
Primarily aboriginal introductions 13
Early post-European-contact introductions 24
Recent introductions 7
Total 100

In terms of origin, 19 are considered to be indigenous throughout most of their Pacific Island range; 37 are indigenous in parts of their range, usually in the western Pacific, but probably aboriginal introductions into other areas; 13 are probably aboriginal introductions throughout most of their range; 24 are long-established post-European-contact introductions; and 7 are relatively recent introductions (table A.1).

The indigenous species include the coastal or mangrove species Bruguiera gymnorhiza, Calophyllum inophyllum, Guettarda speciosa, Intsia bijuga, Pemphis acidula, Pipturus argenteus, Pisonia grandis, Premna serratifolia, Rhizophora spp., Scaevola sericea, Tournefortia argentea, and Vitex spp., which are either commonly incorporated in, or serve to protect, agroforestry systems from the harmful impacts of the sea and salt spray; coastal or lowland forest species Alphitonia spp., Ficus spp., Pterocarpus indicus, and Rhus taitensis, which are commonly found in active agricultural areas and/or protected when clearing new gardens; and the widespread pioneer species Commersonia bartramia, Glochidion spp., Macaranga spp., and Kleinhovia hospita, which are important, sometimes protected, components of fallow vegetation.

Those that are indigenous to the larger islands of the western Pacific, but possibly aboriginal introductions to some of the smaller islands of the eastern Pacific, include the wild and/or cultivated food plants Adenanthera pavonina, Barringtonia spp., Canarium spp., Dracontomelon vitiense, Ficus spp., Gnetum gnemon, Inocarpus fagifer, Metroxylon spp., Musa troglodytarum, Pandanus spp., Pometia pinnata, Saccharum edule and S. officinarum, and Terminalia catappa. Other culturally important species are Acalypha amentacea, Atu na racemosa, Bischofia javanica, Cananga odorata, Casuarina spp., Cordia subcordata, Cycas spp., Erythrina variegata, Euodia hortensis, Fagraea berteriana, Gardenia taitensis, Garuga floribunda, Heliconia spp., Hibiscus tiliaceus, Morinda citrifolia, Pandanus spp., Polyscias spp., Schizastachyum spp., Securinega flexnosa, Syzygium spp., and Thespesia populnea.

The 13 plants that seem to have been aboriginal introductions throughout their Pacific Island range are mostly domesticated food plants (cultigens). These include breadfruit (Artocarpus altilis and A. marianennsis), shaddock, or pomelo (Citrus grandis), coconut palm (Cocos nucifera), Cordyline terminalis, bush hibiscus spinach (Hibiscus manihot), two banana or plantain cultivars (Musa AAB and Musa ABB triploid), and Malay apple (Spondias dulcis). The remaining aboriginal introductions include candlenut (Aleurites moluccana), betel-nut palm (Areca catechu), paper mulberry (Broussonetia papyrifera), common hibiscus (Hibiscus rosa-sinensis), and kava (Piper methysticum).

The 24 early post-European-contact introductions, all of which have been integrated into agroforestry systems in the Pacific, are mainly food or beverage plants. These include jakfruit (Artocarpus heterophyllus), soursop (Annona muricata), pigeon pea (Cajanus cajan), papaya (Carica papaya), five Citrus species (lime [C. aurantiifolia], sour, or Seville, orange [C. aurantium], lemon [C. Iimonl hystrix?], mandarin orange [C. reticulata], and sweet orange [C. sinensis]), coffee (Coffea spp.), mango (Mangifera indica), horseradish, or drumstick, tree (Moringa oleifera), the Cavendish and lady's finger bananas (Musa AAA and Musa AAB), guava (Psidium guajava), avocado (Persea americana), tamarind (Tamarindus indicus), and cocoa ( Theobroma cacoa). Other important post-European-contact introductions include common bamboo (Bambusa vulgaris), kapok (Ceiba pentandra), flamboyant, or poinciana (Delonix regia), leucaena (Leucaena leucocephala), plumeria, or frangipani (Plumeria rubra), and rain tree, or monkey-pod (Samanea saman).

More recent introductions with increasingly widespread or localized importance in agroforestry systems include Albizia spp., the flowering cassias, or shower trees (Cassia spp.), gliricidia (Gliricidia septum), a range of eucalyptus trees (Eucalyptus spp.), Caribbean pine (Pinus caribaea), big-leaf mahogany (Swietenia macrophylla), and the jambolan (Syzygium cumini).

The common names, scientific names, and synonyms, related species, origin, range and antiquity status, brief physiognomic descriptions, place in Pacific Island agroforestry systems, and information on the cultural and ecological importance of these 100 tree or tree-like perennial species or groups of related species are detailed in table A.2.

Table A.2 One hundred (100) cultivated and wild tree or tree-like perennial species of widespread or considerable local economic, cultural, and ecological importance in Pacific Island rural and urban agroforestry systems (Note: Information provided on: 1) distribution, range, and antiquity status in the Pacific Islands, i.e. whether a given species is indigenous or an aboriginal or recent post-European-contact introduction [Malesia refers to the Malaysian "phytogeographical region," which includes all of the Malay and Indonesian archipelago, the Philippines, and the Island of New Guinea]; 2) the physical characteristics [physiognomy] of the species; 3) its habitat and role within the context of agroforestry systems; and 4) its cultural and ecological utility.)

Key

SCIENTIFIC NAMES

  1. Scientific names (binomials) are provided in italics.
  2. The first name listed is that which seems to be the currently most widely accepted (usually the earliest published) binomial for a given species.
  3. The scientific names provided subsequently indicate either other binomial synonyms (syns.) for the species or names of very similar, but less important, species (spp.)
  4. The authorities (persons responsible for describing and providing a given name for a species) are provided after each species name.
  5. In cases where given genera are represented by numerous, often endemic, but closely related species (i.e. when a given genera exhibits wide speciation), the first name will consist of the genus name, followed by spp., with the binomials of closely related, usually indigenous Pacific Island species of this genera.

FAMILY

  1. The name of the family to which the species belongs is provided in upper-case letters (e.g. RUBIACEAE) immediately after the widely accepted scientific name.
  2. Where a family is known by two different names, both are listed with a slash between them (e.g. FABACEAE/LEGUMINOSAE or POACEAE/GRAMINAE), or, if a given species is sometimes placed in different families, both are listed (e.g. CLUSIACEAE OR GUTTIFERAE).

COMMON NAMES

  1. The English or common names for the species, with other widely-used non-Pacific Island and Pacific Island names, are included in quotation marks, e.g. "coconut."

Appendix One hundred Pacific Island agroforestry trees (2)

ONE HUNDRED PACIFIC ISLAND AGROFORESTRY SPECIES

1. Acalypha amentacea Roxb. ssp. wilkesiana (Muell.-Arg.) Fosb. EUPHOR BIACEAE
"copper leaf," "Jacob's coat," "beefsteak plant," "fire dragon plant"
syn. A. wilkesiana Muell.-Arg. f. wilkesiana A.C. Smith

Native to Malaysia or western Melanesia; probably an aboriginal introduction to Melanesia and western Polynesia, and a recent introduction to eastern Polynesia and to some atoll countries. Perennial shrub, up to 2 m or higher, with attractive, rather curved and coarsely crisped, dark or bright red, red-green, or green leaves, which are often mottled or variegated with various shades of red, dark pink, white, or bronzy green. Common to abundant in home gardens, often as a hedge or living fence; occasional in rural garden areas and in urban landscaping. Planted ornamental, boundary, and hedge plant; leaves used medicinally; has important ceremonial and spiritual value in Melanesia

2. Adenanthera pavonina L. MIMOSACEAE/LEGUMINOSAE
"red-bead tree," "red sandalwood"

Native to South-East Asia and Malesia; possibly an aboriginal introduction to parts of Melanesia and western Polynesia, and a recent introduction to Hawaii and elsewhere. Medium-sized, deciduous tree, 6-15 m or higher, with bipinnate leaves; greenish white to yellowish, sweetly fragrant flowers; and flattened, linear at first, becoming curved or contorted, brown pods bearing dark or bright red, hard, lensshaped seeds. Common in cultivation and as a protected tree in garden areas and mature fallow forests throughout the Pacific; not generally found on atolls, although two specimens were seen on Funafuti, Tuvalu. Hard, durable wood used in construction and for furniture and firewood; seeds used in necklaces, body ornamentation, and handicrafts; endosperm eaten, mainly by children, the cooked seeds being sold commercially at the market in Western Samoa.

3. Albizia lebbeck (L.) Benth. MIMOSACEAE/LEGUMINOSAE
"albizia," "siris tree," "woman's tongue"
spp. A. falcataria (L.) Fosb.; A. procera (Roxb.) Benth.

Probably indigenous to tropical Asia, but now Palaeotropical from Africa to Australia; a recent introduction to most Pacific Islands; not found on atolls. Large to medium tree, 10-20 m high, with bipinnate leaves; fragrant yellowish to whitish flowers; and flat, pale, straw-coloured, few-seeded pods. Commonly planted as a shade tree or ornamental in rural areas, plantations, home gardens, and along roadsides; abundantly naturalized along roadsides, stream banks, and in secondary vegetation in Vanuatu, Fiji, and Saipan. Planted ornamental; wood used in general construction, for firewood, and for wood carving; pods used occasionally as cattle feed. A. falcataria and A. procera are both important in some areas, such as in Hawaii, where A. falcataria is widely used in reforestation.

4. Aleurites moluccana (L.) Willd. EUPHORBIACEAE
"candlenut"

Indigenous to Malaysia, probably an aboriginal introduction to most of Melanesia, Polynesia, and parts of Micronesia, and a recent introduction into some areas such as

Guam. Tree 10-25 m tall, with light green, downy, palmately-lobed leaves; small greenish yellow to white flowers; and globose fruit containing hard, woody, bony, brownish black seeds with an oil-rich kernel (endosperm). Commonly planted or protected in gardens and on farms, along trails, and in home gardens; often planted as living fencing or as boundary markers; common in mature fallow forest and naturalized in many areas, mainly in disturbed sites, but rarely found in deep forest. Tree features in Polynesian legends and proverbs and is the state tree of Hawaii; planted ornamental or boundary tree; timber used in light construction, canoe hulls, and for firewood; extract of roots, bark, and fruit skin provides black to reddish dye; leaves and flowers used in leis in Hawaii; seed kernels strung together and ignited to provide light; oil from seed kernel used as a hair and skin conditioner; seeds burnt to provide ash used to dye bark cloth black and for tattooing; baked and pounded kernels eaten as a relish in Hawaii; uncooked kernel a strong purgative; oil formerly exported from Hawaii and Fiji.

5. Alphitonia zizyphoides A. Gray RHAMNACEAE
spp. A. incana (Roxb.) T & B ex Kurz; A. neo-calidonica; A. ponderosa Hbd.

Indigenous from Sumatra and the Philippines to the Society Islands (related species found from New Guinea to Hawaii). Medium to large tree, 5-25 m high, with leaves that are bright green above and white tomentose beneath; small white to greenish white flowers; and clusters of small purple fruit. Common in gardens and recent fallow areas, savanna lands, mature fallow forests, open forest, and occasionally in primary forest; often protected when clearing new garden areas; infrequently in home gardens. Durable and straight-grained wood used in construction, for furniture, canoes and canoe paddles, tools and tool handles, digging sticks, and for firewood; bark and leaves used medicinally; leaves used as a soap substitute.

6. Annona muricata L. ANNONACEAE
"soursop"

Native to tropical America; a pre-World War 11 introduction into most of the Pacific Islands. Small tree, rarely over 8 m tall, with bright green oblong or oblongelliptic, glossy leaves; yellowish green flowers; and oblong or ovoid, irregularly kidney-shaped, green to yellowish green fruit, covered with regularly-spaced, short, slightly curved, fleshy spines, and having white, juicy, somewhat acid, aromatic, cotton-like, edible pulp and hard, dark brown to black seeds. Commonly cultivated in home gardens and occasional in rural garden areas on volcanic and raised limestone islands; occasionally naturalized; not found on atolls. Ripe fruit eaten raw, often with ice cream, and made into drinks.

7. Areca catechu L. ARECACEAE/PALMAE
"betel-nut palm," "betel-nut"

Indigenous to southern Asia, Indonesia, and possibly the Philippines; an aboriginal introduction into Papua New Guinea, the Solomon Islands, and western Micronesia; a recent introduction into Fiji, Samoa, and other areas. Singlestemmed erect palm, up to 15 m tall, with a slender, pale gray-green trunk with well-spaced, gray-white rings; pinnate fronds; and small, hard, red or orange ovoidoblong fruit. Commonly planted in garden areas and in home gardens in western Melanesia and western Micronesia; often found in dense groves near villages; planted by South Indians in Fiji. Often planted as an ornamental; timber useful, the outer wood used for walling, flooring, or battens; husk and other parts of the fruit used medicinally; kernel (endosperm) of green and mature fruit chewed as an astringent and stimulant, often with the leaves or fruit of betel-pepper (Piper betle) and lime; fruit a major commercial crop for local sale in Papua New Guinea, the Solomon Islands, and the Caroline Islands of Micronesia.

8. Artocarpus altilis (Park.) Fosb. MORACEAE
"breadfruit"
syns. A. incisus (Thunb.) L. f.; A. communis Forst.
sp. A. mariannensis Trec.

Indigenous to Malaysia, probably an aboriginal introduction into most of Melanesia, Polynesia, and Micronesia, and a more recent introduction into some of the smaller islands and atolls. Medium to large tree, 10-15 m high, with thick, milky sap; large, variable, entire to deeply lobed leaves; and large, globose to oblong, light green to yellow fruit studded with a grid-like network of tubercles and with creamy white to yellowish, edible flesh. Common to abundant in garden areas, mature fallow forests, village tree groves, and in home gardens on both high islands and atolls; absent on drier atolls, such as Tamana and Arorae in southern Kiribati. Trunk used for canoe hulls and occasionally in house construction; inner bark used to make bark cloth in some areas; sap used for caulking canoes, adhesive for bark cloth, and as chewing-gum; leaves used for wrapping food for cooking, for parcelling of fresh food, and as plates; meristem used medicinally; dried inflorescence burnt as a mosquito repellent; fruit eaten cooked as a major or supplementary staple food in most areas of Polynesia and Micronesia, and as a supplementary staple in most of Melanesia; cooked or raw fruit preserved in pits or by sun-drying. Many named cultivars exist in most areas. A. mariannensis, which is reportedly indigenous to Guam, is also common throughout Micronesia and also present, possibly as an aboriginal introduction, in Tuvalu and Tokelau in Polynesia. Interspecific hybrids, A. altilis x mariannensis, exist in most areas of Micronesia.

9. Artocarpus heterophyllus Lam. MORACEAE
"jakfruit," "jackfruit"
syn. A. integer (Thunb.) Merr.

Indigenous to South-East Asia and of ancient cultivation in India; a postEuropeancontact introduction into the Pacific Islands. Tree up to 20 m high, with white, resinous latex; dark green, coriaceous leaves; and a very large, yellow-green to golden-yellow, somewhat oblong-cylindric, occasionally distorted, edible fruit. Commonly planted on smallholder Indian cane farms in Fiji, mostly in home gardens, and occasional in rural gardens and home gardens in other areas of the Pacific. Planted fruit-tree; pulp of ripe fruit eaten raw as a fruit or cooked as a supplementary staple, commonly in curries by Indians in Fiji; sold locally in Fiji; seed edible; heartwood useful for general construction.

10. Atuna racemosa Raf. CHRYSOBALANACEAE
syns. Parinarium laurinum A. Gray; P. glaberrimum Hassk.; Parinari laurina A. Gray; P. glaberrima (Hassle.) Hassk.

Indigenous from Indonesia and the Philippines to Fiji, and to the Caroline Islands in Micronesia; probably an aboriginal introduction from Fiji into Tonga, Samoa, Uvea, and Futuna. Medium to large tree, 5-20 m high, with bright green, ovate to elliptical or lanceolate, somewhat coriaceous, leaves; whitish flowers; and a large, rough, subglobose, light brown fruit. Occasional in garden areas, in home gardens, in mature fallow forest, and common to occasional in lowland forest and grassland thickets; commonly protected when clearing for new gardens. Timber used in light construction for posts, poles, and canoe spars; leafy branchlets, with the leaves attached, used in Fiji to thatch or insulate the outside walls of houses; leaves used for ceremonial decoration of houses in the Solomon Islands; inner bark used medicinally to treat high blood pressure in Fiji and diarrhoea in the Solomon Islands; skins of fruit pounded and shredded to make canoe caulking and to adhere inlay in carvings in the Solomon Islands; crushed seeds used to scent coconut oil in Fiji, Tonga, Samoa, and Uvea.

11. Bambusa vulgaris Schrad. ex Wendl. POACEAE/GRAMINAE
"common bamboo," "feathery bamboo"

Indigenous to tropical Asia, but an early post-European-contact introduction into most Pacific Islands. Giant perennial, clump-forming, woody grass with segmented green to yellowish stems, up to 15 m tall and 8 cm wide. Commonly planted and naturalized and forming large stands, particularly in alluvial sites throughout the high-island Pacific; also introduced but not established on some atolls. Woody stems widely used in construction, for fencing, handicrafts, making rafts, cages, knives, net-mending needles, fishing poles, floats for fishing nets, fruit-picking poles, irrigation conduits, musical instruments, fencing, containers for food, water, and for cooking food over an open fire, and as fuel wood; parts used medicinally.

12. Barringtonia edulis Seem. BARRINGTONIACEAE/LECYTHIDACEAE
"cut nut," katnut (SI pidgin)
spp. B. novae-hibernia Ltb.; B. procera (Miers) Kunth; B. samoensis A. Gray

Indigenous species of Barringtonia are found throughout Melanesia, from which a range of cultivars have been selected and possibly distributed as far east as Fiji and Samoa prior to European contact. Tree, 6-20 m high, with obovate-oblong to oblanceolate leaves, white, yellow, or pink tinged flowers and long, pendant clusters of green to purple fruits containing a white, edible kernel. Commonly to occasionally planted or protected around villages and in home gardens; occasional in garden areas, in mature fallow forest, and in dense and open forest, or on the edge of forests from sealevel to 400 m; generally protected when clearing for new gardens. Light wood used for canoe paddles, casing, light construction, and for quick-burning firewood; bark used medicinally for stomach ailments and gonorrhoea in the Solomon Islands; mature seed kernels eaten raw or cooked and sold at local produce markets; cooked for storage in the eastern Solomon Islands.

13. Bischofia javanica B1. EUPHORBIACEAE/BISCHOFACEAE
"Java cedar," "koka" (Polynesia, Fiji, and Vanuatu)

Indigenous from India and South China through Indonesia and the Philippines eastward into the Pacific Islands; although probably native as far east as Fiji, it may be an aboriginal introduction to Samoa, Tonga, Uvea, Futuna, and the Cook and Society Islands, and probably a recent introduction into Hawaii. Large dioecious tree, 530 m high, with reddish wood; bright green, trifoliate leaves; many small, greenish or yellowish flowers in panicles; and small, globose fruit. Common in lowland forest, on the forest edge, in thickets, along streams, and in degraded savanna and grasslands; one of the commonest trees in garden areas and mature fallow forest, and occasional in villages and home gardens; although commonly severely pruned or pollarded, it is usually protected or survives during the clearance of new garden areas, due to its high resistance to fire, felling, and ringbarking. Durable and water resistant timber used in light construction, as a favoured house post, and as firewood; red-brown dye and tannin for the tape-making process derived from the bark, and sometimes from the roots; leaves and bark used medicinally; branches cut in Tonga to provide trellising for yams; leaves cooked with, often to tenderize, pork; fruits a favoured food of pigeons, doves, and other frugivorous birds.

14. Broussonetia papyrifera (L.) Vent. MORACEAE
"paper mulberry"
syn. Morus papyrifera L.

Indigenous to China, Japan, and probably Burma and Thailand; an aboriginal introduction throughout Malesia, Melanesia, and Polynesia as far east as Hawaii. Slender, erect shrub or tree, up to 8 m high, with densely hairy branches, milky latex, and tough best fibre; variable ovate to cordate, scabrous, lobed or entire, serrate or toothed leaves; small flowers (although rarely flowering) in axillary clusters; and pulpy, club-shaped fruit. Commonly cultivated in garden areas and around villages from near sealevel in Fiji, Tonga, and Samoa and in some areas of lowland and highland New Guinea; found mainly as a relict in other areas of Polynesia; not reported present in Micronesia or on atolls, except as a recent introduction to Yap; common in abandoned gardens and occasionally naturalized; occasional in home gardens. Often planted in monocultural plots, often following the harvest of yams in Tonga; best fibre treated and pounded to make fibre and bark cloth (tape cloth), which is used un-dyed or variously, often intricately, dyed and/or painted; fibre used for toilet paper, bandages, clothing, ceremonial dress, string bags, cordage, interior decorations, and ornamentation; large painted tape cloths a major item of ceremonial exchange in parts of Papua New Guinea, Fiji, and Polynesia; tape cloth an important commercial and export item in Fiji and Tonga, and to a lesser extent Samoa.

15. Bruguiera gymnorhiza (L. ) Lam. f. RHIZOPHORACEAE
"brown mangrove," "Oriental mangrove"
syns. Rhizophora gymnorhiza L.; B. conjugata (L.) Merr.; B. eriopetala W. & Arn.; B. reedit Bl.

Indigenous to East Africa, the Indian Ocean, and tropical Asia to Tonga and Samoa in western Polynesia and to Nauru and the Marshall and Gilbert Islands in eastern Micronesia; a recent introduction into areas of eastern Polynesia, such as Hawaii. Small to large-sized tree, up to 20 m or higher, with inconspicuously buttressed trunks; thick, leathery, brittle, oblong-elliptic leaves; greenish yellow to bright red flowers; and an elongated spindle-shaped, ribbed, pre-germinated fruit. Abundant in tidal swamps and river mouths, and occasional in landlocked sink-holes and depressions; commonly borders agricultural lands, protecting them from sea spray, storm surge, and salt-water incursion. Planted in coastal areas and near fish-ponds for coastal protection and stabilization; hard and durable wood used in general construction, for tool handles and digging sticks, and considered to be one of the best fuel woods; pre-germinated seed (fruit) eaten cooked, after scraping or grating, washing, and drying (to remove tannins) and sometimes mixed with coconut in Melanesia and Nauru; fruit sold as a vegetable at Honiara Market; extract of bark used as a dye and preservative (tannin) for bark cloth; bark used medicinally as an abortifacient and for treating burns in the Solomon Islands; skin of seed used to prepare a black dye for traditional skirts in Nauru; flowers used in garlands.

16. Cajanus cajan (Mill.) Millsp. PAPILIONACEAE/FABACEAE/ LEGUMINOSAE
"pigeon pea," "red gram," arhar dhal (Hind)
syns. C. indicus Spreng.; C. flavus DC.; Cajan cajan (L.) Millsp.; Cystisus cajan L.

Indigenous to India and South-East Asia; a post-European-contact introduction into Fiji and many other countries. Erect, short-lived, perennial pubescent shrub, 1-3 m or higher, with pinnately trifoliate leaves; bright yellow flowers, marked with dark reddish brown to crimson; and linear-oblong, flattened, inflated pods bearing globose, compressed, cream-coloured to reddish or brownish or speckled seeds. Abundant on sugar-cane farms and occasional in home gardens, especially in the dry zone of Fiji; infrequent in other areas of the Pacific, where it has been widely introduced. Cultivated food plant; a major supplementary staple crop and source of plant protein for the Fiji Indian population; important nitrogen-fixing intercrop or supplementary crop on sugar-cane farms; immature pods cooked as a vegetable and mature seeds dried and cooked as a protein-rich pulse; plant remains used as fodder or green manure; dried stems also occasionally used as fuel.

17. Calophyllum inophyllum L. CLUSIACEAE/GUTTIFERAE
"Portia tree," "Alexandrian laurel," "beach mahogany"

Indigenous from tropical Africa to eastern Polynesia and Micronesia; possibly an aboriginal introduction to some islands. Medium to large, hard-wooded, slow growing tree, 10-20 m tall, with a broad, low-branching, spreading crown; dark green, stiff, leathery leaves; showy, waxy, white, very fragrant flowers; and green to purple-black, globose, hard fruit containing a single, somewhat poisonous, oil-rich kernel. Common to abundant in coastal forest and the dominant component of the pre-phosphate-mining vegetation of Nauru, Banaba (Ocean Island), and Makatea; occasionally planted or protected in inland gardens and home gardens and planted along roadsides. Tree sacred in parts of eastern Polynesia, where it features in legends and was planted around temples; timber favoured for construction, wood carving, and canoe hulls; sticky sap used for caulking canoes; leaves used medicinally; kernel of green and mature fruit crushed to yield oil that is used to scent coconut oil and applied to hair to make it long and black; old, decayed fruit skewered on coconut midribs and burned as traditional Nauruan light; mature fruit burned as a mosquito repellent; the seed kernel (known commercially as "punnai nut") yielding a dark green oil formerly exported from Fiji.

18. Cananga odorata (Lam.) Hook. f. and Thoms. ANNONACEAE
"ylang-ylang" (Malaya/Indonesia), "perfume tree"
syns. Canangium odoratum (Lam.) Baill.; Uvaria odorata Lam.

Native to South-East Asia, the Philippines, and northern Australia and possibly as far east as the Solomon and Caroline Islands; an aboriginal introduction in parts of

Melanesia and to Polynesia, but a recent introduction to Hawaii and some other smaller islands of the eastern Pacific. Tree, up to 15 m or taller, with a crooked trunk; smooth, grey bark; drooping, brittle branches; very fragrant, drooping, yellowish green turning yellow, spider-like flowers with wavy, linear-lanceolate petals; and oblong, gray-green to black, fleshy fruits containing numerous seeds. Common in home gardens and occasional as a planted or protected tree in garden areas; occasional to common in fallow forests, open forests, and on the edges of forests where it has become naturalized; generally protected when clearing new areas for cultivation. Timber used in general construction, for canoe making, and occasionally for firewood; flowers used in garlands and for scenting coconut oil; leaves and bark used medicinally; of considerable cultural importance throughout Melanesia and Polynesia; used in the commercial production of perfume and perfumed soap in Fiji and of essential oil in the Philippines and Indonesia; fruit a preferred pigeon food.

19. Canarium spp. BURSERACEAE
"canarium almond," "Java almond," "galip nut," "pill nut"
spp. C. decumanum ?; C. indicum L. (syn. C. commune L.); C. lamii Leenhouts; C. salomonense Burtt.; C. harveyi Seem. (syn. C. mufoa Cristoph.); C. vitiensis A. Gray; C. vulgare Leenh.

C. indicum is indigenous to Malaya, Indonesia, and New Guinea, and C. vulgare to Malaya and Indonesia, both are possibly aboriginal introductions into parts of Melanesia and recent introductions into other islands; other species seem to be native to and cultivated in New Guinea and other islands of Melanesia and Polynesia as far east as Samoa and Niue, and possibly aboriginal introductions into some of these areas. Medium to large trees up to 30 or 40 m high; with oddpinnate compound leaves; white to creamy white or yellow, racemose flowers; and ovoid, green to purplish, bluish, or black, hard-shelled fruit containing an oil-rich seed kernel. Commonly planted around villages and in home gardens and common in mature lowland forests, fallow forests, thickets, and in garden areas throughout Melanesia; occasional in Fiji, Tonga, Samoa, and Niue in lowland forest and garden areas. Occasionally used as a shade tree in plantations. Timber suitable for light construction, canoes, bowls, and firewood; rotten logs a source of edible insect larvae; bark used medicinally for chest pains in the western Solomon Islands; edible, oil-rich seed kernels highly prized and constitute a major component of the diet throughout most of Melanesia, except in Fiji; a major item of inter-island trade and sold commercially, often as snack food; often stored either in the shell or after baking and drying; often mixed with staple root crops, added to soups, or eaten with megapode eggs in the Solomon Islands; oil from kernel used for lighting in past.

20. Carica papaya L. CARICACEAE
"papaya, " "pawpaw"

Indigenous to tropical America; an early post-European-contact introduction to the Pacific Islands. Soh-wooded, un- or few-branched, rather palm-like, small, quick growing tree, up to 4 m or higher, with thick, hollow, tapering, nearly smooth trunks or stems with light bark and numerous, almost heart-shaped, leaf-scars; copious, thick, sticky, irritating milky sap; leaves drop as the tree grows; numerous, white or cream-coloured, fragrant flowers; and variably shaped, subglobose or pear-shaped to cylindrical, green turning yellow or orange fruit, with orange to red-orange edible flesh and numerous small, gray-green, mucilaginous seeds. Abundant throughout the Pacific in home gardens and rural crop lands as an intercrop in shifting agricultural systems; planted in monocultural orchards in Fiji, Tonga, and Rarotonga; widely naturalized in fallow areas and waste places; grows well in the calcareous soils of atolls. Cultivated fruit-tree, grown throughout the Pacific for local sale and in some areas, such as Hawaii, Fiji, Tonga, and the Cook Islands, as an export crop; ripe fruit eaten or made into jam; fruit known to be a laxative; green fruit cooked in curries as a vegetable in Fiji; juice and flesh from green fruit (which contains the enzyme papain) used to tenderize pork, beef, and fish; white sap from small immature fruit used as a cure for ringworm; fragrant flowers used in garlands; hollow leaf petioles used by children as "pea shooters" for the fruit of Tournefortia argentea.

Appendix One hundred Pacific Island agroforestry trees (3)

21. Cassia spp. CAESALPINIACEAE/LEGUMINOSAE
"cassias," "shower trees"

spp. C. alata L. ("candle bush," "golden candelabra tree"); C. fistula L. ("golden shower tree," "Indian laburnum," "pudding-pipe tree"); C. glauca Lam. (syns. C. sulfurea DC. ex Colladon and Senna sulfurea DC. ex Colladon Irwin & Barneby) ("scramble-egg tree"); C. grandis L. f. ("pink shower tree," "horse cassia"); C. javanica L. ("pink and white shower tree," "pink shower tree"); C. siamea Lam. ("kassod tree")

The cassias are all recent introductions to the Pacific Islands; some species (C. fistula, C. glauca, C. javanica, and C. siamea) are native to tropical Asia, whereas others (C. alata and C. grandis) are indigenous to tropical America. They range from shrubs or small trees (C. alata and C. glauca) to large, spreading trees (C. grandis and C. siamea), with pinnately compound leaves; striking racemose clusters or racemes of bright yellow to pink, or pink and white, flowers, and short to long, seed-bearing pods. Occasional to common in home gardens and in rural areas; occasionally naturalized in poorly-drained areas, along streams, and in other disturbed sites. Planted ornamental and roadside trees; occasionally planted as shade trees; wood of some species used in general construction and as firewood; leaves of C. alata widely used as a cure for ringworm; flowers used in garlands.

22. Casuarina equisetifolia L. CASUARINACEAE
"casuarina," "she oak, " "ironwood," "beefwood"
syn. C. Iitorea L.
spp. C. nodiflora Forst. f (syns. Gymnostoma nodiflorum [Thunb.] L.A.S. Johnson; C. vitiense L.A.S. Johnson); C. oligodon ("yar," PNG Pidgin); C. papuana S. Moore

C. equisetifolia indigenous to the Indian Ocean, South-East Asia, Malesia, northern Australia, and most parts of the western Pacific; possibly an aboriginal introduction in some areas of the eastern Pacific, and a recent introduction to Nauru and most of the atolls. Medium to large, hardwooded, fast-growing, pine-like tree, up to 20 m high, with numerous, short-lived, long, thin, drooping, needle-like, gray-green, photosynthetic branchlets; awl-shaped, scale-like leaves; and small, dull green ripening to brown, cone-like fruit. Common on rocky volcanic and limestone coastal areas and commonly naturalized or indigenous in sandy or limestone areas and on de graded, lateritic, highly-eroded interior areas and fern lands of many islands; a naturalized pioneer on some open-cast mined areas in Nauru; occasional in home gardens and urban areas. Formerly a sacred tree in eastern Polynesia, where it was planted in temple grounds; nitrogen-fixing tree, often planted or protected in garden areas; commonly planted as an ornamental, roadside tree, wind-break, or for coastal protection or reclamation; hard, durable wood favoured for general construction, wood carving, and the production of tools, war clubs, tape pounders, and canoe parts; an excellent firewood; parts used medicinally.

Other, more restricted species indigenous from New Guinea to New Caledonia and Fiji, but apparently a recent aboriginal introduction into the Eastern Highlands of Papua New Guinea (c. 1200 BP). Small to medium-sized trees, up to 20 m or higher, with slender, green branchlets (the distal branchlets quadrangular) bearing whorls of scale-like leaves. Common in garden areas and grasslands throughout highland New Guinea, either as nurtured spontaneous, or deliberately planted trees; common in home gardens and as individual trees or groves surrounding villages and ritual pig-killing sites in highland New Guinea; occasional in lowlands and degraded or ultrabasic upland areas in New Guinea, New Caledonia, the Solomon Islands, and Fiji. Tree commonly planted or protected as an improved fallow in highland New Guinea because of its nitrogen-fixing ability; occasionally planted to control erosion and as boundary markers; timber used for light construction, tools, weapons, fencing, and firewood; trees also important in garden magic and death rituals in New Guinea. (L.A.S. Johnson has reclassified the Fijian species as belonging to the genus Gymnostoma, thus indicating that a similar reclassification of other Pacific Island species might be expected.)

23. Ceiba pentandra Gaertn. BOMBACEAE
"kapok tree, "silk-cotton tree"

Indigenous to India or Africa; a pre-World War II introduction into most Pacific Islands. Tall, deciduous, soft-wooded, light-gray-barked tree, up to 25 m or higher, with a buttressed, commonly spiny trunk; palmately compound leaves; and oblongellipsoid, capsular, pendulous, 5-celled fruit filled with numerous long, soft, silky or cotton-like fibres. Common to occasional throughout most of the high islands, including raised limestone islands, in both home gardens and in rural areas, often along roads or bordering agricultural holdings. Planted ornamental or roadside or boundary-marker tree; light-weight timber used in light construction; fibre used for stuffing pillows and mattresses.

24. Citrus aurantiifolia (Christm.) Swingle RUTACEAE
"lime," "West Indian lime"
syns. C. Iima Lunan; C. acida Roxb.; Limonia aurantiifolia Christm.

Probably indigenous to southern Asia or Indonesia; a post-European-contact introduction into the Pacific Islands. Bushy shrub or small tree, up to 4 m high, with sharp spines; variable leaves; fragrant, pink to white-petalled flowers; and small, smooth, globose, green to light yellow-skinned fruit with aromatic, acidic pulp and juice. Common in home gardens in some areas of the Pacific and occasional in rural areas; only occasionally naturalized; planted occasionally in monospecific plantations in Fiji, Niue, and the Cook Islands; the only citrus species well established in Kiribati and some other atolls. Planted fruit-tree; juice of fruit used to marinade raw fish, make drinks, and for flavouring food and desserts; whole fruit pickled as a relish (achar) for curries by Indians in Fiji; fruit sold as a minor cash crop in many countries.

25. Citrus aurantium L. RUTACEAE
"sour orange," "Seville orange"
syn. C. vulgaris Risso

Probably indigenous to South-East Asia; an early post-European-contact introduction into the Pacific Islands. Small tree, up to 10 m high, with short, slender spines; leaves with broad-winged petioles; very fragrant, white flowers; and subglobose, rough-skinned, green fruits that turn bright orange at maturity and contain sour or bitter orange pulp. Occasional in rural garden areas and in home gardens throughout most of Polynesia and in some areas of Micronesia and Melanesia; not found on atolls. Fruit juice used to make drinks and for flavouring food; fruit and skins occasionally used to make marmalade; fruit sold as a minor cash crop in some countries; parts used medicinally.

26. Citrus grandis (L.) Osbeck RUTACEAE
"pummelo," "pomelo," "shaddock"
syns. C. aurantium var. grandis L.; C. auranfium var. decumana L.; C. maxima (Burm.) Merr.; C. decumana (L.) Murr.; Aurantium maximum Burm.

Probably indigenous to South-East Asia and Malesia; probably an aboriginal introduction into many Pacific islands, including Fiji and western Polynesia; a postEuropean-contact introduction into some areas. A medium-sized tree, up to 12 m high, with leaves with winged petioles; pure white flowers with yellow anthers; and large subglobose to pyriform (pear-shaped), thick-skinned, light green to yellowgreen fruit with pale yellow to pink pulp. Occasional in garden areas, mature fallow forests, and secondary forests, where it seems to be naturalized; uncommon in home gardens. Timber used in light construction and considered a good firewood; scraped root used medicinally to cure haemorrhoids in Fiji; pulp eaten as a snack food; whole fruit used to make children's toys (e.g. wheels, etc.); fruit sold occasionally as a minor cash crop in some countries. The pamplemousse, or grapefruit, that is common in French Polynesia and other French territories is probably a hybrid cross between C. grandis and possibly C. sinensis, the sweet orange.

27. Citrus limon (L.) Burm. f. RUTACEAE
"lemon"
syns. C. medica var. limon L.; C. Iimonum Risso; C. Iimonia Osbeck
sp. C. hystrix DC. ("rough lemon"?) (see note below)

Possibly indigenous to South-East Asia; probably an early post-European-contact introduction to the Pacific Islands. Small tree, up to 9 m high, with leaves with narrowly-winged petioles; white flowers tinged with red in bud; and rough, somewhat bumpy-skinned, yellow-green fruit with pale yellow, sour pulp. Common in garden areas, fallow forests, and home gardens; occasionally naturalized; generally protected when clearing fallow vegetation for new gardens. Timber used for tool handles and firewood; leaves boiled to make tea; fruit used to make drinks, marinade raw fish, and season or garnish food; fruit also used medicinally; fruit sold as a minor cash crop in many countries. The most common "lemon" in many areas, referred to as the rough lemon locally, is usually classified as C. limon, but may in fact be the leach lime, or Mauritius papeda (C. hystrix DC.).

28. Citrus reticulata Blanco RUTACEAE
"tangerine," "mandarin orange"
syn. C. nobilis sensu Andrews

Indigenous to South-East Asia, the Philippines, and perhaps other areas of Malesia; a post-European-contact introduction to the Pacific Islands. Small, sometimes spiny tree, up to 9 m high, with leaves with scalloped edges and slightly winged petioles; white flowers; and somewhat flattened globose fruit with rough, loose yellow-green to bright orange skin and sweet to somewhat sour, orange pulp in easy-to-separate segments. Common in garden areas, fallow forests, and tree groves surrounding villages; common in home gardens; not found on atolls. Fruit eaten as a favoured seasonal snack food throughout the Pacific; a major seasonal cash crop in some areas such as New Caledonia, Fiji, Tonga, and Samoa.

29. Citrus sinensis (L.) Osbeck RUTACEAE
"orange," "sweet orange" syns. C. aurantium var. sinensis L.; C. aurantium ssp. sinensis (L.) Engl.

Indigenous to South China or South-East Asia; a post-European-contact introduction to the Pacific Islands. Small to medium-sized, often spiny tree, up to 12 m high, with ovate to elliptic leaves with broad-winged petioles; fragrant, white flowers; and subglobose, smooth fruit with yellow-green to orange, adherent skin and sweet to somewhat sour, orange pulp. Common in garden areas, fallow forests, village tree groves, and home gardens; occasionally naturalized in disturbed areas; usually protected when clearing new gardens; improved varieties planted in monocultural plantations in some areas such as Fiji and the Cook Islands. Wood used in light construction and for firewood; leaves used medicinally and sometimes boiled to make tea; flowers occasionally used in garlands; ripe fruit eaten and made into juice and marmalade; fruit sold as a major seasonal cash crop in many countries; the juice or concentrate is exported and used in commercial fruit-drink manufacture in Fiji, and formerly in the Cook Islands.

30. Cocos nucifera L. ARECACEAE/PALMAE
"coconut palm," "coconut"

Indigenous to southern Asia and the Indian Ocean islands, but probably an aboriginal introduction to most Pacific Islands. Tall, erect, single-stemmed palm, up to 30 m tall, with a slender, more or less curved or inclined, trunk; frond-like leaves, up to 4 m or longer, clustered at top of trunk; flowers in large axillary clusters; and large ovoid, subglobose to ovoid, or ellipsoid fruit, with a thick, fibrous husk surrounding a hard nut filled with hard, white, oily, edible pulp and, when young, with sweet water. Very abundant in all island groups, except for Easter Island; planted in extensive, regularly-spaced monocultural plantations and in haphazard, mixed-age-class plantings, often dominated by naturalized volunteer palms; sometimes undergrazed by cattle or other animals; common on atolls, often as the only cash crop; the most important intercrop or agroforestry species in smallholder mixed cropping systems on most small islands and in coastal areas. The most useful of all plants in the Pacific; features in mythology, legends, songs, proverbs, and riddles throughout the Pacific; of ceremonial importance and its leaves a sign of high rank in Polynesia and Micronesia; specific trees or two trees planted together serving as boundary markers in Tuvalu; trunk used in house construction for poles, rafters, and beams, and for wood carving, in fencing, and for animal pens, for other articles such as food containers, tools, spears and weapons, drums, canoe planking, and small canoe hulls and paddles, walking sticks, fish clubs, and, most recently, for sawn timber using portable timber mills; major source of fuel on most smaller islands, with almost all parts being used; coir and dry leaves important as tinder in making fire by friction and carrying fire; swelling at base of trunk made into food containers and large hula drums in Hawaii; bark used for scenting body oil and for smoking skirts; mature and young leaves used for weaving baskets, food containers and parcels, mats, housing thatch, tables or table mats for feasts, trays, fans, balls, weirs/barricades for communal fish drives, and other plaited ware; young leaves from germinating nut used to make a coconut-tree climbing bandage or foot-harness that is tied between the feet; unfurled immature leaves used to make skirts, body ornamentation, hats/eyeshades, baskets, fans, and fishing lures; leaves used in magic, particularly garden magic, and tied around trees in plantations as boundary markers or to ward off evil spirits and as a sign of no trespassing, or tapu; old dried leaves and husks used as mulching; midrib of leaflets or pinnules used in brooms, in weaving, toy windmills, for fishing lures and shrimp snares, to spear mudworms, small arrows, musical instruments, head-dresses, combs, for stringing fish, and oilrich seed kernels burnt for illumination, fastening thatch segments, for strengthening bonito hooks, cooking skewers, toy canoes, and in a variety of other ways; woody leaf base and midribs of fronds used for house flooring and rafters, for sandals, carrying poles, toy boats, rattles, sledges or clappers, clubs or mallets, and to beat water during fish drives and for pounding and stabilizing banks of taro beds; doubled-over midribs of fronds used as cooking tongs; midribs of young fronds used for fishermen's belts in Tuvalu; burlap-like fibrous sheath at base of fronds used as tinder, toilet paper, gauze, a filter or strainer, and to press medicine or coconut oil and to wrap bait for deep-water fishing and the earth ball on the roots of seedlings when transplanting; flower used in connection with religious ritual in Tahiti; kernel or endosperm of mature nut used raw, cooked, and fermented in a variety of ways as a staple food, as a major food for chickens and pigs and ingredient in locally produced commercial livestock feeds, for fish and rat bait, and dried to provide the socially important scented and unscented coconut oil for soap, skin oil, cosmetics, perfume, and copra, the only export crop in many rural areas; chewed pieces of mature kernel used as popgun ammunition on Tuvalu; kernel of mature nuts hung in house rafters as emergency food for up to 10 years; oil used as a preservative for tape, carvings, and other objects; soft flesh of immature nut an important weaning food and adult food; juice of immature nuts a nutritious local beverage, which is often sold, and considered a sacred offering to visitors in Kiribati and used in divination in Hawaii; oil chewed and spat on the ocean to calm the sea; sap from flower spathe used to make unfermented and fermented toddy and syrup, which are of considerable nutritional importance in Micronesia and on atolls; husk of some cultivars of green nuts eaten in atoll Polynesia and Micronesia; flower spathe sheath and dried fronds used to make torches for night fishing and for major night-time ceremonial occasions; dried sheaths used as splints to set broken bones; flower spathe sheath and frond midribs used to splint broken bones in Tuvalu; coir of husk of both green and mature nuts used to make strong fibre and cordage (sennet) for strainers, affixing tool handles, boat and house lashings, fishnets and lines, measuring tapes for garden lands, hammocks, belts, reef-walking sandals, canoe caulking, corks or stoppers, slings, toilet paper, baskets or carry bags, tying corpses for burial, and commercially to make brooms, brushes, fly whisks, doormats, and other objects; green husks used to cover earthen oven; pieces of green husk used as temporary spoons to scoop meat out of green nuts; charred husk fibre used for black dye in Tokelau; shell of nut used to make cups, bailers, small bowls, cooking vessels, funnels, utensils, storage containers, fish hooks and lures, floats, knee drums, cymbals, lagging discs, toys, a wide range of handicrafts, and high quality charcoal; roots used to make fish traps, floating cages, sand screens, and other objects; very important medicinal plant, with most parts used medicinally; leaning palms with excavated cavities or attached receptacles near the base used for water catchment; numerous named cultivars recognized in all parts of the Pacific, many of which have specialized uses, e.g. for drinking nuts, cup, or coir cordage.

31. Coffea spp. RUBIACEAE
"coffee"
spp. C. arabica L. ("Arabian coffee"); C. canephora Pierre ex. Froehn. (syn. C. robusta Linden.) ("robusta coffee"); C. liberica Bull ex Hiern. ("Liberian coffee")

Indigenous from tropical west to east Africa; an early post-European-contact introduction to many islands; not found on atolls. Shrub or small tree, up to 3 m or higher, with dark green, glossy leaves; axillary clusters of fragrant white flowers; and ellipsoid or oblong, yellow turning bright red fruit, each containing two, deeply -grooved, ellipsoidal seeds. Occasional to abundant throughout the high-island Pacific, where it is cultivated by both smallholders and as a plantation crop under a range of shade trees, such as Erythrina, Albizia, and Leucaena; naturalized, and even invasive in some areas, such as in Samoa and the Marquesas. The major export crop in Papua New Guinea, where some 200,000 smallholder growers produced 70 per cent of some K117.5 million in export earnings in 1985; also an important cash crop in New Caledonia, Vanuatu, and Hawaii and a minor crop, mainly for local sale, in Fiji, Tonga, Western Samoa, the Cook Islands, and French Polynesia. The most important species is C. arabica, the mainstay of the Papua New Guinea highlands industry and which is grown in Vanuatu, Tonga, and Hawaii, whereas C. canephora is important in other areas of Papua New Guinea, New Caledonia, and some coastal areas. C. Iiberica does not seem to have become well established anywhere. Coffee leaf rust (Hemileia spp.), which led to the complete collapse of the coffee industry in Sri Lanka, and to which only C. canephora has resistance, is present in some parts of the Pacific and poses a serious threat to the Papua New Guinea coffee industry.

32. Commersonia bartramia (L.) Merr. STERCULIACEAE
syns. C. echinata Forst.; Muntingia bartramia L.

South-East Asia, Australia, and New Caledonia to Fiji, the Caroline Islands, and the Society and Marquesas Islands; not reported from Tonga and Niue. Shrub or small tree, up to 18 m high, with toothed and sometimes lobed leaves; and small, bristly fruit. Common tree of secondary and dry forest and patches of forest or thickets in grassland; occasional in garden and fallow areas, where it is an important pioneer species. Wood used in light construction, for fishing floats, and as a fast-burning firewood; light wood used to start fire by friction in the Solomon Islands; small poles and sticks used to stake yams; fibrous bark used as cordage for fishing lines, nets, baskets, belts, girdles, headbands, and bark cloth in Melanesia; strips of bark used as a crude rope to carry produce and firewood, and for lashing in construction; roots used medicinally in Fiji.

33. Cordia subcordata Lam. BORAGINACEAE
"sea trumpet," "beach cordia"

Indigenous from the Indian Ocean to the Pacific Islands, but possibly an aboriginal introduction to many of the islands of Polynesia and Micronesia. Medium-sized tree, up to 10 m or taller, with pale greyish, slightly fissured bark; ovate leaves on long petioles; and attractive pale to bright orange, crêpey, trumpet-shaped. scentless flowers. Formerly common in coastal strand forests, particularly in disturbed sites, but increasingly scarce due to overexploitation of wood for carving; occasional in home gardens on atolls and in eastern Polynesia, occasional as a roadside tree. Currently planted as a roadside tree as part of conservation efforts in the Cook Islands; planted as an ornamental shade tree in or near settlements; formerly many famous large groves in Hawaii; a favoured shade tree in ancient Hawaii; features in Polynesian legends and chants, including the origin of fire from the underworld and, in Kiribati, it is the totem of the Karongoa clan and features in mythology; soft but durable, chocolate brown and blond wood used in general construction and is among the most favoured carving woods; also used for making canoe hulls, thwarts, rudders, weather platforms, outrigger booms and paddles, furniture, headrests, bowls, trays, plates, combs, food stirrers, food containers or boxes, airtight reef boxes for fishing equipment, fishnet floats, fishing rods, tools, coconut-climbing sticks, toys, drums and slit-gongs, tobacco pipes, images of gods (tiki), and other carved objects for sale to tourists; young saplings used for fishing rods and flutes; inner bark used as pregnant woman's girdle to provide magical powers in Kiribati and as sail ornamentation on Polynesian voyaging canoes; inner bark soaked in sea water made into dance skirts, hats, fans, baskets, garlands, and, in former times, men's clothing; used for firewood and dry bark and wood used in making fire by friction; leaves, bark, growing tips, and stems used medicinally; leaves used as an aphrodisiac in Kiribati and for love, wave, and protective magic in Micronesia; leaves used to make brown dye in Tahiti and for pig feed in Tokelau; brown dye made from roots in Tokelau; attractive orange flowers used in leis and garlands; seeds eaten, mainly by children, in Fiji, Tokelau, Puluwat, and Ifaluk; seeds used to make paste for bark cloth in Samoa.

34. Cordyline fruticosa (L. ) A Chev. LILIACEAE/AGAVACEAE
"cordyline, " "ti-plant"
syns. C. terminalis (L.) Kunth; Taetsia fruticosa (L.) Merr.

Indigenous to tropical Asia and possibly New Guinea and Malesia; probably an aboriginal introduction to most islands from Melanesia to Hawaii; a recent introduction to some smaller islands, particularly atolls, of Polynesia and Micronesia. Woody, erect shrub or tree-like, branched or unbranched perennial, up to 2 m or taller, with smooth, tough, shiny, dark green to rust or red leaves; numerous small, white, pink, or red flowers; fleshy, globose yellow or reddish fruit; and some cultivars with a large edible tuberous root. Common in garden areas, in home gardens, in ceremonial grounds, and as border plantings throughout the non-atoll Pacific; commonly naturalized in fallow vegetation; rare to occasional on atolls. Planted ornamental and pot plant; very important ceremonial and magico-religious plant, commonly planted as a garden marker or to ward off evil spirits or to discourage black magic or sorcery in garden areas; a traditionally important supplementary food plant and famine food, and important decorative plant, with numerous other cultural uses in Melanesia and Polynesia, where numerous named cultivars and hybrids exist; large, sweet, white tubers of some cultivars baked for up to four days in earthen ovens to be consumed as food, sweets, refreshment, or confectionery, or, in Hawaii, made into an alcoholic beverage known as okolehao; leaves important for parcelling food, dancing skirts, and body ornamentation.

35. Cycas spp. CYCADACEAE
"cycad," "palm fern"
spp. C. circinalis L.; C. rumphii Miq.; C seemannii A. Braun

Indigenous to tropical Asia, Australia, and as far east as Fiji and Tonga, although it is possibly an aboriginal introduction into some of these areas; a recent introduction to eastern Polynesia, Micronesia, and some atolls. Small dioecious, palm-like, rarely-branching plant, up to 2 m or taller, with a sturdy, brown-ringed trunk; dark green, palm- or fern-like, pinnate fronds; apical flower clusters; and hard, somewhat flattened, green turning reddish brown, globose, nut-like fruit. Component in savanna grasslands in New Guinea, occasional in fallow areas, open forests, and savanna or grassland areas in Melanesia; occasional in coastal areas; occasional in home gardens and urban areas on both high islands and atolls. Planted ornamental and ceremonial plant; a ceremonial plant in Vanuatu, where it is planted in sacred meeting grounds (nakamal) and areas where pigs are killed; bark sap used as glue in the Solomon Islands; seed kernels processed into flour as a famine or ceremonial food in areas of Melanesia, Polynesia, and Micronesia (because the seed contains highly toxic hydrocyanic acid, a detoxification process is involved; evidence shows deleterious effects in populations eating the seeds without proper processing); staminate cones also reportedly edible; fruit strung together to make children's toys or rattles; fruit and bark used medicinally in the Solomon Islands.

36. Delonix regia (Bojer) Raf. CAESALPINIACEAE/LEGUMINOSAE
"poinciana," "royal poinciana," "flame tree," "flamboyant," "flame of the forest"

Indigenous to Madagascar; a pre-World War I introduction into most of the Pacific Islands. Rapidly growing, medium-sized, spreading, broad-crowned deciduous tree, up to 10-12 m high, with bipinnate leaves; corymbose racemes bearing large, brilliant crimson to red-orange, showy flowers; and long, flattened, woody, dark reddish brown turning black, many-seeded pods. Common in home gardens and towns, and occasional in rural areas. Planted ornamental and shade tree; one of the most popular roadside and park trees; wood used occasionally in light construction and for firewood; flowers and leaves used in ornamentation.

37. Dracontomelon vitiense Engl. ANACARDIACEAE
"dragon plum"

Possibly indigenous from Vanuatu to Fiji, Rotuma, and Samoa, although possibly an aboriginal introduction; probably not present in Micronesia; not found on atolls. Medium-sized, buttressed tree, 8-20 m high, with pinnate leaves; white flowers; and green or yellow fruit, about 3.5 cm in diameter, with yellowish pulp. Occasional in garden and fallow areas, tree groves, and mature fallow forest or open forest; occasional in home gardens and urban areas. Wood used in light construction and for firewood; ripe fruit an important seasonal snack food and commonly sold at urban markets in Fiji and Vanuatu; parts used medicinally; superstitions concerning the tree exist in Fiji.

38. Erythrina variegata var. orientalis (L.) Merr. FABACEAE/LEGUMINOSAE
"coral tree," "Indian coral tree," "tiger's claw," "dadap"
syns. E. indica Lam.; E. corallodendron var. orienfalis (L.) Merr.

Indigenous from the Indian Ocean to the Pacific Islands; possibly an aboriginal introduction into some areas of Melanesia, Polynesia, and Micronesia, but a recent introduction into some of the smaller islands of the eastern Pacific. Medium to large deciduous tree, up to 25 m high, with the trunk and branches usually bearing coarse, spiny thorns; broadly triangular-ovate trifoliate leaves; numerous attractive, claw-like, bright dark red or scarlet flowers; and black pods containing bright red to brownish red, ovoid seeds. Common in rural agricultural areas as a vegetatively reproduced living fence or boundary marker, as a shade tree for coffee or a windbreak; occasional in home gardens and as a roadside tree; occasionally naturalized. Planted as living fencing, boundary markers, living pig and livestock pens, as a nitrogen-fixing and green manure plant, as shade for coffee in Papua New Guinea, and as an omamental; trunks and thick branches used in light construction and in canoe construction in the past; flowers used in garlands in the past; seeds of ornamental and religious value; bark and seeds used to stupify fish in Vanuatu.

39. Eucalyptus spp. MYRTACEAE
"eucalyptus," "gum tree"
spp. E. camaldulensis Dehnh. ("Murray red gum"); E. citriodora Hook. ("lemon-scented gum"); E. deglupta Blume ("Mindanao gum," "Bagras euca Iyptus"); E. saligna Sm. ("flooded gum"); E. rereticornis Sm. ("forest red gum")

Indigenous to South-East Asia, the Philippines, Australia, and New Guinea; recent introductions into most of the Pacific Islands. Trees, often with smooth, peeling bark, resinous sap, thick, aromatic leaves, umbellate flowers, and woody, capsular fruit. Abundant in some areas such as Papua New Guinea, the Solomon Islands, Samoa, and Tonga, where Eucalyptus spp. are among the main species used in reforestation projects; occasional in experimental plantings and occasional in rural areas, home gardens, and as a roadside tree in some areas. Wood from large plantations produced to be chipped for export, mainly for the production of paper; wood used locally for fencing, firewood, and occasionally for light construction; occasionally planted as an ornamental; used medicinally in New Guinea.

40. Euodia hortensis Forst. RUTACEAE
"island musk"
syn. Evodia hortensis Forst.

Indigenous to South-East Asia and probably to parts of western Melanesia, but probably an aboriginal introduction into Fiji, Tonga, Samoa, and Niue. Shrub or small tree, up to 4 m high, with pungently aromatic, 1- to 3-foliate leaves; small, white, yellowish or greenish white flowers in particulate clusters; and small, subglobose, pale brown fruit. Commonly planted in home gardens and around villages; naturalized or native in rural agricultural areas and as an understorey shrub in thickets, lowland forest, fallow forest, and tree groves. Cultivated ornamental or ceremonial plant and boundary marker; often planted in cemeteries or burial grounds; branches placed near gardens to ward off pigs; leaves and flowers used in garlands, worn behind the ear, and used to scent coconut oil; bark, roots, and leaves used medicinally for a wide range of maladies; used in garden magic in Melanesia.

41. Fagraea berteriana A. Gray ex Benth. LOGANIACEAE/ GENTIANACEAE
"pua" (Polynesia)
syns. F. berteroana A. Gray ex Benth.; F. galilai Gilg. & Bened.; F. racemosa Jack. ex Wall. (sp?); F. sair Gilg. & Bened.

Indigenous from New Guinea, north-eastern Queensland, and New Caledonia to the Marianas and Caroline Islands and as far east as the Marquesas, the Austral Islands, and Hawaii; possibly an aboriginal introduction into some of these areas. Small to large, glabrous, often-branching, commonly epiphytic tree, 1.5-20 m high, with rather thick leaves; very fragrant, fleshy, tubular, creamy white turning rich yellow flowers; and rather succulent, yellow turning orange to bright red fruit. Frequent in open lowland and limestone forests, in poorly drained sites, and in degraded uplands and cloud forest; occasionally planted or protected in active garden areas and occasional in home gardens. Planted or protected ornamental; tree features in legends and is sacred in Hawaii, French Polynesia, the Cook Islands, and other areas of Polynesia; important tree for live fencing or pig pens in the Solomon Islands (F. racemosa); soft but durable timber used for house posts, tools, combs, and in wood carving of idols in Tahiti; inner bark used in treating asthma and diabetes in Fiji; leaves used for wrapping food or sealing earthen ovens; fragrant flowers used in garlands and to scent coconut oil in Polynesia and Melanesia; ripe fruit a favoured food of wild pigeons.

42. Ficus spp. MORACEAE
"banyan," "strangler fig"
spp. F. aoao Warb.; F. microcarpa Linn. f.; F. obliqua Forst. f.; F. prolixa Forst. f.; F. virens Ait.

Most species indigenous from Malaysia, Indonesia, and Australia to as far east as the Line, Marquesas, Tuamotu, and the Pitcairn Islands. Small to very large (depending on the habitat), spreading, broad-crowned trees, up to 20 m or higher, commonly beginning as epiphytes; with adventitious, aerial roots; smooth to coriaceous leaves; and small, globose, fig-like fruits. Occasional to common in open and seaside forests, secondary forests, and occasionally protected in garden areas; occasional in ceremo nial sites and as roadside or trail-side trees; F. prolixa seems to be one of the only species capable of long-term colonization of residual pinnacles in strip-mined areas of Nauru and could become dominant in the disclimax vegetation. Banyan trees are regarded as having spiritual importance in many areas and are the focus of creation mythology and cosmogony in Polynesia and Melanesia; important in rituals and ceremonies in Melanesia; found in and serve as the focus for the ceremonial meeting places (nakamal) in Vanuatu; timber and aerial roots used in light construction, tool making, and for firewood; occasionally used for canoe hulls in Papua New Guinea; large aerial roots used for canoe masts and hauling loads on Ulithi; best fibre used to make bark cloth in Vanuatu, Fiji, Niue, and Tahiti, and for making very large seine nets in Tahiti; sap (latex) used as chewing gum, putty, and caulking, to dye ceremonial sashes and belts, and for waterproofing; leaves used in garlands and ceremonial dress by Pacific Islanders and Indians in Fiji; roots, aerial roots, bark, fruit latex, and leaves used medicinally; fruit cooked and eaten, sometimes mixed with coconut syrup; sap used for chewing gum; fruit important food of birds and fruit bats.

43. Ficus spp. MORACEAE
"native figs," "edible figs," "sandpaper cabbage"
spp. F. aspera Forst. f.; F. copiosa Steud. ("sandpaper cabbage"); F. dammaropsis Diels ("highlands breadfruit"); F. nodosa T. & B.; F. pungens Reinw.; F. scabra Forst. f. ("sandpaper fig"); F. storkii Seem.; F. tinctoria Forst. f. ("Dyer's fig"); F. trachypison Ltb. & K. Schu.; F. vitiensis Seem. ("Fiji fig"); F. wassa Roxb. ("sandpaper cabbage")

Different species indigenous from South-East Asia to Polynesia and Micronesia; many of the cultivated species are probably aboriginal introductions or selected cultivars derived from indigenous species. Shrubs to medium-sized (depending on the habitat), slender to widespreading trees, up to 10 m or higher, with variable, smooth to coriaceous or scabrous (sandpaper-like) leaves; and globose, subglobose, or pearshaped, green fruit that become yellow, orange, dull red, or purple when mature. Cultivated or protected in garden areas and home gardens and around villages; common to occasional in grasslands, open forest, and fallow areas; common to occasional in coastal strand and lagoon-side vegetation; some species well adapted to coastal and atoll environments. Planted food trees in Melanesia and in Kiribati and other parts of Micronesia; some species believed to restore soil fertility; wood sometimes used in light construction and for digging sticks, yam stakes, canoe connectives, fishnet frames, fish-trap parts, earth sieves, and for firewood and making fire by friction; fibrous branches used to clean teeth; roots used in scoop-net frames in Kiribati, and ropes for fish drives in Puluwat; best fibre used as cordage for clothing, lashing, fishnets, strapping bundles; best fibre of roots used for fish lures on Ifaluk and for cordage and chewed to make fuses or tapers used in medical treatment in Tuvalu; leaves of some species used to wrap food for cooking; leaves of some species used as a sandpaper substitute and for scouring pots; leaves and fruit of a range of species are eaten and are of particular nutritional importance in Melanesia, where they are cultivated or gathered from wild plants; leaves also an important pig, horse, and livestock feed; fruit an important food of fruit bats, pigeons, and other birds; ripe or green fruit of F. tinctoria processed or cooked in many ways to produce a minor staple and made into puddings and dried, preserved food in Tuvalu and Micronesia; fruit formerly used to dye bark cloth, hats, mats, etc.; roots used to produce red dye for pandanus in Kiribati; sap used to produce red dye for face in Tahiti; fruit used as ammunition for popguns in Tuvalu; yellow leaves used in body ornamentation in Tuvalu; young leaves and young inner bark used medicinally in Kiribati.

Appendix One hundred Pacific Island agroforestry trees (4)

44. Gardenia taitensis DC. RUBIACEAE
"Tahitian gardenia," "tiare Tahiti" (Tahiti)

Believed to be indigenous from Vanuatu to Fiji, Tonga, Niue, Wallis and Futuna, and possibly Samoa; probably an aboriginal introduction into some of these islands and to eastern Polynesia; a recent introduction into Hawaii, Micronesia, and most atoll countries (endemic species of Gardenia, some of which are found in agroforestry systems, exist in many island groups). A shrub or small gnarled tree, up to 6 m high, with shiny, bright green, obovate leaves; attractive, very fragrant, pure white, tubular, spreading 5- to 8-petalled, solitary flowers; and globose, ribbed, yellow-green fruits. Occasional to common in home gardens and villages, sometimes planted as hedges, especially in the Cook Islands, French Polynesia, and Hawaii; apparently naturalized in some areas such as Hawaii; rare to uncommon on sea cliffs and rocky islets, almost always on limestone. Widely planted ornamental that is cultivated commercially in Tahiti and Hawaii; planted near graves of chiefs in Tuvalu; national flower of the Cook Islands and Tahiti; features in legends and songs in Polynesia and Micronesia; used in love magic and sorcery in Tuvalu; wood carved into bows and cricket balls in Tuvalu, and netting needles and gauges in Tokelau; fragrant white flowers used in leis and garlands and worn in slits in and behind the ear, and in the hair; flowers and fruit used for scenting coconut oil, which is produced commercially in Tahiti and Rarotonga; leis and head garlands sold and exported from Tahiti and Hawaii; used medicinally in Melanesia and Polynesia; selected cultivars with large leaves and flowers recognized in Polynesia.

45. Garuga floribunda Decne. BURSERACEAE
syn. G. pacifica Burkill

Indigenous from the Philippines, Java, and Melanesia to Tonga and Samoa; not reported present in Fiji; possibly an aboriginal introduction from Melanesia to Tonga and Samoa. Medium tree up to 3 m or higher, with leaves commonly with a flush of red somewhere in the crown, crowded near the ends of branches; small flowers in particulate clusters; and small, subglobose, fleshy, green fruits that turn black at maturity. Common in lowland forest and in open or disturbed forest, often in drier sites; occasional in and near garden areas, as living fencing around plantations, and as a protected species near settlements. Planted as living fencing Or boundary markers in Vanuatu and Tonga; timber used in general construction, for fencing, and firewood; bark used medicinally; fruit edible.

46. Gliricidia sepium (Jacq.) Kunth ex Walp. FABACEAE/LEGUMINOSAE
"gliricidia," "madre de cacao (mother of cocoa)," "Nicaragua cocoa shade"
syn. Robinia sepium Jacq.

Indigenous to Central America and northern South America; a recent introduction into the Pacific Islands. Small, semi-deciduous, fast-growing tree, up to 10 m tall, with pale bark and ascending stems; pinnately compound leaves; stiff, short racemes on the older branches bearing rose or paler pink, rose-tinged flowers with a pale yellow, central blotch; and linear-oblong, flattened pods containing purplish brown seeds. Occasional to common in rural areas as living fencing, mainly around pastures; occasional in home gardens and as a roadside tree; seems to grow well on calcareous soils of atolls. Widely used in tropical America and elsewhere as a shade tree for cocoa, bananas, and coffee; an increasingly important living fence and windbreak in Fiji and Vanuatu; valuable nitrogen-fixing and green manure plant; occasional as an ornamental tree in home gardens and along roadsides; used as firewood; leaves used as fodder for pigs, goats, and cattle.

47. Glochidion spp. EUPHORBIACEAE "glochidion"
spp. G. concolor Muell.-Arg.; G. mearianum Muell.-Arg.; G. ramiflorum Forst.
(syns. Phyllanthus ramiflorus [Forst.] Muell.-Arg., G. tannaense Guill.); G. philippicum (Cav.) C.B. Rob.; G. pomiferum?

Indigenous from tropical Asia to the Marquesas, Tuamotus (Makatea) and Henderson Island in eastern Polynesia, and to the Marianas and Caroline Islands in Micronesia; the genus is extremely variable, with endemic species represendng it on many islands. Small to medium-sized trees, up to 12 m or taller, with 2-ranked, elliptic leaves; small, pedicellate, yellow or orange to yellowish green flowers in axillary fascicles; and grooved, depressed-oblate fruit. Common in open or secondary forest, thickets, grasslands, and fallow vegetation; occasional in garden areas and grazing areas, often volunteering in fallow vegetation; commonly protected when clearing for new gardens; occasional in home gardens. Durable wood used in light construction, in tool making, and for firewood; leaves and bark used medicinally throughout the Pacific; red dye made from the bark in New Guinea.

48. Gnetum gnemon L. GNETACEAE
"gnetum," "joint fir"

Indigenous from Assam in India through southern Asia and Malesia to the Caroline Islands and Fiji; possibly a naturalized aboriginal introduction into some islands or the result of a selection process from wild varieties. Small to medium, deep-rooted, shade-tolerant tree, up to 15 m high, with broadly elliptic-lanceolate leaves; and ellipsoid fruits that turn orange-red at maturity. Occasional in lowland, ridge, and mature fallow forest; cultivated in or near gardens and in home gardens in the Solomon Islands; common in planted Artocarpus-Pandanus tree groves in highland New Guinea. Tree serves as a support for yams and other shade-tolerant climbers; wood used for house beams; best fibre used to provide cordage for fishing nets and line, and string bags; fruits, flowers, and young leaves eaten in a variety of ways; cooked, dried fruit often stored in the Solomon Islands; leaf sap used medicinally.

49. Guettarda speciosa L. RUBIACEAE
"guettarda"

Indigenous from eastern Africa and tropical Asia to the Marshall Islands and southeastern Polynesia, but not to Hawaii. Small to medium-sized tree, up to 20 m high, with large, obovate leaves, and fragrant, long, tubular, white flowers born in cymose clusters; and hard, yellowgreen, ovoid fruits that turn black at maturity. Common to infrequent in coastal strand forest, thickets, and open vegetation on rocky and sandy shores; common in regrowth in older stripmined areas on Nauru; common in coconut plantations and garden areas in Kiribati and on other atolls; common to occa signal in home gardens on atolls. Important in Kiribati and Tuvaluan legends and mythology; national flower of the Marshall Islands; names of the leaf and the plant associated with phases of the moon and stations of the sun in Kiribati; hard and durable wood used in light construction, for pilings, fish-trap stakes, stakes to hold garden mulch in place, coconut huskers, fishing poles, floats, spears, thatching needles, fishing rods, fishnet and bird-net handles, stilts, eel traps, fruitharvesting sticks, bowls, slit-gongs, for canoe hulls, supports, steering paddles, bailers, poles for poling canoes, and floats; the most desired wood for tape-beating anvils in Tonga; wood used in games in Fiji; used for firewood and for making fire by friction; leaves used in fires for drying pandanus leaves and for toilet paper in Tokelau; dead wood used to smoke skirts in Tuvalu; bark, leaves, flowers, and fruit used medicinally; leaf litter considered the most important component and source of black topsoil, which is mixed with compost for the cultivation of giant swamp taro, pandanus, and other crops in Kiribati; leaves, either alone or with other leaves, provide one of the most important composts in Kiribati and Tuvalu; all pastes or preserves spread on Cuettarda leaves for sun drying in Kiribati; leaves used to cover earthen oven and as disposable plates in Micronesia; leaves provide a jet-black hair dye in Kiribati; leaves used as a baby's wash cloth in Ulithi; leaves used for pig feed in Tokelau; leaves used in head garlands and worn in ear slits in Tuvalu; flowers used in garlands and for scenting coconut oil; flowers and young leaves soaked in water to provide deodorant or aphrodisiac in Kiribati; parts used as love charms in Ulithi.

50. Heliconia indica Lam. HELICONIACEAE/STRELITZIACEAE
"heliconia"
syns. H. paka A.C. Smith; H. solomonensis Kress.

Considered indigenous to the Palaeotropics and the western Pacific as far east as Fiji and Samoa, but possibly a naturalized aboriginal introduction in some areas and to Tonga; not reported in Micronesia. Erect, coarse, rhizomatous herb, superficially resembling a banana, up to 2 m or taller, with a pseudostem comprised of tightly rolled leaf sheaths; oblong leaves up to 2-3 m long; flowers with overlapping, scarlet and yellow or greenish bracts; and yellow fruits. Common in dense forests, secondary forests, fallow, and garden areas from New Guinea to Samoa and Tonga; occasionally cultivated or protected, in rural and home gardens. Leaves used as thatch for temporary shelter, umbrellas, make-shift sleeping mats, to cover earthen ovens, to wrap food for cooking, especially starchy puddings and the staple "laplap" in Vanuatu (where it is referred to as "laplap leaf"); fibre from petiole and midrib processed into tauanga and used to strain coconut oil in Samoa; flowers cooked as a famine food in Fiji; pseudostems and heated leaves used medicinally.

51. Hibiscus manihot L. MALVACEAE
"bush hibiscus spinach," "edible hibiscus," "slippery cabbage," "sunset hibiscus," "aibika" (PNG Pidgin)
syn. Abelmoschus manihot (L.) Medik.

Indigenous to South-East Asia; an aboriginal introduction to Melanesia and possibly to Tonga from Fiji; probably a recent introduction into other areas of Polynesia and Micronesia. Erect, perennial shrub, 1-5 m high, with slightly fleshy, variably-shaped, entire to deeply lobed or laciniate, bright green to red-green or purplish leaves; yellow, hibiscus-like flowers with dark purple centres; and a beaked, oblong, dehiscent capsule containing numerous pubescent seeds. Common to abundant in both rural and urban food gardens; common in home gardens. A wide range of cultivars planted as an intercrop and uncommonly as a monocrop in small plots as a supplementary food crop; nutritious slippery green leaves and young shoots cooked as one of the major leafy green vegetables throughout Melanesia and in Tonga; an important cash crop sold at local produce markets; recently promoted as a nutritious vegetable in Samoa, the Cook Islands, Tuvalu, Kiribati, and the Federated States of Micronesia; leaves also used medicinally to cure coughs, sore throats, dysentery, and stomach aches; cordage for dancing skirts made from stems in Yap.

52. Hibiscus rosa-sinensis L. MALVACEAE
"hibiscus," "red hibiscus"

Origin uncertain, but probably indigenous to eastern Africa or southern Asia; probably an aboriginal or very early post-European-contact introduction to Melanesia and parts of Polynesia; a pre-World War 11 introduction into most of the other islands of Polynesia and Micronesia. Shrub to small tree, up to 4 m high, with glabrous, dark green to variegated, serrate leaves; and conspicuous 5-petalled, red, pink, white, or yellow flowers. Occasional in rural areas and common to abundant in urban areas as an ornamental and hedge or living fence plant. Planted ornamental; flowers used in garlands and for decoration; leaves fed to goats.

53. Hibiscus tiliaceus L. MALVACEAE
"beach hibiscus tree," "beach mallow"
syns. Pariti tiliaceus (L.) A. St. Hill; P. tiliaceum Britt.; Paritium tiliaceum (L.) A. St. Hill

Pantropical and subtropical and indigenous to the Pacific Islands; possibly an aboriginal introduction to some areas. Spreading, often scrambling tree, 3-10 m high, with cordate, gray-green leaves; attractive, few-flowered clusters of yellow, 5petalled flowers that have a dark maroon to brown centre and age to salmon-pink before falling; and 5-celled, dehiscent, capsular fruits. Abundant in coastal and lowland thickets, along the inner margins of mangroves, often along river banks, common and often invasive in disturbed and open forest and degraded upland areas; common in garden areas, plantations, and fallow areas, and often protected or only severely pruned when clearing new garden plots; occasional to common in home gardens and around villages; common in grazing lands. One of the most useful trees in the Pacific; commonly planted as living fencing and animal pens and in coastal areas, near houses, in gardens, and as an ornamental or shade tree; a creeping variety planted as wind-break in Hawaii; its presence in forested areas considered a sign of former cultivation in Hawaii; features in eastern Polynesian legends and Hawaiian fire-making legends; commoners not allowed to cut branches without permission of chiefs in Hawaii; branches borne in battle by priests as a good omen and allowed to fall in retreat in Hawaii; born by attendants at presentation of first fruits to kings on Easter Island; branches used as tapu markers to delimit restricted areas in Hawaii; used to make spears used in typhoon magic in Ulithi; soft wood used in light construction and wood carving, for house rafters, pig-tethering posts, for canoe outriggers, spreaders, bailers, booms and occasionally hulls, fishing rods, hoists and floats, fishnet frames and handles, bows, fruit-picking rods, tools and tattooing comb handles, kite struts, jackstraw sticks, pestles, breadfruit splitters, coconut huskers, net floats, spears, shore-line posts to delineate fishing zones, fishing gear containers, noddy bird net handles and frigate bird nesting platforms (Nauru), and other purposes; a decent firewood, especially for slow smoking; used in making fire by friction; wood dried for six months used for fireworks in Hawaii; best fibre used as canoe caulking and to make cordage for clothing and dancing skirts and kilts, coconut-climbing bandages or foot harnesses, mats, sandals, sewing tape, bark cloth paint brushes, making fishnets, fishing line and lures, slings, kava strainers, sandals, tying corpses in tape, and cordage for tying, lashing and binding canoes, housing, and other things; bark used to strain kava in Pohnpei to give it its preferred slimy consistency; leaves, terminal buds, unopened flowers, and bark used medicinally, with leaves being used to reduce hemorrhaging and for treating neurological disorders; leaves used to parce] food, especially seafood, as plates, and to line and cover the earthen oven; leaves widely used as toilet paper; flowers used in garlands in Hawaii; bark, shoots, and sapwood eaten in New Caledonia and other parts of Melanesia; leaves occasionally added to compost in Kiribati and Tuvalu; a number of distinct varieties or cultivars recognized in Melanesia and Polynesia.

 

54. Inocarpus fagifer (Park.) Fosb. FABACEAE/LEGUMINOSAE
"Tahitian chestnut"
syns. I. edulis Forst.; I. fagiferus (Park.) Fosb.; Aniotum fagiferum Park.

Indigenous to Malesia and considered indigenous to the Pacific Islands as far east as the Society, Marquesas, and Austral Islands, but possibly an aboriginal introduction into some areas, such as Niue, where it is found mainly associated with human activity; a recent introduction into Hawaii. A medium to large, buttressed tree, up to 30 m high, with leathery, oblong leaves; fragrant, yellowish white or pinkish flowers; and fleshy, somewhat ovoid or kidney-shaped, yellowgreen fruit containing a large, chestnut-like, edible kernel. Common in lowland forests, particularly in poorly drained areas, inner margins of mangroves and along streams; occasional in garden areas and protected when clearing new gardens; occasional in towns and as a roadside and path-side tree. Features in Polynesian mythology and is the sacred tree of the people of Moce, Fiji, who are referred to as Vuata Ivi (fruit of the ivi); to injure the tree in any way was taboo on Moce and the first fruits were offered to priests; traditional calendar associated with its fruiting in Lau, Fiji; commonly planted or protected as boundary markers; wood used in general construction and wood carving, for tool handles, kava bowls, tape beaters, weapons, packing boxes, etc.; used for firewood; bark a source of dye in Tahiti; leaves used for indicating the value of pigs for ceremonial presentation in Vanuatu; leaves, bark, and stems used medicinally; ripe seed, which tastes like chestnut, eaten cooked as a seasonal staple and preserved in the past in Polynesia and Melanesia; cooked seeds an important seasonal cash crop; gum from fruit used for caulking canoes in Uvea.

55. Intsia bijuga (Colebr.) O. Ktze. CAESALPINIACEAE/LEGUMINOSAE
"ipil"
syns. I. amboinicus DC.; Afzelia bijuga Colebr. ex Volkens; Macrolobium bijugum Colebr.

Indigenous from eastern Africa and Madagascar, southern Asia, Taiwan, and Malesia to the Caroline Islands, Fiji, Rotuma, and Samoa; possibly an aboriginal introduction from Fiji to Tonga. Medium to large tree, up to 35 m high, with small but tresses; compound leaves with 1-3 pairs of broadly ovate, but asymmetric leaflets; pure white to pink flowers with a red claw and red to purple stamens in dense terminal panicles; and thick, leathery, oblong pods containing orbicular, black seeds. Occasional in coastal and lowland forests and thickets and on inner margins of mangroves; sometimes found inland and protected in garden areas; occasionally planted in villages in Fiji. Planted ornamental and sacred tree; one of the most sacred trees in Fiji; durable, attractive, dark red-brown wood used in house construction, for canoes and canoe masts, fencing, and furniture, and most desired for wood carving, for food and kava bowls, headrests, containers, tape beaters, combs, walking sticks, war clubs, and a variety of other articles of inter-island trade between Samoa, Tonga, and Fiji; used for firewood; roots and bark used medicinally in Melanesia; seeds used for dancing anklets in Samoa.

56. Kleinhovia hospita L. STERCULIACEAE
"kleinhovia," "puzzle tree," "guest tree"

Indigenous from tropical Africa and Asia through Malesia to the Caroline Islands and Samoa, Tonga, and the Society Islands. Medium tree, up to 20 m high, with ovate to cordate, palmately-nerved leaves; pink or rose-coloured flowers in panicles; inflated, S-parted, papery, pink, capsular fruit containing usually one globose white seed. Common in secondary forest, clearings, and fallow areas, often forming groves; one of the most common pioneer species in Melanesia and Samoa; often felled or ringbarked when clearing new garden plots, but left standing in some areas of Vanuatu. Chosen as worthy of inclusion in hedgerowlalleycropping trials in the Solomon Islands; wood used for light construction, canoe floats, floats for fishnets, yam stakes; considered one of the best firewoods and favoured for making fire by friction; strips of bark provide temporary cordage for binding garden produce or firewood; leaves used to seal earthen ovens, wrap food, and as tobacco wrappers; bark, shoots, and leaves used medicinally; young leaves cooked as a vegetable.

57. Leucaena leucocephala (Lam.) de Wit MIMOSACEAE/LEGUMINOSAE
"leucaena," "koa haole" (Hawaii), "lead tree," "wild tamarind"
syns. L. glauca (L. ex Willd.) Benth.; Mimosa leucocephala Lam.; Mimosa glauca sensu L.; Acacia leucocephala (Lam.) Link

Indigenous to tropical America; pre-World War I introduction throughout most of the Pacific. Erect, slender shrub or small tree, 1-5 m high, with dense wood; bipinnate leaves; pale green to white, globose flowers in dense clusters; and clustered, flat, dehiscent, dark brown pods containing flat, glossy brown seeds. Common to very abundant in rural areas of large islands, where it has become naturalized in extensive stands; occasional as living fencing; occasionally deliberately planted as fuelwood plantations or as shade in coffee plantations; occasional in home gardens. Planted as shade for coffee plantations and in fuelwood plantations surrounding urban areas in Papua New Guinea; nitrogen-fixing ability well-known; timber used in light construction, for fencing, and as a very important source of firewood; firewood sold commercially in Tonga; foliage and green pods an important fodder and green manure; green pods and seeds occasionally eaten in some parts of the Pacific; seeds used in necklaces and handicrafts. Improved, fast-growing "giant" varieties from Hawaii inttoduced into some areas. Heavily defoliated throughout the Pacific Islands in the mid-1980s by psillid insect infestations, which, due to poor recovery in Tonga, have threatened its status as one of the best species for fuel-wood plantations.

58. Macaranga spp. EUPHORBIACEAE
"macaranga"
spp. M. aleuritoides F. Muell.; M. carolinensis yolk.; M. graeffeana Pax & Hoffm.; M. harveyana (Muell.-Arg.) Muell.-Arg.; M. seemannii Muell-Arg.; M. similis Pax & Hoffm.; M. tanarius (L.) Muell.-Arg.; M. thompsonli Merr.

Indigenous to tropical Africa, Madagascar, tropical Asia, and throughout Malesia to northern Australia, the Caroline and Gilbert Islands in Micronesia, and the Cook. Society, and Austral Islands in eastern Polynesia; possibly an aboriginal introduction into some areas, such as the Gilbert Islands, and a recent introduction into some areas, such as the Marshall Islands and Hawaii. Small to medium, monoecious or dioecious, soft-wooded tree, up to 10-20 m high, with, large, variable, but commonly peltate or ovate, leathery, distinctly veined leaves; variable flowers that are sometimes reddish; and small, dehiscent, capsular, often spiny fruits containing seeds with a fleshy testa. Common in secondary forest, fallow vegetation, and patches of forest in open country; one of the main pioneer species in abandoned garden areas; occasional in inland forest, often on limestone; often felled, but occasionally preserved when clearing for new gardens; some species seem well adapted to atolls; infrequent in home gardens. Timber used in house construction for rafters, walling frames, flooring and battens, for wood carving, banana cases, and other purposes; a major source of firewood sold at urban markets; leaves used to seal earthen ovens, to parcel seafood to keep it fresh, and to parcel food before cooking; leaves used medicinally for a range of maladies and to induce abortions.

59. Mangifera indica L. ANACARDIACEAE
"mango"

Probably indigenous to India and Burma; an early post-European-contact introduction in most areas of the Pacific; possibly an aboriginal introduction into some areas? Large, dense, broad-crowned tree, up to 30 m high, with leathery, lanceolate leaves; pinkish white flowers borne in terminal panicles; and green to orange or red fruit containing sweet, juicy, often stringy, yellow-orange to dark orange flesh and a flat, woody, adhering seed. Common to abundant in garden and fallow areas and usually protected when clearing fallow vegetation for new gardens; common as wild, possibly naturalized, trees in mature fallow forest and along rivers in dry areas; common in home gardens in rural and urban areas and as a street or roadside tree; occasionally planted, mainly using improved cultivars, in regularly-spaced orchards; more common and produces more fruit in drier or intermediate climatic areas, and produces fewer fruit in areas of high rainfall. A sacred plant, the leaves being used in Hindu ceremonies in Fiji; timber occasionally used in light construction and for firewood; fruit eaten ripe and green, with ripe fruit occasionally made into jam or chutneys and green fruit into pickles by Indians in Fiji; an important seasonal cash crop for local sale in Polynesia, Melanesia, and the larger islands of Micronesia; exported either as whole ripe fruit, puree, or juice from Fiji, Tonga, the Cook Islands, and Hawaii; bark used medicinally in Fiji and New Guinea.

60. Metroxylon spp. ARECACEAE/PALMAE
"sago palm," "ivory-nut palm"
spp. M. amicarum (Wendl) Becc.; M. sagu Rottb. (syn. M. rumphii [Willd.] Mart.); M. salomonense (Warb.) Becc. (syn. M. bougainvillense Becc.); M. vitiense (H. Wendl.) H. Wendl. ex Hook. f.; Coelococcus spp. Wendl.; Sagus spp. Steck

Indigenous to Indonesia (where now perhaps only cultivated), New Guinea, the Solomon Islands, Vanuatu and Fiji in Melanesia, and the Caroline Islands in Micronesia; probably an aboriginal introduction into some islands, such as Rotuma and Samoa, and a recent introduction into other islands. Medium to tall, single-stemmed, columnar palm, up to 10-20 m or higher, with denselypacked, large, pinnate fronds; stout, often spiny petioles; a large, single, terminal inflorescence that is produced shortly before the tree dies; and subglobose, light brown fruit with overlapping, snake-like scales and a very hard, white kernel (endosperm). Grows extensively in freshwater swamps and along streams and rivers in Irian Jaya, Papua New Guinea, and the Solomon Islands, and in localized concentrations on Viti Levu in Fiji; planted deliberately in some areas of western Melanesia and in non-coastal stream valleys, around villages, and home gardens both in eastern Melanesia and in Vanuatu, Fiji, Rotuma, and Samoa; occasionally naturalized in apparently nonindigenous habitats. Important wild and cultivated staple food plant in Irian Jaya and mainland Papua New Guinea and in localized areas of island PNG and the Solomon Islands, such as in New Georgia, Choiseul, and the Langalanga Lagoon area of Malaita, where over a million people use sago regularly; almost pure starch from the trunks of immature (pre-flowering) trees is removed through a laborious process involving felling, splitting, pounding, washing, kneading, and drying; floral bud often removed to improve starch yields; numerous cultivars/varieties exist where trees have been subject to selection and planting; a minor staple or famine food in most of the islands of Papua New Guinea, Melanesia, the Solomon Islands, and other areas; starch used in making pudding or desserts in most areas, including Rotuma and Samoa, although rarely in Fiji; sago starch a major item in traditional "Hiri" trade networks of the Gulf of Papua; sago pith a pig food in some areas; trunk of rotting palms a source of edible beetle larvae; meristemlheart of palm sold at urban and roadside markets in Fiji and cooked in curries by Indians; fronds considered among the best thatching for roofing and walling, which can last as long as 10 years; seeds of M. amicarum used in necklaces and handicrafts and for buttons in western Micronesia.

61. Morinda citrifolia L. RUBIACEAE
"beach mulberry," "Indian mulberry"

Indigenous from tropical Asia and Australia to south-east Polynesia and Hawaii and the Marshall and Gilbert Islands in Micronesia; probably an aboriginal introduction into at least the eastern part of its range. Shrub or small tree, up to 10 m high, with broadly elliptical to obovate, shiny, dark to pale green leaves; head-like clusters of white flowers; and fleshy, globose-ovoid, somewhat cone-like, yellowish white, somewhat gelatinous when ripe, very strong-smelling fruit. Occasional to common in coastal vegetation, along streams, or on the inner margins of mangroves, as an understorey plant in open forests, and in fallow areas, thickets, and waste places; often an early pioneer in grasslands and abandoned agricultural areas; often planted or protected in garden areas and common to occasional in home gardens and villages, especially in Micronesia. Tree features in Hawaiian, Tahitian, and Tongan mythology; commonly planted in home gardens; planted around houses to dispel evil spirits in Nauru; wood used in light construction, for digging sticks, adze handles, canoe parts, canoe paddles, stilts, and for firewood; poles used as taboo markers on reefs in Namoluk; fruit formerly eaten, especially by older people, but now mostly as an emergency food in Polynesia, but more widely eaten in Micronesia, often with toddy or sugar; fruit cooked and mixed with coconut to make pudding in Nauru; ripe fruit eaten as a stimulant on long sea voyages and used in love and fishing magic in Kiribati; fruit said to be eaten in the Mortlock Islands as a male contraceptive; bark and roots provide red and yellow dyes, respectively; roots mixed with lime to make red hair dye in Tuvalu; one of the Pacific's most important medicinal plants, with the roots, bark, leaves, terminal buds, and fruit used to treat a wide range of maladies; stipules used to treat scorpion-fish puncture wounds in Pohnpei; leaves fed to children as a treatment for vitamin-A deficiency in Kiribati; leaves used in head garlands and as compost in Tuvalu; leaves used to wrap breadfruit seeds for cooking in earthen ovens in Namoluk; juice of fruit mixed with spring water and drunk with kava to counteract unpleasant effects.

62. Moringa oleifera Lam. MORINGACEAE
"horseradish tree," "drumstick tree," "saijan, seijan" (Hind)), "malunggay" (Philippines)

Indigenous to north-western India; a pre-World War II introduction into Fiji and a more recent introduction into most Pacific Islands; seems to thrive on atolls, where seen present in Kiribati. Small tree, up to 10 m high, with pinnately compound leaves; many-flowered, pendulous panicles of fragrant, white flowers; and long, 3-angled pods bearing winged seeds. Common on smallholder Indian sugar-cane farms in Fiji and occasional in rural areas throughout Fiji; common to occasional in urban and rural home gardens in Fiji and occasional in other countries where it has been introduced, either by Indians or Filipino residents or as a vitamin-rich experimental food crop. Planted as an ornamental, a food tree, and for hedges and living fencing; bark and leaves used medicinally by Indians in Fiji, the leaves to treat high blood pressure and diabetes; flowers, leaves, and immature fruit cooked as a vegetable by Indians in Fiji; leaves very high in vitamins A and C, iron, and plant protein.

63. Musa troglodyfarum L. MUSACEAE
"fe'i banana," "mountain plantain"
syns. M. fehi Bert. ex Vieill.; M. semmanni F.v. Muell.

Possibly indigenous to New Guinea or New Caledonia; an aboriginal introduction as far east as the Society Islands and the Marquesas; an early post-European-contact introduction to Hawaii (either conspecific or related to M. maclayi F.v. Muell. of New Guinea and the Solomon Islands). Clump- or stand-forming, giant, perennial herb, 4-5 m tall, with a robust, often purple-tinged, pseudostem with copious, blood-red sap; long, dark green shiny leaves; an erect flower stalk bearing thick, blunt, fleshy, un-seeded or few-seeded fruit that are orange to orange-red when ripe. Found in a naturalized or semi-naturalized state in inland mountain or up-valley forests or old fallow forests; occasional cultivation in gardens, along rivers, and in home gardens. Pseudostem used medicinally; leaves used to wrap food for cooking; dry leaves used as cigarette wrappers; sap used to prepare a purple dye for bark cloth in Samoa; ripe fruit cooked as a supplementary staple or to make desserts or puddings, and occasionally sold at urban produce markets. Two or more cultivars recognized in some areas.

64. Musa* (AAA Group) Simmonds MUSACEAE
"banana," "Robusta," "poyo," "Mons Marie"
syns. M. sapientum L.; M. paradisiaca L. var. sapientum (L.) Kuntze; M. paradisiaca L. ssp. sapientum (L.) Kuntze; M. x paradisiaca spp. sapientum (L.) K. Schum.; M. acuminata Colla cvs

Indigenous to South-East Asia; an early pre-World War II introduction throughout the Pacific Islands. Clump- or stand-forming, giant perennial herb, up to 6 m tall, with large, broad-bladed, broadly feather-shaped, bright green leaves; and a curved, hanging flower stalk bearing large bunches of seedless, blunt-tipped, medium-thick-skinned, greenish yellow fruit that turn bright yellow on ripening. Common banana of commerce found on all high islands, as an intercrop in shifting agricultural areas, in small contiguous stands in garden and fallow areas, and around villages; common in home gardens; often planted along borders or in lines at intervals among other ground crops; planted as a monocultural export crop, often under coconuts or with other trees scattered throughout, or with short-term crops as intercrops; major plantings often in alluvial or colluvial soils. Musa (AAA Group) also includes other cultivars, such as the "Gros Michel" banana, or pisang Ambon (Indonesian), and the "Dwarf Cavendish banana," or "nain," both of which were early introductions and more important in the past, but due to susceptibility to disease are of limited importance today. Important food and export crop in many areas of the Pacific, especially in Tonga, Western Samoa, and the Cook Islands, where bananas are a major export crop; common local cash crop sold at urban produce markets and along roadsides; a major intercrop and staple or supplementary food crop in many areas; pseudostems used medicinally, to wrap or parcel food, and for small dishes or food platters at feasts; leaves used to parcel food and for covering earthen ovens; green fruit cooked as an important staple; ripe fruit eaten raw as a snack food.

*The nomenclature for the genus Musa is confused, with most of the following common seedless cultivars or clones (these do not include M. troglodylarum) being triploid crosses of the fenile species Musa acuminata Colla and M. balbisiana Colla. The Latin binomials M. nana Loureiro, M. sapientum L., and M. paradisiaca L. are commonly used as follows: M. nana for the "dwarf Cavendish," and M. sapientum for the taller bananas, which are generally eaten ripe but which are also cooked throughout the Pacific as starchy staples, and M. paradisiaca for the starchier bananas or plantains, which are usually eaten cooked as a staple starch but occasionally eaten ripe as fruit. The nomenclature most widely used by agronomists is that developed by Simmonds, which classifies all cultivars or clones on the basis of their assumed genetic background, e.g. Musa ABB Group would be a triploid cross of one M. acuminala group and two M. balbisiana groups. Both nomenclature systems are presented here to identify more precisely the clones that are currently of widespread importance in the Pacific Islands.

65. Musa* (AAB Group) Simmonds MUSACEAE
"lady's finger banana," "pisang rajah" (Indonesia)
syns. Musa x paradisiaca L. var. hors. "Pisang raja" (M. acuminata Colla x M. balbisiana Colla)

Possibly indigenous to southern India; a late nineteenth-century introduction into the Pacific Islands, in many cases by missionary societies. Common in Vanuatu, Fiji, most of Polynesia, and parts of Micronesia, including atolls, where it is common in home gardens and occasional as an intercrop in rural garden areas. Clump- or stand-forming, giant perennial herb, up to 7 m tall, with bronze-green pseudostems (trunks) composed of leaf sheaths; broad-bladed, broadly feather-shaped, bright green, spirally arranged leaves; and a curved, hanging inflorescence turning into a tightly-packed bunch of light yellow, short, plump, very thin-skinned, seedless fruit. More resistant to bunchy-top virus and leaf-spot disease than most other Musa cultivars. Ripe fruit are an important supplementary or snack food; immature green fruit cooked as an important supplementary staple food to taro and breadfruit in Samoa.

66. Musa* (AAB Group) MUSACEAE
"Pacific plantain"
syns. Musa x paradisiaca L. ssp. paradisiaca; M. paradisiaca L.; M. sapientum
ssp. paradisiaca Baker

Probably indigenous to tropical Asia; an aboriginal introduction into most of Melanesia and Polynesia; a recent introduction into Kiribati. Clump- or standforming, giant perennial herb, up to 5 m tall, with green, often red or purple-browntinged pseudostems (trunks) composed of leaf sheaths; broad-bladed, broadly feather-shaped, bright to dark green, spirally arranged leaves; and a curved, hanging inflorescence turning into compact bunch of large, cylindrical or slightly 4angled, blunt fruit with thin yellow skin (green when immature) and soft, pinkish yellow flesh. Common to abundant in rural garden areas as an intercrop and in home and urban gardens; common to occasional in banana patches near agricultural areas or villages. A major intercrop in many areas, such as Fiji and Tonga, and a traditional intercrop in yam gardens in Tonga; pseudostems used medicinally, to wrap or parcel food, and for small dishes or food platters at feasts; leaves used to parcel food and for covering earthen ovens; green fruit cooked as one of the most important staples in areas of Melanesia and western Polynesia; ripe fruit cooked or eaten raw in desserts, often with coconut milk; an important cash crop sold at urban produce markets and occasionally exported overseas from Tonga.

67. Musa* (ABB Group) Simmonds MUSACEAE
"cooking banana," "plantain," "bluggoe," "blue Java," "ash plantain"
syns. Musa x paradisiaca L. var. hors. "Bluggoe" (M. acuminata Colla x M. balbisiana Colla)

Indigenous to tropical Asia; an aboriginal introduction into most areas of Melanesia, Polynesia, and Micronesia. Clump- or stand-forming, giant perennial herb, up to 6 m tall, with pale green pseudostems (trunks) composed of leaf sheaths; pale green leaves; and a curved, hanging inflorescence bearing large bunches of light green to blue-gray-green, waxy, thick-skinned, angular fruit with a tapering, blunt tip. Common to occasional in rural garden areas as an intercrop and occasional in home gardens; common in border plantings or along paths and roads on individual agri cultural holdings; a vigorous, easy-to-grow clone that seems to grow well in drought prone areas such as the islands off the west coast of Viti Levu, Fiji, and on some atolls. Important traditional supplementary staple in many areas of the Pacific; pseudostems used medicinally, to wrap or parcel food, and for small dishes or food platters at feasts; leaves are among the most favoured for parcelling food and for covering earthen ovens, green fruit cooked as one of the most important staples in areas of Melanesia and western Polynesia; ripe fruit cooked or eaten raw in desserts, often with coconut milk and a favoured fruit and ingredient in traditional puddings (fekei) in Rotuma; a minor cash crop (fruit, pseudostem, and young leaves) sold at urban produce markets.

Appendix One hundred Pacific Island agroforestry trees (5)

68. Pandanus** tectorius Park. PANDANACEAE
"pandanus," "screw pine"
syns. P. pyriformis Gaud.; P. odoratissimus sensu auct. non L. f.; P. fragrans Gaud. spp. P. odorattssimus L. f. vars.; P. spurius Miq.; P. veitchii Hort.; P. whitmeeanus Mart. (syn. P. corallinus Mart.)

Indigenous and probably an aboriginal introduction to most islands in the case of some cultivars, some of which have possibly originated through selection from wild plants in Melanesia, Polynesia, and Micronesia. Stout, branching tree, up to 5 m or more tall, with numerous aerial roots and thick, forking stems; spirally-arranged, pointed leaves with armed or spiny margins and midribs; a fragrant, pendant, male inflorescence, with cream-yellow bracts and white spikes; a similar but smaller female inflorescence on separate trees; and pineapple-like ovoid fruit bearing many yellow to red-orange, wedge-shaped, fleshy drupes. Common in coastal vegetation bordering garden areas in almost all island groups, occasionally planted or protected in garden areas and home gardens, and planted in discrete groves on Nauru, on most atolls, and in many areas of Micronesia. One of the Pacific's most useful plants; features prominently in Melanesian, Polynesian, and Micronesian creation mythology, cosmogony, proverbs, riddles, songs, chants, and sayings, and a symbol of love and a nature spirit in Hawaii; many famous pandanus groves recognized in Hawaii, with the Kahala area of Honolulu, formerly known for its groves, named after the tree; people of Kiribati referred to as the "Pandanus People"; commonly planted in home gardens and in monocultural and mixed stands in garden areas; trunk and

*The nomenclature for the genus Pandanus is, like that for Musa, confused, with some taxonomists classifying many of the common cultivars and wild clones or species, both edible and non-edible, as forms or varieties of P. tectorius. Other taxonomists consider them to be distinct species, often listing numerous species or varieties for a given area. For example, P. odoratissimus L. f. has long been thought to be synonymous with P. tectorius, but is considered by many authorities not to occur east of Malaysia. Similarly, P. odoratissimus L. f. var. pyriformis Mart. has been used as a synonym for a wild and doubtful variety of P. tectorius, whereas Stone (1970) considers P. fragrans Gaud. to be the common wild species on Guam, and does not consider P. tectorius to be present. Thus, most named cultivars are commonly grouped under P. tectorius. Other widespread forms, such as P. dubius Spreng., a widespread edible species; and P. spurius Miq. cv. "PUTAT" (syns. P. tectorius Warb. var. laevis Warb. and P. odoratissimus L. f. var. laevis Warb.) Mart., which are widely cultivated for their leaves for use in plaited ware, are also present on many islands in the Pacific. prop or aerial roots used in house construction and for ladders, digging sticks, headrests, rat traps, containers, canes, musical bows, and for fuel wood; roots used to make the ukeke musical instrument in Hawaii; chewed pieces of prop root used as popgun ammunition in Tuvalu and dried to make fuses or tapers used in medical treatment in Tuvalu; green wood used in smokeless fires to wilt pandanus for mat making; dead wood used to smoke skirts in Tuvalu; treated leaves of selected varieties used to make mats, baskets, hats, fans, bracelets, pillows, canoe sails, toy boats, weather screens, balls, toys, and other plaited ware, and for cordage; leaves used for compost, bandages, swabs, corks, cigarette wrappers, whistles and ornaments, and for caulking; most parts used medicinally; male flower used to scent coconut oil, to perfume tape cloth, in garlands, as a love charm and aphrodisiac, and worn in ear slits in Tuvalu, and to make fine mats by Hawaiians in the past; the fleshy drupes (keys) of fruit of many varieties or cultivars eaten ripe as a snack food or cooked and/or dried and processed in a variety of ways to make coarse starch or flour, desiccated cakes, and other staple substitutes in Micronesia and atoll areas, but considered an emergency food in most other areas of the Pacific; aerial root tips eaten on some atolls; stalk or receptacle upon which keys are attached fed to pigs in Tokelau; yellow to red immature drupes strung in leis or garlands; fibrous, chewed or dried, mature drupes (or after being chewed by hermit crabs) used as paint brushes for painting tape, for fuel, and as fishing-line floats or markers; stilt roots used to make fish-net floats, red dye, and fibre from stilt roots for ceremonial skirts in Kiribati, jump ropes in Tokelau, and for stringing leis and straining kava in Hawaii; numerous cultivars or distinct species of pandanus exist, many of which are shrub-like cultivars, and not P. tectorius, although some authorities believe that most of the tree-like and edible cultivars could be variants of P. tectorius.

Cultivars used almost exclusively for fibre (e.g. P. odoratissimus, P. spurius, P. veitchii, and P. whitmeeanus) are highly variable, commonly trunkless, and unbranched shrubs to small, branched, and stilt-rooted trees, with variable, thornless or thorny, green or white and greenstriped leaves; and rarely bearing fruit, although some cultivars bear male flowers and/or fruit when reaching maturity. These are widely cultivated throughout Melanesia and Polynesia in both rural and home gardens; commonly in monocultural patches or as border plantings; occasionally naturalized in swampy and grassy areas and on forest margins, possibly as remnants of cultivation. Leaves treated by boiling or soaking, drying, and dyeing, and used to make mats, baskets, hats, fans, bracelets, pillows, canoe sails, toy boats, weather screens, rain capes, balls, toys, and other plaited ware, and for cordage; leaves also used for thatching in some areas; trunks of some cultivars used in light construction for flooring, etc.

69. Pandanus** spp. PANDANACEAE
"edible pandanus"
spp. P. brosimos Merr. and Perry; P. conoideus Lam.; P. dubius Spreng.; P. julianettii Mart. (excluding P. tectorius, described above, of which there are also edible varieties and cultivars)

Different species indigenous to South-East Asia, New Guinea, and the Pacific Islands; P. brosimos, P. conoideus, and P. julianettii all possibly native to, and semidomesticated in, highland New Guinea; P. dubius is native from the sea coasts of lndonesia to the Marianas and Caroline Islands in Micronesia, and probably an aboriginal introduction into some islands such as Rotuma and the south-eastern Solomon Islands. Highly variable trees, bearing edible fruits.

P. conoideus, of which there are many cultivars, is one of the most important cultivated trees in orchards and around houses in highland New Guinea, with large, conical syncarps weighing up to 8 kg bearing numerous, bright red (rarely yellow) drupes; fruits cored and steamed in earthen oven, boiled or roasted and eaten by directly sucking off the edible mesocarp or in the form of a pleasant-tasting, oily, vitamin-A-rich, ketchup-like sauce that is eaten with greens, pumpkin, or bananas; fruits sold at roadside and urban markets; oil from drupes also used as hair and body oil, polish for arrow shafts, and for paints and dyes; leaves sometimes used for thatching.

P. brosimos and P. julianettii, of which there are many cultivars, have large, ovoid syncarps composed of hundreds of finger-sized, nut-like drupes; found both wild and cultivated at high elevations in New Guinea; oily, white kernels or drupes eaten cooked or raw and sold at roadside and urban markets, as a major snack food and seasonal staple food; fruit commonly smoked and stored in house rafters; leaves used to make bush shelters.

P. dubius is found wild and occasionally cultivated in the Marianas and Caroline [slands, Rotuma, and the south-eastern Solomon Islands; drupes of syncarp eaten ripe as a snack food and cooked and stored as a staple food; leaves and other parts used much the same as P. tectorius.

70. Persea americana Mill. LAURACEAE
"avocado," "avocado pear," "alligator pear"
syns. Laurus persea L.; Persea gratissima Gaertn. f.

Indigenous to Mexico and a post-European-contact introduction into the Pacific islands. Medium to large, evergreen tree, up to 12 m or taller, with papery to leathery leaves; small, greenish to yellowish white or yellowish brown flowers; and subglobose to pear-shaped, fleshy fruit with light green to purplish skin and light green to yellow-green, butter-like, edible flesh, and containing a single subglobose seed with a brown seed coat. Occasionally planted and protected in rural garden areas and home gardens. Ripe fruit eaten raw, often as a butter substitute; commonly sold at local produce markets and stores, and a minor export crop in some countries such as Tonga.

71. Pemphis acidula Forst. LYTHRACEAE
"pemphis"

Low, sprawling shrub, up to 4 m high, with small, fleshy leaves; small, solitary, white flowers; and small, turbinate fruit. Common on coastal limestone rocks, cliffs, and on limestone bedrock outcrops on atolls; common on inner margins of mangroves and in contiguous stands on the coastal margins of copra plantations and agricultural areas in Kiribati and other atolls. Ancestral tree of the people of Kabara and Wagava, Fiji, who believe they originated as its fruit; referred to by the title of "Vu" (forefather) on Kabara, where only one tree remained in the 1930s; important in protecting inland areas from sea spray; extremely hard wood favoured for carved objects such as house frames, canoe parts, keels, connecting pegs, and paddles, digging sticks, clam knives, tool handles, thatching needles, pipes, back scratchers, fish hooks, fish-net frames, fishing poles, shuttles and meshing needles, fish, eel, and rat traps, spears, fish clubs, war clubs, darts, food containers, mortars and pestles. pounders, pump drill pieces, coconut huskers, combs, drums, tops, throwing sticks and other toys, etc.; a preferred fuel wood with a very hot flame; wood used for spear of wave magician and as staff for magic dances in Ifaluk; old wood used to smoke skirts in Tuvalu; rotting wood mixed with coconut oil as a cosmetic; bark, leaves, and flowers used medicinally in Tahiti and Micronesia; bark and leaves mixed with toddy as baby food in Ifaluk; fruits sometimes eaten in Kiribati; used as pig feed in Tokelau; scraped bark yields a red dye in Tokelau.

72. Pinus caribaea Morel. PINACEAE
"Caribbean pine"

Indigenous to the West Indies and along the Caribbean seaboard of Central America from Belize to Nicaragua; a recent introduction to the Pacific Islands; intensive selection has taken place in Fiji, from where it has been introduced to other islands. Medium to large tree, up to 20 m or higher, with wide-spreading branches when mature; dark green foliage with needles (leaves) in bundles of three; and cones with small, apical prickles. Found in extensive monocultural plantations in Fiji, Vanuatu, New Caledonia, and Tonga, in village woodlots and as individual plantings in rural and urban areas in a number of countries. Major species used in reforestation of degraded gassland areas in Fiji and southern Vanuatu; sometimes undergrazed with cattle and other livestock. Timber used for construction, fence posts, and fuel wood; major commercial species in Fiji for the export production of wood chips destined for paper production in Japan and timber for export and local sale. Trees have shown susceptibility to tropical cyclone damage in Fiji.

73. Piper methysticum Forst. f. PIPERACEAE
"kava," "kava root"
syn. Macropiper methysticum Miq.

Now considered a likely domesticate from Vanuatu; once assumed to be an aboriginal introduction from farther west. Shrub, up to 2-3 m tall, with a thick, woody rhizome; stems with prominent, swollen nodes; large, round, heart-shaped, palmatelyveined leaves; and solitary greenish white flower spikes. Common in rural gardens as an intercrop and occasionally in monocultural plots in some areas in Papua New Guinea, Vanuatu, Fiji, Tonga, Samoa, and in Pohnpei in Micronesia, with residual plantings in other areas of Polynesia; occasional in home gardens; infrequently naturalized along trails. A major social and ceremonial beverage of considerable cultural importance and an important exchange item in Vanuatu, Fiji, Tonga, Samoa, and Pohnpei, where it is a major cash crop for local sale and limited export; large roots and lower stems crushed or pounded and mixed with water to produce "kava," an alkaloid stimulant that has a mild narcotic, sedative, or soporific effect and that is drunk both ceremonially and as a social beverage; roots, stems, and leaves used medicinally to treat convulsions, stiffness, toothaches, sore throats, respiratory disease, filariasis, intestinal parasites, and venereal disease; kava exported to Germany as an ingredient in medicines to treat high blood pressure.

74. Pipturus argenteus (Forst. f.) Wedd. URTICACEAE
syns. Urtica argentea Forst. f.; P. incanus Wedd.
sp. P. albidus (H. & A.) A. Gray

Indigenous from Malesia through Melanesia to the Marquesas in eastern Polynesia and to the Marshall Islands and Kiribati in eastern Micronesia; a similar endemic species (P. albidus) exists in Hawaii. Shrub to small tree, up to 8 m tall, with rather dense, white, pubescent young growth; coarsely serrate, oval-ovate, 3-nerved leaves that are green above and white-woolly beneath; greenish white flowers in head-like clusters; and fleshy fruit. Frequent on rocky cliffs near the sea, in coastal thickets, and in dry or open forest; common to occasional in garden and fallow areas; an important pioneer species in abandoned gardens; occasional in home gardens. Features in legends of the Polynesian god Maui capturing the sun in Tahiti; important in magic, sorcery, and ritual in Melanesia; wood used in house construction, for fishhooks, rollers for hauling, and firewood and for making tape beaters in Hawaii; strong cordage obtained from best fibre used for fishing lines and nets throughout the Pacific, and for making ceremonial mats in Samoa and for tying the navel of newborn babies in Tokelau in the past; made into tape cloth in Tahiti and Hawaii; bark cloth paste made from bark sap; bark, roots, and leaves used medicinally; leaves used in ceremonial dress and to parcel food and line earthen ovens in Melanesia and as imitation feathers on fishing lures on Puluwat; flowers used to scent coconut oil in Lau, Fiji; seeds eaten by pregnant women and newborn babies in Hawaii and by children in Tokelau; young leaves eaten after cooking in toddy and coconut milk on Ifaluk; bark fed to pigs on Namoluk; a number of different red and greenleaved varieties or cultivars recognized in Vanuatu.

75. Pisonia grandis R. Br. NYCTAGINACEAE
"pisonia," "lettuce tree"
syn. P. alba Span.

Indigenous from Madagascar, tropical Australia, and Taiwan to the Tuamotus in eastern Polynesia and the Marshall Islands and Kiribati (including the Phoenix and Line Islands) in Micronesia. Medium to large, soft-wooded tree, up to 20 m or higher, with thin, light green leaves; small, fragrant, white to greenish yellow flowers in cymose clusters; and glandular fruit. Common in solid groves or thickets on small, uninhabited islets and in original inland and coastal forest on most atolls, small limestone islands, and coastal lowland areas of some big islands; occasionally cultivated in home gardens and agricultural areas on some islands. The favourite nesting or rookery tree for sea birds, including the black noddy, a ceremonial delicacy in Nauru; reportedly a protected sacred grove on Onotoa in Kiribati; leaves considered godlike on Tongareva; occasionally planted as a living bathhouse to provide shade and privacy and as a living pig pen in Tonga; soft timber occasionally used for light construction, fence posts, outhouse flooring, canoes, canoe outriggers, floats and bailers; occasionally used for firewood and to make fire by friction; leaves edible and used to wrap food for cooking and eaten with fish in Vanuatu and with taro on Kapingamarangi atoll; leaves a common pig feed in Polynesia and Micronesia; bark and leaves used medicinally in New Caledonia, Polynesia, and Micronesia; planted recently in Kiribati in home gardens and at the hospital for edible vitamin-rich leaves; leaves used as mulching and green manure in Micronesia and Tokelau; a sterile cultivar with edible leaves, P. alba, is the lettuce tree of Indonesia.

76. Plumeria rubra L. APOCYNACEAE
"frangipani," "plumeria," "temple tree," "graveyard tree"
syns. P. acuminata Ait. f.; P. acutifolia Poir.
spp. P. obtusa Lour.

Indigenous to tropical America from Mexico to Panama; a pre-World War I introduction throughout the Pacific. Small to medium-sized, soft-wooded, deciduous tree, up to 5 m or taller, with thick, fragile branch tips; white, milky sap; gray-green, acuminate leaves that are spirally clustered near the ends of branches; and attractive, fragrant, waxy-white to yellow, pink, deep red, or multicoloured flowers. Common in home gardens, especially in urban areas, throughout the Pacific; occasional in rural agricultural areas; common street or roadside tree; commonly planted in cemeteries. Planted ornamental; easily planted from cuttings; flowers used in garlands and dried in the sun and used to scent coconut oil; leis and garlands of plumeria sold locally to tourists and to residents; leaves used medicinally in Nauru. A more recent introduction, P. obtusa Lour., an evergreen frangipani with obovate leaves and large white flowers, is increasingly common throughout the Pacific.

77. Polyscias spp. ARALIACEAE
"panax," "hedge panax"
spp. P. balfouriana (Andre) Bailey; P. filicifolia (Moore) Bailey; P. fruticosa (L.) Harms (syn. Nothopanax fruticosus [L.] Miq.); P. guilfoylei (Cogn. and March.) Bailey (syn. N. guilfoylei [Cogn. and March.1 Merr.); P. scutellaria (Burm. f.) Fosb. (syn. N. scutellaria [Burm. f.] Merr.); Polyscias tricochleata (Miq.) Fosb. (syns. N. tricochleatus Miq.; Polyscias pinnata Fosb. cv. tricochleata Stone)

Depending on the species, indigenous to tropical Asia or Melanesia or possibly aboriginal introductions into the islands of western Melanesia and Polynesia; recent introductions into other areas. Large, erect shrubs, up to 3 m or taller, with few, rapidly ascending branches; and normally compound pinnate, ovoid to orbicular to deeply toothed or laciniate, green, bright yellow to variously variegated leaves, depending on the variety. Common in home gardens and occasional in rural garden areas in Melanesia, Polynesia, and Micronesia. Planted ornamentals in urban areas; important living fencing or hedges throughout the Pacific; commonly planted as boundary markers; very important on atolls to provide protection against sea spray; leaves of some species cooked as a green vegetable, commonly eaten with pig, turtle, fin fish, and shellfish, in Papua New Guinea, the Solomon Islands, and Vanuatu; one of the most important vegetables in the south-eastern Solomon Islands, where it is eaten regularly, almost always included in earthen oven meals, and fed to lactating women to improve milk production; the consumption of the leaves encouraged to combat vitamin-A deficiency in Kiribati; cup-like leaves of P. scutellaria used as food platters in the Marshall Islands; leaves used as body ornamentation and for dancing costumes; leaves of some species used medicinally; leaves of P. scutellaria used as a stimulant in Papua New Guinea; leaves fed to livestock; stems used as fuel in areas of fuel-wood scarcity.

78. Pometia pinnata Forst. SAPINDACEAE
"Oceanic Iychee," "island Iychee"

Indigenous from the Celebes in Indonesia and the Philippines to as far east as Fiii, Tonga, Samoa, and Niue; probably an aboriginal introduction as far east as the Cook and Society Islands, the Tuamotus, and Hawaii; a recent introduction into some islands; not reported present in Micronesia as an indigenous or aboriginally introduced species. Medium to large, slightly buttressed tree, up to 20 m high, with large, pin nate leaves with prominent veins that are pink to bright red when young; small, white flowers in terminal panicles; and spherical, green to dull red, loose, thinskinned fruit with a gelatinous, sweet, white, translucent pulp surrounding a large seed. Abundant in lowland forests and one of the most common trees in lowland, secondary, and mature fallow forests, open woods, and garden areas; commonly protected when clearing new garden areas; commonly cultivated or protected in active garden areas and home gardens. Important seasonal fruit-tree, of which a number of cultivars or varieties exist in Melanesia and Polynesia; timber used in house construction, tool making, for canoe hulls, and as a high quality firewood; canoe putty extracted from the inner bark; bark used medicinally for a wide range of maladies throughout the Pacific; bark used to produce shampoo in Samoa; leaves used for mulching or fertilizing yam gardens in the Sepik areas of New Guinea; ripe fruit eaten raw and an important seasonal cash crop for local sale; seeds roasted and eaten in parts of the Solomon Islands; fruits a favoured food of fruit bats.

79. Premna serratifolia L. VERBENACEAE
syns. P. obtusifolia R. Br.; P. taitensis Schauer

Indigenous from east Africa, tropical Asia, and Australia to the Tuamotus in Polynesia, and Kiribati and the Marshall Islands in Micronesia. A shrub or small tree, up to 12 m high, with shiny, green, ovate leaves; small, greenish white flowers in terminal panicles; and small, globose fruit that turns black at maturity. Common tree in coastal vegetation and in open or secondary forest, fallow areas, and thickets; common to occasional in garden areas and home gardens. Commonly planted in Fiji as living fencing; emblem of the god Avaro in Tahiti; a symbol of love, affection, beauty, goodness, pleasure, and virtue in Ulithi; wood used in general construction, for canoe connectives in Ulithi, canoe nails in New Guinea, and to make specialized, large fish-hooks in Kiribati; used as firewood and for making fire by friction in Micronesia; best firewood to cook pandanus in earthen ovens in Nauru; straight saplings or branches used as fishing rods; leaves and roots used to perfume coconut oil in Tuvalu, Kiribati, and Nauru; bark, leaves, and fruit used medicinally for a wide range of maladies throughout Melanesia, Polynesia, and Micronesia; leaves used in ceremonial dress in New Guinea and worn in ear slits in Tuvalu; leaves used to arouse love and a mixture of the bark, coconut milk, and Sida fallax flowers used to banish fear in marriage and to promote true love in Kiribati; important in fishing, canoe making, and house-building magic in Ulithi; flowers used in coconut oil for hair in Samoa; flowers used in garlands in Nauru and Puluwat; seeds eaten by children in Tuvalu; ripe fruit formerly eaten with yams, and a food of pigeons and fruit bats in Vanuatu; root provides dye in Tuvalu.

80. Psidium gunjava L. MYRTACEAE
"guava"

Indigenous to tropical America; an early post-European or nineteenth-century introduction into most of the Pacific Islands. Shrub or small tree, up to 8 m high, with smooth, light reddish brown bark; rather thick and slightly brittle, prominentlyveined, gray-green, oval-elliptic leaves; solitary, white flowers with white stamens; and small, globose to slightly pear-shaped, light-green to light-yellow fruit with numerous, small, woody, yellowish brown seeds embedded in fragrant, pink to whitish yellow, edible flesh. Common to abundant in scrub and open secondary forest, fallow areas, thickets, swampy or marshy areas, grasslands, and grazing areas; occasionally protected in rural garden areas and home gardens; a major pioneer species in recently abandoned gardens or plantations; widely naturalized in most countries; seems to grow well on some atolls; a declared noxious, difficult to eradicate, weed in some countries. Timber occasionally used in light construction, for fencing, and for fishing poles in Nauru; very highly regarded as firewood; leaves used to treat diarrhoea throughout the Pacific; fruit widely gathered, mostly from wild trees, and eaten green and ripe and sold at roadside and urban markets, fruit used to make jams and jellies; a major commercial crop in Fiji and Hawaii, where wild fruit are used to produce puree and juice for local sale and export.

81. Pterocarpus indicus Willd. FABACEAE/LEGUMINOSAE
"New Guinea rosewood," "bluewater," "sang dragon," "padouk"

Indigenous from South-East Asia to the Ryukyu Islands, the Caroline Islands in Micronesia, and to Vanuatu and Fiji in Melanesia; possibly an aboriginal introduction into some areas. Medium to large, buttressed tree, up to 40 m tall, with copious, red sap; compound leaves with 5-15 widely-spaced, pale green leaflets; axillary racemes of fragrant, yellow or yellow-orange flowers; and discshaped fruits with a flat, membranaceous wing. Occasional to common in lowland forest, mixed swamp forest7 usually near the sea or along rivers, and in agricultural areas; occasional in home gardens. Fast-growing tree that coppices easily; planted or protected in garden areas for its nitrogen-fixing/nutrientrecycling ability; important living fence plant and for living pig pens; occasionally planted as a boundary marker, shade tree, or ornamental tree; timber used in house construction, canoe building, and for tools, fence posts, paddles, handicrafts, wood carving, yam stakes, and fuel wood: timber exported from the Solomon Islands; bark used medicinally to treat dysentery, anaemia, tuberculosis, headaches, sores, and as a purgative.

82. Rhizophora spp. RHIZOPHORACEAE
"mangrove"
spp. Rhizophora apiculata Bl.; Rhizophora samaensis (Hochr.) Salvoza (syn. R. mangle L. var. samoensis Hochr.); Rhizophora mucronata Lam. ("red mangrove"); Rhizophora stylosa Griff. (syn. R. mucronata var stylosa [Griff.] Schimper)

Indigenous from the Indian Ocean to Samoa, Tonga, and Tuvalu in western Polynesia and to the Gilbert Islands in Micronesia; recent introduction into French Polynesia, Hawaii, and other areas. Small trees with prop roots, up to 4 m tall, with thick, leathery, glossy, dark green leaves; dull white, cymose flowers; and cylindrical, dark brown, pre-germinated fruit. Common to abundant on tidal flats, along stream mouths and estuaries, on the seaward side of mangrove swamps, and less often along beaches, landlocked lagoons, and lakes. Important for providing protection from coastal erosion, salt-water incursion, and sea spray to inland agricultural areas; commonly partially reclaimed for agriculture (e.g. sugar cane and rice farming) and mariculture. Wood and stilt roots used in general construction, for needles and awls, earrings, combs, rat traps, headrests, coconut huskers, canoe parts, digging sticks, throwing sticks for games, for threading coconut shells for shark rattles, and scoop-net frames, and for fish-trap stakes in Kiribati because of its resistance to salt water and shipworm; an excellent firewood and also used to make charcoal in Fiji; supple stilt roots once used for bows and arrows in Polynesia and Melanesia and for house walling on Kosrae; bark an excellent source of tannin and yields blackbrown dye for bark cloth in Fiji and Tonga and for fishing line and nets in Samoa; red dye obtained from roots in Kiribati; bark used in red hair dye and to paint or glaze pottery in Fiji and Polynesia; bark scraped to scent coconut oil in Kiribati; leaves an important pig feed in Tuvalu; bark, leaves, and fruit used medicinally in Melanesia, Polynesia, and Micronesia; fruit a supplementary food and used in body ornamentation in Tuvalu; caulking paste made in past from boiled fruits; hollow fruit used as whistle in Samoa.

83. Rhus taitensis Guill. ANACARDIACEAE
"island sumac"
syn. R. simarubifolia A. Gray

Indigenous from South-East Asia and the Philippines to the Marianas and the Caroline Islands in Micronesia, and the Society Islands in eastern Polynesia. Medium to large tree, up to 20 m tall, with the younger parts more or less pubescent; large, pinnate leaves with oblong, bluntly pointed leaflets; small, white flowers in terminal, compound clusters; and fleshy, globose, black fruits. Occasional to common in lowland forest, most commonly in open forest and secondary, disturbed forest; common in garden areas and sometimes protected when clearing for new gardens. Important pioneer tree species in abandoned garden areas; timber used for general construction, wood carving, banana cases, and a favoured firewood; leaves used medicinally; leaves a source of black dye for the hair in Fiji and for staining the teeth jet-black in the Solomon Islands; fruits a favourite food of doves and pigeons.

Appendix One hundred Pacific Island agroforestry trees (6)

84. Saccharum edule Haask. POACEAE/GRAMINAE
"edible sugar cane inflorescence," "Fiji asparagus," "pitpit" (PNG Pidgin)

Indigenous to Malesia and possibly New Guinea; probably an aboriginal introduction to the Solomon Islands, Vanuatu, and Fiji. Large, perennial, sugar-canelike grass, up to 2.5-4 m high, with erect, yellow or green to reddish stems; nodes marked but obscured by leaf sheaths; leaf blades tapering to a sharp point; a soft, fibrous, white to light yellow, cigar-shaped inflorescence enclosed in non-opening, leaf sheaths. Common in garden areas and widely naturalized, often in extensive stands; common in poorly drained valley bottoms, alluvial plains, or low-lying sites. Commonly cultivated or wild supplementary food plant in Melanesia; unopened (aborted) inflorescence eaten roasted over an open fire or cooked in an earthen oven in the leaf sheathes, cooked in coconut milk, and made into curries by Indians in Fiji; inflorescence a major seasonal cash crop; cooked inflorescence canned commercially in Fiji.

85. Saccharum officinarum L. POACEAE/GRAMINAE
"sugar cane," "noble sugar cane"

Indigenous to Malesia or New Guinea; an aboriginal introduction throughout island Melanesia and high-island Polynesia and Melanesia to as far east as Hawaii and Easter Island; a recent introduction to Nauru and some atolls. Large, perennial grass, up to 4 m tall, with erect, unbranched, yellow or green to dark red or almost black, sometimes striped, stems; nodes close together at base of stems and more widely spaced distally; leaves, long, broad, overlapping at base; and a dense, long, erect or drooping, branched panicle bearing silvery white to pinkish flowers. Common in rural garden areas, often in contiguous stands, as an intercrop or border planting; common in home gardens in both rural and urban areas; uncommonly naturalized; present, but not wellestablished, on atolls. Plant of important ceremonial or legendary importance in many areas of the Pacific, such as in Tonga, where it is presented along with kava in ritual offerings, or Hawaii, where its flowering signals the onset of the octopus season; common supplementary food plant throughout the Pacific, for which numerous named cultivars exist in most localities; juicy stems chewed as a source of refreshment and energy; stems chewed for medicinal purposes in New Britain; chewing the fibrous cellulose is believed to keep teeth and gums in good condition; a major commercial export crop or for industrial processing into sugar for local consumption in Fiji, Papua New Guinea, and Hawaii, where interspecific hybrid cultivars (often crosses between the noble caneslS. officinarum and woodier, more disease-resistant, wild canes such as S. spontaneum L. and S. robustum Brandes and Jeswiet ex Grassl) are cultivated for this purpose; formerly cultivated commercially by the Japanese in the Marianas Islands; by-products of commercial sugar manufacture include molasses, which is fed to livestock, alcohol, which is made into rum, gin, vodka, and whiskey for export and local sale in Fiji, and fertilizer (mill mud); leaves of some varieties are used as durable thatch for roofing; leaves provide fodder for cattle in commercial sugar-cane farming areas of Fiji; fibre from flower stalk used in braided hats in Hawaii.

86. Samanea saman (Jacq.) Merr. MIMOSACEAE/LEGUMINOSAE
"rain tree," "monkey-pod tree"
syns. Albizia saman (Jacq.) F. v. Muell.; Mimosa saman Jacq.; Enterolobium saman (Jacq.) Prain ex King; Pithecellobium saman (Jacq.) Benth.; Inga saman (Jacq.) Willd.

Indigenous to tropical America; a post-European-contact introduction to the Pacific Islands. Large to massive tree, 7-25 m high, the trunk up to 1 m in diameter, the crown rounded, usually broader than tall, with bipinnate leaves; pink flowers with greenish or yellowish lobes; and straight or slightly curved pods containing sweet, sticky, brown pulp and brown seeds. Common in both rural and urban areas; commonly naturalized and invasive in some areas, such as Vanuatu, where it is a major weed in abandoned or poorly maintained plantations, and in Fiji, where it is among the most common trees on smallholder sugar-cane farms, particularly along stream banks; not found on atolls. Important ornamental, shade, or roadside tree; timber used in general construction and for firewood; a favoured species for wood carving and furniture such as small table tops; a nitrogen-fixing species of some value for Breen manure; pods occasionally fed to livestock.

87. Scaevola sericea Vahl. GOODENIACEAE
"scaevola," "saltbush, " "half-flower"
syns. S. taccada (Gaertn.) Roxb.; S. sericea var. raccada Makino; S. frutescens sensu auct. non (Mill.) Krause; S. frutescens var. sericea (Forst. f.) Merr. (nom. nud.); S. koenigii Vahl; S. lobelia Murr.; Lobelia taccada Gaertn.; L. koenigii (Vahl) Wight

Indigenous from East Africa through tropical Asia, southern Japan, northern Australia, and Malesia to eastern Polynesia, Micronesia, and Hawaii. Erect, freely branching, spreading, somewhat succulent, soft-wooded, pithy-stemmed shrub, up to about 2 m high, with slightly fleshy, light bright green leaves crowded near the ends of the branches; white or pale green flowers that appear to be split in two with only half the petals remaining; and grape-like bunches of fleshy, white, subglobose fruit. Very abundant in coastal strand vegetation, often bordering inland garden areas, throughout the Pacific; common in understorey vegetation in coconut plantations on atolls; dominant species and one of first colonizers on strip-mined areas of Nauru. Protected in garden areas on atolls and occasionally planted in home gardens; associated with phases of the moon in Kiribati; features in legends and chants in Hawaii; wood sometimes used for roofing strips, rafters, supports, and house decking, rafts, canoe paddles and poles, scoop-net handles, eel traps, reef markers, net gauges, shark rattles, throwing sticks, and toy darts; hollow branches used as popguns or blowguns in games in Tuvalu, Tokelau, Kiribati, and Nauru; pith of large trees cut into strips and made into paper-like garlands and headbands in Kiribati and Nauru, and to caulk canoes in Tuvalu; pith chewed as gum in Kiribati; leaves made into tuna lures in Tokelau; leaves boiled with women's grass skirts in Kiribati to dye them brown and make them durable; bark, white heartwood, roots, leaves, fruit, and seeds used medicinally; leaves used as a contraceptive in New Guinea; leaves used for wrapping penis in Vanuatu circumcision ceremony; leaves used to wrap food and to cover earthen ovens; leaves used as pig feed in Tokelau; fruits used in magic in Kiribati; fruit used in fishing magic in Ulithi; leaves used for cleansing diving goggles in Polynesia, for shelter in fish traps in Kiribati, to scent coconut oil in Micronesia, in head garlands, and worn in ear slits in Tuvalu, and occasionally for compost or fertilizer; flowers used in garlands in Tuvalu, Kiribati, and Nauru; fruit eaten by pigeons or fed to them in Tokelau.

88. Schizostachyam spp. POACEAE/GRAMINAE
"bamboo," "aboriginal bamboo," "native bamboo"
spp. S. glaucifolium (Rupr.) Munro; S. lima (Blanco) Merr. (syn. Bambusa lima Blanco); S. stenocladum A. Camus; S. tessellatum A. Camus

Indigenous species from tropical Asia through Melanesia (S. glaucifolium, S. stenocladum, and S. tessellatum) to Palau and Yap in the Caroline Islands (S. lima); S. glaucifolium probably an aboriginal introduction from Fiji into Polynesia as far east as the Marquesas and Hawaii; possibly a naturalized aboriginal introduction in parts of Melanesia. Tall, clump-forming bamboos, up to 15 m tall, with hollow, thinwalled stems with branches above the upper nodes; linear-oblong leaves; and inconspicuous flowers. Common to occasional in secondary vegetation, often dominant in intermediate climatic zones of Fiji; common in hillside thickets and along streams and rivers; occasional in garden areas, often in large stands; widely naturalized throughout the Pacific. Stems used in construction, for house walling, rafters, battens for affixing thatch, shelving and scaffolding, fencing and poultry pens, fishing rods, yam stakes or trellising, makeshift spears, cooking and water containers, nose flutes, and pillows; leaves used to cover cooking containers in the Solomon Islands. Widely being replaced in importance by the larger, recently introduced Bambusa vulgaris.

89. Securinega flexuosa Muell.-Arg. EUPHORBIACEAE
syns. S. samoana Croizat.; S. virosa ?; Flueggea flexuosa Muell.Arg.

Indigenous from the Moluccas and Philippines eastward to the Solomon Islands; also present, but probably an aboriginal introduction, in Vanuatu, Fiji, Rotuma, Uvea (Walks Island), Tonga, and Samoa; a recent introduction into Rarotonga in the

Cook Islands. Shrub or small tree, up to 15 m high, with rough bark; ovate leaves; small, petal-less flowers in axillary clusters; and globose berries with many small seeds. Common tree in forests, usually near villages, and in agricultural lands, and occasionally planted near villages and in home gardens, particularly in Samoa; important, apparently naturalized, pioneer species in abandoned gardens. Planted along boundaries; dark, heavy wood favoured for house posts, general construction, and fencing; occasionally used for firewood; shredded root used medicinally in New Guinea; fruit used as a source of dye on Uvea.

90. Spondias dulcis Park. ANACARDIACEAE
"Polynesian vi-apple," "Polynesian plum," "Otaheiti apple," "golden apple"
syn. S. cytherea Sonn.

Probably indigenous to tropical Asia; an aboriginal introduction to Melanesia, Polynesia as far east as the Marquesas, and to the Caroline Islands and Nauru in Micronesia; a post-European-contact introduction to Hawaii. Medium to large, stiffbranched, smooth, gray-barked, deciduous tree, up to 15 m or taller, with pinnately compound leaves; numerous, small, whitish flowers in particulate clusters; and ovalobovate, edible fruit with green to yellow-orange skin and light green to dark yellow pulp. Common tree in rural garden areas, fallow forests, and home gardens; sometimes growing in an almost wild or naturalized state in mature fallow forests; almost always preserved when clearing fallow vegetation for new gardens. Soft, light wood used for interior purposes in house building and formerly made into canoe hulls; immature and ripe fruit eaten throughout the Pacific, occasionally cooked in the Solomon Islands, and a minor seasonal cash crop for local sale; immature fruit grated and mixed with coconut as a refreshment or dessert and made into spiced pickles (achar) by Indians in Fiji; juice of fruit an important food for pregnant women in the Solomon Islands; young leaves eaten in some areas, often with pork or to flavour meat; bark, leaves, and fruit used medicinally.

91. Swierenia macrophylla King MELIACEAE
"West Indian mahogany," "large-leaved mahogany," "Honduras mahogany"

Indigenous to eastern Central and South America from Mexico southward to Peru, Bolivia, and Brazil; a recent introduction into the Pacific Islands. Large tree, up to 34 m high, with hard, heavy, reddish wood; pinnately compound leaves; panicles of numerous, small, white flowers; and ovoid, woody fruit. Major exotic species used in reforestation programmes in the wet zone of Fiji; occasionally planted as a timber tree in Niue, Tonga, Vanuatu, the Solomon Islands, and elsewhere; occasionally planted in agricultural areas, in and around villages, on sugar-cane farms, in home gardens, and as a street tree. Timber used for high quality construction, furniture, veneer, mouldings, panelling, general joinery work, boat planking and decking; other uses, depending on the grade, include carving, light handles, drawing boards, light construction and cases; waste wood used as firewood.

92. Syzygium cumini (L.) Skeels MYRTACEAE
"jambolan," "Java plum"
syns. Myrtus cumini L.; Eugenia cumini (L.) Druce; E. jambolana Lam.

Indigenous or naturalized from India and Sri Lanka to Malesia; a recent introduction into the Pacific Islands. Medium-sized tree, up to 15 m high, with stiff, leathery, oblong to oblong-elliptic leaves; clustered, white flowers with numerous stamens; and small, oblong, thin-skinned, dark red or purple, tart, edible fruit. Occasional to common in rural areas and in home gardens; occasionally naturalized. Planted in some areas, such as Fiji, Niue, the Cook Islands, and French Polynesia, as a windbreak, street tree, or fruit or fuel-wood tree; timber used occasionally in light construction; a good firewood and one of the main sources of fuel in the Marquesas; ripe fruit eaten, usually by children, as snack food and sometimes made into preserves.

93. Syzygium malaccense (L.) Merr. and Perry MYRTACEAE
"Malay apple," "mountain apple"
syns. Eugenia malaccensis L.; Carophyllus malaccensis (L.) Stokes; Jambosa malaccensis (L.) DC.

Probably indigenous to tropical Asia, but now pantropical in cultivation; an aboriginal introduction into most islands of Melanesia, Polynesia, and Micronesia. Medium tree, up to 10 m or higher, with glossy, oval- or elliptic-oblong leaves; flowers in dense, pedicellate cymes with numerous, scarlet to red-pink stamens; and subglobose to oblong, light green to reddish pink, fleshy, edible fruit containing a large seed. Common in mature fallow forests and open forests, and planted or protected in agricultural areas; protected when clearing for new gardens; common to occasional in home gardens; does not seem to grow on atolls. Planted or protected fruit-tree throughout the Pacific; fruit eaten green and ripe; timber occasionally used in construction; one of the most important medicinal plants in the Pacific, with its bark and leaves used medicinally as a general tonic and for a wide range of maladies.

94. Syzygium spp. (Eugenia spp.) MYRTACEAE
spp. S. brackenridgei (A. Gray) C. Muell.; S. carolinensis (Koidz.) Hosok.; S. clusiaefolium (A. Gray) C. Muell.; S. corynocarpum (A. Gray) C. Muell.; S. curvistylum (Gillesp.) Merr. and Perry; S. dealatum (Burkbill) A.C. Smith; S. diffusum (Turrill) Merr. and Perry; S. effusum (A. Gray) C. Muell.; S. gracilipes (A. Gray) Merr. and Perry; S. inophylloides (A. Gray) C. Muell.; S. myriadena Merr. and Perry; S. neurocalyx (A. Gray) Christoph.; S. onesima Merr. and Perry; S. palauensis Kaneh.; S. richii (A. Gray) Merr. and Perry; S. samarangense (Bl.) Merr. and Perry; S. stelechanthum (Diels) Glassman; Eugenia buettneriana K. Schaum.; E. fierneyana F. Muell.

Indigenous species from tropical Asia through Malesia and Melanesia to western Polynesia and the Caroline Islands in Micronesia; some species may have been aboriginal introductions to western Polynesia and some other islands because of their fruit or cultural importance for making leis and medicine (e.g. S. corynocarpum, S. neurocalyx, and S. samarangense). Variable shrubs to large trees, up to 20 m high, usually with prominently-veined leaves; variable, many-stamened flowers; and generally ovoid to fusiform fruit, both of which are often borne along the branches and trunk. Common in lowland forest, mature fallow forest, open forest, coastal forest, and thickets throughout the high-island Pacific; some species planted or protected in agricultural areas and home gardens. Some species very important timber trees for local use, with some being milled commercially; wood used in house construction, tools and handicrafts, and for firewood; bark, leaves, and buds of fruit used medicinally; fruit of some species eaten; fruit and flowers used in leis and garlands; fruits used to produce dyes and skin lotions.

95. Tamarindus indica L. CAESALPINIACEAE/LEGUMINOSAE
"tamarind"

Probably indigenous to tropical Africa and Asia, although exact origin unknown due to its long existence in cultivation throughout this area; a post-Europeancontact introduction into the Pacific Islands. Medium to large, spreading tree, up to 25 m high, with a rounded crown; pinnately compound leaves with many small leaflets; pink to yellow or cream-coloured flowers with red or purple veins; and oblong, curved or straight, brittle-skinned, velvety brown fruit that are irregularly constricted between the seeds and contain hard, dark brown, obovate seeds embedded in greenish to yellow-brown, acid pulp. Commonly cultivated or protected in rural agricultural areas, particularly on smallholder Indian sugar-cane farms in Fiji; occasional in home gardens and in urban areas; uncommonly naturalized in secondary vegetation and near seashores. Commonly cultivated shade, ornamental, and fruit-tree; timber occasionally used for light construction and firewood; acid pulp of ripe fruit eaten uncooked and used in beverages and jams, and in chutneys by Indians in Fiji; balls of extracted pulp sold in produce markets in Fiji; used medicinally by Indians in Fiji.

96. Teminalia catappa L. COMBRETACEAE
"beach almond," "Indian almond," "Malabar almond," "tropical almond," "coastal almond," "sea almond"

Indigenous to tropical Asia through Malesia, northern Australia, and Melanesia to eastern Polynesia and Micronesia: probably an aboriginal introduction into many areas in the eastern parts of its range, such as Hawaii and some of the atolls. Medium to large, deciduous tree, up to 30 m tall, with whorled, horizontal, widespreading branches arranged in tiers; reddish timber; leathery, shiny, dark green leaves that turn red and yellow before dropping, the new leaves appearing almost immediately; and subobovoid, hard, flattened, two-keeled fruit, green turning yellow, then red, containing thin, fleshy pulp surrounding a single, edible, almond-like kernel. Occasional in coastal strand forests and coastal thickets; common in lowland, inland agricultural areas as a protected or cultivated tree; occasional in home gardens and as a roadside or tourist resort tree. Favourite tree of the ancestral goddess Nei Tituaabane in Kiribati; tree important in sorcery in Nauru; commonly cultivated as an ornamental shade and nut-tree; timber used in general construction and for canoe hulls, paddles, kava bowls, tool handles, war clubs, walking sticks, slit-gongs, and drums; used for firewood; bark and leaves occasionally used to make black dye for pandanus leaves in Fiji and Niue; bark of young stems used for cordage; leaves used for wrapping food for cooking in earthen ovens in Kiribati; roots, bark, young leaves, and fruit used medicinally; mature seed kernel widely eaten; seeds preserved twice yearly in the Solomon Islands for storage; ripe fruit surrounding seeds occasionally eaten in Nauru and Puluwat; fruit eaten by fruit bats; desiccated pith of fruit used to rub corpses in Kiribati; necklaces made of fruit in Nauru; seeds occasionally made into oil in Samoa for mixing with coconut oil.

97. Theobroma cacao L. STERCULIACEAE
"cocoa," "cacao"

Indigenous to tropical South America and long cultivated in Central America; a post-European-contact introduction to the Pacific Islands. Shrub or small tree, up to 8 m high, with fine, soft, pubescent, new branches; thin, leathery, oblong-ovate, acu minate leaves; white flowers with white to pale violaceous or reddish sepals and clawed, white petals with purplish or red nerves, a yellow blade and red or purplish stamens, borne on cauliflorous pedicles at the fallen leaf axils or on branches; and large, ovoid-ellipsoid, ridged, red, purple, or yellow (when ripe), indehiscent fruit (pods) containing ellipsoid, white to deep purple seeds embedded in sweet, edible, mucilaginous pulp. Common in smallholder monocultural plantations and sometimes on extensive estates in Melanesia and Samoa; occasionally intercropped with coconuts, bananas, other overstorey tree species or undercropped with shadetolerant food plants such as Xanthosoma taro; infrequent in rural and home gardens; occasionally naturalized near abandoned plantations; not found on atolls. Commercial export crop in Papua New Guinea, the Solomon Islands, Vanuatu, Fiji, and Western Samoa; minor or residual fruit-tree in some other areas; seeds (beans) fermented and roasted for export for the manufacture of chocolate; seeds dried and roasted on a small scale to make local cocoa in Samoa; pulp of mature fruit eaten as a refreshment or snack food; mature fruit and processed cocoa occasionally sold locally.

98. Thespesia populnea (L.) Sol. ex Correa MALVACEAE
"milo" (Hawaii)
syns. Hibiscus pulpulneus L.; Hibiscus bacciferus Forst. f.; Malvaviscus populneus (L.) Gaertn.

Indigenous from eastern Africa and southern Asia through Malesia to eastern Polynesia and the Marshall and Gilbert Islands in Micronesia; probably an aboriginal introduction into some areas, such as Hawaii, in the eastern part of its range. Shrub or medium tree, up to 20 m high, with leathery, heart-shaped leaves; yellow, cupshaped flowers with a crimson blotch at the base of the throat; and globose, capsular fruit containing a prominently-veined seed. Occasional along seashores, on the inner margins of mangroves, in disturbed habitats, and in rural gardens, home gardens, and urban areas. Believed to be the shadow and spirit medium of the god of prayer and chanting in Tahiti and Hawaii; planted in sacred places in Tahiti, surrounding the house of Kamehameha I of Hawaii, and along bunds between taro gardens in Tuvalu; occasionally cultivated as an ornamental shade or street tree; branches attached to masts of canoes in Tahiti as a token of peace; currently planted as part of a replanting programme in the Cook Islands; durable (especially under water) attractive wood highly esteemed for general construction and wood carving of calabashes, containers, pestles, kava bowls, paddles, drums, weapons, fishing poles and fish-net handles, and carved figures, spirit images, ceremonial thrones, and handicrafts throughout Polynesia; used for canoe hulls and floats in Papua New Guinea; provides sticks for stick games in Nauru; branches held by priests when praying and leaves offered to gods as a substitute for kava in Tahiti; leaves used in making fire by friction; bark used for tannin and dye; bark, stems, leaves, and green fruit used medicinally; young leaves edible; leaves provide black dye for pandanus in Tuvalu; inner bark used for cordage in Hawaii; fruit used for tops in eastern Polynesia; flowers used in garlands in Kiribati; leaves occasionally used in compost in Kiribati.

99. Tournefortia argentea L. f. BORAGINACEAE
"beach heliotrope"
syns. Messerschmidia argentea (L. f.) I.M. Johnst.; Argusia argentea (L. f.) Heine

Indigenous from eastern Africa, South-East Asia through Malesia to the Ryukyu Islands, eastern Polynesia, including Hawaii and the Marshall Islands and Kiribati in Micronesia. Small to medium, wide-spreading, short-bunked tree, 2-12 m tall, with rather stout twigs; deeply grooved bark; dense, silvery grey pubescent, leathery leaves; numerous, small, white flowers in branching, cymose clusters; and greenish white to brown, grape-like fruit. Common in coastal strand forests and beach thickets throughout the Pacific; occasional in agricultural areas, often surrounding excavated taro pits, on atolls; occasionally planted or protected in coastal home gardens on atolls and in coastal areas. Tree features in Tuamotuan and Kiribati mythology; timber occasionally used in light construction, occasionally for canoe hulls, connective and other parts, and for tools, cooking equipment, bailers, backrests, ladles, slit-gongs, diving goggles, carved masks, and rat traps, and for wood carving; a favoured firewood and formerly used in making fire by friction in Kiribati; leaves reportedly eaten raw in "salads" by boat crews in Kiribati; important pig feed in Tokelau and Micronesia; leaves, leaf buds, growing tips, roots, bark, stems, and fruit used medicinally; tender leaves and meristem used to cure fish poisoning in Nauru; leaves used as a female deodorant in Kiribati, in ceremonial dress in New Guinea, and as fish bait and to stuff pigs for cooking in Tokelau; seeds shot through hollow branch tubes in children's games; leaves used in compost and fertilizers in Kiribati; branches at ground surface and immature flower stalks used in voyaging and love magic in Micronesia.

100. Vitex spp. VERBENACEAE
"vitex," "beach vitex," "blue vitex"
spp. Vitex negundo L. (syns. V. incisa Lam.; V. paniculata Lam.); V. trifolia L. (syn. V. rotundifolia L. f.)

Indigenous from southern Africa and the Indian Ocean islands through southern Asia, Malesia, and northern Australia; possibly an aboriginal introduction into some islands in the easternmost parts of its range; probably a recent introduction to some of the atolls. Small, aromatic shrub or small, freely branching tree, up to 10 m high, with tri- to 5-foliate, downy, aromatic leaves, blueviolet flowers in narrow panicles, and small, drupe-like, blue to black fruit. Occasional in beach thickets, along margins of mangroves, in disturbed sites, often along roads near the sea, and in fallow areas; rarely in undisturbed forest; common weed in pastures in the dry zone of Fiji; occasional in home gardens. Sometimes cultivated in home gardens; V. trifolia widely planted as mosquito-repellent hedges and wind-breaks on Majuro and Kwajalein in the Marshall Islands; wood used in light construction and for tools and axe handles in Melanesia; used for firewood; branches used for fishing rods in Nauru; leaves burned as an insect/mosquito repellent; leaves, bark, and fruit used medicinally; leaves eaten with dried coconut and leaves and wood made into tea in Hawaii; flowers, fragrant leaves, and growing tips used in garlands.

Sources

General: Backer and Bakuizen van den Brink 1963; Mernll 1943,1945; Purseglove 1975a and 1975b; Sterly 1970; Whistler 1980a, 1983a, 1991; Yen 1991; and personal records, in-field observations and interpretation of available literature by the author.

Papua New Guinea: Clarke 1971; Henty 1982; Powell 1976a; Sterly 1970; Tarepe and Bourke 1982.

New Caledonia: Jardin 1974; Rageau 1973; Schmid 1981.

Solomon Islands: Kirch and Yen 1982; Sterly 1970; Whitmore 1966, 1969; Yen 1974.

Vanuatu: Cabalion 1984; Gowers 1976; Lebot and Cabalion 1986; Mescarm 1989; Sterly 1970.

Fiji: J.W. Parham 1972; Seemann 1873; Smith 1979, 1981, 1985, 1988, 1991.

Tonga: Sykes 1981; Thaman 1976; Yuncker 1959.

Samoa: Amerson et al. 1982; B.E.V. Parham 1972; Setchell 1924; Whistler 1980b, 1983a, 1983b, 1984.

Niue: Sykes 1970.

Wallis and Futuna: St. John and Smith 1971.

Cook Islands: Sykes 1976; Wilder 1931.

Tuvalu: Chambers 1975; Hedley 1896, 1897; Koch 1983; Woodroffe 1985.

Tokelau: Parham 1971; Whistler 1987.

French Polynesia: Decker 1971; Guerin 1982; Oliver 1974; Petard 1986; Sachet 1983.

Easter Island: Metraux 1940.

Hawaii: Handy et al. 1972; Kaaiakamanu and Akina 1922; Krauss 1974; Neal 1965; Rock 1974; St. John 1973.

Kiribati: Catala 1957; Fosberg and Sachet 1987; Koch 1986; Luomala 1953; Moul 1957; Overy et al. 1982; Polunin 1979; Thaman 1987b, 1990, 1992.

Nauru: Tharnan et al.1985; Thaman 1992.

Micronesia: Fosberg et al. 1979, 1982, 1987; Fosberg and Sachet 1984; Kanehira 1933; Lessa 1977;

Manner 1987; Stemmermann 1981; Wiens 1962.

Marshall Islands: Bryan 1972; Fosberg and Sachet 1975; Lamberson 1982.

Pohnpei: Lessa 1977; Niering 1956; St. John 1948.

Truk: Lessa 1977; Marshall and Fosberg 1975.

Palau: Alkire 1974; Fosberg et al. 1980.

Guam and Marianas: Fosberg et al. 1975; Stone, 1970.

References (A-E)

ADAB. 1986. A review of the hardwood plantation reforestation subsector in Fiji. Canberra: Australian Development Assistance Bureau (ADAB).

Agarwal, V.S. 1986. Economic Plants of India. Calcutta: Kailash Prakashan.

Alex, K.K. 1965. Taro and its relatives in Truk: Field training and interchange in root crops. Koror, Palau. Unpublished report.

Allen, B.J. 1985. "Dynamics of fallow successions and introduction of Robusta coffee in shifting cultivation areas in the lowlands of Papua New Guinea." Agroforestry Systems 3: 227-238.

Allen, W. 1965. The African Husbandman. Edinburgh: Oliver and Boyd.

Ali, 1. 1976. Urban gardening in the Kalabo Housing Estate. School of Social and Economic Development, University of the South Pacific, Suva. Unpublished research paper.

-. 1986. Polyculture to monoculture: A case-study of changing agriculture in the Yaladro Cane Sector, Ba Province, Viti Levu, Fiji. Unpublished M.A. thesis in Geography, University of the South Pacific, Suva.

Alkire, W.H. 1974. "Numbers of plant, insect and land bird species on nineteen remote islands in the southern hemisphere." Journal of the Linnaean Society: Biology 6: 143-152.

Amerson, A.B., W.A. Whistler, Jr., and T.D. Schwaner. 1982. Wildlife and Wildlife Habitat of American Samoa: 1. Environment and Ecology. Washington, D.C.: United States Department of the Interior Fish and Wildlife Service.

APU (Agricultural Planning Unit). 1978. Current State and Prospects of the Coolc Islands Citrus Industry: A Summary. Agricultural Planning Unit, Ministry of Agriculture and Fisheries, Rarotonga.

-. 1979. Results: Banana Survey 11, Aitutaki, October 1979. Agriculture and Fisheries Statistical Working Paper No. 4, Agricultural Planning Unit, Ministry of Agriculture and Fisheries, Rarotonga.

Arnold, J.E.M. 1984. "Economic constraints and incentives in agro-forestry." In: J.K. Jackson, ed. Social, Economic, and Institutional Aspects of Agro-forestry. Tokyo: The United Nations University, pp. 3-10.

Backer, C.A., and R.C. Bakhuizen van den Brink, Jr. 1963. Flora of Java (Spermatophytes Only): Vol. 1. Groningen, Netherlands: N.V.P. Noordhoff.

Baines, G.B.K. 1979. Mangroves for national development: A report on the mangrove resources of Fiji. (Unpublished manuscript).

Baker, P.T. 1979. "Migration and human adaptation." In: C. Fleming, and I. Prior, eds. Proceedings of the 1979 Seminar on Migration, Adaptation and Health in the Pacific. Wellington: Wellington Hospital, pp. 3-13.

Bakker, M.K. 1977a. "The population problem in the South Pacific: Population growth in the South Pacific." South Pacific Bulletin Third Quarter: 7-13.

-. 1977b. "The population problem in the South Pacific: The exploding 'population bomb'." South Pacific Bulletin Fourth Quarter: 17-22.

Balgooy, M.M.J. van. 1971. Plant-geography of the Pacific. Blumea Supplement VI, Rijksherbarium, Leiden, Netherlands.

-. 1976. "Phytogeography." In: K. Paijmans, ed. New Guinea Vegetation. Amsterdam: Elsevier Scientific Publishing Co., pp. 1-22.

Barrance, A. 1986. Stumps versus Seedlings as Planting Stock for Cordia alliodora in Vanuatu. Forest Research Report 3/86, Vanuatu Forest Service, Port Vila.

-. 1988. Assistant Forest Research Officer, Vanuatu Forest Service, Port Vila. Personal communication.

Barrau, J. 1958. Subsistence Agriculture in Melanesia. Bulletin 219, Bernice P. Bishop Museum, Honolulu.

-. 1961. Subsistence Agriculture in Polynesia and Micronesia. Bulletin 223, Bernice P. Bishop Museum, Honolulu.

Basha, S., P. Chand, K. Seniloli, U. Naugavule, P. Lakhan, D. Prasad, J. Henson, E. Vaqewa, A. Rua, J. Narayan, H. Kumar, S.C. Lal, and M. Samuela. 1974. Food production in Suva. University of the South Pacific, Suva, Fiji. Unpublished research paper.

Baumer, M. 1990. The Potential Role of Agroforestry in Combating Desertification and Environmental Degradation, with special reference to Africa. Technical Centre for Agricultural and Rural Co-operation, Wageningen, Netherlands.

Baxter, M.W.P. 1977. Food in Fiji: Aspects of the Produce and Processed Foods Distribution Systems. Centre for Applied Studies in Development, University of the South Pacific, Suva.

Bayliss-Smith, T.P. 1980. "Population pressure, resources and welfare: Towards a more realistic measure of carrying capacity." In: Population-Environment Relations in Tropical Islands: The Case of Fiji. MAB Technical Notes 13, Unesco, Rome, pp. 61-93.

-. 1982. The Ecology of Agricultural Systems. Cambridge: Cambridge University Press.

Beer, J. 1988. Project Investigator (Consultant to the Fiji-German Forestry Porject, Suva). Centro Agrondmico Tropical de Investigación y Ensehanza (CATIE), Turrialba, Costa Rica.

Beets, W.C. 1989. The Potential Role of Agroforestry in ACP Countries. Wageningen: CTA/EC.

Bell, T.I.W. 1981. Tree Spacing and Cattle Grazing in Young Pinus caribaea Plantations in Fiji. Fiji Pine Research Paper No. 9, Fiji Pine Commission, Lautoka, and Fiji Forestry Department, Suva.

Bell, T.I.W., and T. Evo. 1982. Energy Plantations in the Fiji Dry Zone. Fiji Pine Research Paper No. 10, Fiji Pine Commission, Lautoka, and Fiji Forestry Department, Suva.

Bene, J.G., H.W. Beall, and A. Cote. 1977. Trees, Food and People: Land Management in the Tropics. Ottawa: International Development Research Centre.

Bellwood, P. 1978. The Polynesians: Prehistory of an Island People. London: Thames and Hudson.

Benjamin, C., and I. Wapi. 1982. "Subsistence gardening on the Hoskins Oil Palm Scheme. " In: R.M. Bourke, and V. Kesavan, eds. Part 1, pp. 168-175.

Bindon, J.R. 1982. "Breadfruit, bananas, beef, and beer: Modernization of the Samoan diet." Ecology of Food and Nutrition 12: 49-60.

Blaut, J.M. 1962. "The nature and effects of shifting agriculture." In: Proceedings: Symposium on the Impact of Man on Humid Tropics Vegetation (Goroka, Territory of Papua New Guinea, September, 1960). Unesco Scientific Co-operation Office for South-East Asia, Jakarta, pp. 185-198.

-. 1967. "Geography and the development of peasant agriculture." In: S.B. Cohen, ed. Geography and the American Environment: Voice of America Forum Lectures. Washington, D.C.: U.S.I.A., pp. 217-238.

Bonnemaison, J. 1984. "Les jardins magiques: Le géosystème de ['horticulture vivriere dans une 'ile mélanésienne du Pacifique Sud." In: Le Développement Rural en Question. Memoires ORSTOM No. 106, ORSTOM, Paris.

-. 1985. "The tree and the canoe: Roots and mobility in Vanuatu societies." In: M. Chapman, ed. Mobility and Identity in the Pacific. Special issue Pacific Viewpoint 26(1): 30-62.

-. 1986. Magic gardens on Tanna. Paper presented at a departmental seminar, 26 March. Department of Human Geography, Research School of Pacific Studies, Australian National University, Canberra.

Bottrell, G. 1978. Integrated Pest Management. Council on Environmental Quality, Washington, D.C.

Bourke, R.M. 1976. "Food crop farming systems on the Gazelle Peninsula of New Britain." In: K. Wilson, and R.M. Bourke, eds. 1975 Papua New Guinea Food Crops Conference Proceedings. Port Moresby: Department of Primary industry, pp. 81-105.

Bourke, R.M., B. Carrad, and P. Heywood. 1982. "Papua New Guinea's food problems: Time for action." In: R.M. Bourke, and V. Kesavan, eds. Proceeding of the Second Papua New Guinea Food Crops Conference, Part l, p. 37. Abstract.

Bourke, R.M., and V. Kesavan, eds. 1982. Proceedings of the Second Papua New Guinea Food Crops Conference, Parts 1-111. Port Moresby: Department of Primary Industry.

Brookfield, H.C., with D. Hart. 1971. Melanesia: A Geographical Interpretation of an Island World. London: Methuen and Co.

Brown, E.D.W. 1931. "Polynesian lets." American Anthropologist 33(4); 615-619.

Brown, L.R. 1982. "Can we live better on less?" lnternational Wildlife 12(1): 3135.

Bryan, E.H., Jr. 1972. Life in the Marshall Islands. Honolulu: Pacific Science Information Center, Bernice P. Bishop Museum.

Buck, L. 1989. Agroforestry Extension Training Source Book. New York: CARE International.

Budd, W.W., L.H. Hardesty, and W.T. Hinds. 1990. "A conspectus for agroforestry." In: W.W. Budd, 1. Duchart, L.H. Hardesty, and F. Steiner, eds., pp. 321-335.

Budd, W.W., I. Duchart, L.H. Hardesty, and F. Steiner, eds. 1990. Planning for Agroforestry. Amsterdam: Elsevier.

Budowski, G. 1982. "Applicability of agro-forestry systems." In: L. MacDonald, ed. Agroforestry in the African Humid Tropics. Tokyo: The United Nations University, pp. 13-16.

Burden, O.J., and D.G. Coursey. 1977. "Bananas as a food crop." In: C.L.A. Leaky, and J.B. Wills, eds. Food Crops of the Lowland Tropics. London: Oxford University Press, pp. 97100.

Burkill, I.H. 1901. "The flora of Vava'u, one of the Tonga Islands." Journal of the Linnaean Society - Botany. 35: 19-65.

Cabalion, P. 1984. "Les noms plantes en biclamar: Origines formations et determinations botaniques." Journal de la Société des Océanistes 78 (tome XL): 107-120.

Carlquist, S. 1965. Island Life: A Natural History of the Islands of the World. Garden City, N.Y.: American Museum of Natural History, The Natural History Press.

-. 1980. Hawaii: A Natural History (2nd edn). Pacific Tropical Botanical Garden, Lawai, Kauai, Hawaii.

Carrad, B. 1982. "Economic aspects of smallholder practices: Mixed cropping of food and coffee." In: R.M. Bourke, and V. Kesavan, eds. Part 11, pp. 294-302.

Carter, J., ed. 1981. Pacific Islands Yearbook (14th edn). Sydney: Pacific Publications.

-. 1984. Pacific Islands Yearbook (15th edn). Sydney: Pacific Publications.

Catala, R. L. A. 1957. "Report on the Gilbert Islands: Some aspects of human ecology." AtolJ Research Bulletin 59: 1-187.

Central Planning Department (Kiribati). 1983. National Development Plan 19831986. Central Planning Department, Bairiki, Tarawa.

Central Planning Department (Tonga). 1981. Kingdom of Tonga Fourth Five-year Development Plan 1980-1985. Nuku'alofa: Government Printing Department.

Central Planning Office (Fiji). 1985. Fiji's Ninth Development Plan 1986-1990: Policies, Strategies and Programmes for National Development. Suva: Government Printer.

Chambers, A. 1975. Nanumea Report: A Socio-economic Study of Nanumea Atoll, Tuvalu. Victoria University of Wellington Rural Socio-economic Survey of the Gilbert and Ellice Islands. Department of Geography, Victoria University, Wellington.

Chambers, R. 1983. Rural Development: Putting the Last First. London: Longmans.

Chan, H. 1987. "Mangrove forest management in the ASEAN region." Tropical Coastal Areas Management 2(3): 6-8.

Chapman, M., and P. Pirie. 1974. Tasi Mauri: A Report on Population and Resources of the Guadalcanal Weather Coast. East-West Center Population Institute and the University of Hawaii, Honolulu.

Cheeseman, T.F. 1903. "The flora of Rarotonga." Transactions of the Linnaean Society, London, Second Series, Botany 6:161-213.

Christophersen, E. 1927. Vegetation of Pacific Equatorial Islands. Bulletin 44, Bernice P. Bishop Museum, Honolulu.

Clarke, W.C. 1965. "From extensive to intensive shifting cultivation: A succession from New Guinea." Ethnology 5: 347-359.

-. 1971. People and Place: An Ecology of a New Guinean Community. Berkeley: University of California Press.

-. 1973. "The dilemma of development." In: H.C. Brookfield, ed. The Pacific in Transition. London: Edward Arnold, pp. 275-298.

-. 1977. "The structure of permanence: The relevance of self-subsistence communities for world ecosystem management." In: T. Bayliss-Smith, and R. Feachem, eds. Subsistence and Survival: Rural Ecology in the Pacific. New York: Academic Press, pp. 363-384.

-. 1978. "Progressing with the past: Environmentally sustainable modifications to traditional agricultural systems." In: E.K. Fisk, ed. The Adaptation of Traditional Agriculture: Socio-economic Problems of Urbanization. Monograph No. 11, Development Studies Centre, Australian National University, Canberra, pp. 142- 157.

-. 1991. "Agro-deforestation in Melanesia." Cultural Survival Quarterly 15(2): 4548.

Clements, R., ed. 1988. Proceedings of the Technical Meeting on Agroforestry in Tropical Islands. (Meeting held in IRETA, USP School of Tropical Agriculture' W. Samoa, 23-27 February 1987). Institute for Research, Extension and Training in Agriculture (IRETA), University of the South Pacific, School of Agriculture, W. Samoa.

Cole, T.G., M. Falanruw, C.D. MacLean, C.D. Whitsell, and A.H. Ambacher. 1987. Vegetation Survey of the Republic of Palau. Research Bulletin PSW-22, Pacific Southwest Forest and Range Experiment Station, U.S. Department of Agriculture.

Committee on Food Supplies (BSIP). 1974. Food and Self-reliance: Report of the Committee on Food Supplies. Government House, Honiara, British Solomon Islands Protectorate.

Commoner, B. 1971. "Evaluating the biosphere." In: T. Detwyler, ed. Man's Impact on Environment. New York: McGraw-Hill, pp. 50-60.

Connell, J. 1978. Taim Bilong Mani: The Evolution of Agriculture in a Solomon Island Society. Monograph No. 12, Development Studies Centre, Australian National University, Canberra.

-. 1984. Diets and Dependency: Food and Colonialism in the South Pacific. Occasional Paper No. 1, Ideas Centre Research and Education Unit, Australian Freedom from Hunger Campaign, Sydney.

Connell, J., and R. Curtain. 1982. "The political economy of urbanization in Melanesia." Singapore Journal of Tropical Geography 3(2): 119-136.

Corner, E.J.H. 1962. "Botany and prehistory." In: Proceedings: Symposium on the impact of Man on Humid Tropics Vegetation, Goroka, Territory of Papua AJew Guinea, September, 1960. Unesco Scientific Co-operation Office for South-East Asia, Jakarta, pp. 38-41.

Coyne, T. (ea. by J. Badcock, and R. Taylor). 1984. The Effect of Urbanization and Western Diet on the Health of Pacific Island Populations. Technical Paper No. 186, South Pacific Commission, Noumea.

C.P.O. (Central Planning Office, Fiji). 1980. Fiji's Eighth Development Plan 19811985, Vol. 1. Suva: Central Planning Office.

Cruz, R.E. de la, and N.T. Vergara. 1987. "Protective and ameliorative roles of agroforestry: An overview." In: N.T. Vergara, and N.D. Briones, eds. Agroforestry in the Humid Tropics: Its Protective and Ameliorative Roles to Enhance Productivity and Sustainability. Environment and Policy Institute, East-West

Center, Honolulu/Southeast Asian Regional Center for Graduate Study and Research in Agriculture, College, Laguna, Philippines.

Cuzent, G. 1983. Archipel de Tahifi: Recherches sur les principales productions végétales. Papeete: Editions Haere Po No Tahiti.

Dahl, A.L. 1980. Regional Ecosystems Survey of the South Pacific Area. Technical Paper No. 179, South Pacific Commission, Noumea.

Darwin, C.R. 1895. On the Origin of Species by Natural Selection, or the Preservation of Favoured Races in the Struggle for Life. London: John Murray.

Dasman, R.F., J.P. Milton, and P.H. Freeman. 1973. Ecological Principles for Economic Development. New York: John Wiley.

Decker, B.G. 1971. Plants, man and landscape in Marquesan valleys, French Polynesia. Ph.D thesis, Department of Geography, University of California, Berkeley.

Denevan, W.M., J.M. Treacy, J.B. Alcorn, C. Padoch, J. Denslow, and S. Flores Paitan. 1984. "Indigenous agroforestry in the Peruvian Amazon: Bora Indian management of swidden fallows." Interciencia 9(6): 346-357.

Denevan, W.M., and C. Padoch, eds. 1988. Swidden-Fallow Agroforestry in the Peruvian Amazon. (Advances in Economic Botany, Vol. 5). New York: The New York Botanical Gardens, Bronx.

Department of Agriculture (Rarotonga). 1979. Statistics of Commercial Agricultural Production and Prices 1978. Department of Agriculture, Rarotonga. Cook Islands.

Department of Economic Planning (Western Samoa). 1984. Western Samoa's Fifth National Development Plan 1985-1987. Apia: Department of Economic Planning.

Djogo, A.P.Y. 1992. The Possibilities of Using Local Drought-resistant Multipurpose Tree Species as Alternatives to Lamtoro (Leucaena leucocephala) for Agroforestry and Social Forestry in West Timor. Working Paper No. 32, Environment and Policy Institute, East-West Center, Honolulu.

Dilmy, A. 1962. "The effect of fire used by early man on the vegetation of humid tropics." In: Proceedings: Symposium on the Impact of Man on Humid Tropics Vegetafion, Coroka, Territory of Papua New Guinea, September, 1960. Unesco Scientific Co-operation Office for South-East Asia, Jakarta, pp. 119-122.

Dorst, J. 1972. "What humans are doing." In: N. Polunin, ed. The Environmental Future. London: Macmillan, pp. 67-99.

Douglas, Norman and Ngaire, Douglas, eds. 1989. Pacific Islands Yearroook (16th edn). North Ryde, Australia: Angus and Robertson.

Drake, del Castillo, E. 1893. Flore de la Polynesie française Paris: Libraire deI'Academie de Medecin.

Drysdale, P.J. 1982. Cattle Grazing and Pine Forests. Technical Report No. 4, Fiji Pine Commission, Lautoka.

-. 1988a. Rainforest Management and Conservation in Fiji: A Prescription for Action. Vol. 2 of Fiji's Rainforests: Our Heritage and Future. Proceedings of the 2nd National Conservation Congress, Suva, Fiji, 9-10 June 1988. Suva: National Trust for Fiji.

-. 1988b. General Manager. Fiji Pine Commission, Lautoka. Personal communication.

Eaton, C. 1988. Directed smallholder farming in Fiji: A case study of Virginia tobacco production. Unpublished M. Phil. thesis in Geography, University of the South Pacific, Suva.

Eckholm, E.P. 1976. Losing Ground: Environmental Stress and World Food Prospects. Oxford: Pergamon Press.

Eele, G. 1978. 1974-1975 Agricultural Statistics Survey: A Sample Survey of Solomon Islands Smallholder Agriculture. Honiara: Statistics Office.

Egler, F.E. 1939. "Vegetation zones of Oahu, Hawaii." Empire Forestry Journal 18(1): 44-57.

Etherington, D.M. 1983. "The design and economic evaluation of agroforestry technologies for the South Pacific." In: A. Schirmer, ed. The Role of Agroforestry in the Pacific. Hamburg: German Foundation for International Development, pp. 71-88.

References (F-R)

Fa'anunu, H.'O. N.d. Vanilla intercropping systems. Ministry of Agriculture, Forest and Fisheries, Nuku'alofa. Mimeo.

Fairbairn, I.J. 1971. "Pacific island economies." Journal of the Polynesian Society 80(1): 74-118.

Falanrnw, M.V.C. 1985. Traditional agriculture and resource management systems in the high islands of Micronesia. Prepared for the Office of Technology Assessment Project on Integrated Renewable Resource Management for US Insular Areas. Yap Institute of Natural Science, Yap, Federated States of Micronesia.

-. 1990. "The food production system of the Yap islands." In: K. Landauer, and M. Brazil, eds. Tropical Home Gardens. Tokyo: United Nations University Press, pp. 94-104.

Falanruw, M.C., T.G. Cole, A.H. Ambacher, K.E. McDuffie, and J.E. Maka. 1987a. Vegetation Survey of Moen, Dublon, Fefan, and Eten, Federated States of Micronesia. Resource Bulletin PSW-20, Pacific Southwest Forest and Range Experiment Station, U.S. Department of Agriculture.

Falanruw, M.C., C.D. Whitesell, T.G. Cole, C.D. MacLean, and A.H. Ambacher, 1987b. Vegetation Survey of Yap, Federated States of Micronesia. Resource Bulletin PSW-21, Pacific Southwest Forest and Range Experiment Station, U.S. Department of Agriculture.

FAO. 1980a. Coconut Industry of Fiji: An Analysis of the Viability of the Coconut Industry with Recommendations for its Improvement. (TCPIFIJ/8901(1)). Rome: FAO.

-. 1980b. Use of Coconut Lands in Fiji: An Analysis of Alternative Land Uses for Coconut Areas. (TCP/FIJl001(I)). Rome: FAO.

Farrell, B.H. 1972. "The alien and the land in Oceania." In: R.G. Ward, ed. Man in the Pacific Islands. London: Oxford University Press. pp. 34-73.

Fiji Pine Commission. 1984. Annual Report 1983. Lautoka: Fiji Pine Commission.

Firman, I.D. 1972. "Susceptibility of banana cultivars to fungus leaf diseases in Fiji." Tropical Agriculture 49(3): 189-196.

Firth, R. 1936. We, the Tikopia. London: George Allen and Unwin.

-. 1940. The Work of the Gods in Tikopia. (2 vole). Monograph on Social Anthropology, The London School of Economics and Political Science, London.

-. 1959. Social Change in Tikopia. New York: Macmillan.

-. 1970. Rank and Religion in Tikopia. Boston: Beacon Press.

Fischer, J.L., and A.M. Fischer. 1970. The Eastern Carolines. New Haven: Human Relations Areas Files Press.

Fisk, F.K. 1972. "Motivation and modernization." Pacific Perspective 1(1): 21-33.

-. 1977. "Adapting traditional agricultural systems to serve urban food markets." In: M. Lambert, ea., pp. 203-208.

Fitzroy, G.J. 1981. "Commentary: Influence of development factors on nutritional patterns in the Solomon Islands." Ecology of Food and Nutrition 10: 87-119.

Fleckenstein, F. von. 1978. Dooryard Food Gardens in Port Moresby: An Original Study of Morata together with Comparison of Other Studies Past and Present. Occasional Paper, Economics Department, University of Papua New Guinea, Port Moresby.

Flenley, J.R., and S.M. King. 1984. "Late Quaternary pollen records from Easter Island." Nature 307: 47-50.

Foist, I.A. 1974. Fiji Pine Scheme: Cattle Grazing under Trees Feasibility Investigation. Lautoka: Fiji Pine Commission.

Foreman, D.B. 1971. A Checklist of the Vascular Plants of Bougainville with Descriptions of Some Common Forest Trees. Botany Bulletin No. 5, Division of Botany, Department of Forests, Lae, Papua New Guinea.

Forum Secretariat. 1991. Profiles of the Forum Island Countries. Suva, Fiji: Forum Secretariat.

Fosberg, F.R. 1960. "The vegetation of Micronesia." Bulletin of the American Museum of Natural History 119(1): 1-75.

-. 1965. "The island ecosystem." In: F.R. Fosberg, ed. Man's Place in the Island Ecosystem: A Symposium. Honolulu: B.P. Bishop Museum Press, pp. 1-6.

-. 1982. "Geographical checklist of the Micronesian Pteridophyta and gymnosperms." Micronesica 18(1): 23-82.

-. 1987. "A geographical checklist of the Micronesian monocotyledonae." Micronesica 20(1-2): 19-130.

Fosberg, F.R., M.V.C. Falanruw, and M.-H. Sachet. 1975. Vascular Flora of the Northern Marianas Islands. Smithsonian Contributions to Botany No. 22, Smithsonian Institution, Washington, D.C.

Fosberg, F.R., D. Otobed, M.-H. Sachet, R.L. Oliver, D.A. Powell, and J.E. Canfield. 1980. Vascular Plants of Palau with Vernacular Names. Department of Botany, Smithsonian Institution, Washington, D.C.

Fosberg, F.R., and M.-H. Sachet. 1975. Flora of Micronesia, 2: Casuarinaceae, Piperaceae, and Myricaceae. Smithsonian Contributions to Botany No. 24, Smithsonian Institution, Washington, D.C.

-. 1984. "Micronesian Poaceae: Critical and distributional notes." Micronesica 18(2): 45-120.

-. 1987. "Flora of the Gilberts, checklist." Atoll Research Pulletin 29: 1-30.

Fosberg, F.R., M.-H. Sachet, and R. O1iver. 1979. "A geographical list of the Micronesian dicotyledonae." Micronesica 15(1-2): 41-295.

-. 1982. "Geographical checklist of the Micronesian Pteridophyta and gymnosperms." Micronesica 18(1): 23-82.

-. 1987. "A geographical checklist of the Micronesian monocotyledonae." Micronesica 20(1-2): 19-130.

Fownes, J.H., and W.C. Raynor. In press. "Seasona]ity and yield of breadfruit cultivars in the indigenous agroforestry system of Pohnpei, Federated States of Micronesia." Tropical Agriculture (Trinidad).

Fox J. W., and K.B. Cumberland. 1962. Western Samoa: Land, Life and Agriculture in Tropical Polynesia. Christchurch, N.Z.: Whitcombe and Tombs.

Gallasch, H. 1976. "Integration of cash and food cropping in the lowlands of Papua

New Guinea." In: K. Wilson, and R.M. Bourke, eds. 1975 Papua New Guinea Food Crops Conference Proceedings. Port Moresby: Department of Primary Industry, pp. 101-116.

Gholz, H.L., ed. 1987. Agroforestry: Realities, Possibilities and Potentials. Dordrecht: Martinus Nijhoff (in cooperation with ICRAF).

Gibson, G. 1988. Agronomist. Tanna Coffee Development Project, Tanna, Vanuatu. Personal communication.

Gliessman, S.R. 1990. "Applied ecology and agroecology: Their role in the design of agricultural projects for the humid tropics." In: R. Goodland, ed. Race to Save the Tropics. Washington, D.C.: Island Press, pp. 33-47.

Goodenough, W. 1951. Property, Kin and Community on Trak. Publication No. 46, Yale University Publications in Anthropology, New Haven.

Gowers, S. 1976. Some Common Trees of the New Hebrides and their Vernacular Names. Forestry Section, Department of Agriculture, Port Vila.

Gregersen, H., S. Draper, and D. Elz, eds. 1989. People and Trees: The Role of Social Forestry in Sustainable Development. EDI Seminar Series. Washington, D.C.: The World Bank.

Gregor, E.W. 1972. Integration of grazing in tropical forestry - An experiment in combining cattle raising with pine plantation forestry in Fiji. Paper presented at the Seventh World Forest Congress, Buenos Aires. Mimeo.

Groome (J.G.) and Associates. 1981. Namosi Provincial Forest Development. Report prepared for the Fiji Government under the New Zealand Government's Bilateral Aid Programme, Groome and Associates, Wellington.

Grossman, L. 1981. "The cultural ecology of economic development." Annals of the Association of American Geographers 71(2): 220-236.

Groube, L.M. 1971. "Tonga, Lapita pottery, and Polynesian origins." Journal of the Polynesian Society 80(3): 220-236.

Guerin, M. 1982. "The flora of the atolls of French Polynesia." In: M. Lambert, ed. Regional Technical Meeting on Atoll Cultivation, Papeete, Tahiti, French Polynesia, 14-19 April 1980: Collected Papers. Technical Paper No. 180, South Pacific Commission, Noumea, pp. 77-89.

Guillaumin, A. 1911. Catalogue des plantes phanerogames de la Nouvelle-Caledonie et dependences lies des Pins et Loyalty: Apetales et monocotyledones. Extract des Annales du Musee colonial de Marseille Tome XIX, 1911. Marseille: Institut Colonial.

-. 1931. "Contributions to the flora of the New Hebrides: Plants collected by S.E. Kajewski in 1928 and 1929." Journal of the Arnold Arboretum XII: 221-264.

-. 1932. "Contributions to the flora of the New Hebrides: Plants collected by S.E. Kajewski in 1928 and 1929." Journal of the Arnold Arboretum XIII: 1-126.

-. 1948. Flore Analytique et Synoptique de la Nouvelle-Caledonie Phanerogames. Editions Larose. Office de la Recherche Scientifique Coloniale, Paris.

Haddon, A.C., and J. Hornell. 1975. Canoes of Oceania. Bernice P. Bishop Special Publications 27, 28, and 29, Bishop Museum Press, Honolulu.

Haen, H.A. de. 1988. Protection of Agricultural Land Resources through Interaction of Agriculture and Forestry: Economic Assessment and Policy Recommendations. Technical Report No. 4, Fiji-German Forestry Project (Veutsche Gesellschaft fur technische Zusammenarbeit (GTZ), Suva.

Hamilton, L.S., and S.C. Snedaker. 1984. Handbook for Mangrove Management.

East-West Center Energy and Policy Institute, Honolulu, in cooperation with IUCN, Unesco, and UNEP.

Hammon, R. 1980. Multiple Resources Nomination Form for Kaho'olawe Archaeological Site. Washington, D.C.: National Register of Historic Sites.

Hamnett, M.P., R.J. Surber, D.E. Surber, and M.T. Denoncour. 1981. "Unbalanced books: Economic vulnerability in the Pacific." East-West Perspectives 2(3): 6-12.

Handy, E.S.C., E.G. Handy, with M.K. Pukui. 1972. Native Planters of Old Hawaii: Their Life, Lore, and Environment. Bulletin 233, Bernice P. Bishop Museum, Honolulu.

Hansell, J.R.F., and J.R.D. Wall. 1976. Land Resources of the Solomon Islands: Volume 1. Introduction and Recommendations. Land Resource Study 18, Land Resources Division, Ministry of Overseas Development, Surbiton, Surrey.

Hardaker, J.B., E.M. Fleming, and G.T. Harris. 1984a. "Smallholder modes of agricultural production in the South Pacific: Prospects for development." Pacific Viewpoint25(2): 19~211.

-. 1984b. "Modes of agricultural production in the South Pacific: A rejoinder." Pacific Viewpoint 25(2): 223-226.

Hardin, G. 1968. "The tragedy of the commons." Science, 162: 1243-1248.

Harlan, J. 1975. "Our vanishing genetic resources." In: P. Abelson, ed. Food: Politics, Economics, Nutrition and Research. Washington, D.C.: American Association for the Advancement of Science, pp. 157-160.

Harris, G.T. 1977. Subsistence Food Gardening in a Port Moresby Suburb: Gerehu. Discussion Paper No. 32, Economics Department, University of Papua New Guinea, Port Moresby.

Hautofa, E., A. Proctor, R.G. Ward, and D.E. Yen. 1980. "Other infrastructural conditions." In: R.G. Ward, and A. Proctor, eds. South Pacific Agriculture: Choices and Constraints. South Pacific Agricultural Survey 1979. Canberra: Asian Development Bank in association with Australian National University Press, pp. 181-193.

Haynes, P.H. 1977. "Rootcrop production and use in Fiji." In: M. Lambert, ed. Collected Papers: Regional Meeting on Production of Root Crops, 24-29 October 1975, Suva, Fiji. Technical Paper No. 174, South Pacific Commission, Noumea, pp. 4450.

Hedley, C. 1896. The Atoll of Funafuti, Ellice Group: Its Zoology, Botany, Ethnology, and General Structure: Part 1. General Account of the Atoll of Funafuti. Memoir III. Sydney: Australian Museum.

-. 1897. The Atoll of Funafuti, Ellice Group: Its Zoology, Botany, Ethnology, and General Structure: Part 4. The Ethnology of Funafuti. Memoir III. Sydney: Australian Museum.

Henderson, C.P., and I.R. Hancock. 1988. A Guide to the Useful Plants of the Solomon Islands. Research Department, Ministry of Agriculture and Lands, Honiara.

Hendren, G.H. 1976. "Recent settlement changes on Ulawa, southeast Solomon Islands." In: R.C. Green, and M.M. Cresswell, eds. Southeast Solomon Islands Cultural History: A Preliminary Survey. Bulletin 11, Roya] Society of New Zealand, Wellington, pp. 149-159.

Henty, E.E. 1982. "Some nut-bearing plants in Papua New Guinea." In: R.M. Bourke, and V. Kesavan, eds. Part 1, pp. 78-85.

Herklots, G. 1972. Vegetables in South-east Asia. London: George Allen and Unwin. Heywood, P. 1991. "Food dependency and food and nutrition in the South Pacific." In: M. Bakker, and R.R. Thaman, eds. Population, Food and Development. (USP/SPC/UNFPA Regional Symposium, University of the South Pacific, Suva, Fiji, 12-16 November 1990). Population Studies Programmes, University of the South Pacific, Suva, Fiji, pp. 74-80.

Heywood, P., and M. Nakikus. 1982. "Protein, energy and nutrition in Papua New Guinea." In: R.M. Bourke and V. Kesavan, eds. Part 11, pp. 303-324.

Hiroa, Te Rangi (P.H. Buck). 1930. Samoan Material Culture. Bulletin 75, Bernice P. Bishop Museum, Honolulu.

-. 1932a. The Ethnology of Tongareva. Bulletin 92, Bernice P. Bishop Museum, Honolulu.

-. 1932b. The Ethnology of Manihiki and Rakahanga. Bulletin 99, Bernice P. Bishop Museum, Honolulu.

Holdsworth, D.K. 1975. Medicinal Plants of Papua New Guinea. Technical Paper No. 175, South Pacific Commission, Noumea.

Holdsworth, D., and P. Mahana. 1983. "Traditional medicinal plants of the Huon Peninsula, Morobe Province, Papua New Guinea." International Journal of Crude Drug Research 21: 123-133.

Hommon, R. 1980. Multiple resources nomination form for Kaho'olawe Archaeological Sites. National Register of Historic Sites, Washington, D.C.

Howcroft, N.H.S. 1974. Forest Pasture Introductions and Improvement Trials at Bulolo. Papua New Guinea field information sheet, Department of Forests, Boroko.

-. 1983. Agroforestry in Papua New Guinea. Paper presented at Workshop on Present and Future Role of Agroforestry in the Pacific, 12-17 December, University of the South Pacific, Suva.

Hunter-Anderson, R. 1984. Notes on a comparative study of traditonal horticulture in five high island groups in western Micronesia: Palau, Yap, Truk, Ponape, and Kosrae. Micronesian Area Research Center, University of Guam, Mangilao, Guam. Unpublished paper.

-. 1987. Indigenous Fresh Water Management Technologies of Trak, Ponape and Kosrae, Eastern Caroline Islands, and of Cuam, Mariana Islands, Micronesia. Technical Report 65, Water and Energy Research Institute of the Western Pacific, University of Guam, Mangilao, Guam.

Jacober, F. 1977. "Evolution recense de l'alimentation en Polynesie Francaise et ses consequences." Medecine Oceanienne 7: 8-86.

Jacovelli, P.A., and P.E. Neil. 1984. A First Look at Agroforestry in Vanuatu. Forest Research Report 2/84, Vanuatu Forest Service, Port Vila.

Jansen, A.A. 1973. Report on a Nutrition Survey in the Kingdom of Tonga. Manila: World Health Organization Western Pacific Regional Office.

Jansen, A.A.J., and J.V. Wilmott. 1971. Nutrition and Dietary Survey of Urban and Rural Populations in the British Solomon Islands Protectorate. South Pacific Health Service WHO Survey 3, WHO Publications, Suva.

Janzen, D.H. 1973. "Tropical agroecosystems." Science 82(4118): 212-219.

-. 1990. "Sustainable society through applied ecology: The reinvention of the village." In: R. Goodland, ed. Race to Save the Tropics: E;cology and Economics for a Sustainable Future. Washington, D.C.: Island Press, pp. ix-xiv.

Jardin, C. 1974. Kulu, Kuru, Uru: Lexicon of Names of Food Plants in the South Pacific. Noumea: South Pacific Commission.

Jeffries, D.J. 1979. From Kaukau to Coke: A Study of Rural and Urban Food Habits in Papua New Cuinea. Report No. 1, Centre for Resource and Environmental Studies, Australian National University, Canberra.

Johnson, J.S. 1981. "Fiji's food supply: Where is it coming from?" Fiji Food and Nutrition Newsletter 2(2): 1 and 7.

Johnson, J.S., and J.N. Lambert. 1982. The National Food and Nutrition Survey of Fiji. National Food and Nutrition Development Programme Field Document No. 6 (FIJ/79/004), National Food and Nutrition Committee, United Nations Development Programme, and the Food and Agricultural Organization of the United Nations, Suva.

Johnson, W.B. 1951. "The citrus industry of the Cook Islands." New Zealand Ceographer 7(2): 121-138.

Johnston, K.M. 1967. Village Agriculture in Aitutaki, Cook Islands. Pacific Viewpoint Monograph No. 1, Department of Geography, Victoria University, Wellington.

Kaaiakamanu, D.M., and J.K. Akina. (A. Akana, translator). 1922. Hawaiian Herbs of Medicinal Value. Honolulu: Territorial Board of Health.

Kanehira, R. 1933. Flora Micronesica. Tokyo: Southseas Bureau under the Japanese Mandate.

Kartasubrata, J., S.S. Tjitrosomo, and R.C. Umaly, eds. 1990. Symposium on Agroforestry Systems and Technologies, Bogor, Indonesia, 19-21 September, 1989. Biotrop Special Publication No. 39, Southeast Asian Regional Center for Tropical Biotogy (SEAMEOBIOTROP), Bogor, Indonesia.

Kayser, P.A. 1934. "Des pandanus auf Nauru." Anthropos 29: 775-791.

Keith-Reid, R. 1982. "The eating catastrophe." Islands Business 8(9): 12-16.

Keju, D., and G. Johnson. 1982. "Kwajalein: Home on the 'range'." Pacific Magazine 7(6): 2328.

Kesavan, V. 1979. Backyard gardening in urban centres. Paper presented at the Waigani Seminar, 24-28 September 1979, Port Moresby.

Khoo, K.M. 1976. "Optimal utilisation and management of fisheries resources." In: Proceedings of a Seminar on the Development of the Fisheries Sector in Malaysia. Journal of the Malaysian Economics Association 13: 40-50.

King, B. 1987. "Fuelwood planting in urban Papua New Guinea." Ples: An Environmental Journal for the South Pacific Region 3: 1-9.

King, K.F.S. 1989. "The history of agroforestry." In: P.K.R. Nair, ea., pp. 3-11.

King, K.F.S., and M.T. Chandler. 1978. The Wasted Lands: The Programme of Work of ICRAF. Nairobi: International Council for Research in Agroforestry (ICRAF).

King, W.B. 1981. "The world's rarest birds." International Wildlife 11(5): 12-19.

Kirch, P.V. 1982. "Ecology and adaptation of Polynesian agricultural systems." Archaeology in Oceania 17(1): 1-6.

-. 1984. The Evolution of Polynesian Chiefdoms. Cambridge: Cambridge University Press.

Kirch, P.V., and D. Yen. 1982. Tikopia: The Prehistory and Ecology of a Polynesian Outlier. Bulletin 238, Bernice P. Bishop Museum, Honolulu.

Kirkpatrick, J.B., and D.C. Hassall. 1981. "Vegetation of the Sigatoka sand dunes, Fiji." New Zealand Journal of Botany 19: 285-297.

Klee, G.A. 1980a. "Traditional wisdom and the modern resource manager." In: G.A. Klee, ed. World Systems of Traditional Resource Management. London: Edward Arnold, pp. 283-285.

-. 1980b. "Oceania." In: G.A. Klee, ed. World Systems of Traditional Resource Management. London: Edward Arnold, pp. 245-281.

Knudtson, P., and D. Suzuki. 1992. Wisdom of the Elders. North Sydney: Allen and Unwin.

Koch, G. (G. Slatter, translator). 1983. The Material Culture of Tuvalu. Institute of Pacific Studies, University of the South Pacific, Suva. Originally published in German in 1961 by Museum fur Volkerkunde, Berlin.

-. 1986. The Material Culture of Kiribati. Institute of Pacific Studies, University of the South Pacific, Suva. Originally published in German in 1965 by Museum fur Völkerkunde, Berlin.

Kooijman, S. 1972. Tapa in Polynesia. Bulletin 234, Bernice P. Bishop Museum, Honolulu.

Kow, F. 1982. Report on Nutrition Survey - Wasu 1980: Comparative Nutritional Status Study. Department of Chemical Studies, PNG University of Technology, Lae.

Kraemer, J.H. 1944. Native Wood for Construction Purposes in the Western Pacific Region. Bureau of Yards and Docks, Department of the Navy, Washington, D.C.

Kramer, A. 1929. "Palau." In: G. Thilenius, ed. Ergebnisse der Sddsee [Expedition, 1908-1910:11. Ethnographie: B. Mikronesien. Vol. 3. Friederichsen, de Gruyter, Mangurg.

Krauss, B. 1974. Ethnoootany of Hawaii. Honolulu: University of Hawaii Press.

Kunzel, W. 1989. Agroforestry in Tonga: A Traditional Source for Development of Sustaninable Farming Systems. Occasional Paper 12, South Pacific Smallholder Project, University of New England, Armidale, N.S.W., Australia.

-. 1992. Fiji-German Forestry Project, Suva, Fiji. Personal communication.

Lal, P.N., A. Swamy, and P. Singh. 1983. "Mangroves and secondary productivity: Fishes associated with mangroves in Wairiki Creek, Fiji." In: P.N. Lal, ed. Proceedings of an Interdependent Workshop, 24 February 1983, Suva, Fiji. Technical Report 5, Fisheries Division, Ministry of Agriculture and Fisheries, Suva.

Lamberson, J.O. 1982. A Guide to Terrestrial Plants of Enewetak Atoll. Honolulu: Pacific Science Information Center, Bernice P. Bishop Museum.

Lambert, J.N. 1979. Population Growth, Nutrition and Food Supplies: Populations of Papua New Guinea. ESCAP Country Monograph, United Nations Economic and Social Council, New York.

Lambert, M. 1982. "The cultivation of 'taro' (Cyrtosperma chamissonis) Schott in Kiribati." In: M. Lambert, ed. Regional Technical Meeting on Atoll Cultivation, Papeete, Tahiti, French Polynesia, 14-19 April 1980: Collected Papers. Technical Paper No. 180, South Pacific Commission, Noumea, pp. 163-165.

Lebot, V., and P. Cabalion. 1986. Les kavas de Vanuatu. Travaux et documents No. 205, Editions de l'ORSTOM, Paris.

Lessa, W.A. 1977. "Traditional uses of the vascular plants of Ulithi atoll, with comparative notes." Micronesia 13(2): 129-190.

Leung, W.W., R.R. Butrum, and F.H. Chang. 1972. Food Composition Table for Use in East Asia, Part 1. FAO, Rome, and U.S. Department of Health, Education, and Welfare, Bethesda, Maryland.

Lundgren, B.O. 1987. "Institutional aspects of agroforestry research and development." In: H.A.teppler, and P.K.R. Nair, eds., pp. 43-51.

Luomala, K. 1953. Ethnobotany of the Gilbert Islands. Bulletin 213, Bernice P. Bishop Museum, Honolulu.

McArthur, N. 1967. Island Populations of the Pacific. Canberra: Australian National University Press.

McCoy, P.C. 1976. Easter Island Settlement Patterns in the Late Prehistoric and Protohistoric Periods. Bulletin 5, Easter Island Committee, International Fund for Monuments, New York.

McCuddin, C.R. 1974. Samoan Medicinal Plants and Their Usage. Department of Medicinal Services, Government of American Samoa, Pagopago.

McCutcheon, M.S. 1981. Resource exploitation and the tenure of land and sea in Palau. Ph.D. thesis, Anthropology, University of Arizona, Tucson.

MacDicken, K.G., and N.T. Vergara, eds. 1990. Agroforestry: Classification and Management. New York: John Wiley & Sons.

McDonald, M.A. 1978. Ma Lei: The Leis of Hawaii. Honolulu: Topgallant Publishing.

Macfarlane, D.C. 1980. "Pasture trials at Poitete and implications for development." In: Cattle under Trees Seminar, 8th August. Honiara: Ministry of Agriculture and Lands.

Macfarlane, D.C., and P.C. Whiteman. 1983. Grazing under Eucalyptus deglupta Reafforestation on Kolombangara Islands, Solomon Islands. Final Report (19771982). Ministry of Home Affairs and National Development, Honiara, and Australian Development Assistance Bureau, Canberra.

McGee, T.G. 1975. Food Dependency in the Pacific: A Preliminary Statement. Development Studies Centre Working Paper No. 2, Research School of Pacific Studies, Australian National University, Canberra.

McGregor, A. 1984. A Programme for the Development of the Batiri Citrus Project. School of Social and Economic Development, University of the South Pacific, Suva.

McKee, H.S. 1957. Some Food Problems in the Pacific Islands. Technical Bulletin No. 106, South Pacific Commission, Noumea.

-. 1985. Les plantes introduites et cultivées de la Nouvelle-Caledonie: Supplement a la flore de la Nouvelle-Calédonie et dépendences. (Volume horsseries) Laboratoire de Phanéogamie, Musée National d'Histoire Naturelle, Paris.

MacLean, C.D., T.G. Cole, C.D. Whitesell, M.C. Falanruw, and A.H. Ambacher. 1986. Vegetation Survey of Pohnpei, Federated States of Micronesia. Resource Bulletin PSW-18, Pacific Southwest Forest and Range Experiment Station, U.S. Department of Agriculture.

Maenu'u, L.P. 1979. An indicative list of Solomon Islands medicinal plants. Pacific Collection, University of the South Pacific, Suva. Unpublished manuscript.

MAP (Ministry of Agriculture and Fisheries) (Fiji). 1973. Cattle under Coconuts. MAF, Suva.

-. 1974. Commodity Study Paper: Cocoa. Working paper, MAF, Suva. Mimeo. MAFF (Ministry of Agriculture, Fisheries, and Forests) (Tonga). 1979. Report of the Ministry of Agriculture, Fisheries, and Forests for the Year 1978. Nuku'alofa: Government Printer.

-. 1981. Report of the Ministry of Agriculture, Fisheries, and Forests for the Year 1980. Nuku'alofa: Government Printer.

-. 1985. Report of the Ministry of Agriculture, Fisheries, and Forests for the Year 1984. Nuku'alofa: Government Printer.

Mangenot, G. 1965. "The effect of man on the plant world." In: F.R. Fosberg, ed. Man's Place in the Island Ecosystem: A Symposium. Honolulu: B.P. Bishop Museum, pp. 117-132.

Manner, H.I. 1980. "Buma subsistence agriculture: Patterns, processes and problems on Malaita." O'o: A Journal of Solomon Islands Studies 1(1&2): 101-112.

-. 1981. "Ecological succession in new and old swiddens of montane Papua New Guinea." Human Ecology 9(3): 359-377.

-. 1983. "Institutional agroforestry in the Pacific: A preliminary survey." In: A. Schirmer, ed. The Role of Agroforestry in the Pacific. Hamburg: German Foundation for International Development, pp. 35-43.

-. 1987. "Atoll flora and vegetation." Alafua Agricultural Bulletin 12(3): 6780.

Manner, H.I., R.R. Thaman, and D.H. Hassall. 1984. 'iPhosphate-mining induced vegetation changes on Nauru Island." Ecology 65(5): 1454-1465.

-. 1985. "Plant succession after phosphate mining on Nauru." Australian Geographer 16: 185-195.

Maragos, J.E., A. Soegiarto, E.D. Gomez, M.A. Dow. 1983. "Development planning for tropical coastal ecosystems." In: R.A. Carpenter, ed. Natural Systems for Development: What Planners Need to Know. Macmillan, New York, and Collier Macmillan, London, pp. 229-298.

Marshall, M., and F.R. Fosberg. 1975. "The natural history of Namoluk Atoll, Eastern Caroline Islands: With identifications of vascular flora." Atoll Research Bulletin 189: 1-65.

Martin, F.W., and R.M. Ruberte. 1975. Edible Leaves of the Tropics. Mayaguez, Puerto Rico: Mayaguez Institute of Tropical Agriculture.

-. 1978. Survival and Subsistence in the Tropics. Mayaguez, Puerto Rico: Antillian College Press.

Massal, E., and J. Barrau. 1956. Food Plants of the South Sea Islands. Technical Paper No. 94, South Pacific Commission, Noumea.

Maude, A. 1965. Population, land and livelihood in Tonga. Unpublished Ph.D. thesis, Department of Geography, Australian National University, Canberra.

Maydell, H.J. von. 1987. The Potential Role of Agroforestry in Forestry Extension in Fiji. Technical Report No. 3, Fiji-German Forestry Project (Deutsche Gesellschaft fur technische Zusammenarbeit [GTZ]), Suva.

Merlin, M.D. 1985. "Woody vegetation in the upland region of Rarotonga, Cook Islands." Pacific Science 39(1): 81-99.

Merrill, E.D. 1943. Emergency Food Plants and Poisonous Plants of the Pacific. Technical Monograph 10-42O, United States Printing Office, Washington, D.C.

-. 1945. Plant Life of the Pacific World. New York: Macmillan.

Mescarm, G. 1989. Pentecost: An Island in Vanuatu. Vanuatu Extension Centre, Vila, and Institute of Pacific Studies, University of the South Pacific, Suva.

Metraux, A. 1940. Ethnology of Easter Island. Bulletin 160, Bernice P. Bishop Museum, Honolulu.

Miller, C.D. 1927. Food Values of Breadfruit, Taro Leaves, Coconut and Sugar Cane. Bulletin 64, Bernice P. Bishop Museum, Honolulu.

Miller, C.D., K. Bazore, and M. Bartow. 1965. Fruits of Hawaii: Description, Nutritive Value, and Recipes. Honolulu: University of Hawaii Press.

Montagnini, F. 1990. "Ecology applied to agroforestry in the humid tropics." In: R. Goodland, ed. Race to Save the Tropics. Washington, D.C.: Island Press, pp. 49-58.

Moul, E.T. 1957. "Preliminary report on the flora of Onotoa Atoll, Gilbert Islands." Atoll Research Bulletin 57.

Mowbray, D. 1984. A Review of Pesticide Use in the South Pacific with a Proposal for Establishing a Pesticide Monitoring and Evaluation Network: Preliminary Draft. South Pacific Regional Environmental Programme Pesticide Project, University of Papua New Guinea, Waigani.

Mueller-Dombois, D. 1975. "Some aspects of island ecosystem analysis." In: F.B. Colley, and E. Medina, eds. Coastal Ecological Systems: Trends in Aquatic and Terrestrial Research, Vol. 11. Ecological studies series: Analysis and synthesis. New York: Springer Verlag, pp. 353-366.

Murai, J., F. Pen, and C.D. Miller. 1958. Some Tropical South Pacific Island Foods. Honolulu: University of Hawaii Press.

Nair, P.K.R. 1984. Fruit Trees in Tropical Agroforestry Systems. Working Paper, Environment and Policy Institute, East-West Center, Honolulu.

-. 1987. "Agroforestry systems inventory." Agroforestry Systems 5: 301-308.

-. 1989. "Agroforestry defined." In: P.K.R. Nair, ea., pp. 13-18.

-. ed. 1989. Agroforestry Systems in the Tropics. Dordrecht: Kluwer (in cooperation with ICRAF).

NAS (National Academy of Sciences). 1975. Plant Studies in the Peoples Republic of China: A Trip Report of the American Plant Studies Delegation. National Academy of Sciences, Washington, D.C.

-. 1977. Leucaena: Promising Forage and Tree Crop for the Tropics. Washington, D.C.: National Academy of Sciences.

National Planning Office (Papua New Guinea). 1983. The National Expenditure Plan 1984-1987. Port Moresby: Papua New Guinea Government Printer.

National Planning Office (Vanuatu). 1982. Republic of Vanuatu First National Development Plan 1982-1986. Port Vila: National Planning Office.

National Wildlife. 1982. 20(2): 29-36.

Naveh, Z., and A.S. Lieberman. 1984. Landscape Ecology: Theory and Application. New York: Springer-Verlag.

Neal, M.C. 1965. In Gardens of Hawaii. Special Publication 50, Bernice P. Bishop Museum, Honolulu.

Neil, P.E. 1984. Cordia alliodora in Vanuatu: A Position Paper. Forest Research Report 1/84, Vanuatu Forest Service, Port Vila.

-. 1985a. Gmelina arborea in Vanuatu. Forest Research Report 3/85, Vanuatu Forest Service, Port Vila.

-. 1985b. Agathis macrophylla in Vanuatu: A Position Paper (Including a Proposal for a Kauri Reserve). Forest Research Report 1/85, Vanuatu Forest Service, Port Vila.

-. 1986a. Pinus caribaea var. hondurensis in Vanuatu. Forest Research Report 1/86, Vanuatu Forest Service, Port Vila.

-. 1986b. Swietenia macrophylla (Mahogany) in Vanuatu. Forest Research Report 4/86, Vanuatu Forest Service, Port Vila.

-. 1986c. The Performance of Eucalypts Tested in Vanuatu. Forest Research Report 6/86, Vanuatu Forest Service, Port Vila.

-. 1986d. Early Results from Acacia Species Research. Forest Research Report 7/86, Vanuatu Forest Service, Port Vila.

-. 1986e. Sandalwood in Vanuatu. Forest Research Report 5/86, Vanuatu Forest Service, Port Vila.

-. 1987a. Preliminary Report on Agathis macrophylla Tree Improvement Research. Forest Research Report 5/87, Vanuatu Forest Service, Port Vila.

-. 1987b. Notes on Some Indigenous Tree Species. Forest Research Report 7/ 87, Vanuatu Forest Service, Port Vila.

-. 1987c. Fuelwood Trials. Forest Research Report 2/87, Vanuatu Forest Service, Port Vila.

NFNC (National Food and Nutrition Committee). 1979. Food Balance Studies Report. Suva: Ministry of Health.

Niering, W.A. 1956. "Bioecology of Kapingamarangi Atoll, Caroline Islands: Terrestrial aspects." Atoll Research Bulletin 49: 1-32.

Nunn, P.D. 1989. Effects of global warming on the South Pacific islands, 1900-2100. Draft of paper prepared for the Environment and Policy Institute (EAPI) Workshop on "Responding to the Threat of Global Warming: Options for the Pacific and Asia," East-West Center, Honolulu, 21-27 June.

Oedekoven, K.H. 1962. "Saving our vanishing forests." Unasylva 16: 55-58.

Okabe, M. 1940. "Investigation of the medicinal plants found on the Palau islands, their virtues and popular remedies." Bulletin of Tropical Indusry, Palau 5 Tokyo (Translated by Hisayoshi Takeda, December 1952).

Oliver, D.L. 1955. A Solomon Island Society: Kinship and Leadership among the Sinai of Bougainville. Cambridge, MA: Harvard University Press.

-. 1974. Ancient Tahitian Society, Vol. 1: Ethnography. Canberra: Australian National University Press.

Oomen, H.A.P.C., and G.J.H. Grubben. 1978. Tropical Leaf Vegetables in Human Nutrition. Communication 69, Department of Agricultural Research, Koninklijk Instituut voor de Tropen, Amsterdam.

OTA (Office of Technology Assessment). 1987. Integrated Renewable Resource Management for U. S. Insular Areas. Washington, D. C.: Congress of the United Sates.

Overton, J. 1989. Land and Differentiation in Rural Fiji. Pacific Research Monograph No. 19, National Centre for Development Studies, Australian National University, Canberra.

Overy, R., 1. Polunin, and D.W.G. Wimblett. 1982. Some Plants of Kiribati: An Illustrated List. Tarawa: National Library and Archives.

Paijmans, K. 1976. "Vegetation." In: K. Paijmans, ed. New Cuinea Vegetation. Amsterdam: Elsevier Scientific Publishing Co., pp. 23-105.

Papua New Guinea Housing Commission. 1975. Urban programme in Papua New Guinea. Information paper for the 5th Congress of the Eastern Organization for Planning and Housing, Manila.

Pargeter, K.A., R. Taylor, H. King, and P. Zimmet. 1984. Kiribati: A Dietary Survey. Noumea: South Pacific Commission.

Parham, B.E.V. 1971. "The vegetation of the Tokelau Islands with special reference to the plants of Nukunonu Atoll." New Zealand lournal of Botany 9(4): 576-609.

-. 1972. Plants of Samoa. Information Series No. 85, Department of Scientific and Industrial Research, Wellington.

Parham, J.W. 1972. Plants of the Fiji Islands (revised edition). Suva: The Government Printer.

Parkinson, S.V. 1973. "Some observations on the cause of malnutrition in Pacific urban communities." In: J. Harre, ed. Living in Town: Problems and Priorities in Urban Planning in the South Pacific. Suva: School of Social and Economic Development, University of the South Pacific, pp. 85-91.

-. 1982. "Nutrition in the South Pacific: Past and present." Journal of Food and Nutrition 39(3): 121-125.

-. 1984a. Preservation and Preparation of Root Crops and Some Other Traditional Foods in the South Pacific. Field Document 1, UNDP/FAO-SPC Project on Strengthening Plant Protection and Root Crops Development in the South Pacific (RAS83001). Food and Agricultural Organization of the United Nations in association with the South Pacific Commission, Suva.

-. 1984b. "The contribution of aroids in the nutrition of people in the South Pacific." In: S. Chandra, ed. Edible Aroids. Oxford: Clarendon Press, pp. 215224.

Partridge, I.M. 1977. Dry Zone Pasture Work. Annual Research Report, Ministry of Agriculture, Suva.

Percival, M., and J.S. Wormersley. 1975. Floristics and Ecology of the Mangrove Vegetation of Papua New Guinea. Botany Bulletin No. 8, Papua New Guinea National Herbarium, Department of Forests, Lae.

Perry, L.M., and J. Metzger. 1980. Medicinal Plants of East and Southeast Asia: Attributed Properties and Uses. Cambridge, Massachusetts: MIT Press.

Petard, P. 1986. Quelques plantes utiles de Polynesie Francaise et ruau Tahiti. Papeete: Edition Haere Po No Tahiti.

Pickering, C. 1876. "The geographical distribution of animals and plants, Part II: Plants in their wild state." U.S. Exploring Expedition, Botany (2). Salem.

Pollock, N.J. 1970. Breadfruit and breadwinning on Namu Atoll, Marshall Islands. Unpublished Ph.D. thesis, Anthropology, University of Hawaii, Honolulu.

-. 1974. "Breadfruit or rice: Dietary choice on a Micronesian atoll." Ecology of Food and Nutrition 3: 107-115.

Polunin, I. 1979. A Study of Local Medicinal Plants, Tarawa, Kiribati. Assignment Report, Regional Office for the Western Pacific, World Health Organization, Suva.

Pompey, S.L. 1974. The Aitutaki Banana Scheme. Papers on the South Pacific Islands No. 98, Harresbury, Oregon.

Powell, J.M. 1976a. "Ethnobotany." In: K. Paijmans, ed. New Guinea Vegetation. Amsterdam: Elsevier Scientific Publishing Co., pp. 106-183.

-. 1976b. "Pollen, plant communities, and prehistory in the Solomon Islands." In: R.C. Green, and M.M. Cresswell, eds. Southeast Solomon Islands Cultural History: A Preliminary Survey. Bulletin 11, Royal Society of New Zealand, Wellington, pp. 75-106.

Pulea, M. 1984. "Environmental legislation in the Pacific region." Ambio 13(5-6): 369-371.

Purseglove, J. W. 1975a. Tropical Crops: Vol. 1. Monocotyledons. London: Longmans.

-. 1975b. Tropical Crops: Vol. 11. Dicotyledons. London: Longmans.

Pyle, R.M. 1981. "Butterflies: Now you see them. . ." International Wildlife 11(1): 4-11.

Quartermain, A.R. 1980. "Livestock." In: R.G. Ward, and A. Proctor, eds. South Pacific Agriculture Choices and Constraints: South Pacific Agricultural Survey 1979. Canberra: Asian Development Bank in Association with Australian National University Press, pp. 261-292.

Rageau, J. 1973. Les plantes médicinales de la Nouvelle-Calédonie. Travaux et Documents de l'O.R.S.T.O.M. No. 23, Editions de l'office de la Recherche Scientifique et Technique Outre-mer, Paris.

Rainbow, R.G., and S. Teteha. 1983. Shifting Cultivation in Solomon Islands. Honiara: Curriculum Development Centre.

Raintree, J.B., ed. 1987a. D & D User's Manual: An Introduction to Agroforestry Diagnosis and Design. Nairobi: International Council for Research in Agroforestry.

-. 1987b. Land, Trees and Tenure: Proceedings of an International Workshop on Tenure Issues in Agroforestry. International Council for Research in Agroforestry, NairobilLand Tenure Center, University of Wisconsin, Madison.

-. 1990. "Theory and practice of agroforestry diagnosis and design." In: K.G. MacDicken, and N.T. Vergara, eds., pp. 58-97.

Raintree, J.B., and K. Warner. 1986. "Agroforestry pathways for the intensification of shifting cultivation." Agroforestry Systems 4: 39-54.

Rambo, A.T., with L. Hamilton. 1991. "Social trends affecting natural resource management in upland areas of Asia and the Pacific." In: S. Naya, and S. Browne, eds. Development Challenges in Asia and the Pacific in the 1990s. Honolulu: EastWest Center, pp. 111-126.

Ravuvu, A. 1983. Vaka i Taukei: The Fijian Way of Life. Institute of Pacific Studies, University of the South Pacific.

Raynor, W.C., and J.H. Fownes. 1991a. "Indigenous agroforestry of Pohnpei: 1. Plant species and cultivars." Agroforestry Systems 16: 139-157.

-. 1991b. "Indigenous agroforestry of Pohnpei: 2. Spatial and successional vegetation patterns." Agroforestry Systems 16: 159-165.

Richards, P. 1985. Indigenous AgriculturaJ Revolution: Ecology and Food Production in West Africa. Boulder, CO: Westview Press.

Richards, P.W. 1979. The Tropical Rainforest: An Ecological Study. London: Cambridge University Press.

Richardson, S.D. 1981. Forestry and the environment in the South Pacific. Topic review paper, South Pacific Regional Environment Programme (SPREP), South Pacific Commission, Noumea.

-. 1983. "Agroforestry in the Pacific islands." In: A. Schirmer, ed. The Role of Agroforestry in the Pacific. Hamburg: German Foundation for International Development, pp. 53-70.

Rizer, J.P., J. Lin, M. Waqavonovono, S. Saumatua, and A.G. Marjoram. 1982. The Potential Impact of a Namosi Copper Mine: A Case Study of Assimilation Planning. Centre for Applied Studies in Development, University of the South Pacific, Suva.

Rocheleau, D., F. Weber, and A. Field-Juma. 1989. Agroforestry in Dryland Africa. Nairobi: International Council for Research in Agroforestry.

Rock, J.F. 1974. The Indigenous Trees of the Hawaiian Islands. Pacific Tropical

Botanical Garden, Lawai, Kauai, Hawaii, and Charles E. Tuttle Co., Rutland, Vermont.

Rody, N. 1982. "Food for all in Micronesia." Pacific Magazine 7(5): 27-32.

Rogers, B. 1980. The Domestication of Women: Discrimination in Developing Societies. London: Tavistock Publications.

References (S-Z)

Sachet, M.-H. 1983. "Botanique de File de Tupai, Iles de la Societe." Atoll Research Bulletin 276: 1-35.

Sachet, M.-H., and F.R. Fosberg. 1955. Island Bibliographies. Washington, D.C.: Smithsonian Institution.

-. 1971. Island Bibliographies Supplement. Washington, D.C.: Smithsonian Institution.

St. John, H. 1948. "Report on the flora of Pingelap Atoll, Caroline Islands, Micronesia, and observations on the vocabulary of the native inhabitants: Pacific plant studies 12." Pacific Science 2: 97-113.

-. 1973. List and Summary of the Flowering Plants in the Hawaiian Islands. Memoir No. 1, Pacific Tropical Botanical Garden, Lawai, Kauai, Hawaii.

St. John, H., and W.R. Philipson. 1962. "An account of the flora of Henderson Island, south Pacific Ocean." Transactions of the Royal Society of New Zealand. Botany 1(14): 175194.

St. John, H., and A.C. Smith. 1971. "The vascular plants of the Horne and Wallis Islands." Pacific Science 25: 313-348.

Sarasin, F., and J. Roux. 1921. Nova Caledonia forschungen in Neu-Caledonien und auf den Loyalty-lnseln/Recherches scientifiques en Nouvelle-Calédonie et aux ales Loyalty. Berlin and Wiesbaden: C.W. Kreidel's Verlag.

Sarlin, P. 1954. Bois et Fôrets de la Nouvelle-Caledonie. Nogent-sur-Marne, France: Centre Technique Forestier Tropical.

Saner, C.O. 1952. Agricultural Origins and Dispersals. New York: American Geographical Society.

Schieffelin, E.L. 1975. "Felling the trees on top of the crop: European contact and the subsistence ecology of the Great Papuan Plateau." Oceania 46(1): 25-39.

Schirmer, A., ed. 1983. The Role of Agroforestry in the Pacific. Hamburg: German Foundation for International Development.

Schmid, M. 1973. Flore des Nouvelles Hebrides. Noumea: Office de la Recherche Scientifique et Technique d'Outre-mer.

-. 1981. Fleurs et plantes de Nouvelle-Calédonie. Papeete: Editions du Pacifique.

Schumacher, E.F. 1973. Small is Beautiful: A Study of Economics as if People Mattered. London: Abacus (Sphere Books).

Sedjo, R.A. 1989. "Forests: A tool to moderate global warming?" Environment 319(1): 14-20.

Seemann, B. 1873. Flora Vitiensis: A Description of the Plants of the Viti or Fiji Islands, with an Account of Their History, Uses and Properties. London: L. Reeve.

Setchell, W.A. 1924. American Samoa: Part 1. Vegetation of Tutuila Island; Part 11. Ethnobotany of the Samoans; Part 111. Vegetation of Rose Atoll. Department of Marine Biology, Carnegie Institution, Washington, D.C.

Sevele, F.V. 1981. An Economic Overview of the South Pacific Region. Noumea: South Pacific Commission.

Singh, A., and S. Siwatibau. 1980. Medicinal Plants in Fiji and Other Pacific Islands

(Preliminary Results). Sehool of Natural Resources, University of the South Pacific, Suva.

Siwatibau, S. 1981. Rual Energy in Fiji: A Survey of Domestic Rural Energy Use and Potential. IDRC-157e, International Development Research Centre, Ottawa.

Small, C.A. 1972. Atoll Agriculture in the Gilbert and Ellice Islands. Tarawa: Department of Agriculture.

Smith, A.C. 1979. Flora Vitiensis Nova: A New Flora of Fiji (Spermatophytes Only): Vol. 1. Pacific Tropical Botanical Garden, Lawai, Kaaai, Hawaii.

-. 1981. Flora Vitiensis Nova: A New Flora of Fiji (Spermatophytes Only): Vol. 2. Pacific Tropical Botanical Garden, Lawai, Kauai, Hawaii.

-. 1985. Flora Vitiensis Nova: A New Flora of Fiji (Spermatophytes Only): Vol. 3. Pacific Tropical Botanical Garden, Lawai, Kauai, Hawaii.

-. 1988. Flora Vitiensis Nova: A New Flora of Fiji (Spermatophytes Only): Vol. 4. Pacific Tropical Botanical Garden, Lawai, Kauai, Hawaii.

-. 1991. Flora Vitiensis Nova: A New Flora of Fiji (Spermatophytes Only): Vol. 5. Pacific Tropical Botanical Garden, Lawai, Kauai, Hawaii.

Smith, K.R. 1985. Biomass Fuels and Development. East-West Resource Systems Institute, East-West Center, Honolulu.

Soemarwoto, O.M., I. Soemarwoto, S. Karyono, W.M. Tadiredia, and A. Raman. 1985. "The Javanese home garden as an integrated agro-ecosystem." Food and Nutrition Bulletin 7(3): 44-47.

Sofer, M. 1985. "Yaqona and the peripheral economy." Pacific Viewpoint 26(2): 415- 436.

Solomon Islands Government. 1979. 1976 Census of Population: Basis Tabulations Nos I to 8 Inclusive. Statistical Bulletin No. 19/79, Statistics Office, Honiara.

-. 1982. 1982 Statistical Yearbook. Bulletin No. 33/82, Statistics Office, Honiara.

Sommers, P. 1990. "Advancing Pacific Island food gardening systems: Some observations and suggestions." In: K. Landauer, and M. Brazil, eds. Tropical Home Gardens. Tokyo: United Nations University Press, pp. 193-202.

Soucie, E.A. 1983. Atoll Agriculture for Secondary Schools: Soils and Major Agricultural Crops of Micronesia. Ponape: Ponape Agriculture and Trade Sehool.

SPEC (South Pacific Bureau for Economic Cooperation). 1982. Pacific Energy Programme Review Meeting, SPEC Headquarters, Suva 23-24 September 1982: Agreed Record. SPEC (82) 14, South Pacific Bureau for Economic Cooperation, Suva.

Spoehr, A. 1949. "Majuro: A village in the Marshall Islands." fieldiana: Anthropology 39: 1266.

SPREP (South Pacific Regional Environment Programme). 1982a. Draft Convention for the Protection and Development of the Natural Resources and Environment of the South Pacific Region. Noumea: South Pacific Commission.

-. 1982b. Report on the Conference on the Human Environment in the South Pacif c, Rarotonga, Cook Islands, 8-11 March 1982. Noumea: South Pacific Commission.

Spriggs, M.J.T. 1980. "Taro irrigation in Oceania: A call for more research." South Pacific Bulletin 30(1): 15-18.

-. 1981a. Bombs and Butter: The Revival of Ancient Irrigation Techniques for a Market Economy: A Pacific Example. Papers in Prehistory No. 2, Department of Prehistory, Australian National University, Canberra.

-. 1981b. Vegetable kingdoms: Taro irrigation and Pacific pre-history. Unpublished Ph.D. thesis, Australian National University, Canberra.

Sproat, M.N. 1968. A Guide to Subsistence Agriculture in Micronesia. Agricultural Extension Bulletin No. 9, Trust Territory of the Pacific Islands, Division of Agriculture, TTPI Publications Office, Saipan.

Standal, B.R., J.M. Street, and R.M. Warner. 1974. "Tasty, protein rich, and easy to grow: the new edible 'sunset' hibiscus." Hawaii Farm Science Second Quarter: 2-3.

Stemmermann, L. 1981. A Guide to Pacific Wetland Plants. Honolulu: U.S. Army Corps of Engineers.

Steppler, H.A. 1987. "ICRAF and the decade of agroforestry development." In: H.A. Steppler, and P.K.R. Nair, eds., pp. 13-21.

Steppler, H.A., and P.K.R. Nair, eds. 1987. Agroforestry: A Decade of Development. Nairobi: ICRAF.

Sterly, J. 1970. Heilpflanzen der Einwohner Melanesiens: Beitrage zur Ethnobotanik des sudwestlichen Pazifik. Hamburg: Arbeitsstelle fur Ethnomedizin.

Stone, B.C. 1967. "Studies of Malesian Pandanaceae 1." Gardens Bulletin, Singapore 22(2): 231-257.

-. 1970. "The flora of Guam." Micronesica 6 (complete): 1-659.

Streimann, H. 1983. The Plants of the Upper Watut Watershed of Papua New Guinea. Sydney: National Botanic Gardens.

Suggs, R.C. 1961. Archaeology of Nuku Hiva, Marquesas Islands, French Polynesia. Anthropological Papers of the American Museum of Natural History 49, American Museum of Natural History, New York.

Sugiura, K. 1942. "Taro culture of the Palauans." Geographic Research 1(8): 10171035. (Translated 1956 from Japanese by Office of the Engineer, US Army Air Forces, Far East, Tokyo.)

Sykes, W.R. 1970. Contribufions to the Flora of Niue. Bulletin 200, New Zealand Department of Scientific and Industrial Research, Wellington.

-. 1976. Vegetation reports of the southern Cook Islands. Department of Scientific and Industrial Research, Christchurch. Unpublished typescript.

-. 1981. "The vegetation of Late, Tonga." Allertonia 2(6): 323-353.

Tanaka, T. (S. Nakao, ea.) 1976. Tanaka's Cyclopedia of Edible Plants of the World. Tokyo: Keigaku Publishing Co.

Tarepe, T., and R.M. Bourke. 1982. "Fruit crops in the Papua New Guinea Highlands." In: R.M. Bourke, and V. Kesavan, eds. Part 1, pp. 86-100.

Thaman, R.R. 1975. "Tongan agricultural land use: A study of plant resources." In: E. Stokes, ed. Proceedings of the International Geographical Union Regional Conference and Eighth New Zealand Geography Conference, Palmerston North, December 1974. Wellington: New Zealand Geographical Society (Inc.), pp. 153-160.

-. 1976. The Tongan Agricultural System: With Special Emphasis on Plant Assemblages. Suva: University of the South Pacific. (Published version of 1975, Ph.D. thesis, University of California, Los Angeles).

-. 1976t77. "Plant resources of the Suva Municipal Market, Fiji." Ethnomedicine IV(1/2): 2362.

-. 1977a. "Urban gardening in Papua New Guinea and Fiji." In: J.H. Winslow, ed. The Melanesian Environment. Canberra: Australian National University Press, pp. 146-168.

-. 1977b. "Urban root crop production in the southwest Pacific." In: M. Lambert, ed. Collected Papers: Regional Meeting on Production of Root Crops, 24-29 October 1975, Suva, Fiji. Technical Paper No. 174, South Pacific Commission, Noumea, pp. 73-82.

-. 1978. "Cooperative yam gardens: An adaptation of a traditional agricultural system to serve the needs of the developing Tongan market economy." In: E.K. Fisk, ed. The Adaptation of Traditional Agriculture: Socioeconomic Problems of Urbanization . Monograph No. 11, Development Studies Centre, Australian National University, Canberra, pp. 116-128.

-. 1979. "Food scarcity, food dependency and nutritional deterioration in small Pacific island communities." In: W. Moran, P. Hosking, and G. Aitken, eds. Proceedings of she Tenth NZ Geographical Conference and 49th ANZAAS Congress. New Zealand Geographical Society Conference Series No. 10, Auckland, pp. 191-197.

-. 1982a. "Hurricane Isaac and Tonga: A natural or cultural disaster?" Review (Journal of the School of Social and Economic Development, University of the South Pacific, Suva, Fiji) 3(8): 22-34.

-. 1982b. "The foods that came first." Alafua Agricultural Bulletin 7(3): 105116.

-. 1982c. "Deterioration of traditional food systems, increasing malnutrition and food dependency in the Pacific islands." Journal of Food and Nutrition 39(3): 109-121.

-. 1983a. "Food for urbanizing Polynesian peoples." Proceedings of the Nutrition Society of New Zealand 8: 26-37.

-. 1983b. "Food and national development in the Pacific islands: An introduction." In: R.R. Thaman, and W.C. Clarke, eds. Food and National Development in the Pacific Islands. Suva: University of the South Pacific, pp. 1-16.

-. 1984a. "The poisoning of paradise: Pesticides, environmental pollution, and increasing dependency in the Pacific islands." South Pacific Forum 1(2): 165-200.

-. 1984b. "Intensification of edible aroid cultivation in the Pacific islands." In: S. Chandra, ed. Edible Aroids. London: Clarendon Press, pp. 102-122.

-. 1984c. "Urban gardening in Fiji: A direct road to development and independence." Transactions of the Fiji Society for the Years 1978 to 1980 14: 1-28.

-. 1984d. The Firewood Crisis and Smaliholder Fuelwood Systems on Tongata-pu Island, Tonga: Present Systems and Development Potential. PEDP Report: Tonga 85-1, United Nations Pacific Energy Development Programme (UNPEDP), Suva.

-. 1984e. Trees, Tongans, and Reafforestation in Tonga 1965-1984. Occasional Paper No. 4, Institute of Rural Development, University of the South Pacific, Atele, Tonga.

-. 1985a. Mixed homegardening in the Pacific islands: Present status and future prospects. Paper presented at the First International Workshop on Tropical Homegardens, Bandung, 2-9 December.

-. 1985b. "Pacific islands health and nutrition: Trends and areas for action." In: Development and Change: Issue Papers. Pacific Islands Conference, Rarotonga, Cook Islands, August 1985. Pacific Islands Development Programme (PIDP), East-West Center, Honolulu, pp. III- A. 1-27.

-. 1985c. "Fuel famine dilemma." Islands Business 11(2): 49-51.

-. 1986a. "Trees, conflict resolution, and peace: The preservation of trees as a precondition for environmental and social stability." In: J. Maas, and R.A.C. Stewart, eds. Toward a World of Peace: People Create Alternatives. Suva: The University of the South Pacific, pp. 379-396.

-. 1986b. "The role of root crops, bananas and plantains in Pacific island food security." Contribution to document on The Role of Roots, Tubers, Plantains and Bananas in the Pacific Islands. Rome: Food and Agricultural Organization of the United Nations.

-. 1986c. "Microparks in the Pacific islands: The relevance of traditional and modern small-scale conservation areas in the context of nature conservation in the Pacific islands." In: P.E.J. Thomas, ed. Report of the Third South Pacific National Parks and Reserves Conference (Apia, Western Samoa): Vol. II. Collected Key Issue and Case Study Papers. South Pacific Regional Environmental Programme (SPREP), South Pacific Commission, Noumea, pp. 215-236.

-. 1987a. "Urban agroforestry: The Pacific islands and beyond." Unasylva 39(55): 2-13.

-. 1987b. "Plants of Kiribati: A listing and analysis of vernacular names." Atoll Research Bulletin 296: 1-42.

-. 1988a. "The forgotten fruit: Overspecialisation in agricultural and forestry development and the need for professional reversals and a whole-farm approach." In: P. Drysdale, ed. Agricultural Extension in Fiji: Proceedings of the Congress of the Fiji Institute of Agricultural Science, February 4 and 6, Suva. Lautoka: Fiji Institute of Agricultural Science, pp. 10.1-33.

-. 1988b. "Fijian agroforestry." In: J. Overton, ed. Rural Fiji. Institute of Pacific Studies, University of the South Pacific, Suva, pp. 31-34.

-. 1988c. "Rainforest management within the context of existing agroforestry systems." In: J. Heuveldop, M. Homola, H.J. von Maydell, and C. von Tuyll, eds. Proceedings: GTZ Regional Forestry Seminar, Fiji, 3-14 October 1988. FijiGerman Forestry Project, Ministry of Forests, Suva, on behalf of Deutsche Gesellschaft fur technische Zusammenarbeit, Hamburg, pp. 354-371.

-. 1988d. "Health and nutrition in the Pacific Islands: Development or underdevelopment." GeoJournal 16(2): 211-227.

-. 1988e. Environmental Issues in the Pacific Islands: Constraints to Sustainable Island Development: Pacific Issues 1. Woden, Canberra, Australia: Pacific Circle Consortium.

-. 1988f. Pacific island ecosystems and the need for their protection and maintenance. Background paper prepared for the South Pacific Regional Environmental Programme (SPREP) Workshop on Environmental Education Curriculum Development, Institute of Education, University of the South Pacific, Suva.

-. 1988g. "By the people and for the people: Home gardening and national development in the Pacific islands." In: B. Allen, Y. Byron, J. Hirst, and J. Overton, eds. Small Scale Agriculture. Commonwealth Geographical Bureau, London, and Department of Human Geography, Australian National University, Canberra, pp. 167-182.

-. 1989a. Agrodeforestation and the Neglect of Trees: Threat to the Well-being of Pacific Societies. SPREP Working Paper No. 5, South Pacific Regional Environment Programme, South Pacific Commission, Noumea.

-. ed. 1989b. Applied Atoll Research for Development: Proceedings of the Kiri bati Applied Atoll Research for Development Consultation, Tarawa, 27 February to 2 March, 1989. Ministry for Natural Resource Development, Bairiki, Tarawa, Kiribati.

-. 1989c. Coastal reforestation and agroforestry as immediate ameliorative measures to address global warming and to promote sustainable habitation of lowlying islands and coastal areas. Paper presented at the Environment and Policy Institute (EAPI) Workshop on "Responding to the Threat of Global Warming: Options for the Pacific and Asia," East-West Center, Honolulu, 21-27 June 1989.

-. 1990. "Kiribati agroforestry: Trees, people and the atoll environment." Atoll Research Bulletin 333: 1-29.

-. 1992. "Vegetation of Nauru and the Gilbert Islands: Case studies of poverty, degradation, disturbance and displacement." Pacific Science 46(2): 128-158.

Thaman, R.R., and T. Ba. 1979. "Energy needs and forest resources of small islands." In: W. Moran, P. Hosking, and G. Aitken, eds. Proceedings of the Tenth and 49th ANZAAS Congress. New Zealand Geographical Society Conference Series No. 10, Wellington, pp. 198-204.

Thaman, R.R., and W.C. Clarke, eds. 1983. Food and National Development in the Pacific Islands. Suva: IJniversity of the South Pacific.

Thaman, R.R., and W.C. Clarke. 1987. "Pacific island agrosilviculture: Systems for cultural and ecological stability." Canopy International 13(1): 6-7; 13(2): 8-10; 13(3): 6-9. (three-part series).

Thaman, R.R., H.I. Manner, and D.C. Hassall. 1985. Nauruan Plant Names. Environmental Studies Report No. 21, Institute of Natural Resources, University of the South Pacific, Suva.

Thaman, R.R., M. Meleisea, and J. Makasiale. 1979. "Agricultural diversity and traditional knowledge as insurance against natural disasters." Annex 10 in Natural Disaster Prevention Preparedness and Rehabilitation Meeting 26-30 November 1979: Summary Record. Suva: South Pacific Bureau for Economic Cooperation, pp. 75-96.

Thaman, R.R., and P.M. Thomas. 1982. "The cassava invasion: The cultural, nutritional and ecological impact of cassava on Pacific island food systems." In: R.M. Bourke, and V. Kesavan, eds. Proceedings of the Second Papua New Cuinea Food Crops Conference: Part Two. Port Moresby: Department of Primary Industries, pp. 330-350.

-. 1985. "Cassava and change in Pacific island food systems." In: D.J. Cattle, and K.H. Schwerin, eds. Food Energy in Tropical Ecosystems. New York: Gordon and Breach, pp. 191-228.

Thompson, L. 1940. Southern Lau, Fiji: An Ethnography. Bulletin 162, Bernice P. Bishop Museum, Honolulu.

TTHPDA and MIHPO (Trust Territory Health Planning and Development Agency and Marshall Islands Health Planning Office). 1984. Republic of the Marshall Islands FiveYear Comprehensive Health Plan 1983-1988. Office of Health Planning, Ministry of Health, Majuro.

Tudge, C. 1977. The Famine Business. Harmondsworth: Penguin Books.

Tudor, J., ed. 1972. Pacific Islands Yearbook (eleventh edn). Sydney: Pacific Publications.

Tuyll, C. van. 1988. Forestry extension in Fiji. Paper prepared for Seminar on Extension in Fiji organized by the Fiji Institute of Agricultural Science (FIAS), Suva.

Twyford, I.T., and A.C.S. Wright. 1965. The Soil Resources of the Fiji Islands. Suva: Government Printer.

Uhe, G. 1974. "Medicinal plants of Samoa." Economic Botany 28(1): 1-30.

Underwood, R.A. 1987. American education and the acculturation of the Chamorros of Guam. Unpublished Ph.D. thesis in Education, University of California, Los Angeles.

Uphof, J.C. Th. 1968. Dictionary of Economic Plants. Lehre, Germany: Verlag Von. J. Cramer.

Vasey, D. 1981. "Agricultural systems in Papua New Guinea: Adapting to the humid tropics." In: D. Denoon, and C. Snowden, eds. A Time to Plant and a Time to Uproot: A History of Agriculture in Papua New (Juinea. Port Moresby: Institute of Papua New Guinea Studies, pp. 17-31.

-. 1985. "Household gardens and their niche in Port Moresby, Papua New Guinea." Food and Nutrition Bulletin 7(3): 37-47.

Vergara, N.T., and N.D. Briones, eds. 1987. Agroforestry in the Humid Tropics: Its Protective and Ameliorative Roles to E:nhance Productivity and Sustainability. Environment and Policy Institute, East-West Center, Honolulu/Southeast Asian Regional Center for Graduate Study and Research in Agriculture, College, Laguna, Philippines.

Vergara, N.T., and P.K.R. Nair. 1985. "Agroforestry in the South Pacific region An overview." Agroforestry Systems 3: 363-379.

Vincent, A.J. 1971. An Experiment to Determine the Effect on Tree Growth Resulting from Grazing under Pinus caribaea var. hondurensis Plantations. Experiment Report No. 2, Department of Forestry, Suva.

Walker, E.H. 1954. Important Trees of the Ryukyu Islands. Special Bulletin No. 3, U.S. Civil Administration of the Ryukyu Islands.

-. 1976. Flora of Okinawa and the Southern Ryukyu Islands. Washington, D.C.: Smithsonian Institution Press.

Walker, F.S. 1962. The Forests of the British Solomon Islands Protectorate. Noumea: South Pacific Commission. (reprinted 1967).

Wall, J.R.D., and J.R.F. Hansell. 1974. Land Resources of the British Solomon Islands Protectorate: Vol. 3. Malaita and Ulawa: Land Resource Study 18, Land Resources Division, Ministry of Overseas Development, Surbiton, Surrey.

Walsh, A.C. 1982. Migration, Urbanization and Development in South Pacific Countries. Report prepared for PacificlESCAP, Bangkok.

Waqavonovono, M. 1980. Traditional Medicine and Practices: An Alternafive Health Care System for Women (A Case Study of Traditional Types of Healing Relating to Women's Ailments and Conditions in Seven Villages in Nadarivatu, Fiji). Centre for Applied Studies in Development, University of the South Pacific, Suva.

Ward, R.G. 1965. Land Use and Population in Fiji: A Geographical Study. London: H.M.S.O.

-. 1980a. "The environmental context." In: R.G. Ward, and A. Proctor, eds. South Pacific Agriculture: Choices and Constraints: South Pacific Agricultural Survey 1979. Canberra: Asian Development Bank in Association with Australian National University Press, pp. 3-25.

-. 1980b. "Agricultural options for the Pacific islands." In: R.T. Shand, ed. The island States of the Pacific and Indian Oceans: Anatomy of Development. Monograph No. 23, Development Studies Centre, Australian National University, Canberra, pp. 23-40.

-. 1980c. "Plus c,a change: Plantations, tenants, proletarians or peasants in Fiji." In: J.N. Jennings, and G.J.R. Linge, eds. Of Time and Place: Essays in Honour of O.H.K. Spate. Canberra: Australian National University Press, pp. 134-152.

-. 1982. "Dilemmas in south Pacific agriculture." South Pacific Journal of Natural Science 3: 9-30.

-. 1984. "Production or management- Where is the problem?" Pacific Viewpoint 25(2): 212-217.

-. 1986. "Reflections on Pacific Island agriculture in the late 20th century." The Journal of Pacific History 21(4): 217-226.

Ward, R.G., and A. Proctor, eds. 1980. South Pacific Agriculture Choices and Constraints: South Pacific Agricultural Survey 1979. Canberra: Asian Development Bank in Association with Australian National University Press, pp. 293-323.

Watling, D. 1985. A Mangrove Management Plan for Fiji: Phase 1. Zonation Requirements and a Plan for the Mangroves of the Ba, Labasa and Rewa Deltas. Suva: Government Printer.

-. 1988. "The effects of logging on Fijian wildlife." In: Fiji's Rainforests: Our Heritage and Future. Proceedings of the Second National Conservation Congress, Suva, Fiji, 9-10 June 1988, vol. 1, pp. 293-323.

Watt, G.R. 1980. "Forestry." In: R.G. Ward, and A. Proctor, eds. South Pacific Agriculture Choices and Constraints: South Pacific Agricultural Survey 1979. Canberra: Asian Development Bank in Association with Australian National University Press, pp. 293-323.

Watters, R.F., and K. Banabati. 1977. Abemama Island Report. Ministry of Local Government and Rural Development, Tarawa.

Weiner, M.A. 1971. "Ethnomedicine in Tonga." Economic Botany 25(4): 423-450.

-. 1984. Secrets of Fijian Medicine. Berkeley: University of California Press.

Wester, L. 1985. "Checklist of the vascular plants of the northern Line Islands." Atoll Research Bulletin 287: 1-38.

Western Samoa Government. 1972. The Report of the Survey of Household Conditions in Western Samoa, 1977-82. Apia: Department of Statistics.

Whistler, A.W. 1980a. "The vegetation of eastern Samoa." Altertonia 2(2): 45-190.

-. 1980b. Coastal Flowers of the Tropical Pacific. The Pacific Tropical Botanical Garden, Lawai, Kawai, Hawaii.

-. 1983a. Weed Handbook of Western Polynesia. Schriftenreihe der GTZ No. 157, German Agency for Technical Cooperation (GTZ), Eschborn.

-. 1983b. "The flora and vegetation of Swains Island." Atoll Research Bulletin 262: 1-25.

-. 1984. "Annotated list of Samoan plant names." Economic Botany 38(4): 464-489.

-. 1987. Ethnobotany of Tokelau: The Plants, Their Tokelauan Names and Their Uses. The Pacific Tropical Botanical Garden, Lawai, Kauai, Hawaii.

-. 1991. "Polynesian plant introductions." In: P.A. Cox, and S.A. Banack, eds. Islands, Plants and Polynesians: An Introduction to Polynesian Ethnobotany. Portland, Oregon: Dioscorides Press.

Whitesell, C.D., C.D. MacClean, M.C. Falanruw, T.G. Cole, and A.H. Ambacher. 1986. Vegetafion Survey of Kosrae, Federated States of Micronesia. Resource Bulletin PSW-17, Pacific Southwest Forest and Range Experiment Station, U.S. Department of Agriculture.

Whitmore, T.C. 1966. Guide to the Forests of the British Solomon Islands. London: Oxford University Press.

-. 1969. "The vegetation of the Solomon Islands." Philosophical Transactions of the Royal Society: Series B. Biological Sciences 255(8W): 259-270.

-. 1972. Tree Flora of Malaysia: A Manual for Foresters. Kuala Lumpur: Longman Malaysia.

Whitmore, T.C., and I.G.M. Tantra, eds. 1982. Tree Flora of Indonesia: Check List for Sumatra. Forest Research and Development Centre, Ministry of Forestry, Bogor.

Wiens, H.J. 1962. Atoll Environment and Ecology. New Haven: Yale University Press.

Wilder, G.P. 1931. Flora of Rarotonga. Bulletin 86, Bernice P. Bishop Museum, Honolulu.

-. 1934. The Flora of Makatea. Bulletin 120, Bernice P. Bishop Museum, Honolulu.

Wilkes, C. 1845. Narrative of the United States Exploring Expedition: Vol. 5. Philadelphia: Lea and Blanchard.

Willis, J.C. (revised by H.K. Airy Shaw). 1966. A Dictionary of the Flowering Plants and Ferns (seventh edn). Cambridge: Cambridge University Press.

Wilson, K., and R.M. Bourke, eds. 1975. Papua New Guinea Food Crops Conference: Proceedings. Port Moresby: Department of Primary Industry.

Wilson, W.S. 1968. Land, activity and social organization of Lelu, Kusaie. Ph.D. thesis in Anthropology, University of Pennsylvania, Philadelphia.

Winslow, J.H., ed. 1977. The Melanesian Environment. Canberra: Australian National University Press.

Withington, D., K.G. MacDicken, C.B. Sastry, and N.R. Adams. 1988. Multipurpose Tree Species for Small-Farm Use: Proceedings of an International Workshop held Novemoer 2-5, 1987, in Pattaya, Thailand. Winrock International Institute for Agricultural Development, Arlington, Virginia/International Development Research Centre of Canada, Ottawa.

Wood, P.J., and J. Burley. 1991. A Tree for All Reasons: The Introduction and Evaluation of Multipurpose Trees for Agroforestry. Nairobi: International Centre for Research in Agroforestry.

Woodroffe, C.D. 1985. "Vegetation and flora of Nui atoll, Tuvalu." Atoll Research Bulletin 283: 1 -28.

Yang, Y.H. 1976. Home gardens: A low-input and high-output food resource for rural families in Asian and Pacific countries. Paper presented at the Planning Seminar on Agriculture for Nutritional Improvement, 16-27 August, East-West Center, Honolulu.

Yao, C.E. 1986. "Mangrove reforestation in central Visayas." Canopy International 12(2): 6-9.

Yen, D.E. 1971. "The development of agriculture in Oceania." In: R.C. Green, and M. Kelly, eds. Studies in Oceania Culture History: Yol. 2. Pacific Anthropological Records No. 12: 1-12.

-. 1973. "Agriculture in Anutan subsistence." In: D.E. Yen, and J. Gordon, eds. Anuta: A Polynesian Outlier in the Solomon Islands. Pacific Anthropological Records No. 21, Bernice P. Bishop Museum, Honolulu, pp. 112-155.

-. 1974. "Aboriculture in the subsistence of Santa Cruz, Solomon Islands." Economic Botany 28(3): 247-286.

-. 1976a. "Agricultural systems and prehistory in the Solomon Islands." In: R.C. Green, and M.M. Cresswell, eds. Southeast Solomon Islands Cultural History: A Preliminary Survey. Bulletin 11, Royal Society of New Zealand, Wellington, pp. 61-74.

-. 1976b. "Inland settlement on Santa Cruz island (Nendo)." In: R.C. Green, and M.M. Cresswell, eds. Southeast Solomon Islands Cultural History: A Preliminary Survey. Bulletin 11, Royal Society of New Zealand, Wellington, pp. 203-224.

-. 1980a. "Food crops." In: R.G. Ward, and A. Proctor, eds. South Pacific Agriculture Choices and Constraints: South Pacific Agricultural Survey 1979. Canberra: Asian Development Bank in Association with Australian National University Press, pp. 197-234.

-. 1980b. "Pacific production systems." In: R.G. Ward, and A. Proctor, eds. South Pacific Agriculture Choices and Constraints: South Pacific Agricultural Survey 1979. Canberra: Asian Development Bank in Association with Australian National University Press, pp. 73-106.

-. 1990. "Environment, agriculture and the colonisation of the Pacific." In: D.E. Yen, and J.M.J. Mummery, eds. Pacific Production Systems: Approaches to Economic Prehistory. Department of Prehistory, Research School of Pacific Studies, The Australian National University, Canberra.

-. 1991. "Polynesian cultigens and cultivars: The questions of origin." In: P.A. Cox, and S.A. Banack, eds. islands, Plants and Polynesians: An Introduction to Polynesia Ethnobotany. Portland, Oregon: Dioscorides Press, pp. 67-97.

Yuncker, T.G. 1959. Plants of Tonga. Bulletin 220, Bernice P. Bishop Museum, Honolulu.

Zimmet, P.Z., A. Seluka, J. Collins, P. Currie, J. Wickering, and W. Deboer. 1981. `'The prevalence of diabetes mellitus in urbanised, isolated populations. The Funafuti Survey." Diabetes 30: 40-44.

Contributors

RANDOLPH R. THAMAN. Professor of Pacific Islands Biogeography at the University of the South Pacific, Suva, Fiji, where he has been since 1974. He also serves as Chairman of the Fiji National Food and Nutrition Committee and was founder and Chairman (1982-1986, 19891990) of the South Pacific Action Committee for Human Ecology and the Environment (SPACHEE). His B.A. and M.A. in Geography are from the University of California, Berkeley, and his Ph.D. from the University of California at Los Angeles. His research interests include Pacific Island agriculture, agroforestry, food systems, nutrition and health, Pacific vegetation and ethnobotany, environmental management and education, and remote sensing. He has carried out field research in Papua New Guinea, Vanuatu, Fiji, Rotuma, Tonga, Nauru, the Cook Islands, Kiribati, Tuvalu, Indonesia, and China.

WILLIAM C. CLARKE has served as Professor of Geography at the University of the South Pacific, Suva, Fiji, at Monash University in Melbourne, and at the University of Papua New Guinea. Prior to those appointments, he was Senior Research Fellow in the Department of Human Geography, Research School of Pacific Studies, the Australian National University, and Associate Professor of Geography at the University of Hawaii, Honolulu. His B.A. in Anthropology and his M.A. and Ph.D. in Geography are from the University of California, Berkeley. His research interests include Pacific agriculture and agroforestry, tropical biogeography, soil erosion, human ecology, and tourism. He has carried out field research in the Caribbean, Cape York, Papua New Guinea, Fiji, Rotuma, and Vanuatu.

HARLEY I. MANNER. Professor of Geography, University of Guam, where he has been since 1985; formerly Senior Lecturer in Geography at the University of the South Pacific, Suva, Fiji. He has a B.A., M.A., and Ph.D. in Geography from the University of Hawaii, Honolulu. His research interests include Pacific agriculture and agroforestry, vegetation degradation and retrogression, effects of reforestation on soils and hydrology, environmental management, medicinal plants, and Pacific ethnobotany. He has carried out field research in Hawaii, Papua New Guinea, the Solomon Islands, Fiji, Rotuma, Nauru, and several islands in Micronesia.

BRYCE G. DECKER received his Ph.D. from the University of California in 1970, with a dissertation on the plant cover in and near the inhabited valleys of the Marquesas Islands. Following appointments at the University of Delaware and the Department of Botany at the Smithsonian Institution's National Museum of Natural History, he joined the Department of Geography at the University of Hawaii in Honolulu, where he presently serves. His research interests include historical geography of the Pacific Islands from a biological and ecological perspective, vegetation and flora, plant geography, Hawaii and French Polynesia, and problems of conservation and land use, with special reference to Hawaii.

IMAM ALI received his B.Ed. in 1976 from the University of the South Pacific. While serving as the Principal of a high school in Fiji, he took up postgraduate studies in Geography and received an M.A. from the University of the South Pacific with a thesis on changing patterns of land use among a community of sugar-cane farmers in Fiji. He is presently a Lecturer in Geography at the University of the South Pacific, after recently completing four years as a Ph.D. scholar in Human Geography at the Australian National University, carrying out research on changes among highland vegetable farmers in Sabah.