|Priorities for Water Resources Allocation (NRI)|
|Watershed management and land use|
Ian R. Calder
Institute of Hydrology, Wallingford
Summary: We live in a changing world and the effects of the changes are of interest to us all. On a global scale the most significant land-use change in terms of land area, and arguably also in terms of hydrological effects, involves afforestation and deforestation practices. In the tropics, the deforestation of indigenous forests continues as land is converted to agriculture to feed increasing populations though the balance of forested land is being partially redressed through commercial afforestation of fast growing, often exotic, tree species. In contrast, in the developed world, and particularly within Europe, the balance of forested land is likely to increase as a result of improved agricultural productivity and food surpluses and a move to 'set aside' policies for agricultural land. Planting trees creates concern that they will intercept more rainfall during wet periods and, because of their deeper root systems, transpire more water during dry periods and thus deplete groundwater and downstream surface water resources; acidification may also result. Cutting down trees raises concerns of erosion, siltation of streams and increased leaching of soil nutrients. Forests are also likely to have beneficial effects on climate, at all scales ranging through micro to meso and global. These and other issues are discussed in relation to recent experimental studies into the hydrological impact of temperate forests in the UK, indigenous tropical forests in Indonesia and Brazil and of Eucalyptus plantations in southern India. A summary of the expected impacts of forests in relation to water yield, floods, low flows, water quality, erosion and climate is presented.
Forests are generally regarded as being beneficial to the environment. Bio-diversity and global climate issues in relation to forests received high priority at the UNCED conference. However forest impacts on the environment may not always be beneficial; although forests, through increased evaporation, generally have a favourable effect on climate, forests, because they evaporate more water than other vegetation types, are likely to deplete surface and groundwater resources. Both water and forests are central to the development of many LDC's economies; wood is required for buildings and for fuel for local people and timber is required for paper and rayon-based industries. Plantation forests, with high water-use efficiency, can meet these needs and take the pressure off remaining indigenous forests, whilst minimizing the effects on water resources.
The impacts of forests on the environment are not always easy to assess because many competing processes are often at work and the net result cannot always be predicted accurately with current knowledge. Research may still be required. Nevertheless many of these impacts are now fairly well understood and this paper attempts to summarise these impacts so that the environmental implications of forests are better understood in relation to development projects.
The hydrological impact of forests has always been a contentious issue. Within the UK the effects of coniferous afforestation of the uplands on water quantity and water quality stimulated many studies both at the process study and catchment scale. More recently the effects of broadleaf afforestation of the lowlands of the UK and the European Community as a result of 'set aside' policies have received more prominence. The issues raised in relation to the hydrological impact of forestry in the tropics and in developing countries are perhaps the most serious. It is often in these countries that water represents one of the most important constraints on development and where any adverse effects on water resources should be viewed with concern. The effects of eucalypt plantations on water quantity have aroused controversy in many tropical and subtropical countries including India, Kenya, Uganda, South Africa and Portugal and have stimulated a large ODA-funded research projects in Karnataka, southern India. The water relations and climatic impacts of tropical rainforests have also received great interest and ODA have funded projects in Indonesia and more recently in Brazil with the Anglo-Brazilian Amazonian Climate Observation Study (ABRACOS). Concerns over the hydrological impacts of tropical plantation forestry are not restricted to eucalypts, tropical pines are also under scrutiny. One of the principal objectives of the ODA forestry programme, involving pine plantations, in Sri Lanka was to 'regulate' the flows to the Victoria water supply and hydropower reservoirs and, thereby, to reduce erosion. However, current hydrological knowledge would suggest that the impacts of the plantations, in the areas where they are currently being planted, are likely to reduce flows overall and may even increase erosion. Nevertheless, planting at higher altitudes in Sri Lanka, where cloud deposition to forests may be a significant process, holds the promise of improving water resources.
In this paper current knowledge on some of the hydrological impacts of forests is outlined, particularly in relation to ODA-funded research projects in Indonesia, Brazil and India. Further details of the hydrological impacts of land-use change including the impacts of forestation, agricultural intensification, and the drainage of wetlands are available in recent publications (Carder, 1990; Calder, in press).
Hydrological impacts and processes
The higher water use of forests compared with shorter vegetation is due principally to two processes. In wet areas of the world, such as the uplands of the UK, the high aerodynamic roughness of forests leads to greatly enhanced evaporation rates in wet conditions (interception) and evaporation can, on an annual basis, be as much as twice that for grass. In drier climates the deep root systems of forest and their greater water availability during dry seasons leads to higher transpiration losses. The water-use studies carried out under the ODA-funded eucalypts project in the dry zone of southern India have established that the total evaporation (transpiration plus interception) from forest is nearly 1.5 times greater than from agricultural crops.
Annual flow: Annual flow results from catchment experiments have been reviewed by Hewlett and Hibbert (1967) and Bosch and Hewlett (1982). From an analysis of results from 94 catchments world-wide Bosch and Hewlett concluded that:
· Pine and eucalypt types cause an average change of 40 mm in annual flow for a 10% change in cover with respect to grasslands, that is, a 10% increase in forest cover will decrease annual flow by 40 mm, a 10% decrease in cover will increase annual flow by the same amount.
· The equivalent response on annual flow of a 10% change in cover of deciduous hardwood or scrub is 25-10 mm, that is, if 10% of a grassland catchment is converted to hardwood trees or scrub vegetation, the annual runoff will decrease by 10-25 mm.
Although the impacts on annual flow are related to local climate and soil characteristics, an overall reduction in flow is to be expected, with few exceptions, from forests world-wide. Better quantification of the impacts in a particular area can be achieved if the limits on forest evaporation can be identified. The uplands of the UK, subject to a maritime climate typified by high rainfall, a high number of raindays per year and high windspeeds, are an example of a situation where large-scale advection is the principal limit on forest evaporation. In the UK uplands, the total evaporative losses from forest can consume an amount of latent heat that easily exceeds the radiant energy input to the forest (Table 1).
Table 1 Observations of the annual water and energy balance of moist tropical and temperate forests
The wet evergreen forests of the tropics represent another situation where climatic demand is likely to limit forest evaporation. However, climate circulation patterns in the tropics do not favour large-scale advection of energy to support evaporation rates and here evaporation rates are likely to be closely constrained by the availability of solar radiation (Table 1). As humid rain forest is able to convert, on an annual basis, virtually the equivalent of all the net radiation into evaporation it is unlikely that any other land use will be able to evaporate at a higher rate and conversion of forest to annual crops in these areas will increase annual flows.
In very low rainfall areas the principal limit on annual evaporation is soil water availability. Studies in Karnataka, southern India (Harding et al., 1992), show that the available soil water capacity of both indigenous, dry deciduous forest and Eucalyptus plantation is of the order of 480 mm whereas, in the same region, the available water capacity for finger millet, an annual agricultural crop, is 150 mm. The annual evaporation from the indigenous and plantation forests is, within the errors of measurement (10%), equal to the rainfall of 800 mm/year the evaporation from the finger millet, with a reduced soil water reservoir to exploit, is 500 mm/year. Conversion from forest to agricultural crops in this area will therefore increase annual flow (or catchment recharge) by this difference in annual evaporation. The studies also demonstrated the importance of tree size and age as limiting factors on evaporation. Measurements made on these (young) Eucalyptus plantations have established a new and surprisingly close correlation, Figure 1, (Carder et al., 1992) between the transpiration rate of an individual tree and its stem cross-sectional area (a better correlation than was found with leaf area). This relationship, when expressed in terms of the total stem cross-sectional area of the stand per hectare, and with the use of a suitable soil moisture regulating function, enables the stand evaporation to be calculated and has been used in models to predict the evaporation, the soil moisture deficit and the volume growth (Carder, 1992) and will be used in the future to improve the water-use efficiency of the stands. The only meteorological data that are required are daily rainfall. Meteorological demand, although providing the driving force for evaporation, is not thought to be a limiting factor during most of the year; the principal limitations on transpiration are thought to be soil moisture availability and tree size. These results from semi-arid Karnataka, which indicate that evaporation is limited principally by soil water availability and plant physiological controls, are therefore in direct contrast to the observations from the wet uplands of the UK where evaporation is principally limited by atmospheric demand and physical, aerodynamic controls.
Figure 1 Transpiration rate of Eucalyptus tercticornis trees in conditions with little soil moisture stress at sites in Southern India - measured using the deuterium tracing method plotted against the basal (stem) cross sectional area of the tree measured at 1.2 m above ground level
Seasonal flow: Afforestation may affect seasonal flow through two principal mechanisms. Firstly, the higher interception losses from forests in wet periods and increased transpiration losses in dry periods (because of deeper root systems) both tend to increase soil moisture deficits in dry periods compared with those under shorter crops. These increased deficits lead to reduced dry-season flows where part, at least, of the dry-season flow is derived from the soil reservoir. Secondly, land drainage operations, which are often part of the management associated with afforestation in wet, temperate climates, tend to increase flows as a result both of the initial dewatering (which may take a number of years) and through the long-term effects of the alteration of the drainage regime. The two mechanisms are opposing and the net effect on low flows may result in either higher or reduced low flows but in the long term, when trees have reached maturity, it is expected that the effects of increased evaporation will predominate and low flows will be reduced.
Cloud forest: For high-altitude forest, or cloud forest, which is above the cloud base for a significant proportion of the year, the deposition of cloud water onto the forest is likely to be a significant hydrological process. Because of the reduced aerodynamic transport of water vapour above forest, and increased leaf area of forest, compared with shorter crops, the cloud deposition rates onto forest will be many times greater than those onto short vegetation. For cloud forest in locations such as the Andes, Hawaii and Sri Lanka cloud-water deposition may provide a significant component of the dry-season flow in rivers.
A further example of a situation in which forests may assist in supporting dry-season flows is where forests are being used to reclaim degraded lands. There is some evidence to suggest that, where forests have been planted in India in degraded areas with laterite outcrops, the increased infiltration of rainfall into the soil beneath the forest exceeds the extra evaporation from the forest and recharge to groundwater aquifers is increased.
There is greater awareness that there are not only water quantity but also water quality implications of afforestation; forestry has been associated with catchment acidification. Interestingly, process studies have identified that the same process is responsible for both increased evaporation in wet conditions and increased acidification. The higher aerodynamic transfer of water vapour and heat between the surface of the forest vegetation and the atmosphere, in comparison with shorter vegetation, allows the high evaporation rates of intercepted water (interception) and higher deposition rates of pollutants in the dry form as reactive gases and particles and in the wet form as pollutants contained within cloud and mist droplets. Cloud and fog water contain significantly larger ionic concentrations than rain with peak concentrations up to 50 times greater (UK Review Group on Acid Rain, 1990). Recent studies (Fowler et al., 1989) indicate that for high-altitude forests in the UK (- 500 m), altitudes sufficient for forests to intercept cloud and mist droplets frequently, the deposition of sulphur particles contained within cloud droplets (5-10 mm radius) may make a large contribution to the total annual deposition. Because cloud droplets, as opposed to sub-micron-sized dry particles, are efficiently captured by vegetation surfaces, and as forests have lower aerodynamic resistances compared with shorter vegetation, deposition rates of cloud-borne pollutants onto forests will be greater than deposition onto shorter crops.
The most disruptive effects of forestry on water quality arise through intensive management practices associated with harvesting, site preparation and site management. In particular, clearcutting can result in large increases in nutrient concentrations in watercourses. The highest concentrations reported in the USA are from forests in New Hampshire. Hornbeck et al. (1975) and Pierce et al. (1972) report values of 26 mg/dm³. More commonly values of about 1 mg/dm³ have been reported for other forests in America. The increased nutrient concentrations affect lake and stream eutrophication and increase the outbreaks of phytoplankton blooms.
Forestation and erosion
Forestry operations are often associated with increased erosion. Land drainage operations prior to afforestation, the construction of access roads, felling operations involving soil compaction and disturbance all increase erosion as they do flooding. The presence of the forest also affects erosion. Principally these are through the effects on slope stability and on splash detachment. In relation to slope stability O'Loughlin and Ziemer (1982) state that the positive influences of forests on erosion depend upon the reduced soil pore water pressure caused by the forest evaporation, accumulation of an organic forest floor layer and mechanical reinforcement of the soil by tree roots. Negative influences result from windthrow of trees and the weight of the tree crop itself.
Vegetation canopies influence splash detachment through the modification of the natural raindrop size spectrum. Contrary to popular belief forest canopies do not necessarily 'protect' the soil from raindrop impacts. For storms with small raindrop sizes, usually low intensity storms, canopies tend to amalgamate drops until vegetation elements are fully wetted and larger drops are released as net rainfall. Depending upon the height of the vegetation above the ground (drops of up to 6 mm diameter will reach terminal velocity within 12 m) drops may approach terminal velocity and acquire a higher kinetic energy than those in the natural rainfall (Morgan, 1985). The potential for greater splash detachment from bare mineral soils is therefore greater under tall forest canopies than under shorter vegetation. Conversely, for storms with the largest drop sizes, usually the higher intensity storms, vegetation canopies may break up the large drops and reduce both the mean drop size and the mean kinetic energy of the incident rain. The Eucalyptus water-use studies in India (Hall and Calder, in press) have shown that vegetation canopies have characteristic net rainfall spectra. For Pin us caribaea, irrespective of the drop size spectra of the incident rain the throughfall spectra remain essentially unchanged (Figure 2) and retains a 'signature' characteristic of this particular vegetation type. For three tree species studied, Pinus caribaea, Eucalyptus camaldulensis and Tectona grandis, median volume drop diameters of the throughfall ranged from 2.6 to 4.6 mm (Figure 3) whilst corresponding drop kinetic energies, assuming the drops reached terminal velocity, ranged by a factor of 7 with Pinus caribaea having the least and Tectona grandis the greatest kinetic energies.
Figure 2 Cumulative frequency distribution of throughfall drop spectra beneath Pinus caribaea subject to spray with median volume drop diameter (the drop diameter for which 50% of the volume was in drops less than this value) of 3.2 mm and 1.9 mm
Figure 3 Cumulative frequency distribution of throughfall drop spectra for three tree species subject to spray with median volume drop diameter (the drop diameter for which 50% of the volume was in drops less than this value) of 3.2 mm.
Splash detachment mobilises soil particles which can be transported if there is surface runoff. These small soil particles can dog surface micropores and macropores leading to an impermeable crust which itself reduces infiltration and enhances the production of surface runoff. In natural mixed forests, where a surface vegetation cover or a deep litter layer is usually present which helps to protect the soil surface from raindrop impact, and where infiltration capacities are high, surface runoff and surface erosion are usually minimal. For plantation forest the understorey cover of vegetation is often reduced by shading or through competition for soil water or nutrients. For some plantations outbreaks of fire are a common occurrence which destroy both understorey vegetation and litter layers. Plantations which have both tree species with large net raindrop spectra, such as Tectona grandis, and a lack of understorey or a litter layer have the potential for particularly high rates of soil erosion.
Land use affects climate. Depending upon the scale of the land-use change the effect can occur on a micro, meso or global scale. The effect occurs principally through the different inputs, into the atmosphere, of heat, water vapour and radiation from the different land surfaces. The variation with height of temperature, humidity and windspeed close to a surface is the result of a balance between externally applied climatic variables, the surface fluxes of heat and water vapour, and the aerodynamic properties of the surface. Differences in the water availability at the evaporating surface will produce marked differences in micro-climate as a result of altering surface fluxes of heat and water vapour. An extreme example is the cool, moist micro-climate found over a forest which has a deep root system and readily available soil water as compared with the hotter, drier micro-climate found above a short-rooted crop or a bare soil (where evaporative fluxes will be much less). For land-use changes occurring over areas extending for tens of kilometres the height of the planetary boundary layer (the height of the cloud base) may be altered and meso-climate change may occur. The scale of the effect is poorly understood at present and warrants further research. Similarly, the alteration of surface fluxes of heat and water vapour as a result of land-use change may have an impact on global climate. The Brazilian ABRACOS project, funded by ODA, is seeking to parameterise the surface fluxes from Amazonian rain forest for use in Global Climate Model (GCM) predictions of climate change.
The question of whether the effects of a land-use change can alter rainfall is still controversial. Kitteridge (1948) concluded that the influence of forests on rainfall generation is small, less than a 3% increase in temperate climates in rainfall over forests as compared with grassland, which is caused by the increased orographic effect resulting from the height of the trees raising the effective height of the topography. Some 40 years later it is possible to say little more on the effects of land use on rainfall generation on the meso-scale, although recent developments in mesoscale climate modelling indicate that the increased evaporation of intercepted water from forests can humidify the planetary boundary layer and can lead to a 5-10% increase in the regional rainfall. Further experimental and modelling studies are required to provide information on this important and contentious topic.
A summary of the hydrological impacts associated with land-use change is given in Table 2.
The impact of forests and of forestry management practices is likely to have profound effects on hydrology and climate at both the local and regional scale. There may still be a requirement for research to quantify these impacts for a given environment; one such environment is the dry tropics where the major part of the worlds tropical forests reside and which support large populations but which are, at present, very poorly researched. Perhaps more importantly, research should be directed not just to quantifying the impacts, as has largely been the case in
Table 2 Summary of the major hydrological effects of land-use change
Table 2 contd.
CALDER, I. R., SWAMINATH, M. H., KARIYAPPA, G. S., SRINIVASALU, N. V., SRINIVASA MURTY, K. V. and MUMTAZ, J. (1992) Deuterium tracing for the estimation of transpiration from trees. 3) Measurements of transpiration from Eucalyptus plantation, India. Journal of Hydrology, 130, 37-48.
FOWLER, D., CAPE, J. N. and UNSWORTH, M. H. (1989) Deposition of atmospheric pollutants on forests. Philosophical Transactions of the Royal Society, 324, 247-265.
HALL, R. L. and CALDER, I. R., (In press) Drop size modification by forest canopies measurements using a disdrometer. Submitted to Journal of Hydrology.
HARDING, R. J., HALL, R. L., SWAMINATH, M. H. and SRINIVASA MURTHY, K. V. (1992) The soil moisture regimes beneath forest and an agricultural crop in southern India measurements and modelling. In: Growth and Water Use of Forest Plantations. Proceedings of the International Symposium on the Growth and Water Use of Forest Plantations, Bangalore, 7 - 11 February 1991. CALDER, I. R., HALL, R. L. and ADLARD, P. G. (eds), John Wiley & Sons, Chichester, UK.
HEWLETT, J. D., and HIBBERT, A. R. (1967) Factors affecting the response of small watersheds to precipitation in humid areas. In: International Symposium on Forest Hydrology. SOPPER, W. E. and LULL, H. W. (eds) Pergamon Press, Oxford.
HORNBECK, J. W., PIERCE. R. S., LIKENS, G. E. and MARTIN, C. W. (1975) Moderating the impact of contemporary forest cutting on hydrologic and nutrient cycles. In: International Symposium on Hydrological Characteristics of River Basins. Tokyo, Japan, December 8-11, 1975. International Association of Hydrological Sciences Publication 117, 423-433.
KITTREDGE, J. (1948) Forest Influences. McGraw-Hill Book Company, Inc. New York.
MORGAN, R. P. C. (1985) Establishment of plant cover parameters for modelling splash detachment. In: Soil Erosion and Conservation. EL-SWAIFY, S. A., MOLDENHAUER, W. C. and LO, A. (eds) Soil Conservation Society of America.
O'LOUGHLIN, C. L., and ZIEMER, R. R. (1982) The importance of tree root strength and deterioration rates upon edaphic stability in steepland forest. In: Carbon Uptake and Allocation: a Key to Management of Subalpine Ecosystems. WARING, R. H. (ed), Corvallis, Oregon, USA.
PIERCE, R. S., MARTIN, W. C., REEVES, C. C. et al. (1972) Nutrient loss from clearcuttings in New Hampshire. In: Natl. Symp. Watersheds in Transition Proc., American Water Resources Association, Urbana, III.
SHUTTLEWORTH, W. J. (1988) Evaporation from Amazonian rainforest. Proceedings of the Royal Society of London B. 233, 321-346.
UNITED KINGDOM REVIEW GROUP ON ACID RAIN (1990) Acid deposition in the United Kingdom, 1986-1988. Warren Springs Laboratory, Department of Trade and Industry, Stevenage, UK.
Where transpiration rates under forest are 1.5 times the rainfall it could be assumed that the trees were mining the previous years' rainfall. In response to a question on whether the natural acidity of the soil influences the uptake of pollutants it was said that the Institute of Hydrology was looking at the transference of pollutants including the effect of the canopy roughness. It was pointed out that land-use management decisions involve numerous factors and it is not simply a question of forestry versus hydrology. Also we should not discount local perceptions such as the conventional wisdom that trees do enhance or regulate stream flow. In response it was said that anecdotal material is unreliable; 100 catchments world-wide show reduction in flow when forest is removed. Important and understated benefits of large forest blocks are the prolongation of wet season by a few days and temperature moderation with the potential to induce considerable land-use changes. Since there is a linear correlation between evaporation rates and trunk cross section it was suggested that tree size could be an excellent measure of water use, simplifying our understanding of hydrological processes on a grand scale. It was reported that in Australia water use is being estimated from the indigenous sparse Eucalyptus forest by remote sensing based on the assumption that leaf areas grow to use the available water.