ICRISAT did not attempt to re-invent the wheeled toolcarrier, but rather evaluated a variety of pre-existing models, including the Kenmore (UK), the Otto Frame (India), the Polyculteur (Senegal) and the Tropiculteur (France). The preferred design was the Tropiculteur, manufactured in France by Mouzon, and the 1975-1976 ICRISAT Annual Report contained three photographs of this toolcarrier looking remarkably similar to present-day models. ICRISAT obtained the services of the French agricultural engineer Jean Nolle, who since starting his pioneering work in Senegal had designed several wheeled toolcarriers including the Tropiculteur, and who therefore was the world's leading specialist in this field. Jean Nolle carried out consultancy assignments for ICRISAT in 1976 and renamed his design Tropicultor to make it more international (Nolle, 1986). ICRISAT subsequently purchased the rights to allow the local manufacture in India of the Tropicultor (ICRISAT, 1979).

Originally designed in 1963, the Tropicultor has been modified and refined over the years, but essentially it consists of a strong chassis made of steel tube supported on wheels with pneumatic tyres. The wheels which are mounted on stub axles give an adjustable track and can be fitted either inside or outside the chassis. A wide range of implements can be clamped to a square section toolbar hinged to the chassis, which can be raised and lowered with a lever, The Tropicultor can carry one or more operator and a one tonne payload. Following several years of technically successful on-station trials and some on-farm evaluation, in 1985 ICRISAT published a detailed and well illustrated manual on the use of the Tropicultor. This covers implement assembly and a range of field operations including plowing, tine cultivation, harrowing, making broadbeds, seeding and weeding. This manual was designed for publication in different languages, to aid the adoption of the Tropicultor in different areas (ICRISAT, 1985).

Even in the early years of research at ICRISAT there was concern over the cost of wheeled toolcarriers which were technically efficient but also too expensive for most farmers in the semi-arid tropics. Efforts to "decrease the cost" of the Tropicultor started as early as 1975 (ICRISAT, 1976) and subsequently three attempts were made by ICRISAT to develop cheaper toolcarriers.

4.3.3 The Akola cart-based wheeled toolcarrier (1978-1982)

One attempt to develop a low-cost toolcarrier started in 1978, and was the only toolcarrier to be developed at ICRISAT that was not derived from a French or British design. The toolcarriers were based on the relatively small and lightweight passenger bullock carts made of wood by artisans in the Akola region of the Maharashtra State of India. Akola carts were purchased and their axles were converted to take the implements designed for the Tropicultor. Four units were tested, and during on-station trials in 1978 and 1979 they performed operations with a precision comparable with that of the more expensive Tropicultor. Lal (1986) considered the cart-based toolcarriers were an important development, being based on existing artisanal technology and at an estimated cost of about $ 300 (primarily the cost of the implements) they were less than one third of the cost of the Tropicultor. Although it was based on traditional cart axles and wooden spoke wheels, the cart-based toolcarrier was not designed to allow easy conversion between cart and toolcarrier. Nevertheless loadbearing platforms could have been built onto the axle if required.

The initial trials with the Akola cart-based toolcarrier were sufficiently optimistic to justify a season of comparisons with 22hp tractors at `'operational" level (2-3 ha) in 1979 and one objective of these on-station trials was to "study the economics" of the cart-based toolcarrier (ICRISAT, 1980). Operations using the cart-based toolcarrier were easier and more trouble-free than with the tractor (ICRISAT, 1980) but work on the Akola toolcarrier was not continued. The reasons for the rejection of this toolcarrier were not given in the optimistic report of Lal (1986), who considered that it was due primarily to his own departure and the fact that no one else was sufficiently interested in taking on research on lower cost implements. Other researchers at ICRISAT cited problems of standardization of dimensions, structural weakness, limited endurance and rising costs of wood (ICRISAT, 1984; Bansal, Awadhwal and Takenaga, 1986).

It should be noted that the Akola cart toolcarrier was a hybrid of traditional and modern technologies, for it had been designed to use all the tools of the Tropicultor. The main reasons for its rejection seem to have been related to the engineering problems (and costs) of the hybridization process. This necessitated reliably adapting the carts to take precision implements made of steel. No attempt had been made to adapt other artisanal technology (such as traditional "Des)" plows and blade harrows) to the toolcarrier concept, or develop village-level artisanal solutions to the perceived engineering problems.

4.3.4 The NIAE/ICRISAT (Nikart) wheeled toolcarrier (1979 - 1986)

A second initiative to develop a cheaper toolcarrier started in 1979 when the British National Institute of Agricultural Engineering (NIAE) with funding from the British Overseas Development Administration (ODA) started to collaborate with ICRISAT on the design of a new toolcarrier intended to be simpler and of lower cost (ICRISAT, 1979).

A review of existing models was carried out, and it was found that none of these were being marketed at an acceptable cost.

Four major design problems were identified on existing toolcarriers:

As a result of the review, a design philosophy was adopted that would attempt to combine multipurpose use with simplicity yet would intentionally limit some of the options for versatility in favour of lower production costs. Among the design specifications were the capability to perform conventional tillage as well as the broadbed cultivation, one-man ride-on operation, on-the-move depth adjustment and rapid conversion to a one tonne cart (NIAE, 1981).

Early prototypes of the new wheeled toolcarrier, which became widely known by the name Nikart, were made at NIAE and were tested by ICRISAT at Patancheru in 1979. In 1980 four slightly modified units were successfully tested at Patancheru, and there was then a need for further examples for on-farm testing. The British Intermediate Technology Development Group (ITDG) was contracted in 1981 to supervise the start of local production at the privately owned Mekins Agro Industrial Enterprises workshop at Hyderabad. The ITDG consultant found that although the Nikart had been designed to be made from locally available materials, there had still been the need to make certain design changes to take account of the actual availability of different steel sizes and qualities. The consultant concluded that the original Nikart design had in practice been separated from the realities of the resources and skills available to the small-scale producers (Barwell, 1983), although one specific objective of the design team had been to avoid this problem (Kemp, 1980). NIAE considered that there had been no contradiction between design philosophy and practice, as the contracting of IT-Transport to assist in establishing the manufacturing process and identifying any necessary changes had been an integral part of the research and development programme (D. Kemp, personal communication, 1987).

The Mekins workshop produced about 100 Nikart wheeled toolcarriers during the period 1981 to 1984 (Fieldson, 1984; Kshirsagar, Fieldson, Mayande and Walker, 1984) and 32 in the period 1985 - 1987 (Agarwal, personal communication, 1986). During the same time it also manufactured about 1100 Tropicultors. Almost all sales have been to development projects and institutions, some of which have lent them to farmers or sold them with 50-80% subsidies. Several other workshops in India including Medak Agricultural Centre, Kale Krish Udyog (Pune) and Sri Lakshmi Enterprises (Bangalore) made small numbers of Nikart-type prototypes between 1981 and 1984, but all preferred to manufacture Tropicultors, and all subsequently stopped making toolcarriers.

ICRISAT and NIAE have also promoted the Nikart design in other semi-arid areas. The version most widely distributed has been the GOM Toolcarrier manufactured in the U.K. by Geest Overseas Mechanization. Between 1981 and 1986 about 100 GOM Toolcarriers were sold to aid agencies and development projects in at least twenty different countries including Botswana, Burma, Ethiopia' Mali, Mexico, Mozambique, Philippines, Sudan and Zimbabwe. Most were sold in small numbers for evaluation, and by early 1987 there had not been any significant follow-up orders. By 1986 Geest was pessimistic about the prospects for its own manufacture of these toolcarriers due to the inability of small farmers to afford them, and the prohibitive costs of manufacturing such items in the U.K.

The British ODA, in cooperation with NIAE, assisted the start of production in Mexico of the Yunticultor toolcarrier based on the Nikart research and development. A smaller initiative, also with technical support from NIAE, was started in Honduras, and a prototype Yunticultor Mk II was developed to make local fabrication more easy. By 1986 about 100 Mexican and 20 Honduran Yunticultors had been made. Few had been bought by farmers and most sales were to government agencies, development projects and research stations. These Latin American experiences are discussed further in Chapter 6.

One of the main objectives of the Nikart project had been to reduce the price of the basic toolcarrier by at least $ 150 in comparison with the cost of the Tropicultor. Initially this objective appeared to have been achieved for in 1985 prices quoted by the Mekins workshop were $ 400 for the Nikart without implements and $ 600 for the Tropicultor without implements. (The implements were interchangeable, and the basic set for either was about $ 500 excluding a seeder.) This price differential had been maintained in deference to the assistance the workshop had received to start Nikart, but the quoted prices were largely theoretical as there was negligible demand for the Nikart.

In practice the savings in manufacturing cost of the Nikart due to lower weight and lack of wheel track adjustment, had been offset by the relatively complex system of height adjustment and the amount of precision welding required to manufacture the frame. In addition the early cost-saving device of the use of old car tyres for the Nikart had ceased due to problems of supply, quality and convenience of manufacture, and Indian-manufactured Nikarts were supplied with new Animal-Drawn Vehicle (ADV) tyres. The Mekins Director considered that the actual manufacturing costs of both the Nikart and the Tropicultor were similar, and by October 1986 the Mekins price differential between the basic toolcarriers almost disappeared in India at Rs 5 750 for the Tropicultor frame and Rs 5 500 for the Nikart frame ($ 500 for export sales). Other Indian manufacturers had previously also shown preference for the Tropicultor over the Nikart and unpublished data of Ghodake and Mayande (1984) suggested that even with economies of scale and the stimulus of competition, the supply prices of the Tropicultor and Nikart would be within 3% of each other.

In Europe in early 1987, the anticipated price savings of the Nikart design might be indicated if one were to compare the price of a GOM Toolcarrier with a simple set of implements (about US $ 1250) with a comparable Mouzon Tropicultor set (about US $ 1450). However any such price comparisons should be treated with great caution, since both the products and also the sales conditions of the two firms are very different, and both prices are liable to fluctuate with currency movements.

Operationally the Nikart was found to be effective, although even at an early stage it was found that few users changed between the cart mode and the cultivation mode (Kemp, 1983). While it was at first cheaper than the Tropicultor, at $ 400 - 500 for the basic carrier (without cart or implements) it was still very expensive. Thus even before the Nikart project had been completed, in 1978-1979 an even simpler tool, the Nolbar or Agribar was being developed.

4.3.5 The Agribar (Nolbar) wheeled toolcarrier (1978 - 1986)

The Agribar was the name given in 1981 to a derivative of the Nolbar. The Nolbar (presumably named after the designer Jean Nolle) had been tested at Patancheru in 1978, and in 1979 comparative trials had been carried out between the Nolbar, the Akola cart-based toolcarrier, the Tropicultor and a 22hp tractor. The Nolbar/Agribar had been designed to simplify still further the toolcarrier concept, and reduce cost (and flexibility) still further. It was designed as a simple, transverse toolbar (rather than a full chassis) pulled with a long, integral steel draw-pole. The bar is supported on two small (30 cm) wheels, with independent levers that raise or lower each end of the bar. On early models there was no operator's seat, and when one was provided it tended to give the driver a feeling of insecurity and instability. Handles in the centre of the bar can be used for implement guidance by an operator walking behind the toolcarrier. There is no provision to convert the bar to a cart. The attachments are the same as those for the Tropicultor or Nikart except that, being lighter, it cannot support as many soil preparation implements at the same time. In some respects the simplicity of the Agribar gives it some resemblance to the Ariana intermediate type of toolbar, but it differs significantly in that it uses a drawbar, and the toolbar can be raised and lowered.

In comparative trials in 1979-1980 the Nolbar/Agribar was found capable of all broadbed operations, but the time and effort required to raise and lower the implements at the end of each row made it less efficient in operation than the other toolcarriers. From 1978 to 1984 the Agribar was tested and adapted at Patancheru and was also (briefly) tested at Sotuba and Cinzana Research Stations in Mali. In 1985 it was tested by farmers in India but to date it does not appear to have been tested by farmers in Africa (ICRISAT, 1984 and 1985). In theory the Agribar is being commercially manufactured at the Mekins workshop, but to date total sales have been only thirty, of which fifteen have been exported for evaluation West Africa and Somalia. Priced at Rs 1 500 in India and at $ 200 for export (without implements), the Mekins Agribar is only 25-33% of the cost of a Tropicultor. Although it has been under development for nine years at ICRISAT farmer evaluation, sales and promotion have been minimal. In 1987 ICRISAT will publish a manual on its use, using the style of the Tropicultor manual.

On the ICRISAT station at Patancheru, the preferred toolcarrier has been the Tropicultor, and the on-station uses of this have been further diversified with the development and testing of prototype high-clearance pesticide sprayers and dust applicators and rolling crust breakers (ICRISAT, 1984 and 1985). At the ICRISAT research stations in Mali and Niger, the Nikart is preferred for its greater precision of depth control (see Chapter 5).

4.3.6 On-station and on-farm "verification" trials

Since 1975 wheeled toolcarriers have been used to cultivate over 100 ha of crops a year at ICRISAT's Patancheru research station (Bansal and Srivastava, 1981). From 1976 to 1981 the Farming Systems Research Program and the Economics Program of ICRISAT combined to evaluate at an operational scale the use of a complete package of "improved watershed-based technology" of which wheeled toolcarriers were considered an integral component (Virmani, Willey and Reddy, 1981; Ryan and Sarin, 1981). Small watersheds were systematically developed on the research station and from the carefully recorded and monitored trials it was clear that the combination of watershed bunding, the broadbed and furrow system using wheeled toolcarriers and the use of fertilizers and high yielding varieties produced significantly greater yields than traditional agricultural practices. This on-station work gave rise to great optimism, and a series of on-farm "verification" trials were initiated, in collaboration with Indian national programmes.

In 1978- 1979 ICRISAT supervised small plot experiments in the villages of Aurepalle in Andhra Pradesh and Shirapur and Kanzara in Maharashtra State. These were followed in 1979 to 1981 with the development of watersheds of about 12 ha in each village, and the use by farmers of the broadbed and furrow technology. ICRISAT provided all relevant inputs of equipment, fertilizers, seeds and pesticides (Sarin and Ryan, 1983). Early results suggested some problems with the technology, which had not proved successful in Alfisols (red soils), medium-deep Vertisols (black soils) or in areas affected by variable rainfall. Emphasis was therefore placed on the use of the technological package in deep Vertisols in regions of assured rainfall. The village of Taddanpalle (or Taddanpally) 40 km northwest of the ICRISAT Center was selected as representative of the appropriate conditions and in 1981 a watershed of 15 ha was developed by fourteen cooperatting farmers, with intense scientific and technical guidance from ICRISAT scientists (Ryan and von Oppen, 1983). The relative success of the first season's work at Taddanpalle led to a similar scheme in the nearby village of Sultanpur in 1982. A great deal of information was collected from the village studies and this showed there were both advantages and disadvantages to the new technology. As discussed in a following section, there are several examples of the positive aspects of the "on-farm verification" being selectively reported. However the final outcome in all the villages in which watersheds were developed is that none of the farmers continued with the technology and in general farmers were not prepared to buy or hire the wheeled toolcarriers, even at subsidized prices.

Thus by 1986 ICRISAT was not aware of any villages in India in which the wheeled toolcarrier and broadbed and furrow system had been proven by sustained farmer use and adoption. There were only a few examples of any use of broadbeds or wheeled toolcarriers. In one village, Antwar, about 100 km from Patancheru, three land-owning brothers had been experimenting with the broadbed and furrow system for three years and had obtained six Tropicultors and one Nikart. In this scheme the toolcarriers had been used on family land and had been loaned to twenty farmers without charge. The toolcarriers were only used as carts when all traditional carts were unavailable and in December 1986 the dry season cultivation of fields was being undertaken with traditional Desi plows due to the high draft of the toolcarriers. In 1986 visiting dignitaries were taken to this village as an example of the ICRISAT technology in use. Other examples of users of the technology in 1986 were also atypical and included a community research farm at Adgenar near Aurangabad, where the organizers and farmers are interested in the wheeled toolcarriers but none of the three toolcarriers provided by a development project had been used in 1986 (a dry year), and previous utilization rates had never been high. At the village of Neoli near Latur, the father of an ICRISAT researcher had purchased a Tropicultor and in 1986 used it for plowing about three hectares of upland rice (not on the broadbed system) and for fifty days of transport.

Such isolated examples indicate that to date "verification" (in the sense of farmers proving that the claimed benefits of a technology are real) has not yet been achieved. However this has not prevented some highly optimistic reports being produced as recently as September 1986 claiming that the wheeled toolcarrier technology has been "verified".

4.3.7 Optimistic economic studies on wheeled toolcarriers (1979 - 1986)

Relatively early in the ICRISAT research programme, studies were carried out on the economic costs and benefits of the use of wheeled toolcarriers (Binswanger, Ghodake and Thierstein, 1980). This study tried to estimate the hire rate a contractor would have to charge to pay for a toolcarrier over a period of ten years assuming he bought the toolcarrier with a commercial loan, and required a 10 - 20% profit over his actual out-goings. Several models with different assumptions were presented but although the toolcarriers were assumed to have significant working rate advantages over traditional implements, even a low cost toolcarrier (with steel wheels) with high utilization rates for agricultural use (eighty days a year) and transport (one hundred days a year), and only a 10% margin of profit appeared more expensive than existing hire rates for traditional cultivation services. Thus, while the toolcarrier could undoubtedly save time and drudgery, it was concluded "even under the most favourable circumstances assumed such machines cannot compete on a cost basis with the traditional implements in traditional agriculture."

Binswanger et al. noted that there was a social cost involved, for wheeled toolcarriers would make 1.5 bullock drivers unemployed on each 15 ha on which it was assumed they would operate. However the authors noted that the toolcarriers might become both socially and economically justified if there were compensatory yield increases. If such increases were large enough they could gene" rate sufficient extra work to offset the unemployment of the bullock drivers. In on-station research carried out between 1976 and 1978, significant yield advantages had been attributed to the soil management systems associated with the toolcarriers, and while these had not been fully verified in on-farm conditions, there was an indication that particular benefits might be achieved on the deep Vertisols (black soils). Thus the authors concluded that on-farm research relating to wheeled toolcarriers was amply justified, but cautioned that wheeled toolcarriers would not be competitive unless they could generate yield advantages in excess of 200-400 kg/ha (Binswanger, Ghodake and Thierstein, 1980).

Binswanger et al. intentionally avoided the problem of relating farm size to toolcarrier ownership by assuming that a contractor would be able to hire out such an implement to several farmers and thereby cultivate a to: tal of 15 ha. This has been considered to be a realistic maximum for the area that could be cultivated with a toolcarrier and-this figure allows costs per unit area to be minimised. However another ICRISAT worker discussed this particular problem, noting that the majority of farmers in India have much smaller farms than 15 ha (Doherty, 1980). Doherty argued that small group ownership of toolcarriers would be sociologically difficult and if large groups could be formed they might find greater benefits from tractor ownership. He also argued that farmers prefer individual ownership of implements to hiring from entrepreneurs. Doherty pointed out that some of the assumed potential yield advantages of the toolcarrier would come from the associated soil management techniques involving developing small watershed areas. However he highlighted the likely social problems of redeveloping drainage patterns between farms owned by different families of different social and economic backgrounds. Thus, while also advocating more on-farm research in this area, he emphasised the need for developing low cost implements that could be afforded by individual farmers, on-farm yield increases that could justify the investment and socially viable systems for transferring such technology (Doherty, 1980).

Despite the cautions of Doherty voiced in 1979, from 1979 to 1985 ICRISAT economists continued to base economic assessments of toolcarriers on the "optimising" assumptions of Binswanger et al. (1980).

An example of the optimism of ICRISAT economists is seen in the paper of Ryan and Sarin (1981) who stated: "We discuss the economics of the improved technologies that have been evolving from research at ICRISAT Center and in villages, aimed et enabling crops to be grown in deep Vertisols in the rainy season. . . This improved system utilizing graded broadbeds and furrows has generated profits . . . These profits represent a return to land, capital and management, as the cost of all human and animal labor, fertilizers, seeds and implements have been deducted . . . Based on these figures the extra profits from the new system could pay for the wheeled toolcarrier in one year provided that it was utilized along with improved technology on at least four hectares." Although the details of the cost assumptions used in the calculations were not provided in these papers, the profits quoted were based on "annual costs of implements". Towards the end of the paper the high cost of the toolcarrier was acknowledged, but it was pointed out that attractive rates of return would be available to entrepreneurs hiring out wheeled toolcarriers for 180 days a year.

Perhaps the most optimistic economic analyses by ICRISAT were presented by Ryan and von Oppen in 1983 and were based on initial on-farm verification. Referring to results from the village of Taddanpalle for 1981 - 1982, the authors stated: "These data show a 244% rate of return on the added expenditure, confirming the experience at ICRISAT Center (250%), and giving us confidence about the technology options on village farms . . . The relative success of the Taddanpalle experiment led to a further experimental area in adjoining Sultanpur village in 1982-83."

They then attempted to make a benefit-cost analysis, admitting that at the early stage of adoption, this was a hazardous exercise. The assumptions included an annual growth of toolcarriers of 45% per year (rising to 0.5 million units in use in the year 2003) and additional profits based on Taddanpalle experience of Rs 1434/ha. This gave a benefit to cost ratio of 5: I by the year 2000 if each toolcarrier could work on ten hectares (a 300% internal rate of return), and 7: 1 if the toolcarriers were used on fifteen hectares. The additional costs of the provision of extra agricultural officers, fertilizers stores and banks to service the new technology were not included, nor were any benefits attributable to soil conservation considered (Ryan and von Oppen, 1983).

Highly optimistic economic statements relating to wheeled toolcarriers continued to be made by ICRISAT economists until 1985. Ghodake (1985) drew heavily on the content of Ryan and Sarin (1981) and repeated the suggestion that a wheeled toolcarrier' could be paid for in one year on four hectares although he did note that the wheeled toolcarrier might not actually be an essential component of broadbed technology for which it was being advocated.

The agricultural engineers at ICRISAT have seldom included any economic data in their reports and papers. However, in 1985, a paper was published giving an economic comparison of the Akola toolbar, the Tropicultor, the Nikart and the Agribar. Assumptions were based on 14 ha annual use, plus 400 transport hours for the toolcarriers that could be used as carts. With these assumptions the Tropicultor had the best marginal benefit-cost ratio attributable largely to the reduction in hourly cultivation costs achieved by assumed transport operations. However, in terms of simple cost per hectare, the Agribar appeared most promising, and was suggested as a low cost alternative to the heavier machines for the broadbed technology (Mayande, Bansal and Sangle, 1985). In another approach, wheeled toolcarrier technology was promoted for its energy efficiency (Bansal, Kshirsagar and Sangle, 1985).

4.3.8 General promotion of toolcarriers by ICRISAT (1981-1982)

While most of ICRISAT's work on wheeled toolcarriers had actually been based on the broadbed and furrow system of cultivation, and their economic justification derived from on-station trials using that system, ICRISAT publications started to consider wheeled toolcarriers as a valuable technology in their own right. Thus Information Bulletin No. 8 on "The Animal-Drawn Wheeled Tool Carrier" (ICRISAT, 1981) stated:

"The animal-drawn wheeled tool carrier . . . is able to perform virtually all operations that can be done with a tractor, thus providing to many farmers the versatility and precision previously available to only a few. . . The present multipurpose machine permits farmers to carry out their basic operations of tillage, planting, fertilization and weeding in a timely and precise manner to increase productivity and, as a bonus, it can be used as a cart to provide transportation. . .

Such a system of machinery promotes agriculture by increasing farmers' income and making available to them machinery that enables:

The picture presented in this Bulletin of what seemed almost ideal equipment, perhaps a panacea of agricultural engineering, was short-lived, as feedback reached ICRISAT from village experiences.

In response to the general promotion of wheeled toolcarriers by ICRISAT and cooperating manufacturers, a large toolcarrier project (the largest to date in India) was undertaken in Nasik District of Maharashtra State. In the planning stages it was envisaged that 350 Nikart toolcarriers would be sold, but when offered the choice of Tropicultors and Nikarts, the farmers opted for Tropicultors. In 1982/1983 about 266 farmers had been sold Tropicultors at 80% subsidies under the Maharashtra Integrated Rural Energy Project. The toolcarriers had been supplied complete with plow bodies, tines and carts, and in line with the promotion for general use there had been a clear emphasis on the transport potential of the toolcarriers (Fieldson, 1984; Kshirsagar, Fieldson, Mayande and Walker, 1984).

It is illuminating to follow the progress of this scheme. After only one or two seasons, by 1984 few farmers used the Tropicultors on any significant scale for cultivation, generally perceiving them as too heavy and the implements not suited to local soil conditions. By 1986 it was relatively difficult to find any farmers who used their Tropicultors for cultivation. One farmer was specially contacted because he reportedly still used his toolcarrier, but in practice he only used the Tropicultor on one small plot and it was clear from the lack of wear on the implements that they had not been extensively used since manufacture. Many farmers had stopped using their Tropicultors even as carts, preferring the more stable and more easily repairable traditional carts. During village visits in 1986 several Tropicultor carts were seen to be still in use, but more significantly, considering the cost of the toolcarriers and research predictions concerning potential for transport use and life expectancy, abandoned frames and cart bodies were also seen. Thus this general promotion project showed a pattern very similar to some of the early African schemes: an early rejection of toolcarriers for cultivation and a slower abandonment for transport purposes. This has implications for both technical and economic assessment, for if farmers actually own implements but stop using them, the problem is not simply one of cost or profitability for they have already invested in the technology. It implies some technological problems relating to the use of wheeled toolcarriers in local farming systems and village life.

As the results of the on-farm trials and promotional schemes became known to ICRISAT scientists, doubts slowly started being expressed in papers and publications.

4.3.9 Doubts relating to wheeled toolcarriers (1981-1986)

Doubts about the overriding economic advantages of wheeled toolcarriers only slowly entered the ICRISAT literature. Ghodake, Ryan and Sarin (1981) warned that exacerbated labour bottlenecks could lead to the rejection of broadbed technology. Sarin and Ryan (1983) noted that on-farm verification trials in Alfisols (red soils) in Aurepalle village near Hyderabad had failed to show advantages for the broadbed and furrow technology. In Shirapur village in Maharashtra State the deep Vertisols (black soils) were too hard to allow plowing with wheeled toolcarriers and a single pair of bullocks, and the toolcarrier could not control weed infestation on the raised beds. In medium-deep Vertisols at Kanzara village in Maharashtra State plowing with the wheeled toolcarrier required multiple pairs of bullocks and did not lead to greater profitability when compared with traditional techniques. It was concluded that while wheeled toolcarriers were efficient, less costly alternatives should be explored (Sarin and Ryan, 1983).

Further questioning of the applicability of the station-derived technology was provided by von Oppen, Ghodake, Kshirsagar and Singh in 1985. The authors confirmed that the Vertisol technology had been consistently successful on station but admitted that "the continuing need for management support, and input supplies and the emergence of further constraints seem to impose much narrower limits on the technology than had earlier been anticipated.)' Constraints identified by on-farm trials included exacerbated human labour peaks, bullock power and fodder constraints, inadequate credit, difficulties in fertilizer supply, increased weed growth and technical problems of repairs and maintenance of wheeled toolcarriers. It was noted that the farmers involved in the on-farm verification trials did not consider wheeled toolcarriers as indispensable to the broadbed and furrow technology, and were not prepared to pay realistic hire or purchase costs for the wheeled toolcarriers. It was concluded that further research was needed into the various components of Vertisol technology, including the development of lower cost wheeled toolcarriers. It was also suggested that such research should be carried out in closer cooperation with farmers, perhaps by national programmes rather than by ICRISAT.

Hints of possible doubts entered the Information Bulletin No. 8 on "The Animal-Drawn Wheeled Tool Carrier" between the 1981 and 1983 editions (ICRISAT, 1981 and 1983). Many changes between the two editions were small and provided additional technical information relating to the toolcarriers, such as weight, use of the Nikart and the additional operation of land shaping. Small subtle changes were related to possible problems when the toolcarriers are used off the research station, for example indicating that farmers must adjust the load to the capacity of their animals. However perhaps the most important change was that, while the 1983 booklet was still very positive and stressed the potential benefits of toolcarriers, it also had a new heading "Drawbacks of the toolcarrier" which noted that they cost more than small farmers could normally afford and their maintenance might be difficult under village conditions. The 1981 conclusion that "such a system of machinery promotes agriculture by increasing farmers' income" was subtly modified to "in the long run it can increase agricultural production and farmers' income particularly in regions where there is a high ratio of land per farmer."

This last change is interesting as in much of India holdings are small, and the ratio of land to farmer is generally higher in Africa and Latin America than Asia. The 1983 toolcarrier promotional booklet (ICRISAT, 1983) was also given a very distinct change in its overall impression through the inclusion of photographs of toolcarriers in use in Brazil, Botswana, Mexico and Mozambique in addition to India. This reflected the increasing interest of ICRISAT in the potential for toolcarriers in other parts of the world, in addition to their use in India. However it also tended to create the impression that the technology had diffused worldwide.

The greatest doubts to date have been expressed in the report of the British NIAE by Fieldson (1984) and the resulting paper by Kshirsagar, Fieldson, Mayande and Walker (19.84). These observed that few wheeled toolcarrier machines had been sold in India without large subsidies; annual utilization had been low; hire markets had not developed; farmers did not perceive that the wheeled toolcarriers had overriding advantages over traditional implements; most manufacturers had stopped making wheeled toolcarriers due to insufficient market demand and future prospects were not bright.

4.3.10 Continued optimism (1985-1986)

Despite the doubts expressed in internal papers, few externally circulated ICRISAT papers have shown any indication of the problems being faced in the field by wheeled toolcarriers. In a paper presented at a seminar at IRRI in 1985, ICRISAT staff managed to cite Fieldson's very pessimistic report and still present a very optimistic overall picture: "Now the farmers in SAT regions of India have started appreciating the usefulness of WTC. This trend is rather encouraging. It reflects the collaborative efforts by the Government extension agencies and national research institutions. Occasional subsidies from the Government also assist. As a result of all this the sale of WTC in India is improving, even though direct purchase by individual farmers and non-governmental agencies is only about 11% (Fieldson, 1984)." (Awadhwal, Bansal and Takenaga, 1985).

In April 1986 an article in the newsletter of the Regional Network on Agricultural Machinery (RNAM) described the farm machinery research of ICRISAT and the development of wheeled toolcarriers (Bansal, 1986). No mention was made of farmer response to the wheeled toolcarriers, and the impression was given that they were being increasingly used by Indian farmers. Most recently three ICRISAT scientists participated in the "Animal Power in Farming Systems" workshop in Sierra Leone in September 1986 and presented a highly positive picture of the progress of wheeled toolcarriers in India (Bansal, Klaij and Serafini, 1986). The overall optimistic tone presented can be gauged by the following quotations:

"In the past decade a successful "technology package" for Vertisols was developed . . . The WTC has been used to overcome the problems of working these soils . . . After the successful experiences at the ICRISAT Centre with the Tropicultor . . . The Mkart is about $ 80 (US) less expensive than the Tropicultor that cost $ 500 (US) . . . It is also well suited to the manufacturing capabilities of small industries in developing countries. At the ICRISAT Centre animal-drawn WTCs have been successfully integrated in improved farming systems developed for the management of Vertisols. On-farm verification has been carried out in different regions of the Indian SAT. Data from two villages, Taddanpally in Andhra Pradesh and Farhatabad in Karnataka State, illustrate the role of improved farm equipment in a new farming system. . . In Taddanpally . . . the use of the WTC led to substantial labor savings for field operations . . . higher yields . . . increased labor productivity . . . ICRISAT has demonstrated that a properly conceived animal-traction-based crop management strategy can have significant impact on productivity." (Bansal, Klaij and Serafini, 1986).

There is no hint in the paper of the problems being experienced with the adoption of wheeled toolcarriers in India or that farmers at Taddanpally and Farhatabad did not continue to use the "successful" technology, after its "verification". Nor was there any indication that the "$ 500" Tropicultor had no implements and would actually cost four times this figure shipped with implements to a West African port.