Household survival in marginal environments such as mountains, dry lands, and flood prone regions requires tremendous creativity. As was noted in Alice in Wonderland, you have to move very fast and work very hard even to remain where you are. The choice for large number of households is to sustain the livelihood support systems such as the catchments, biodiversity, other natural resources, etc., in a manner that they do not get trapped in downward spiral of erosion of resources, self-esteem, and of course, economic opportunities. The fact that despite various odds, including lack of policy support, so many communities and individuals manage not only to conserve resources but also augment them is something that this monograph is all about. The Honey Bee perspective builds upon what poor people are rich in i.e. their knowledge, creative potential, and institutional heritage. The discourse on participation often is restricted to the concept of either physical participation in terms of labour or social participation in implementation of externally designed policies and programmes. In this study, we draw attention to the scope of intellectual, moral, and institutional participation of local communities in reconceptualizing the watershed approach and implementation process. The greatest irony of watershed projects is that they founder after they are ‘handed over’ to the people by the project implementation authorities. If the watershed projects are designed, owned and implemented by the people, why should the question of handing over arise at all. Unless we, the external facilitators, learn to participate in peoples’ own plans (Gupta, 1995), the possibility of building upon peoples’ knowledge is very remote.

It is extremely opportune that international and national institutions are recognizing the need for incorporating indigenous knowledge and institutional heritage in the design and implementation of modern watershed projects. This blending of traditional knowledge and contemporary innovations developed by people without outsiders help will not take place unless we understand the policy and institutional context of technology generation and diffusion for rainfed, mountain, and dry regions. The macro policy and the framework for organizing incentives to ensure peoples’ participation in design and implementation of watershed are discussed in part one. In part two of the paper we critique the formal models of technology development and transfer. We argue that technology development process in highly ecologically heterogeneous environments cannot take place in the classical lab to land framework. It will require land to lab to land, and land to land approaches (Gupta, 1987, 1989, Richards, 1985. The last part three deals with the framework for institution building in watersheds. The contention here is that self regulating behaviour is essential for managing natural resources in the long run. We deal with the institutional aspects of watershed development. Here we focus on two particular aspects, (a) institutional triggers for technological solutions and (b) technological triggers for institutional innovations. This is a relationship which has not been adequately appreciated while designing policies and programmes for watershed management in various countries. In part four, we provide illustrations of more than fifty technological and institutional innovations from Himalayan region as well as western Indian dry regions.

The traditional models of on-station development of technology and its transmission to farmers are no longer feasible, since high ecological variability demands niche-specific solutions. Local solutions developed by farmers themselves need to be identified and their scientific bases understood. The value-added scientific principles have to be shared back with farmers, who would then be able to develop technologies through their own research and experimentation. Thus transferring ‘science’ and not just technology (Gupta, 1989a & 1994b). Supporting and developing such experimentation is an important task for scientists and outsiders. Perhaps the most crucial challenge is for scientists to realize that how they can participate in people’s programs rather than asking how people can participate in formal outside initiatives.

This change in outlook, within less than three decades of the onset of the green revolution, is a result of the increasingly complex interactions between local socioecological and institutional conditions, and externally-induced technological change. In other words, the challenge technology designers face today is how to move away from delivering fully-tailored cloth towards supplying semi-stitched cloth which may be tailored by users themselves, keeping local specifications in mind (Kumar, 1985 p c ). This requires both an understanding of the tailoring process on the part of the people, and an understanding of local preferences, criteria and specifications on the part of researchers.

Another reason for seeking participation is that it provides opportunities to scientists to recalibrate their scales of measurement and co-ordinates of perception. Perhaps what is more important is developing in scientists the ability to learn how to participate in the plans, programs, experiments and missions of farmers themselves (Gupta 1980, 1987b, 1995d). Ashby et al. (1987) had rightly criticized the excessive emphasis on the so-called diagnostic research methods that treated farmers as objects of investigation and in the process lost the farmers’ voice. She emphasized that participatory research should involve farmers as co-investigators and researchers, and demonstrated, through farmer-managed trials, creative ways of understanding farmers’ criteria for selecting varieties. Gupta (1987d), while describing the dynamics of homestead utilization by women, provided examples of the criteria used by poor women in the management of sweet potato seedlings, that had never formed a part of formal scientific research. There are many other examples, including the excellent research of Richards (1985, 1987), that demonstrate the need for scientists to participate in farmers’ own research programs.

However, any process of collaborative learning can be meaningful and mutually enjoyable only when the classificatory schemes or taxonomies used by the partners are matched. It is not necessary to synthesize these taxonomies, but it is essential to understand the various vectors on which each knowledge system organizes information and generates patterns of knowledge. Does it matter in a dialogue between farmers and scientists in Peru whether the potato is distinguished by its local name, Puka suytu, or only by its Latin name, Solanum tuberosum (Vasquez 1996)? It does not when two classificatory schemes are mere tools to highlight the strengths of the knowledge systems on which they are based. But when one system’s superiority is asserted, or when scientists use scientific language to mask their inability to understand the richness of the vernacular, there is a problem.

A second aspect of matching taxonomies is the need for formal science to realize that an indigenous taxonomy would be extremely rich when the variance in any phenomenon critical for the survival of that community is high. The community breaks down the phenomenon into a larger number of discrete categories, and characterizes each category by a different name. Thus, for instance, Eskimos have a large number of words for snow, and fisher folk many names for varieties of waves. Each category symbolizes not only a pattern but also a theory underlying the classification and interrelationship of different categories.

Often, uncovering the farmers’ own experimental approaches and heuristics may be sufficient to help them to redefine the problem and devise appropriate solutions (Gupta 1989c, Gupta 1989d, Pastakia 1995). But in some cases, farmers cannot devise solutions on their own. On-station research becomes necessary and farmers will have to merely participate in evaluating results or monitoring the experiments for any counter-intuitive observations. Normatively, we should not consider one form of participation superior to the other. Thus, farmers’ participation in the scientists’ own experiments need not necessarily be superior to scientists’ participation in farmers’ research. Both forms have their own advantages and limitations. In order to evolve a contingent framework, it is necessary to match the different methods of participation with the different approaches to defining the purpose of participation. The same method, say on-farm research, may not address all kinds of problems.

It is a truism that the proper definition of a problem is half the solution. And yet, very often, we do not know whether our definition of the problem is correct or not. Let us take the case of weeds, which are considered to be a menace in rainfed crops. In the conventional definition, weeds are plants out of their place. But in nature, no plant can truly be out of its place. It is possible that we may not know the significance or role of a particular weed as a companion plant. For instance, the distribution of minerals in a field may help certain plants grow faster or slower. Thus, weeds may act as indicators of soil mineral properties (Hill & Ramsay, 1977). If we know the variability in the soil nutrient profile, we can follow precision farming that will lead to economy and efficiency in input use. Once the existing heterogeneity of nutrients is known, it is possible to study the reasons and take remedial action. Another way to look at weeds is to ask ourselves why farmers are selective in removing weeds. They obviously must be recognizing the allelopathic interactions of various plants. A good example is a weed (companion plant) called Sama (Echinocloa colonum) which grows on its own in paddy fields, or is cultivated in certain parts of the country. Why would farmers conserve a ‘weed’? There may be several reasons: (a) it is an extremely nutritious grain suitable for consumption during fasting (b) a review of literature shows that it provides an alternative host for a few insects including leaf roller which do not affect paddy crop but get attracted to Sama and (c) some other ecological function which we are not aware of as yet. It is not without significance that farmers have conserved this weed through sociocultural mechanisms such as a particular festival, Sama pancham, when only grains like Sama are eaten. If sustainability requires a long time frame and a wide variety of heuristics through which our choices should be processed, then a strong case exists for understanding how farmers define a particular problem (Gupta 1981, Gupta et al., 1995).

Primarily drawing upon the Honey Bee database, Pastakia (1996) studied grassroots innovators involved in sustainable pest management in order to understand their decision making processes. He identified two particular heuristics which were not reported in the formal scientific repertoire: (i) use of insect and plant material for repelling pests and (ii) increasing the growth of a crop to minimize economic damage by a pest instead of controlling the pest itself. The heuristics that the innovators used to derive such solutions included various combinations of materials (or products), methods (or processes) and products, each of which had a sustainability dimension determined by the renewability of the resources involved (Figure 1). An analysis of a farmers’ heuristics in these three dimensions of Product, Process and Purpose as shown helps us in understanding firstly, where the innovation was actually done and secondly, how best the modern science can intervene to improve upon.

Figure 1. Combinational heuristics

Old methods, old material and old products. Old methods, old materials and old products signify the traditional wisdom which may have relevance even for the contemporary context. For instance, Virda is an age-old technology for conserving rain water in a saline arid region with saline ground water. In a predominantly flat region, rain water gets stored in minor depressions or tanks. Within these tanks, the pastoralists dig shallow wells lined with frames of wood of Prosopis juliflora and grass. Just ten inches of rainfall provide sufficient fresh water which remains above the saline ground water inside the wells. The virdas are covered with silt and sealed. They are opened, one at a time, depending upon the need. The water remains sweet for two to three months, after which it turns saline due to the upward movement of saline water. This technology has enabled the pastoralists in Banni pastures to survive for several centuries. The season’s rain may fall within a few days, hence the need for a robust, efficient and adaptive strategy (Chokkakula & Gupta, 1995; Ferroukhi & Suthar, 1994).

In such a case, modern science does not merely help explain the functional viability of the technology, but also provides a basis for abstraction and generalization. For instance, once the properties of wood and grass, the pressure that the walls will need to cope with, the infiltration rate and the functions of the saline soil in holding the salts are explained, the search for other materials and methods for similar outputs may begin. There is very little advantage that the prior art of knowledge in modern science can provide while dealing with such complex questions of survival in difficult regions.

Old methods, old materials and new products : The hair which constitutes the mane of camels is known to be very hardy and resistant to corrosion. Traditionally, the pastoralists make different kinds of ropes, carpets and bags out of this hair. Once science figured out the use of these carpets as oil filters in oil refineries, a new product was developed from the old method and material. Similarly, sisal rope has been used in various activities, both for commercial and domestic purposes. It was found that these ropes can withstand corrosion better than any other material in the sea. Thus a new use for material grown in poor soils is generated. The processing of sisal is very painful because of the various tannins released into the water in which sisal plants are immersed for some time. When the fibre is taken out, these tannins cause blisters on the hand. Simple technologies have been developed to take the fibre out without hurting the hands. Modern science can blend in with the traditional methods while leaving other choices intact.

New methods, old materials and old products : In many of the cumin-growing regions, farmers had observed that the plots on the roadside were more productive than the ones in the interior. They figured out that the dust which settled on the plants saved them from certain pests and fungal diseases. Some other farmers observed a similar phenomenon near brick kilns. Dusting with ash or fine soil thus became a new method for controlling pest and fungal diseases in this crop. In many other crops, the use of ash as a dusting material is well known.

Similarly, the case of termite control using cut immature sorghum stalks in irrigation channels, reported earlier in this paper, opens up a new field of research. So far, sorghum breeders had been looking for landraces with a low hydrocyanide content. This innovation opens up the opportunity for selecting high hydrocyanide content sorghum lines. If this technology works in different parts of the world, dry farmers may very well grow a small patch of such sorghum for pest control purposes.

Old methods, new materials and new products or uses :Some innovative farmers have used a drip of castor oil (a tin box with a wick hanging over an irrigation channel). The oil drips into the water and spreads into the soil, adding luster to the banana crop. This drip is also used in other crops for soil-based pest control.

Examples for other combinations are listed in the table below. What these examples show is that farmers can be extremely creative in solving local problems. But the issue is whether their knowledge systems can be blended with formal scientific research. One block may possibly be the tension between the farmers’ interest in solving the problem and the scientists’ interest in developing a new theory. For instance, a farmer, Khodidasbhai, after reading about three different practices for controlling a pest in a local version of Honey Bee, used all three on the same crop, in the same season, but sequentially. It is quite possible that scientists would not attempt such an experiment in order to avoid a complicated design with confusing results. Learning to break old rules, which formal training does not easily permit, can be a useful purpose of participatory research.

1.3 The Threats to Local Knowledge: The Case of Honey Bee Network

Erosion of knowledge is as much, if not more serious problem than the erosion of natural resources. We can probably reverse the declining productivity of natural resources like soil through watershed projects or other resource conservation strategies. However, erosion of knowledge can not be easily reversed once lost. The regeneration of resources and knowledge associated with these resources have to be seen in a single as well as multiple generation framework (Gupta, 1990, 1992, 1996, Gupta et al , 1994).

Consider first the single generation situation. The ideal sustainable situation occurs when both resources and knowledge have been conserved, but what happens when one or the other is eroded.

When the resources are conserved and the knowledge is eroded (as in the case of state-controlled conservation of resources through parks or sanctuaries keeping people out of the resource), the sustainability of the system becomes endangered. If knowledge is eroded, the erosion of resource can’t be far behind.

When the knowledge is conserved but the resources are eroded, the sustainability of the system is more likely if local knowledge is incorporated in strategies of regeneration. The knowledge will also be eroded, however, if it is not used.

The least sustainable single generation situation occurs when both the resources and the knowledge become eroded. The folk knowledge once eroded may be almost impossible to reconstruct or rejuvenate. Erosion of knowledge was never so rapid as in our generation because of declining inter-generational communication.

As bleak as the single generational picture is, consider now, the multi-generational situation. Again, the ideal situation occurs when both knowledge and resources have been conserved.

The situation where knowledge has eroded and resources have been conserved is not a likely scenario. This is so because a re source cannot be sustained over generation without drawing upon local knowledge at all. Under conditions of no human intervention or access, certain resources like forests may be conserved over generations without incorporating local knowledge. But with the increasing influence of human-made factors on the survivability of forests through acid rains, global warming, and erosion of upper catchments etc., as well as increasing population pressures, we doubt such a situation could occur.

The case of erosion of resources and the conservation of knowledge over several generations leads to a possibility of sustain ability if knowledge has been documented through efforts like the Honey Bee network and is available to people, regeneration of resources is possible within a long time frame.

The worst case of all occurs when both knowledge and resources have become eroded over several generations. Only rare repositories of knowledge may exist among some bypassed communities.

Whether the analysis is performed in a single or multiple generational setting, the key is the same. The conservation of knowledge is as important as the conservation of resources, if not more so. Thus, any system of conservation should be directed not only at rewarding communities for the conservation of resources, but also at rewarding them for the valuable knowledge they hold, create and recreate.

In the context of the biologically rich, low-mean/high-variability income areas discussed earlier, emphasis is placed on providing short-term relief, employment, and other means of subsistence in high-risk environments in order to alleviate poverty. The economic stress on the community erodes their self-respect and dignity. The will of the people to struggle and innovate gets subdued. Both the resource and, the knowledge around this re source get eroded.

In order to stem knowledge and resource erosion, the Honey Bee network, a global voluntary initiative was launched nine years ago. Its purpose is to network the people and the activists engaged in eco-restoration and reconstruction of knowledge about precious ecological, technological, and institutional systems used by other people.

This network aims at identifying the innovators (individuals or groups) who have tried to break out of existing technological and institutional constraints through their own imagination and effort. What is remarkable about these innovations is the fact that most of these require very low external inputs, are extremely eco-friendly and improve productivity at very low cost.

It is necessary to note here that organizations of creative people, which take the form of networks or informal cooperatives or just loose associations, would generate a very different kind of pressure on society for sustainable development. The spirit of excellence, critical peer group appraisal, competitiveness and entrepreneurship so vital for self-reliant development, may emerge only in the networks of local ‘experts’, innovators and experimenters. It is true that every farmer or artisan does experiment. But not every one is equally creative and not in the same resource-related fields. The transition of the developmental paradigm from ‘people as victim’s perspective to that of the people as potential victor’s is the answer. Former may generate patronizing and externally driven initiatives where as latter may spur endogenous initiatives by people themselves.

Honey Bee network newsletter is brought out in seven languages in India (English, Hindi, Gujarati, Kannada, Tamil, Punjabi and Telugu) and Dzonkha in Bhutan so that dialogue with the people takes place in their own language. The creative people of one place should be able to communicate with similar people elsewhere to trigger mutual imagination and fertilize respective recipes for sustain able natural resource management. The Honey Bee network is head quartered at SRISTI (Society for Research and Initiatives for Sustainable Technologies and Institutions c/o Prof Anil K Gupta, Indian Institute of Management, Ahmedabad),an autonomous NGO.

It is realized that the technological innovations cannot survive without institutional innovations and support structures. Hence we have been documenting the ecological institutions which have been evolved by the people to manage knowledge and resources as common property.

Honey Bee insists that two principles are followed without fail: one) whatever we learn from people must be shared with them in their language, and two) every innovation must be sourced to individuals/communities with name and address to protect their intellectual property rights.

It is possible to take the current global debate on biodiversity and peasant knowledge beyond rhetoric. Our network extends into 75 countries at present. Some of the colleagues have started similar documentation in their respective regions. Offers have been received from Nepal, Sri Lanka, Uganda, Paraguay and Mali for local language versions.

Honey Bee also appeals to fellow researchers, activists and planners in other developing countries to identify native wisdom both to inspire and also to provoke the young minds to explore. In every country a very strong oral tradition of knowledge generation, validation, scrutiny and diffusion exists. Honeybee strongly believes that boundaries between formal and informal knowledge systems may often be false. The informal system may have formal rules waiting to be discovered. The formal system may have informal beliefs, accidents, or conjectures providing impetus for further inquiry.

Honey Bee has already collected more than five thousand innovative practices predominantly from dry regions to prove that disadvantaged people may lack financial and economic resources, but are very rich in knowledge resource. That is the reason we consider the term ‘resource poor farmer’ as one of the most inappropriate and demeaning contributions from the West. If knowledge is a resource and if some people are rich in this knowledge, why should they be called resource poor? At the same time, we realize that the market may not be pricing peoples’ knowledge properly today. It should be remembered that out of 114 plant derived drugs, more than 70 per cent are used for the same purpose for which the native people discovered their use (Farnsworth, 1988). This proves that basic research linking cause and effect had been done successfully by the people in majority of the cases. Modern science and technology could supplement the efforts of the people, improve the efficiency of the extraction of the active ingredient or synthesize analog of the same, there by improving effectiveness (Gupta, 1991).

The scope for linking scientific search by the scientists and the farmers is enormous. We are beginning to realize that peoples’ knowledge system need not always be considered informal just because the rules of the formal system fail to explain innovations in another system. The soil classification system developed by the people is far more complex and comprehensive than the USDA soil classification systems. Likewise, the hazards of pesticides residues and associated adverse effects on the human as well as entire ecological system are well known. In the second issue of Honeybee out of ninety four practices thirty four dealt with indigenous low external input ways of plant protection. Some of these practices could extend the frontiers of science. For instance, some farmers cut thirty to forty days old sorghum plants or Calotropis plants and put these in the irrigation channel so as to control or minimize termite attack in light dry soils. Perhaps hydrocyanide present in sorghum and similar other toxic elements in Calotropis contributed towards this effect.

Honeybee in that sense is an effort to mould markets of ideas and innovations but in favor of sustainable development of high risk environments. The key objectives of SRISTI thus are to strengthen the capacity of grassroots level innovators and inventors engaged in conserving biodiversity to (a) protect their intellectual property rights, (b) experiment to add value to their knowledge (c) evolve entrepreneurial ability to generate returns from this knowledge and (d) enrich their cultural and institutional basis of dealing with nature.

Of course no long term change in the field of sustainable natural resource management can be achieved if the local children do not develop values and a world view which is in line with the sustainable life style. Thus education programs and activities are essential to perpetuating reform. That is also the reason why we have organised biodiversity contests among school children to identify little eco-genius.

Sustainability of some of the traditional soil and water conservation structures in many mountain regions, dry regions and other areas has come under stress in recent times. And yet, there are few contemporary institutional models that have survived one generation without any decline in the quality of leadership or management of resource. Many of the traditional institutions have worked successfully for several generations and through small innovations - or improvements from time to time in technology as well as institutional processes. Many of the modern projects seem to be designed for failure after the project management team withdraws from the scene. How do we avoid spawning failure and ensure not just success but a sustainable success in watershed project is the purpose of this note.

Our contention is that there are time tested processes of institution building which somehow have never received adequate attention in watershed projects. The results are obvious. Extremely good and effective watershed projects have faltered when external interventions or incentives are withdrawn as if people were implementing somebody else’s project. In some cases when projects have indeed sustained their effectiveness, the cost at which the success has been achieved has been ignored. In still other cases, the innovations in the process underlying the successes have never diffused even to the neighboring villages. This evidence puts the question mark on the very strategy of establishing demonstration watershed projects. Nobody ever expected in the canal irrigated regions that after looking at the advantages of canal irrigation, farmers will on their own design and manage secondary and tertiary irrigation channels (Gupta, 1996). And yet, in watershed projects such an assumption is made despite considerable evidence to the contrary. This paper therefore also suggests the limits of institution building process and need for complementarity between internal and external incentives for managing watersheds in stressed environments.

About eight years ago in an action research project in dry-land regions of Karnataka, we asked a question in a village meeting, "What were the activities which villagers have done collectively without any outside help?" The answers were very instructive as expected. Different villagers had a strong tradition of collective action in religious, cultural and socio-economic fields. In one village, the people had organised a rotating saving and credit association. The discount money from the chits was not distributed as dividend. This was used to build temple and buy necessities for the local primary school. In many other villages people have managed common breeding bull, a tank, common land for compost pits, common drainage, temples, etc. And yet, when we design watershed projects, we never look into the processes and the dynamics of these existing institutions.

There is a considerable research done on crafting institutions (Ostrom, 1992). And very little on ‘grafting’ institutions. Whenever we initiate a collective institution in any village we obviously don’t begin in vacuum. There is a history of people working and not working together and watershed project has to deal with this history explicitly. The so-called participatory technique by missing the issue have failed in generating an organic fusion or blend between traditional and modern institutions. Fifteen years ago we came across an interesting example of this fusion in a village in Ahmednagar district of Maharashtra. In a dry land village, people had planned planting of tree seedlings on an auspicious day as a part of watershed project. They wanted to carry the seedlings in a cradle, normally used for carrying idol of the local deity on religious festivals. Important dignitaries had been invited next day for the function. However, during the previous night when discussions were going on in the temple premises about the arrangements, somebody raised the issue of impurity of soil and thus impossibility of using the cradle meant for deities for this purpose. Everybody was perplexed. They did not know what to do in the available time. A carpenter’s son belonging to lower cast was standing at the gate of the temple and listened to this question. Being a person of lower caste, he was not allowed to participate in the discussions. However, he pleaded with the people to be given a chance to solve the problem. He knew of a cradle lying in somebody’s house unused. This cradle originally meant for the children was in a bad shape. However, he could repair it during the night and thereby make it available before the function so that people could carry the seedlings in this cradle in a procession without changing any programme. Everybody liked the idea and accordingly an excellent function was held and tree seedlings were planted. Such a fusion sometimes takes place serendipitously. But can it also be planned?

Sometimes grafting of tradition and modern cultural and institutional values can be planned. In Gujarat, a very large scale movement of water recharge has been triggered by Swadhyaya Movement, building upon people’s cultural and religious values without any injection of external resources. In many traditional situations, the place of origin of a natural spring or a stream in mountain areas is considered a sacred site and sometimes would have temple to signify it. There was an interesting case in Bhutan which went to the court on the ground of violation of sacred space. A farmer had cut a tree from a sacred space from the upper reaches of a stream. When people protested, he did not confess his fault or do anything to atone for the mistake. Eventually, the case went to the higher court where the judge held the offender guilty and asked him to plant trees as a part of the punishment in the sacred space and take care of them regularly till the trees were established. Incorporating respect for such institutions in modern jurisprudence may help in recognising that sustainability without involvement of the spirit was not possible in the long term. The functional attributes of a technology was not sufficient to generate the kind of respect that is called for in an inter-generational time frame.

Institutions seldom evolve in isolation. Link across resource and property regimes evolve to generate cross sectoral incentives for sustainability of institutions. During our recent visit to Himalayas, we came across an excellent institution in Belehra, a remote village in the Kangra district of Himachal Pradesh. Way back in 1954, the then Punjab government offered the villagers usufruct rights of grass on a 80 acre degraded forest land in order to provide them with regular supplies of grass for their livestock. However, the government insisted that the farmers would have to generate the necessary funds to regenerate the degraded land and also maintain it. The farmers agreed and on the advice of the government, they pooled one tenth of their individual land holdings and formed a joint farming society. They decided that the land pooled would be cultivated collectively and the revenues thus generated will be used to regenerate the degraded forest land as well as manage it. The forest land was thus regenerated and the fodder from the forest distributed among the farmers. The surplus funds are deposited in the name of the joint farming society and are spent on common facilities such as school, a dam on a nearby stream, guest house etc. Unless a farmer participates in the joint farming of the land, he is not allowed to claim a share in the grass from the forest land. Grass is an important resource for the livestock during dry seasons and a farmer cannot afford to lose his share. The institution is particularly interesting because of the inter-locking arrangement between two resource management systems actually contributing to its sustainability. Thus fusion between two or more institutions can generate generalized reciprocities (Gupta, 1995) among the communities- a step considered necessary for generating cooperation among heterogeneous communities.

The institution building process also involves recognising the boundaries of the common properties and the relationship between common, public and private properties within and outside the watershed areas. During 1988, first author was invited by the state planning board to look at the dry land development programmes of the state. During the visit to Mittmerri watershed in a dry-land region, it was noted that several farmers had experienced increase in the water table in their private wells in the downstream of a water storage structure. This was to be expected. The project design and management structure, however, did not discuss how would the gains from the rise in water table to private individuals be shared with the community. The gains were obviously not a consequence of the contribution by well owners alone. Large number of non-well owning dry farmers and land less pastoralists had also contributed to the conservation of the catchment area by not grazing their animals. The benefits were restricted to only a few. In the same watershed several second generation problems of maintenance of water ways, weirs and spill ways had arisen. The common fund that did exist did not require contribution from such individual well owning beneficiaries and therefore was limited in its scope.

Let us extend the same example to look at how resource utilization is affected by the technology used vis-a-vis the change in property right regimes. In one of the watershed projects in Andhra Pradesh, an open tank was converted into a percolation tank in order to increase water table level. But the result in the next few years was exactly contrary to the expectations, a drastic fall in the overall ground water table level was experienced. The reason being that, once the level in the private wells began raising due to the recharging of the ground water, farmers started over-extracting water from the wells. In other words, once the regime under which the control of access to water shifted from a common property in a tank to a private property in a private well, the sustainability of the resource itself was at stake.

There are many cases where we have looked at the issues in management of common property right regimes with the framework of commons ignoring the interface of such regimes with private and public resources (Gupta, 1985, 1990).