Chapter 2
Some key agricultural resource issues

The starting point for SARM, particularly with regard to poor and marginal farmers, must be to look at the resources available to individual rural households that they can use for agricultural purposes. Two broad categories of agricultural resource circumstances can be recognised: those that are internal and external to the farm household. The internal resources refer to those on-farm resources to which the household members have direct access and by and large control (eg. their own land, labour, planting material, livestock and knowledge). The external ones are those that are obtained from off-farm sources and may require the assistance of people and organisations external to the household (eg. advice from government extension services, chemical fertiliser and pesticides from private traders). Tables 1 and 2 show the key agricultural resources available to farm households. The poorer and more marginal the farm household, the greater the reliance on their own internal resources.

Goals for SARM

During the past 50 years, agricultural development policies promoted by governments and donor agencies within the Asia Pacific region have emphasised the use of external resources as the means to increase agricultural production. The end result of such policy initiatives has been a remarkable growth in global consumption of pesticides, inorganic fertiliser, manufactured animal feedstuffs and tractors and other machinery (Pretty 1995).

These external inputs have however substituted for natural control processes and resources, rendering them more vulnerable. Pesticides have replaced biological, cultural and mechanical methods to control pests, weeds and diseases; inorganic fertilisers have substituted for livestock manure, compost and nitrogen-fixing crops; information for management decisions comes from input suppliers, researchers and extensionists rather than from local sources; and fossil fuels have substituted for locally generated energy sources (Ibid). The specialization of agricultural production and associated decline of traditional mixed farming have also contributed to this situation. What were once valuable internal resources have now become waste products which can be cause for environmental concern. In Fiji, manure from many of the commercial piggeries and dairy farms is dumped adjacent to and in local rivers with adverse effects on downstream water quality (Jai Kumar personal communication).

The basic challenge for SARM is how to make better use of these internal resources. The need is to help farm households review their internal resources and find ways to use them more effectively while minimizing the use of external inputs. This does not necessarily mean doing away with external inputs so much as using them when and as necessary to enhance the effectiveness of internal resources. SARM should therefore seek to pursue the following goals (Pretty 1995):

• More thorough incorporation of natural processes such as nutrient cycling, nitrogen fixation and pest predator relationships into agricultural production processes.
• Reduction in the use of those off-farm, external and non-renewable inputs with the greatest potential to damage the environment or harm the health of farmers and consumers, and a more targeted use of the remaining inputs used with a view to minimizing variable costs.
• More equitable access to productive resources and opportunities, and progress towards more socially-just forms of agriculture.
• Greater productive use of the biological and genetic potential of plant and animal species.
• Greater productive use of local knowledge and practices, including innovative approaches not yet fully understood by scientists or widely adopted by farmers.
• Increase in self-reliance among farmers and rural people.
• Improvement in the match between cropping patterns and the productive potential and environmental constraints of climate and landscape to ensure long term sustainability of current production levels; and
• Profitable and efficient production with an emphasis on integrated farm management, and the conservation of soil, water, energy and biological resources.

Table 1. Some key agricultural resources

Table 2. Source and type of organic manure used on-farm in the Asia Pacific Region

Geographic units for agricultural resource management

It is widely accepted that blanket agrotechnology recommendations disseminated at a national or regional level may not be appropriate at the local level. Thus extensionists, researchers and planners have sought to divide the broader landscape into discrete geographic units for agricultural development purposes. This has often taken the form of delineating different agro-ecological zones (AEZ). This is based on the assumption that technologies used successfully in one AEZ could be adopted by farmers in similar AEZs and that policies and plans successfully implemented in one country, region or district can be readily transferred between or within countries where the range of AEZs is similar. However AEZs are largely based on biophysical parameters such as soils, climate and topography and have tended to ignore the human element of the development of agricultural resources.

For farming systems development (FSD) purposes attempts have been made to define discrete geographic areas known as farming systems zones (FSZ). The aim behind the zoning is to identify workable units which are homogeneous entities with strong similarities in terms of the farming systems within zones, and differences between and among zones. The concept of FSZs is based on the principle that farmers with similar problems and development potentials have similar objectives, resource availability and utilisation, strategies and practices (FAO 1990). It also assumes a high degree of correlation between agro-ecological and farming systems variables, such as cropping pattern, planting dates, farmers' problems and possible solutions. At a very broad level this assumption is valid as all farm households within the same agro-ecological zone face the same biophysical constraints and potentials and therefore have their choice of crops limited by the same natural environmental circumstances.

The criteria for determining AEZs and FSZs are highly interrelated but are clearly not the same. The former is based on the physical and biological characteristics of an area, the latter on the people farming within an area, their circumstances, practices, problems and solutions. In practice, FSZs may well be synonymous with agro-ecological zones or subdivisions of them. However, again at the broad level, the choice of crops, livestock and land management practices may be determined more by the socio-cultural environment in which the farm households operate. For instance two different tribal/ethnic groups occupying adjacent zones may have different farming systems for historical, cultural and religious reasons, e.g. one group may have been traditionally pastoralists whereas the other have always been primarily farmers.

More recently those involved in researching natural resource management have promoted the concept of resource management domains (RMD). The proposed working definition is that (Dumanski & Craswell 1996):

The aim is to define discrete RMDs whose bio-physical and socio-economic characteristics can be stored on computer and used for research, extension and planning purposes with the aid of geographic information systems and decision-support systems.

In an ideal world all farm households within an ecologically homogeneous area and the same socio-cultural environment would form one homogeneous group. In practice, the internal household socio-economic circumstances may vary even when the biophysical and external socio-economic circumstances do not. Whereas it is possible to subdivide an area into discrete land units according to differences in bio-physical conditions (e.g. AEZs), and it may be possible to differentiate areas on an ethnic/tribal community basis, it is rarely possible at the local area level to separate farm households into homogeneous groups occupying mutually exclusive geographic areas (e.g. FSZs).

It is not uncommon to find that neighbouring households within a village will pursue different land use enterprises and have markedly different socio-economic circumstances hence have different objectives and resources, and different problems and development potentials. This means that they may not respond in the same way to the same recommendation and will need to be considered, for development purposes, as members of different homogeneous groups. The concepts of AEZs, FSZs, and RMDs may be useful for recognising broad scale differences, but SARM at the field level calls for the grouping of farm households at the local level on the basis of similarities in their socio-economic circumstances rather than on a geographic area basis. Hence it is important to recognise that there may be different broadly homogeneous farm household groups within the same geographic area, and that each group will require separate consideration.

Farm household level resource management domains

It is important for SARM to recognise that no two farm households will have exactly identical circumstances or goals. Hence each household will be faced with a unique set of production constraints and potentials as well as sustainability problems. As a result different households will be interested in different ways of using the land for crop, livestock and/or tree production. Even if using the land for the same purpose, differences between households, with regard to their agricultural resources and socio-economic circumstances, may mean that they are unable to adopt the same recommended land management practices.

Within a farming community it is normal to find considerable variation between farm households with regard to goals, needs, circumstances and resources available. As a result neighbouring households within the same village area may have different development interests and opportunities. For instance livestock owners will have different interests from those farmers without livestock. It is therefore highly unlikely that one standard set of improved farm management recommendations could address the production and sustainability problems and constraints faced by all the households farming within a specific area.

It is impractical for research and extension workers to develop and disseminate improved agricultural resource management recommendations for each farm household. Hence the pragmatic value in grouping individual farm households into common resource management domains (CRMD)¹¹ for whom a particular set of recommendations would be appropriate. A CRMD can be defined as one in which the constituent farm households:

• share broadly similar bio-physical and socio-economic circumstances, i.e. have a similar resource base (eg. land, labour, equipment, livestock, skills, cash etc) from which to undertake their farm operations;
• follow similar farming systems (i.e. engage in the same range of farm enterprises);
• face the same production constraints and sustainability problems.

An assessment of the circumstances of the households within a CRMD would reveal a range of common problems, constraints and development potentials. Different CRMDs can be distinguished on the basis of obvious differences in the circumstances and farming systems of the constituent households. The underlying assumption is that the specific set of common problems, constraints and development potentials shared by the households of each identified CRMD would be significantly different. In the context of SARM, assessing and defining the characteristics of different farm household CRMDs provide the basis for the development of locally specific SARM practices suitable for adoption by all the constituent households of an individual CRMD.

Unit of intervention

The debate over the appropriate unit of intervention for SARM focuses on two issues. The first is whether the geographic boundaries of the area in which to promote improved SARM activities should be delineated using topographic features (eg a watershed) or correspond to administrative units. The second issue is whether the target unit of action for field level technical interventions should be individual farm households or community groups.

Limitations of a watershed planning focus

Much of the planning for soil and water conservation purposes in upland areas has relied on the adoption of a watershed management approach in which plans are formulated and implemented within the catchment area of a natural watershed.¹² The reasoning is that since each watershed is an independent unit with respect to its surface hydrology, selecting the topographic watershed as the unit of work enables the movement of surface water to be controlled and managed effectively (Shaxson 1992a). The aim is to ensure that the way the land is used upstream will not have adverse consequences downstream.

Because a watershed is a recognisable natural hydrological unit it has been assumed that its geographic boundaries can serve as a socio-economic or socio-political unit for planning and implementing resource management activities (Easter et al 1991). From a purely technical perspective (e.g. where the concern is to prevent siltation in a downstream irrigation scheme by reducing soil erosion in the upper watershed) the use of watershed boundaries makes sense as an appropriate unit for defining the location and extent of a planning area. However, except where high mountain ridges make it physically difficult to cross the watershed boundary, the cultural, administrative and political boundaries in which farming and forest dwelling communities operate are almost always different (see box 3).

Because the watershed is essentially an `artificial' unit that is surveyed, mapped and planned by outsiders, it leads inevitably to the enforcement of a top-down physical planning approach. This bypasses local people's priorities and skills and leaves them as onlookers (IFAD 1992). Much of the work undertaken under the label of watershed management has involved detailed and complex land evaluations and resource surveys to indicate what sort of use should be applied where, and then devising means of rearranging current land use until it fits the maps. This may work in sparsely settled areas but will be unacceptable in dense agricultural settlements. The latter is the more typical situation within the developing countries of the Asia Pacific region, in areas where soil degradation is a problem. Pushing people around (especially those with small, scattered plots) to make their land use conform to a rational pattern on a map, is not a politically popular exercise at any level of society, usually generating mistrust, apathy and antagonism (Shaxson 1992a). It also encourages planners to see the people as part of the problem rather than the key to the solution.

It is not unusual for farmers to cultivate more than one plot of land and different plots may be located in different watersheds. If projects are planned rigidly on an individual watershed basis the result may be that farmers can obtain technical advice and inputs for only some of their fields, ie. those located within the project watershed. The control of shifting cultivation in one watershed may lead to excess pressure on neighbouring areas, where households also have traditional cultivation rights. The land use problems associated with shifting cultivation, eg. deforestation and erosion, are merely transferred as farmers transfer their crop production to areas outside the project's control. Likewise total protection of natural woodlands in one watershed (eg. a ban on logging) may lead to increased exploitation of woodland areas in neighbouring watersheds considered to be the common property of the rural community. Similarly closing a watershed to grazing, without reducing stock numbers, merely increases grazing pressure on neighbouring areas as the herds are moved (Douglas 1994).

It is clear that where small-scale farmers are the primary beneficiaries, the boundaries of a SARM project or programme should correspond to the boundaries of the land area in which the members of the target communities have land use rights. That is, the cultural, administrative and political boundaries in which the individual farm families that make up particular communities operate.

Many national and local government services are organised along similar lines, e.g. agricultural extension and social welfare, and this will aid co-ordination with their activities. Such boundaries may also be artificial creations, having been imposed on communities for administrative and political reasons. These are likely to be more closely related to the social and cultural boundaries within which the `natural' communities live and operate than the hydrological watershed.

It was notable when reviewing documentation concerning both FAO and non-FAO watershed management projects that the maps of the project areas clearly demarcate the topographic features of the area, showing the project boundary conforming to the watershed. It was very rare to find any maps showing the boundaries between the individual villages or communities within the watershed. The conclusion is that, despite the new participatory rhetoric, most watershed management projects still consider the social boundaries, of the people they are working with, of less importance than the hydrological boundary.

Focusing on administrative/community boundaries does not mean ignoring "watershed management" principles but requires that watersheds be looked at within the context of the boundaries of the land over which a community has direct land use rights and interests. Thus the outer limits of a project area could be defined with respect to social and administrative boundaries, whereas individual field level activities could still, when appropriate, could be implemented on a local (micro) watershed basis within these limits, e.g. for the improved management of the catchment of a `protected' water source. However the boundaries of individual farmholdings, rather than watersheds, can expect to be more important in determining the location of many other potential SARM activities, e.g. promotion of improved conservation-effective crop husbandry practices, on-farm fodder production and tree planting.

There are reports, particularly from India (e.g. Mishra & Sarin 1988, Aruna 1990, Grewal et al 1995), of successful conservation projects planned on a watershed basis. The "protected" watersheds are invariably small in area and fall within the social and cultural boundaries of one community. Furthermore such projects have shown the participants a clear link between the "upstream" costs and "downstream" benefits. What these "successes" do not report is the impact these projects had on the community land outside the watershed area - were these also protected or were the problems merely transferred?

Problems can arise when a watershed management plan calls for the costs to be borne by communities in the upper watershed areas whereas the benefits (e.g. irrigation water) accrue to communities in the lower watershed areas. Such problems are further compounded when watershed boundaries cut across political and administrative boundaries. There are often practical difficulties in implementing watershed management programmes that depend on inter-departmental, community and political co-operation across them.

Social unit of intervention

SARM projects which have concentrated on individual land users, and their privately managed cropland, have been more successful than those that focus on the catchment or watershed (IFAD 1992, Pretty 1995). For the individual farm household a watershed is a less natural unit of perception and action than the boundaries of it's own holdings. More often than not the factors that determine the location and boundaries of a farm household's fields will be the social and administrative circumstances of the community in which it operates and its internal resources (particularly labour). Rarely will these be determined by the physical characteristics of a watershed (Shaxson 1981).

Individual watersheds may contain many farm households with separate holdings and marked differences in farming skills, education, interests and needs. It is difficult to implement a conservation package that requires them all to work together for the conservation of resources not solely their individual responsibility. It is also rarely successful given the need for the complete package to be operational if it is to work technically (Douglas 1989, 1992b). When the appeal is to the individual and the benefits accrue directly to the land user, the picture changes dramatically (IFAD 1992).

The priority for sustainable agriculture has to be the motivation of individual farm households to adopt good land use and improved management practices within the boundaries of their own plots. The ideal is therefore to be able to work with households on an individual basis. In reality there are many practical difficulties in doing this. Given the low staffing levels and budgetary limitations of most extension services, an individual approach means that only a few farmers within a community can be contacted at any one time. This is why most agricultural extension services have sought alternative ways of disseminating messages to larger numbers of farmers, such as the use of contact farmers, farmers clubs and T&V (teach and visit). Whereas work may be done on individual fields conservation messages will have to be disseminated, and training conducted, on a group or community basis.

If the technical emphasis is on physical earthworks to control and safely dispose of excess runoff, an individual approach will not work with small-scale farmers because the installation of the full catchment plan will require diversion ditches and waterways to run through neighbouring plots. Diverting runoff in one field may cause problems and social conflict if it is discharged into neighbouring untreated fields. Where such an approach is appropriate, conservation activities will require the organisation of groups of farmers, given the need to coordinate activities on a wider farm basis. Likewise other activities may require action at the community level, particularly where the community will receive tangible benefits which can only be achieved through a community effort, e.g. impoundment and control of water for the irrigation of several land holdings.

¹¹ Farming systems research (FSR) workers commonly use the term recommendation domain for this. Experience has shown that inclusion of the word recommendation in the name can be confusing, particularly to extension workers. There is a tendency for those not familiar with FSR work to immediately want to give a recommendation. The priority should be to first assess and understand the local farm household circumstances in order to group farm households into common resource management domains that subsequently become the focus for participatory SARM development planning, extension and research work.

¹² There is a variation in terminology (Hudson 1992). British usage is `catchment' for the area which catches run-off, and `watershed' for the boundary of the catchment, i.e. the divide which sheds water on either side. American usage is `watershed' instead of `catchment' and this is increasingly adopted in the international literature on `watershed management' and has therefore been followed here.