Smallholder
Adoption of Soil Conservation Technologies:
Evidence From Upland Projects in The Philippines
Rob Cramb
University of Queensland
Introduction
Soil erosion due to smallholder agriculture in upland areas of the Philippines is widely regarded as the country's most serious environmental problem. There has been a plethora of upland development projects implemented by government and non-government agencies, many with international funding, and all with a significant component for the promotion of soil conservation and agroforestry measures among upland farmers (notably, contour hedgerow intercropping, in a package known as Sloping Agricultural Land Technology or SALT). Yet it is widely felt that adoption of such practices has been minimal and that the problem of soil erosion and land degradation is not abating.
The University of Queensland and the Southeast Asian Regional Centre for Agriculture (SEARCA) were commissioned by the Australian Centre for International Agricultural Research (ACIAR) to investigate the technical and socio-economic factors limiting adoption of recommended soil conservation technologies by upland farmers. The project, which ran from 1993 to 1998, involved case studies of seven locations where conservation farming had been promoted and (at least for a time) adopted. These studies were supplemented with bio-economic modelling to assess the on-farm costs and returns of alternative conservation measures.
This paper summarises the project's findings regarding the farm-level factors associated with the adoption of recommended soil conservation technologies within the case-study sites. Its focus is on the attributes of the farm-household influencing the adoption-decision process, and the consequences of adoption at the level of the farm-household system. It is acknowledged that the concept of "adopter" (and its converse, "non-adopter") can be problematic, not least because of the "pro-innovation bias" it tends to introduce into the analysis. However, the use of these categories involves no presumption that those identified as non-adopters could or should adopt the technology in question. The starting point is merely the observation that some farmers in the case-study sites implemented the soil conservation technologies which were being promoted, and others did not. How did they differ? What was the impact? These questions, though limited in scope, are clearly relevant to understanding the problem of soil erosion in the uplands.
Methods
The research methods involved a combination of reconnaissance or "rapid rural appraisal" methods followed by a questionnaire survey of a sample of farmers from within the project area. The reconnaissance methods included: a review of existing municipal and village documents; direct observation; semi-structured interviews with focus groups and key informants; resource mapping; time lines; seasonal diagrams; and community histories. The formal survey was administered to a random sample of 70-120 farm-households divided into three categories: "adopters" of the recommended soil conservation measures; "non-adopters" within the same setting, i.e., other farmers in the project village(s) who were thus exposed to the conservation project but chose not to implement the technologies; and (where feasible) a "control group", i.e., farmers in a similar setting but from outside the project area, most of whom were not aware of the conservation technologies being promoted hence had not yet commenced the adoption-decision process.
In the analysis which is summarised below, the attributes of the adopters (or factors affecting adoption) were primarily examined in relation to the attributes of non-adopters within the project site, while the consequences of adoption were assessed where possible by reference to a control group. Individual and group perceptions of changes over time were used to supplement these cross-sectional comparisons. The meaning of "adopters" obviously varies somewhat from site to site, but in general the term refers to farmers who at some point implemented contour hedgerows and/or bench terraces (sometimes with rock walls) on all or a substantial part of their farms. It is likely that the adoption process was largely complete in most sites, the studies being conducted 10-14 years after the commencement of the conservation project; this is important if cross-sectional comparisons are to be valid.
The Study Sites
The "adoption process" in each study site began with the commencement of a project which sought, among other things, to promote conservation technologies. Though many farmers had already invented or adopted various conservation practices (including such measures as ploughing across the slope and piling rocks or crop debris in lines across the slope), none had adopted contour hedgerows or dryland bench terraces (with or without rock walls), the two principal technologies promoted. Hence the role of the project in initiating the adoption process was crucial and needs to be taken into account in analysing farmers' adoption-decisions. In some cases the prior decision was to participate in the project in order to receive whatever benefits were on offer (e.g., livestock dispersal, stewardship contracts, farm tools and inputs, subsidies) and a concomitant of that decision was the implementation of recommended conservation technologies. In general, however, the decision to adopt contour hedgerows or bench terraces, while brought to a head and facilitated by the presence of a project, and often involving group activity, was what the adoption literature terms an "individual-optional" decision, as opposed to a "collective" or "authoritarian" decision. The following paragraphs summarise the available data on the extent and rate of adoption in each case study site.
At Tabayag, a World Neighbours project (which eventually gave rise to a farmer organisation, the Mag-uugmad Foundation Inc. [MFI]) began in 1981 with the formation of a labour-exchange group (alayon) of five members (although two pulled out after two weeks, objecting to the high labour requirement for rock wall construction). By 1993, around 50% of the 159 households in Tabayag had adopted rock-wall terraces and/or hedgerows on their maize farms and the diffusion process within the village was largely complete (though follow-up visits indicated there was still some adoption occurring). There was little evidence of diffusion beyond the project village.
At Guba, another World Neighbours-MFI site, extension began in 1981 with the recruitment of a progressive farmer (who had already received a "best farmer" award) and the formation of a working group of five farmers (his siblings) to implement conservation measures - primarily contour bunds, canals and hedgerows. The main crops grown in the alleys were maize (for subsistence) and, increasingly, vegetables and flowers for the Cebu market. By 1982, 23 farms had been developed and in the following two years there was rapid uptake, facilitated by the employment of part-time farmer-trainers. By the mid-1990s adoption of the recommended technologies was reported for over 1,000 farm-households in 10 villages spread over 78 sq.km. This represented perhaps 30% of the population of potential adopters in those villages. However, a reconnaissance survey in 1996 found that adoption had reached a ceiling and, in many cases, hedgerows were not being maintained or re-established.
At Pananag, the project was more drawn out. Initial contact with Mindanao Baptist Rural Life Centre (MBRLC) - the developers of Sloping Agricultural Land technology (SALT) - occurred in 1980, extension efforts began in 1984, and an intensive extension effort was undertaken between 1989 and 1992. By 1991 around 50% of farm-households had adopted contour hedgerows (not the full SALT package) on at least part of their maize farms, and by 1994 this figure was around 70%. There had been some diffusion to relatives in a neighbouring village.
At Managok, an Integrated Social Forestry (ISF) Project of the Department of Environment and Natural Resources (DENR) began in 1983, and the MUSUAN project of Central Mindanao University (CMU) operated from 1988 to 1992, promoting contour hedgerows. The project reported 60 adopters, but in 1994 there were only 47, the decline being due to death, outmigration, discontinuance, or double-counting. This represented 20-40% of upland households within the project's target area. Adopters generally established hedgerows throughout their maize farms. There was no evidence of wider diffusion.
At Salogon, the government's Upland Stabilisation Project began in 1982 and wound up in 1990. The major conservation measures promoted were contour hedgerows and bench terraces for both upland rice (the staple crop) and maize (the main cash crop). Most adoption occurred in 1985 and by 1990 the adoption curve had levelled out. In the 1995 survey, 54% of the sample were classified as adopters on the basis that they had implemented contour hedgerows or bench terraces on part (mostly 10-50%) of their farms. In many cases, however, the technologies were not being maintained, confirming the view of a key informant that only 5-10% of the land development existing at the end of the project was still in evidence.
At Magdungao, an ISF Project began in 1979 and the government's Magdungao Agroforestry Project was implemented from 1982 to 1991, by which time 87 households had participated in the project, involving among other things the adoption of contour hedgerows and/or bunds on their maize and vegetable farms. This represented 80% of the potential adopters within the village. By the time of the survey in 1995 the level of adoption was still around 80%, though many adopters had not maintained their contour bunds. A significant proportion (28%) of the adopters surveyed had not participated in the project but had adopted of their own accord or had inherited farms from adopters. There was evidence that diffusion to farmers in neighbouring villages was minimal, depending mainly on contact between close relatives.
At Domang, an ISF Project operated from 1986 to 1993, mainly promoting contour hedgerows. From 1989 to 1991 farmers were paid P6 (US$0.25) per linear metre of hedgerow established (e.g., one community leader received a total of P16,000). The site was upgraded to a Model Site in 1990 and received more intensive extension. By 1991 the majority of residents were reported to have adopted hedgerows. In 1993 the site was "devolved" to the municipal government, after which extension activity became practically non-existent. However, at the time of the survey in 1996 there were 78 adopter-households or 90% of the Domang population. Hedgerows were being maintained but there was no expansion onto additional land. The alleys were being used for maize, upland rice, and a range of commercial vegetable and field crops. Diffusion beyond the village was almost non-existent and where adoption did occur it was not well implemented due to poor understanding of the principles and techniques involved. It should be noted that bunded irrigated rice terraces (an indigenous technology for the Ifugao members of the village population) were being constructed before the project began and continued to be developed at the time of the survey.
Summary Of Findings
The process of adoption of conservation farming technologies in the case-study sites was complex and highly variable and could not be separated from the process of intervention in the villages by government and non-government agencies implementing upland development projects. The case-study sites represented a sample of the better-resourced and more successful projects, yet in general adoption rates were low and diffusion beyond the project site almost non-existent. Even at Guba, the most successful example of long-term, widespread adoption, there was clear evidence of discontinuance. The implications for the vast areas of the uplands not in proximity to the "nodes of diffusion" created by well-resourced upland projects are not encouraging.
Factors Affecting Adoption
Three broad sets of farmer attributes were examined in relation to adoption: personal attributes, perceptions, and farm attributes. In general, readily measurable personal attributes of farmers (age, education, etc) were not important in explaining adoption. However, age and gender were factors where strenuous work was required to implement the technology (i.e., terracing and rock walls), older farmers and female farmers being less inclined to adopt such measures. Less easily measured traits related to "innovativeness" and "managerial ability" were clearly important, particularly among early adopters. Such traits are usually readily identified by community members and experienced extension workers, and were probably correlated with initial levels of farm diversification and cash income.
Awareness of soil erosion was relatively high, but farmer perceptions of soil erosion as a problem varied, partly due to differences in farmers' knowledge but mainly because of objective differences in soil and farming conditions (e.g., fallow systems versus continuous cropping). Perceptions of erosion clearly had an effect on adoption behaviour. Perceptions of and attitudes to the recommended technologies appeared to be well-informed, based on the direct or indirect acquisition of site-specific knowledge over several years. Adopters and non-adopters shared perceptions regarding the labour requirements for establishing and maintaining the technologies, the loss of plantable area, and the delay in obtaining benefits, as well as the undesirable side-effects of some forms of hedgerow on weeds, pests and diseases. Nevertheless, contour hedgerow technology in particular was widely seen (at least within project villages) as useful and necessary, easy to learn, and easy to acquire (though acquisition of planting materials was clearly a problem in some cases). Beyond project villages, however, there was very little awareness or knowledge of the recommended technologies (particularly regarding methods of implementation), indicating that diffusion does not occur without facilitation.
There were many farm-specific factors which influenced adoption. Adoption was more likely on larger farms. The physical features of individual fields were also important: adoption of hedgerows in particular was more likely on fields which were larger, steeper, had more erodible soils, were located close to the homestead, had relatively uniform terrain, and were oriented down rather than across the slope. In the case of rock walls, obviously the rockiness of the field was a predisposing factor, whereas the prior construction of rock lines discouraged further development. Ownership of lowland fields discouraged investment in upland fields in some contexts.
Land tenure was a major factor but its influence was highly conjunctural. The classification of the land as public or private was only important where farmers occupying public land felt their tenure was insecure vis-a-vis the state, in which case stewardship contracts with the government (and the conditions that went with them, including conservation measures) appeared attractive. Otherwise tenancy was the major issue. In general, tenancy discouraged adoption or led farmers to adopt less costly measures such as grass hedgerows, but the crucial factors were the awareness and attitude of the landowner, which varied enormously.
The labour requirements of the technologies were an important consideration but the farm-household's labour supply was not a major factor in itself. Rather, it was related to cash flow concerns and the need to use spare labour off-farm in preference to implementing conservation measures. Relatedly, cash income was an important factor promoting adoption, particularly for a sub-group of adopters with high cash income. While all farmers experienced a credit constraint, farmers with higher cash income were less bound by it and therefore able to invest more labour and working capital in their own farms.
In short, differences in personal attributes and perceptions as such were not a major factor explaining differences in adoption within project villages; rather it was the appropriateness or relevance of the technologies to the farmer's specific circumstances (the farm factors) which was the key, emphasising the need for technologies to be adapted to different sets of farmer circumstances. Analysis of these farm-specific constraints to adoption can assist in the development and promotion of a wider range of technology options (e.g., low-cost, quick-return options for resource-poor, tenant farmers).
Consequences of Adoption
The consequences of adoption for the farm-household were difficult to disentangle; in general they were positive, though not substantial. The impact on the farming system was mainly in terms of labour use. There was an unavoidable requirement for a high initial investment of labour which, particularly in the case of hedgerows, created an early-season labour peak. There was also a redistribution (from land preparation to hedgerow maintenance) and a net increase in on-going labour requirements, though this varied from site to site depending on the specific form of hedgerow technology used. There was some evidence that the effectiveness of fertiliser use on conservation plots was increased, encouraging some farmers to apply more and some to apply less; in general, the nutrient cycling aspects of hedgerow technology were overshadowed by use of purchased fertiliser. The balance of farming activities was changing, but not as a direct consequence of adoption. Some farmers (e.g., in Guba and Tabayag) were expanding commercial vegetable production on conservation plots and in some sites hedgerow technology was tied to intensive goat-rearing.
The consequences of adoption for food production were largely indirect. Where adopters obtained increased yields and output this was largely because they had stabilised their sloping land and so were willing to invest more in seed-fertiliser technology. Sometimes the latter phenomenon was a general effect of the project, so that farmers in the project village who were non-adopters of the recommended conservation technology nevertheless produced more food than farmers in neighbouring villages, due to increased use of improved production technology. To assess the longer-term consequences of conservation technology for food crop yields and production requires a modelling approach, such as reported by Nelson and Cramb (1998).
The consequences for farm cash income were also indirect and not very important, except in specific cases where hedgerows were productive in their own right (e.g., providing fodder for goats or seed for sale), but these gains could perhaps have been achieved more efficiently in other ways. Most differences in income between adopters and non-adopters were not related to the use of conservation measures but to differences in farm resources and management ability. The main direction of causation was thus in reverse: better-off, more commercial farmers were more likely to adopt conservation measures.
Conclusion
In conclusion, conservation farming technologies, particularly hedgerows, are widely seen by farmers who are aware of them as useful and even necessary, but it has required resource-intensive project intervention to get the adoption process going, and adoption is often constrained by farmers' specific circumstances (rather than their personal attributes and perceptions). A wider range of more profitable and less demanding conservation technologies is needed, promoted more flexibly and with greater on-going support for farmers in their efforts to experiment with improved farming systems. This implies a need to pursue a major program of adaptive research and extension in the uplands which better accommodates the range of farmers' goals and circumstances. Guidelines are available for the kind of program required, incorporating soil conservation and land husbandry into a process of farming systems development. However, the higher-order, politico-economic constraints to implementing such a program in the Philippines are formidable.
This summary is an overview of a larger set of studies. For details of other publications, contact Rob Cramb at R.Cramb@mailbox.uq.oz.au
The SEA News index is at http://welcome.to/seanews