level of fluidity in land-use change. Second, local perception could be an important factor in adoption of sustainable land uses, and there was no perceived damage caused by shrimp culture and no perception of land dereliction being a problem. Third, there were clear preferences for the types of employment across all sites.
The design of this study was based on the concept that a farmer would convert his or her land from another agricultural pursuit to intensive shrimp culture, and then convert to another use when shrimp production failed. The process we found was not linear, but rather cyclic (Figure 1). The number of farmers that had failed at shrimp farming and had returned, either by relocating or by waiting for a period of time, indicated that land use decisions were not permanent, and failure at one time did not diminish the allure shrimp farming generated for future economic return.
The cyclical pattern of abandonment complicated quantification of abandoned, converted, or underused land as the extent of dereliction varied regionally and temporally. At times and in certain locations, shrimp culture may appear to be completely successful and sustainable. This assessment could lead to the erroneous conclusion that there was no problem with land dereliction in that area. Interviews and field observations indicated there were large areas of shrimp ponds not in use, and at any point, they could be brought back into use and just as quickly become idle again. Although land dereliction was difficult to precisely quantify, there were large areas of former shrimp ponds with little or no production of any sort. Given the importance of the agricultural sector in Thailand and its limitation by cultivable area, land dereliction on any scale presents a barrier to development.
Solving the land dereliction problem requires that the conversion cycle be broken. This study identified two factors that influenced the type of interventions likely to arrest rapid and unsustainable land-use change. First, issues that motivate a community to participate in an intervention, adopt technology, or make lasting changes must be considered. Secondly, the intervention should have its foundation in desires of the community and build on activities that are acceptable locally. Success of development activities has been linked to inclusion of local opinions (Blanchet, 2001; GESAMP, 2001; Barg, 1992).
Two important results related to how communities mobilize to address locally defined problems. The Òperceptions of community developmentÓ theme was developed to offer insight into social impacts of shrimp culture. We assumed that perception of social degradation would provide impetus for a community to stop unsustainable shrimp culture and create pressure to move to other forms of income generation
that had less severe effects. There must be local perception of a problem in order to facilitate local action. Therefore, the problem perception theme for questions inquired about local views of shrimp culture, pond conversion and land dereliction. Strong local perception of a problem would be another motivator to break the conversion cycle.
Many references on shrimp aquaculture predict that shrimp farming communities will experience dramatically negative social impacts from shrimp farming (Scott, 1998; Gujja and Finger-Stich, 1996; Quatro et al., 1996). Though this may occur in some locations, these impacts were not detected by this study. Respondents felt their communities were better off in a number of ways than they had been in the past. This study did not establish a causal link between shrimp culture and positive community development, but from the perspective of community members, it demonstrated improvement in some indicators of quality of life during the shrimp boom. However, it was possible that villagers did not recognize, or were unwilling to discuss, negative social impacts.
Data from all three study sites supported a trend toward further marginalization of the poorest shrimp farmers. The debt cycle created through shrimp culture caused farmers who failed to convert or sell land. In many cases, conversion required that farmers accept lower incomes (e.g., culture of fish or other aquatic species), or make additional capital investment (e.g., rice culture, housing, or industry). Poorest farmers are often unable to accept these conditions, and land sale is the most economically sensible option for them. After repaying debt, these farmers remain impoverished, landless laborers (Flaherty and Karnjanakesorn, 1995) left to seek employment outside the village where they farmed.
The perception that the community was better off than in the past may simply be an artifact of further marginalization of the poorest people. Conditions for these people may not have improved, but since they could have been forced out of the community by debt and need for employment, they would not be included in interviews.
There was no perception that communities were harmed by displacement of people, nor that the social legacy of shrimp culture was fundamentally negative. Therefore, social concerns were not sufficient by themselves to cause conversion of abandoned or poorly functioning farms.
There was no perception that land dereliction was a problem in any study site. The definition of "abandonment" played a role in how land dereliction was perceived. Land for which there were future intended uses was not perceived as derelict, regardless of time of fallow or uncertain