Optimal resource management with the multiple goals of citrus-based farming systems at farm and watershed levels, Chinag Mai Province

Abstract

The objectives of this study were to find the optimal resource management with the multiple goals of citrus-based farming systems in the Fang watershed at the farm and the watershed levels and to find the relationships and trade-offs between the natural resource management at the farm and the watershed levels. The citrus farms in this study were divided into four types: small farms using chemicals (SFC), small farms using chemicals and bioextract (SFCB), large farms using chemicals (LFC) and large farms using chemicals and bioextract (LFCB). A set of questionnaires was designed for data collection covering 149 households in Chai Prakan, Fang and Mae Ai districts, Chiang Mai province. The study used a multiple goal linear programming model at the farm level which contained 80 decision variables and 64 constraints. The watershed level model contained 84 decision variables and 67 constraints. The models have 3 sets of goals: economic, social and environmental. These were further broken down into 7 goals at the farm level, and 8 goals at the watershed level. An analysis using multiple-goal programming was conducted by weighting the goals using analytic hierarchy process (AHP) method. The study used the IMGP-LPlan software program.

In the case of small farms using chemicals, small farms using chemicals and bioextract and large farms using chemical and bioextract, the results showed that the equal weight method gave a more environmentally-friendly patterns of land use and had high goal achievement rates. The results suggested that fruit tree cultivation should contain mixed cropping of coffee, lychee and citrus. The AHP method produced the weight values derived from discussion sessions among the stakeholders who gave their relative importance of the various goals. The multiple goal linear programming model results using such AHP method recommended that the only fruit trees grown should be citrus. They also recommended increasing use of household labor and decreasing use of hired labor. For all farm types, their net annual revenue from crop production increased compared with the results from the equal weight method. For the large farms using chemical and bioextract, both methods gave similar results and goal achievement rates.

For the sensitivity analyses, the study examined 4 scenarios using equal weight method: 1) an increase in the weight value of the annual equivalent value goal by 20 percent 2) a decrease in the goal level of off-farm work by 30 per cent 3) an increase in the wage rate from 200 to 300 baht per day and 4) an increase in the hired labor availabililty by 30 per cent. All the sensitivity analyses results found that the goal achievement rates varied slightly from the base results.

The trade-off analyses between goals at the farm level are found the tradeoffs to be rather low, e.g. if there were a decrease in the dependence on external inputs or loan investment by 60 percent, or revenue variance by 80 percent, these would only reduce the annual equivalent value by only 16 percent for both the small farms using chemicals and the small farms using chemicals and bioextract. For both the large farms using chemicals and the large farms using chemicals and bioextract, if the same trade-off were applied it would have a greater negative impact on the annual equivalent value. However, for all farm types, there were high trade-offs between hired labor and the annual equivalent value e.g.if hired labor was reduced by 10 percent, it would have a serious negative impact on the annual equivalent value.

An analysis using multiple goal linear programming at the watershed level showed that the results of the AHP method gave a more environmentally-friendly pattern of land use and higher goal achievement rates. They recommended a mixed fruit trees at 68 percent of total area and an increase of the forest area to 30 percent. The recommended fruit trees were coffee, tea, longan, lychee and citrus. The equal weight method results recommended fruit trees at 87 percent of total area being citrus and coffee in equal proportion, however, it did not recommend any forest area. The AHP method recommended doubling of off-farm work, increasing employment by 40 percent, reducing loans from the BAAC by 16 percent while attaining higher net annual revenue from crop production compared to the results obtained from the equal weight method.

At the watershed level, this study took into account 4 sensitivity analyses from the equal weight method: 1) a decrease of 30 percent in the dependence on external inputs 2) an increase of 30 percent in the revenue from non-timber forest products 3) an increase in the wage rate and 4) an increase in the hired labour availability. All the sensitivity analyses gave similar patterns of land use.

The trade-off analyses conducted between the economic and the environmental goals, and between the social and the environmental goals showed rather low trade-offs between such goals. If there was a reduction in the environmental goal through reduced expenditure on pesticides, nitrogen use or soil erosion by 30 percent, it would have a small negative impact on the annual equivalent value (the economic goal) or employment (the social goal) by only 10 percent. The trade-off between the economic and the social goals was however rather high. If hired labor decreased by 20 percent, it would have a serious negative impact on the annual equivalent value. This result is similar to the results found with large farms.

Land use management using multiple goal linear programming at the watershed level gave a more environmentally-friendly pattern of land use compared with that recommended by the extrapolated results from the farm-level. It recommended a mix of fruit trees and forest area for 98 percent of land use. The extrapolated results to the watershed level showed a contrast which recommended more cash crops than fruit trees on 40 percent of the land but did not recommended any forest area.

The results of this study could contribute to a drive to secure sustainable livelihoods in the Fang watershed. Local groups and networks should be established, working together with the local authorities to establish strong networks of participatory land use planning. There should be land use zoning as a result of participation and acceptance of stakeholders at all levels. Local monitoring committees and networks should be established. Furthermore, the government should promote usage of bioextract and organic matter to in place of chemical use and promote an increase in research and development budget for watershed development to relevant partners.