Evaluating strip farming systems for arid ecosystems: A stochastic approach.

Abstract

The objectives of this study were: (a) to delineate a methodology for determining the feasibility of a water harvesting system for an arid ecosystem, (b) to develop a simulation model for the water balance of a water harvesting strip farming system, and (c) to perform a sensitivity analysis of the simulation model. The discussion focuses on the feasibility as constrained by risk. A computer program was developed for the computations. The results indicate that light textured soils cannot be used for this type of farming system since their hydrodynamic characteristics do not allow runoff from light rains (i.e. the threshold value is high). Medium textured soils require a minimum of 50% of initial soil water content and a CCAR of 2:1. Heavy soils require a minimum of 20% of initial soil water content and a CCAR greater than 5:1. As a general rule, the system is more efficient, in terms of water use efficiency (eᵤ), as the conditions become dryer. In this way, for CCAR ranging from 0 to 5 eᵤ is around 95% for all three textures. Beyond that point, as the CCAR increases eᵤ decreases being more efficient light soils and less efficient heavy soils. The shapes of the curves of profit, reflect the impact of weather conditions on the response of output to varying CCAR. Lack of rainfall prior to and during the growing season results in poor outputs. The risk (probability of loss) will depend on both the farmers own criteria and the economical situation. Thus, these criteria can be divided into two broad categories: risk-taking (those farmers who pursue a high level of investment relative to the probability of failure), and risk-adverse (those farmers with low level of investment relative to the probability of failure). Curves of minimum risk-maximum net benefit (minimax) were developed as a guide for decision making processes. According to the minimax curves, for medium soils, the minimum risk maximum net benefit is achieved with CCAR 2:1 and initial soil water content of 100%. This implies a risk of 0.2 and net benefit of $510.00. For heavy textured soils, the minimum risk maximum net benefit is achieved with CCAR 10:1 and initial soil water content of 100% with a risk of 0.55 and net benefit of $580.00. The sensitivity analysis indicate that the model is sensitive to: (1) initial soil water content, (2) soil depth and (3) soil texture in terms of reduction on yield (actual/potential yield).