Effects of agroecological optimization and decision-making on threshold behavior

Background/Question/Methods

Modern agroecological systems need to be designed to produce food, fiber, feed and fuel in the most cost efficient and pollution free means possible. Optimizing efficiency over large acreages requires reducing per-unit-area costs and judicious use of petrochemical inputs, however current recommendations ignore the influence of ecological factors that vary on small scales.  In order to enhance the use of ecological management principles by large farms, it is crucial to account for the spatial and temporal variability inherent in farming systems.  Furthermore, these sources of variability must be placed within the context of all the driving factors that influence farmers’ management choices.  This analysis examined the combined ecological, climatic and economic sources of variability, using site-specific yield and protein monitor data combined with spatiotemporal environmental datasets to assess the most influential factors determining farmers’ decisions.  The analysis was performed within a dryland wheat farming system.  To understand the implications of different sources of variability for whole-farm sustainability, a simulation analysis was performed to determine the probability of crossing critical economic and ecological thresholds.

Results/Conclusions

Linear and non-linear hierarchical mixed-effects models were parameterized for three fields over three growing seasons.  The fit of these models suggest only a minimal (0.058 bu. spring wheat gained from each additional lb. of N-fertilizer) response of grain yield and protein to applied fertilizer.  The crop response to weed competition was negligible, however the exact sensitivity was dependent upon the within-year precipitation patterns and historical management regimes.  When these responses were combined within the simulation model, the resilience of on-farm profitability was shown to be most sensitive to changes in baseline wheat crop prices.  A distinct non-linear threshold was found at $9.48/bushel, below which the farmer would optimally apply fertilizer in only a few productive locations, and above which they would apply the maximum of the recommended rate.  This abrupt shift in management strategies suggests that policies targeted to reduce nitrate pollution could easily incentivize low levels of fertilizer use when prices are near the threshold.  Furthermore, the observed weak response to weed pressure suggests that interspecific competition may be less of a limiting factor than expected when environmental heterogeneity is high.