Investigating the Effects of Agent Rationality Assumptions on Models of Agricultural Adaptation

While the deep uncertainties characterizing both future greenhouse gas emissions pathways and the climatic impact of those emissions are widely acknowledged, less attention has been paid to the role of autonomous adaptation in contributing to uncertainty in the social and economic impacts of climate change. This uncertainty is particularly important in agricultural systems both because they are highly exposed to climate change and because farmers can and do use multiple strategies to manage changing weather risks. However, because long-term environmental change at the rate and scale of projected climate change has few historical analogies, the ability of farmers to perceive and adapt to those changes remains unclear. This study uses a highly stylized, agent-based model of agricultural adaptation to investigate how various representations of farmer behavior (optimizing, risk-averse, lagged information, networked information) affect estimates of climate change impacts in agriculture. These models result in very different spatial and temporal patterns of adaptation uptake and therefore variation in the modeled impact of climate change on incomes and food production. Since it is unclear which, if any, of these models best describes farmers' responses to future climate change, these results can be thought of as a partial description of one aspect of the uncertainty surrounding autonomous adaptation in agriculture. A sensitivity analysis is used to identify parameters that significantly affect patterns of adaptation uptake.