Scaling climate change to human behavior predicting good and bad years for Maya farmers

Objectives Human responses to climate variation have a rich anthropological history. However, much less is known about how people living in small-scale societies perceive climate change, and what climate data are useful in predicting food production at a scale that affects daily lives. Methods We use longitudinal ethnographic interviews and economic data to first ask what aspects of climate variation affect the agricultural cycle and food production for Yucatec Maya farmers. Sixty years of high-resolution meteorological data and harvest assessments are then used to detect the scale at which climate data predict good and bad crop yields, and to analyze long-term changes in climate variables critical to food production. Results We find that (a) only local, daily precipitation closely fits the climate pattern described by farmers. Other temporal (annual and monthly) scales miss key information about what farmers find important to successful harvests; (b) at both community- and municipal-levels, heavy late-season rains associated with tropical storms have the greatest negative impact on crop yields; and (c) in contrast to long-term patterns from regional and state data, local measures show an increase in rainfall during the late growing season, indicating that fine-grained data are needed to make accurate inferences about climate trends. Conclusion Our findings highlight the importance to define climate variables at scales appropriate to human behavior. Course-grained annual, monthly, national, and state-level data tell us little about climate attributes pertinent to farmers and food production. However, high-resolution daily, local precipitation data do capture how climate variation shapes food production.

Automated summary

Local, daily precipitation closely fits the climate pattern described by farmers. Heavy late-season rains associated with tropical storms have the greatest negative impact on crop yields at both community- and municipal-levels. Fine-grained data are needed for accurate inferences about climate trends, as coarse-grained data tell us little about climate attributes pertinent to farmers and food production.