American Maize: Climate Change, Adaptation, and Spatio-Temporal Variation in Temperature Sensitivity

Abstract

Agricultural production is vulnerable to climate change. However, this vulnerability can be reduced by adapting food crops to a hotter climate. Many studies have ignored adaptation when quantifying the effect of climate change on crop yield, which has likely overestimated yield losses. Therefore, it is necessary to quantify agriculture's adaptive potential to climate change. Such work is challenging because there are no historical analogues to current or future warming. In place of such a precedent this work explores the varying sensitivity of maize yield to elevated temperatures through a suite of multiple linear regression models. These models use high resolution yield and crop development data available since 1981 in the United States to account for overlooked features of maize physiology and agricultural management. The results of these models substantially alter estimates of how crops will respond to a warming environment. The studies here illustrate how finer scale details can be incorporated into broader regional models. Temperature sensitivity is found to vary with local climatology indicating that maize cultivars are adapted to their particular environment. Incorporating this historical adaptation into estimates of yield loss substantially reduces the effect of a modest warming. A physiological basis for spatial adaptation is apparent when maize development data are incorporated into the model -- cooler regions accelerate through sensitive development phases faster than hotter areas. The development data also suggest that crop development has been adapted to the seasonal cycle and that a non-trivial portion of the temporal trend in maize yield has resulted from management adjustments. Finally, the importance of spatio-temporal variation in temperature sensitivity is highlighted through case studies of recent years with record-setting yield losses. Spatial and/or temporal variation in temperature sensitivity is necessary to reduce bias in estimates of yield loss in these years. This work builds from previous conclusions regarding the negative effects of hot temperatures, and suggests that while hotter temperatures will harm maize yields there are steps that farmers might take to manage and reduce these losses. Taken together these results quantify how extant adaptation may help to ameliorate yield losses in a hotter future.