Heterogeneous effects of warming and drought on selected wheat variety yields

Abstract

Climate change is likely to significantly impact agricultural production in the Great Plains region of the Central United States. This study estimated the impact of changes in temperature and precipitation on wheat (triticum aestivum) variety yield distributions using the moment-based maximum entropy (MBME) model. This approach allows for quantification of potential weather impacts on the yield distribution, and allows these effects to vary across varieties. The unique data set matches wheat variety trial data for 1985 to 2011 with weather data from the exact trial site for 11 locations throughout Kansas. Ten widely-planted varieties with a range of biotic and abiotic characteristics were included for comparison. Weather scenarios were simulated for baseline, increased temperature (one-degree Celsius warming), decreased precipitation (tenth-percentile rainfall outcome), and a combination warming and drought scenario. Warming resulted in an 11 % yield reduction, drought a 22 % reduction, and warming and drought a cumulative 33 % reduction. These effects vary across varieties. Alternative measures of yield risk (e.g. yield variance and coefficient of variation) were also constructed under each scenario and a similar pattern of heterogeneous impacts emerges. The key findings are that (i) exposure to warming and drought lead to mean yield reductions coupled with increased yield risk for all varieties, and (ii) newer (post 2005) seed varieties have a yield advantage over older varieties, however this advantage is reduced under warming and drought conditions.