It has been experimentally determined that maize is particularly sensitive to drought during the flowering period of development. However, this has not been studied across highly varied geographic regions growing different cultivars of maize. Furthermore, the exact pattern of sensitivity will determine how much control farmers have to avoid particularly damaging weather. This will be particularly relevant as the climate warms. Here we examine 20 states in the United States Department of Agriculture/National Agriculture Statistics Service (USDA/NASS) database which contains data on six distinct phases of the maize development cycle: planting, silking, doughing, dented, mature, and harvested. We combine these development phases with data from 549 weather stations to determine the sensitivity in each phase of development to the number of Growing Degree Days, Killing Degree Days, and precipitation totals using a multiple linear regression model.
Results/Conclusions
We find highly variable sensitivity both spatially and temporally. The silking and dentings stages appear to be the most sensitive and may be as much as an order of magnitude more sensitive than the least though the significance of this result remains to be more fully assessed. Furthermore, the pattern of sensitivity varies by state though several distinct patterns appear to be geographically linked. Finally, we were able to determine distinct sensitivity to both precipitation deficit and hot temperatures, which has been challenging for regression models of this type given the close correlations between hot temperatures and low precipitation. These results suggest that insofar as there appears to be a broad or double peak in sensitivity it may be difficult to simply adjust planting schedules to avoid damaging late season heat. Despite this, an improved quantification of developmental sensitivity may still help to produce improved management schedules and illustrate management types that will be more appropriate to a hotter environment.