Perennial grasslands have well-documented sensitivity to changing precipitation regimes generally and drought in particular. Recent analyses indicate that relative to average or wet years, extreme drought years are characterized by first, an almost complete absence of large daily precipitation events, but also long periods between precipitation events (extended periods of consecutive dry days) and fewer precipitation events. In ecosystems where most annual precipitation occurs during the growing season, drought can be imposed experimentally by reducing a fixed proportion of each rainfall event throughout the growing season. This “chronic drought” eliminates large daily events but only moderately affects event number and the period of time between precipitation events. Alternatively, precipitation reductions can be imposed by completely removing all precipitation events for a proportion of the growing season. In this case, precipitation can be reduced by a similar amount, but this shorter, more “intense drought” extends the period of time between events and reduces event number dramatically. We contrasted aboveground net primary production (ANPP) responses to these two types of drought over three consecutive years in four native grasslands spanning a precipitation gradient in the Central US. Grasslands included the semi-arid shortgrass steppe in Colorado, northern mixed grass prairie in Wyoming, southern mixed grass prairie in Kansas, and mesic tallgrass prairie in Kansas.
Results/Conclusions
The imposed droughts resulted in a 66% reduction in growing season rainfall and an approximate 50% reduction in annual precipitation for all sites. Sensitivity to drought regardless of type varied substantially with ANPP in the northern mixed grass prairie insensitive to either type of precipitation reduction, but the other grasslands responding more strongly to precipitation removal. The maximum reduction in ANPP relative to ambient (49%) was measured in the driest grassland (shortgrass steppe). Across the three sites that responded to drought, drought type (intense vs. chronic) was least important in the semi-arid shortgrass steppe. Here, ANPP was reduced by a similar amount regardless of drought type. In contrast, intense drought, imposed by a shorter period of complete rainfall exclusion, led to almost twice the reduction in ANPP in the more mesic grasslands compared to chronic drought. These results are consistent with results from previous precipitation experiments indicating that arid ecosystems are primarily sensitive to alterations in event size whereas mesic ecosystems respond more to alterations in the time between precipitation events.