Background/Question/Methods Decades of research have documented that the sensitivity of dryland aboveground net primary productivity (ANPP) to inter-annual variation in precipitation declines as average precipitation increases. Although several hypotheses have been proposed to explain this pattern in ANPP-precipitation sensitivities, the mechanisms driving the pattern still remain unclear. Our ability to identify the prevalence of each hypothesis has been limited, in part, because analyses of ANPP dynamics have often focused at the aggregate community level, yet many of the hypotheses invoke mechanisms operating at the level of individual species. Using data on individual species ANPP dynamics from four long-term ecological research (LTER) sites across a precipitation gradient in the central and western US, we used multivariate auto-regressive models to quantify how individual species dynamics lead to observed community-level ANPP patterns. We evaluated support for four non-exclusive hypotheses explaining lower sensitivity of ANPP in mesic locations: 1) greater secondary resource limitation, 2) compensatory dynamics, 3) lagged responses, and 4) changes in species life history. Results/Conclusions Preliminary results indicate that declining community-level ANPP sensitivity to precipitation is driven in part by compensatory dynamics: in the most mesic sites a nearly equal number of species respond positively and negatively to precipitation. This led to a lower aggregate community sensitivity than in more xeric sites where species sensitivities to rainfall were nearly uniformly positive. However, somewhat surprisingly, one of the strongest drivers of ANPP sensitivity patterns appears to be differences in species life history across sites. In more xeric sites, increases in ANPP are closely coupled with increases in species richness, indicating that many species (e.g. annual plants) remain dormant during dry years but capitalize on wet years for rapid growth and reproduction, leading to large increases in ANPP in wet years. Our results indicate that a consideration of interactions between life history strategies and climate variability may be critical to understanding differences in community and ecosystem dynamics across climate gradients.