COS 195-4 - Sensitivity vs. asynchrony: Responses of grassland net primary production to precipitation variability

Abstract: Two common themes characterize much recent research in grassland ecology. The first is primarily associated with community ecology and focuses on species asynchrony as a mechanism to stabilize aboveground net primary production (ANPP) in response to precipitation variability over time. In most cases stability is quantified by the variance:mean ratio of ANPP over time. Asynchrony occurs when the decline of some species in a community is compensated for by increases among other species resulting in low temporal variability in ANPP. Asynchrony is often positively correlated with species richness. A second theme more often associated with ecosystem ecology assesses sensitivity of ANPP to interannual variation in precipitation. Sensitivity can be measured as the slope of the relationship between annual precipitation and ANPP, where a strong positive slope indicates high sensitivity to variability in precipitation. These two concepts, asynchrony and sensitivity, are potentially contradictory and in need of reconciliation. We analyzed long-term ANPP data from the Cedar Creek and Jena biodiversity experiments to quantify sensitivity and asynchrony in relationship to species richness and interannual variability in precipitation. We hypothesized that (1) sensitivity to interannual variability would decrease with increasing species richness at Jena and CDR, and (2) asynchrony was correlated with stability.



We found high interannual variability in species abundances at both Jena and CDR. At CDR, total ANPP was insensitive to interannual variation in precipitation, with temporal slopes close to zero. Surprisingly, slopes became slightly steeper in the high diversity plots. Furthermore, relationships based on the change in ANPP relative to the change in PPT from year to year resulted in negative slopes in monocultures and positive slopes in high species richness plots. That is, sensitivity of ANPP increased with species richness. Nevertheless, stability and asynchrony are strongly correlated at CDR. We found similar patterns at Jena, although the slope of the relationship between change in ANPP relative to change in PPT from year to year was steeper at Jena than CDR. Together, these results show that asynchrony is correlated with stability but that these communities are generally stable to begin with because ambient variation in PPT in these mesic grasslands rarely leads to prolonged water deficits that would lower ANPP. In the future, investigators should combine analysis of sensitivity with analysis of asynchrony to assess the degree to which grassland ANPP fluctuates over time, and to assess the relative importance of asynchrony for reducing inherent interannual variability in aboveground production.