Assessing the stochastic and deterministic factors contributing to spatial and temporal turnover in California grasslands

Background/Question/Methods

Patterns of turnover in species composition can provide useful insights into the mechanisms that maintain diversity over space and time. Variation in both time and space may be driven by sampling effects, stochastic mechanisms such as dispersal, or by deterministic factors such as environmental filtering and plant-plant interactions. While this framework often takes a “bottom-up” approach (i.e. focusing on the relative importance of abiotic factors and plant-plant interactions), it can also be expanded to address how “top-down” interactions such as grazing can influence turnover over space and time. Here, we examined the relationship between turnover and key environmental drivers, specifically soil properties and precipitation, previously shown to influence species composition within California grasslands. In three East Bay Regional Parks in California, species composition was recorded along four transects in 500 m² circular plots from 2005-2011 in grazed and ungrazed areas. We coupled spatial and temporal turnover measurements with null modeling and regression analysis to test the predictions that 1) grazers maintain diversity by promoting spatial heterogeneity in composition (turnover over space), 2) precipitation maintains diversity by promoting temporal heterogeneity in composition (turnover over time), and 3) the influence of precipitation on temporal heterogeneity is tempered by grazing.  

Results/Conclusions

Observed spatial and temporal turnover were positively related with one another, such that plots with high spatial turnover also had high temporal turnover.  This pattern held irrespective of grazing, and is consistent with the expectation of a sampling effect: an area that supports high diversity is also likely to show greater turnover in diversity over space and time. Interestingly, after accounting for these sampling effects with null modeling, the relationship within the ungrazed communities shifted such that areas with high spatial turnover were more stable over time.  For instance, low rainfall caused composition to shift (high temporal turnover) but also converge over space (low spatial turnover). This signature of precipitation was only evident in ungrazed communities; in grazed communities, spatial and temporal turnover were no longer related once differences in diversity were taken into account. Grazed communities maintained high spatial turnover across all years, tempering the influence of precipitation.  Our results show that sampling effects may obscure the ecological processes influencing community structure within California grasslands and that management practices such as grazing may produce patterns that resemble stochastic processes by maintaining spatial variability in composition across time.