Null models reveal the relationship between local and regional diversity and stability in pond metacommunities

Background/Question/Methods

Despite the wealth of information on how communities vary through space (beta diversity), temporal patterns of community variation have received comparatively little investigation. Variation in local and/or regional diversity can influence the values of most spatial beta diversity metrics, necessitating null models that can disentangle changes in beta diversity due to sampling versus biological processes across diversity gradients. We applied this concept to patterns of community compositional variation through time (termed temporal beta diversity), by first creating a simulation model that varied extinction rate, local diversity, and regional diversity, and then calculating temporal beta diversity. Next, we modified and applied this null model to data from zooplankton in pond metacommunities to reveal whether observed temporal beta diversity values deviate from those expected due to stochastic sampling effects across a regional and local diversity gradient. 

Results/Conclusions

The simulation model showed that the relationship between temporal beta diversity and regional or local diversity depends on the relationship between diversity and extinction rate. Temporal beta diversity can increase, decrease, or be unrelated to local and regional diversity in the absence of any deterministic processes. The importance of this null model is emphasized when applied to zooplankton in pond metacommunities. Prior to application of the null model, observed temporal beta diversity was negatively related to local and regional diversity. After accounting for stochastic sampling effects, however, the trend reversed, indicating that more diverse systems are less compositionally stable.