The mechanics of ecological diversity through time: Revisiting colonisation and extirpation process

Background/Question/Methods

Studies of species-time relationships (STR) often employ approaches developed for species-area relationships. One assumption of using these approaches is that the order of temporal sampling is arbitrary, so that changes in diversity observed between any two periods in time would have been equally likely had the temporal order of sampling been reversed. In metacommunities, this assumption only holds if species diversity is at equilibrium, such that sampling through time captures stochastic extinction and colonization events that balance each other. However, most studies of species-time relationships fail to quantify patterns of diversity specific to temporal dynamics, such as successional dynamics and environmental seasonality. We developed a metacommunity approach to understanding species-time relationships by decomposing temporal changes in diversity into total turnover and its two components: colonization (C) and extinction (E) (each measured relative to total turnover).This decomposition allows us to separate total turnover (C+E)from directional turnover, which we define as relative colonization and relative extinction. Using thirteen metacommunities of mites associated with rubber trees (Hevea brasiliensis) in Brazil, we 1) test the assumptions of the STR approach; 2) test trends in directional turnover with time; and 3) identify the environmental drivers of directional and non-directional turnover. 

Results/Conclusions

Our diverse metacommunities (separated by up to thousands and kilometers and sampled over a five year period) all showed temporal patterns of species diversity that were inconsistent with the assumptions of the STR approach. Our metacommunity metrics of total turnover and its components (C and E) indicated why the STR approach was incorrect and also allowed us to identify the environmental conditions that likely structure temporal shifts in colonization and extinction dynamics. Relative colonization was strongly structured by time since leaf emergence and decreased consistently over the season, whereas total turnover was at its lowest at the midpoint between leaf emergence and senescence. These temporal shifts coincided with distinct environmental conditions. Colonization increased with large increases in temperature, and was at a maximum when metacommunities were close to the highest temperatures experienced. In contrast, total turnover decreased with temperature but increased when relative humidity was high. Our results indicate that temporal diversity can be understood by separating colonization and extinction processes from total turnover. When this separation is successful, the environmental conditions that favor high turnover are distinct from those that favor high relative colonization or low extinction.