Examining spatio-temporal patterns of functional diversity to deconstruct Darwin's naturalization conundrum

Background/Question/Methods Darwin’s naturalization conundrum describes the paradox that community assembly is regulated by two contrasting processes, environmental filtering and competitive interactions, which predict both similarity and distinctiveness of species to be important for invasion success.  Here, we examine how functional diversity has changed and determine whether environmental filtering or competitive interactions regulate the Laurentian Great Lakes fish community from 1870 to 2010, which has experienced frequent introductions of non-native species and extirpations of native species.  We determined the most appropriate temporal scale necessary to understand community dynamics and subsequently analyzed how functional diversity has changed over time by decade from 1870 to 2010 at three spatial scales (regional, lake, habitat) to account for potential non-interactions between species at the regional and lake level.  We also determined which process, environmental filtering or competitive interactions, is more important in regulating community assembly and maintenance by comparing observed patterns to what we would expect due to random chance.

Results/Conclusions With the exception of one lake, all analyses show that functional diversity and species richness has increased over time.  By comparing observed functional diversity to what we expect by chance, we have found that environmental filtering regulates community assembly at the regional and lake level; however, when examining functional diversity at the habitat level, the regulating processes become more context dependent.  This study is the first to examine diversity patterns and Darwin’s conundrum by integrating long-term, large-scale, and high-resolution temporal data and multiple spatial scales.  Taken together, our results suggest that examining Darwin’s conundrum is highly context dependent, and more studies at the appropriate spatial and temporal scales are needed.