Separating environmental filtering and biotic resistance of invaders using the functional species pool concept

Background/Question/Methods

Theory suggests that limiting similarity should cause invaders more closely related to the resident community to be less successful. However, empirical results often show positive or neutral relationships. These results are usually attributed to habitat matching, although the exclusion of conserved weak phenotypes could also cause this pattern. Differentiating among these mechanisms is challenging and will require separating the effects of environmental and biotic filters. Here, we adapt the functional species pool concept to explore relatedness-invasion relationships in a native grassland. Using small scale community composition data, we estimated which absent species could co-occur with the resident community; these dark species therefore belong to the functional species pool. Next, we used phylogenetic relatedness between observed and dark species and between functional pool members and the rest of the regional species list to estimate the likelihood of different species passing both biotic and environmental filters invading based on their relatedness to observed and functional pool species respectively. We used survival data for 1065 individuals of 18 species transplanted into bare ground to test our predictions of environmental filtering and the survival of another 1065 individuals transplanted into intact vegetation to test our predictions of biotic resistance.

Results/Conclusions

After four years, >400 transplants survived when neighbouring vegetation was removed, but <250 when the resident community was intact indicating significant environmental filtering and biotic resistance. In general, species that were more closely related to the functional species pool were more likely to pass the environmental filters (i.e. survive without neighbours) and our models successfully predicted the majority of these events. The patterns were far less consistent when considering biotic resistance, with evidence for limiting similarity, weak phenotype exclusion, and neutral dynamics. Despite this contingency, our model still successfully predicted the strength of biotic resistance in most cases. However, both environmental filtering and biotic resistance were still in large part driven by underlying environmental conditions. Phylogenetic relatedness played a less important role, consistent with previous work showing phylogeny to be a poor proxy for function in this system. Further, our results cast doubt on the likelihood of any larger-scale general relationship between relatedness and invasion success given the high degree of variability in these relationships seen over relatively small scales.