Conserved microhabitat affinities drive phylogenetic clustering among Psychotria on Barro Colorado Island, Panama

Background/Question/Methods Tropical forest plant genera with high alpha-diversity, such as Neotropical Psychotria, pose a challenge to coexistence theory because ecological similarity, and hence the likelihood of competitive exclusion, is expected to vary inversely with phylogenetic distance between species. Two hypotheses could explain the coexistence of numerous congeneric species. Competitive exclusion could be avoided if related species have diverged evolutionarily to exploit limited resources in ways that are sufficiently different to permit coexistence at equilibrium. Alternatively, diffuse competition in hyperdiverse communities may select for convergence on a general tolerance of the biotic and abiotic environment, in which case co-occurrence would be transient and greatly influenced by population stochasticity. We evaluated the distributions of Psychotria species along moisture and light gradients on Barro Colorado Island (BCI), Panama, using a canonical correspondence analysis and community phylogenetic analysis to answer three questions. First, do microhabitat distributions differ among the 21 congeneric species? Second, do species differences in physiological traits contribute to segregation along environmental gradients? And third, do phylogenetic patterns to species distributions and trait evolution suggest a role for competitive interactions in community structure? Results/Conclusions Rates of co-occurrence were high, with at least five species encountered in 16% of plots. One species was associated with high-light environments, one with dry microsites, and five with wet microsites, while the remainder of Psychotria species differences in distribution were limited. Of the physiological traits examined, only leaf water potential showed significant phylogenetic signal to its evolution, while phylogenetic signal for the soil moisture habitat association was marginally significant with only 16 species. Mean values of three photosynthetic traits co-occurring species varied significantly less among co-occurring species than in null communities. Co-occurring Psychotria were significantly more closely related than predicted by chance (i.e. were phylogenetically clustered). Interspecific differences in microhabitat distributions support the predictions of equilibrial coexistence owing to species differences in resource use, yet the limited differences observed here may not be substantial enough to maintain the coexistence of 21 congeneric species. The conclusion that co-occurring species are both physiologically similar and closely related suggests that species similarities do not preclude coexistence on an ecological timescale. These results suggest that both extreme views of the role that species differences play in coexistence among congeners are partially correct in regards to BCI Psychotria.