Higher beta-diversity observed for herbs than for woody species in a neo-tropical forest understory is driven by stronger resource-based habitat filtering

Background/Question/Methods

Most of Earth’s plant diversity is housed in tropical forests. However, the majority of information that we have on the local-scale ecology of tropical plants is for woody communities only. Here, we calculate spatial variation in alpha and beta-diversity of a tropical herb community, and determine whether this variation is driven by resource-based habitat associations (i.e. response to overstory shade, soil chemistry, and topography). We address two questions:1) How do alpha and beta-diversity differ between herbs and woody seedlings? 2) do the processes generating observed spatial patterns of alpha and beta-diversity differ between these growth-forms? Herbaceous and woody species abundances and environmental variables were measured within 115 20×20m quadrats. Previously delimited habitat categories were used to compare species richness and community compositions for understory herbs, trees, and lianas. Variance partitioning was next used to measure the correlations between environmental and spatial variables with species richness and community composition. Results were compared between herb and woody communities to determine which processes were most important for each plant type. Finally, we tested whether the same sites harbor similar relative diversity levels by correlating species richness and the local site contribution to beta-diversity (LCBD) among the herbaceous and woody communities.

Results/Conclusions

Sampling the herbaceous community increased total understory richness by ~20%. The swamp habitat harbored much higher relative herb richness than woody richness, while the slope habitat contained the highest liana richness. Beta-diversity was much high in the herbaceous layer, regardless of the metric used, and after controlling for differences in gamma-diversity. Interestingly, plots that had high species richness or beta-diversity (i.e., LCBD) were not the same between herbaceous and woody communities (correlations between communities were either negative or weakly positive). Habitat type explained up to four times the amount of variation in herb composition than tree or liana composition (PERMANOVA for Herbs: F = 13.28, R² = 0.29, Tree: F = 5.40, R² = 0.13, Liana: F = 2.65, R² = 0.07). Similar results were obtained when using variance partitioning on measured environmental variables and spatial descriptors (total variation for herbs, trees, and lianas was 24%, 9%, and 7%, respectively). Focusing on individual environmental variables revealed that different gradients were correlated with herb and woody richness and composition. Our results suggest that contrasting mechanisms drive herbaceous and woody communities, which helps inform our understanding of tropical forest community assembly and diversity.