Discovering the forest among the trees: Testing community assembly theory in the forgotten layers of temperate and tropical ecosystems

Forests are the most diverse and productive terrestrial ecosystems on Earth, so sustainably managing them for the future is a major global challenge. Our understanding of forest plant diversity, however, relies almost exclusively on the study of trees. Here, I show that the other growth-forms (shrubs, lianas, herbs, epiphytes) in fact make up the majority of vascular plant species in both tropical and temperate forests. By comparing the distribution of species among growth-forms for over 3,400 species in 18 forests in the Americas, I construct a high-resolution quantification of growth-form diversity across two important ecological regions. I also quantify the physical distribution of plant species among forest layers, or where among the vertical strata plants ultimately live their adult lives. Plant diversity is strongly concentrated on the forest floor in temperate forests, but is evenly distributed among the forest strata in tropical forests. I then use three large-scale field experiments to test drivers of community assembly specifically in these understudied yet species-rich plant communities. First, I simultaneously test the effects of browsing pressure, understory competitors, and the controversial practice of salvage logging on herb regeneration following a large-scale windthrow in Pennsylvania. Although salvaging was thought to be inimical to forest recovery, my results demonstrate that the herbaceous layer is surprisingly resilient to intense salvage logging, interspecific competition, and vertebrate browsing. I also show that salvage logging does not eliminate tip-up mounds, one important structural legacy of windthrow. Importantly, the patterns I show for woody species do not match those of herbaceous plants, which make up 80% of the temperate forest species. Second, I experimentally test hypotheses regarding the assembly of diverse epiphyte communities in a Panamanian cloud forest. In particular, I test the degree to which host tree substrate traits regulate the early germination and survival of epiphytes. All in all, my research provides evidence that the major drivers of tree community change do not necessarily drive forest community change. Thus, forest management plans that encompass a broader suite of plant growth-forms will be essential to global biodiversity conservation efforts.