Taxonomic and functional diversity responses to soil nutrients, natural enemies, and community composition
Soil resources, herbivory, and plant pathogens act as strong determinants of plant community composition. Community assembly theory predicts that these factors act as performance filters on species diversity, and that their relative strengths vary temporally. Therefore, examining temporal dynamics of species composition alongside plant physiological traits in response to experimental treatments can yield mechanistic insight into community assembly. Using experimental old field communities, we examined the responses of plant assemblages to fertilization, insect herbivore and plant pathogen (enemy) exclusion, and priority effects of extant vegetation. Six native, perennial herbaceous species were transplanted into denuded 1m2 plots in an old field in Durham, NC in 2011 at two different diversity levels (monocultures and polycultures), and factorially manipulated with soil fertilization (NPK) and enemy exclusion We collected species abundance data at seven time points, and measured species’ specific leaf area (SLA) and maximum vegetative height in each plot to examine how treatments acted as performance filters. We use linear mixed models to examine the response of species composition and trait data to the experimental treatments. Species were treated as random intercepts in trait analyses to distinguish interspecific and intraspecific variation. Repeated measure ANOVAs were used to quantify temporal dynamics of species composition.
Enemy exclusion and fertilization reduced species richness (p<0.001). A significant interaction was also observed (p<0.01) between the treatments, showing fertilization effects overwhelmed enemy exclusion effects when in tandem. Analysis of Bray-Curtis dissimilarity index across time within plots revealed that fertilization led to increased species turnover (p<0.001), while enemy exclusion led to decreased species turnover (p<0.001). Community-level SLA increased with fertilization (p<0.001), decreased with enemies exclusion (p<0.01), and showed no response to the diversity treatment (p=0.630). When random intercepts were included to distinguish interspecific and intraspecific variation, model fit was improved, and enemy exclusion showed no significant effect (p=0.631), while fertilization and polycultures increased SLA (p<0.05; p<0.01). Maximum height increased in response to fertilization and enemy exclusion both with and without species’ random intercepts included (all: p<0.001). Polyculture treatments had decreased height (p<0.01) without random effects, but no difference with them (p=0.895). Shifts in the significance of treatments between models suggests that treatments differentially affect inter- and intraspecific trait diversity. Collectively, these results capture complex assembly dynamics occurring in the experiment. The distinction of enemy and fertilization effects on temporal species turnover indicates they decrease richness through different pathways. Moreover, trait based analysis demonstrates distinct treatment selection effects within communities.