White-tailed deer are a biotic filter during community assembly, reducing species and phylogenetic diversity

Abstract

Community assembly entails a filtering process, where species found in a local community are those that can pass through environmental (abiotic) and biotic filters and successfully compete. Previous research has demonstrated the ability of white-tailed deer (Odocoileus virginianus) to reduce species diversity and favour browse-tolerant plant communities. In this study, we expand on our previous work by investigating deer as a possible biotic filter altering local plant community assembly. We used replicated 23-year-old deer exclosures to experimentally assess the effects of deer on species diversity (H'), richness (SR), phylogenetic community structure and phylogenetic diversity in paired browsed (control) and unbrowsed (exclosed) plots. Additionally, we developed a deer-browsing susceptibility index (DBSI) to assess the vulnerability of local species to deer. Deer browsing caused a 12 % reduction in H' and 17 % reduction in SR, consistent with previous studies. Furthermore, browsing reduced phylogenetic diversity by 63 %, causing significant phylogenetic clustering. Overall, graminoids were the least vulnerable to deer browsing based on DBSI calculations. These findings demonstrate that deer are a significant driver of plant community assembly due to their role as a selective browser, or more generally, as a biotic filter. This study highlights the importance of knowledge about the plant tree of life in assessing the effects of biotic filters on plant communities. Application of such knowledge has considerable potential to advance our understanding of plant community assembly.

Keywords: Browsing; herbivory; phylogenetic clustering; phylogenetic community ecology; plant–animal interactions; species diversity..

Figure 1.

Rooted phylogenetic tree (using ITS, rbcL and trnL–trnF regions) of all species present at the study site. Phylogeny was estimated using maximum likelihood in RAxML, as described in the Methods.

Figure 2.

Mean species richness (A), Shannon–Weiner diversity (B) and mean pairwise phylogenetic distance (C) in exclosure (grey bars) and control (blue bars) areas across all years (2006–2012, excluding 2007). Different letters indicate statistical significance between groups at the P < 0.05 level, as tested using ANOVA in a linear mixed effects model. Error bars are ±1 SE.

Figure 3.

Net relatedness index of exclosure areas (squares) and control areas (circles) from 2006 to 2012 (excluding 2007). Points highlighted in green indicate NRI values significantly greater than expected by chance (P < 0.05, based on 999 random permutations of the tip states on the phylogeny).

Figure 4.

A stacked bar graph representing the mean calculated DBSI of each species present in more than one exclosure in two or more years. Light grey bars indicate the proportion of the species present in control areas, while dark grey bars indicate the proportion of the species present in exclosures. Species are ordered left to right from the least susceptible (S. purpurascens) to most susceptible (T. canadensis).