Ungulate exclusion and tree community database

The dataset presents data collected from an inventory of trees and shrubs in a novel experimental layout to investigate the contrasting effects of ungulate herbivory on woody vegetation in nutrient-rich versus nutrient-poor sites (i.e. termite mounds versus off-mound sites with and without large mammalian herbivores) from Lake Mburo National Park in western Uganda. We hypothesized that (1) the resource-rich plots (i.e. termite mounds) would have a higher woody plant species richness and diversity (Støen et al., 2013; Joseph et al., 2015), and a lower evenness (Støen et al., 2013) compared to the resource-poor adjacent savanna matrix, while ungulate exclusion would not affect these diversity measures on any treatment (Støen et al., 2013; Joseph et al. 2015); (2) within-treatment beta-diversity would be higher on the off-mound areas compared with termite mounds and decrease with ungulate exclusion on the off-mound areas only (Støen et al., 2013); (3) woody plant assemblages on resource-rich mounds would be more similar than the resource-poor savanna matrix areas and herbivory would reduce contrasts between mound and matrix sites (Seymour et al., 2016). Findings from the analyzed data indicated that excluding large herbivores had little discernible effect on the species richness and plot-level diversity of woody species on nutrient-rich mounds but increased both richness and diversity on the relatively nutrient-poor off-mound sites. We conclude that ungulates reduce woody species diversity in resource-poor savanna areas but have little influence in resource-rich areas.