Consequences of Fragment for Woody Plant Communities: A Study of Reservoir Islands

Forest fragmentation has been overwhelmingly cited as a major threat to the biodiversity and conservation of forested plant communities worldwide. Here I take a community- and species-specific approach to examine how species richness and composition respond to forest fragmentation. I conducted this research on a series of 35 small forested islands that were fragments of continuous forest created by impoundment of the Savannah River in the Southeastern Piedmont of the United States. I paired the islands with 10 mainland forest sites representing large remnant forest with only a single edge exposed along two reservoirs. Species richness was positively related to island area, as predicted by the species-area relationship, and islands in general had greater species richness than mainland sites because of an addition of liana and shrub species uncommon to oak-hickory forests. Due to the increase of lianas, shrubs and pioneer trees in small forested fragments I detected large differences in the plant communities that have developed on these sites over the last 40 to 70 years. In addition, it appears that islands will continue to diverge from mainland forest over time, likely degrading to the point that the species of oak-hickory forest may disappear and convert entirely to disturbance-tolerant early seral communities. Moreover, these fragments had invasive plants species that were less common in the more intact forest. I used a seedling outplanting experiment to determine if two of the more common, non-native woody invaders are likely to invade intact forest interior sites. Indeed, the non-native woody liana, Lonicera japonica appears to have the ability to not only survive, but also thrive under interior forest. In contrast, Albizia julibrissin does not appear to be a major threat to forest interiors unless there is a large canopy disturbance that increases light to seedlings. Finally, I used this study system to test whether species richness estimators provide accurate estimates and should continue to be used to examine important ecological patterns. I found that all 10 of the ones examined were so imprecise that none of them detected the true species-area relationship found across the forest fragments. Use of species richness estimators, in place of true richness, should therefore be used with extreme caution if the goal is to describe patterns in species richness across a set of sites. Overall, my research highlights how much we still have yet to learn about generalities associated with forest fragmentation and species richness estimation techniques