Effects of connectivity and patch shape on patterns of plant species spillover change over time

Background/Question/Methods

Land use change and associated fragmentation threaten biodiversity globally, but habitat connectivity via corridors increases plant species richness in remnant habitat patches and helps mitigate biodiversity losses. The connectivity and shape of habitat remnants also influences diversity in the surrounding matrix, but it is unclear how these associations change through time. We have large scale, well replicated, experimental landscapes to test the effect of habitat connectivity and patch shape (i.e., edge to area ratio) in the longleaf pine community of South Carolina. These landscapes are composed of savanna patches surrounded by a pine forest matrix. There are three patch types surrounding a central patch: 1) connected (i.e., corridor to center) 2) winged (high edge to area) 3) rectangular (low edge to area). We asked how the effect of patch type on plant species richness spillover (i.e., colonization) into the surrounding forest matrix that we detected in 2007 has changed since then and how they will relate to within patch changes. To address these,  we resampled species presence/absence along 50m transects starting at the patch edge and moving into the forested matrix in 2014 and compared patterns of species spillover to those in 2007

Results/Conclusions

In 2014, we found a change in the patterns of plant species spillover. In 2007, connected patches showed the highest species richness spillover compared to winged and rectangular patches. In 2014, connected patches no longer showed the highest species richness spillover. Connected patches had no significant difference of species spillover between winged or rectangular patches. However, winged patches show a 15% increase in species over rectangular patches (p = 0.048). Winged patches show a marginally significant 20% increase in wind dispersed species compared to connected and rectangular patches (p = 0.081), suggesting that winged patches may be more amenable to wind dispersed species. However, among animal dispersed species, both connected and winged show a significant 15% increase over rectangular species (p< 0.05), suggesting that larger edge to area ratio may be driving animal dispersal into the matrix. Our results show that processes associated with connectivity and patch shape driving species spillover may be dynamic and driven by changes of within patch processes over time.