COS 2-5
Local and regional drivers of plant species richness in Ozark glades
Plant species richness can be influenced by multiple factors including local environment, landscape connectivity, and disturbance history, but understanding the relative importance of these factors for plant communities remains a challenge for ecologists. Previous studies of the effects of habitat spatial structure on plant communities have reached inconsistent conclusions, but often do not (1) control for additional factors influencing communities, such as local environmental conditions, or (2) consider differential responses of unique plant guilds, such as habitat specialists. Here, we ask what factors drive total species richness and richness of specialist species in Ozark dolomite glades—fire-adapted xeric grasslands that vary significantly in spatial structure and environmental characteristics. We hypothesized that total richness is driven primarily by local environment, and that landscape connectivity has a greater effect on specialist richness than total richness. To test these hypotheses, we sampled herbaceous plant communities at 56 glades across the Missouri Ozarks that are managed with frequent, low-intensity fire, and used structural equation modeling (SEM) to analyze the effects of local soil characteristics, topographic slope, landscape connectivity (measured using the proximity index metric), and fire history on plant community response variables.
Results/Conclusions
We found that both local and regional drivers contribute to local species richness. Local soil fertility was the strongest predictor of total species richness in glades, and landscape connectivity and time since fire also had significant effects. As we hypothesized, landscape connectivity had a stronger influence on glade specialist richness than it did on total species richness, although we were able to explain less of the overall variation in specialist richness than total species richness. Topographic slope and soil were also significant predictors of glade specialist richness, but fire history did not have a significant influence on specialists. Our results underscore the importance of considering multiple predictors and indirect effects in studies of plant diversity; our SEM indicated significant relationships that were not apparent in bivariate analyses. Our results suggest that local environmental conditions may set an upper limit for restoration of plant species richness in glades but that landscape connectivity may be especially important for the conservation of specialist species.