Fire-maintained plant species diversity in wet pine savannas: Release from competition or release from dormancy?

Background/Question/Methods - Fire-maintained pine savannas of the Coastal Plain of the southeastern US are recognized as having extraordinarily species-rich communities of endemic plants. The role that fire plays in maintaining such diversity remains unclear in part because of the difficulty in distinguishing release from competition from release from dormancy, which are not necessarily equivalent. I used two complementary field experiments to test the hypothesis that increases in apparent plant species diversity driven by frequent simulated fires are associated with increases in survivorship of species that increase in response to frequent simulated fires. Specifically, in one experiment I simulated the effect of frequent fire on reproductive species diversity using a combination of clipping and litter removal in different months (mix of October and May/June vs. May/June exclusively; four clippings over four years; n = 20 4-m² plots). In the second experiment, I monitored the survival of transplants of five perennial forb species in treated and untreated plots in 2017 and 2018 to determine whether repeated clipping increased their survival. The five species (Bigelowia nudata, Rhynchospora latifolia, Carphephorus pseudoliatris, Eriocaulon compressum, and Triantha racemosa) were chosen because they increased significantly in response to frequent clipping and flowered more after clipping.

Results/Conclusions - Initially, October clipping increased apparent diversity by increasing ephemeral (e.g., annual) species, but after four years of clipping, these species did not differ significantly between clipping season treatments. The overall reproductive species diversity and the abundance of the five perennial target species were significantly greater in clipped plots than in adjacent unclipped plots by the end of the study. Nevertheless, survival of the transplants was not significantly greater in clipped than in unclipped plots and was not affected by the clipping season. Flowering of these species was greater in clipped plots than in unclipped plots. October clipping was more effective than May/June clipping at increasing flowering in B. nudata and E. compressum, but natural densities of these species did not differ in response to clipping season in 2018, suggesting no cumulative effects over time. Results suggest that ultra-frequent simulated fires in this system, at least in the short term, can give the appearance of maintaining high plant species diversity, perhaps by releasing subordinate fire-adapted species from dormancy. Results do not support the hypothesis that frequent simulated fire increases survival of subordinate species following their emergence, nor was there evidence that increased fire-stimulated flowering resulted in significant increases in density.