Semester of Graduation

Fall 2020


Master of Science (MS)


Biological Sciences

Document Type



Savanna tree populations typically share two demographic characteristics, high mortality of juvenile stages and low mortality of established trees. Juvenile savanna tree species escape demographic bottlenecks when they reach a safe size and annual mortality becomes low. Longleaf pine (Pinus palustris Mill.) is the dominant tree species in pine savannas of the southeastern United States. The seeds and juvenile seedlings experience bottlenecks to recruitment to a safe size, the grass stage. Low intensity, frequent savanna fires facilitate establishment of seedlings, but also kill juvenile seedlings that have not grown sufficient above ground biomass. We hypothesized that juvenile seedlings escape the bottleneck by recruiting in safe sites within open areas away from overstory pines. These safe sites are small-scale microhabitats in the ground layer that reduce fire effects and competition with vegetation, thus facilitating seedling survival and growth to the grass stage. Microhabitats chosen for study at the Wade Tract in Thomasville, GA were pine-generated microhabitats, naturally occurring fallen crowns and stumps of dead pine trees that have burned in successive fires and hardwood-generated microhabitats, patches with abundant stems of oaks and other hardwood species. Plots were established both within and outside of these microhabitats in open areas away from overstory pines to study aspects of our hypothesis in two field studies. Results of the first study show that the microhabitats have reduced grass cover compared to the surrounding ground layer matrix. Among microhabitat variation included reduced grasses with increased shrub cover in oak patches and reduced shrubs with increased grass cover in pine microhabitats. Recruitment of grass stage longleaf pine occurred at a higher rate in the microhabitats than in the surrounding ground layer matrix. The second field study was an investigation of seedling demography in the plots in each microhabitat and in the areas outside those microhabitats, but within open areas v away from overstory pines. Seeds collected in a mast year were placed in caged and uncaged quadrats within plots. These were censused for survival and after germination were censused for survival and growth over over the course of 1.5 years. The number of germinating seeds was greater in caged than uncaged plots. Over the course of the experiment, two prescribed fires reduced survival of seeds similarly inside and outside the altered microhabitats. Larger seedlings had higher survival of the second fire than smaller seedlings. Seedlings grew largest in pine microhabitats, especially around stumps. These microhabitats may serve as safe sites to recruitment in frequently burned savannas. The results of this study indicate that altered microhabitats, especially pine stumps, may constitute safe sites for recruitment of new longleaf pine into the old-growth stand on the Wade Tract.

Committee Chair

Platt, William