INTRODUCTION
Herbivory serves as a critical top-down mechanism within plant communities that regulates biodiversity, productivity, and ecosystem function (Picala and Crawley, 1992; Mortensen et al., 2018; Moorhead et al., 2017). Herbivory often influences local-scale spatial ecology of plant communities via direct and indirect impacts (Picala and Crawley, 1992). Herbivores regulate plant communities directly via consumption of plant biomass. Plant communities are impacted indirectly, however, through changes in resource availability by the removal of dominant, highly productive plant species by herbivores that may facilitate the establishment of rare or subordinate species, potentially increasing richness and community evenness (Pacala and Crawley, 1992; Holt, Grover and Tilman, 1994; Becerra, Noge, and Venable, 2009; Mortensen et al., 2018).
Herbivory by vertebrates in temperate ecosystems can reduce plant productivity, increase plant diversity, and alter lower trophic levels and soil chemistry (Sala et. al, 1986; Ritchie, Tilman, and Knops, 1998; Schowalter, 2012; Wilson et. al., 2012; Moorhead et al. 2017; Ramirez et al., 2020). In North American forests, the density of white-tailed deer (Odocoileus virginianus ), a ruminant herbivore, is often strongly related to reductions in understory plant productivity, diversity, and tree recruitment (Ritchie et al., 1998; Shelton et al., 2014; Habeck and Schultz, 2015). Herbivory by mammals can also have cascading impacts on ecosystem function (Van Der Wal et al, 2004; Moorhead et al. 2017; Ramirez et al., 2020). For instance, even small mammal consumption and redistributing of plant biomass have been shown to alter woody biomass and plant community composition leading to indirect effects on plant tissue chemistry and soil enzyme activities (Moorhead et al. 2017). While the ecological impacts of mammalian and insect herbivores have garnered much attention in the literature (Ritchie et al., 1998; Goheen et. al, 2010; Wilson et. al, 2012; Hodkinson, 2012), fewer studies have investigated the role of herbivorous reptiles in regulating plant communities (King, 1996; Olesen and Valido, 2004; Fourqurean et al., 2010).
Herbivorous reptiles are relatively uncommon because of the physiological constraints imposed by ectothermy on thermoregulation and the digestion of nutrient-poor plant materials. However, herbivorous reptiles can serve as the dominant herbivore in a variety of environments, such as arid deserts, forest canopies, tropical rainforests, and both marine and freshwater aquatic systems, particularly on island systems (Nagy and Shoemaker, 1975; Moll and Jansen, 1995; King, 1996; Olesen and Valido, 2003; Fourqurean et al., 2010). In the southeastern US, herbivorous chelonians can be observed in both aquatic (e.g. , Pseudemys floridanus andChelonia mydas ) and terrestrial environments (e.g. ,Gopherus polyphemus and Terrapene carolina ) (Bonin et al., 2006; Fourqurean et al., 2010). However, we have a minimal understanding on the role of chelonian herbivores as regulators of plant community diversity and composition (Estill and Cruzan, 2001; Tuberville et al., 2005).
The gopher tortoise (Gopherus polyphemus ) is an herbivore found throughout the upland pine forests of the southeastern Coastal Plain in North America. Gopher tortoise herbivory and burrow excavation increase habitat heterogeneity by altering light exposure, temperature, soil-pH, soil-moisture content, and the distribution of nutrients in sandhill substrates (Diemer, 1986; Kaczor and Hartnett, 1990; Pike and Mitchell, 2013). Browsing gopher tortoises may remove dominant, palatable plant species and increase the relative abundance of rarer species (MacDonald and Mushinsky, 1988). Thus, gopher tortoises may impact plant community composition and primary production, however, there have only been a few attempts to document these impacts (Diemer, 1986; Breininger et al., 1991; McCoy et al., 2006).
Gopher tortoise herbivory and burrowing provide frequent, low-intensity bouts of disturbance due to low metabolic costs and intimate home range. (McRae et al., 1981, King, 1996). Studies that quantify the impacts of herbivory commonly deploy experiments that selectively exclude herbivores and/or control grazing intensity (Ritchie et al., 1998; Hester et. al., 2000; Habeck and Schultz, 2015). Exclusion experiments allow researchers to understand thresholds of grazing intensity, effects on primary productivity, herbivore-herbivore interactions, and impacts on biodiversity and whole-system dynamics (Noy-Meir, 1975; Mitchell and Kirby 1990; Milchunas and Lauenroth, 1993; Ritchie et al., 1998; Hester et. al., 2000; Habeck and Schultz, 2015; Fourqurean et al., 2010; Richardson and Stiling, 2019). In the case of the gopher tortoise, Richardson and Stiling, 2019, completed a two-year tortoise-exclusion on coastal Florida shrubland islands devoid of mammalian herbivores finding a reduction in both plant richness and species diversity. An exclusion experiment in upland longleaf pine forest that is significantly more rich in understory plant species alongside mammalian herbivores offers an ideal system to test how the removal of a low-intensity, frequent disturbance of a reptilian herbivore effects: 1) univariate measures of richness, evenness, and diversity, 2) aboveground plant community production, and 3) the overall effects on plant community composition. We hypothesize that gopher tortoise exclusion will increase aboveground plant production but will reduce plant richness, evenness, and diversity as fast-growing, dominant species quickly respond to the removal of herbivory. This will result in a significant shift in overall plant community composition between access and exclosure plots.