DISCUSSION
Following a single growing season of gopher tortoise exclusion, we found a 35% increase in total plant cover, 45% increase in plant height, 15% reduction in plant species richness and 12% reduction in plant diversity. Even short-term gopher tortoise exclusion can have significant measurable impacts on the understory plant communities of upland pine forests. Our results also suggest that gopher tortoise exclusion led to an increase in small stature woody species (Geobalanus oblongifolius and Toxicodendron radicans ), grasses (Sorghastrum secundum ), and ferns (Pteridium pseudocaudatum ), many of which are known dietary components of the gopher tortoise (MacDonald and Mushinsky, 1988; Mushinsky, Stilson and McCoy, 2003). Our results suggest that the persistence of gopher tortoises may be useful for maintaining the heterogeneous and biodiverse understory plant assemblages of coastal upland pine forests alongside fire (Kaczor and Hartnett, 1990; Breininger et al., 1991; McCoy et al., 2006).
The effects of tortoise exclusion on plant richness and diversity over six months are comparable to longer-term studies in scrub-shrub habitats (Richardson and Stiling, 2019). After two years of gopher tortoise exclusion in a coastal Florida shrubland, Richardson and Stilling (2019) found a 37% reduction in plant richness (9 plant species total) and a 38% reduction in Gini-Simpson diversity. While our 15% reduction in richness and 12% in plant diversity were about half the effect size found by Richardson & Stilling, we suspect that future studies will reveal that the effects of tortoise exclusion will increase with time. Significant impacts of herbivore exclusion experiments after one year of exclusion are rare in the literature, and significant effects build up after multiple years (Moorhead et al. 2017, Shelton et al. 2014, Frank, Rathfon and Saunders, 2018; Schäfer et al., 2019; Bloodworth, Ritchie and Komatsu, 2020). In longleaf pine forests in southern Alabama, white-tailed deer exclusion had no effect on plant diversity and cover after four years of exclusion (Brockway and Lewis, 2003). As a potential working hypothesis to explain the differences between gopher tortoise and mammal exclusion, most mammalian herbivores have significantly larger home ranges relative to gopher tortoises. For example, white-tailed deer have home ranges up to 1000 ha and eastern cottontail rabbit have home ranges of 4–8 ha, while gopher tortoise home ranges are typically limited to 0.004–3.2 ha (Trent and Rongstad, 1974; Gálvez-Bravo et al., 2011; Bowers, 1993; Guyer at al., 2012). The strong effects of gopher tortoise exclusion may be driven by the more concentrated impacts of gopher tortoise herbivory relative to the diffuse impacts of grazing mammals (McRae et al., 1981, Bowers, 1993).
We found that plot spatial location explained 48% of the variation in plant community composition, which was substantially more when compared tortoise exclusion, which was ~5%. Plant species richness ranged between 7-24 individual species and percent cover varied between 0.5% - 55% coverage yielding high amounts of heterogeneity among our plots. We found that 24% of observed plant species were found in one or two plots and only 13 species were found in more than half of our plots. High heterogeneity is commonly reported within understory communities of upland longleaf pine forests (Drew, Kirkman and Gholson, 1998; Kirkman et al., 2001; Mitchell et al., 2006; Blaustein, 2008; Agrawal, Lau and Hambäck, 2006).
While gopher tortoises directly impact plant assemblages by defoliation, they also influence plants by burrow excavation, urine and feces deposition, and routine trampling of vegetation during foraging and frequent burrow-to-burrow migrations (Diemer, 1986; Kaczor and Hartnett, 1990; Pike and Mitchell, 2013). The small size of our exclosures (1 m2) likely fail to account for the full impacts of gopher tortoises on understory plant communities. To assess for impacts other than direct ingestion of plant material, larger exclosures might be better equipped coupled with experimental manipulation to isolate trampling effects, from nutrient deposition and burrow construction to quantify impacts by these alternative disturbances. In other herbivore species that live in social networks and burrow systems, large areas are used in quantifying impacts to the surrounding plant communities and not necessarily herbivore-exclusion. For example, European rabbits and their warren (burrow) systems were used to show that rabbit activity promoted vegetation productivity and diversity by increase in soil disturbance and resources provided by latrines in (15.5 m2) open influence areas (Gálvez-Bravo et al., 2011). Similarly, gopher tortoises have a life history that relies on complex interacting social networks and frequent movements amount multiple burrows which would need a larger area to assess to understand the full impacts of how these reptiles regulate the forest floor understory (McRae et al., 1981; Boglioli, Guyer, and Michener, 2003; Guyer, Johnson and Hermann, 2012).
We found that exclusion of gopher tortoise herbivory had effects on plant diversity and productivity that are similar to large and small mammalian herbivores. While reptilian herbivores are relatively rare in most ecosystems and have been traditionally under-represented in the herbivory literature, some studies have shown reptilian herbivory to impact arid deserts (Nagy and Shoemaker, 1975), tropical rainforests (Moll and Jansen, 1995), and coastal upland forest (King, 1996; Olesen and Valido, 2003). Future studies should investigate the role of gopher tortoises in maintaining biodiversity of longleaf pine understories in comparison to other commonly used management strategies, like prescribed burning. (Pacala and Crawley, 1992). Gopher tortoises may work in conjunction with fire to promote and maintain biodiversity in the understory of longleaf pine ecosystems. While arguments for gopher tortoises as a keystone species have largely focused on the animal assemblages that access gopher tortoise burrows (Guyer and Bailey 1993, Catano and Stout 2015), our results suggest that gopher tortoises have a role as regulators of plant diversity and productivity through herbivory as well. Gopher tortoises may play a key role in maintaining the productivity and diversity of longleaf pine understory communities; however, human development and habitat loss are having negative effects on gopher tortoise populations throughout the species range. Thus, our results highlight the importance of continued protections and conservation efforts for gopher tortoises in the southeastern United States, because tortoises play an important role in regulating diversity of plant communities.