2011 SOUTHEASTERN NATURALIST 10(4):741–750 Restoration of Plant Communities in Former Pine Tree Plantations Anthony M. Rossi1,*, Ryan C. Meyer1, Keith Stokes2, and Daniel C. Moon1 Abstract - As human populations encroach on more remote areas, tree plantations are frequently converted into suburban developments or conservation lands. The purpose of the current study was to compare the effects of restoration treatments on the abundance and biodiversity of two former pine plantations in northeast Florida. Restoration methods, which varied greatly in both environmental impact and effort from minimal (non-manipulated control plots) to thinned (removal of 90–97% of Slash Pines and exotic plant species) to cleared (mechanical removal of all vegetation to the substrate surface), were compared using replicated 100-m2 plots. Although sites exhibited variation in plant densities and relative guild abundance, restoration treatment produced a significant increase in plant abundance and relative composition of plant guilds. Plant abundance was 2–3 times higher in cleared plots compared to thinned and control plots, respectively; not surprisingly this was because cleared plots had large numbers of small herbaceous seedlings as well as a shift in their plant communities to graminoid species, which are typically less shade tolerant than other plant guilds. Moreover, the relative abundance of vines was substantially lower in cleared plots compared to control and thinned plots, which retained all or some of their tree and shrub cover, thereby providing support structure for climbing plants. While plant diversity (Shannon index) exhibited a significant time-by-treatment interaction because cleared plots initially had only 65–80% of the plant diversity compared to control and thinned plots, cleared plots reached or exceeded them in biodiversity by the last two sampling dates, although cleared plots at one site showed a large decline on the final sampling date. In general, both study sites reflected high levels of graminoids in cleared plots, while thinned plots showed more even increases across plant guilds. Lastly, pine tree diameter (i.e., plant growth rate) was greater and canopy cover was lower in thinned plots compared to controls, although the differences were only significant for one site. Results from this study indicate that thinning former pine plantations with selective removal of exotics will provide maximal plant biodiversity and tree growth rather than utilizing more cost- and labor-intensive clearing of sites. Introduction From 2000 to 2006, the population of Florida increased an estimated 13.2% from 15.9 million to over 18 million people (US Census Bureau 2006). Growing human populations result in increased encroachment into natural ecosystems; to ameliorate lost natural habitat, tree plantations are often converted into conservation lands. Pinus elliottii Engelm. (Slash Pine) plantations are a good choice for restoration in northeast Florida because they are ubiquitous and semi-natural. The benefits of green space include reducing the “heat island” effect in which cities or metropolitan areas have significantly higher temperatures than surrounding rural 1 University of North Florida, Department of Biology, 1 UNF Drive, Jacksonville, FL 32224. 2 University of South Florida, Department of Integrative Biology, 4202 East Fowler Avenue, Tampa, fl33620. *Corresponding author- arossi@unf.edu. 742 Southeastern Naturalist Vol. 10, No. 4 or non-developed land, offering inexpensive storm water control, and lowering atmospheric CO2 (Beckett et al. 1997, Chang et al. 2006, Cunningham and Saigo 1999, San-Salazar and Rausell-Koster 2007). Urban forests also encourage physical activity and provide increased quality of life for citizens who utilize those areas (Chiesura 2003). Because local governments or non-profit organizations are frequently given the task of managing public lands, developing inexpensive guidelines for tree-farm restoration in north Florida may facilitate rejuvenation of native habitat more quickly in these urbanizing areas. Florida tree farms offer relative ease in restoration because the farmed tree species (generally Slash Pine or P. palustris P. Mill [Longleaf Pine]) are native, and herbaceous seed banks may still be intact, depending on prior land-use practices, including agriculture and silviculture, which may have included altered fire regimes, use of pesticides, herbicides, etc. (Andreu et al. 2009, Hedman et al. 2000). Terrestrial habitat restoration generally involves re-establishing a portion of the native plant community characteristic of the area prior to human disturbance (Miller and Hobbs 2007). The goals of restoration projects may be multifaceted; for instance, restoration projects may be utilized to increase ecosystem function and provide native vegetation for endemic animal communities. Plants are easy to manipulate relative to motile organisms, which may re-colonize quickly by natural dispersal. Increasing the plant diversity of an area is likely to result in a greater diversity of both primary consumers (i.e., herbivores) and secondary consumers (Michel et al. 2007, Moro and Gadal 2006). Thus, plant diversity will directly or indirectly influence all other trophic levels in an ecosystem (Chen et al. 2006, Panzer and Schwartz 1998, Provencher et al. 2003). Typically, the historic cycle of fires can be reinstated as a restoration tool in the southeastern US; however, fire is not always a realistic tool due to climate, forest conditions, and proximity to homes or commercial property. Mechanical clearing (e.g., row thinning) may provide an alternative restoration technique that offers some of the benefits of fire, such as opening the overstory canopy. In addition, thinning is a disturbance that reduces dominance by one or a few species and releases resources, reduces competition, and opens spaces which may promote the establishment of non-dominant or rare species (Levine and Paine 1974). In both natural and restored ecosystems, new patches are periodically created by disturbances, which are subsequently colonized by plants, both native and exotic, via dispersal mechanisms, seed banks, etc. (Wiens 1976). The purpose of this study was to compare the effects of restoration techniques (control, thinning, and clearing) on subsequent plant abundance and biodiversity in two former pine plantations in north Florida. Study Sites The current study utilized two partially restored pine plantation sites in Jacksonville, Duval County, FL, 30°19'56"N, 81°39'20"W. Both sites were purchased by the City of Jacksonville (COJ) and designated for preservation in perpetuity for local residents as part of the Preservation Project Jacksonville (PPJ) initiated 2011 A.M. Rossi, R.C. Meyer, D.C. Moon, and K. Stokes 743 in 1999. Both study sites are former pine plantations, heavily planted with Slash Pine and both sites lacked robust herbaceous growth due to high tree densities and leaf-litter accumulation. Betz/Tiger Point (aka: Tiger Point) is located northeast of the city center and is surrounded by salt marsh vegetation which drains into the Trout River and eventually the Atlantic Ocean. This site is part of the Pumpkin Hill Buffer State Park and the greater Timucuan Ecological and Historic Preserve. Although the ages of the remnants stands of Slash Pine at the two sites are unknown, tree dbh was approximately 30–50 cm. At Tiger Point, Slash Pine stands were very dense and averaged 40–60 trees per 100 m2. Sites also had a thick undergrowth of Serenoa repens (W. Bartram) Small (Saw Palmetto) and Ilex glabra (L.) A. Gray (Gallberry). Herbaceous growth included Galactia elliottii Nutt. (Elliott’s Milk Pea), the bunch grass Dicanthelium spp., and Andropogon virginicus var. glaucus (Chalky Bluestem Broomsedge), and the vine guild included several species of Smilax (greenbrier) and Gelsemium sempervirens (L.) (Aiton f.) (Yellow Jessamine). This site is on a small point oriented north into Edwards Creek, is surrounded for many kilometers to the southeast, south, and southwest by former tree farms and pinelands, and is surrounded to the west, north, and east by the waterway. This sandhill community, which includes large numbers of Slash Pines, Quercus laevis Walt. (Turkey Oak), Quercus falcata Michx. (Southern Red Oak), and shrubs such as Morella cerifera (L.) Small (Wax Myrtle) and Lyonia ferruginea (Walt.) Nutt. (Rusty Lyonia), tends to be very sunny and dry, with few public visitors. The second site, McGirt’s Creek, is located closer to the city center in the southwest portion of Jacksonville. The McGirt’s Creek site is part of the St. Johns River watershed and drains initially into the Ortega River and eventually to the Atlantic Ocean. This site has less of an herbaceous layer, slightly more leaf litter, and about 18–20 trees per 100 m2. This site is surrounded to the northeast, east, and southeast by a former tree farm and wetlands, while surrounded to the south, west, and north by suburban development and buffered by a relatively contiguous urban forest. This site also has seasonal standing water present and two drainage ditches approximately 2 m deep, that run 20 m along the entrance road. Some study plots were within 20 m of these ditches. Canopy vegetation at McGirt’s Creek is similar to Tiger Point, with Slash Pine dominating, but also includes Quercus nigra L. (Water Oak) and Prunus umbellata Elliot (Hog Plum). However, it lacks Saw Palmetto and Gallberry. Prominent species included Yellow Jessamine and Gaylussacia frondosa var. tomentosa (L.) A.Gray ex Torr. var. tomentosa (Huckleberry), and some plots contain large stands of Lachnanthes caroliana (Lam.) Dandy (Carolina Redroot), an obligate wetland species. It was expected that rapidly growing plants with highly dispersive seeds such as graminoids and forbs (annuals), which are characteristic of early successional stages, would colonize quickly in cleared and thinned plots compared to control plots. Eventually, slower growing species with larger and less dispersive seeds, including trees, shrubs, and woody vines, were expected to establish after the first growing season. 744 Southeastern Naturalist Vol. 10, No. 4 Materials and Methods Within each study area, large contiguous areas of existing Slash Pine stands were chosen so that the study plots were situated along one side of an existing logging road at each site. Square (10 m × 10 m) study plots were established parallel to the logging road approximately 5 m from the edge of the road. Plots were randomly assigned to one of three groups: thinned, cleared, or control. Fifteen (n = 15) study plots were established at the McGirt’s Creek site (n = 5 per treatment) and nine study plots (n = 9) at the Betz / Tiger Point site (n = 3 per treatment). Plots were 100 m2, which has been shown to be an effective size for study when using multiple plots to gather a representative sample unit (Peet et al. 1998). In thinned plots, two or three randomly selected living Slash Pine trees were left to achieve 20% canopy cover determined by densiometer (1–2 trees per plot); all the other Slash Pines were cut as close to the soil surface as possible using chain saws, and all non-native species were also removed mechanically. Cleared plots were razed to bare soil and raked, with shrub and tree roots left in the ground; no herbicides were applied, and debris from thinning and clearing was left several meters from each plot. Control plots received no treatment and were used as comparison for the treatment groups. Corners of the plots were marked using wooden stakes and two permanent, parallel transects randomly selected to occur at 2.5, 5.0, or 7.5 m were established in each plot; no two transects overlapped. To determine how thinning and clearing treatments affect herbaceous plant diversity, a 1-m-wide belt transect was used to assess plant diversity twice per year; surveys were conducted in the growing season and during the dormant season from 2006 to 2008. Along each transect, all plants were counted and identified to species when possible. During the dormant season, herbaceous plants were not counted if they lacked photosynthesizing tissues or green leaves. Tree characteristics, including canopy cover and diameter at breast height (dbh, cm), were also measured. Canopy readings were taken at a height of approximately 1.5 m in the center point of each plot facing each cardinal direction. Tree diameter was assessed by dbh (cm) during 2006 and again in 2008. Only thinned and control plots were assessed in this manner because cleared plots had no canopy cover during the period of the study. Plant species richness (s) and diversity were assessed using a two-way ANOVA with treatment nested within site (SPSS ver. 11.0). Shannon diversity index values (H') were also calculated for each site. Shannon index, which provides a measure of diversity based on the proportional abundance of each species present in the sample, was calculated as H' = -Σ I = 1 (pi )× (lnpi ), where pi = proportion of the total species the ith species occupies (= ni / N ). H' is a good index for diversity because it is sensitive to rare or uncommon species and therefore may be more useful in conservation or restoration projects. The change in relative abundance in guilds from 2006 to 2008 was assessed using a two-way ANOVA with treatment and survey date as fixed effects. A t-test was used to assess changes in tree diameter (dbh) between thinned and control plots only from 2006 to 2008. Lastly, changes in tree canopy were examined from November 2006 to November 2008, and ANOVA was used to test for significant differences among treatment groups. Data were checked using an F-test to ensure that the met the homogeneity of 2011 A.M. Rossi, R.C. Meyer, D.C. Moon, and K. Stokes 745 variance assumptions of ANOVA prior to analysis, and Tukey’s post-hoc tests were used to compare treatment means when appropriate. Results At both sites, mean number of individual plants per transect was higher in cleared plots than in either thinned or control plots in 2008 (McGirt’s: F2,48 = 3.20, P = 0.045; Tiger Point: F2,24 = 3.38, P = 0.057; Fig. 1). This increase was largely represented by small seedling individuals. Shannon diversity index values were not significantly affected by treatment at Tiger Point (F2,48 = 2.02, P = 0.189; Fig. 2), but differed significantly at McGirts’ Creek (F2,48 = 8.16, P = 0.001), as cleared plots had lower diversity than thinned or control plots in all surveys except winter 2007 (Fig. 3). At McGirt’s Creek, the decrease in diversity in control and thinned plots in winter 2007 resulted in a significant effect of survey (F3,48 = 9.34, P < 0.001) as well as a marginal interaction between treatment and survey (F6,48 = 2.07, P = 0.081). Neither survey date nor the treatment × survey interaction was significant at Tiger Point (P > 0.2; Fig 2). Graminoids and vines were the most abundant functional groups at both sites, accounting for more than 50% of all plants in all surveys at this site, but treatments did not result in significant changes in the relative abundances of functional groups (P > 0.2 in all cases). Trees were dominated by Slash Pine and Diospyros virginiana L. (Persimmon) at Tiger Point while Water Oak and Acer rubrum L. (Red Maple) dominated at Figure 1. Mean species abundance from 2006–2008. Error bars indicate ±1 SEM. 746 Southeastern Naturalist Vol. 10, No. 4 McGirts Creek. Slash Pine diameter (dbh) changed significantly during this study (Fig. 4). At McGirt’s Creek, dbh increased by approximately the same amount in control plots and thinned plots (P > 0.2). At Tiger Point, however, dbh of trees in thinned plots increased significantly more than did dbh of trees in control plots (t = 2.578, df = 7, P = 0.033). Trees in thinned plots increased in diameter by more than 4 cm, while those in control plots increased by only 2 cm (Fig. 4). Discussion Plant community composition and plant abundance were significantly altered by clearing and thinning treatments, which is consistent with other studies (Dodson et al. 2007, Laughlin et al. 2006, Levin and Paine 1974). This increase was largely caused by a greater density of seedlings (especially graminoids) in cleared plots, and is likely a result of higher levels of light and resources available in cleared plots. Thus, removing the dominant vegetation probably facilitated germination of seeds, thereby encouraging seedling recruitment and stimulated re-sprouting from underground vegetative sources. On the other hand, thinned and control plots tended to have a significantly greater abundance of vines such as greenbrier and Galactia (milkpea). This result is consistent with the presence of support plants for vines such as trees and shrubs in control and thinned plots compared to cleared plots from which these support plants were removed. Although Figure 2. Tiger Point Shannon diversity index. Error bars indicate ±1 SEM. 2011 A.M. Rossi, R.C. Meyer, D.C. Moon, and K. Stokes 747 the two study sites showed generally similar trends, the impact of restoration treatment varied substantially between sites; cleared plots at Tiger Point rapidly increased in plant diversity and were similar to thinned and control plots within six months, while cleared plots at McGirt’s Creek remained substantially lower in plant diversity than those treatments on three of four survey dates. The difference in response between the two sites was possibly related to the differences in the ages and density of tree stands at the two sites. As previously mentioned, Tiger Point was planted with Slash Pines much more recently (based upon tree density and tree dbh) than McGirt’s Creek; as a result, removal of trees from the latter site returned it to a much earlier successional stage compared to Tiger Point. Thus, the response was much less dramatic at Tiger Point. Changes in the plant community will have direct effects on local primary consumers including birds, rodents, and insects (Waltz and Covington 2004). For instance, in cleared and thinned plots, the increase is annual plants should favor some animals such as songbirds that specialize on seeds of early successional annual plants (Triquet et al. 1990). Conversely, species such as Meleagris gallopavo L. (Wild Turkey) or Coturnix coturnix (L.) (Common Quail) that require larger fruits are expected to benefit from thinned plots with high levels of fruit-producing shrubs. Interestingly, herbaceous plant species richness was not affected by clearing plots, although plant abundance (primarily in the form of seedlings) increased Figure 3. McGirt’s Creek Shannon diversity index. Error bars indicate ±1 SEM. 748 Southeastern Naturalist Vol. 10, No. 4 in cleared plots at both sites compared to thinned and control plots. This result suggests an intact seed bank and/or a source for recruitment of appropriate native species nearby. Additionally, only a handful of non-native species were recorded at either of our study sites, suggesting that these treatments may be effective in preventing colonization by exotics in restoration of former pine plantations in northeast Florida. Natural resource management techniques suggest that replanting is not necessary at these sites, thereby lowering the cost of restoration. Thinned plots generally displayed common species between control and cleared plots. Thinned plots frequently experienced the benefits of clearing vegetation (increased abundance of pioneer plants), while maintaining relatively high abundances of later successional plant groups as well. Trees responded with robust growth when the habitat was opened; Slash Pines in the thinned treatment grew significantly faster than those in control plots, as expected (Baldwin et al. 2000). Surprisingly, the relative abundance of tree species at Tiger Point shifted from Slash Pine to Persimmon in a relatively short time after thinning. At McGirt’s Creek, cleared plots showed highest abundance for Water Oak, another non-intuitive trend which suggests that an additional restoration step may be necessary. For instance, if hardwood reduction is a primary management goal, then initial restoration treatments may require follow-up monitoring or action such as herbicide, fire, or mechanical removal. Figure 4. Mean change in tree diameter (dbh, cm) at two study sites. Error bars indicate ±1 SEM. 2011 A.M. Rossi, R.C. Meyer, D.C. Moon, and K. Stokes 749 Other species revealed high levels of reproductive potential and may be considered for use in post-planting. Abundance of species such as Solidago spp. (goldenrod) exhibited large increases in the short term. Ferns were another important pioneer and most likely emerged from underground sources following treatment application. One species that was ubiquitous at both sites was Dichanthelium dichotomum (L.) (Panicgrass), which also proved a fast colonizer, along with Chalky Blue Broomsedge at Tiger Point and the mesic Carolina Redroot at McGirt’s Creek. At both sites, shrubs in high abundance included the ever-present Gallberry along with Hypericum spp. (St. John’s wort) at Tiger Point and Vaccinium spp. (blueberries) at McGirt’s Creek. The combined abundance of vines and graminoids represented more than 50% of all individuals on several sampling dates at Tiger Point, and their importance across treatment regimes should not be underestimated. Species richness within plots is generally correlated with the local species pool (Dickson and Foster 2008). It has therefore been suggested that the most common pioneer species may be suitable for use in regional restoration of plant communities (Lane and Texler 2009), thereby facilitating species selection. Prior to establishing a restoration plan, former land practices (e.g., agriculture, sylviculture, etc.) of the area should be considered because cropland likely has less potential for successful restoration than pine plantations (Hedman et al. 2000). Results from the current study suggest that thinning of former pine plantations in northeast Florida may provide an important restoration and management tool near urban areas. This is because thinning allows goals of restoration projects to be realized (e.g., increased biodiversity of plants and animals endemic to an area) without the risk associated with prescribed fires, which may be problematic near populated areas. Thus, restoration of these former plantations should provide multiple benefi ts to urban areas, including ecosystem functions such as water purification and absorption of water runoff as well as increased recreational activities. Acknowledgments We would like to thank the University of North Florida (UNF), which provided support for this research through a Graduate Scholar’s Grant to R.C. Meyer, as well as the UNF Coastal Biology Program for financial support. This project was also funded by a grant from the City of Jacksonville and the Florida Fish and Wildlife Conservation Commission to A.M. Rossi and D.C. Moon. We also thank two anonymous reviewers for their insightful comments that improved the manuscript. Literature Cited Andreu, M.G., C.W. Hedman, M.H. Friedman, and A.G. Andreu. 2009. Can managers bank on seed banks when restoring Pinus taeda L. in southwest Georgia? Restoration Ecology 17:586–596. Baldwin, V.C., Jr., K.D. Peterson, A. Clark III, R.B. Perguson, M.R. Strub, and D.R. Bower. 2000. The effects of spacing and thinning on stand and tree characteristics of 38-year-old Loblolly Pine. Forest Ecology and Management 137:91–102. Beckett, K.P., P.H. Freer-Smith, and G. Taylor. 1997. Urban woodlands: Their role in reducing the effects of particulate pollution. Environmental Pollution 99:347–360. 750 Southeastern Naturalist Vol. 10, No. 4 Chang, C.R., M.H. Li, and S.D. Chang. 2006. A preliminary study on the local coolisland intensity of Taipei city parks. Landscape and Urban Planning 80:386–395. Chen, Z., K. Grady, S. Stephens, J. Villa-Castillo, and M.R. Wagner. 2006. Fuel reduction treatment and wildfire influence on carabid and tenebrionid community assemblages in the Ponderosa Pine forest of Northern Arizona, USA. Forest Ecology and Management 225:168–177. Chiesura, A. 2003. The role of urban parks for the sustainable city. Landscape and Urban Planning 68(1):129–138. Cunningham, W.P., and B.W. Saigo. 1999. Environmental Science: A Global Concern, 5th Edition. McGraw-Hill, Boston, MA. 650 pp. Dickson, T.L., and B.L. Foster. 2008. The relative importance of the species pool, productivity, and disturbance in regulating grassland plant species richness: A field experiment. Journal of Ecology 96(5):937–946. Dodson, E.K., D.W. Peterson, and R.J. Harrod. 2007. Understory vegetation response to thinning and burning restoration treatments in dry conifer forests of the eastern cascades, USA. Forest Ecology and Management 255:3130–3140. Hedman, C.W., S.L. Grace, and S.E. King. 2000. Vegetation composition and structure of southern coastal plain pine forests: An ecological comparison. Forest Ecology and Management 134:233–247. Lane, C.P., and H.D. Texler. 2009. Generating quantitative regional plant community descriptions for restoration. Restoration Ecology 17(1):42–50. Laughlin, D.C., M.M. Moore, J.D. Bakker, C.A. Casey, J.D. Springer, P.Z. Fule, and W.W. Convington. 2006. Assessing targets for the restoration of herbaceous vegetation in Ponderosa Pine forests. Restoration Ecology 14(4):548–560. Levin, S.A., and R.T. Paine. 1974. Disturbance, patch formation, and community structure. Proceedings of the National Academy of Sciences 71(7):2744–2747. Michel, N., F. Burel, and P. Legendre. 2007. Role of habitat and landscape in structuring small-mammal assemblages in hedgerow networks of contrasted farming landscapes in Brittany, France. Landscape Ecology 22:1241–1253. Miller, J.R., and R.J. Hobbs. 2007. Habitat restoration: Do we know what we’re doing? Restoration Ecology 15(3):382–390. Moro, D., and S. Gadal. 2006. Benefits of habitat restoration to small-mammal diversity and abundance in a pastoral agricultural landscape in mid-Wales. Biodiversity and Conservation 12:3543–3557. Panzer, R., and M.W. Schwartz. 1998. Effectiveness of a vegetation-based approach to insect conservation. Conservation Biology 12(3):693–702. Peet, R.K., T.R. Wentworth, and P.S. White. 1998. A flexible, multipurpose method for recording vegetation composition and structure. Castanea 63(3):262–274. Provencher, L., A.R. Litt, and D.R. Gordon. 2003. Predictors of species richness in northwest Florida Longleaf Pine Sandhills. Conservation Biology 17(6):1660–1671. San-Salazar, S.del, and P. Rausell-Koster. 2007. A double hurdle model of urban green areas valuation: Dealing with zero responses. Landscape and Urban Planning 84:241–251. Triquet, A.M., G.A. McPeek, and W.C. McComb. 1990. Songbird diversity in clearcuts with and without a riparian buffer strip. Journal of Soil and Water Conservation 45:500–503. United States Census Bureau. 2006. State and county quickfacts. Available online at http://quickfacts.census.gov/qfd/states/12000.html. Accessed july 2010. Waltz, A.E.M., and W.W. Covington. 2004. Ecological restoration treatments increase butterfly richness and abundance: Mechanisms of response. Restoration Ecology 12(1):85–96. Wiens, J.A. 1976. Population responses to patchy environments. Annual Review of Ecological Systematics 7:81–120.