2007 SOUTHEASTERN NATURALIST 6(3):523–534 Influences on Oak and Pine Establishment with Time Since Fire in Sandhills Pinus paulustris (Longleaf Pine) Forests Jaclyn L. Schnurr1,* and Beverly S. Collins2 Abstract - We investigated biotic and abiotic mechanisms that affect Quercus falcata (southern red oak) establishment from seeds in southeastern Pinus paulustris (longleaf pine) forests along a time-since-fire chronosequence (unburned since 1950, the mid-1990s, or winter 2001) and slope gradient. To determine if seed availability limits southern red oak recruitment, we measured seed production, populations of small mammals, and seed removal across sites. To determine if seedling survival limits establishment, we outplanted greenhouse-raised southern red oak seedlings and followed their survival throughout one growing season. We also measured abiotic conditions such as canopy openness, soil moisture, and soil temperature, and censused longleaf pine recruitment. Contrary to expectations based on species’ fire-tolerances, longleaf pine recruitment was consistent over the chronosequence, but there was little oak seedling establishment in the long-unburned sites. Sites last burned in the mid-1990s had the lowest oak seed rain, highest small-mammal populations, and highest seed -removal rates; yet, they had the highest oak seedling establishment. Seedling survival was highest in the pre- 1950 sites, and was most influenced by intermediate light levels and high soil moisture. We conclude that biotic factors, specifically, seed removal by small mammals, are less important to seedling recruitment than abiotic factors in these forests. Further, the assumption that longleaf pine forests will eventually undergo to succession to hardwoods in the absence of fire is complicated by abiotic conditions associated with landscape position. Introduction In the Piedmont-Coastal Plain fall-line region of the southeastern US, upland Pinus palustris P. Mill. (longleaf pine) forests grade from sandhills longleaf-scrub oak communities on ridgetops through mixed pine-hardwood forests on slopes (Christensen 1988, Collins et al. 2006, Jacqmain et al. 1999, Skeen et al. 1993, Ware et al. 1993). Upland oak species that codominate with longleaf pine in these forests are distributed from sandhills species, such as Quercus margaretta Ashe (sand post oak) and Quercus laevis Walt. (turkey oak), which regenerate primarily by sprouting, through more mesic species such as Quercus falcata Michx. (southern red oak), the focus of this study, Quercus nigra L. (water oak), and Quercus alba L. (white oak), which tend to regenerate from seeds (Collins et al. 2006). Vegetation composition and regeneration of canopy species in upland longleaf forests has been shown to be related to both topography and fire regime. Decreased frequency of natural or prescribed fires has been linked to 1Wells College, Aurora, NY 13026. 2Western Carolina University, Cullowhee, NC 28723. *Corresponding author - jschnurr@wells.edu. 524 Southeastern Naturalist Vol. 6, No. 3 increased recruitment of hardwoods, including oaks (Brockway and Lewis 1997; Hartnett and Krofta 1989; Heyward 1939; Jones et al. 1981, 1984; Monk 1965, 1968; Quarterman and Keever 1962; Shelton and Cain 1999; Walters 1991). Lower light levels in infrequently burned forests can favor regeneration of more shade tolerant hardwoods over less shade tolerant longleaf pine (Gardiner and Hodges 1998, Larsen and Johnson 1998, Li and Ma 2003, Phares 1971), and the lack of fire will allow the hardwood seedlings to persist. Although birds and small mammals can have large effects on tree seedling recruitment (Elkinton et al. 1996; Hulme and Hunt 1999; Janzen 1970; Jensen 1985; LoGiudice and Ostfeld 2002; McCracken et al.1999; Mosandl and Kleinert 1998; Nilsson 1985; Ostfeld et al. 1996, 1997; Schnurr et al. 2002, 2004; Schupp 1988; Sork et al. 1993; Whelan et al. 1991; Wolff 1996), the effects of animals on regeneration have been relatively unstudied in longleaf pine forests of the fall-line sandhills region. Seed production and input are the first biotic filters (Clark et al. 1998, Houle 1998, Schnurr et al. 2004, Shibata and Nakashizuka 1995) to regeneration, but even if a seed arrives at a site, it must escape seed predators (Hulme and Hunt 1999; LoGiudice and Ostfeld 2002; Ostfeld et al. 1997; Schnurr et al. 2002, 2004; Whelan et al. 1991). Oaks, for example, are subject to small-mammal and insect effects on seed viability, dispersal, and longevity, and these influences may interact with environmental drivers and/or forest management practices, including time since fire, to determine the proportion of oak species in the regeneration pool and, ultimately, the canopy of sandhills longleaf pine forests. The objective of our research was to determine the relative effects of seed predation and abiotic conditions (canopy openness, soil moisture, soil temperature) on regeneration of oaks, with a focus on southern red oak, in upland longleaf pine stands that differ in forest management and fire history (time since fire). These study species are dominant oak species in upland pine and pine-hardwood forests on slopes in the fall-line sandhills region (e.g., Collins et al. 2006, Dilustro et al. 2002, Workman and McLeod 1990). They grade upslope into longleaf pine-scrub oak communities and downslope into upland hardwoods (Collins et al. 2006, Workman and McLeod 1990). To compare regeneration over the slope gradient, we included longleaf pine-scrub oak and upland hardwoods stands. We asked, does seed production limit recruitment of oaks, either directly or through an increase in small-mammal seed predators? We predicted sites with greater acorn production would have greater densities of seed predators, specifically small mammals, and that these mammals would cause increased seed predation and decreased seedling recruitment. We used southern red oak to examine seed removal and seedling survival over the topographic gradient. We also predicted that, in the absence of biotic effects on seed and seedling survival, oak recruitment would increase down the topographic gradient from sandhills to mesic hardwood 2007 J.L. Schnurr and B.S. Collins 525 forest, and with time since fire in the upland pine stands. We sampled longleaf pine regeneration to determine if patterns of oak and pine regeneration coincide. Methods Site description and sampling The Savannah River Site (SRS) is a 78,000-ha facility of the United States Department of Energy, located on the upper coastal plain physiographic province along the South Carolina/Georgia border. When the SRS was acquired in 1950, the majority of land was forested (67%), while the rest was cropland (33%), although most of the forests had been logged (Workman and McLeod 1990). Starting in 1952, the US Forest Service (USFS) planted pine (primarily Pinus taeda L. (loblolly pine), but also P. palustris P. Mill. (longleaf pine), P. elliotii Engelm. (slash pine), and some crosses among these) seedlings on abandoned sites throughout the SRS. Today, forest communities grade from swamp forests and bottomland hardwoods at lower elevations, through sloped upland hardwood forests, to upland managed pine forests, to more xeric fire-adapted sandhill longleaf-scrub oak communities at 80–130 m elevation (Workman and McLeod 1990). Pine forests are currently managed using even- aged techniques, such as prescribed fires and thinning (Kilgo and Blake 2005). In spring of 2001, we established ten 1-ha sites over a topographic gradient from upland hardwoods (HDW, N = 2) upslope through upland longleaf pine (N = 6) to sandhills longleaf pine-scrub oak stands (SND, N = 2). Composition and structure of the selected stands were representative to those along topographic gradients on the SRS generally (Workman and McLeod 1990) and to previously sampled stands in other sites along the Piedmont-Coastal Plain fall-line region (e.g., Collins et al. 2006). The upland longleaf pine forests differed in fire management and time since fire; they were unburned since 1950 (B50, N = 2), the mid-1990s (B90, N = 2), or winter 2001 (B01, N = 2). Our sampling areas were located within a large homogeneous patch of each stand, with at least 30 m separating them from a different habitat type; however, there was a potential for seeds and animals to disperse into the sites from the surrounding variable habitat matrix. Due to the labor-intensive nature of our sampling regime, we were only able to sample two 1-hectare sites in each habitat. To characterize the sites, understory vegetation was surveyed along a 5-m line-intercept transect at 20 locations within each site in summer of 2001. We included all herbaceous and woody vegetation less than 2 m tall, and individuals were identified to species when possible. Species richness and the Shannon-Weiner diversity index were calculated for the herbaceous layer in each site. We determined percent cover for each species in each site, and then averaged the values for the 2 sites per treatment. 526 Southeastern Naturalist Vol. 6, No. 3 Seed inputs Seed input into the sites was measured using 8 litter traps per site (0.37 m x 0.5125 m x 0.70 m high) that were emptied monthly from August 2001 through August 2002. Although the seed traps were accessible to small mammals, we found no seed fragments or small-mammal feces that usually indicate that seeds had been removed. All collected seeds were sorted by species, and live seeds (those with green embryos) were counted to estimate total seed fall. Since seed fall of oak species was low, all Quercus sp. acorns were summed for the analyses. To determine the relationship between seed inputs and small-mammal populations, trapping was conducted using large, folding Sherman live traps located at 12.5-m intervals throughout each hectare. We trapped each grid for 3 days in August and November 2001, and February and May 2002. After a 3-day prebaiting period, each trap was set using black-oil sunflower seeds and was checked each morning at 0700 h and each evening at 1700 h. Captured small mammals were identified to species, sexed, weighed, eartagged, and released at the point of capture. Due to extremely low capture rates, traps remained open for the entire 3-day trapping period. Southern red oak seed and seedling survival We quantified seed removal rates experimentally by placing southern red oak acorns directly on the ground at 15 locations randomly chosen throughout each site in December 2001. Acorns were collected in November 2001 from sites throughout the SRS, and only acorns that had no indication of insect infestation were used. We used 2 acorns per location: locations were marked using fiberglass flags and the acorns were located 50 cm on either side of the flag and marked with a stick found in the area. Acorn removal was monitored daily for 3 days, then monthly for a total of 60 days of exposure. Gloves were worn when handling acorns and twigs to decrease olfactory cues to potential seed predators. To assess seedling survival, the southern red oak acorns collected in fall 2001 were stored at 5 °C in a coldroom until early February 2002, when seeds were germinated in the greenhouse. Successful germinants were planted in homogenized soil collected from all sites. Seedlings were kept in the greenhouse until danger of frost passed, and planted in the field in April 2002. Five seedlings were planted in a 1 x 1 m quadrat at six locations randomly chosen throughout each site, for a total of 30 seedlings per site. To assess the environment, soil gravimetric water content was measured at each seedling transplant quadrat in April, May, July, and October 2002. Canopy photos were taken 1 m above each quadrat in July 2002 and analyzed using GLA software (Frazer and Canham 1999). Soil temperatures (± 1 °C) were recorded at 2 locations within each site using iButtons® (Dallas Semiconductor Thermochron, Sunnyvale, CA) buried at 10 cm. Data collection began in October 2001 and continued until July 2002. Soil temperature was logged every 1.5 hours, but data were averaged monthly. Seedling survival was monitored until October 2002. 2007 J.L. Schnurr and B.S. Collins 527 Data analyses Data were analyzed using SYSTAT v. 10 and SAS v. 9. Analyses of variance (􀁟 = 0.05) were used to determine if mean seedling cover, seed inputs, or seed removal differed among sites, and Bonferroni-adjusted t-tests were used to examine pairwise differences between sites. Logistic analysis (SAS, Proc Catmod) was used to examine the relative effect of site on southern red oak seedling survival, i.e., to test the hypothesis that survival in a site did not differ from mean survival averaged over all sites. Results The sites that were most recently burned in the winter of 2001 (B01) had the highest species richness of the herbaceous layer, while the sites burned in the mid-1990s (B90) had the highest Shannon-Weiner diversity indices (Table 1). The dominant species in the herbaceous layer varied among sites, and usually made up about 30% of the total species composition (Table 1). The dominant species among the burned sites varied from Andropogon spp. (bluestem grass) in the most recently burned sites through the lichen Cladonia leporina in the sites burned pre-1950 (Table 1). The sites most recently burned had the largest portion of the ground covered by plants, with the B90 sites having 55% coverage and B01 having 52%. The hardwood sites (HDW) (14%) and sites burned pre-1950 (B50) (19%) had the least ground covered, although the B50 sites also had a thick duff layer, and the sandhills (SND) sites had an intermediate level (40%). Pooled over species, oak seedling percent cover was greatest in the B90 sites compared with all other sites (F = 8.856, P < 0.001; Fig. 1A). Longleaf pine seedling cover did not differ significantly among sites, but the B50 sites had the greatest longleaf recruitment, while the HWD sites had none (F = 1.818, P = 0.16; Fig. 1A). Maximum seed rain occurred in November 2001, when 29 of the 60 total seeds were collected. Seed rain did not correspond with the seedling cover patterns among sites (􀁲2 = 0.342 df = 1, P = 0.559; Fig. 1B). Oak seed rain did not differ significantly among sites and was lacking or low in all sites except HDW sites. B90 sites, which had the greatest oak seedling cover, had no oak seed rain (Fig.1B). Longleaf pine seed rain did not significantly vary among sites (Fig.1B). Table 1. Percent cover of the most abundant herbaceous-layer species, total species richness, and Shannon-Weiner diversity index value for each stand type. Total Stand herbaceous Species Shannontype Most abundant species % cover % cover richness Weiner H' SND Vaccinium stamineum L., deerberry 31 40 30 2.19 B01 Andropogon spp., bluestem 37 52 31 2.07 B90 Gelsemium sempervirens (L.) St. Hil., 36 55 28 2.35 evening trumpet flower B50 Cladonia leporina, lichen 32 19 22 2.14 HDW Vitis rotundifolia Michx., muscadine grape 30 14 13 2.04 528 Southeastern Naturalist Vol. 6, No. 3 Contrary to our prediction, small-mammal capture patterns did not correspond to seed rain (Fig. 1C). Total number of animals captured were: SND = 0, B01 = 10, B90 = 23, B50 = 6, and HDW = 6. The majority of small mammals 2007 J.L. Schnurr and B.S. Collins 529 were captured at the B90 sites (Fig. 1C), which had no oak seed rain. There were no animals captured in the SND sites throughout the study (Fig. 1C). In the other sites, Peromyscus gossypinus LeConte (cotton mouse) was the most abundant animal captured (89%), followed by Ochrotomys nuttalli Harlan (golden mouse; 8%) and Reithrodontomys humilis Audobon and Bachman (eastern harvest mouse; 2%). Removal of southern red oak acorns differed among sites (F = 5.814, P = 0.04; Fig. 1D), but pair-wise comparisons were not significant. B90 sites, which had the highest acorn removal, also had the greatest number of small mammals captured, but not the most seed rain, and there were no acorns produced in these sites. Although some acorns were removed in the SND sites, no small mammals were captured there. Survival of southern red oak seedlings differed among the sites (􀁲2 = 32.96, P < 0.001; Table 2). Survival was greatest in the sites last burned in the 1950s (B50; 􀁲2 = 22.75, P < 0.001), and lowest in sites last burned in 2001 (B01; 􀁲2 = 22.75, P < 0.001). Seedling survival was most variable within the HDW sites (Table 2). Seedling survival patterns reflected environmental differences among sites (Table 2). Specifically, soil moisture (April, May, July, and October) was significantly different among sites, with B50 and HDW sites having consistently higher soil moisture than the other sites (April: F = 5.986, P < 0.001; May: F = 2.82, P = 0.034; July: F = 18.298, P < 0.001; Oct.: F = 2.559, P less than 0.049; Table 2). Canopy openness also was significantly different among sites Table 2. Mean Quercus falcata (southern red oak) seedling survival, canopy openness (%), soil moisture (April, May, July, and October; %) and soil temperature (April–July, °C) in southeastern forests spanning a topographic gradient (mesic hardwoods [HDW] through upland pine forests [B] and dry sandhills [SND]) and time-since-fire (last burned prior to 1950 [B50], in the mid-1990s [B90], or in 2001 [B01]). Values are means ± standard error, N = 2. Seedling Canopy Soil Soil survival openness moisture temperature Stand type (% ± SE) (% ± SE) (% ± SE) (°C ± SE) SND 48 ± 8.7 30.9 ± 2.40 2.1 ± 0.14 22.9 ± 2.5 B01 20 ± 9.5* 28.2 ± 0.67 1.6 ± 0.17 23.4 ± 2.6 B90 65 ± 7.0 18.0 ± 10.80 2.3 ± 0.17 21.5 ± 2.4 B50 75 ± 9.5* 11.4 ± 0.66 3.6 ± 0.25 21.2 ± 2.3 HDW 50 ± 10.5 9.8 ± 0.16 2.9 ± 0.15 21.0 ± 2.2 * = significantly higher or lower survival. Figure 1 (opposite page). A) Average cover of oaks (Quercus falcata, Q. laevis, and Q. nigra) and longleaf pine (Pinus palustris) in m per each 5-m transect in five upland forests. B) Total number of viable acorns (Q. falcata, Q. laevis, Q. nigra) and Pinus palustris seeds collected in seed traps from September 2001–January 2002. C) Average number of small-mammal seed predators captured from August 2001 through May 2002. D) Total percent removal of acorns at each site after the entire 60-day exposure period. Different letters indicate significant differences based on Bonferroni pair-wise comparisons at 􀁟 = 0.05. Error bars are standard deviation, N = 2. 530 Southeastern Naturalist Vol. 6, No. 3 (F = 60.518, P < 0.001), with B50 and HDW sites having the most shade in July (Table 2). Soil temperatures mirrored the canopy-openness data, with the B50 and HDW sites having the lowest average soil temperatures. Discussion We found that longleaf pine recruitment was consistent across sites differing in slope position and fire history, while recruitment of oak species was greatest in sites with an intermediate time-since-fire (B90). Contrary to our expectations, the sites with the most oak seed rain (HDW) did not have the highest numbers of small-mammal seed predators, and seed removal was highest in sites with few small-mammal captures and low seed rain (B90). Southern red oak seedling survival was influenced by the abiotic site conditions found along the chronosequence, with the highest survival in the sites with the longest time-since-fire (B50) and lowest in the most recently burned sites (B01), corresponding with more favorable light and soil-moisture levels (i.e., summer shade and higher summer soil moisture). Ground-layer coverage ranged from 14% in the mesic hardwood stands to over 50% in the most recently burned stands. Similarly, species richness was greatest in the sites that had been recently burned along with the sandhills sites. As expected, there was no measurable longleaf seedling cover in the upland slope hardwood (HDW) stands. However, longleaf seed rain was measurable over all sites, with more longleaf seeds than acorns falling into the seed traps. Seed production by oaks in our sites was relatively low, suggesting our study may have been conducted in a nonmast year. Although much is known about masting in northern oak species (Elkinton et al. 1996, Greenberg and Parresol 2000, McCracken et al. 1999, Ostfeld et al. 1996, Schnurr et al. 2002, Wolff 1996), little is known about southern oaks. More research on the topic is necessary. The presence of pine and oak seedlings in all the upland longleaf pine stands suggests that time since fire does not result in environmental or biotic conditions that would limit regeneration. Longleaf seedling cover in recently burned stands and in those that have not been burned since before the SRS was created does not agree with results from other sites, which show that longleaf regeneration declines with time since fire in upland forests in the southeastern US (Hartnett and Krofta 1989). The relatively high cover of oak seedlings with no recorded oak seed rain, and highest amounts of seed removal in sites last burned in the 1990s, suggests that seedlings may have accumulated from prior mast years or that seeds have been dispersed into the stands by jays or woodpeckers. Our results indicate that a complex mixture of abiotic site conditions and biotic factors determine oak, specifically southern red oak, recruitment with time-since-fire. Close agreement between short-term seedling-survival patterns and environmental conditions related to evapotranspiration, with higher survival in the cooler, wetter, and more shaded long-unburned stands and lower survival in the hotter, drier, and more open recently 2007 J.L. Schnurr and B.S. Collins 531 burned stands, agrees with the abiotic controls on oak regeneration found by other researchers (Garcia et al. 2002, Jacqmain et al. 1999, Jones at el. 1994, Monk 1965, Schumann et al. 2003) and suggests the potential for oaks to increase in the absence of fire. However, we found that seed production or survival could potentially limit oak recruitment after fire, at least in years of low acorn production. Acorn abundance was low in all the upland longleaf pine stands, and was not measurable in the stands burned in 2001 and the 1990s. Over 50% of acorns were removed from all sites, with stands burned in the 1990s (where the largest numbers of small mammals were captured) having 98% acorn removal. Although high seed rain during mast years may lead to episodic recruitment, the lack of seedling cover in the stands last burned before 1950 suggests low seed rain, and low, but consistent, rates of seedling removal, along with long-term seedling mortality may offset short-term seedling survival and lead to low oak regeneration in long-unburned sites. Overall, results of our research suggest environmental conditions, including light, associated with canopy type and soil moisture differences related to topography have a greater influence than time since fire on southern red oak and longleaf pine regeneration in the upland forests we sampled. Lack of longleaf seedling cover despite seed production and low populations of small-mammal seed predators in the hardwood stands, where light levels were at 10% open sky and were the lowest among the sampled stands, suggests that shade may limit longleaf regeneration. Although light levels decreased among the upland longleaf sites with time since fire, levels only approached those of the hardwoods forest in the longest unburned (B50) sites. Low southern red oak seedling survival in the most recently burned sites and the sandhills sites, where the canopy was the most open and soil moisture was lowest in July, suggests that summer drought may limit regeneration of this species. Although where conditions are more mesic, as in the hardwood and longer-unburned upland longleaf sites, biotic factors may have a relatively stronger influence on oak regeneration, relatively consistent acorn production (J.L. Schnurr, pers. observ.) and low smallmammal populations (Langley and Shure 1980; Loeb 1999; B. Reinhart, Savannah River Ecology Lab, Aiken, SC, pers. comm.; Smith et al. 1984.) may combine to produce low overall biotic effects. Results of our research did not support our prediction that sites with greater acorn production would have greater densities of seed predators, increased seed predation, and decreased seedling recruitment of southern red oak. However, as predicted, southern red oak recruitment did increase down the topographic gradient from sandhills to mesic hardwood forest, and with time since fire in the upland pine stands, mainly in response to soil moisture and light availability. Acknowledgments Field assistance was provided by Brian Hough, Chris Crawford, Nate Huber- Fliflet, Pamela Weisenhorn, Matt Opdike, and John Dilustro. Brian Hough provided 532 Southeastern Naturalist Vol. 6, No. 3 comments on a draft of the manuscript. Financial Assistance Award Number DEFC09- 96SR 18456 from the US Department of Energy to the University of Georgia Research Foundation supported this research. Literature Cited Brockway, D.G., and C.E. Lewis. 1997. Long-term effects of dormant-season prescribed fire on plant community diversity, structure, and productivity in a longleaf pine-wiregrass ecosystem. Forest Ecology and Management 96:167–183. Christensen, N.L. 1988. Vegetation of the southeastern coastal plain. Pp. 317–364, In M.G. Barbour and W.D. Billings (Eds.). North American Terrestrial Vegetation. Cambridge University Press, Cambridge, UK. 434 pp. Clark, J.S., E. Macklin, and L. Wood. 1998. Stages and spatial scales of recruitment limitation in southern Appalachian forests. Ecological Monographs 68:213–235. Collins, B.,R. Sharitz, K. Madden, and J. Dilustro. 2006. Comparison of sandhills and mixed pine hardwood communities at Fort Benning, Georgia. Southeastern Naturalist 5(1):93–102. Dilustro J.J., B. Collins, L. Duncan, and R. Sharitz. 2002. Soil texture, land-use intensity, and vegetation of Ft. Benning upland forest sites. Journal of the Torrey Botanical Society 129(4):280–297 Elkinton, J.S., W.M. Healy, J.P. Buonaccorsi, G.H. Boettner, A.M. Hazzard, H.R. Smith, and A.M. Liebhold. 1996. Interactions among gypsy moths, white-footed mice, and acorns. Ecology 77:2332–2342. Frazer, G.W., and C.D. Canham. 1999. Gap Light Analyzer (GLA): Imaging software to extract canopy structure and gap light transmission indices from truecolour fisheye photographs. Copyright 1999. Simon Fraser University, Barnaby, BC, Canada and the Institute of Ecosystem Studies, Millbrook, NY. Garcia, D., M.-J. Banuelos, and G. Houle. 2002. Differential effects of acorn burial and litter cover on Quercus rubra recruitment at the limit of its range in eastern North America. Canadian Journal of Botany 80:1115–1120. Gardiner, E.S., and J.D. Hodges. 1998. Growth and biomass distribution of cherrybark oak (Quercus pagoda Raf.) seedlings as influenced by light availability. Forest Ecology and Management 108:127–134. Greenberg, C.H., and B.R. Parresol. 2000. Acorn production characteristics of southern Appalachian oaks: A simple method to predict within-year crop size. Research Paper SRS-20. US Department of Agriculture, Forest Service, Southern Research Station, Asheville, NC. 16 pp. Hartnett, D.C., and D.M. Krofta. 1989. Fifty-five years of post-fire succession in a southern mixed-hardwood forest. Bulletin of the Torrey Botanical Club 116:107–113. Heyward, F. 1939. The relation of fire to stand composition of longleaf pine forests. Ecology 20:287–304. Houle, G. 1998. Seed dispersal and seedling recruitment of Betula alleghaniensis: Spatial inconsistency in time. Ecology 79:807–818. Hulme, P.E., and M.K. Hunt. 1999. Rodent post-dispersal seed predation in deciduous woodland: Predator response to absolute and relative abundance of prey. Journal of Animal Ecology 68:417–428. Janzen, D.H. 1970. Herbivores and the number of tree species in tropical forests. American Naturalist 104:501–528. 2007 J.L. Schnurr and B.S. Collins 533 Jacqmain, E.I., R.H. Jones, and R.J. Mitchell. 1999. Influences of frequent coolseason burning across a soil-moisture gradient on oak community structure in longleaf pine ecosystems. American Midland Naturalist 141:85–100. Jensen,T.S. 1985. Seed-seed predator interactions in European beech, Fagus sylvatica and forest rodents, Clethrionomys glarelous and Apodemus flavicollis. Oikos 44:149–156. Jones, R.H., R.R. Sharitz, S.M. James, and P.M. Dixon. 1994. Tree population dynamics in seven South Carolina mixed-species forests. Bulletin of the Torrey Botanical Club 121:360–368. Jones, S.M., D.H. van Lear, and S.K. Cox. 1981. Composition and density-diameter pattern of an old-growth forest stand of the Boiling Springs Natural Area, South Carolina. Bulletin of the Torrey Botanical Club 108:347–353. Jones, S.M., D.H. van Lear, and S.K. Cox. 1984. A vegetation-landform classification of forest sites within the upper coastal plain of South Carolina. Bulletin of the Torrey Botanical Club 111:349–360. Kilgo, J.C. and J.I. Blake (Eds.). 2005. Ecology and Management of a Forested Landscape: Fifty Years on the Savannah River Site. Island Press, Washington, DC. 479 pp. Langley, A.K., Jr., and D.J. Shure. 1980. The effects of loblolly pine plantations on small mammal populations. American Midland Naturalist 103:59–65. Larsen, D.R., and P.S. Johnson. 1998. Linking the ecology of natural oak regeneration to silviculture. Forest Ecology and Management 106:1–7. Li, Q., and K. Ma. 2003. Factors affecting establishment of Quercus liaotungensis Koidz. under mature mixed oak forest overstory and in shrublands. Forest Ecology and Management 176:133–146. Loeb, S.C. 1999. Responses of small mammals to coarse woody debris in a southeastern pine forest. Journal of Mammalogy 80:460–471. LoGiudice, K., and R.S. Ostfeld. 2002. Interactions between mammals and trees: Predation on mammal-dispersed seeds and the effect of ambient food. Oecologia 130:420–425. McCracken, K.E., J.W. Witham, and M.L. Hunter, Jr. 1999. Relationships between seed fall of three tree species and Peromyscus leucopus and Clethrionomys gapperi during 10 years in an oak-pine forest. Journal of Mammalogy 80:1288–1296. Monk, C.D. 1965. Southern mixed hardwood forest of northcentral Florida. Ecological Monographs 35:335–354. Monk, C.D. 1968. Successional and environmental relationships of the forest vegetation of north central Florida. American Midland Naturalist 79:441–457. Mosandl, R., and A. Kleinert. 1998. Development of oaks (Quercus petraea (Matt.) Liebl.) emerged from bird-dispersed seeds under old-growth pine (Pinus sylvestris L.) stands. Forest Ecology and Management 106:35–44. Nilsson, S.G. 1985. Ecological and evolutionary interactions between reproduction of beech Fagus sylvatica and seed eating animals. Oikos 44:157–164. Ostfeld, R.S., C.G. Jones, and J.O. Wolff. 1996. Of mice and mast: Ecological connections in eastern deciduous forests. BioScience 46:323–330. Ostfeld, R.S., R.H. Manson, and C.D. Canham. 1997. Effects of rodents on survival of tree seeds and seedlings invading old fields. Ecology 78:1531–1542. Phares, R.E. 1971. Growth of red oak (Quercus rubra L.) seedlings in relation to light and nutrients. Ecology 52:669–672. 534 Southeastern Naturalist Vol. 6, No. 3 Quarterman, E., and C. Keever. 1962. Southern mixed hardwood forest: Climax in the southeastern coastal plain, USA. Ecological Monographs 32:167–185. Schnurr, J.L., R.S. Ostfeld, C.D. Canham. 2002. Direct and indirect effects of masting on rodent populations and tree seed survival. Oikos 96:402–410. Schnurr, J.L., C.D. Canham, R.S. Ostfeld, and R.S. Inouye. 2004. Neighborhood analyses of small-mammal dynamics: Impacts on seed predation and seedling establishment. Ecology 85:741–755. Schumann, M.E., A.S. White, and J.W. Witham. 2003. The effects of harvest-created gaps on plant species diversity, composition, and abundance in a Maine oak-pine forest. Forest Ecology and Management 176:543–561. Schupp, E.W. 1988. Seed and early seedling predation in the forest understory and in treefall gaps. Oikos 51:71–78. Shelton , M.G., and M.D. Cain. 1999. Structure and short-term dynamics of the tree component of a mature pine- oak forest in Southeastern Arkansas. Journal of the Torrey Botanical Society 126:32–48. Shibata, M., and T. Nakashizuka. 1995. Seed and seedling demography of four co-occurring Carpinus species in a temperate deciduous forest. Ecology 76:1099–1108. Skeen, J.N., P.D. Doerr, and D.H. van Lear. 1993. Oak-hickory-pine forests. Pp. 1– 34, In W.H. Martin, S.G. Boyce, and A.C. Echternacht (Eds.). Biodiversity of the Southeastern United States Upland Terrestrial Communities. John Wiley and Sons, Inc., New York, NY. Smith, M.W., W.R. Teska, M.H. Smith. 1984. Food as a limiting factor and selective agent for genic heterozygosity in the cotton mouse Peromyscus gossypinus. American Midland Naturalist 112:110–118. Sork, V.L., J. Bramble, and O. Sexton.1993. Ecology of mast-fruiting in three species of North American oaks. Ecology 74:528–541. Walters, J.R. 1991. Application of ecological principles to the management of endangered species: The case of the Red-cockaded Woodpecker. Annual Review of Ecology and Systematics 22:505–523. Ware, S., C. Frost and P. Doerr. 1993. Southern mixed hardwood forest: The former longleaf pine forest. Pp 447–493, In W.H. Martin, S.C. Boyce and A.C. Ecternacht (Eds.). Biodiversity in the Southeastern United States: Lowland Terrestrial Communities. John Wiley and Sons, New York, NY. 502 pp. Whelan, C.J., M.F. Willson, C.A. Tuma, I. Souza-Pinto. 1991. Spatial and temporal patterns of postdisperal seed predation. Canadian Journal of Botany 69:428–436. Wolff, J.O. 1996. Population fluctuations of mast-eating rodents are correlated with the production of acorns. Journal of Mammalogy 77:850–856. Workman, S.W., and K.W. McLeod. 1990. Vegetation of the Savannah River Site: Major Community Types. Savannah River Site National Environmental Research Park Program No. 19.