Effects of habitat, burial, age and passage through birds on germination and establishment of chinese tallow tree in coastal South Carolina

Journal of the Torrey Botanical Society, Apr-Jun 2001 by Renne, Ian J, Spira, Timothy P, Bridges, W C

STUDY SITE. Field studies were conducted from 1995-1996 and 1998-2000 in the 3077 ha Hobcaw Forest (33 deg 20'N, 79 deg 15'W) located on the outer Coastal Plain in Georgetown County, South Carolina. Mean annual precipitation was 1315 rum and mean January and August air temperatures were 9 deg C and 27 deg C, respectively, from 1930-1996 (Clemson Forestry Department, Georgetown County). In the plot study (see below), mixed pine-hardwood stands (MPH) were sampled. Common canopy species included loblolly pine (Pinus taeda L.), live oak (Quercus virginiana Mill.), and sweetgum (Liquidambar styraciflua L.). Soils ranged from moderately well to poorly drained. Two spoil areas (SA) separated by about five km were also sampled. The latter were highly disturbed areas with fine-- textured soil dikes enclosing a settling pond for dredged material. Tallow trees were the most common large woody plant on these dikes (pers. obs.). Lastly, we worked in longleaf pine-turkey oak forests (PTO). These dry areas had sandy, excessively drained soil and contained few tallow trees (pers. obs.).

In the transect study (see below), five common coastal forest types that contained established tallow tree populations were sampled. The maritime evergreen forest (MAR) had a relatively open 25 m tall canopy of loblolly pine and live oak, high shrub and vine density, and moderate to poorly drained dark gray sandy soil. The loblolly pine forest (LP) contained a continuous 20 m tall canopy, poorly developed shrub and herb layers, a 5-10 cm litter layer of pine needles and moderately well drained grayish brown sand. Grasses, wingstem (Verbesina occidentalis Walt.) and tallow trees dominated a recent mixed pine-hardwood forest clearcut (CC). Soil was moderately well drained gray brown loamy sand. The young mixed pine-hardwood forest (YMPH) had a 7-11 m tall canopy of loblolly pine, water oak (Q. nigra L.) and tallow trees; wax myrtle (Myrica cerifera L.) was the dominant shrub, and a dark gray well to moderately poorly drained sandy soil was present. A welldeveloped bottomland hardwood forest (BLH; blackwater type) had a diverse 25-30 m tall canopy, a thick shrub layer dominated by dwarf palmetto (Sabal minor (Jacq.) Persoon) and a poorly drained black loam that was intermittently flooded. Community and soil descriptions follow Schafale and Weakley (1990) and Stuckey (1982), respectively, and nomenclature follows Radford et al. (1968).

FIELD PLOT STUDY. The effects of habitat type and planting date were tested on seedling emergence and survival by using an unbalanced 3 X 2 factorial, with three habitat types (MPH, SA and PTO), two planting dates (19 December 1995, 23 February 1996) and five replications. Dates correspond to times when tallow tree seeds are dispersed by birds (Renne et al. 2000). On each date, 50 seeds with no aril damage were collected from 30 widely separated trees and mixed in a bag. Seeds were then sown at 5 cm intervals in a 10 by 10 grid (n = 100 seeds per plot) and covered with ca. 0.5 cm of soil to decrease the chance of seed removal by animals, gravity or water. The 0.25 m^sup 2^ plots were fenced with chicken wire to deter feral hogs (Sus scrofa L.) from rooting in them. To standardize local microsite differences and test the effect of soil type among habitats on germination and seedling survival, vegetation and litter were removed from plots before seeds were sown. On 23 February 1996, a fenced 0.25 m^sup 2^ plot was constructed adjacent to each of the 15 plots and seeds were similarly sown. One pair of plots in SA was destroyed by feral hogs, thus data from 28 plots were analyzed (n = 2800 seeds). As a negative control, a 0.25 m^sup 2^ area next to each set of plots was randomly chosen so that emergence from a soil seed bank could be estimated.