Understory forage production, species composition, and nutrient concentrations were studied in relation to overstory structure and site characteristics in thinned loblolly pine (Pinus taeda) plantations in the Virginia Piedmont. Stands exhibited a wide range in overstory basal areas (18.8 to 43.5 m²/ha) and site indices (14.2 to 23.8 m at base age 25 years). Understory forage production for the 0 to 2 m stratum averaged 610 kg/ha and ranged from 154 to 1690 kg/ha. Initially, differential models were used to develop prediction equations relating understory production to overstory characteristics, but an empirical prediction equation proved to be somewhat superior. Forage production was most predictable from total overstory basal area, canopy cover, and slope position. Understory species composition was analyzed with respect to overstory structural and site gradients, using vegetation ordination techniques. Species composition was less closely related to these gradients than was production; however, the greatest species diversity appeared to occur during the period following thinning but before crown closure, when successional and mid-tolerant species coexisted. Nutrient concentrations in forage material averaged 4643 cal/g for gross energy, 38.3 percent for invitro dry matter digestibility by white-tailed deer (Odoooiteus virginiana) and 8.4 percent for crude protein. There were no apparent trends of these nutrient concentrations with respect to overstory structural or site characteristics. Forage based carrying capacities for white-tailed deer were computed using values from this study for production and nutrient concentrations, and values from the literature for deer forage preferences and nutrient requirements. Results indicated that the pine habitat could support 0.03 to 0.19 lactating does per ha during the summer season, and that energy and not protein is likely the limiting nutritive parameter. Suggestions are made for future research in pine overstory-understory and wildlife habitat relationships. These include the use of experimental overstory manipulation followed by periodic remeasurements to directly observe changes in understory production and species composition, intensive sampling to determine specific local wildlife forage preferences, and the quantification of wildlife movements and population dynamics. / Ph. D.

Wild turkey mortality rates, habitat use patterns, and recruitment were investigated on intensively managed pine forest of the Central Piedmont. Thirty-two adults and poults were captured with adult dosages of alpha-chloralose. Poults were easily sedated and usually recovered 1 day sooner than adults. The annual mortality rate for marked turkeys was 85%; September-February mortality was 65% (N=17), and March-August mortality was 57% (N=21). Turkeys that died during hunting seasons were recovered significantly (p < 0.05) closer to roads than those that died out of hunting seasons. High mortality during the hunting seasons and the association of hunting season mortality with roads suggests a crippling loss to legal harvest ratio as high as 4:1. Use of young pine plantations, hardwood leave strips, large stands of mixed hardwoods, mature pine stands, and fields was not significantly different (p > 0.1) from availability. Turkeys were associated with edges while on their winter and prenesting ranges. Nests were located in a variety of stand types, near edges, and in heavier cover than generally available. Nesting success was 48%. Post critical period brood survival was 90%. These data indicate that wild turkeys can adapt to the early stages of forest conversion for intensive pine management. / Master of Science

Destructive sampling was undertaken to estimate the contribution of the herb stratum and shrub stratum to above-ground phytomass and nutrients (N, P, K, Mg and Ca) in four different vegetation types in an Appalachian oak forest in southwestern Virginia. The study area is located at midslope on the southeast aspect of Potts Mountain in four non-contiguous compartments which are to be clearcut and whole-tree harvested in the near future. A partial random sampling scheme was utilized to select 20 x 20m study areas, within which 48 1 x 1 m and 16 5 x 5 m herb and shrub stratum plots were nested respectively. Statistical tests indicated that significant differences in above-ground phytomass and nutrients exist between vegetation types in both strata. The trend is towards greater amounts of understory vegetation on progressively drier sites. Total amounts (kg/ha) of nutrient elements show a similar trend. In contrast, nutrient element concentrations show a reverse trend, with higher concentrations of nutrients found on the more mesic sites. Ordination techniques were utilized to investigate differences in nutrient concentrations among species. Species groupings derived from the ordinations substantiated the sorting of vegetation types along a moisture gradient. Regression equations for individual / M. S.