<p> There has been an unusual shift in timberland ownerships in the United States over the past few decades, in which mostly Forest Product Companies have divested properties to institutional owners, like Timber Investment Management Organizations and Real Estate Investment Trusts. Northeast region of the country has been influenced by this trend. Studies have suggested changes in harvesting, silviculture, and conservation efforts under new ownerships may alter forest structure and resiliency. However, there is little documentation on the spatial pattern of such ownership change and its effect on forest dynamics. This dissertation tries to address some of these knowledge gaps by applying a variety of spatial and statistical analyses to Forest Inventory and Analysis (FIA) data from 2003 to 2012. In Chapter 2, I assessed spatial pattern of change in timberland ownerships and linked these incidences with socio-ecological variables. Largest observed shift was from industrial to institutional ownership, a net increase of only 1% in institutional timberlands. However, there was a significant clustering pattern, and clusters were significantly related to forest type, distance from urban center and distance from road. In Chapter 3, I explored changes in aboveground biomass (?AGB) among different ownerships, harvesting, and forest types, and other selected factors. Overall, a positive ?AGB (671 lb/ac/yr) was observed, with Non Industrial Private Forest (NIPF) timberlands having higher growth than industrial and institutional timberlands. Among forests, Elm-Ash-Cottonwood had the best growth, and among harvesting regimes, plots harvested before first measurement had highest growth. Ownership, harvest, disturbance, silvicultural treatment, forest type, stand age, site index, and precipitation were significantly related to &Delta;AGB. In Chapter 4, I compared three indices used to characterize similarity between overstory-understory tree species composition and assessed potential future change in forest composition. Ownership, forest types, precipitation, stand age, site index, stand origin, slope, elevation, proximity to road and urban centers all contributed to explaining variation in change in similarity indices. Among ownership categories, industrial and institutional ownerships had greater dissimilarity over time, in contrast to other ownerships. The final chapter discusses some potential implications of these results to northern forest structure, resiliency and sustainable production.</p>

I compared composition of plant communities in second growth (60 - 80 years old) and old growth rich coves for evidence of reduced abundance or presence of plant species in the second growth communities. I evaluated the utility of both univariate and multivariate analytical methods. Data were collected from twenty-six 1000 m2 plots located across three distinct mountain ranges and geologies in the southern Appalachians of North Carolina. Comparison of nested and non-nested sampling protocols found no statistical differences in the parameters of species-area curves. Furthermore, I found no difference between the species-area curves of old growth and second growth; but, multiple comparison procedures identified 13 of 79 species with either lower abundance or incidence of occurrence in second growth rich coves. Nonmetric Multidimensional Scaling (NMS) revealed no age class differences, but detected significant block differences associated with the three mountain ranges. Elevation and base cation concentrations were strongly correlated with the first and third NMS axes, respectively, but there was no strong correlate with the second NMS axis. In comparing univariate and multivariate analysis methods, I found that (1) aggregate univariate methods (like species diversity and species-area relationships) facilitate comparison of unrelated communities (2) multivariate methods support comparison of related communities and (3) univariate, multiple comparison procedures readily detect specific species changes within a community .

Two studies were conducted to investigate factors influencing graft success and subsequent growth of Fraser fir [Abies fraseri (Pursh) Poir]. The traditional time of grafting (late winter/early spring) was compared with eight summer/early fall grafting dates from mid-July through mid-October. Optimal grafting success (95%) was in the late winter/early spring (April) while the scions were dormant and the rootstocks were becoming active. Success of subsequent grafting dates decreased from 52% (14 July) to 0% (20 Oct.). Shade improved summer graft success (52% with, 38% without). Irrigation did not affect graft success or growth. Grafting of stored dormant scion material in summer/early fall was not successful (< 1%). In a second study, success and subsequent growth of Fraser fir cleft grafts were studied in relation to season of grafting (late summer vs. spring), grafter, and origin of scion material (height in the tree and lateral branch order (first vs. second). Grafting in early September yielded only 3% success compared to 70% for mid-April. Grafters had significantly different graft success (86% for Grafter 1 with 5 years experience vs. 54% for Grafter 2 with 1 year experience). First-order laterals from the upper crown yielded the best graft success and growth (except plagiotropism). First-order laterals were better than second-order laterals for all growth measurements.

Positive assortative mating (PAM) may substantially enhance genetic variance in a breeding population (BP). This creates potential for additional genetic gains available through production populations (PP) to forest plantations. Open-nucleus strategies (NB) have been incorporated in forest tree breeding programs. In NB, the BP is subdivided into two hierarchical levels, a nucleus and a main population, and can be considered a less rigorous form of PAM. First, PAM was compared to NB by stochastic simulation considering jointly genetic gain and diversity within the framework of a long-term breeding program. Test effort was either assumed constant throughout the entire BP or was redirected according to the rank of each mate. The simulation revealed that PAM results in larger gains in the PP compared to NB under both situations and at any target PP diversity. Second, the test effort during PAM was redirected by varying family sizes as a linear function of mid-parent BLUP values. The actual distribution of mid-parent BLUP values was standardized by a constant value, which was varied in simulation scenarios to cover the entire range of the distribution of family sizes. When equal numbers of progenies were selected per family and the variation in family sizes was maximized, only a minimal reduction in BP diversity was observed, compared to cases with constant family sizes. Under such favorable conditions, the redistribution of resources increased genetic response and variance in the BP, causing substantially greater genetic response in the PP. These conclusions were verified under a mixed-inheritance model with a major-gene locus contributing to variation in a quantitative trait. Finally, the investigation of PAM was extended by considering correlated traits within the framework of a clonal forestry program. The success of somatic embryogenesis in families generated by crossing elite genotypes developed in the breeding program was either considered exponentially distributed or constant. The distribution of success caused non-significant differences in genetic gain of PP. These conclusions were verified over a range of correlation, heritabilities and economic weights of traits.

Gypsy moth (Lymantria dispar: Linnaeus) is a polyphagous non-native forest pest first introduced to Medford, Massachusetts in 1869. It has since spread as far south as North Carolina and as far west as Wisconsin. Gypsy moth is responsible for defoliating approximately 100,000 hectares of forest annually, resulting in mortality in a small percentage of trees, averting behavior by recreators, and creating a nuisance to the general public. Limiting the spread of gypsy moth is beneficial because it delays the onset of costs and losses associated with quarantine, tree defoliation and mortality, and nuisance. Gypsy moth is believed to disperse naturally up to 2.5 km/yr (e.g. early instar ballooning) but has been observed to disperse much greater distances. The scientific consensus is that this longer distance dispersal occurs through anthropogenic vectors (e.g. egg masses being transported on firewood). Despite the resources that United States Department of Agriculture and state agencies dedicate to eliminating and managing new infestations resulting from long distance dispersal, there has been limited empirical research on the relationship between the dispersal of gypsy moth and the movement of people and their goods. This thesis develops a conceptual framework of the anthropogenic factors that could affect dispersal, measures these factors with secondary data at the county level from a variety of sources, and estimates the presence or absence of gypsy moth using logistic regression models. The dependent variable is drawn from trap catch records archived by the United States Department of Agriculture Forest Service program, Slow-the-Spread, in areas distal to the 1 moth/trap line between 1999 and 2007. Through step-wise logistic regression estimating sub-models that include variables representing each broad anthropogenic factor, a final empirical model is specified. The variables of the model are estimated independently for each year from 1999 to 2007, resulting in a mean Pseudo R square of 0.568. Consistently significant ( ) anthropogenic variables are the number of households using wood for heating fuel and mean household income. These findings are discussed with regard to invasion theory and quarantine policy. One key implication is the continual importance of regulating and raising awareness about the risk of moving firewood from infested to uninfested zones.

Alley cropping may prove useful in the southeast U.S., providing multiple products and income streams, as well as affording sustainable land use alternatives to conventional farming and forest planting. Such systems in this region are of particular interest because they can help in soil conservation and nutrient retention and aid in sustaining and improving valued but degraded farmland. In the current study triple row single-species strips of loblolly pine (Pinus taeda), longleaf pine (Pinus palustris) and cherrybark oak (Quercus pagoda) were planted as 1-year-old seedlings separated by 12 or 24 m wide areas of soybean in spring 2007. Select individual tree seedlings of each species were treated with on-farm wastes, used as mulch in a circular area around each stem. These waste/mulches were hog bedding (corn stover + hog waste removed from swine houses), old hay (year-old rolled/slightly spoiled bermudagrass hay - Cynodon dactylon) and black plastic bedding film. After the first season of growth with the applied mulches, tree seedling growth rates were higher for cherrybark oak and longleaf pine seedlings mulched with old hay applied at 7.5 cm deep in 30 cm radius around each seedling. Other mulches had varying effects on soil conditions, but no significant impact on tree growth as compared to the untreated control seedlings. These first-year findings suggest that mulching with specific on-farm wastes may be a valuable management tool in temperate alley cropping systems. Longer term tree growth in this system and with regard to these initial mulching treatments will be studied.

As the doubling time of the global human population decreases, increasing emphasis is placed on sustainable development by both policy makers and scientists. Sustainable forest management is one part of the overall picture of sustainable development. One method to assess sustainable forest management is through the use of criteria and indicators. Criteria represent sustainable management goals. Indicators are measurable quantities that designate whether the goals are being met. The maintenance of forest health and vitality is a criterion of the Montréal Process Criteria and Indicators for the Conservation and Sustainable Management of Temperate and Boreal Forests. Measures of air pollution, forest disturbance, and change in ecological integrity provide indicators of how well forest health and vitality are being maintained. Using national databases, I assess air pollution in the United States, demonstrate the use of epidemiological approaches to examine forest disturbances, and develop an analytical technique to identify gaps and target priorities in reserve networks. The analyses in this dissertation offer new approaches to large-scale analysis of Montréal Process Criteria and Indicators. The results can be summarized as follows. (1) From 1994 through 2000 air pollution was highest in the northeastern United States and the oak-hickory and loblolly-shortleaf forest type groups were consistently exposed to more air pollution than other forest types. Conversely, the western white pine and larch forest type groups were consistently exposed to less air pollution than all other forest types. (2) Examination of the southeastern United States revealed high rates of forest fragmentation in the piedmont and coastal plain region. In the Pacific North west, insect and pathogen activity was analyzed and recurring clusters of high rates of activity were identified. (3) Although protected areas of the Douglas-fir forest type group occurred throughout much of the species range, most existed in colder and drier parts of the range. To conserve representative habitats, future conservation efforts would be most effective in warmer and wetter areas of western Oregon, northwestern Washington, and northwestern California.

In North Carolina, non-industrial private forestland owners control approximately 78% of the state?s forested resource (18.8 million acres). Because they provide benefits that contribute to the health and economic well-being of the state, it is critically important that this diverse group of people is supplied with research-based forestry information and education to ensure the sustainable management of North Carolina?s natural resources. Researchers have evaluated the efficacy of non-traditional methods of information exchange and have determined overall preferences for information delivery methods, but have been unable thus far to satisfactorily connect particular information delivery preferences with other characteristics of landowners. The objective of this study was to seek out and describe groups within the population of non-industrial private forestland owners with particular information delivery method preferences. Identification of these groups will allow educational efforts to be more directed, making outreach efforts more efficient and cost-effective. A questionnaire was mailed to a stratified random sample of 2600 non-industrial private forestland owners from both urban and rural counties found within each of North Carolina?s seven Cooperative Extension districts. A cluster analysis was performed to identify distinct clientele groups among landowners based on their preferences for delivery methods. These groups were used to correlate preference for information delivery methods with more easily identifiable socio-demographic, land, or management characteristics of individuals. Based on the results, recommendations are made to assist Extension Forestry at North Carolina State University increase its impact on the citizens of the state.

We sampled seven intact nonriverine wet hardwood forests to establish target ranges for vegetation, soils and hydrology and to examine trends in plant species composition along a wetness gradient. Although quantitative vegetation analysis for this community has been published, broad drainage classes were used to represent a moisture gradient. We investigated trends along a finer-scaled wetness gradient utilizing a novel wetness index that incorporated indicators of saturated soils. Understanding small-scale patterns in plant community composition is useful in planning wetland restoration projects. Although no strong relationship was found between wetness index classes and plant community composition, these data represent the vegetative community supported by soils specific to each wetness class. Absence of this relationship is most likely a result of the dominance of A. rubrum and L. styraciflua in all wetness classes as well as a history of disturbance in each forest and other unknown stochastic variables. For the restorationist hoping to restore a historic nonriverine wet hardwood forest, once known as oak flats, we can suggest guidelines for restoring hydrology to the wetness classes in which we found oaks with high importance.

Restoring longleaf pine ecosystems is essential for managing rare plant and animal species and protecting biological diversity in the southeastern Coastal Plain of the United States. Natural longleaf pine ecosystems range from xeric uplands to poorly-drained flatwoods and savannas. Most existing stands, however, occur on xeric to dry-mesic sites and approaches to restoring longleaf pine to wetter sites traditionally utilize intensive practices. There is little information available on the efficacy of these practices to establish longleaf pine seedlings on poorly-drained sites and their impacts on soil properties, seedling survival and growth, and the understory plant community. A research project was established at Marine Corps Base Camp Lejeune, NC to evaluate the effects of site preparation methods for returning longleaf pine on hydric soils with no natural seed source. Various site preparation treatments were evaluated in a field experiment, and results revealed greater growth and earlier emergence from the grass stage with more intensive site preparation. There was a marginal increase in soil nutrients, and a slight increase in foliar nutrients found with the more intensive treatments. Site preparation influenced seedling growth in the short-term and this was likely due to the cumulative effects of controlling competition and modifying the planting site. Marine Corps Base Camp Lejeune and other Department of Defense installations include both former and remnant longleaf pine ecosystems that support federally protected plants and animals such as the red-cockaded woodpecker - and thus face the challenge of restoring former, poorly-drained longleaf pine ecosystems. A land use history revealed that, largely due to its poorly-drained status and inaccessibility, the majority of disturbance on the research area occurred after the 1920âs and was largely due to forestry activities. Since purchasing the land area of the project in 1996, the Marine Corpsâ challenge has been to balance the mission of training and readiness with the need for restoration and long-term management of longleaf pine ecosystems. The results of this work provide natural resource managers with a scientific foundation for assessing choices to assist in this restoration and management effort.