In forested ecosystems, salamanders occupy important ecological roles as predator, prey and as potential regulators of ecological processes. The effects of forest management, particularly clearcut harvesting, on salamanders have been well documented; removal of overstory trees negatively affects abundances of salamanders. However, the length of time that salamander populations remain depressed following forest harvesting and factors limiting population recovery have been a source of controversy in the literature and are the goal of this dissertation. As part of the Southern Appalachian Silviculture and Biodiversity (SASAB) project (Chapter 1), a long-term replicated field experiment designed to evaluate a range of silvicultural treatments on biodiversity, I evaluated specific hypotheses related to salamander populations, their prey, and their habitat. First, I examined long-term trends in salamander abundances across a range of silvicultural treatments to determine whether negative effects of forest harvesting persisted for 13-years after harvest (Chapter 2) and to document the effects of multiple harvests on salamanders (Chapter 3). The relative abundances of terrestrial salamanders were quantified in six silvicultural treatments and an unharvested control and on six replicated field sites with night-time, area-constrained searches. Across 13-years of post-harvest data, terrestrial salamander abundances generally were lower in silvicultural treatments with some disturbance to the canopy (group selection harvest through silvicultural clearcut). Further, a comparison of demography of common species of salamanders suggested that differences in habitat quality existed between harvested and unharvested experimental units (EUs). A second harvest in the shelterwood plots to remove overwood had a cumulative negative effect on salamanders at one of two sites studied. Additionally, I conducted a sensitivity and elasticity analysis for eastern red-backed salamanders (Plethodon cinereus) and modeled population growth to evaluate the contribution of demographic parameters to population recovery. These analyses indicated that adult survival was the parameter with the greatest influence on the population growth rate and that >60 years would be required for recovery of salamander populations to preharvest levels even if habitat conditions were restored to preharvest conditions immediately. Next, I quantified the bioenergetics of salamanders across a disturbance gradient to evaluate whether changes to (1) invertebrate prey, (2) energy expenditure for basic maintenance costs, and or (3) an index to body condition could help explain observed changes to abundances or demography of salamanders from forest harvesting (Chapter 4). Although I did not detect a difference in abundances of invertebrates along the disturbance gradient, I determined that salamanders in recently disturbed forest stands expended approximately 33% more energy for basic maintenance costs in an active season and the body condition of salamanders was greater at one of two sites in disturbed EUs. Thus, the bioenergetics of terrestrial salamanders may have been affected by increasing temperatures from silvicultural disturbance and may cause salamanders to allocate less energy to reproduction or growth because of increased maintenance costs. In collaboration with Eric Sucre, Department of Forestry at Virginia Tech, I examined the effects of salamanders on invertebrates and ecosystem processes, specifically leaf litter decomposition. We constructed 12 in situ field mesocosms and I manipulated densities of red-backed salamanders into zero, low, and high density treatments. From June 2006-June 2008, I estimated invertebrate abundances, rates of leaf litter decomposition and food habits of salamanders across treatments. I found that invertebrate abundances were more affected by season than by the density of salamanders and that rates of leaf litter decomposition did not differ among salamander treatments. Salamanders were euryphagic and consumed more (by abundance and volume) herbivorous invertebrates than predators or detritivores. Finally, I modeled habitat relationships of terrestrial salamanders at two spatial scales on the SASAB study sites (Chapter 6). I quantified abundance of salamanders with area-constrained searches during warm rainy nights and measured forest characteristics related to foraging or refugia habitats or that described the overstory and understory of forest stands. At the scale of the 30 m2 transect and the 10 m2 sub-transect, abundance of salamanders was best described by models that incorporated descriptors of canopy cover and woody and herbaceous understory vegetation. Thus, terrestrial salamanders may have responded positively to forest stands with a mature overstory and structurally diverse understory for foraging habitat. Collectively, these data suggest that recovery of salamander populations after forest harvesting will take approximately 60 years, and that life history characteristics (low fecundity, late sexual maturity) and possibly changes to bioenergetics may contribute to the slow recovery. Further, silvicultural practices that retain some canopy trees through a rotation may have a more rapid return of salamander populations to preharvest levels and may encourage development of understory structure for salamander foraging. Although these data fill some gaps in knowledge of relationships between silviculture and terrestrial salamanders, many questions about long-term effects of disturbances on populations and habitats remain. My modeling of recovery of salamander populations depended on estimates of a survival from a congener, and I did not document whether forest harvesting decreases survival of terrestrial salamanders. Lastly, the influence of stochastic events on population dynamics particularly in disturbed stands was not examined in this dissertation. Therefore, future research on the SASAB study sites should continue to track abundances and demography across the disturbance gradient, acquire age-specific estimates of survival, determine whether salamanders exhibit density dependence, develop estimates of entire energy budgets, and use manipulative laboratory experiments to describe the role of plethodontid salamanders in ecosystem functions. / Ph. D.

Composite panel structure optimization is commonly decomposed into panel optimization subproblems, with specified local loads, resulting in manufacturing incompatibilities between adjacent panel designs. Using genetic algorithms to optimize local panel stacking sequences allows panel populations of stacking sequences to evolve in parallel and send migrants to adjacent panels, so as to blend the local panel designs globally. The blending process is accomplished using the edit distance between individuals of a population and the set of migrants from adjacent panels. The objective function evaluating the fitness of designs is modified according to the severity of mismatches detected between neighboring populations. This lays the ground work for natural evolution to a blended global solution without leaving the paradigm of genetic algorithms. An additional method applied here for constructing globally blended panel designs uses a parallel decomposition antithetical to that of earlier work. Rather than performing concurrent panel genetic optimizations, a single genetic optimization is conducted for the entire structure with the parallelism solely within the fitness evaluations. A guide based genetic algorithm approach is introduced to exclusively generate and evaluate valid globally blended designs, utilizing a simple master-slave parallel implementation, implicitly reducing the size of the problem design space and increasing the quality of discovered local optima. / Ph. D.

A typical approach to the synthesis/design optimization of energy systems is to only use steady state operation and high efficiency (or low total life cycle cost) at full load as the basis for the synthesis/design. Transient operation is a secondary task to be solved by system and control engineers once the synthesis/design is fixed. This thesis considers the system dynamics in the process of developing the system using a set of transient thermodynamic, kinetic, and geometric as well as physical and cost models developed and implemented for the components of a 5 kW PEMFC (Proton Exchange Membrane Fuel Cell) system. The system is composed of three subsystems: a stack subsystem (SS), a fuel processing subsystem (FPS), and a work recovery and air supply subsystem (WRAS). To study the effect of control to the optimization, State Space control design is used in a looped set of optimizations. These results are compared to those resulting from a more direct optimization of the controller designs in which the gains for the controllers are part of the decision variable set for the overall optimization. Then, dynamic optimization results are obtained and compared with steady-state optimization results to illustrate the advantages of dynamic optimization. Also, a multi-level optimization technique, dynamic iterative local-global optimization (DILGO), is utilized for the optimization of the PEMFC system by separating the system into three subsystems and the results are compared with the single level optimization results, in which the whole system is optimized together. / Ph. D.

A typical approach to the synthesis/design optimization of energy systems is to only use steady state operation and high efficiency (or low total life cycle cost) at full load as the basis for the synthesis/design. Transient operation as reflected by changes in power demand, shut-down, and start-up are left as secondary tasks to be solved by system and control engineers once the synthesis/design is fixed. However, transient regimes may happen quite often and the system response to them is a critical factor in determining the system's feasibility. Therefore, it is important to consider the system dynamics in the creative process of developing the system. A decomposition approach for dynamic optimization developed and applied to the synthesis/design and operation/control optimization of a solid oxide fuel cell (SOFC) based auxiliary power unit (APU) is the focus of this doctoral work. Called DILGO (Dynamic Iterative Local-Global Optimization), this approach allows for the decomposed optimization of the individual units (components, sub-systems or disciplines), while taking into account the intermediate products and feedbacks which couple all of the units which make up the overall system. The approach was developed to support and enhance current engineering synthesis/design practices by making possible dynamic modular concurrent system optimization. In addition, this approach produces improvements in the initial synthesis/design state at all stages of the process and allows any level of complexity in the unit's modeling. DILGO uses dynamic shadow price rates as a basis for measuring the interaction level between units. The dynamic shadow price rate is a representation of the unit's cost rate variation with respect to variations in the unit's coupling functions. The global convergence properties of DILGO are seen to be dependent on the mathematical behavior of the dynamic shadow price rate. The method converges to a "global" (system-level) optimum provided the dynamic shadow price rates are approximately constant or at least monotonic. This is likely to be the case in energy systems where the coupling functions, which represent intermediate products and feedbacks, tend to have a monotonic behavior with respect to the unit's local contribution to the system's overall objective function. Finally, DILGO is a physical decomposition used to solve system-level as well as unit-level optimization problems. The total system considered here is decomposed into three sub-systems as follows: stack sub-system (SS), fuel processing sub-system (FPS), and the work and air recovery sub-system (WRAS). Mixed discrete, continuous, and dynamic operational decision variables are considered. Detailed thermodynamic, kinetic, geometric, physical, and cost models for the dynamic system are formulated and implemented. All of the sub-systems are modeled using advanced state-of-the-art tools. DILGO is then applied to the dynamic synthesis/design and operation/control optimization of the SOFC based APU using the total life cycle cost as objective function. The entire problem has a total of 120 independent variables, some of which are integer valued and dynamic variables. The solution to the problem requires only 6 DILGO iterations. / Ph. D.

Understanding and establishing reaction mechanisms is an important area in heterogeneous catalysis. This dissertation describes the use of in situ laser Raman spectroscopy combined with kinetic measurements and dynamic experiments to determine the mechanism of catalytic reactions. Two cases involving ozone reactions on manganese oxide catalysts were treated. Manganese oxide was chosen because it is the most active of the transition metal oxides for ozone decomposition and because it is a well-known catalyst for complete oxidation reactions. The first case studied was that of the ozone decomposition reaction on a supported manganese oxide catalyst. An adsorbed species with a Raman signal at 884 cm-1 was observed and assigned to a peroxide species based on results of in situ Raman spectroscopy, 18O isotopic substitution measurements, and ab initio MO calculations. The reaction pathway of ozone decomposition was elucidated with carefully designed isotopic experiments. The reaction sequence was found to involve two irreversible, kinetically significant steps: 1) dissociative adsorption of ozone to form a peroxide species and an atomic oxygen species, and 2) desorption of the peroxide intermediate. The kinetic behavior of the peroxide species and the overall decomposition reaction were investigated to test the validity of the proposed sequence. The transient kinetics were found to be consistent with the steady state kinetics, and both were well represented by the two-step sequence, indicating that the proposed reaction sequence accurately described the mechanism of decomposition. The surface was found to be non-uniform, with activation energies that varied linearly with coverage. At zero surface coverage the activation energy for ozone adsorption was found to be 6.2 kJ mol-1, while that for desorption of the peroxide species was found to be 69.0 kJ mol-1. The second case investigated was that of ethanol oxidation using ozone on alumina and silica supported manganese oxide catalysts. Ethanol was found to react with ozone at lower temperatures than with oxygen, and also with a lower activation energy. The reaction kinetics was found to be well described by a power law equation with the reaction orders on ozone and ethanol being 0.89 and 0.81 respectively. The oxidation reactivity was found to be closely related to that of ozone decomposition, suggesting an important role of ozone decomposition in the reaction mechanism. In situ laser Raman spectroscopic studies showed the existence of adsorbed ethoxide species on the catalyst surface under reaction conditions, however, at a much lower concentration than when oxygen alone was used as the oxidant. Transient experiments provided direct evidence that surface peroxide (an adsorbed species due to ozone) and surface ethoxide (an adsorbed species due to ethanol) reacted with each other on the catalyst surface. / Ph. D.

Many research projects concerning the possible deleterious effects of ozone on forest health have been conducted on individual tree species. The common goal of these projects has been to identify mechanisms of damage by ozone, and then extrapolate research results to forests. Results from seedling studies are used to parameterize process-based tree growth models which are used to project mature tree responses to different levels of ozone. This approach has been criticized because nothing is known about differences in seedling and mature-tree responses to ozone. Another problem is that few projects have examined the effects of ozone on below ground processes; therefore, very little data exists for parameterizing the models. In order to address the problem of scaling seedling results to mature trees, and increase our level of understanding of ozone effects on below ground processes, an ozone fumigation experiment on northern red oak seedlings and mature trees was conducted. It was hypothesized that carbon reallocation to replace foliage damaged by ozone would decrease fine-root production and turnover. The red oak trees and seedlings were fumigated for three years with three levels of ozone (subambient, ambient, and 2X ambient) in open-top chambers. After two seasons of exposure, 2X ozone (0.082 ppm 7hr-mean conc.) reduced mature- tree cumulative net fine-root production and turnover by 31 and 41 %, respectively, relative to ambient ozone (0.042 ppm 7hr-mean conc.). For the same time period, ozone had no effect on seedling cumulative fine-root turnover; fine-root production was 25% higher under ambient ozone relative to subambient and 2X ambient ozone. During the summer, 1994, mature tree BUE was reduced by 2X ozone. Decreased fine-root production, turnover, and BUE under 2X ozone for the mature trees indicates that ozone can alter the dynamics of belowground carbon allocation in mature red oak. Since the seedlings were not sensitive to ozone, use of seedling results for modelling purposes may underestimate mature tree responses to ozone. / Master of Science

Coarse woody debris (CWD) plays an influential role in forested ecosystems by adding organic matter to soils, stabilizing the soil environment, providing wildlife habitat, preventing soil erosion, providing seedling establishment habitat, and involvement in the nutrient cycle. Most CWD research has been conducted in old-growth and unmanaged, second-growth forests. However, less is understood about CWD in intensively managed ecosystems, such as industrialized southern pine plantations. The objectives of this study were to determine the climatic and ecological factors that affect the decomposition rate of CWD, to predict the decomposition rate, specific gravity, and time since death (TSD) using multiple linear regression in industrial loblolly pine (Pinus taeda L.) plantations in the southeastern United States. The study sites for this project were part of a long-term, loblolly pine thinning study maintained by the Loblolly Pine Growth and Yield Research Cooperative at Virginia Tech. Measurements included piece size, position, and decay class. Samples of CWD were collected and analyzed to determine their mass and density. Decomposition rate of CWD was significantly different across position classes and decay classes: disk decomposition rates were significantly negatively correlated with disk diameter, large and small end piece diameter, estimated disk height, and disk dry weight. Average annual precipitation and average annual temperature were not significantly correlated with CWD disk decomposition rate. / Master of Science

The subgroup decomposition method (SDM) has recently been developed as an improvement over the consistent generalized energy condensation theory for treatment of the energy variable in deterministic particle transport problems. By explicitly preserving reaction rates of the fine-group energy structure, the SDM directly couples a consistent coarse-group transport calculation with a set of fixed-source "decomposition sweeps" to provide a fine-group flux spectrum. This paper will outline the implementation of the SDM into the three-dimensional, discrete ordinates (SN) deterministic transport code TITAN. The new version of TITAN, TITAN-SDM, is tested using 1-D and 2-D benchmark problems based on the Japanese designed High Temperature Engineering Test Reactor (HTTR). In addition to accuracy, this study examines the efficiency of the SDM algorithm in a 3-D SN transport code. / Master of Science

This thesis presents both a shared memory parallel version of the hybrid classification algorithm IGSCR (iterative guided spectral class rejection) and a novel data reduction technique that can be used in conjuction with pIGSCR (parallel IGSCR). The parallel algorithm is motivated by a demonstrated need for more computing power driven by the increasing size of remote sensing datasets due to higher resolution sensors, larger study regions, and the like. Even with a fast algorithm such as pIGSCR, the reduction of dimension in a dataset is desirable in order to decrease the processing time further and possibly improve overall classification accuracy. pIGSCR was developed to produce fast and portable code using Fortran 95, OpenMP, and the Hierarchical Data Format version 5 (HDF5) and accompanying data access library. The applicability of the faster pIGSCR algorithm is demonstrated by classifying Landsat data covering most of Virginia, USA into forest and non-forest classes with approximately 90 percent accuracy. Parallel results are given using the SGI Altix 3300 shared memory computer and the SGI Altix 3700 with as many as 64 processors reaching speedups of almost 77. This fast algorithm allows an analyst to perform and assess multiple classifications to refine parameters. As an example, pIGSCR was used for a factorial analysis consisting of 42 classifications of a 1.2 gigabyte image to select the number of initial classes (70) and class purity (70%) used for the remaining two images. A feature selection or reduction method may be appropriate for a specific lassification method depending on the properties and training required for the classification method, or an alternative band selection method may be derived based on the classification method itself. This thesis introduces a feature reduction method based on the singular value decomposition (SVD). This feature reduction technique was applied to training data from two multitemporal datasets of Landsat TM/ETM+ imagery acquired over a forested area in Virginia, USA and Rondonia, Brazil. Subsequent parallel iterative guided spectral class rejection (pIGSCR) forest/non-forest classifications were performed to determine the quality of the feature reduction. The classifications of the Virginia data were five times faster using SVD based feature reduction without affecting the classification accuracy. Feature reduction using the SVD was also compared to feature reduction using principal components analysis (PCA). The highest average accuracies for the Virginia dataset (88.34%) and for the Amazon dataset (93.31%) were achieved using the SVD. The results presented here indicate that SVD based feature reduction can produce statistically significantly better classifications than PCA. / Master of Science

A decomposition methodology based on the concept of "thermoeconomic isolation" applied to the synthesis/design and operational optimization of a stationary cogeneration proton exchange membrane fuel cell (PEMFC) based total energy system (TES) for residential/commercial applications is the focus of this work. A number of different configurations for the fuel cell based TES were considered. The most promising set based on an energy integration analysis of candidate configurations was developed and detailed thermodynamic, kinetic, geometric, and economic models at both design and off-design were formulated and implemented. A decomposition strategy called Iterative Local-Global Optimization (ILGO) developed by Muñoz and von Spakovsky was then applied to the synthesis/design and operational optimization of the fuel cell based TES. This decomposition strategy is the first to successfully closely approach the theoretical condition of "thermoeconomic isolation" when applied to highly complex, non-linear systems. This contrasts with past attempts to approach this condition, all of which were applied to very simple systems under very special and restricted conditions such as those requiring linearity in the models and strictly local decision variables. This is a major advance in decomposition and has now been successfully applied to a number of highly complex and dynamic transportation and stationary systems. This thesis work presents the detailed results from one such application. / Master of Science