I examined the abundance, composition, and distribution of 34 fishes within the first 24 km below Philpott Dam on the Smith River, a hydropeaking system in Virginia. Fish were sampled at 12 sites over 8 time periods ranging from 2000 to 2002 across April, June, and October. I evaluated spatial and temporal change in fish community characteristics. Species demonstrated persistent trends in abundance, diversity, and composition throughout the duration of the study. Fish abundance and diversity generally increased with increasing distance from the dam. Fish composition changed minimally across seasons and years, indicating consistent fish assemblages. Distributional patterns suggested a strong response to thermal gradients and presence of tributaries. I concluded that temperature and tributary location directly influence fish community patterns in the Smith River and that the patterns are persistent over space and time. I characterized spawning microhabitat use and availability, and tested transferability of spawning microhabitat criteria for Etheostoma flabellare and Nocomis leptocephalus, two of the most common species in the Smith River. E. flabellare preferentially selected small and large cobble size rocks for their spawning rock. N. leptocephalus selected areas with slower demersal and mean water column velocities in which to build their mounds. Transferability tests were conducted using spawning microhabitat criteria from the unregulated Roanoke River (Smith 1999). The logistic regression model developed for E. flabellare by Smith (1999), using information on the diameter of the spawning rock, silt, and embeddedness, transferred with most success with over half of the spawning sites and available sites correctly classified in the Smith River. / Master of Science

A Circular Depletion Sampler (the CDS) was constructed in order to collect removal data for estimating benthic sample populations and species catchability. The removal sampling method was compared to Modified Hess, Surber, and Kicknet sampling methods, and gave consistently higher and less variable estimates than other methods. Taxa showing the lowest catch abilities included the case-building Trichoptera, the Mollusca, and the Annelida. The highest catch abilities were shown by the smaller Chironomidae and the nonburrowing Ephemeroptera. Taxa with high catchability generally showed high emigrability for Surber and Kicknet methods, particularly the aquatic Insecta. The determination of species richness and evenness diversity was significantly biased by low catchabilities or emigration or both with Modified Hess, Surber, and Kicknet sampling methods. Removal samples have a higher probability of representing rare species, and as expected, pooled removal collections accumulated species at a higher rate than the pooled collections of other methods. The dominance diversity of pooled CDS collections became stable more rapidly than those of other methods, and gave higher resolution between benthic communities of different structure. DPLETE, a computer program for making removal population estimates by the method of Carle and Strub (1976) and making dominance diversity estimates based on information theory is presented, and is suitable for use in any study where removal data has been collected. / Master of Science

The structure and function of microbial communities are critical to the maintenance and sustainability of terrestrial ecosystem processes. Consequently, there is substantial interest in assessing how microbial communities respond to various land management practices, and if alterations to the characteristics of microbial communities has the potential to disrupt ecosystem processes. This thesis was conducted to identify the long term effects of fertilisation and different levels of post-harvest organic matter removal on the characteristics of the FH litter and soil microbial communities in six, second rotation Pinus radiata plantation forests located around New Zealand. The six sites, established between 1986 and 1994, were sampled in 2002 and 2003. Various physical and chemical properties of the sites were measured, and litterfall production was determined. The microbial biomass in the FH litter layer and soil was determined by chloroform fumigation-extraction, and Biolog plates were used to assess the relative differences in microbial community diversity, based on patterns of substrate utilisation. Fertilisation substantially altered the physical and chemical properties of the forest floor, including FH litter moisture content, mass, carbon content, nitrogen content and carbon: nitrogen ratio and soil pH, nitrogen content and carbon: nitrogen ratio. The same range of FH litter and soil properties were also significantly changed by different levels of organic matter removal. The biomass and diversity of the FH litter and soil microbial communities were significantly altered by fertilisation and organic matter removal, and the differences in the microbial community characteristics were significantly correlated to the effects of the fertilisation and organic matter removal treatments on the physical and chemical environment in the majority of cases. The physical and chemical properties of the sites were significantly correlated to estimates of wood production, and it was also found that the characteristics of the microbial community were strongly related to productivity at several sites. The results demonstrated that fertilisation and organic matter removal regimes have had long term effects on the microbial communities at the sites. The persistence of the effects of the organic matter removal treatments were particularly noteworthy, as these treatments were applied at site establishment, and despite no subsequent reinforcement over the life of the trials, were still substantially influencing the physical, chemical and microbiological properties of the FH litter and soil up to 17 years later. The results of this thesis also emphasised the value of long-term experiments in assessing the effects of disturbance on the physical, chemical and microbiological characteristics of forest ecosystems. Further research into the specific nature of the relationship between site productivity and microbial community characteristics was suggested as an important focus for future studies.

The gut microbiome is a community of closely interacting microbes living in the gastrointestinal tract. Its structure has direct relevance to health. Disturbances to the microbiome, such as due to antibiotic use, have been implicated in various diseases. The goal of this study was to determine how the gut microbiome reacts to and recovers from disturbance caused by antibiotics. Because diet also influences the microbiome, this study included the interaction between diet and antibiotics. Half of the mice in each diet treatment were given antibiotics to disturb their microbiomes. After cessation of antibiotics, mice were paired in combinations within diets to determine whether the microbiomes of control mice influenced the disturbed microbiomes of formerly antibiotic mice. Chapter 1. Diet significantly altered the structure of the gut microbiome but its effect was significantly smaller than the effect of antibiotics. There was a significant interaction between diet and antibiotics; the antibiotic effect was larger in the cornstarch diet than in the glucose diet. Dysbiotic microbiomes resulting from antibiotics were characterized by an increase in Bacteroidetes and Proteobacteria, and a decrease in Firmicutes. Antibiotic administration also resulted in an initial increase OTU diversity, mainly because it reduced the abundance of dominant OTUs, resulting in greater evenness. Chapter 2. Seven weeks after the cessation of antibiotics (experiment termination), the effect of the antibiotics on the microbiome was still evident. The structure of the dysbiotic microbiome had not returned to that of control mice. Antibiotics significantly increased the relative abundance of some taxa and significant decreased the relative abundance of others. It was unexpected that the taxonomic hierarchy within the microbiome did not recover after 7 weeks following cessation of antibiotics. It would appear, therefore, that antibiotics established a new, semi-stable hierarchy. Chapter 3. When paired together, the assumption was that dysbiotic microbiomes of antibiotic mice would be positively influenced by microbiomes of control mice, based on the assumption that the control mouse would act as a probiotic for the antibiotic mouse, either via coprophagy or consumption of food contaminated by feces. Contrary to that hypothesis, the microbiomes of control mice became more similar to that of antibiotic mice. One can offer at least two hypotheses to explain this result, but neither was tested. First, compared to the control microbiome, the dysbiotic microbiome may have been more stable and thus more resistant to change due to invasion by OTUs from the control microbiome. Other research has shown that dysbiotic microbiomes have a high degree of stability. If this were true, the use of probiotics is questionable. Second, one or more of the antibiotics could still have been active at the initial phase of pairing, and coprophagy caused the microbiome of the control mice to rapidly become dysbiotic. If this is true, the experiment should have been conducted with a waiting period between the cessation of antibiotic administration and pairing.

In 2011, several wild North American bumble bee pollinator species were reported to have declined by up to 96% in relative abundance in comparison to historic estimates, and one species was speculated to be extinct. None of these species have yet been documented to have recovered from these declines and additional species are now suggested to be at risk. Imperiled species in particular show increased specificity to narrow climatic envelopes, as opposed to putatively stable species. My dissertation describes patterns of population genetic diversity, structure, and gene flow pathways associated with climate variation and historical biogeography of bumble bees distributed in western North America. The results of my dissertation research suggests that (1) historic climate variability predicts contemporary patterns of population genetic structure and divergence in an economically important species, (2) color variability in bumble bees is likely associated with lineage diversification and phylogeography, (3) bumble bee community structure across evolutionary time is likely driven by Müllerian mimicry at narrow spatial scales, and (4) bumble bees inhabiting specialized ecological niches are associated with high levels of genetic fixation at regional spatial scales in the Pacific Northwest. The results of my research directly contribute to current efforts to effectively manage, conserve, and advocate for wild bumble bee pollinators in the context of global change.

Density, species diversity, species richness, and evenness of bivalve mollusks were measured in the deep (0.2km to 3.7km) northern Gulf of Mexico to describe the community structure of benthic bivalve mollusks. Density decreased gradually from shallow continental slope depths, with remarkably high values in the Mississippi canyon, to the deepest sites. Diversity of bivalve mollusks increased from shallow continental slope depths, with low values in the Mississippi canyon, to a maximum at intermediate depths (1-2km), followed by a decrease down to the deepest locations (3.7km). Nine distinct groups were formed on the basis of the similarity in species composition. The pattern varied more abruptly on the slope compared to the deeper depths, possibly due to steeper gradients in physical variables. ANOVA indicated that the density of bivalve mollusks was not significantly different at different depths, was not significantly different on different transects, was not significantly different between basin and non-basin, but was significantly different in canyon and non-canyon locations. Similar distinctions were observed in diversity, except that basins were lower than non-basins. The patterns observed reflect the intense elevated input of terrigenous sediments accompanied by high surface-water plankton production from the Mississippi River to the north central gulf.

The present study used beam trawl net to collect crabs by Ocean Researcher III at four sites, Wanggong, Taisi, Cigu and Jiading, in western waters off Taiwan from July 2006 to October 2008. In total, 3205 crabs were obtained, including 11 families, 22 genus, 41 species and 6 unknown species. Among these families, Portunidae contained the most species (21 species, 45¢H), followed by families Majidae (4 species, 9%), Leucosiidae (3 species, 6%) and Dorippidae (3 species, 6%). Six dominant species were Portunus hastatoides (60.1%), Portunus argentatus (13.7%), Portunus sanguinolentus (7.2%), Matuta victor (3.8%), Charybdis truncata (3.0%) and Portunus pelagicus (2.1%) contributing 89.8% of the total catch. P. sanguinolentus (35.6%), P. pelagicus (16.3%), P. hastatoides (12.0%), Calappa philargius (11.5%), P. argentatus (8.0%) and Charybdis feriatus (3.1 %) were the six dominant species (86.5%) by biomass. Spatial distributions of the benthic crabs by number were significantly different. P. argentatus was the most dominant species in Cigu, whereas in the other three stations was P. hastatoide. Among the six species, P. sanguinolentus, P. pelagicus and C. feriatus are the most important species in terms of economic value. The former appeared in large numbers in Jiading and composed 35.6% in weight of the total catch from all sites, while the latter frequently occurred in the other three sites and contributed 16.3% to the overall sample weight. Moreover, P. sanguinolentus was the dominant species in terms of weight in Jiading, but the rest of three sites dominated by P. pelagicus. P. hastatoides occurred mostly in the substrate with very fine sand. P. sanguinolentus increased in number and weight with decreasing of salinity and water depth. P. argentatus had a narrowest optimum temperature range from 24.3 to 28.3¢J.

Macrobenthos of the deep, northern Gulf of Mexico have been sampled with large box cores along multiple cross-depth transects extending from depths of 200 m out to 3700 m. Four major depth zones have been identified based on the faunal similarities (beta diversity) between geographic sites, with the two intermediate-depth zones being divided horizontally down the middle of the basin. The input of food resources appears to control the observed patterns. Each zone and sub-zone can be described by a characteristic animal density, biomass and biodiversity (alpha diversity). Highest densities and biomass occurred in two large submarine canyons, the Mississippi and De Soto Canyon, but the two habitats are markedly different. The alpha diversity displays an intermediate depth maximum. Species richness (gamma diversity) is highest on east mid-slope, due, we suggest, to habitat complexity, but alpha diversity is lowest at the canyon head due to extreme dominance by amphipods. Small mean individual size and low densities encountered are a reflection of the meager surface water primary production, albeit with exceptional isolated habitats in which detrital material is concentrated, such as canyons on the upper continental slope.

Density, species diversity, species richness, and evenness of bivalve mollusks were measured in the deep (0.2km to 3.7km) northern Gulf of Mexico to describe the community structure of benthic bivalve mollusks. Density decreased gradually from shallow continental slope depths, with remarkably high values in the Mississippi canyon, to the deepest sites. Diversity of bivalve mollusks increased from shallow continental slope depths, with low values in the Mississippi canyon, to a maximum at intermediate depths (1-2km), followed by a decrease down to the deepest locations (3.7km). Nine distinct groups were formed on the basis of the similarity in species composition. The pattern varied more abruptly on the slope compared to the deeper depths, possibly due to steeper gradients in physical variables. ANOVA indicated that the density of bivalve mollusks was not significantly different at different depths, was not significantly different on different transects, was not significantly different between basin and non-basin, but was significantly different in canyon and non-canyon locations. Similar distinctions were observed in diversity, except that basins were lower than non-basins. The patterns observed reflect the intense elevated input of terrigenous sediments accompanied by high surface-water plankton production from the Mississippi River to the north central gulf.

The structure and function of microbial communities are critical to the maintenance and sustainability of terrestrial ecosystem processes. Consequently, there is substantial interest in assessing how microbial communities respond to various land management practices, and if alterations to the characteristics of microbial communities has the potential to disrupt ecosystem processes. This thesis was conducted to identify the long term effects of fertilisation and different levels of post-harvest organic matter removal on the characteristics of the FH litter and soil microbial communities in six, second rotation Pinus radiata plantation forests located around New Zealand. The six sites, established between 1986 and 1994, were sampled in 2002 and 2003. Various physical and chemical properties of the sites were measured, and litterfall production was determined. The microbial biomass in the FH litter layer and soil was determined by chloroform fumigation-extraction, and Biolog plates were used to assess the relative differences in microbial community diversity, based on patterns of substrate utilisation. Fertilisation substantially altered the physical and chemical properties of the forest floor, including FH litter moisture content, mass, carbon content, nitrogen content and carbon: nitrogen ratio and soil pH, nitrogen content and carbon: nitrogen ratio. The same range of FH litter and soil properties were also significantly changed by different levels of organic matter removal. The biomass and diversity of the FH litter and soil microbial communities were significantly altered by fertilisation and organic matter removal, and the differences in the microbial community characteristics were significantly correlated to the effects of the fertilisation and organic matter removal treatments on the physical and chemical environment in the majority of cases. The physical and chemical properties of the sites were significantly correlated to estimates of wood production, and it was also found that the characteristics of the microbial community were strongly related to productivity at several sites. The results demonstrated that fertilisation and organic matter removal regimes have had long term effects on the microbial communities at the sites. The persistence of the effects of the organic matter removal treatments were particularly noteworthy, as these treatments were applied at site establishment, and despite no subsequent reinforcement over the life of the trials, were still substantially influencing the physical, chemical and microbiological properties of the FH litter and soil up to 17 years later. The results of this thesis also emphasised the value of long-term experiments in assessing the effects of disturbance on the physical, chemical and microbiological characteristics of forest ecosystems. Further research into the specific nature of the relationship between site productivity and microbial community characteristics was suggested as an important focus for future studies.