Search

The central theme of my dissertation is the influence of spatial structure and trophic interactions on ecological community dynamics and subsequent patterns of community diversity. I studied this phenomenon in different ecological context. In a synthesis of protist microcosm experiments, I studied how spatial network structure influenced predator-prey persistence (chapter 1). Subsequently, I shifted my focus to Sierra Nevada, CA high elevation aquatic systems. In high elevation lake-stream networks, I investigated whether patterns of macroinvertebrate diversity matched predictions from established stream ecology frameworks and the underlying processes driving these diversity patterns (chapter 2). Lastly, in order to understand how spatial structure and trophic interactions interactively structure communities, I studied how macroinvertebrate and zooplankton communities in Sierra Nevada streams and lakes are structured along spatial gradients and as a function of fish (chapter 3).

Acid rain alters freshwater pH and ion composition, preventing organisms from performing essential bodily functions causing mortality. Macroinvertebrate communities in acidified streams are characterized by species loss in response to physiological stress and altered food quality resulting from the degradation of microbial (e.g., fungi on leaves) communities. Although freshwater acidification in the U.S. is lessening following reduced industrial emissions, little is known about macroinvertebrate recovery. Often, biotic recovery is assessed by looking at changes to what taxa and how many individuals are present in the community (e.g., richness, density). While providing a metric for change, changes in "who" is there (i.e., richness) doesn't necessarily tell us changes in "what" they are doing (i.e., function). The relationship between diversity and function requires linking a "who" to their "what" with direct measurements or as indicated by their traits. Traits are attributes of an organism that aggregate biological, morphological, and behavioral information and may relate to their success in a particular environment. For example, taxa that cannot survive with stream drying (not desiccation resistant) may only be found in streams with permanent water. Trait-based taxonomic metrics could bridge "who" and "what" and expand the impact of stream recovery assessments. My objective was to assess trends over time in water chemistry and macroinvertebrate taxonomic and trait richness and density following reduced industrial emissions. To do so, I studied two long-term data sets from Shenandoah National Park to assess trends in water chemistry and macroinvertebrate taxa and trait composition over a 30-year period to identify taxa and traits that are sensitive to acidification. I also measured how much biomass macroinvertebrates produced in a year (i.e., secondary production) in two streams (1 acidified; 1 not acidified) to determine taxa and traits that are functionally sensitive to acidification. I used these structural and functional measures of sensitivity to determine if changes in trait richness or density predict changes in the function of that trait (e.g., secondary production). Changes over time show that streams have some recovery from acid rain with increasing stream pH and a greater number of taxa and traits present in the community. Changes in taxa were greater than changes in traits over time. While this result was expected as multiple taxa make up each trait category, it may also suggest minimal or delayed functional recovery over time. Still, macroinvertebrate secondary production indicated that function did differ with differences in acidification. Therefore, observed small changes in traits over time mirror prior studies that found other variables, such as competition for food or space, delay or inhibit macroinvertebrates from returning to the recovering streams. Additionally, there were similarities between traits changing over time and the secondary production of traits that differed between more and less acidified streams. Taxa characterized by long life spans and large body size (e.g., semivoltine, long adult life, slow seasonal development) appeared to be the most sensitive to changes in acidification. These findings suggest that some compositional attributes, like taxonomic or trait richness, may predict functional changes measured as secondary production while others, such as density, do not.

Community structure and ecosystem function are often influenced by both food web interactions (e.g., competition, predation) and environmental factors (e.g., temperature, oxygen availability). Identifying the drivers of food web interactions and their response to environmental change can therefore shed insights into how communities and ecosystems are regulated, which in turn can benefit the management of valued ecosystem services (e.g., fisheries). Food web interactions in north-temperate reservoirs, which are the focal ecosystems of this research, are thought to be driven primarily by a planktivorous fish, gizzard shad (Dorosoma cepedianum), by its ability to regulate the abundance of organisms at both higher and lower trophic levels. This current conceptualization, however, overlooks the potential role of bottom hypoxia, which is typically extensive during of the spring through summer growing season and can shift the spatial distributions of organisms and subsequently competitive and predator-prey interactions. It also overlooks the role of the vertically migrating macroinvertebrate, Chaoborus, which is tolerant of bottom hypoxia, and highly abundant in north-temperate reservoirs. Large populations of Chaoborus have been shown to limit zooplankton availability to other intermediate consumers (e.g., planktivorous fish) in other ecosystems, and thus have the potential to also influence food web interactions and fish recruitment in north-temperate reservoirs.Currently, the effects of both Chaoborus and bottom hypoxia on food web interactions in north-temperate reservoirs are largely unknown. To better understand the factors that influence food web interactions, energy flow through the ecosystem, and the processes that limit sport fish recruitment, my collaborators and I set out to determine how hypoxia and Chaoborus alter food web structure, function, and dynamics in north-temperate reservoirs. Specifically, we sought to answer: 1) Does north-temperate reservoir food web structure differ between periods of normoxia to hypoxia?; and 2) How do hypoxia and intermediate consumers (i.e., gizzard shad and Chaoborus) combine to influence prey availability to zooplanktivorous fishes? To answer these questions, we developed field studies aimed at improving methods for quantifying food web structure (Chapters 2 and 3), determining how the spatiotemporal distributions of organisms in the food web vary from normoxia to hypoxia (Chapter 4), and evaluating food web interactions from normoxia to hypoxia, with particular attention paid to understanding the potential for Chaoborus to limit zooplankton availability to other secondary consumers such as larval sport fishes (Chapter 5). This research led to many novel insights. First, we found that hydroacoustic estimates of the conventionally accepted driver of food web interactions in north-temperate reservoirs, gizzard shad, were biased high, owing to Chaoborus migrating into the water column and being misconstrued as gizzard shad (Chapter 2). Second, our ability to improve the use of hydroacoustics to estimate gizzard shad (Chapter 3), and simultaneously develop a method to estimate Chaoborus abundance (Chapter 2), allowed us to determine that the spatiotemporal distributions of zooplanktivores changed from normoxia to hypoxia, leading to especially high nighttime spatial overlap among gizzard shad, Chaoborus, and their zooplankton prey when bottom hypoxia is present (Chapter 4). This finding suggests the potential for alternative drivers (besides gizzard shad) of zooplankton prey availability during summer. Finally, our evaluation of food web interactions during normoxia and hypoxia showed that Chaoborus can at times potentially limit zooplankton availability to other fish consumers during the summer hypoxic period, and therefore should also be considered as a driver of food web interactions in north-temperate reservoirs (Chapter 5). Collectively, our findings provide important and novel insights into food web interactions in north-temperate reservoirs, identifying previously unconsidered processes (Chaoborus and bottom hypoxia) that can potentially regulate energy flow and fish recruitment in these ecosystems.

Streams in southeastern Ohio are affected by acid mine drainage (AMD) as a result of historical and present-day mining; effluents from mining sources contribute to elevated acidity and metals loadings within impaired streams. Remediation efforts in the area have been successful in lowering acid and metals concentrations in impaired streams, and biological recovery has been tracked and documented for the past decade. Macroinvertebrate communities in these streams are also influenced by other factors, including hydrological conditions, especially floods and droughts. These two stressors (AMD and hydrology) can have interactive affects that could harm macroinvertebrate communities within impaired streams to varying degrees. I analyzed seven years of data from macroinvertebrate communities within AMD impaired streams, streams that have recovered as a result of remediation projects, and reference unimpaired streams to detect any patterns in variation related to AMD condition and hydrological regime. I found that impaired sites supported fewer macroinvertebrate taxa; abundances were markedly lower. The greatest seasonal variation in AMD chemical stressor intensity occurred at impaired sites during low flow in summer and fall. Recovered and unimpaired sites also experienced lower pH and increased conductivity at low flow, but conditions were not as severe. Impaired stream sites that experienced a wide range of chemical variation also had larger coefficients of variation around annual biotic indices. Recovered and unimpaired reference sites exhibited less chemical variation, and the macroinvertebrate communities had greater persistence and stability than impaired sites. Stability and persistence of the macroinvertebrate community were influenced more by AMD than hydrological events. Abundance and richness were significantly higher in hydrologically wet years than in mid-range and drought years. Drought did not appear to exacerbate AMD effects.Episodic events of increased acidity or metals occur in other AMD impaired streams across the globe, so I recorded continuous chemical data in sites along a gradient of impairment to detect any episodic events and to understand the overall relationship between discharge and chemical concentration. No episodic events were detected; in most instances, dilution mechanisms were driving the AMD chemistry-discharge relationship. I also examined the interactive effect of seasonal and annual hydrology and AMD on macroinvertebrate community compositions. Impaired communities exhibited a wide range of variation in community composition, while composition in recovered and unimpaired communities was significantly less variable. AMD strongly structured the communities, while hydrological conditions appeared to have less effect. I was unable to identify any indicator taxa of specific AMD or hydrological conditions, but this may be due to additional factors such as other land-use and nutrient concentrations. To understand the discrete effects of a flood event on AMD impaired, recovered, and unimpaired macroinvertebrate and periphyton communities, I experimentally disturbed the stream substrate and tracked biotic recovery for 30 days. I found that both macroinvertebrate and periphyton communities were resistant and resilient to the disturbance across all AMD conditions, though macroinvertebrate abundance in recovered and unimpaired sites were faster to recover than impaired sites. Periphyton communities in recovered streams were slightly but significantly slower to recover than those in unimpaired streams, suggesting there may be some occasional chemical stress in recovered streams that affects the periphyton, but not the macroinvertebrates. The results of my dissertation suggest that AMD strongly structures the macroinvertebrate communities, while communities are usually robust against any interactions between AMD and hydrological disturbances. Communities within impaired streams have a wide range of variation, most likely due to more sensitive species entering but not permanently colonizing these streams. Recovered communities are similar to unimpaired communities most of the time, indicating that chemical concentrations are not harmful to the stream inhabitants, even in extremely low- or high-discharge conditions.

My study centered on a bioindicator songbird, the Louisiana Waterthrush (Parkesia motacilla), hereafter waterthrush, an organism that co-occurs in both forested and aquatic habitat across the aquatic-terrestrial interface. This enabled the opportunity to quantify demographic, spatial, and epigenetic (i.e., DNA methylation) responses in a highly forested watershed of the Central Appalachians, the areas that have undergone the most rapid transformations over the last decade from unconventional shale gas development and activities. I organized my dissertation into 4 parts (Part 1: Introduction, Part 2: Louisiana Waterthrush Demography, Part 3: Spatial Assessment of Louisiana Waterthrush Foraging, Part 4: Louisiana Waterthrush Molecular Ecology) including 6 chapters that indicate multiple biotic and abiotic factors interacted with or were altered by shale gas development resulting in atypical, negative disturbances that drove a steep decline in a waterthrush population in West Virginia. Part 1 includes Chapter 1 and is an introduction to my dissertation. I introduce the reader to the rationale for my study, the focal species, research objectives, and the study area. I also mention some limitations to my study that can be considered in any future research endeavors. Part 2 comprises Chapters 2–3 which are a comprehensive examination of demographic parameters over a six-year period (2009–2011, 2013–2015). In Chapter 2, I examined demographic response to shale gas development for nest abandonment, nest survival, nest productivity, a source-sink threshold, riparian habitat quality, and territory density and length. Nest productivity was lower in areas disturbed by shale gas where a source–sink threshold suggested these areas were more at risk of being sink habitat. Overall results suggest a decline in waterthrush site quality as shale gas development increased. In Chapter 3, I focused on first-year return rates (site fidelity), site fidelity factors, and apparent survival. I related natal fidelity and pairing rates to territory density, and also compared # of breeding attempts between return and non-returning females with and without territory shale gas disturbance. The study identified potential conflicts between factors that influence adult survival and site fidelity that may affect long-term population persistence. Part 3 includes Chapters 4–5 and focuses on utilizing and accounting for spatial properties intrinsic to stream ecosystems to make informed decisions regarding waterthrush foraging. Chapter 4 was a follow-up to a waterthrush aquatic prey study at our site in 2011 that suggested shale gas development negatively affected waterthrush demography from alterations in their aquatic prey at a watershed scale. During 2013–2014, I quantified waterthrush demographic response and nest survival in relation to potential changes in its aquatic prey due to shale gas development. I utilized spatial generalized linear mixed models that accounted for both spatial and non-spatial sources of variability. I found waterthrush aquatic prey was negatively affected by shale gas development at the nest and territory level, and that there may be a disturbance threshold at which waterthrush can no longer adapt and respond negatively to changes in its aquatic prey. In Chapter 5, I used spatial stream network models (SSNMs) to explore relationships among the waterthrush, stream channel and monitoring data, and the aquatic prey of the waterthrush. I compared the spatial models to traditional regression models to see which ones performed best. We sampled aquatic prey in waterthrush territories and collected wetted perimeter stream channel and water chemistry data along a 50m fixed point stream grid that mapped the foraging substrate or stream channel where waterthrush forage. By relating foraging observations and data collected to the stream grid, I was able to develop a foraging probability index that determined what conditions or variables create or affect ideal foraging locations. Spatial models outperformed traditional regression models and made a statistical difference in whether stream covariates of interest were considered relatable to waterthrush foraging. My study also indicated waterthrush forage in areas of higher biotic stream integrity. Lastly, Part 4 includes Chapter 6 where I examined epigenetic modifications. These are alterations to genes without changing the gene sequence and can be thought of as an evolutionary "soft" inheritance of gene expression that can either be adaptive or maladaptive for the individual. DNA methylation is one type of epigenetic modification that may vary in response to environmental stressors. We examined the association between DNA methylation and demographic characteristics in addition to potential differential methylation from shale gas development. There was differential methylation for demographic characteristics as well as for adult males between shale gas undisturbed and disturbed areas. Barium (Ba) and strontium (Sr) data were collected in 2013 feather samples where adult males had fewer methylated sites at higher concentrations of Ba and Sr, while nestlings displayed no correlation of methylation to Ba and Sr concentrations. Females displayed increased methylation with increased Ba and Sr, a trend reflected in adult female recaptures. Overall, results of our study suggest sex-specific influences of shale gas development on gene expression that may affect long-term population survival and fitness.

Biomonitors are a popular tool to determine biological impairment of an ecosystem. Body burdens (i.e., measuring excess pollutants assimilated in organisms) are a novel tool that could contribute to detection and identification of impairment, but this tool requires more validation in ecosystems with complex stressors. For a study site, I selected the Cuyahoga River watershed, and the 23 sample sites in 8 streams span a region from heavily impaired to unimpaired streams. I sampled metals (Cd, Cu, Fe, Pb, & Zn) in surface water, suspended solids, and Trichopteran bodies in urban streams to determine if Trichopteran body burdens were useful as predictors of impaired macroinvertebrate communities or useful to identify the cause of impairment. I evaluated the status of impairment in the region against the Environmental Protection Agency’s (EPA) statement of metal impairment. Our final redundancy models using Ohio EPA Invertebrate Community Index (ICI) data were not significant in predicting Trichopteran body burdens or macroinvertebrate community assemblages, which suggests that our macroinvertebrate community do not reflect the EPA definition of metal impairment in the region. Our final models predicting the measured community abundances (redundancy analysis with forward stepping) suggest that Trichopteran body burdens are worse predictors of macroinvertebrate abundance than environmental pollutants (17% and 28% of variance explained, respectively). However, in our models Trichopteran body burdens were able to explain shifts in abundance for different macroinvertebrate taxa than changes predicted by environmental variables. In conclusion, Trichopteran body burdens were not found to be better predictors of the macroinvertebrate community than environmental samples, but can play a supporting role in identifying impairment for a portion of the macroinvertebrate community not accounted for from environmental models and identifying the source of impairment in a more biologically relevant manner that environmental models.

Invasive species are of global importance due to their impacts on ecological communities, habitat structure, native community dynamics, and ecosystem processes. Scientists and conservation managers are increasingly focusing on the biological impacts of invasive species and devising management practices that emphasize the health of ecosystems based on measured biological processes. Lonicera maackii is a highly successful invasive shrub in forests of eastern and Midwestern North America. We investigated how riparian invasion of L. maackii influenced (1) the availability of in-stream leaf litter resources, algal growth, above stream canopy cover, and light available to the stream, (2) the functional and taxonomic diversity and community composition of aquatic macroinvertebrate communities, (3) the effects of L. maackii on throughfall chemistry. In summary, the removal of an invasive riparian shrub influenced the timing, deposition, quality, and abundance of leaf litter habitat into a headwater stream, ostensibly driving bottom-up effects on aquatic primary producer biomass and the macroinvertebrate community. Patterns in macroinvertebrate community and functional trait dynamics were influenced by seasons and the L. maackii riparian forest. These findings suggest that functional traits were driven by life history strategies linked with seasonal patterns in temperature and food resources that are also influenced by L. maackii riparian forests. In addition, riparian L. maackii has the potential to alter nutrient subsidies during rain events that enter aquatic systems as throughfall, and suppress stream algal growth early in the growth season, impacting nutrient cross-system subsidies and one of the basal food resources in aquatic systems. Based on these findings we have developed a predictive framework for understanding how this terrestrial invasive shrub influences aquatic ecosystems.

Many streams throughout the United States are negatively impacted by excess fine sediments (sand, silt, and clay). Benthic macroinvertebrates are a commonly-used tool to assess stream condition; however, current methodologies typically are not able to distinguish among stressors. Previous studies have correlated macroinvertebrate communities and traits with excess fine sediments, demonstrating that aquatic macroinvertebrates are sensitive to deposited fine sediment and the assemblages will shift in response. Western Lake Superior streams have a wide range of fine sediment amounts due to clay and sand soils, but have low amounts of other stressors, and thus are a good region to investigate relationships between macroinvertebrate traits and fine sediments. Data were collected from 22 stream sites located along the north shore of Lake Superior in 2010. The data collected in 2010 did not have the desired gradient of fine sediment due to wet conditions that year; therefore, the data were supplemented with data collected by NRRI personnel in earlier years (1997 – 2008). The five sediment stressors used in analyses included percent embeddedness, depth of fine sediments, total percent fine sediments, percent sand, and a combined sediment index created using normalized and transformed embeddedness, depth of fine sediments and total percent fine sediments. Fifty-seven specific taxonomic groups and macroinvertebrate physical and behavioral characteristics (traits) were tested as potential response metrics in linear regressions. In addition, TITAN analyses were used to look for thresholds or sediment stressor values at which a taxon increases greatly, decreases greatly, or disappears from a community. Both the linear regressions and TITAN analyses showed a change in the community structure under conditions of excess sediment in the form of embeddedness, total fines, depth of fines, and/or the combined sediment index. The TITAN analyses also showed a change in the community structure due to increasing proportion sand in the streambed. Furthermore, the analyses identified potential characteristics that may specifically make a particular macroinvertebrate more or less vulnerable to excess fine sediments.

Lime doser treatment for acid mine drainage (AMD) is often used in areas with insufficient space for passive treatment systems and in rural areas where more complex treatment systems would be impractical. In this study, assessment of pH, conductivity, sulfate, iron, manganese and aluminum was conducted for 8-11 miles downstream of each doser. Grain size profile and trace metal concentration in stream sediments were also analyzed. Doser downstream reaches’ field parameters, water quality profile and sediment were analyzed with R application and compared to assess differences in biological recovery in the four watersheds. Results of the study showed that precipitation of dissolved metals, high alkalinity content, high average grain size of sediment downstream of the doser treatment led to better biological improvement. Model output of multivariate regression showed strong relationship between biological score and iron and alkalinity. The adjusted R2 was 0.47. The parameter linkage in the streams in the different watersheds shows that the influential factors are applicable to many different systems to enhance biological score improvement.

Land-use practices have long been associated with alterations in stream ecosystem structure and function, however, 'exurbanization' and its impact on streams is poorly understood. This study compares the ecosystem structure and function of 9 southern Appalachian streams of differing land-use (forested, exurban, agricultural). Impacts of land-use on leaf breakdown are examined in Chapter 1. Leaf breakdown rates were significantly related to land-use. Forested streams exhibited the slowest breakdown rates, followed by exurban streams, with agricultural streams having the fastest rates. Leaf breakdown was most strongly related to discharge (white oak) and some fine sediment metrics (red maple). Our results suggest that the altered hydrological regimes in agricultural streams, as well as the influx of fine sediments into streams from exurban development, can play a role in altering in-stream organic matter processing. The taxa and number of shredders present may play a role to a lesser extent. Impacts of land-use on macroinvertebrate assemblages are examined in Chapter 2. Shannon diversity, %EPT, and NCBI were significantly related to land-use regime. There were significant negative relationships between macroinvertebrate diversity and conductivity and temperature. In addition, biotic integrity had a significant negative relationship with conductivity. Canonical Correspondence Analysis (CCA) showed that agricultural streams were characterized by temperature and flow, forested streams by MPS and standing stock course particulate organic matter (SSCPOM), and two of three exurban streams by conductivity and temperature. Principal Coordinates Analysis (PCoA) revealed that while macroinvertebrate communities overlapped, some differences in community assemblage could be seen between land-use types.

Abstract: Resource development is expanding in Canada, particularly in the Arctic. In Canada, damage to stream ecosystems as a result of development requires habitat restoration or compensation measures. A compensation project, focused on improving ecosystem connectivity and aquatic habitat for fish within three small fish-bearing Barrenland lakes and their outlet streams, was conducted in the Lac de Gras, NWT watershed as a result of diamond mine development at Diavik Diamond Mine Inc. Habitat manipulations to the three outlet streams used two general fishway designs: gabion-style pool-weir and nature-like choke-pool structures. I used a Before-After-Control-Impact (BACI) design to assess stream habitat attributes, their ecosystem functions, and the macroinvertebrate assemblages before and after the compensation project. Many impacts resulted from the manipulations, but few were relevant to the long term ecological function of the system. The removal of riparian vegetation during manipulation construction best explains the reduction in stream organic matter, specifically course particulate organic matter (CPOM), observed at the gabion-weir treatment streams after manipulations. However, given the reasonable organic matter retention rates observed, CPOM can be expected to increase as vegetation reestablishes. For future compensation projects, it is recommended to take added measures to preserve and actively re-establish the riparian vegetation as much as possible during and after construction activities. There were many shifts in abundance of the macroinvertebrate assemblages for both treatments (fishway designs) in riffle and pool habitat, but no easily distinguished overall patterns emerged from ordinations, in part because shifts in at reference streams in riffle habitat also occurred. The observed decrease in Simuliidae at the gabion-weir treatment may be of concern as it is a potential food source for fish; however the affected stream was fishless. Although a reduction in some organic matter was observed at the gabion-weir treatment, and there were shifts in the macroinvertebrate assemblages, my research suggested that the compensation manipulations were largely successful in maintaining ecosystem structure and function, with some room for improvement.

Abandoned mines in the United States are one source of aqueous pollution due to acid mine drainage which can leave streams heavily impacted by trace metals. Middleton Run in Little Raccoon Creek sub-watershed has a very poor biological community and impacts the biological community downstream in Little Raccoon Creek. It was hypothesized that the high aluminum loadings of more than 30 mg/L from Middleton Run were the cause of biological impairment. Field studies were undertaken to characterize the biological community in Middleton Run and lab studies were performed to test the impact of aluminum chloride on a target macroinvertebrate, crayfish. Macroinvertebrate community assessment followed the Macroinvertebrate Aggregated Index for Streams method. Impairment of growth in crayfish from the ingestion of aluminum chloride was the primary focus of the lab studies. Crayfish were separated into four groups, a control group and three treatment groups, and fed food pellets with dosed with 25 mg/g, 50 mg/g, and 75 mg/g of aluminum chloride. Crayfish carapace length, total length, and mass were recorded pre-trial, at 2, 4, and 6 weeks of treatment. Analysis of the data revealed that there was no significant difference or changes in growth between the control group and those given aluminum chloride spiked pellets. It is likely that ingested aluminum is excreted in waste or accumulated in tissues without impairing growth. While the macroinvertebrate community in Middleton Run is certainly impaired by AMD, this study suggests that aluminum toxicity due to ingestion is not the mechanism of impairment.

Beaver were at the brink of extinction in the early 20th century, but over the past century populations have been rebounding. This increase in range and population size has raised new concerns over possible impacts on aquatic ecosystems. Beaver are considered ecosystem engineers because they change the physical structure of aquatic systems through the construction of dams, which turn naturally flowing sections of rivers and streams into standing water. Such changes are known to influence habitat heterogeneity, influence organic matter transport and storage, and consequentially influence biological diversity. To my knowledge, no studies of the effects of beaver impoundments on benthic invertebrate assemblages and benthic organic matter standing stocks have been conducted in the SE Coastal Plain ecoregion. Furthermore, few studies have explored the influence of beaver dams during drought conditions. The goal of this study was to assess the effects of beaver dams on organic matter standing stocks and benthic invertebrate assemblages in a Southeastern Coastal Plain stream during extreme to severe drought conditions. This goal was addressed by comparing stream reaches currently impacted by beaver activity (i.e., presence of dams) with segments where no beaver activity was evident (i.e., reference reaches), and comparing upstream and downstream segments within impacted reaches. Beaver impacted reaches retained significantly higher standing stocks of several major organic matter categories, including fine particulates (250 µm) ranged from 29-154 g ash free dry mass (AFDM) m¯² at beaver impacted reaches compared to 9-67 g AFDM m¯² at reference reaches (P=0.004), and coarse particulates (>1 mm) ranged from 473-1,540 g AFDM m¯² at impacted sites compared to 206-1335 g AFDM m¯² at reference sites (P=0.04). Furthermore, there were no differences in benthic organic matter standing stocks within impacted reaches (i.e., upstream and downstream of beaver dams). Benthic macroinvertebrate abundance or biomass did not differ between impacted reaches and reference reaches, or within impacted reaches. Taxonomic diversity as calculated by the Shannon-Wiener Index (H’) was significantly higher (P=0.03) at reference reaches (H’=1.16) over impacted reaches (H’=0.99), but did not differ within impacted reaches (P=0.30). Collector-gatherers consistently dominated macroinvertebrate assemblages accounting for more than half (65-91%) of the total community. In conclusion my study suggests that beaver activity can have an effect on the retention of organic materials in streams, with impacted sites retaining and storing significantly higher levels of benthic organic matter. In contrast with other studies, my results indicate that beaver dams had little effect on macroinvertebrate assemblages, at least during low flow conditions in SE Coastal Plain streams.

The focus of this research was the mechanistic relationship between fluid flow, spatial density, distribution, and metabolism of stream-dwelling insect larvae of the caddisfly genus Glossosoma (Glossosomatidae: Trichoptera). Glossosoma are ecologically important for many reasons, including that they are dominant grazers in lotic food webs, are capable of suppressing stream periphyton, and act as an indicator of stream health. A description is herein presented of fluid flow and stream bathymetry environments where Glossosoma are often present. Predictive relationships for Glossosoma spatial density are proposed from local fluid flow and channel bathymetry variables. The research was conducted through a series of four separate studies. First, habitat was quantified for larval Glossosoma in three coastal mountain streams in northern California. Applying dimensional analysis, a functional relationship was developed for predicting larval abundance. Variogram analysis of Glossosoma spatial density and bed topography revealed overlap in the separation distance above which point measurements were statistically independent. Second, field measurements were conducted in Valley Creek, Minnesota and data were compared with the measurements in the coastal mountain streams. Third, a computational model was employed to estimate high resolution fluid flow variables along a riffle in Valley Creek. Simulation results were verified and implemented in a predictive model for Glossosoma abundance. Fourth, a laboratory study of Glossosoma metabolism under varying flow conditions revealed generally increasing oxygen consumption with increasing fluid flow velocities. The proposed research will be instrumental for predicting not only how Glossosoma respond to changes in fluid flow and stream bathymetry conditions, but also how these variables influence larval spatial density, distribution, and behavior in lotic environments.

Vernal pools are temporary wetlands and local-scale biodiversity hot spots for a variety of amphibians, macroinvertebrates, and plants because their seasonal drying prevents the establishment of predatory fish populations. Vernal pools are often of conservation concern because of the amphibian populations; however, the emphasis on these organisms often eclipses the macroinvertebrates, which are important predators, prey, and nutrient cyclers in wetlands and the surrounding habitat. Hydroperiod and water chemistry are thought to be the primary regulators of vernal pool macroinvertebrates, but the surrounding habitat also affects these organisms. Specifically, canopy cover and forest composition can alter the autochthonous and allochthonous carbon sources for wetland food webs. My research objectives were to understand how variations in these factors affect macroinvertebrate diversity and community composition. In 2009, I conducted a field survey of fifteen vernal pools that varied in area, depth, hydroperiod, canopy and surrounding land use. I measured several habitat conditions, assessed the biotic communities of these wetlands, and found that canopy cover influenced bottom-up productivity and macroinvertebrate diversity. I used the results of this study to determine how known macroinvertebrate communities respond to variation in canopy cover in mesocosm wetlands. The low canopy treatments sustained the highest macroinvertebrate abundance, family richness, and Shannon diversity, as greater algal productivity increased resources available to support the macroinvertebrate communities. I conducted a second mesocosm experiment to explore how variation in canopy cover (low or high) and litter species (oak, maple, or a mixture of the two litters) affect vernal pool macroinvertebrates. Macroinvertebrate abundance and family richness were greater under low canopy and with oak litter, but the effects of canopy often depended on litter species. Canopy limits the amount of light that reaches a wetland, but litter and its impact on dissolved organic carbon (DOC) limit light transmittance and producer communities. Because forest composition and structure are changing from anthropogenic activities, these results suggest they will also alter the food webs of vernal pools. However, as some organisms utilize pools considered to be low quality for macroinvertebrates, vernal pools should be protected in a variety of habitat contexts to preserve regional species diversity.

Macrobenthic invertebrates are important indicators of water quality in many aquatic systems, but accurately identifying these organisms is a challenge, particularly for citizen scientists, due to insufficient funding for training programs and events and a dwindling pool of training personnel. Accurately identifying these organisms is key to the accuracy of the models used to generate the water quality assessments. To help address these issues, this project has developed an interactive online photographic key for macrobenthic invertebrates, designed to improve classification accuracy with minimal training for use by citizen scientists. The online digital photographic field guide was created and tested against a widely used paper key (the Izaac Walton key). Classification exercises and user satisfaction surveys were conducted at different events over two years in order to compare the two identification methods. Results indicate that most participants preferred the online field guide, but that the digital field guide did not necessarily increase classification accuracy within all user groups. The results suggest the need for some basic experience or training in using macrobethic invertebrates for water quality assessment, similar to what high school or college students receive in aquatic ecology or field science courses. Another obstacle to improve water quality assessment is the dissemination of data. A tool currently being developed as part of this project is a digital map and linked database created using ArcGIS 10. This map incorporates spatial data helpful for water quality analyses (e.g. land use/cover) from different sources and has web-enabled links to other water quality databases through hyperlinks built into the map features. This will help account for multiple variables that can affect water quality and to help scientists with outreach, data dissemination, and networking. The map project at this point is still conceptual, but is intended to demonstrate its potential utility.

Although North American sturgeon have been the focus of extensive research in the last several decades, more research is essential to ensure their conservation. The free-flowing Lower Mississippi River (LMR) is occupied by two sympatric sturgeon, pallid [ Scaphirhynchus albus (Forbes & Richardson)] and shovelnose sturgeon [Scaphirhynchus platorynchus (Rafinesque)]. These species are considered 'endangered' and 'threatened,' respectively. Recent studies have documented the life history of adult sturgeon in the Mississippi River, but studies focusing on young-of-year and juveniles are limited because young fish are difficult to collect and identifications are problematic. Spawning sites in the LMR are unknown and though extensive effort has been put forth to capture young sturgeon for scientific study, specimens are seldom collected and rarely in large numbers. This gap in knowledge is substantial because recruitment success is important for the recovery and survival of both species. This study takes an ecosystem approach in exploring microhabitat associations of larval and juvenile pallid and shovelnose sturgeon based on their diets. During systematic sampling (2001-2010) of the LMR (river kilometer 131.32-1361.18), 75 total specimens (pallid and shovelnose) were obtained using a 3.05 m modified Missouri trawl. Gut contents were analyzed and prey items were identified to the most refined taxonomic level possible. By examining microhabitats and behaviors of sturgeon prey items (mostly benthic macroinvertebrates), the microhabitat-feeding associations of sturgeon were predicted. These findings suggest that both species are specific in the type of microhabitat in which they feed. The majority of prey items (71.8%) of both sturgeon species belong to a single subgroup of Chironomidae (Harnischia complex) that are predominantly burrow in sand substrates, typically in shifting sediments of large river systems. Several other sturgeon prey items also occupy this microhabitat. These data can be paired with collection data to assess habitat use, availability, and threats, and to make recommendations for conservation. Also studied was general early life history information related to feeding of young-of-year sturgeon. Based on this study, I conclude that larval and juvenile pallid and shovelnose sturgeon have different feeding habits than adults. This study also provides an updated checklist of invertebrates known to occur in the Mississippi River.

An invasive insect, hemlock woolly adelgid (HWA), has initiated widespread hemlock decline and mortality in the Great Smoky Mountains National Park (GSMNP). Riparian hemlock mortality impacts on vegetative and aquatic systems of first-order, headwater streams were evaluated. Reference sites for this study were representative of the best available conditions within the GSMNP, with initial stages of HWA presence. Impacted sites were defined as areas with over 90 percent hemlock mortality. Impacted streams had decreased canopy coverage and increased light availability. Residual red maple, yellow birch, and sweet birch capitalized on the loss of hemlock, with increases in relative basal area and species importance values. Rosebay rhododendron responded with increased density and height at impacted sites, thereby preventing woody regeneration. Since long-term regeneration and post-mortality canopy recruitment are limited, alterations of vegetative composition and structure in the stream riparian zone are expected. A seasonal assessment (Sept. 2009 – March 2010) of aquatic impacts revealed increased diurnal variation in stream temperature and exhibited cooler temperatures during the colder months than reference streams. Impacted streams exhibited greater amounts and larger size classes of large woody debris (LWD). Higher concentrations for several nutrients, including silicon, sodium, potassium, magnesium, chlorine, and copper were detected in impacted streams than detected in reference streams. Impacted streams were characterized by higher pH and increased acid neutralization capacity, while reference streams exhibited nitrate concentrations three times higher than impacted stream concentrations. Reference streams were experiencing the initial stage of HWA-induced defoliations, increasing stream nitrate concentrations, while impacted sites had levels suggesting nitrate concentrations have returned to pre-infestation levels. A seasonal assessment of macroinvertebrates found species diversity, abundance, and taxa richness were not affected by hemlock mortality. Impacted streams had a lower density of Chironomidae and Ephemeroptera when compared to reference streams, while Pleuroceridae snails were virtually eliminated at impacted stream sites. Reference streams were dominated by the collector/filter functional feeding group (FFG), while impacted streams were dominated by the scraper FFG. Hemlock mortality induced by the presence of HWA has resulted in short-term impacts to vegetative and aquatic dynamics in stream riparian areas of the Great Smoky Mountain National Park.

Watershed urbanization has been associated with declines in biodiversity and decreases in the proportion of pollution intolerant organisms. Some of these changes in community composition may be due to increased stormwater runoff from urbanized watersheds. In order to better understand how streamflow patterns change along a putative stormwater impact gradient and which abiotic factors may be driving shifts in macroinvertebrate community composition, we studied the hydrology, chemistry, and biology in six small streams in the Rocky River Watershed, northeastern Ohio. According to Ohio EPA classification, three of these streams were primary headwater streams, draining 0.41 to 1.22 km2; three headwater streams drained 3.47 to 7.25 km2. We measured flow at each stream using an Isco 2150 area-velocity flow meter from January 17 to December 9, 2010. We sampled each stream for macroinvertebrates and water and sediment chemistry three times: June 21 - 26, August 9 - 11, and September 30 - October 10, 2010.Stream flashiness increased with watershed urbanization. The seasonality of high flow events was also altered in the more urbanized streams relative to the least urbanized streams, with least urbanized streams having the highest and urban streams having the smallest percentage of high flow days in the winter. The most urban primary headwater stream had larger average event peak flows than the least urbanized primary headwaterstream during the growing season, but not the dormant season. Peak flows did not differ among the larger headwater streams. Macroinvertebrate biodiversity, as measured by the Shannon Index, declined along the urbanization gradient. Chemical, hydrologic, and physical habitat variables were significantly correlated with macroinvertebrate community characteristics; however, the variable most strongly correlated with each macroinvertebrate community metric was a hydrologic variable. Potential multiple linear regression models to describe aspects of macroinvertebrate community composition were created from combinations of abiotic factors. The five models with the lowest Schwarz Bayesian Information Criterion and Mallow’s Cp statistics for each macroinvertebrate metric were examined further, and those that met the model assumptions were retained. In one multiple regression model, skewness of streamflow and total phosphorous concentration in stream water were highly predictive of Shannon’s Index, while in the another regression model skewness of streamflow, flow during sampling, and days since the ninety-fifth percentile of streamflow was exceeded were significant predictors of Shannon’s Index. The relative abundance of intolerant macroinvertebrate taxa was best described by skewness of streamflow. The increased flashiness of urban streams relative to least urbanized streams, the larger average flood peaks of urban primary headwater streams in the growing season, and the shifts in seasonality of high flow events provide evidence of hydrologic impairment of the more urban streams in this study. This hydrologic impairment appears to be one of the mechanisms by which urbanization influences stream macroinvertebrate communities. Therefore, increased use of lowimpact design and more rigorous stormwater detention technologies within the watersheds may be needed to restore these streams.

Eutrophication has been the most challenging global threat to the quality of our freshwater resources (Likens, 2010). Reversing the effects of eutrophication can be difficult and takes a great deal of time and money. A hypereutrophic Midwestern reservoir underwent a restoration from 2000-2002, which added a sediment basin at the inlet, three islands, five rock jetties, and the removal of rough fish. The primary objective of this study was to determine changes in the benthic macroinvertebrate community in response to the restoration. Water quality and benthic macroinvertebrate samples were collected once per month in June, July, August, and September in 2009; benthic invertebrate densities were compared from: i) 1970 and 2003 to determine if the macroinvertebrate community was reset to its original composition, ii) 1970 and 2009 to determine the impact of the restoration and iii) 2003 and 2009 to determine if the community has remained the same or if it has changed. Benthic invertebrate mean densities were significantly different between 1970 and 2003, and between 1970 and 2009. However, they were not significantly different between 2003 and 2009. Overall, water quality seemed to have had no effect on the benthic community, whereas the restoration increased the diversity of the benthos by increasing the amount of habitat available.

Dams have been built around the world for various uses since the beginning of civilization. Lately, many more modern dams do not serve their original purpose and/or are in need of repair or removal. Due to the drastic altering of the river-scape that dams have created, removing dams is also a high disturbance event. Unfortunately there have been few studies documenting the impacts of dam removal. This study focuses on the removal of a low-head dam in Southeast Michigan and the impacts the removal has the macroinvertebrate communities at 3 sites upstream and 1 site downstream. Qualitative samples were collected, then using the Hilsenoff Biotic Index paired t-tests were conducted. The results show a significant, slight increase in HBI score occurred after the dam removal, as well as an increase in the average EPT Richness and % EPT Species Composition. These findings are similar to other studies in that macroinvertebrate communities improved quickly after dam removal, leading to the conclusion of increased water quality.

Land use is a key factor influencing the health of a stream. Agricultural land use, in particular, can have detrimental effects on water quality, habitat, and biological assemblages. Artificially high nitrogen loading from fertilizer runoff can lead to eutrophic conditions in affected streams, and may significantly alter the composition of aquatic vertebrate, invertebrate and microbial communities. In this study, six streams of varying land use type (three impacted by high intensity agriculture and three in mixed-use watersheds) in Indiana and Ohio were sampled in summer and fall for a variety of chemical, physical and biotic factors. Nested ANOVAs determined differences in these factors longitudinally within each stream, and subsequent one-way ANOVAs determined significant differences in these factors based on season and land use type. The abundance and diversity of the whole bacterial community and the denitrifying community were examined using quantitative PCR and terminal restriction fragment length polymorphism (T-RFLP) analysis of the 16S rRNA gene (all bacteria) and the nosZ gene (specific to denitrifiers). Detrended canonical correspondence analysis and multiple linear regression were used to elicit possible predictive relationships between various factors and denitrification rate and denitrifier abundance. Streams impacted by high intensity agriculture were found to have significantly higher nitrate, soluble reactive phosphorus and dissolved oxygen concentrations than mixed-use streams during both seasons. These streams also contained a significantly higher proportion of pollution tolerant macroinvertebrate taxa. Mixed-use streams had significantly higher dissolved organic carbon concentrations and higher total bacterial abundances than agricultural streams. Multivariate analyses suggested nitrate concentration was an important predictor variable for both denitrification rate and denitrifier abundance. Denitrification rates and denitrifier abundance were significantly higher in agricultural streams than mixed-use streams, and generally higher in summer than fall. Redundancy analysis of T-RFLP data suggested that a larger portion of the variation among stream bacterial communities was explained when streams of differing land use type were compared. Overall, the data suggested possible relationships between denitrifier community abundance, diversity, and function, as well as the possibility to predict stream denitrification rates using chemical, physical and community factors.

Clear, cool headwater streams in the mountains of North Carolina are inextricably linked to the surrounding landscape. Trickling perennial streams drain precipitation from their catchment and are capable of sustaining excellent water quality to support rich aquatic biodiversity that feeds and beneficially contributes to the stream network below. However, headwater ecosystems can be easily compromised by even seemingly insignificant anthropogenic impacts. Small headwater streams were not mapped until recently, and are now known to be ubiquitous. Although the NC Mountains contain some of the highest headwater streams densities in the nation, they remain very susceptible to changes in the catchments that sustain them. Understanding effects from changes in the catchment can ameliorate future impacts, prioritize preservation efforts and inform restoration trajectories. Although a variety of stakeholders have preserved and passively managed unimpaired stream systems, others have attempted to enhance or restore streams with limited success. Without consideration to the surrounding landscape, these efforts may not be successful. This dissertation offers a starting point for determining thresholds of anthropogenic impacts to sustain biological integrity in headwater streams and offers examples of successful Extension education outreach efforts.

Hunnicutt Creek watershed drains to the eastern edge of Clemson University campus. The landscape of this part of campus changes from a natural park like botanical garden to one dominated by academic buildings. The campus area has changed in time as well with increasing impervious areas such as roads, parking lots, roof tops and decreasing the amount of forested land leaded to the impairment of water quality in this creek. The objectives of this study were to evaluate the relationships between water quality indicators and percent imperviousness at a watershed level. Eight sampling sites were assigned at each sub-watershed within the Hunnicutt Creek watershed. Each sub-watershed was evaluated for the percent imperviousness and a variety of water quality indicators including physical, chemical and biological parameters. The correlation between the imperviousness and these water quality parameters was determined. The results indicated the mean percent imperviousness within study site was 21%. The values of habitat score and water quality parameter (pH, dissolved oxygen, conductivity and nitrate) were significantly different among sub-watersheds. Also, macroinvertebrate matrices showed the significant differences among sub-watershed sampled sites. Percent imperviousness showed significant correlations with habitat scores, dissolved oxygen, conductivity, fecal coliform bacteria, taxa richness, EPT richness, biotic indices and percent oligochaeta at p-values ≤ 0.05. The headwaters of Hunnicutt creek, upstream of the McMillan road and the academic zone tended to be more impacted than the headwaters within the South Carolina botanical garden.

The previous spills and releases of polychlorinated biphenyls (PCB) into the Shenango River caused the Westinghouse Site to become a superfund. With the superfund concluded, our goal was to compare the dredged areas of the Westinghouse Superfund site in the Shenango River to the channel areas which were not remediated. Replicate ponar grab samples were taken in remediated areas, in the non-remediated areas and at an upstream reference site on three days between summer, fall 2007 and spring 2008 seasons. Thirty-five total taxa were collected and were dominated by the chironomid genera. Also identified were a chironomid chironomus mentum deformity incidence, tolerance index value. A multivariate analysis of variance was done on date and site type for abundance, richness and tolerance index. Site type was not significant, meaning there was no statistical significance between reference, non-remediated and remediated sites. Date was found significant as expected by the phenology of macroinvertebrates emergence. The depression of abundance and richness values, chironomid deformity incidence and tolerance index suggest mild stress on the site sampled. This stress could be from one or more sources such as urbanization or the upstream landfill. This study can prove to be a valuable baseline for future studies at the Westinghouse Site or the Shenango River.

Herein, I describe research that quantifies how native and non-native (henceforth exotic) benthic organisms influence community and ecosystem processes. As aquatic ecosystems are recovering from years of excessive inputs of nutrients and industrial pollution, the influence of benthic food webs on the overall ecosystem likely will increase. By conducting a series of laboratory and outdoor experiments, observational studies, and field assessments in small reservoirs, I quantified how benthic organisms transfer material to higher trophic levels. For the native omnivore, gizzard shad Dorosoma cepedianum, growth and survival depended on the quality of sediment detritus, suggesting that detritus quality ultimately can regulate community and ecosystem productivity, mediated by its influence on gizzard shad biomass available for trophic transfer to piscivorous fish (Chapter 2). The addition of an exotic, benthic fish, round goby Neogobius melanostomusto the Lake Erie ecosystem, by being preferentially consumed over native prey by smallmouth bass Micropterus dolomieuappears to transfer benthic energy and contaminants to the pelagic food web (Chapter 3). A field study and historical data in Lake Erie revealed biomagnification of PCBs by an exotic species component (comprising round goby and dreissenid mussels Dreissenaspp.) to native terminal predators, smallmouth bass and largemouth bass Micropterus salmoides(Chapter 4). Recovering benthic macroinvertebrate communities in Lake Erie appear influenced by dreissenid mussels and dreissenid mussel interactions with PCBs and organic content of sediments (Chapter 5). As nutrient and contaminant inputs continue to decline and exotic species continue to proliferate, I predict an increase in the relative importance of such benthic transfer pathways in influencing variability in transfer of energy and contaminants from to the pelagic food web.

The performance of OEPA’s Invertebrate Community Index (ICI) and a family-level Macroinvertebrate Aggregate Index for Streams (MAIS) was compared in 40 streams of varying impact throughout the Western Allegheny Plateau (WAP). In addition, the effect of 3 levels of subsampling on MAIS scores was evaluated. There was a significant correlation (rho = 0.719) between the ICI and MAIS. When grouped in OEPA’s biological condition categories the indices were only significantly correlated (rho = 0.703) at severely degraded sites. The indices rated 81% of sites similarly as “attaining” or “non-attaining”, but there was less agreement (37.80%) when comparing the 4 biological condition categories of each index. Subsampling had little effect on mean MAIS scores and no consistent effect on within-category score variation. Findings give support for the reliability of family-level indices, and suggest that the MAIS may be useful in degraded streams of the WAP and other streams upon recalibration.

In-stream habitat structure and water chemistry have significant influence on the structure and composition of stream macroinvertebrate assemblages. Habitat at this local scale can be significantly affected by the geomorphology of a stream or region. Both in-stream habitat and geomorphology are, in turn, influenced by other factors operating at the landscape scale (e.g., land use, connectivity of habitat patches, etc.). It is unclear which of these three scales of habitat has the greatest influence over lotic assemblage structure. Anthropogenic disturbance to a stream ecosystem can occur at all three scales of habitat, and is particularly common in predominantly agricultural systems. The Sugar Creek watershed in northeastern Ohio represents several different types of anthropogenic disturbance, including dairy farming, crop production, urbanization, and industrialization.The South and Middle Forks of the Sugar Creek watershed, dominated by agriculture and a mix of agriculture and industry, respectively, were sampled in early summer 2005 for habitat and macroinvertebrates. Richness, evenness, diversity, family-level biotic index, percent Diptera Chironomidae, and the number of macroinvertebrates were all similar across the drainages. The percent Ephemeroptera, Plecoptera, and Trichoptera was significantly larger in the Middle Fork than in the South Fork. There were no significant differences in habitat or macroinvertebrate assemblages between the two drainages overall. In-stream habitat structure and water chemistry explained 58.8% of the variation between sites among macroinvertebrate taxa. Geomorphology explained 10.4% and land use 9.4% of the variation. Shared variances between different scales of habitat did not explain substantial amounts of variation among macroinvertebrate taxa. These results have, however, identified several sites in the South Fork with good potential for Best Management Practice implementation and several sites in the Middle Fork for preservation.

The importance of streamside management zones (SMZ) in minimizing the impact of non-point source pollution from silvicultural operations is recognized by the forestry Best Management Practices of most states. However, research concerning the SMZ width and harvesting intensity required to maintain water quality and biotic communities is limited. The goal of this study is to evaluate the efficacy of different SMZ widths and forest harvesting intensities within SMZs, in maintaining the water quality and biotic communities of 22 headwater streams located in the mountains of East-central West Virginia. Streams were organized in four blocks and randomly assigned one of six silvicultural treatments involving variation of SMZ width and harvesting intensity within the SMZ; 30.5 m SMZ with no residual harvest, 30.5 m SMZ with 50% residual harvest, 15.3 m SMZ with no residual harvest, 15.3 m SMZ with a 50% residual harvest, 4.5 m SMZ and control (no harvest within the watershed). Stream water chemistry parameters (in particular, NO3, NH4, Ca, Mg, conductivity and total dissolved solids) as well as aquatic macroinvertebrate communities were monitored from June 2003 through March 2005. Average nitrate concentration in streams harvested with a 4.5 m SMZ was more than 4 times as high as that of control streams. Average summer and fall stream temperatures were inversely related to SMZ width. Mean values for a number of macroinvertebrate community metrics were indicative of poorer water quality in streams harvested with a 4.5 m SMZ. During this short-term study SMZs of at least 15.3 m appeared to be sufficient to maintain water quality. However, harvesting was restricted to one side of the stream and logging induced stream disturbances were observed even with SMZs of 30.5 m. For these reasons SMZs of at least 30.5 m are recommended as a cautionary measure to minimize the potential for impacts to biotic communities. In addition, residual harvest of up to 50% of the basal area within the SMZ did not appear to impact water quality during the temporal scope of the study.

Regulated river flows and land-use activities have altered the floodplain ecosystem within the Upper Missouri River basin. These changes have led to decreased connectivity between the floodplain and main river and enhanced land-use activities (i.e., agriculture) within the floodplain. Knowledge of floodplain wetland attributes and their response to anthropogenic disturbances is limited. In this study, I developed a macroinvertebrate index of biotic integrity (M-IBI) that evaluated the effects of disturbances in seasonal floodplain wetlands. I also examined differences in macroinvertebrate community structure in forested and emergent seasonal floodplain wetlands. As part of a U. S. EPA REMAP study, aquatic macroinvertebrates were sampled to develop biomonitoring criteria for seasonal floodplain wetlands in the Garrison Reach of the Missouri River. Twenty-six seasonal wetlands were sampled in the summers of 2000-2001. Wetlands were designated as reference (n=6) or impaired (n=5) based on land-use characteristics. Candidate metrics were first identified using univariate t-tests (e.g., reference versus impaired) and then subjected to a stepwise discriminant function analysis to develop a predictive suite of variables that distinguished reference and impaired wetlands. The final macroinvertebrate index of biotic integrity consisted of seven metrics that included: 1) proportional abundance of Chironomidae, 2) proportional abundance of Lymnaeidae, 3) proportional abundance of predators, 4) proportional abundance of dominant taxa, 5) proportional biomass of Hydraenidae 6) proportional biomass of univoltine taxa and 7) number of Odonata taxa. These metrics were then scored from 0-10 (high scores indicated similarity to the reference condition), summed and standardized for a final M-IBI that ranged from 0 to 100. The M-IBI was found to be robust to inter-year variation, but displayed considerable intra-year variation implying that time of year wetlands are sampled is critical for accurate wetland assessment. Randomly selected wetlands (n = 15) were used to determine the overall condition of seasonal floodplain wetlands within the Garrison Reach and yielded an average condition rating of fair (mean M-IBI score = 52). This study demonstrated the usefulness of macroinvertebrate data for assessing floodplain wetlands of the upper Missouri River. Riparian forests and their associated wetlands have disappeared throughout much of the Upper Missouri River basin. Information on seasonal forested wetland invertebrate communities and how they compare to surrounding emergent wetlands are limited. Twelve seasonally flooded forested (n = 6) and emergent (n = 6) floodplain wetlands were sampled in the summer of 2001. Physical attributes, periphyton biomass macrophyte coverage and water chemistry were analyzed along with invertebrate community attributes. Forested wetlands had significantly higher pH (8.16) and lower macrophyte coverage than emergent wetlands. Numerous differences in taxa richness and biomass occurred between the invertebrate communities of forested and emergent wetlands. In general, emergent wetlands had higher taxa richness and biomass than forested wetlands. The most important finding was the general increase in total invertebrate biomass as macrophyte coverage increased. Macrophytes are known to increase invertebrate abundance by affecting the physical environment (i.e., habitat complexity, shading, etc.), food availability and predation. Differences observed in invertebrate community structure suggest that biomonitoring criteria be developed separately for forested and emergent wetlands.

Litter decomposition is a poorly understood function of constructed and natural forested wetlands. This study compared rates of litter mass loss, changes in litter morphology, and associated macroinvertebrate populations in constructed and natural non-tidal wetlands. Two sets of wetlands (constructed vs. natural) were studied in eastern Virginia; a 9 year-old riparian set near Fort Lee, (FL), and a 2 year-old wet flat set in Charles City County, (CC). Mixed deciduous forest litter collected from the FL natural wetland decayed more rapidly in the created wetlands than the adjacent forested wetlands. Mixed emergent marsh litter collected from the FL created wetland exhibited a similar relationship, although marsh litter decomposed slower than forest litter. Litter area and weight loss followed a similar pattern, although area loss lagged behind weight loss, consistent with an initial leaching phase of decomposition. Both the FL and CC created wetlands exhibited faster litter decomposition than their adjacent forested wetland, however, the FL created wetland had a lower weight:area ratio and higher detritivore abundance than the adjacent forested wetland, while the reverse was true for the CC wetland pair. These relationships suggest macroinvertebrates played an important role in decomposition in the FL created wetland, while other factors were more significant at CC. Faster decomposition in the created wetlands may be of concern for long-term soil organic matter accumulation, or conversely, may indicate quick recovery of the litter decomposition function. Overall, these findings point out the difficulties involved in using certain functional indicators to compare very young and mature systems.

The structure and function of stream ecosystems may be strongly influenced by land-use practices within watersheds. The magnitude of land-use impacts may depend upon the spatial arrangement of development in watersheds. This study examines the relationship between land-cover patterns and stream structure and function in 9 southern Appalachian headwater basins. Using a GIS/remote sensing approach, land-cover patterns were quantified at several spatial scales, including the entire watershed, riparian corridor, and riparian sub-corridors extending upstream in 200 m increments for 2 km. In-stream physico-chemical variables were related to land-cover patterns at different spatial scales. Dissolved constituents (e.g., TIN, alkalinity) were frequently related to physical features or land-cover patterns at the watershed scale. Conversely, mean substrate particle size and stream temperature were most strongly related to land-use practices at the entire riparian corridor scale. Finally, maximum stream temperature was best explained from land-cover patterns at the 200 m sub-corridor scale. The relationship between land-cover patterns and benthic macroinvertebrate communities is examined in Chapter 1. Macroinvertebrate assemblage structure was quantified using the slope of rank-abundance plots, and further described using standard diversity and evenness indices. Macroinvertebrate taxa richness ranged from 24 to 54 among sites, and the analysis of rank-abundance curves defined three distinct groups with high, medium, and low diversity. In general, other macroinvertebrate indices were in accord with rank-abundance groups, with richness and evenness decreasing among sites with maximum stream temperature. Macroinvertebrate indices were most strongly related to land cover patterns evaluated at the 200 m sub-corridor scale, suggesting local, streamside development effectively alters assemblage structure. The relationship between land-cover patterns and leaf breakdown is examined in Chapter 2. Breakdown rate for American sycamore (Plantanus occidentalis) leaf packs varied significantly among sites (k = 0.0051 to 0.0180 d-1), but fell within the range reported in the literature for sycamore. Leaf breakdown rate increased among sites with shredder density and biomass. Further, breakdown rate and shredder density and biomass were positively related to mean substrate particle size. Though several instream variables were related to watershed-scale features, leaf breakdown rate was not related to land cover at the watershed scale. Leaf breakdown rate was inversely related to % non-forest within riparian sub-corridors of approximately 1 km. Results suggest that the distribution of shredders is critical to leaf processing in these streams. In some streams, increased sediment inputs as a result of agricultural activity or residential development in riparian corridors may limit the distribution of shredders and thus influence leaf breakdown rates. Alternatively, near stream development may reduce the quantity and/or quality of allochthonous inputs to streams, and thus indirectly influence the distribution of shredders.

Macroinvertebrate assemblage structure was compared from 12 streams differing in urbanization type and degree. Urbanization, both historical and current, was measured using several variables generated from GIS overlays of land cover, aerial photographs, and field exploration in the study watersheds. Quantitative benthic macroinvertebrate samples were taken, and a variety of physicochemical characteristics were measured. Increasing urbanization resulted in a decline in diversity and abundance of intolerant organisms. Streams in industrial areas had greater invertebrate density due to large increases in a few tolerant groups. Urbanization in the watersheds was coupled with changes in the physical and chemical structure of the streams suggesting some possible mechanisms for urbanization impact on stream biota. Multivariate analysis grouped streams based on a number of pollution-sensitive taxa suggesting the utility of this type of approach in analyzing community data. Primary funding for this project was from the Coweeta Hydrologic Lab NSF-LTER grant. Additional funding was provided through a Graduate Research and Development Project grant from the Graduate Student Assembly of Virginia Polytechnic Institute and State University. Matching funds were provided by the Biology Department.

I investigated the role of land use in structuring benthic invertebrate assemblages in agricultural streams in the French Broad River drainage in western North Carolina. I sampled six agricultural streams (3 with cleared headwaters and 3 with forested headwaters) at three points along a gradient (headwaters, a midpoint, and a downstream site). At each site, I measured a variety of physico-chemical parameters, including temperature, chlorophyll a, discharge, nutrients, and suspended solids. Invertebrates were sampled at all sites in October 1996 and April 1997. Riparian vegetation was assessed for each site at mutiple spatial scales using GIS data from the 1950s, 1970s, and 1990s. Forested agricultural (FA) streams had more riparian vegetation than cleared agricultural (CA) streams in both the 1950s and the 1970s. Cleared agricultural streams had less organic matter, more primary production, higher nitrates, and warmer temperatures than FA streams. Total and EPT taxa richness was greater in FA streams. Pollution-sensitive Plecoptera were relatively more abundant in FA streams, while tolerant Diptera were more abundant in CA streams. High diversity and Plecoptera abundance was related to high habitat quality, more riparian vegetation, low nitrates, and low summer temperatures. Higher invertebrate diversity was related to the land use 25-50 years as well as the current land use (forested, moderate agriculture, or heavy cattle impact). These results indicate a long-term legacy of agricultural influences on stream invertebrate assemblages.

A wetland passive treatment system (PTS) was used to treat mine effluent flowing into East Prong Creek, Virginia. Prior to treatment, copper concentrations in the stream ranged from 8.9 to 32.0 μg/L at the impacted sites and from 0.1 to 7.7 μg/L at the reference site. In kick samples, insect abundance (n) and the number of taxa (#) were reduced at the impacted sites (n = 31, 22, 33, 24 and # = 190, 246, 266, 345, at sites 2 - 5, respectively) relative to the reference site (n = 52, # = 973). Red maple (Acer rubrum) leaves broke down twice as fast at the reference site (k = -0.029) than at sites receiving the untreated effluent (k = -0.016, -0.013, -0.013, -0.013 at sites 2 - 5, respectively). In benthic samples and leaf packs, abundances of all functional groups were reduced at the impacted sites relative to the reference site. In terms of abundance, scrapers and predators were most and least affected by copper, respectively. Following treatment, copper concentrations at the impacted sites in East Prong Creek ranged from 0.1 to 14 μg/L. The wetland PTS was most effective at reducing metal concentrations in the effluent from July to December, when dissolved oxygen concentrations in and flow through the wetland cells were low. Functional recovery preceded recovery of community structure by at least six months. Decomposition rates were similar at all sites in Fall 1994, after six months of treatment (k = -0.012, -0.011, -0.011, -0.012, -0.012) at sites 1 - 5, respectively). Recovery of community structure was not complete after 1 year of treatment. Abundance of collector-filterers and predators recovered quickly following treatment, while recovery of collector-gatherers and shredders was slower.

This study was conducted to gather information regarding aquatic macroinvertebrate habitat in the middle James River. Macroinvertebrate abundance, standing stock biomass, and production were quantified and compared among major instream habitats at two widely separated river sites. Three habitats that dominated the total substrate available for macroinvertebrate colonization were sampled: snags (fallen trees and woody debris), rock (areas of stony substrata), and mud (areas lacking instream features with streambeds consisting of fine unconsolidated sediments). Sampling stations were established in areas where instream habitat was dominated by a single habitat feature (e.g. snag, rock, mud). Both river sites contained three stations from which sampling was conducted every three weeks from May 1990 through October 1990. The size-frequency method was applied to individual taxa to determine total macroinvertebrate community production. Total macroinvertebrate densities, standing stock biomass, and production were much higher on snags than in rock or mud benthos. Mean annual densities of total macroinvertebrates ranged from 56,761 m2 on snags to 12,279 m2 in rock substrates. Abundant taxa varied among the three habitats. Glyptotendipes spp. (Chironomidae), Cyrne/lus .fraternus (frichoptera), and Naididae predominated on snags while Caenis spp. ( Ephemeroptera), chironomids, and oligochaetes predominated in rock habitats. The mud benthos was dominated numerically by Dero spp. (Oligochaeta), immature Tubificidae and Procladius spp. (Chironomidae). Mean annual standing stock biomass ranged from 4.4 g DM m2 on snags to 1.0 g DM m2 in mud substrates. Numerically abundant taxa made up most of the biomass in each habitat. However, Anodonta grandis (Pelecypoda) was an important contributor to the biomass of the rock benthos and Chironomus spp. ( Chironomidae) made up much of the biomass in the mud benthos. Mean annual production on snags (329.8 g DM m2 yr1) was far greater than in rock (46.0 g DM m2 yr1) or mud (46.2 g DM m2 yr1) benthos. On a substrate surface area basis, standing stock biomass on snags was only 1.3-3.1 times higher than in rock habitat and 4.3-4.5 times higher than in mud habitat. Production on snags was 5.2-12.2 times higher than in the benthic habitats, with higher annual P/B in the former. The production estimates for snag habitat are among the highest yet reported for lotic ecosystems, and it appears that production on snags is limited by available substrate. Relative density, biomass, and production for each habitat for a stretch of river was approximated. Although snag habitat accounted for only 1 % of the effective habitat substrate over a stretch of river, it was responsible for 5.4-8.7% of production. Taxa within each habitat were categorized by functional feeding groups, and habitat-specific functional groupings were evaluated using numbers, biomass, and production. Shredders predominated on snags and gathering collectors in benthic habitats. When corrected for habitat abundance, community structure consisted primarily of gathering collectors and to a lesser extent predators. As a high-order (> 6) stream, the distribution of biomass or production among functional groups in the James River resembles that predicted by the River Continuum Concept, having a high percentage of collectors. However, the dominance by shredders on snags strongly suggests that allochthonous materials are an important food source, and that limitations imposed by availability of snags determines the composition of the macroinvertebrate community. Results of this study suggest that efforts should be taken to preserve instream features and habitat diversity in warmwater streams. Special consideration also should be given to the protection of streamside forests.