Phytoplankton account for at least half of all primary production in estuarine waters and are at the center of biogeochemical cycles and material budgets. Environmental managers use water column chlorophyll-a (chl-a) concentrations as a basic water quality indictor, as the problems of eutrophication and hypoxia are intrinsically linked to excessive phytoplankton growth. Evidence suggests that the distribution and frequency of harmful algal blooms may be increasing worldwide. For the most part, phytoplankton communities follow a standard seasonal pattern, with specific groups dominating the assemblage during the time of year when environmental conditions correspond to their requisites for growth. However, climate change will result in incremental but consistent shifts in some environmental factors known to affect phytoplankton production and biomass accumulation. Mean surface temperatures in North American mid-Atlantic coastal and estuarine regions are steadily rising, and the frequency and severity of drought and storm events are projected to fluctuate, potentially increasing the severity of extreme weather events. Anthropogenically-induced nutrient loading, especially from non-point sources, is one of the largest consistent contributors to coastal marine eutrophication. The consequences of changes in these environmental factors to estuarine ecosystems and phytoplankton community dynamics are unclear. Because different phytoplankton groups respond to environmental changes in distinctive ways, some classes thrive during periods of environmental stability and others at times of temporary or sustained disturbance. to predict how phytoplankton and therefore water quality might respond to changes in climate and land use, we built mathematical phytoplankton kinetics sub-models that differentiate phytoplankton groups using taxonomic classes with well-defined functional characteristics. Then we integrated them into a reduced-complexity estuarine ecosystem model. The sub-models were designed to simulate daily biomass of diatoms, dinoflagellates, cyanobacteria, and raphidophytes in the New River Estuary, NC. We calibrated and validated the model using data collected from 2007 – 2012 through the Aquatic Estuarine monitoring module of the Defense Coastal/Estuarine Research Program. The model was a relatively good predictor of total chl-a and primary production, and a fair predictor of group dynamics. The model was employed in heuristic simulations of changes in temperature, nutrient loading, and freshwater delivery to predict their effects on overall phytoplankton biomass, productivity, and community composition. Increases in temperature had a modest effect on mean daily simulated phytoplankton production and chl-a, but considerably decreased the relative abundance of diatoms and simultaneously increased the relative abundance of cyanobacteria. The seasonal phenology of phytoplankton abundance also shifted in response to increased temperatures: chl-a concentrations were larger in the winter and spring and smaller in the summer and fall. The model was most sensitive to changes in the watershed nutrient load. Nutrient influx had a dramatic effect on the temporal and spatial extent of phytoplankton blooms. The relative abundance of dinoflagellates and raphidophytes increased in response to elevated nutrient loading, regardless of whether load was increased directly as in nutrient simulations or indirectly as in freshwater simulations. Initially, greater freshwater discharge increased total chl-a, productivity, and the frequency of phytoplankton blooms. However, these relationships leveled off or were reversed as flow continued to increase due to greater rates of flushing and light attenuation. Results demonstrated how models like this can be important tools for both heuristic understanding and environmental management. A benefit of this model is how easy it is to update to other estuarine systems through the re-parameterization of the phytoplankton groups.

Habitats of Chesapeake Bay have been altered due to anthropogenic impacts and climate change. Due to these human disturbances, seagrasses have been extirpated from many areas in lower Chesapeake Bay and persisting beds face future losses as water temperatures continue to rise. Further loss of seagrass habitat will negatively impact juvenile blue crabs (Callinectes sapidus) that use seagrass beds as nursery grounds. Habitat degradation allows for more successful introductions of exotic species, and the communities formed from the mixing of native and exotic species are known as emerging ecosystems. Gracilaria vermiculophylla, an exotic macroalga, may be an emerging nursery habitat for juvenile blue crabs in Chesapeake Bay; however the extent to which the alga is present and used as a nursery by juvenile blue crabs are largely unknown. I investigated algal distribution in the shallow littoral areas of the York River, a subestuary of Chesapeake Bay, over two years (2013 – 2014) and found that G. vermiculophylla presence correlated with salinity and that algal presence and biomass increased with seagrass presence, although biomass was generally low. The alga was present in areas where seagrasses have been lost, and is therefore likely providing nursery habitat in these areas of high megalopal recruitment. Benthic epifaunal communities had lower species richness and were less abundant in G. vermiculophylla relative to seagrass, while benthic infaunal communities had lower species richness but similar abundance in the alga relative to seagrass. Juvenile blue crab densities were similar in the alga and seagrass, although seagrass supported about 3 times as many first and second instar crabs than G. vermiculophylla. Young juvenile blue crabs preferred seagrass, which may be due to epifaunal prey preference, and G. vermiculophylla likely represents a secondary nursery habitat. Juvenile blue crab growth rates of crabs 15 – 50 mm carapace width were similar in the alga, native seagrass, and unvegetated habitat, indicating that growth does not drive ontogenetic shifts in habitat use by larger (20 – 30 mm carapace width) juveniles. Similar growth rates also suggest that G. vermiculophylla performs similarly to seagrass as a nursery habitat in terms of providing resources for growth. Simulations of density-dependent migration of young juvenile blue crabs between habitat types suggest that G. vermiculophylla may mediate continued seagrass loss, at least in part. Together, these results increase our understanding of an emerging Chesapeake Bay ecosystem and the impacts that changes to nursery habitats have on the juvenile component of the blue crab population.

The finfish communities using the intertidal surfaces of a bay-exposed marsh and a sheltered, channel marsh of the Goodwin Islands, York River, Virginia were studied from May to November 1994. Samples were taken approximately twice monthly using a flume weir at three stations in each marsh. Species composition, abundance and biomass were estimated, and growth patterns and production rates were described for the dominant species. A total of 3001 fish were collected from 11 species and 8 families. Fundulus heteroclitus dominated both marshes and accounted for 83.07% of the total number of fish caught, and 84.83% of the total biomass collected. Menidia menidia was the second most abundant species at 9.00% of the total number of fish, but F. majalis was second in the total biomass at 7.51 % of the total. Abundance first peaked in June, and a second peak occurred in November. Biomass had a small peak in the beginning of August, and a large peak in late September. Analysis of variance showed the number of species captured did not differ significantly between the marshes. However, both the number of individuals and biomass were significantly higher in the protected marsh. Correspondence analysis showed that species composition in the protected marsh was mostly cyprinodontids, but in the open marsh species composition varied more throughout the sampling season. Species densities were highest at 9.8 ± 3.3 fish/m2 for F. heteroclitus in the protected marsh. Production for F. heteroclitus over the summer in the protected marsh was high at 10.1 g dry wt.fm2 due to the large number of rapidly growing larval and juvenile fish.

Marine turtles are distributed in temperate, sub-tropical and tropical waters and beaches worldwide, often in areas heavily impacted by humans. Although there are many threats to marine turtle populations, the growing threats of nutrient pollution and harmful algal blooms are relatively understudied despite their widespread impacts on coastal marine ecosystems that marine turtles depend on. By studying juvenile green turtles (Chelonia mydas) in the Indian River Lagoon, Florida, where nutrient pollution and HABs are a widespread and longstanding issue, I aimed to conduct a case study of how these threats may affect this federally Threatened species. In Chapter 2, I used four concurrent, 18-year data sets to characterize and assess the interrelatedness of long-term trends in seagrass cover, macroalgae occurrence, juvenile green turtle abundance, and juvenile green turtle growth rates. From 2000 to 2018, IRL seagrass cover declined precipitously, macroalgae rose slowly through 2011 then declined during two severe HABs, juvenile green turtle abundance declined slowly, and growth rates declined through 2011 then rose through 2018. In Chapter 3, I conducted a 9-year study of juvenile green turtle foraging ecology using a comparative stable isotope approach. I found that carbon and nitrogen stable isotopic variance declined during and after two severe HABs in the IRL. In Chapter 4, I used two complementary methods to assess the diet of juvenile green turtles after two severe algal blooms in order to assess changes compared to previous diet studies. Visual identification of forage items showed that juvenile green turtle diet remained dominated by nutrient-tolerant red macroalgae with smaller components of seagrass and green algae; metabarcoding techniques largely failed to resolve their diet. My results highlight the web of complex effects and responses that factor in to determining the effects of nutrient pollution and HABs on juvenile green turtles. Future studies of habitat selection, foraging ecology, and the effects of these on juvenile green turtle growth and survival are needed to fully assess the threat of nutrient pollution.

Through genome wide association of nonvolatile metabolites and leaf ecophysiological traits, historic breeding practices were found to have led to germplasm divergence within the cultivated sunflower Helianthus annuus. In genome-wide analyses of single nucleotide polymorphisms (SNPs) in relation to flower petal carotenoid content across the cultivated H. annuus germplasm, alternative methods of analysis proposed differing genetic architectures, which suggests that these methods can be used as complementary approach in prioritizing SNPs for function analysis. Leaf hyperspectral reflectance was leveraged in a machine learning framework to predict herbivore- and volatile induction across the genus with 95% accuracy, while characterizing changes in volatile metabolites. The body of work in this dissertation represents the first characterization of the standing genetic variation for nonvolatile specialized metabolite diversity in cultivated sunflower in the context of modern breeding practices, and the first assessment of hyperspectral reflectance and volatile metabolite diversity across the genus Helianthus.

Coastal habitats provide crucial nursery habitat for predatory fishes, but they are in decline worldwide, impacting economically important fisheries. Habitat restoration can simultaneously mitigate the effects of habitat loss and benefit predators (e.g., sportfish), although this relationship is understudied. Here, the response of juvenile sportfish to oyster reef and living shoreline restoration is compared to controls in the Indian River Lagoon, Florida prior to and following restoration for up to three years through examination of community (diversity, assemblage), population (abundance), biometric (size, body condition), ontogenetic, and trophic (gut contents, stable isotopes) dynamics. Stable isotopes were used to quantify dietary history, beyond the temporary record of gut content, using carbon (δ13C) and nitrogen (δ15N). This study found juvenile sportfish abundance and richness was higher at restored reefs compared to natural and degraded reefs and was comparable between stabilized and natural living shorelines. Biomass was higher at oyster reefs compared to living shorelines, which was likely driven by an ontogenetic shift of larger mangrove snapper Lutjanus griseus toward oyster habitat. Snappers and prey at live reefs were carbon (δ13C) enriched, with a smaller isotopic niche area compared to dead reefs, while restored reefs were intermediate and contracted over time. Therefore, sportfish at higher quality reefs likely exhibit greater site fidelity, particularly among less-mobile juveniles. Moreover, stabilized shorelines can achieve trophic equivalence to natural shorelines, as shown through their similar isotopic signatures and niche area. These responses were best predicted by benthic habitat (e.g., oyster density), prey abundances, and site location, demonstrating the connection between habitat quality and setting to provide prey that support sportfish populations. This study validates the benefits of habitat restoration to economically important fisheries, by augmenting various attributes of juvenile sportfish life history, which can improve their survival and recruitment into the adult population, and thus extend restoration outcomes.

<p>Elucidating the mechanisms structuring communities has been a challenge for community ecology since its beginnings. One theory argues that assembly rules structure communities by means of deterministic mechanisms arising from biological interactions. Another view maintains that patterns seen in community composition and species abundance result from stochastic processes such as migration and extinction. The dilemma has yet not been resolved unambiguously. The main issue is that communities shaped by deterministic mechanisms can produce stochastic patterns via priority effects. The main goal of this study was to determine whether assembly rules structure communities. My strategy was to minimize priority effects by controlling timing of colonization. To do this I used a null community by combining communities of 17 rock pools. This null community was later divided among experimental communities. I conducted three experiments: (1) Experimental communities were exposed to the same external conditions. (2) Communities were exposed to different environments, disturbance, dispersal and habitat heterogeneity. (3) Replicated null communities were connected to allow inter-replicate dispersal. After 4 months, communities (experiment 1) formed alternative states, suggesting the lack of assembly rules control in community structure. The second experiment showed that adding factors results in more alternative states. The increasing number of alternative states among replicate communities indicates that diversified environment and migration are needed to reproduce qualitative patterns observed in nature. The last experiment (3) showed that patterns observed among connected replicate communities resemble patterns that emerged in the presence of biological interactions in unconnected communities. Similarity of patterns between connected and unconnected groups of communities suggest that local biological interactions can be sufficient to structure communities to a considerable degree. Nevertheless, the regional processes appear necessary in their role of supplying species for local communities.</p> / Doctor of Philosophy (PhD)

<p>Aquatic vegetation in the pristine coastal marshes of eastern Georgian Bay (GB) provides critical spawning and foraging habitat for fish species, with complex habitat supporting the greatest diversity. These wetlands are threatened by a changing water level regime and forecasted lower water levels. To monitor and conserve these wetlands, we must understand how they function and respond to this stressor. The overall goals of this thesis are to determine the impact of declining water levels on both wetland fish habitat and the fish community as well as identify the spatial scale of habitat utilization by fishes.</p> <p>We first delineate all coastal wetlands in eastern GB, identifying 3771 wetlands that provide habitat for Great Lakes fishes. Using satellite imagery, we develop an object-based classification method to classify four types of wetland vegetation. Since submerged aquatic vegetation (SAV) is not visible from satellite imagery in GB, we develop a model to predict potential area of this important habitat. The model suggests that the response of SAV to declining water levels depends on wetland geomorphology, but generally, the area of SAV decreases. To assess the response of fish habitat coverage and structure to sustained low-water levels, we classify vegetation in images collected in 2002 and 2008. The result is increasingly homogeneous habitat, a net loss of fish habitat and a decrease in fish species richness. Finally, mark-recapture and radio-tracking are used to evaluate fish movement among closely situated wetlands. Results suggest that the current distance used to group and protect small wetlands provincially (750 m), likely protects most resident fish species, but does not cover movement patterns of a top predator. This research will advance our scientific understanding of freshwater coastal ecosystems and aid in the creation of conservation strategies to mitigate future threats from declining water levels.</p> / Doctor of Philosophy (PhD)

<p>Coastal wetlands are hydrologically connected to their watershed and the lake. Water levels in Georgian Bay have been at a sustained low for thirteen years and thus connectivity of wetlands to the lake is being threatened as water levels decline. Decreased connectivity has likely caused changes in ecological and chemical characteristics. Future climate change models predict further water declines and potentially increasing the number of wetlands that will be hydrologically disconnected. The over-arching goal of this thesis is to investigate the role of connectivity between the lake and coastal marshes in eastern Georgian Bay on the amount of potential fish habitat, water chemistry and larval amphibian habitat.</p> <p>Bathymetric information is needed in order to estimate fish habitat and two approaches were utilized in order to collect these data. A site-specific method completed in 2009 used an intensive field survey in seven wetlands to create a digital elevation model and calculated the amount of fish habitat at 10 cm increments. A second, regional method, selected 103 sites by using a stratified random sample in 18 quaternary watersheds. In both methods, changes in water levels between 173 and 176 m asl resulted in the most drastic loss of habitat. Approximately 24% of the current fish habitat has already been lost due to low water levels.</p> <p>Water chemistry in coastal marshes is influenced by hydrologic connection. In the summers of 2010 and 2011, 35 coastal marshes were sampled, 17 of which had been impounded and disconnected by a beaver dam. Beaver-impounded marshes resulted in significantly lower pH, conductivity, dissolved oxygen and sulphate concentrations, but had significantly higher soluble reactive phosphors concentrations. These conditions are indicative of the lack of connection and reduced mixing with lake water. This altered habitat was shown to support breeding area for 7 species of amphibians, the most common being green frogs and the least common being American Toads and chorus frogs.</p> / Master of Science (MSc)

The tidal-fresh James River experiences recurring blooms of toxin-producing cyanobacteria, including Microcystis aeruginosa. However, cyanobacteria cell densities in the James are relatively low. Our purpose was to identify key factors suppressing cyanobacteria growth in the face of eutrophication. A mesocosm experiment was designed to test the effects of light, mixing and grazing on cyanobacteria abundance, with nutrients held constant. We predicted that toxic cyanobacteria would be most abundant under stagnant conditions, with enhanced light, with no bivalve grazers present. Abundances of indicator gene copies and phytoplankton counts supported this hypothesis. However, chlorophyll-a, phycocyanin, and the toxin microcystin were all found to be most abundant under mixed conditions with ambient light. Statistically, light and mixing were important in controlling toxic cyanobacteria abundance, with little to no effect observed for bivalve grazers. Our findings suggest that toxin production may be regulated by factors independent from those driving algal growth and cyanobacteria abundance.