Climate change impacts in the Chesapeake Bay will limit the efficacy of nutrient reduction efforts and decrease dissolved oxygen, but uncertainties associated with the magnitude of these effects remain. An understanding of underlying mechanisms that have driven recent warming trends will narrow uncertainties for future pathways of temperature change. Additionally, future simulations of climate impacts in the estuary are dependent on multiple different sources of uncertainty, many of which have not yet been fully evaluated. This dissertation used a three- dimensional coupled hydrodynamic-biogeochemical model to investigate recent warming trends as well as underlying uncertainties likely to influence regional projections of changes to estuarine dissolved oxygen.Recent warming trends over the past 35 years were analyzed using a combination of long- term observations, hindcast simulations, and model sensitivity tests. Robust agreement between model results and in situ sampled temperatures supported the use of the numerical model to calculate warming trends. Additional sensitivity tests that isolated the impact of different factors on long-term temperature trends demonstrated the dominance of atmospheric warming on the Bay, while also identifying the importance of ocean warming on summer temperature increases within the estuary’s southern reaches. The relationship between future climate impacts on watershed processes and estuarine hypoxia were also investigated by varying a multitude of climate input factors. These factors that modified watershed forcings for the estuarine model included the choice of Earth System Model, downscaling methodology, and watershed model. Results showed that each of these factors contributed substantially to the total uncertainty with respect to changes to hypoxia. Simulations also showed that the largest remaining uncertainty for dissolved oxygen is tied to the successful implementation of watershed nutrient reductions, which will decrease hypoxia by an order of magnitude more than increases due to watershed climate impacts alone. Further uncertainties associated with estimates of future Chesapeake Bay hypoxia are due to climate projection scenario design. These were studied by analyzing multiple common approaches for generating climate scenarios. Simulations of climate impacts on mid-21st century hypoxia derived using continuous, delta, and time slice approaches were compared. Key findings demonstrated that the commonly used delta method doubled the projected change in hypoxia relative to the continuous and time slice approaches. Differences in experimental results suggest that when assessing changes in estuarine hypoxia continuous simulations should be favored over time slice experiments whenever possible, and delta approaches should be avoided. This dissertation set out to generate scientific knowledge applicable to environmental managers working to integrate the unknowns of a future climate into decision-making for a more resilient ecosystem. The evidence produced by this research affirms the capabilities of numerical modeling techniques to identify unobservable causal mechanisms and constrain a future range of uncertain changes to dissolved oxygen levels. Hopefully, this work will also help to optimize future observational sampling efforts, and identify nutrient reduction priorities in the face of mounting climate stressors.

Mixotrophic protists, which combine the use of photosynthesis and prey ingestion to obtain nutrients for growth, comprise a substantial portion of the plankton community. However, there is a major gap in our understanding of how mixotroph prevalence varies spatially and temporally and under what conditions they dominate. I utilized a recently developed molecular technique to experimentally identify active mixotrophs (taxa identified to be grazing when samples were collected) and combined this with microscopy data to estimate active mixotroph abundance and proportion at two locations in a temperate estuary over a year. Active mixotroph abundance was compared to potential mixotroph (taxa that have demonstrated mixotrophic capability in previous peer-reviewed studies) abundance to assess potential overestimations of mixotrophs when not accounting for which taxa are actively ingesting. Measurements of potential mixotrophs demonstrated overestimations of mixotroph abundance. More importantly, analyses demonstrated that presence of taxa with mixotrophic capability does not necessarily mean they are engaging in mixotrophic activity. Constraining the identification of mixotrophs to known active mixotrophs present in the same environment combined with environmental conditions conducive to mixotrophy provides the most accurate estimation of the abundance and proportion of mixotrophs. The abundance of active mixotrophs at both locations was correlated to the abundance of major taxonomic groups. At one location, dinoflagellates were the dominant mixotrophic ASVs throughout the whole year, and the environmental patterns associated with a high abundance of active mixotrophs were similar with the patterns associated with a high abundance of dinoflagellates. At the other location, dinoflagellates and cryptophytes were the dominant mixotrophic ASVs depending upon the time of year, and the environmental patterns associated with a high abundance of mixotrophs were similar to patterns associated with a high abundance of cryptophytes. The results obtained suggest that in situ mixotroph abundance might not be only regulated by environmental conditions favorable to mixotrophy but, instead, environmental conditions favorable to each specific taxon’s utilization of phagotrophy. These findings substantially increase our understanding of how in situ mixotrophic abundance and proportion are influenced by the planktonic community composition in combination with environmental factors.

The influence of bottom-up and top-down controls on the formation and persistence of phytoplankton blooms has been well studied. However, the relative importance of these bottom-up and top-down controls vary spatially and temporally. In the tidal tributaries and mainstem of Chesapeake Bay, the summer dinoflagellate population follows a succession of bloom-producing species. The dinoflagellate species Margalefidinium polykrikoides and Alexandrium monilatum are currently considered the end of this succession. These species form near-annual blooms in the lower half of Chesapeake Bay and are considered harmful algal bloom (HAB) species due to their negative ecological impacts. However, analysis of long-term monitoring data and previous field samples suggest that Ceratium furca, a non-toxic dinoflagellate, might be an overlooked species in this dinoflagellate succession. My objective was to explore the influence of bottom-up and top-down controls on the species succession of the late summer phytoplankton blooms in the lower Chesapeake Bay. In the laboratory I used cultures of A. monilatum and C. furca isolated from the York River to evaluate the abiotic drivers influencing the succession of the late summer bloom from A. monilatum to C. furca. These experiments showed that each species exhibited differing light and temperature preferences but neither had a preferred N:P ratio. Lower light levels and lower temperatures favor non-toxic C. furca over toxin-producing A. monilatum in the York River. This information can help oyster aquaculturists identify regions of Chesapeake Bay that are unlikely to favor A. monilatum and are safer for oyster grow-out practices. Next, I used a combination of weekly field sampling along with in situ experiments during the late summer to assess top-down control of the copepod Acartia tonsa on harmful M. polykrikoides blooms. Sampling occurred during M. polykrikoides blooms in the lower York River in 2021 and 2022 and prey removal experiments were conducted using the water and copepods collected. I found that at M. polykrikoides abundances above 2000 cells mL-1 A. tonsa experienced >50% mortality in the prey removal experiments over 24-hours. Furthermore, A. tonsa abundances within the lower York River declined over the course of the bloom. This suggests that at high concentrations, M. polykrikoides may act as its own grazing deterrent, reducing the influence of top-down control and supporting bloom proliferation and maintenance. The time it takes for the A. tonsa population to recover from the negative impacts of M. polykrikoides blooms may also result in a “window of opportunity”, where a decrease in grazing pressure could support the formation of other blooms, like A. monilatum. This research provides insight into the bottom-up and top-down controls influencing the species succession in late summer phytoplankton blooms in the lower Chesapeake Bay. Evidence supports environmental conditions aiding in the transition from A. monilatum to C. furca. However, decreased grazing potential, as during an M. polykrikoides bloom, may also aid in the succession of species and provide opportunities for blooms to form. These findings can help environmental managers better predict when these blooms will occur and help to mitigate the negative impact of these blooms.

No description available.

No description available.

No description available.

No description available.

As fisheries management moves away from single-species approaches and towards more holistic, ecosystem-based approaches, physiological and ecological interactions need to be explicitly considered and mechanistically understood. Accurate portrayals of food web interactions and the direction and magnitude of energy flow between predator and prey populations are fundamental components to further develop ecosystem-based fisheries management (EBFM). to bolster information that is required within an EBFM framework in the Chesapeake Bay, I conducted research designed to advance traditional dietary studies and better understand the form and structure within the Bay's food web. This research relied on controlled feeding experiments, comprehensive sampling of predator and prey communities, and over 10 years of data from the Chesapeake Bay Monitoring and Assessment Program (ChesMMAP) and the Juvenile Fish and Blue Crab surveys. The dissertation presented here has two main objectives: 1) incorporate additional methodologies to improve stomach content identification, and 2) examine the drivers of trophic interactions and consumption within a suite of abundant and economically valuable predatory fishes in the Chesapeake Bay. Prey that is considered unidentifiable is often ignored in stomach content analyses, but can account for a significant proportion of fish diets. in Chapter 1, I demonstrate the use of molecular techniques to detect specific prey consumed by Atlantic croaker (Micropogonias undulatus) and evaluate factors that influence the rate of gastric evacuation. Molecular protocols developed to identify specific prey DNA from stomach contents determined that DNA from blue mussel (Mytilus edulis) can be detected as long as prey resides in the stomach (~30 hours), which is long after prey can be considered visually identifiable. Furthermore, temperature significantly influenced gastric evacuation rates and therefore should be considered throughout the collection process to ensure accurate identification of prey. Chapter 2 evaluated prey selection patterns among three sympatric predators in the Chesapeake Bay: weakfish (Cynoscion regalis), summer flounder (Paralichthys dentatus), and Atlantic croaker. Comprehensive sampling of predator and prey (midwater, zooplankton, benthic) populations revealed selection patterns on dominant prey selection taxa driven by a variety of mechanisms. Bay anchovy selection was significantly influenced by predator size in both weakfish and summer flounder. Mysid selection was driven by both fish size and Julian Day in weakfish and by temperature in summer flounder. Atlantic croaker select for both polychaetes and bivalves, with selection patterns relating to predator size and Julian Day. to evaluate how trophic linkages and environmental conditions influence consumption, bioenergetics models were developed in Chapter 3 for young-of-the-year Atlantic croaker and weakfish. Annual consumption from 2006 – 2016 was estimated and subsequent analyses demonstrated that prey abundance metrics significantly influenced the observed consumption patterns. This research represents a comprehensive study on predator-prey interactions within the Chesapeake Bay and contributes to a broader understanding of fish ecology and production patterns. The results from this dissertation provides a better understanding of food web structure and aids in the development EBFM strategies towards the sustainable use of marine living resources.

Freshwater inflow influences numerous physical, chemical, and biological characteristics of estuaries. The influx of freshwater to an estuary typically serves as an important source of allochthonous material from which primary producers derive their energy and transfer this energy to higher trophic levels. Any changes to freshwater flow subsequently impacts nutrient delivery and indirectly impacts organisms across multiple trophic levels. Anthropogenic changes to coastal land use and climate both act to threaten the integrity of estuarine systems by influencing freshwater inflow and dissolved nutrient input. Watershed loading models such as the Regional Nutrient Management (ReNuMa) model offer the ability to estimate freshwater inputs and dissolved nutrient loads to estuaries under current and future conditions. This tracking is important because it allows scientists to better understand how watershed delivery is currently impacted by anthropogenic activities and natural environmental variability, which allows for a better understanding of how watershed delivery is likely to be affected by anthropogenic changes in land use and climate. This research aims to assess how changes in climate and coastal land cover will impact streamflow and loads of total dissolved nitrogen (TDN) and total dissolved phosphorus (TDP) to the New River Estuary (NRE), NC. We applied the ReNuMa model to the NRE watershed to estimate streamflow, TDN, and TDP loads. We used in situ data to calibrate (2009-2011) and validate (2012-2014) modeled streamflow and dissolved nutrient loads within 10 subwatersheds located on Marine Corps Base Camp Lejeune (MCBCL), which surrounds the estuary, and one subwatershed in the off-base portion of the NRE watershed. Following model calibration and validation, model parameters were scaled up from these subwatersheds to estimate loads from the entire NRE watershed. Model results confirm the ability of ReNuMa to capture seasonal variability in streamflow, TDN, and TDP for >50% of the subwatersheds. Under current conditions, most (71-98%) streamflow and dissolved nutrient loads are sourced from the off-base portion of the NRE watershed, while a smaller percentage of loads (2-29%) are sourced from MCBCL. Projected changes in climate revealed that changes in precipitation, even when compounded with changes in temperature, will have the greatest impact on resulting streamflow, TDN, and TDP. Streamflow and dissolved nutrient loads generally increased under anticipated climate projections through the year 2100 and such increases were further amplified under hypothetical increases in land use, especially agricultural land. Watershed delivery patterns for the NRE may therefore be substantially altered under projected changes in climate and land use. The potential impacts of changes in these loads on estuarine physical, chemical, and biological processes highlights the necessity for research assessing the impacts of land use and climate changes on watershed delivery.

Unusual mortalities of cultured Crassostrea virginica in late spring have been reported from farms in the Chesapeake Bay from 2014 to 2017. None of the usual causes (e.g. disease, poor husbandry) were likely responsible, and mortalities occurred without clear signs of biological or physical stressors. Mortalities in the spring of 2014 were particularly high on over a dozen farms in Virginia, most of which were on the bayside of the Eastern Shore. Estimated losses were over 50%; however, mortalities only occurred within a four-week period between mid-May and early June. Farmers that had unusually high mortality in their crop were exclusively growing triploid oysters, which implicated triploidy as an important factor and even led to calling these events "triploid mortality." Many affected farmers were also growing a "northern cross," oysters made by crossing brood stock from New England to brood stock from Virginia, and relatively high fecundity was found in some triploid oysters sampled from these farms. It was hypothesized that the genetic contribution from the New England parent of the northern cross was causing aberrant gametogenesis in triploid oysters during the late spring, and that this caused triploid mortality. A controlled field test was conducted to evaluate this hypothesis and to further examine the role of genetics and gametogenesis in triploid mortality. to investigate the effect that ploidy and brood stock origin have on susceptibility to triploid mortality, four triploid and four diploid types of oysters, produced by crossing different combinations of brood stock of Virginia, Louisiana, and Maine origin, were deployed to three commercial oyster farms that experienced unusually high mortality in 2014, as well as to a control site lacking reports of unusual spring mortality. The survival and growth of oysters from each group were measured throughout the spring and summer of 2016. A mortality event (>20% mortality) was observed in late spring at one site. The mortality event was only associated with triploid groups, with cumulative mortality ranging from 12 to 24% among groups. Mortality in the "northern cross" was not especially high (23%). The effect that gametogenesis had on the mortalities was investigated by examining histological cross sections of triploid and diploid oysters from the site where triploid mortality occurred, as well as from the control site where there was no triploid mortality. Diploid oysters at the affected site became gravid and spawned earlier than diploid oysters at the control site. Gonad development in triploid oysters was abnormal and variable, so categories were developed for classifying gonad development in triploid oysters and making comparisons between sites. No obvious difference was observed in gonad development between triploid oysters at the two sites. Cross-sections from male triploids did not typically contain any spermatozoa, and cross sections from females usually contained only a few oocytes. Gonad development may not be strongly associated with triploid mortality, but the physiology associated with abnormal gametogenesis in triploid oysters could be a major contributing factor. More work is needed on the physiological consequences of gametogenesis in triploid oysters to explain triploid mortality.