The global climate is strongly regulated by the oceans, which store carbon away from the atmosphere for long periods. In an effort to understand the role of the oceans in the carbon cycle, it is necessary to understand the nuances of specific regional and functional marine ecosystems. The continental shelf of the West Antarctic Peninsula (WAP) is one particularly important regional ecosystem that plays a vital role in the Southern Ocean carbon export. Within the seasonally productive marginal ice zone of the WAP, I sought to identify the long-term drivers of particulate organic carbon (POC) flux. The vast majority of exported POC on the WAP was previously found to be made up of krill fecal pellets. I provide evidence that supports the hypothesis that the inherent life cycle of krill drives the observed 5-year oscillation in POC export. At the end of their life cycle, when krill are at their largest body size, the WAP experiences anomalously high POC export events through the production and sinking of large, carbon-rich krill fecal pellets. Conversely, when krill are young and small, POC export is anomalously low. This pattern shows that ecology exerts a first-order control on the the biogeochemical cycles of the WAP. Upon identifying the source and driver of POC export on the WAP, I set out to determine the role heterotrophic bacteria play in POC flux attenuation. I collected krill fecal pellets on the WAP over three years and measured bacterial metabolic activity in terms of bacterial production and respiration, thereby identifying the amount of organic carbon within the sinking fecal pellets that is lost due to bacteria. Overall, fecal pellet POC turnover rate by bacteria is very low. The relationship between bacteria and POC is complex with each having an affect on the other. Despite varied reactions of the free-living bacterial populations to the presence of krill fecal pellets, a consistent pattern emerged in the concentration of nucleic acid within each bacterial cell. Access to fecal pellets increased the metabolic activity of the free-living bacterial population. This finding shows that the egestion of krill fecal pellets metabolically stimulates the surrounding bacterial community, even though bacteria play a minor role in fecal pellet POC flux attenuation. Though bacteria were found to play a minimal role in organic carbon uptake on krill fecal pellets, they are still vital members of the WAP ecosystem and biological pump. I next sought to identify which bacteria in particular were responsible for colonizing and consuming the fecal pellet POC. Krill fecal pellets were genetically sequenced after timed exposure to the free-living water column bacterial community. I found that there is an endemic population of bacteria that are associated with each population of krill and their fecal pellets. This community of fecal pellet-associated bacteria does not change over time, indicating little colonization by free-living bacteria. Krill fecal pellets, aside from being good agents of POC export, seem to be selective environments for certain specialized copiotrophic bacteria. Further, I find that only a small subset of these endemic copiotrophs actively partake in carbon consumption on krill fecal pellets. Overall, these results show that a small endemic, specialized bacterial community play an outsized role in krill fecal pellet POC degradation and flux attenuation, but that krill fecal pellets remain efficient agents of carbon export to the deep ocean.

Sewage pollution of surface waters is a pressing issue of global concern, even in regions with extensive wastewater and sewage treatment infrastructure. Contaminants, like harmful bacteria that can cause gastrointestinal disease and hinder economic growth and development, enter natural waters through a variety of point and non-point source discharges that range from treated to untreated. With increasing urbanization, aging infrastructure, and changing precipitation patterns due to climate change, it is increasingly important to understand and predict the persistence and transport of sewage-derived bacterial pollution in surface waters. To effectively monitor and predict these contaminants, it is critical to understand sewage-derived bacteria’s extra-enteric ecology, or the ecological dynamics they experience after transitioning from a primary habitat (like the human gastrointestinal system) to a secondary habitat (like natural waters). Dynamics of fecal bacteria are assumed to be driven by loss, as commonly observed for the fecal indicator bacteria (FIB), Enterococcus sp., with sunlight exposure as the dominant driver (i.e., greatest impact on population dynamics). However, particle association of FIB may alter their persistence and transport in natural waters, though this aspect of extra-enteric ecology is rarely included in predictive models. Models predicting persistence and transport of fecal bacteria and pathogens could be improved by incorporating information on the impacts of particle association on dominant loss rates of FIB and the population dynamics of various indicated pathogenic groups. Further, it is important to understand the variation in and drivers of surface water optical properties, like water transparency, due to the likely importance of light penetration to fecal bacteria environmental persistence. This dissertation aims to address the critical knowledge gaps of how particle association influences the extra-enteric ecology of various sewage-derived bacteria and how optical properties relevant to the light-dependent mortality of FIB vary spatially and temporally in an urban-influenced water body. To do so, I employ a combination of empirical, modeling, and observational techniques. The Hudson River Estuary (HRE) is an ideal field site for this research because of its consistent problems with sewage pollution, especially following precipitation events, despite significant improvements following the Clean Water Act in 1972. Managing human health risks associated with sewage pollution is especially important for this water body that runs through the NY/NJ metropolitan area with its 19 million stakeholders. Further, previous research quantifying FIB dynamics has predominantly been conducted in clear, low turbidity water columns. Experiments constraining the dynamics of FIB in water with low clarity, like in the HRE, would fill this important knowledge gap in the field of sewage pollution monitoring and modeling. Chapter 1 assesses the impact of particle association on dominant growth rates and persistence of the brackish fecal indicator bacteria, Enterococcus sp. (also called enterococci). In this chapter, I conducted a series of natural water microcosm laboratory experiments to quantify dominant growth rates of enterococci. I then used these growth rates to parameterize a 1-dimensional advection-diffusion-decay model to simulate enterococci persistence in waters ranging from clear, quiescent lakes to turbid, turbulent waters. This combined empirical and mathematical modeling approach led to four major conclusions related to the persistence and transport of enterococci in natural waters: 1) particle association increases dominant growth rates (light-induced and dark, temperature-dependent growth) and induces sinking of enterococci, 2) particle association increases simulated enterococci persistence, 3) simulated enterococci persist longer in more turbid and/or more turbulent waters, and 4) discharge timing later in a diel cycle increases simulated Enterococcus sp. persistence. Results from this chapter demonstrate the importance of distinguishing free-living and particle-associated Enterococcus sp. in models of their persistence and transport and provide empirical data for independently constraining their population dynamics. Further, the simulated persistence indicates that sewage-derived fecal bacteria discharged into water bodies like the HRE will last longer than discharges in clear, calm waters (e.g. Lake Tahoe) and even clear, turbulent waters (e.g. coastal ocean in California). This information is broadly applicable to water quality management and indicates how variability in turbidity or turbulence within a water body could alter sewage discharge persistence and exposure risk for the public. Model sensitivity testing confirmed the consistent impact of particle association on enterococci persistence and reaffirms the need for FIB models to include particle association. An adapted version of Chapter 1 was published in Water Research (Myers and Juhl 2020). Because particle association increases enterococci persistence and growth rates, it is important to determine if particle association similarly affects co-occurring pathogenic bacteria and if particle association prevalence is similar. Chapter 2 is a valuable complement to Chapter 1 and addresses key knowledge gaps related to ambient pathogen abundance, particle association, and correlation with FIB in surface waters, in addition to the effect of particle association on dominant pathogen growth rates. In this chapter, I report multi-year observations of abundance and particle association proportions in the HRE for four bacterial genera: the fecal indicator Enterococcus sp., two enteric pathogens (Salmonella sp. and Shigella sp.) and a naturally-occurring, marine pathogen (Vibrio sp.). I found that mean particle association ranged from 34% to 49% and that overall abundances were significantly positively correlated across all genera. The second major goal of this chapter was to determine if particle association impacted dominant growth rates of pathogens similarly to the effect observed for enterococci (Chapter 1). In experiments similar to those in Chapter 1, I quantified the fraction-specific (free-living, particle-associated, and total) temperature and light-dependent growth of the three pathogenic genera. Overall, particle association consistently increased temperature- and light-dependent growth rates across genera, similarly to Chapter 1, though particle association did not benefit Vibrio sp. as much as the enteric genera. I found that Salmonella sp. had similar temperature- and light-dependent growth rates to enterococci. By contrast, Shigella sp. growth rates were greater than those of enterococci. As expected due to its different origin, Vibrio sp. also had dissimilar growth rates to enterococci. Interestingly, Shigella sp. behaved more similarly to Vibrio sp., with increasing dark period growth with temperature, which is opposite of the trend observed for the other two enteric organisms (Salmonella sp. and enterococci). The disparities and similarities of dominant growth rates between enterococci and 2 co-occurring fecal pathogens, together with the finding that abundances were positively correlated across all genera, suggests that enterococci are good indicators of recent sewage pollution, but have limitations in their use for assessing extended water column persistence of some co-occurring pathogenic bacteria. Information in this chapter is important for our understanding of FIB use to monitor sewage pollution persistence and for water quality management to minimize human exposure risk, especially in water bodies where environmental persistence is likely longer, like in turbid and turbulent waters (as shown in Chapter 1). Persistence simulations in Chapter 1 demonstrated that water transparency (modeled as diffuse attenuation of light, 3), was critical for determining the persistence of enterococci. It is then important to understand how water transparency varies throughout a water body to eventually predict how sewage bacteria persistence timescales vary. Chapter 3 examines the spatial and temporal variability of water transparency and its primary drivers (suspended particulate matter (SPM), chlorophyll, and colored dissolved organic matter (CDOM)) throughout the terrestrially- and marine-influenced HRE using observational and laboratory techniques. Data in this chapter indicate that water transparency in the HRE is predominantly controlled by SPM (measured as turbidity) and, to a lesser degree, chlorophyll. Despite some dramatic changes in inputs affecting the primary drivers (e.g. decreased sewage pollution - NYCDEP 2012, decreased chlorophyll concentrations - Caraco et al. 1997; Smith et al. 1998, and increased dissolved organic carbon transport - Findlay 2005) in the HRE, the dominance of turbidity in determining water transparency found in this study was consistent with work in the 1980s (Stross and Sokol 1989). Together, the findings by Stross and Sokol 1989 and in this chapter suggest that future work on understanding water transparency variability and its impact on sewage bacteria persistence in the HRE should focus on quantifying the variability in SPM. In contrast to other estuarine systems, CDOM absorption in this chapter minimally impacted water transparency (measured here via KdPAR). This chapter also documents spatial and temporal variability of water transparency and its primary drivers in the HRE. Turbidity and chlorophyll fluorescence varied seasonally, generally consistent with trends in other estuarine systems. Turbidity, chlorophyll, and CDOM absorption were all elevated after increased river flow from Tropical Storm Isaias. All three primary drivers of water transparency were also commonly higher for nearshore and tributary sites, as opposed to mid-channel sites, possibly due to increased shallow bed resuspension and terrestrial runoff. In the upper, freshwater portions of the estuary, CDOM absorption was highest, indicating a greater relative importance of CDOM on water transparency in this region. Data in this chapter also demonstrated that Wastewater Treatment Plant (WTP) outfalls commonly had elevated optical brighteners and (Sl(275-295), a CDOM slope ratio that indicates excess smaller CDOM molecules at a treated discharge, and a contrasting influence on CDOM absorption. This information could then allow CDOM absorption and optical brightener fluorescence to be used as indicators of treated or untreated discharges that could be measured on a faster timescale than current bacteria monitoring via culture-based techniques. Appended to this dissertation are results from experiments examining bacterial community composition for free-living and total populations in the HRE, which provide additional context for the research presented in this dissertation. The free-living and and total bacteria communities were not found to be significantly different, which indicates that there are not distinct communities in free-living and particle-associated fractions. Together with results from Chapters 1 and 2, this indicates that particle association may increase growth rates for bacteria that become particle-associated, instead of particles supporting the development of a unique and more resistant bacteria community. Four of the five fecal core families (bacteria commonly found in human fecal samples) were also identified in samples for these experiments, though relative abundance of these groups for free-living and total fractions were largely uncorrelated with each other over time. This finding demonstrates a notable influence of sewage inputs on the bacterial community in the HRE. Finally, these observations demonstrate that bacteria community composition varied seasonally, as noted by the significant influence of temperature and salinity on bacterial community composition. A wide variety of genera were strongly associated with colder (<12°C) water temperatures and samples from colder water generally exhibited higher alpha diversity. The findings from this dissertation have significantly contributed to our understanding of the extra-enteric ecology of the fecal indicator bacteria Enterococcus sp. and multiple co-occurring potential pathogens. This dissertation demonstrates that particle association must be considered in models of sewage-derived bacteria persistence. This dissertation also deepens our current understanding of water transparency drivers and their spatial and temporal heterogeneity in the turbid and turbulent system of the HRE. The results from this dissertation are useful for improving predictions of sewage pollution persistence and, by extension, minimizing human exposure risk to potentially harmful bacteria. These findings are broadly applicable beyond the HRE, to water bodies of varying turbidity and turbulence conditions.