Marine microorganisms are a key vector in global carbon cycling, supporting an annual flux of 5 – 12 gigatons of carbon to the ocean interior via the biological carbon pump. While methodological advances over the last half century have greatly advanced our understanding of the factors influencing variability in this flux, the contributions of individual components in the microbial food web remain poorly resolved. Utilizing a combination of laboratory, field and remote sensing studies, this dissertation addresses several different aspects of this challenge. In the second chapter, unsupervised learning methods are applied to a global bio-optical data set from biogeochemical Argo floats to identify six oceanic biomes characterized by distinct seasonal trends in vertical phytoplankton distributions.



This study demonstrated the great potential for using data from autonomous profiling floats to generalize seasonal trends in vertical phytoplankton distributions across vast regions of the global ocean, while also providing new insight on the hydrological and biogeochemical drivers of this variability. The third chapter reports the development of a novel method for the direct measurement of chlorophyll a attributable to individual phytoplankton groups in natural samples via cell sorting by flow cytometry. Critically, this approach makes it possible to evaluate phytoplankton community structure in terms of a parameter measured by autonomous platforms, while simultaneously quantifying sources of variability not captured by existing methods.



The fourth chapter investigates the environmental drivers of phytoplankton distributions within the Western Tropical South Pacific, providing a case study for the biogeographical provinces identified in chapter 2 while also investigating how biogeochemical gradients influence linkages between heterotrophic groups central to carbon cycling within the microbial food web. Chapter five reports series of experiments investigating cell physiology as a driver of predator-prey interactions between heterotrophic bacteria and algal phagomixotrophs—eukaryotic algae that supplement requirements for carbon and/or nutrients by ingesting smaller cells.



By validating the predictions of a gene-based model of algal trophic modes, the results from these experiments point toward the potential widespread occurrence of phagomixotrophy amongst green algae, while highlighting potential sources of bias in field and laboratory studies of bacterivory. With global climate change expected to produce rapid changes in ocean circulation and biogeochemistry, the urgency of understanding the role of marine microbes in global biogeochemical cycling has never been greater. This dissertation represents an advance in this larger goal, providing an expanded framework for the broad distribution of microbial communities in addition to novel insight into the environmental and physiological drivers of microbial community structure from the global to cellular scale.