Abstract
Understanding the mechanisms acting on particulate organic matter (POM) and the relative importance of those mechanisms as the material sinks is key to building accurate predictive models of oceanic carbon. POM is also a basal resource for marine food webs; thus, processes affecting POM have implications for higher trophic levels. I use geochemical tools to investigate the major transformation and degradation processes acting on POM, the dominant biological inputs to POM, and how anthropogenic pollutants cycle through POM. Using nitrogen isotopes, I distinguish three dominant mechanisms affecting POM as it settles: remineralization and waste generation by heterotrophs, particle solubilization by microbes, and particle disaggregation. I employ an isotope-based Bayesian mixing model to estimate the composition of fecal pellets, phytoplankton, and microbially degraded organic matter in particles, contributing real-world data to test model estimates particle composition. Using the carbon isotopic compositions of amino acids and phytol, as well as enantiomeric ratios of alanine, I probe the biosynthetic origins of carbon in this same POM. I find that heterotrophic bacteria contribute significant biomass to POM at this site and identify isotopic fingerprints of the stratified phytoplankton community in the upper versus lower euphotic zones. I then apply these techniques to size-fractionated POM samples from a site in the California Current Ecosystem, where I find similarities in the mechanisms of POM cycling, with greater contribution of fecal pellets to large particles in the surface ocean in the latter site. Finally, I explore two ways that this framework for interpreting particle data presents a unique opportunity to interpret distributions of anthropogenic pollutants in marine water columns: how particle alteration mechanisms might affect mercury distributions across particle size fractions and the feasibility of studying persistent organic pollutants in size-fractionated POM.