ESTIMATING MARINE PHYTOPLANKTON BIOMASS AND PRODUCTIVITY FROM AUTONOMOUS PROFILING FLOATS

Abstract

Knowledge on the biomass and productivity of ocean phytoplankton is fundamental to our understanding of life on Earth. Phytoplankton are autotrophic microbes at the base of the marine food web, that, through photosynthesis, produce organic matter that sustains higher trophic levels – a rate termed net primary productivity. Conventional approaches to measuring the biomass and productivity of phytoplankton often involve the use of satellite remote sensing. Satellites provides daily, global images at kilometer-scale resolution, offering an unprecedented view of the ocean. However, satellites only observe a small portion of the sunlit surface ocean, missing out on biomass and productivity below the surface. In this thesis, I investigate the uncertainties relating to subsurface biomass and productivity by using the fleet of Biogeochemical-Argo floats. These robotic profiling platforms are distributed across the globe and provide proxy bio-optical observations of chlorophyll-a (from fluorescence) and carbon biomass (from particle backscatter) throughout the water-column. In Chapter 2, I assess the quality and quantity of the biogeochemical data collected by the Biogeochemical-Argo program. I provide a census of this data for each the primary variables that the program measures, including chlorophyll-a fluorescence and particle backscatter. I identify interannual trends in data quality, and areas where more data could be collected in the future. In Chapter 3, I design a method for estimating net primary productivity from daily cycles of particulate carbon. In this approach, I construct the daily cycle of particulate carbon from quality-controlled particle backscattering taken at ~5 or 10 days intervals. I demonstrate that the primary productivity inferred from daily cycles varies seasonally and regionally, producing estimates that are comparable to satellite models. With this chapter, I argue that this approach could 1) constrain uncertainties in satellite-based models with regard to the vertical structure of productivity, and 2) identify climate-related, basin-scale trends in ocean productivity. In Chapter 4, I use the global BGC-Argo array to estimate Earth’s stock of phytoplankton. I also describe the phenology and biogeography of phytoplankton carbon and chlorophyll-a. I highlight how in the vast majority of the ocean the spatiotemporal distribution of carbon substantially differs from the metric of chlorophylla, which is commonly used as a proxy for phytoplankton biomass. With these results, I make the point – like others have before – that to properly describe the basic naturalistic tendencies of Earth’s phytoplankton stocks, the proper metric of carbon must be used and must include information from throughout the water-column. The combination of these chapters underscores how profiling robots can provide a more accurate, holistic view of ocean phytoplankton.