Integrative Computational Tools for Mapping and Projecting Ocean Phytoplankton

Marine phytoplankton are dominant contributors to global primary production and hence play a major role in carbon cycling. Gephyrocapsa huxleyi, the most abundant and globally-distributed species of calcifying phytoplankton (coccolithophore), is known to have vast genomic variability that allows it to grow in very different environments. We grew 12 different strains of this diverse coccolithophore in the lab and coupled genome sequencing with physiological and gene expression measurements. We used these results to predict the habitat of diverse genetic variants in a global biogeochemical model. Our modeling exercise demonstrates how intraspecific trait variation can help predict the distributions of critical phytoplankton taxa, and how their distributions may shift with changing environmental conditions. I will also show how time-series community level gene expression data coupled to statistical techniques can reveal shifts in coccolithophore habitat. Via an environmental dataset from Cape Cod Bay, I will demonstrate how the phosphorus use physiology of coccolithophores responds to phosphate availability. I will discuss how these findings are relevant for ecosystem and physiological modeling, and hence why collecting time series -omics measurements is an important approach for documenting shifts in phytoplankton biology and identifying mechanisms implementable in models.