Project 3: Energy landscapes.

Deep-sea sediments and subseafloor environments host slowly growing, low-density microbial populations. Microbial persistence in systems with such extreme energy limitation remains poorly understood. Hydrothermal vents, on the other hand, are veritable oases in the deep-sea desert, supporting rich biological communities with primary productivity driven by chemosynthetic autotrophs. Energy landscapes in hydrothermal vents are shaped by chemical gradients resulting from the mixing of seawater and vent fluids.

Sampling of hydrothermal fluids from UiB - Universitetet i Bergen on Vimeo. Sampling of hydrothermal fluids pooling under a flange at the Loki’s Castle vent field, at nearly 2500 m depth, during the 2017 summer cruise. Copyright: K.G. Jebsen Centre for Deep Sea Research

Numerical models are an important tool for studying deep-sea hydrothermal and subseafloor systems given the difficulties of in situ sampling and culturing. Data for this project come from ongoing studies of hydrothermal vents and associated sediments along the Arctic Mid-Ocean Ridge. Geochemical data include concentrations of electron donors/acceptors from sediment cores, seawater, and vent fluids. Microbial data include relative abundance of taxonomic groups based on 16S rRNA gene copy numbers, and functional gene abundances from metatranscriptomic sequencing. Given the available observations, we turn to modeling approaches to shed light on mechanisms underpinning statistical associations in the data.

In this project, we build models that capture the dynamic interactions between geochemical energy landscapes and microbial populations. By coupling bioenergetic parameters such as catabolic rates, maintenance energy, and biomass yield with microbial population growth, we wish to predict the relative proportion of microbial metabolisms, their spatial distribution, and the impact of microorganisms on the composition of hydrothermal fluids. However, given the range of uncertainties regarding the most appropriate model choice and the extent to which available data can help constrain unmeasurable parameters (e.g. flows, or fluxes), this project will also address the inverse problem: what are the most probable models and their associated uncertainty?