About CBIOMES

Microbial communities in the sea mediate the global cycles of elements including carbon, sulfur, and nitrogen. Photosynthetic microbes in the surface ocean fix these elements into organic molecules, fueling food webs that sustain most other life in the ocean. Organic matter is transported into the deeper, dark ocean, where it is consumed and respired by microbes, maintaining a reservoir of carbon substantially larger than the atmospheric inventory of CO₂. The organisms that sustain these global-scale cycles are functionally and genetically diverse, non-uniformly distributed and sparsely sampled in space and time.

The Simons Collaboration on Computational Biogeochemical Modeling of Marine Ecosystems (CBIOMES, pronounced “Sea Biomes”) is a multi-institution effort seeking to characterize biogeography of key marine microbes at basin and seasonal scales, to ask how organismal characteristics and interactions shape these patterns, and to understand and quantify the relationship between biogeography and elemental cycles. To this end the team is developing models that represent key traits of marine microbes, compiling and interpreting diverse observational datasets, and formally combining data and models in order to map out microbial biogeography, interpret the organizing principles, and infer large-scale biogeochemical fluxes.

The project brings together an eclectic, multi-disciplinary group of investigators from oceanography, statistics, data science, ecology, marine biogeochemistry and remote sensing from institutions across the US, the UK, Germany and Canada. We connect with and complement other Simons Collaborations including SCOPE, SCOPE-Gradients and PriME.

Project Motivation and Goals

With four key overarching scientific goals, CBIOMES seeks to:

• better characterize the biogeography of marine microbial populations.
• develop interpretive and predictive mathematical and computational models of the system.
• develop and implement biologically meaningful representations of organisms in those models and simulations.
• work towards ecological ocean data assimilation, combining observations and models to produce state estimates of the planktonic ecosystem.

But why these goals?

Mathematical Models and Simulations

Despite the fascinating ecology, mediation of the global carbon cycle, and climatic significance of marine microbial systems, our ability to interpret and predict their structure and function with mathematical theory and simulations remains limited.

Most current models of marine ecosystem structure and carbon cycling remain rooted in ‘black box’ mathematical parameterizations developed in the 1960s or earlier. Yet, because the observational characterization of ocean ecosystems and biogeochemistry is still sparse, such models remain important tools to help quantify and interpret the integrated pools and flows of carbon in the ocean. However, the current generation of ocean carbon cycle models is relatively crude and lacks mechanistic veracity.

One example of something that is missing is any explicit and mechanistic representation of the variable elemental composition of organic particulate (a cornerstone of global carbon cycling). We observe marine microbes to be highly flexible in their trophic strategies though efforts to describe and interpret this with models and simulations remain very modest. Current basin- and global-scale simulations of marine microbial systems and biogeochemical cycles generally do not attempt to resolve, interpret or even parameterize, the extremely rich observed functional and taxonomic diversity. Likewise, principles of conservation of mass, energy, and other quantities enable the powerful interpretive and predictive capabilities of theory and simulations in many branches of physics, but we do not yet exploit the available conservation laws for energy and electron flow in ecological and biogeochemical simulations.

In summary, to advance the value of biogeochemical and ecological models we must understand and capture the essential organization of marine microbial ecosystems and how they shape key system characteristics including the elemental composition of sinking particulate matter that regulates carbon sequestration, nutrient cycles, and more. This, in turn, will allow us to quantitatively describe the system we observe today, interpret past changes, and inform explorations of future scenarios.

Empirical Characterization

The foundation for the development of theory and models is the observational characterization of marine microbial biogeography. To gain interpretive and predictive skill with respect to the large-scale oceanographic distribution of microbial communities and their function, we need baseline data sets that characterize current biogeography at basin and global scales.

Relevant data sets generated via underway sampling, and autonomous platforms, along with the establishment of continuous global remote-sensing over the past decades are rapidly expanding the potential for large-scale observational coverage and model-data synthesis. Such data provide excellent opportunities to develop, test, and constrain biogeochemical and ecological models. However, there are hurdles that hinder the optimal use of these data.

Firstly, many highly relevant data were published before electronic repositories were common and require compilation and digitization. Secondly, interdisciplinary studies require data from diverse sources, with relevant data commonly dispersed across multiple repositories. Thirdly, unlike physical or chemical data, measures of microbial populations and activity come in many forms that are not always easy to inter-calibrate.

It is a challenge to translate these varied forms of biological information – e.g., cell counts from microscopy or flow cytometry, pigment densities, and a variety of ‘omic metrics – into a common biomass currency. Such an inter-calibration would allow the resulting theoretical models and simulations to capitalize on the powerful conservation constraints (mass, energy, electron flow).

In short, the generation, compilation, and accessibility of large-scale marine microbial biogeography remain important goals and, to this end, inter-calibration and innovation present key challenges and opportunities which could enable us to simultaneously exploit diverse metrics of population size and structure.

Model-data Synthesis

Over the past few decades, models and observations of the fluid Earth environment have matured. In the physical and chemical realms, synthesis of models and observations (data assimilation) has come to fruition, driven in part by the clear economic value of meteorological forecasting, which in turn has motivated the development of geophysical fluid dynamics and atmospheric physics simulations, as well as global-scale observing platforms, including remote sensing.

Computational models provide a means to interpolate sparse observations while bringing to bear powerful conservation constraints and algorithmically encoded knowledge of system dynamics. This blending of data and models provides quantitative characterization of dynamic systems (state estimates) including weather and climate: our best estimates of the physical state of the atmosphere over the past several decades, with quantified uncertainties, as well as our daily weather forecast.

Since the 1990s and the World Ocean Circulation Experiment global survey, physical oceanography has followed suit with global state estimates of ocean circulation and climate. In parallel, the associated Joint Global Ocean Flux Study has characterized chemical distributions on the global scale, with the associated development of biogeochemical models having, in recent years, enabled state estimation for ocean chemistry. One of our goals in CBIOMES is to develop regional and global-scale ocean state estimation derived from emerging microbial biogeography and trait-based ecological modeling platforms.