Cohesive trends in carbon cycling over the last 66 million years

Paper: Reconciling atmospheric CO2, weathering, and calcite compensation depth across the Cenozoic

Featured image: Figure 1 from a related study: Boudreau et al., 2018 – a schematic which illustrates the carbonate/calcite compensation depth (CCD). Just as snow accumulates on mountains above the snowline and melts at lower elevations, white calcium carbonate shells and minerals (the sinking green discs in this image) accumulate on the seafloor above the CCD and dissolve below this depth.

Authors: Nemanja Komar and Richard E. Zeebe

For multiple decades, we have known that temperatures have largely cooled over the last 66 million years (during the Cenozoic, our current geological era). This insight comes from measuring oxygen isotopes in microfossil shells from ocean sediment cores that extend hundreds of meters into the deep ocean seafloor. Slight increases in the heavier oxygen isotope (which contains ten neutrons) relative to the lighter isotope (which contains eight neutrons) in these shells over time indicates cooling. However, it has been significantly more difficult to understand how the long-term geological carbon cycle has been intertwined with this temperature change. Since carbon and climate are inherently connected under modern and projected future climate change, it is crucial to understand these linkages. A new study by Komar and Zeebe expands a multi-faceted geological carbon and climate model to show how geological and geochemical evidence from ocean sediments that initially appears to be incompatible actually tells a cohesive story of carbon cycling and changes over the Cenozoic.

The authors’ aim was to see whether their long-term carbon cycle model (which they name LOSCAR-P: Long-term Ocean-Atmosphere Sediment CArbon cycle Reservoir with Phosphorus) could simultaneously explain three separate datasets that illustrate three different components of Earth’s carbon cycle over the Cenozoic. These three datasets are estimates of atmospheric carbon dioxide concentrations, the carbon isotope composition of ocean microfossils, and the depth in the ocean where calcium carbonate minerals are almost completely dissolved. This depth is termed the calcite compensation depth, or CCD – calcium carbonate minerals and shells remain intact above this depth but dissolve deeper in the ocean. In theory, the three combined datasets should outline how carbon in the ocean and atmosphere evolved in tandem across the cooling Cenozoic.

However, previous research had noted a striking inconsistency in these records – the CCD has become deeper over the Cenozoic even though decreasing atmospheric carbon dioxide levels and cooling conditions should cause it to become shallower. In theory, cooler climates result in lower weathering of terrestrial minerals, so less dissolved continental material flows into the sea. This decrease in dissolved ions (most crucially calcium) would provide less ability for the shells of calcifying organisms to accumulate on the seafloor. In the geological record, this lowered accumulation of calcium carbonate sediments would appear as a shallowing of the CCD, so the observed CCD deepening has long been a mystery. Komar and Zeebe show that realistic paleoclimate scenarios envisioned with LOSCAR-P can explain this CCD discrepancy, and can also fully explain the history of atmospheric carbon dioxide and microfossil carbon isotope composition when they include two new, key constraints to their model.

First, the authors set their sights on how their model described the eventual fate of sinking organic remains of organisms. Most of this dead carbon from the surface of the ocean is consumed by plankton and bacteria as it sinks down through the water. Previous research has found that the rate of this carbon consumption is not constant, but instead depends on ocean temperature. Therefore, the authors modeled the rate of  organic carbon consumption as temperature-dependent. For example, under warmer climates, organic carbon consumption occurred at a greater rate. This greatly improved the fit of their model data to the observed atmospheric carbon dioxide estimates and microfossil carbon isotope compositions, showing that this temperature dependency is crucial to explaining the carbon cycle over long timescales! However, they still needed to consider additional factors to explain what was going on with the CCD over the Cenozoic.

The second part of the model that the authors targeted to explain the geological CCD observations had to do with the relative proportion of calcium carbonate minerals and shells accumulating in shallow vs. deep regions of the ocean over time, a quantity that they call the shelf-deep partitioning factor. Previous studies had indicated that the relative amount of calcium carbonate deposited in shallow seas was significantly higher in the earlier Cenozoic compared to the present. This is because sea levels back then were much higher than today, submerging more area on shallow continental shelves where sediments could accumulate. The authors term this sea level effect the “longitudinal expansion” of carbonate accumulation in the early Cenozoic. However, also important for their model was a modeled “latitudinal expansion” of carbonate accumulation during this time. Carbonates dissolve under colder seawater conditions, but the authors propose that under the warmer climates of the early Cenozoic, carbonate sediments could remain well-preserved at more polar, colder latitudes. To tie this longitudinal and latitudinal expansion together, they suggest that the rise in diversity of calcifying organisms in the early Cenozoic allowed a much higher number of these calcifiers to spread and accumulate in the warmer, shallower seas near the continents. Therefore, as Cenozoic climate cooled and sea levels dropped, the “shelf-deep partitioning” also decreased as the surface area of shallow, warm seas shrank to a smaller range of latitudes. Even if weathering decreased on land over time (still a hot point of debate) and less overall carbonate sediments could accumulate in the ocean, the CCD had to deepen because the available space for these sediments was now located mostly in the deep ocean rather than the wide expanses of shallow oceans from the early Cenozoic. Once the authors incorporated this carbonate sediment partitioning effect into LOSCAR-P, their model was able to reproduce the CCD observations very well!

A crucial success of this research is that the authors were able to synchronously reproduce three very distinct carbon-cycle datasets with their model, while including new realistic temperature and sedimentation constraints on the model processes. As new geological and geochemical observations come to light, this study will be an important benchmark for scientists who aim to fully capture how carbon and climate coevolve over millions of years. From this study we get a taste of how complex and fascinating Earth’s climate system is: long-term carbon dynamics are affected by temperature, changes in sea level, and biological diversity! What additional factors will future research outline?


Cohesive trends in carbon cycling over the last 66 million years by Lloyd Anderson is licensed under a Creative Commons Attribution 4.0 International License.

Leave a Reply

Your email address will not be published.