Measures of prehistoric terrestrial net ecosystem productivity and carbon sink function

Abstract

Indicators of past biological productivity, or ‘palaeoproductivity proxies’, offer ways to indirectly measure key aspects of Earth's deep-time ecosystem and carbon cycle functioning. Given that plants have been the principal primary producers on land for hundreds of millions of years, the abundances of fossil plants in the rock record can indicate past ‘net terrestrial ecosystem productivity’ (NTEP). This is the net carbon storage or release by a terrestrial ecosystem, and a measure of whether an ecosystem is a carbon sink or source. When applied on a global scale, NTEP represents a major component of Earth's carbon cycle. Moreover, since plants are particularly sensitive to rapid climatic events, measuring NTEP with fossil plants should indicate how land carbon sinks have been impacted by these climatic changes. Herein, we compare and contrast three potential proxies of NTEP changes in deep time: total organic carbon (TOC), terrestrial organic microfossil concentrations ($c_t$) and terrestrial organic carbon (TrOC). However, the preservation pathways of terrestrial organic microfossils (hence, $c_t$ and TrOC) are complex and poorly understood. In this review, we have: 1, adapted and applied a framework of modern net ecosystem productivity (NEP) to prehistoric settings by incorporating post-burial effects; 2, summarized the factors that influence the preservation and accumulation rates of land-derived organic carbon in the fossil record; and 3, explored the conditions under which $c_t$ and TrOC may provide valid estimates of prehistoric NTEP. Given their specificity, we conclude that the fossil-based metrics ($c_t$ and TrOC) hold greater potential than TOC as proxies of NTEP. Moreover, $c_t$ and TrOC can be readily applied to the rock record to estimate the relative changes of past NTEP, if several key parameters are standardized between samples: lithofacies, depositional environment, sample thickness (as a surrogate for accumulation time), burial history, basin type and palaeolatitude. In practice, these can be addressed by targeting similar facies within the same basin over relatively short stratigraphic intervals. Moreover, the most precise fossil count techniques should be utilised to minimize data collection error. While these proxies can provide valid metrics of NTEP changes, they are not yet capable of providing absolute burial rates of terrestrial carbon. However, we produce a roadmap towards refined proxies of absolute deep-time NTEP, which would constrain biogeochemical models since the emergence of large land plants >360 million years ago.