MODELING NET PRIMARY PRODUCTIVITY IN ANCIENT TERRESTRIAL ECOSYSTEMS: INVESTIGATIONS INTO THE CRETACEOUS CRITICAL ZONE

Net primary productivity is a measure of the annual flux of carbon that an ecosystem experiences through autotrophic processes. This value is a master control on ecosystem biogeochemistry, is used to assess spatiotemporal variation in carbon cycling, and is an important factor in many global climate models. However, obtaining a value of net primary productivity (NPP) for ancient ecosystems from deep-time terrestrial sediments has proven difficult. Previous workers have proposed various proxies to estimate NPP of paleo-ecosystems, all of which rely upon geochemical analysis of paleosols to estimate mean annual temperature (MAT) and mean annual precipitation (MAP). While empirical models demonstrate that MAT and MAP are valuable predictors of NPP in modern ecosystems, such relationships likely fall short for paleo-ecosystems in which the plant communities might be vastly different from modern analogs.

Based upon the reciprocal relationship between soil carbon stocks, NPP, and litter decay rates that was established by Todd-Brown et al. 2013, we present a simplified model that uses total organic carbon (TOC) values to estimate NPP. While not used for previous proxies of paleo-NPP, paleosols do contain evidence of the plant communities that they supported in the form of pyrogenic carbon and recalcitrant organic carbon, which in modern soils constitutes up to 80% of the TOC. Therefore, in exceptionally preserved paleosols, TOC may be assumed to be representative of the carbon stocks of that soil prior to burial. We applied our model to modern data sets from the Eastern U.S., limiting that application to regions with geomorphologies that encourage preservation of paleosols. Our model was also applied to data from the Big Cedar Ridge (BCR) plant locality in Wyoming. BCR, in the Campanian Meeteetse Formation, preserves in situ plant fossils and the paleosol that supported them beneath an ashfall. Finally, we explored the limitations of our simple model on paleosol samples from the Early Cretaceous Western Interior Basin that were neither exceptionally preserved nor sampled with this model in mind.