Dynamic sedimentary ancient DNA models of Fennoscandic Holocene plant communities reveal the role of temperature and competition

Ecosystems are constantly responding to shifting environmental conditions, and the observed structure of ecological communities at any given discrete observation timepoint is part of a larger dynamic response. However, we lack long-term time series of ecosystem responses allowing understanding of the processes driving community dynamics. Sedimentary ancient DNA (sed aDNA) offers novel opportunities to link empirical data with ecological processes, providing near continuous records of plant and animal community changes over millennia. Here, we analyzed metabarcoding data from 10 lakes from Northern Fennoscandia characterizing plant and mammal communities at temporal resolution of 100 years over several millenia. We used sed aDNA data to test how biotic processes, including temperature-dependent plant growth, competition, shading, and herbivory, may have driven the dynamics of Fennoscandic Holocene plant communities since the Last Glacial Maximum. We compared the explanatory power of ordinary differential equation-based models simulating community changes and including these processes. The fit of the models improved with the consideration of temperature dependent densification and biotic interactions, most notably shading. When considering individual taxa, the importance of shading and herbivory was greatest for the dwarf shrubs, while tree and shrub taxa were more often best predicted by models without competition. Our study demonstrates how using sed aDNA time series together with process-based inverse modelling can uncover the mechanisms of species response to climate change, offering potential for more realistic predictions.