Current climate change is a major threat to marine biodiversity with severe effects on marine ecosystem function and stability. Phytoplankton, at the base of marine food webs, already start showing shifts in species composition and abundance as a consequence.
The PhytoArk project, financed by the Leibniz Association, was born in 2021 as a collective of scientist from diverse institutions all over Germany and international cooperation partners in Sweden, Finland and the United States. The main ambition of PhytoArk is to predict future impacts of global warming on the Baltic Sea ecosystem by studying the biodiversity patterns of marine primary producers up to the Holocene period. Namely, the biodiversity in marine sediments is unlocked through a combination of DNA metabarcoding, paleo-genomics, molecular organic proxies, resurrection ecology and ecological modeling shared among the several work packages of PhytoArk. Such an integrative approach allows for a leap through the history of the Baltic Sea in a “time machine” fashion.
The PhytoArk Work Package 1 is led by Prof. Dr. Miklós Bálint with the goal of evaluating the long-term impact of climate change on the structure, composition and function of Baltic phytoplankton communities from sediment cores, by comparing sedimentary ancient DNA (sedaDNA) with established paleoecological biomarkers.
The current DAAD-financed research proposal aims at complementing the work package 1 by reconstructing microbial communities from the Baltic Sea through the Holocene history. Namely, this project plans on estimating the biotic interactions going through the benthic communities into a temporal succession of trophic networks. Using marine sediments as natural archives that store the traces of benthic communities over the course of geological ages, information not only on species presence, but also on their co-occurrence are encoded in sedaDNA. So, the imprint of the “ghost of past” biotic interactions can be retrieved from sedaDNA time series, resulting in ecological networks. The community-wide response to environmental change will be inferred with a two-folded analysis of the co-occurrences:
- Latent Variable Models (LVMs) together with Joint Species Distribution Models (JSDM), predicting the occurrences of a species by using both environmental covariates and the occurrences of all other species. (Figure 1)
- Integrating taxonomic and functional information for a trait-based network analysis, testing the relationship between changes in network topology with environmental change through generalized linear models. (Figure 2)
Figure 1: Example of residual correlation matrix based on latent factor loadings for a LVM with environmental covariates (from Niku et al. 2019, Methods Ecol Evol, Volume: 10, Issue: 12, Pages: 2173-2182)
Figure 2: Example of taxonomic co-occurrence network (modified from Seymour et al. 2020, Science of The Total Environment, Volume: 729, Issue: 138801)
JSDM based on LVM loadings will disentangle the co-occurrence patterns due to variation in species occurrence from the co-occurrence due to the influences of environmental variables and species-to-species associations.
Said results will tell the role played by biotic interactions and environmental change in assembling the Baltic plankton communities.
Trait-based interaction network will assess the response of the marine benthic communities to environmental changes from the emergent properties of the ecological network.
Thus, a framework to reconstruct interaction networks of past ecosystems from sedaDNA time series will provide an informative tool for understanding the consequence of climatic alterations to the Baltic Sea ecosystem in the future.