Unlocking the past: Genetic clues of historical shifts in forest community composition from ancient DNA in lake sediments

Background/Question/Methods

Paleoecological reconstructions based on fossil pollen records show that climate has driven shifts in forest community composition in North America. Such long-term records are necessary to make better prediction of forest responses to future climatic changes. However, inherent limitations to pollen analysis (long-distance dispersal and similarities in pollen morphotypes among species can potentially limit the inferences we can make about historical shifts in forest composition. Here, we present a genetic tool that potentially addresses these limitations: analysis of chloroplast DNA preserved in lake sediments for thousands of years. We extracted chloroplast DNA (cpDNA) from macrofossils and bulk sediments as old as ca. 6,000 years from lakes in Upper Michigan. To examine different tree species composition across different time periods, species-specific chloroplast markers were first developed from extant tree species present in Michigan to generate species sequence database. Probes or ‘baits’ were designed from these sequences and then used to capture ancient DNA fragments from potentially different species in sediment samples. These captured DNA fragments were enriched and sequenced. Presence of different tree taxa was determined by documenting the presence/absence of species-specific cpDNA haplotypes in each time period.

Results/Conclusions

Results show that we can reliably distinguish different species across various time periods, suggesting that ancient DNA is a useful tool for elucidating long-term forest changes. For example, we identified the presence of multiple tree taxa (e.g. Fagus, Quercus, Acer, Betula) and differentiated between species that share the same pollen morphotypes (Quercus spp.) from bulk sediments and macrofossils from ca. 1500- 2,000 year-old lake sediments. Moreover, as one persistent problem with ancient DNA analysis is low yield, capture and enrichment of ancient DNA fragments resulted to a much higher DNA yield (up to 10 E^7 copies per uL), allowing us to target multiple cpDNA regions for greater taxonomic resolution. These results, when combined with paleoecological data (e.g. pollen and macrofossil analysis), provide a better and more comprehensive signal for reconstruction of past shifts in forest community. Furthermore, we can examine the individual responses of species to past environmental changes as ancient DNA differentiates species sharing similar pollen morphotypes that are often lumped together in pollen analysis.