Compatibility of Diatom Valve Records With Sedimentary Ancient DNA Amplicon Data: A Case Study in a Brackish, Alkaline Tibetan Lake

Abstract

Lake sediments represent valuable and widely used archives for tracking environmental and biotic changes over time. Past aquatic communities are traditionally studied via morphological identification of the remains of organisms. However, molecular identification tools, such as DNA metabarcoding, have revolutionized the field of biomonitoring by enabling high-throughput and fast identification of organisms from environmental samples (e.g., sediments and soil). Sedimentary ancient DNA (sedaDNA) metabarcoding, an approach to track the biodiversity of target organisms from sediment cores, spanning thousands of years, has been successfully applied in many studies. However, researchers seldom explore how well the signals from sedaDNA data correlate with the fossil records of target organisms. This information is essential to infer past environmental conditions and community changes of bioindicators when the increasingly popular molecular identification method, metabarcoding, is desired instead of a morphological identification approach. In this study, we explore the correlations of diatom valve records across the last ∼940 years with the diatom sedaDNA metabarcoding data from the same sediment core from lake Nam Co (Tibetan Plateau). Overall, the results from valve vs. sedaDNA data revealed concordant diatom richness as well as community patterns. However, several mismatches in the diatom taxonomic composition existed between the data sets. In general, sedaDNA data harbored much higher diatom diversity, but due to the lack of reference sequences in public databases, many molecular units (amplicon sequence variants) remained unclassified to lower taxonomic levels. As our study lake, Nam Co, is characterized by brackish water and alkaline pH, some likely cases for the observed taxonomic composition mismatches may be due to a valve dissolution issue. Nevertheless, significant drivers for the diatom richness and community structure largely corresponded between data sets. Both valve and sedaDNA data demonstrated similar breakpoints for historical diatom community shifts. A particularly strong shift in the diatom community structure occurred after ∼1950 CE, which may be associated with abrupt environmental changes on the Tibetan Plateau. Altogether, our study indicates that environmentally driven signals reflected by the diatom communities are successfully recovered via microfossil as well as molecular identification methods.