Disentangling the effects of glaciers and climate on vegetation is complicated by the confounding role that climate plays in both systems. We reconstructed changes in vegetation occurring over the Holocene at Jøkelvatnet, a lake located directly downstream from the Langfjordjøkel glacier in northern Norway. We used a sedimentary ancient DNA (sedaDNA) metabarcoding dataset of 38 samples from a lake sediment core spanning 10,400 years using primers targeting the P6 loop of the trnL (UAA) intron. A total of 193 plant taxa were identified revealing a pattern of continually increasing richness over the time period. Vegetation surveys conducted around Jøkelvatnet show a high concordance with the taxa identified through sedaDNA metabarcoding. We identified four distinct vegetation assemblage zones with transitions at ca. 9.7, 8.4 and 4.3 ka with the first and last mirroring climatic shifts recorded by the Langfjordjøkel glacier. Soil disturbance trait values of the vegetation increased with glacial activity, suggesting that the glacier had a direct impact on plants growing in the catchment. Temperature optimum and moisture trait values correlated with both glacial activity and reconstructed climatic variables showing direct and indirect effects of climate change on the vegetation. In contrast to other catchments without an active glacier, the vegetation at Jøkelvatnet has displayed an increased sensitivity to climate change throughout the Middle and Late Holocene. Beyond the direct impact of climate change on arctic and alpine vegetation, our results suggest the ongoing disappearance of glaciers will have an additional effect on plant communities.
Sedimentary ancient DNA (sedaDNA) has rarely been used to obtain population‐level data due to either a lack of taxonomic resolution for the molecular method used, limitations in the reference material or inefficient methods. Here, we present the potential of multiplexing different PCR primers to retrieve population‐level genetic data from sedaDNA samples. Vaccinium uliginosum (Ericaceae) is a widespread species with a circumpolar distribution and three lineages in present‐day populations. We searched 18 plastid genomes for intraspecific variable regions and developed 61 primer sets to target these. Initial multiplex PCR testing resulted in a final set of 38 primer sets. These primer sets were used to analyse 20 lake sedaDNA samples (11,200 cal. yr BP to present) from five different localities in northern Norway, the Alps and the Polar Urals. All known V. uliginosum lineages in these regions and all primer sets could be recovered from the sedaDNA data. For each sample on average 28.1 primer sets, representing 34.15 sequence variants, were recovered. All sediment samples were dominated by a single lineage, except three Alpine samples which had co‐occurrence of two different lineages. Furthermore, lineage turnover was observed in the Alps and northern Norway, suggesting that present‐day phylogeographical studies may overlook past genetic patterns. Multiplexing primer is a promising tool for generating population‐level genetic information from sedaDNA. The relatively simple method, combined with high sensitivity, provides a scalable method which will allow researchers to track populations through time and space using environmental DNA.
Metabarcoding is the most widely applied method for studying plant communities using environmental DNA, with shotgun sequencing and capture probes being alternative methods that aim to retrieve multiple markers or genome‐wide information. Any method's ability to detect and correctly identify plant taxa varies with DNA preservation, DNA reference library, and the diversity of the local flora, making it difficult to compare results from different environments. Here we compare these three methods using lake surface‐sediments from Northern Fennoscandia with the PhyloNorway genome skim reference library (1500 taxa) that includes nearly all species of the regional flora. We also undertook vegetation surveys from around the lakes to estimate the true positive detection rate, identify false positive detections, and provide optimal filtering cut‐off thresholds for the three methods. Applying these thresholds, the rate of false positives was too high for reliable identification at the species level based on shotgun (49%) and capture probes (62%), whereas it was low for metabarcoding (5%–12%). All methods were reliable at genus and family levels after applying the optimal filtering thresholds (< 4% false positives). Our results show that in these lake sediments, metabarcoding on average detects 2.1 times as many true positive taxa as shotgun sequencing and 6.4 times as many taxa as capture probes. The proportion of a taxon's sequenced reads for the metabarcoding and shotgun methods was significantly related to the taxon's abundance category from the vegetation surveys, but this was not the case for capture probe data. We expect the false positive rate of shotgun sequencing to decrease with increasing genome completeness in the reference libraries and the method to be advantageous for highly degraded DNA with fragments too short for metabarcoding. At present, metabarcoding provides the highest detectability and taxonomic resolution for correct identification and quantification of vascular plants.
Inferring community composition from shotgun sequencing of environmental DNA is highly dependent on the completeness of reference databases used to assign taxonomic information as well as the pipeline used. While the number of complete, fully assembled reference genomes is increasing rapidly, their taxonomic coverage is generally too sparse to use them to build complete reference databases that span all or most of the target taxa. Low‐coverage, whole genome sequencing, or skimming, provides a cost‐effective and scalable alternative source of genome‐wide information in the interim. Without enough coverage to assemble large contigs of nuclear DNA, much of the utility of a genome skim in the context of taxonomic annotation is found in its short read form. However, previous methods have not been able to fully leverage the data in this format. We demonstrate the utility of wholeskim , a pipeline for the indexing of k‐mers present in genome skims and subsequent querying of these indices with short DNA reads. Wholeskim expands on the functionality of kmindex , a software which utilises Bloom filters to efficiently index and query billions of k‐mers. Using a collection of thousands of plant genome skims, wholeskim is the only software that is able to index and query the skims in their unassembled form. It is able to correctly annotate 1.16× more simulated reads and 2.48× more true seda DNA reads in 0.32× of the time required by Holi , another metagenomic pipeline that uses genome skims in their assembled form as its reference database input. We also explore the effects of taxonomic and genomic completeness of the reference database on the accuracy and sensitivity of read assignment. Increasing the genomic coverage of the genome skims used as reference increases the number of correctly annotated reads, but with diminishing returns after ~1× depth of coverage. Increasing taxonomic coverage clearly reduces the number of false negative taxa in the dataset, but we also demonstrate that it does not greatly impact false positive annotations.
Metabarcoding is the most widely applied method for studying plant communities using environmental DNA, with shotgun sequencing and capture probes being alternative methods. Any method’s ability to detect and correctly identify plant taxa may vary with DNA preservation, DNA reference library and the size of the local flora, making it difficult to compare results from different environments. Here we compare these methods using lake surface-sediments from Northern Fennoscandia with the PhyloNorway genome skim reference library (1500 taxa) that includes almost all species of the regional flora. We also undertook vegetation surveys from around the lakes to estimate the true positive detection rate, identify false positive detections, and provide optimal filtering cut-off thresholds for the three methods. Optimal filter cutoffs were one PCR replicate and 3 reads for metabarcoding, and 3.2×10e-6 and 5.6×10e-6 of queried reads, respectively, for the shotgun and capture probe data. Applying these thresholds, the rate of false positives was too high for reliable identification at species level based on shotgun (49%) and capture probes (62%), whereas it was low for metabarcoding (5-12%). All methods were reliable at genus and family levels after applying the optimal filtering thresholds (<4% false positives). Our results show that metabarcoding on average detects 2.1 times as many true positive taxa as shotgun sequencing, and 6.4 times as many taxa as capture probes. Proportions of filtered metabarcoding and shotgun reads were significantly related to abundance categories from vegetation surveys, but this was not the case for capture probe data. We expect the false positive rate of shotgun sequencing to decrease with increasing genome completeness in the reference libraries, and be advantageous for highly degraded DNA with fragments too short for metabarcoding. At present, metabarcoding provides the highest detectability and taxonomic resolution for correct identification of vascular plants.