Assessing metabarcoding methods for re-analysis of pond DNA samples: Towards a new metric for assessment of pond condition.
Abstract
The aims of this study were to determine whether the quality and quantity of plant and eukaryote DNA present in the district level licensing (DLL) samples is sufficient to allow the re-analysis of the samples using amplicon-based metabarcoding using novel sequencing methodologies suitable for generating longer, more taxonomically informative sequence data. A total of 120 DLL samples collected across England, UK, were included in this study as part of the great crested newt (Triturus cristatus) eDNA District Level Licensing (DLL) survey. The quality and quantity of DNA present in samples collected under the DLL scheme was assessed using three different assays: capillary electrophoresis to determine DNA integrity, fluorimetry to quantify double-stranded DNA and UV-VIS spectrophotometry to determine the purity of DNA extracts. This was benchmarked against the quality of a subset of community DNA samples collected from an artificial pondscape using methodologies optimised for community DNA metabarcoding. Next, amplification of metabarcoding regions using primer sets of varying amplicon length (700, 1100 and 4500 basepairs) was attempted in order to determine the performance of the DLL samples using amplicon-based methods for ONT metabarcoding. This was followed by the preparation of sequencing libraries from successfully amplified samples, sequencing on the ONT MinION platform, and bioinformatic analysis of sequence data output in order to determine the suitability of the DLL samples for metabarcoding analyses. Quality control assessment of the DLL samples showed that while intact, high molecular weight DNA was present in many samples, DNA concentration was very low and the concentration of contaminant compounds and non-target nucleic acids was high, especially in contrast to the community DNA samples which have high DNA concentrations and few impurities present. Attempts at amplification from the DLL samples using all three primer sets were unsuccessful using best practice PCR conditions and was only possible using a universal eukaryote primer set when using a very high number of PCR cycles. Library preparation and sequencing using an ONT MinION sequencer was attempted; however, carryover of inhibitory compounds affected library preparation efficiency and greatly reduced sequencing success. Analysis of data produced showed very low diversity to be recovered from the samples. This could be as a result of the small volume sampled, the nature of sample processing and required amplification conditions and was likely impacted by the limited success of sequencing using these methods.