Persistence of marine fish environmental DNA and the influence of sunlight

Abstract

Harnessing information encoded in environmental DNA (eDNA) in marine waters has the potential to revolutionize marine biomonitoring. Whether using organism-specific quantitative PCR assays or metabarcoding in conjunction with amplicon sequencing, scientists have illustrated that realistic organism censuses can be inferred from eDNA. The next step is establishing ways to link information obtained from eDNA analyses to actual organism abundance. This is only possible by understanding the processes that control eDNA concentrations. The present study uses mesocosm experiments to study the persistence of eDNA in marine waters and explore the role of sunlight in modulating eDNA persistence. We seeded solute-permeable dialysis bags with water containing indigenous eDNA and suspended them in a large tank containing seawater. Bags were subjected to two treatments: half the bags were suspended near the water surface where they received high doses of sunlight, and half at depth where they received lower doses of sunlight. Bags were destructively sampled over the course of 87 hours. eDNA was extracted from water samples and used as template for a Scomber japonicus qPCR assay and a marine fish-specific 12S rRNA PCR assay. The latter was subsequently sequenced using a metabarcoding approach. S. japonicus eDNA, as measured by qPCR, exhibited first order decay with a rate constant ~0.01 hr -1 with no difference in decay rate constants between the two experimental treatments. eDNA metabarcoding identified 190 organizational taxonomic units (OTUs) assigned to varying taxonomic ranks. There was no difference in marine fish communities as measured by eDNA metabarcoding between the two experimental treatments, but there was an effect of time. Given the differences in UVA and UVB fluence received by the two experimental treatments, we conclude that sunlight is not the main driver of fish eDNA decay in the experiments. However, there are clearly temporal effects that need to be considered when interpreting information obtained using eDNA approaches.

Fig 1. Deployment of dialysis bags in tank.

23 bags suspended at surface (middle of bags 5 cm beneath water surface) and 22 bags suspended at depth (middle of bags 70 cm beneath water surface).

Fig 2. S. japonicus eDNA concentration (pg/ml seawater) as a function of time (hours) at surface (solid squares) and depth (open circles).

Error bars represent standard deviation of triplicate qPCR reactions; triplicate samples shown as separate symbols at each time point (4 replicates for T6 surface and T7 depth, 1 replicate for T7 surface). Some error bars are small and hidden by overlapping symbols.

Fig 3. Genera identified as present using eDNA metabarcoding over the course of the experiment.

Solid squares indicate presence of the genus in at least 1 biological replicate from surface samples; open circles indicate presence of the genus in at least 1 biological replicate from depth samples.