Environmental DNA (eDNA) is sensitive to decay during water sample collection, potentially influencing molecular species detection. When eDNA water sample processing is delayed during remote field collections, whole water samples can be stored frozen to minimize further eDNA degradation. However, the freezing and thawing processes may influence eDNA integrity thus affecting species identification. We split bulk seawater samples to compare the marine taxa from water that had been filtered immediately to water that had been stored frozen for one month. Using DNA metabarcoding we evaluated the effect of freezing on the inferred marine invertebrate communities based on a 313bp section of the cytochrome c oxidase subunit I (COI) mitochondrial barcode gene. Previously frozen samples had significantly lower species richness and different community composition compared to the non-frozen samples. Particularly, amplified sequence variants (ASVs) of the copepod Pseudocalanus were significantly more abundant in non-frozen samples. These data suggest that freezing water samples intended for eDNA analyses may impede species detection for select organisms and significantly alter the inferred community structure, however this effect is not consistent across genes. Studies using pre-frozen water samples should consider downstream implications, such as PCR bias and reduced species detection, when analyzing eDNA biodiversity.
10.21966/bm17-c870
