Is DNA metabarcoding an option for formaldehyde-preserved zooplankton time series?

The recovery of amplifiable DNA from formaldehyde{square}fixed (FF) zooplankton samples has long been considered unfeasible. Nevertheless, advances in DNA sequencing and methods for retrieving highly degraded genetic material have demonstrated that even million{square}year{square}old samples and FF museum specimens can yield usable DNA. To access the biological information preserved in long{square}term zooplankton time series, we assessed methodologies for extracting amplifiable DNA from community samples stored for up to 28 years in formaldehyde at room temperature. On one hand, we report the failure of a method previously described as successful for FF zooplankton samples, likely due to the cold{square}storage conditions (4{square}{degrees}C) used in the original study. On the other hand, by adapting two extraction protocols designed for FF museum specimens--representing harsher and softer alternatives (HHA and HPC, respectively)--we successfully amplified and sequenced a subset of FF zooplankton samples. As expected, DNA integrity and sample pH were inversely related to preservation time, and only short DNA fragments were recovered, ruling out the use of commonly employed [≥]300{square}bp metabarcoding markers. While DNA integrity appeared to be a better predictor than DNA yield for amplification success, the presence of a gel band of the expected size did not always guarantee congruence with microscopy{square}based assessments. Although amplifiable DNA was recovered from most samples, including some of the oldest, community compositions concordant with microscopy were consistently recovered only from samples preserved for up to two years. Beyond this point, the HHA and HPC methods produced divergent results, reflecting a trade{square}off between the removal of formaldehyde{square}induced cross{square}linkages and the avoidance of additional DNA damage. Among the small universal markers tested ([~]120-170{square}bp), including one nuclear rRNA marker and two mitochondrial markers, only the 18S rRNA V9 region consistently amplified. We conclude by providing a set of recommendations aimed at improving the methods presented here.