Can DNA-metabarcoding of multi-species samples from routine egg surveys be used to monitor the ichthyoplankton assemblage?

<p>Broadscale ichthyoplankton surveys covering over 350,000 km² are routinely undertaken in shelf waters off south-eastern Australia to support management of the Commonwealth Small Pelagic Fishery and South Australian Sardine Fishery. Each survey is designed to sample eggs of target species and obtain estimates of spawning biomass using the Daily Egg Production Method. Several hundred samples are taken per survey, but only limited additional information is collected to monitor the marine environment. Our project evaluated the potential for using mtDNA-metabarcoding (mtDNA-mb) from multi-species egg and larval samples to obtain information to support management of Australia’s marine estate, including the spawning habitat of data-poor species and responses of pelagic ecosystems to climate change. Two cytochrome oxidase sub-unit I (COI) mtDNA-mb assays were used to identify eggs and larvae of fish species and compared with a morphological approach. We found that morphological identifications provided overall-estimates of abundance (albeit often only to family level), and required high levels of taxonomic skill, are time consuming and expensive, and can include errors. The two mtDNA-mb assays performed similarly and identified more species and genera compared to the morphological approach. However, mtDNA‑mb only provides a species list per sample with no quantitative abundance information (e.g., sequence read number ≠ abundance). Other limitations include: abundant species may swamp assays causing rarer species to go undetected, and individual species respond to assay amplification differently. The best ways forward may involve hybrid approaches that integrate abundance data from coarse morphological identifications and species lists from mtDNA-mb. </p>