Assay selection and design (in silico)
The peer-reviewed scientific literature was queried for existing assays (i.e. PCR/qPCR/dPCR molecular assays targeting relevant marker/taxon combinations) by searching public repositories (i.e. Google Scholar, Web of Science, Pubmed) using key words “PCR”, “qPCR”, “dPCR” in combination with the common and scientific names of target species, regardless of whether studies used DNA isolated directly from tissue of the target species or environmental samples (e.g. plankton, sediment or water). Subsequently, primer-pairs were aligned to the reference DNA data and manually examined for mismatches with target and off-target species. The same alignments were also used to identify suitable conserved but taxon-specific DNA regions for the design of new primers. The suitability of promising primer sets was tested using Oligoanalyzer™ (IDT, Coralville, IA, USA), following the criteria detailed inTable 2 . Where possible, multiple markers per species were designed/chosen to increase species-specific marker development success rate. To further test for assay specificity, primer pairs that satisfied the above criteria were also evaluated using a machine learning tool that accurately predicts qPCR amplification (eDNAssay), as described in The default settings were used and, as a model output, the lowest assignment probability value (AMP value) was reported including the mismatches with that specific target on each primer. Primer pairs that fitted all relevant criteria and had desirable AMP value (<0.5) (where possible) were retained for subsequent in vitro testing.