Assays overview
Specificity and sensitivity . The high degree of molecular differentiation of the three presented probes compared to the homologous regions in the sister species -Hawaiian monk seal- genome suggested high specificity of the proposed assays, assumption that was confirmed in all positive and negative control samples sets. High specificity to target species is a fundamental prerequisite for the dissemination of an innovative method given the conservation relevance of the study species. Our experimental design and trial suggest that there is little room for false positives (except those attributable to lab contamination, for which maximum caution should be posed, as in any experiment that is based on eDNA analysis), while incidence of false negative should be more of a concern, as PCR inhibitors present in environmental matrices is recognized as one the main limitations of eDNA-based detection approaches (Loeza-Quintana et al ., 2020).
Reliability . We reckon that also the samples being positive in only 1 or 2 of the three amplification replicas represent genuine monk seal detection, considering the following points:
- The high homology of the probe to -and only- monk seal DNA (see above), supported by the specific melting temperatures recorded, suggests that any positive detection has to denote monk seal DNA presence.
- Failing amplification in replicas of a positive sample may be ascribable to the extremely low concentration of the target molecule. Marine vertebrate DNA counts for only about 0.004% of the eDNA retrieved from marine samples (Stat et al ., 2017). In the DNA metabarcoding analysis carried out on the ferry sample subset (2018) and same loci, over 90% of the detected reads was attributable to bony fish community (decreasing of one more order of magnitude the incidence of rare vertebrates, such as marine mammals, on the overall marine vertebrate fauna), and no read was attributable to monk seal (Valsecchiet al ., in prep.). Replications (from sampling to DNA amplification) reduces the rate of false negatives; however, the optimal level of replication is difficult to be estimated, since it strongly depends on the detection probability of the target taxon.
- With the exception of MmoM+1, collected in proximity of the entrance of the cave most densely populated by monk seals (n=8 ca) and that returned positiveness for all three loci, all remaining Madeira control-eDNA samples (n=7, 87.5%) resulted positive to monk-seal DNA either only in some sample/reaction replicas and/or for only some of the loci (typically the two 12S-rDNA ones), although seals were definitely around in those waters.
In some cases, a positive detection was found in only one of the sample replicates (each originating from one of the three filters used to process a single sample). This effect can be more easily explained, considering the specific characteristics of the animal and the eDNA state released (Lacoursière‐Roussel & Deiner, 2019; Turner et al ., 2014). Large animals, such as marine mammals, release in the surrounding water biological traces often consisting of large clusters of cells rather than of single or a few cells (Valsecchi et al ., 1998). In this way, when the sample is divided into two or more subsamples, it is likely that the cells, being clustered, are not distributed homogeneously among the various subsamples.
Imperfect detection is an unavoidable feature of most data on species presence/absence even during traditional ecological field studies, when individuals and species that are present at one site are not always all detected, and failure in accounting for imperfect detection may result in biased inference. eDNA assays are clearly imperfect, relying on small amounts of degraded DNA, and still laboratory and sequencing costs limit replication (Ficetola et al ., 2015). However, the high sensitivity and specificity of well-developed eDNA-based approaches can counteract these limitations.
Readiness . Quantitative PCR assays allow to obtain quick outcomes as opposed to DNA metabarcoding approach (requiring library preparation and high-throughput DNA sequencing).