ObiTools filtering analyses for taxonomic assignments and comparison of the two areas
We applied a first bioinformatic workflow that optimizes the ability to detect identified taxonomic entities. The sequencing reads were processed to remove errors and analyzed using programs implemented in the ObiTools package (http://metabarcoding.org/obitools, Boyer et al. 2016) following a published protocol (Valentini et al. 2016). The forward and reverse reads were assembled using the ILLUMINAPAIREDEND program using a minimum score of 40 and retrieving only joined sequences. The reads were then assigned to each sample using the NGSFILTER software. A separate data set was created for each sample by splitting the original data set in several files using OBISPLIT. After this step, we analysed each sample individually before merging the taxon list for the final ecological analysis. Strictly identical sequences were clustered together using OBIUNIQ. Sequences shorter than 20 bp, or with occurrences lower than 10 were excluded using the OBIGREP program. The OBICLEAN program was then run within a PCR product. We discarded all sequences labelled ‘internal’ that correspond most likely to PCR substitutions and indel errors. Taxonomic assignment of the remaining sequences was performed using the program ECOTAG the sequences extracted from the release 142 (standard sequences) w of the European Nucleotide Archive (ENA). Taxonomic assignments were corrected as follows to be more conservative: for an identification match > 98% identity, we validated a species level, for a 96-98% match, genus level if available and for an 90-96% match, family level if possible. Considering the bad assignments of a few sequences to the wrong sample due to tag-jumps (Schnell et al. 2015), all sequences with a frequency of occurrence below 0.001 per taxon and per library. We further corrected for Index-Hopping (MacConaill et al., 2018) with a threshold empirically determined per sequencing batch using experimental blanks (i.e. combinations of tags not present in the libraries), for a given sequencing batch between libraries (Polanco-Fernández et al. 2021).  From the taxonomic assignment recovered from the ObiTools analyses, we compared the species recovered in each area. We further compared the species recorded by eDNA with other species distribution sources, including a compiled set of species distribution maps for the Caribbean region (Robertson and Van Tassell, 2019). Differences in species recovered between the two areas using eDNA were further compared with those of the UVC transects.