Hidden biodiversity in Neotropical streams: DNA barcoding uncovers high endemicity of freshwater macroinvertebrates at small spatial scales

the manufacturer’s instructions. A standard sequencing protocol [40] was used to amplify the full-length 658 base pair (bp) of the COI barcode region using the following primers sets: LCO1490/HCO2198 [41] and LepF1/LepR1 [42]. All PCR products were verified on a 1% aga- rose gel, and purified with EXO-SAP-IT (USB Corp., Cleveland, Ohio, U.S.A.). The protocol included adding 1 μl of 1 U/μl Shrimp Alkaline Phosphatase (SAP), 0.5 μl of 20 U/μl Exonucle- ase I (EXO), and 5 μl of amplified product, and then incubating as indicated in the manufac- turer’s protocol [43]. This product was sequencing using an Applied Biosystems Genetic Analyzer (ABI 3130xl, Applied Biosystems, Carlsbad, California). Sequences were aligned in Geneious V7.03 [44], using the MAFFT 7.313 [45] tool and the L-INS-i algorithm. Sequence alignments were also inspected by eye in Geneious to confirm overall sequence quality. We did not find gaps or stop codons in any of the sequences. Project sequences, together with the information on collected specimens, are available on the Barcode of Life Data System (BOLD systems, under project code INVPA: http://www.boldsystems.org/). Project sequences are also available in GenBank (accession numbers: KX039451-KX039650, KU980966-KU981004). Data analysis To confirm morphological identification for our sequenced specimens, we performed BLAST searches for publicly available sequences in GenBank. We created a final dataset comprising unique haplotypes from our study and including three COI sequences retrieved from Genbank

that were used as an outgroup. The retrieved sequences were Thermobia domestica (GenBank NC006080), Atelura formicaria (GenBank NC01119), and Tricholepidion gertschi (GenBank NC005437). We estimated phylogenetic relationships among taxa using maximum likelihood (ML) searches in IQ-TREE v 1.6 [46] and Bayesian inference (BI) in BEAST v 2.4.6 [47] as implemented on the CIPRES Science Gateway [48]. The best-fit model of nucleotide substitution for the dataset, selected using jModelTest 2.0 [49] based on the Bayesian Information Criterion, was GTR+I+G. To determine node support for the IQ-TREE we used 10000 ultrafast bootstraps [50] and 1000 Shimodaira-Hasegawa-like approximate likelihood ratio test replicates [51]. BI analysis was exe- cuted with an uncorrelated lognormal relaxed clock and coalescent prior, with the default settings of BEAUti for the remaining parameters. We performed two runs of 2.0×107 generations, and sampled trees every 5000 generations. Trace logs and species trees for the two runs were combined using LogCombiner v 2.4.8 [52]. We used Tracer v. 1.6 [53] to ensure that effective sample size (ESS) values for all parameters were above 200 and to determine the burn-in. Finally, output trees were summarized as maximum clade credibility (MCC) trees using mean node heights after dis- carding 25% of generations as burn-in using TreeAnnotator v1.8.4 [54]. We then assessed the diversity of “molecular species” by estimating molecular operational units (MOTUs; [55] using the software MEGA 7.0 [56] and the BOLD analyses tools [57].

Sequence divergence was estimated using the Kimura-2-Parameter (K2P) model with 1000 bootstrap estimates in MEGA7. This a standard model that has been extensively used in bar- coding studies [58,59]. The Barcode Index Number (BIN) system [57] was used as delimitation criterion for the assignment of MOTUs across the full dataset. This method uses a 2.2% in sequence divergence cut-off, but updates this value according to the distribution of divergence among sequences in the dataset. Recent studies suggest that using a single divergence cut-off may not be appropriate for every organism (reviewed in [60], such as in the case of diverse non-tropical chironomids [60,61]. But, given that our sample of this taxon was small, we assumed a single cut-off value for the entire dataset. However, future work should evaluate the optimal level of genetic divergence to delimit biological species in Neotropical macroinvertebrates. PLOS ONE Hidden biodiversity in Neotropical streams PLOS ONE | https://doi.org/10.1371/journal.pone.0231683 August 7, 2020 4 / 13