3.1 Loci selection, DNA extraction, and primer efficacy
We tested 28 primer pair combinations by performing 2056 individual PCR reactions and provided 74 new sequences of Facetotecta that have been deposited in NCBI GenBank with the accession numbers XXXX-YYYY (Table S1).
We also tested the efficacy and amplification success of the two DNA-extraction protocols (Fig. 2A-C; Tables 1, S2). We performed 815 PCR reactions with the GeneReleaser kit and 1241 reactions with the simplified DNeasy kit (Table S2). The minimum and maximum amplification rates varied between 0% (DNeasy method, 16S 1471F/1427R, COI_Palerointernal, both n=6; Table S2) and ~98% (Dneasy method, 18Sface2F/2R, n=109; Table S2). Both protocols produced nearly identical overall amplification rates (~68%; Fig 2A, C; Table S2). The GeneReleaser protocol yielded at least ~42% and at most ~90% amplification success across loci (Fig 2; Table S2). Extraction and PCR amplification with the simplified DNeasy method yielded a slightly higher minimum amplification success (~47%) and a slightly lower maximum (~88%) (Fig 2A; Table S2); because different primer combinations were tested with the two protocols, these results may be biased. Combining the data from the two extraction methods, we computed an overall amplification success rate of 68% across all primer combinations (2096 successfully amplified fragments; Tables 1, S2; Fig 2A). Average (mean) successful amplification rates were ~60% for COX1 (296 out of 495 reactions; Fig. 2; Tables 1, S2), ~87% for 18S (515 of 589 reactions; Fig. 2C; Tables 1, S2), ~85% for 28S (361 of 426 reactions; Fig. 2C; Tables 1, S2), ~52% for 12S (73 of 139 reactions; Fig. 2C; Tables 1, S2), and 56% for 16S (168 of 300 reactions; Fig. 2C; Tables 1, S2). The amplification rate difference between the two extraction methods was notably higher for 16S and H3 than any other locus (Fig 2B, C; Table 1, S2), with 12S, COX1 and 18S yielding more similar amplification success rates between the two methods (Fig 2B, C; Tables 1, S2). For all primer combinations, 18S and 28S expectedly had the highest amplification rates, but we did not observe this pattern for H3 (~40% with GeneReleaser and ~75% with the simplified DNeasy method; Tables 1, S2).
Based on the highest observed primer efficacy (i.e., amplification success), as well as the overlap with publicly available primers, the fragment lengths obtained, and the quality of the sequences (e.g., the presence of clear and distinct chromatogram peaks), we recommend sequencing the following five “legacy” loci in future investigations of facetotecta systematics. COX1: F1new and R1 with the internal “Leray”-segment mlCOIInF/jgHCO2198 (~93% amplification success; ~750bp; Leray et al., 2013; Fig. 3; Tables 2, S2). 18S: initially with Face1F/2R (~85% success; ~1800bp; Fig. 3; Tables 2, S2); subsequently, depending on the amplification rate of the latter, Face1F/1R (~83% success; ~300 bp; Fig. 2; Tables 2, S2) and Face 2F/2R (~98% success; ~300bp; Fig. 3; Tables 2, S2). 28S: 84F (~74% success; ~880bp; Fig. 3; Tables 2, S2) with the internal primer Face3F/3R (~94% success; ~350bp; Fig. 3; Tables 2, S2). 12S: 3F (~83% success; ~300bp; Fig. 3; Tables 2, S2). 16S: SF/SR (~60% success; 700bp; Tsang et al., 2009; Fig. 3; Tables 2, S2) with 137F/548R (~50% success; ~400bp; Fig. 3; Tables 2, S2).