Dna

DNA Extraction, Amplification, and Sequencing

We obtained total DNA from thoracic/leg muscle tissue of both adults and larvae by extraction with QIAquick PCR purification kit (QIAGEN, Alda et al. 2007) (see Valladolid et al. 2023 for the complete process). We amplified a 660-bp region of the mitochondrial COI gene (cytochrome oxidase I) with the primers C1-J-1718 (5’-GGAGGATTTGGAAATTGATTAGTTCC-3’) and C1-N-2329 (5’-ACTGTAAATATATGATGAGCTCA-3’) ( Folmer et al. 1994; Simon et al. 1994). We used three microlitres of the DNA solution as a template in 25 μl PCR reactive: 1x of the corresponding buffer [75 mM Tris HCl, pH 9.0, 50 mM KCl, and 20 mM (NH 4) 2 SO 4], 2 mM MgCl 2, 10 mM dNTPs mix, 0.1 μM of both primers, 0.02% BSA, and 0.125 units AmpliTaq Gold® DNA Polymerase (Applied Biosystems) . We assessed PCR success through visualization with SYBR Safe™ DNA Gel Stain (Invitrogen) under ultraviolet light following gel electrophoresis (1.5% agarose gel). We purified PCR products by treatment with ExoSAP-IT (USB Amersham, Buckinghamshire, UK) incubating them at 37ºC for 15 min, followed by 80ºC for 15 min to inactivate the enzyme. We used purified PCR product (10 μl volume, containing 15–20 ng of purified product and 3 pmol of primer) to sequence in both directions using the BigDye Terminator v3.1 sequencing kit (Applied Biosystems Inc., Foster City, USA). Sequencing was performed at the CAI Biological techniques, Genomic Unit, Complutense University of Madrid (UCM). We BLASTed ( Altschul et al. 1997) the obtained sequences against GenBank (2024) to assess sequencing quality.

The alignment of all Rhyacophila COI gene sequences generated in our lab was performed and edited manually using MEGA X ( Kumar et al. 2018). Fine adjustments (for individual, undefined nucleotides) were made by eye, as the COI does not present any gaps.

We undertook phylogenetic analyses using Maximum Likelihood (ML) and Bayesian inference (BI) methods in the CIPRES Science Gateway v.3.3 online server ( Miller et al. 2010). We ran jModeltest v.2.1.9 ( Darriba et al. 2012; Guindon & Gascuel 2003) on the partitioned data to identify the best-fitting substitution model that we used in the ML and BI tree inferences. We performed phylogenetic inference under ML in RAxML-HPC BlackBox ( Stamatakis 2014), using 10,000 bootstrap replicates to assess node support. We performed BI in MrBayes-3.2.7 ( Ronquist et al. 2012) using the best-fitting substitution model and setting lset nst = mixed rates = invgamma, for 5×106 generations, and reconstructed Majority Rule trees (with 50% threshold, made with Geneious Prime® 2022.0.1) after discarding the first 25% of the generations as burn-in. We visualized the best tree using FigTree v1.4.4. ( Rambaut 2019), and we labelled the nodes with posterior probabilities. We calculated distances among and within species using the maximum composite likelihood and p-distance models with MEGA X. We deposited nucleotide sequences in GenBank [accession numbers and Tissues and DNA Collection numbers (MNCN: ADN) in Table 1, and in the dataset by Valladolid et al. 2025].