Genome-wide markers test the status of two putative species of North American bumble bees

Accurate species delimitation is critical to identifying the conservation status of species. Molecular species delimitation methods have revealed previously unrecognized cryptic species across the taxonomic spectrum. However, studies vary in the molecular markers selected, analytical approaches used, and taxon sampling, which sometimes results in conflicting conclusions. We tested a two-species hypothesis of the Bombus occidentalis complex using nuclear (ultraconserved elements, UCE) and mitochondrial (cytochrome c oxidase I, COI) markers to infer maximum likelihood and Bayesian phylogenies for the taxa. We extracted tissue and sequenced 102 specimens from across the geographic range of the species complex. Through our analyses, we concluded that the complex actually represents two species, B. occidentalis and B. mckayi. Here, we include the raw sequences for the UCE and COI analyses and the final newick-formatted trees from the analyses.

Methods
UCE methods:

Methods generally followed those in Branstetter et al. (2021). We extracted DNA from the mid and hind legs of specimens using a Zymo Quick-DNA Miniprep Plus extraction kit and stored extracts in -80°C freezers at PIRU. Specimens were collected between 1956 and 2017, with one specimen from 1920.

We used a Tapestation 4150 automated electrophoresis system (Agilent, 5301 Stevens Creek Blvd., Santa Clara, CA 95051, USA) to measure the size of DNA fragments extracted from the specimens and Qubit 3.0 to quantify DNA concentrations. The size of fragments varied among specimens due to their variable ages, collection methods, and storage histories. We sheared the DNA fragments to target fragment sizes of 400 to 600 base pairs using a Q800R2 acoustic sonicator (Qsonica, Newtown, CT, U.S.A.). We varied shearing times from 0 seconds to 120 seconds with a 10-second on, 10-second off pulsing pattern. Samples with small fragment sizes were sheared for less time and samples with large fragment sizes were sheared for more time. Once sonicated, we purified the DNA samples using a homemade paramagnetic bead solution (Rohland and Reich 2012).

We captured and sequenced UCE loci from our sample specimens following the methods described in Branstetter et al. (2021). We prepared Illumina sequencing libraries using Kapa Hyper prep kits and custom 8 bp dual indexing adapters (Glenn et al. 2019). We amplified the libraries using 12 cycles of PCR, cleaned the amplified DNA using 1.0 to 1.2x SPRI beads to remove contaminants and fragments smaller than 200 bp, and quantified the DNA using Qubit. Samples with low measured volumes of DNA were re-amplified for 14 to 16 PCR cycles from an aliquot of the pre-PCR library.

We enriched the samples using an existing UCE bee-ant specific baitset (bee-ant-specific Hym-v2, Branstetter et al. 2017, Grab et al. 2019) identified and optimized for use in the order Hymenoptera. The baitset was developed using seven genomes from hymenopteran species, including two species from the bee families Apidae (the family that contains all bumble bees)  and Halictidae. We enriched the pooled libraries following a combination of the Arbor Biosciences v3.02 protocol (enrichment day 1) and a protocol based on Blumenstiel et al. (2010). We pooled up to ten samples per library at equimolar concentrations for enrichment. Finally, we repeated the PCR amplification, purification, and quantification steps previously described for the pooled enriched samples. Enriched pools were combined into a final sequencing pool and sent to Novogene Inc. for sequencing on an Illumina HiSeq X instrument (PE150).

COI Methods:

We extracted COI barcodes from the B. occidentalis and B. terricola UCE targeted sequences using the Phyluce program assembly_match_contigs_to_barcodes and a sequence downloaded from BOLD as a bait sequence (BBHYL247).