2014-05-05_REVISION - C1996083 - Laserfiche WebLink

Home Browse Search

HISTORICAL DNA REVEALS EXTINCTIONS AND INVASIONS 5 Table 1 For museum fish, the number of localities sampled within each drainage, year range of collection and number of individual successfully sequenced are listed Drainage Localities Years Individuals sequenced South Platte 5 1871 -1889 9 Arkansas 4 1889 8 Rio Grande 2 1873 -1889 2 San Juan 1 1874 2 Gunnison 1 1889 2 Colorado 3 1889 4 White /Yampa 1 1889 1 Unknown 1 1857 2 Total 18 1857 -1889 30 Thirty museum fish were sampled and successfully sequenced from seven drainages across a total of 18 different streams and lakes. Details on each museum fish, including an additional 14 samples that did not produce sufficient DNA for sequencing, can be found in Table S1. exists. In the case of two samples with uncertain locality information (Behnke 2002) —the type specimens of Oncorhynchus clarkii stomias [Academy of Natural Sciences in Philadelphia (ANSP) 7825 and 78261—addi- tional historical research was conducted (results in Supporting Information). Tissue and bone material from these samples were collected using disposable, sterile metal blades and placed into sterile glass or plastic vials in 70 % ethanol and transported to the University of Colorado, Boulder (CU). To ensure endogenous DNA was recovered from each historical sample and to minimize risk of contami- nation, all pre - polymerase chain reaction (PCR) steps were performed in a building that never housed mod- ern cutthroat trout samples, DNA or PCR product (Knight lab, Porter Sciences, CU). Clean clothes and shoes were designated solely for use in the laboratory. Furthermore, all pre -PCR steps were performed while wearing a disposable lab coat, facemask and two pairs of gloves. Before DNA extraction, samples were repeat- edly rinsed with 70 % ethanol and allowed to dry in a PCR workstation equipped with UV lights and HEPA- filtered airflow (Fisher Scientific). Once samples were dried, total genomic DNA was extracted using a Qiagen animal blood and tissue extraction kit, for which we modified two steps of the protocol. We added two times the recommended amount of buffer ATL and eluted DNA from the silica column with three 30 -µL aliquots of Qiagen buffer AE. One mock extraction con- trol was included for every four or five samples. Both extraction controls and no- template PCR controls were amplified with every four or five samples. Replicate DNA extractions, PCR and sequencing were carried out © 2012 Blackwell Publishing Ltd for a subset of samples (listed in Table Sl) at CU and in a specialized historical DNA laboratory at the Austra- lian Centre for Ancient DNA (ACRD), University of Adelaide, Australia. Owing to the degraded nature of the DNA in the museum fish samples, multiple short, overlapping mito- chondria) DNA fragments were amplified from each successful DNA extraction and subjected to Sanger sequencing. We generated data for five ND2 gene frag- ments and one COI gene fragment, ranging in size from 91 to 135 nucleotides including primers (primer sequences are listed in Table S2). These gene fragments included several diagnostic mitochondrial single nucleo- tide polymorphisms (SNPs) characterized in extant cut- throat trout subspecies. At CU, PCR amplicons were cloned and 5-8 clones were sequenced per successful amplification, whereas at ACAD, amplicons were sequenced directly. Sequence data were imported and edited in Sequencher 4.6 (Gene Codes Corporation). Primer sequences were removed and ND2 sequence fragments were grouped by sample. Disagreements between base calls were resolved by majority rule (best two out of three) or left ambiguous (e.g. Y = C or T). Modern tissue collection, DNA extraction and sequence generation For comparison to the historical samples, we sequenced a 648- or 889 - basepair region of the ND2 gene for at least 10 individuals from 53 contemporary populations (Table S3). Fish were sampled from across the species' range in the Southern Rockies as part of previous stud- ies (Metcalf et al. 2007; Rogers 2008), and ND2 sequenc- ing methods were performed as described in Metcalf et al. (2007). Genetic analysis For the four extant lineages of cutthroat trout, we iden- tified unambiguously diagnostic SNPs in the 430 -bp 'museum subset' of ND2 and COI gene sequence data. Given the difficulty referring to lineages that have dif- ferent geographical locations in different time periods and the confusion of pre- existing taxonomy in relation to those lineages, we refer to three of these lineages by colour (as shown in Fig. 2) and resolve these lineages to taxonomy later in this study. As Rio Grande cutthroat trout (shown in orange in Fig. 2) do not appear to have been propagated and stocked extensively (Pritchard et al. 2009), we refer to this lineage by name. We assigned museum samples to one of the extant lineages based on the presence of lineage- defining SNPs. Second, we generated a statistical parsimony network of haplo- types that were present in both museum and modern