Species identification (ID) is crucial in many wildlife conservation and management applications. Given the conservation interest and ecological relevance of the order Carnivora, several molecular species ID tests have been developed for this taxonomic group including sequencing of diagnostic DNA markers (Farrell et al. 2000; Chaves et al. 2012), mitochondrial DNA (mtDNA)-RFLP (Paxinos et al. 1997; Mills et al. 2000), amplification with mtDNA species-specific primers (Palomares et al. 2002; Dalén et al. 2004; Fernandes et al. 2008), amplification of nuclear markers (Pilot et al. 2007; Oliveira et al. 2010), and real-time PCR (Berry and Sarre 2007; Matejusová et al. 2013). Recent approaches have moved toward more efficient methods to detect entire carnivore communities (Oliveira et al. 2010) and the use of standardized loci for the order (Chaves et al. 2012).

We report the development of a rapid, cost-effective molecular species ID test for 16 wild carnivores found in North America, South America and Europe. The method consists of co-amplification and fragment analysis of two short segments of the mtDNA control region using a single dye-labeled forward primer SIDL 5′-TCTATTTAAACTATTCCCTGG-3′ (Murphy et al. 2000), and two different reverse primers H16145 5′-GGGCACGCCATTAATCGACG-3′ (corrected sequence from Murphy et al. 2000) and H3R 5′-CCTGAAGTAGGAACCAGATG-3′ (Dalén et al. 2004).

We validated the method on 274 known hair (n = 60), fecal (n = 115), and blood/tissue samples (n = 99) of 21 wild carnivores from five families and the domestic dog (Canis familiaris) and cat (Felis catus) (Table 1). We analyzed 58 samples from 11 mammalian prey to evaluate the possibility of misidentification of carnivore fecal samples due to amplification of prey (Table 1) and tested human DNA to evaluate potential misidentification from human contamination. Samples were preserved in 95 % ethanol (tissue, feces), lysis buffer (blood, tissue), DETs buffer (feces), and silica desiccant (hair), and extracted using a Qiagen DNeasy Blood and Tissue Kit (tissue, blood, hair) and QIAmp DNA stool kit (feces) (Qiagen Inc., Valencia, CA). The 7 μl PCR contained 1× Qiagen Multiplex PCR Master Mix, 0.5× Qiagen Q Solution (Qiagen Inc.), 0.29 µM SIDL, 0.2 µM H16145, 0.1 µM H3R, 1 µl DNA extract. Thermocycling conditions included 15 min denaturation at 94 °C, 35 (hair, feces) or 30 (blood, tissue) cycles of 30 s at 94 °C, 90 s at 46 °C, 60 s at 72 °C and 30 min elongation at 60 °C. DNA extraction and PCR set up for hair and feces were performed in a laboratory dedicated to low quantity DNA samples. Negative controls were used in each extraction and PCR to monitor contamination. PCR products were loaded on an ABI3130xl DNA sequencer and fragments scored with Genemapper 3.7 software (Applied Biosystems) using size bins for each species.

Table 1 The number and locality of samples analyzed for each carnivore and prey species
Full size table

A PCR product was observed for 16 of the 22 wild predators plus the dog (Table 1). Primers SIDL and H16145 produced a 115–220 bp fragment (Fragment 1) in 13 predators, and primers SIDL and H3R produced a 315–401 bp fragment (Fragment 2) in 15 predators (Table 2). Ten species (American marten, fisher, black bear, Andean bear, river otter, gray fox, kit fox, swift fox, red fox, and Andean fox) exhibited a species-specific fragment size pattern (Table 2). Gray wolf (C. lupus), wolverine (Gulo gulo), and dog samples produced the same fragment or fragments (Table 2) and could not be distinguished. Also, coyotes (C. latrans) and red wolves (C. rufus), plus brown (Ursus arctos) and polar (U. maritimus) bears produced the same fragment size pattern respectively (Table 2). However, these primers will still be useful in areas where their ranges do not overlap. No amplification was observed for any felid species or for raccoons (Procyon lotor).

Table 2 Size ranges in base pairs for Fragment 1 and Fragment 2 for each species
Full size table

To further investigate these patterns, we performed in silico PCR using the program ecoPCR (Ficetola et al. 2010) and mammalian sequences from EMBL nucleotide library (release 117) and examined mismatches using PrimerBLAST (NCBI). These analyses revealed that ≥2 base pair mismatches within the last 4 bases on the 3′ end of a primer generally prevented PCR amplification. The horse was the only prey species that produced a PCR product in the predator species size range, but the 375 bp fragment did not overlap with predator fragments. Six other prey species also produced fragments, however, the size of the fragments was longer than observed for the predator species and unlikely to amplify in low quality samples (Table 2). Human DNA amplified at 573 base pairs.

This species ID method provides multiple benefits. First, 16 wild predator species of wide distribution and conservation interest can be identified with a single PCR and capillary sequencer run, allowing simple, rapid, and inexpensive processing of large sample numbers (ca. USD$1.5 in supplies after DNA extraction). This will be particularly useful for species identification of hair, feces, saliva or degraded tissue samples in areas where canids, ursids and mustelids are sympatric. Moreover, additional primers could be added to the multiplex and co-amplified for target species presently not detected (i.e. felids). Other benefits include the amplification of relatively short, high copy number mtDNA fragments favoring the analysis of degraded DNA, and the co-amplification of two markers producing distinct PCR products in most species.