A noninvasive method for distinguishing among canid species: amplification and enzyme restriction of DNA from dung

Faecal DNA detection of invasive species: the case of feral foxes in Tasmania

Early detection of biological invasions is critical to reducing their impact, but because invading organisms are initially at low densities, detection and eradication can be challenging. Here, we demonstrate the utility of faecal DNA analysis for the detection of an elusive invasive species – the red fox, Vulpes vulpes, which was illegally introduced to the island of Tasmania in the late 1990s. Foxes are a devastating pest to both wildlife and agriculture on the Australian mainland, and would have a similarly serious impact in Tasmania if they became established. Attempts to eradicate foxes from Tasmania have been hampered by unreliable distribution data derived mostly from public sightings. In response, we developed a highly accurate and reliable DNA-based PCR-multiplex test that identifies foxes from field-collected faeces. We also developed a sexing test, but it was reliable only for faeces less than three weeks old. Faeces are a useful target for DNA-based diagnostics in foxes because they are deposited in prominent locations and are long-lasting. The species identification test formed a key component of a Tasmania-wide detection and eradication program. In all, 1160 geo-referenced carnivore scats were analysed; of these, 78% contained DNA of sufficient quality for species identification. A single scat from the north-east of the island was identified as belonging to fox, as was a nine-week-old roadkill carcass from the north coast, and a blood sample from near Hobart, triggering increased control and surveillance in these regions. The accuracy, reliability, and cost-effectiveness of non-invasive tests make them a critical adjunct to traditional tools for monitoring cryptic invasive species that are at low density in the early stages of invasion and when eradication is still an option.

Assessing reliability of microsatellite genotypes from kit fox faecal samples using genetic and GIS analyses

Noninvasive faecal DNA sampling has the potential to provide a wealth of information necessary for monitoring and managing endangered species while eliminating the need to capture, handle or observe rare individuals. However, scoring problems, and subsequent genotyping errors, associated with this monitoring method remain a great concern as they can lead to misidentification of individuals and biased estimates. We examined a kit fox scat data set (353 scats; 80 genotypes) for genotyping errors using both genetic and GIS analyses, and evaluated the feasibility of combining both approaches to assess reliability of the faecal DNA results. We further checked the appropriateness of using faecal genotypes to study kit fox populations by describing information about foxes that we could deduce from the 'acceptable' scat genotypes, and comparing it to information gathered with traditional field techniques. Overall, genetic tests indicated that our data set had a low rate of genotyping error. Furthermore, examination of distributions of scat locations confirmed our data set was relatively error free. We found that analysing information on sex primer consistency and scat locations provided a useful assessment of scat genotype error, and greatly limited the amount of additional laboratory work that was needed to identify potentially 'false' scores. 'Acceptable' scat genotypes revealed information on sex ratio, relatedness, fox movement patterns, latrine use, and size of home range. Results from genetic and field data were consistent, supporting the conclusion that our data set had a very low rate of genotyping error and that this noninvasive method is a reliable approach for monitoring kit foxes.

Fecal DNA Analysis for Identifying Species and Sex of Sympatric Carnivores: A Noninvasive Method for Conservation on the Tsushima Islands, Japan

Fecal analysis is a useful tool for the investigation of food habits and species identity in mammals. However, it is generally difficult to identify the species based on the morphological features and contents of feces deposited by mammals of similar body size. Therefore we developed noninvasive DNA analysis methods using fecal samples for identification of the species and sex of four small sympatric carnivores living on the Tsushima Islands of Japan: the leopard cat (Felis bengalensis), Japanese marten (Martes melampus), Siberian weasel (Mustela sibirica), and feral cat (Felis catus). Based on DNA sequence data from previous phylogenetic studies, we designed species-specific primers for polymerase chain reaction (PCR) amplification of the partial mitochondrial cytochrome b gene (112-347 bp) to identify the species and primers for the partial SRY gene (135 bp) to determine the sex. Due to the adjustment of PCR conditions, those specific DNA fragments were successfully amplified and then applied for species and sex identification. Nucleotide sequences obtained from the PCR products corresponded with cytochrome b sequences of the carnivore species expected. The protocol developed could be a valuable tool in the management and conservation of the four carnivore species occurring on the Tsushima Islands.

Nucleic acid isolation from ecological samples--animal scat and other associated materials

Noninvasive sampling is very attractive to field biologists and has tremendous potential for studying secretive species and being a cost-effective method of increasing sample sizes in studies of large, dangerous animals. Extracting DNA from noninvasively collected samples can be challenging, and the methods have been developed mainly through modification of previously developed protocols for other sample types. We present the most commonly used methods along with modifications used by some researchers to deal with the problem of coextraction of polymerase chain reaction (PCR) inhibitors. Although it is difficult to generalize about which methods should be used on particular sample types, we discuss the success of the methods in studies to date. We close with general suggestions for dealing with potential problems associated with the analysis of DNA obtained from noninvasively collected samples.

Using faecal DNA sampling and GIS to monitor hybridization between red wolves (Canis rufus) and coyotes (Canis latrans)

The US Fish and Wildlife Service's (USFWS) Red Wolf Recovery Program recognizes hybridization with coyotes as the primary threat to red wolf recovery. Efforts to curb or stop hybridization are hampered in two ways. First, hybrid individuals are difficult to identify based solely on morphology. Second, managers need to effectively search 6000 km(2) for the presence of coyotes and hybrids. We develop a noninvasive method to screen large geographical areas for coyotes and hybrids with maternal coyote ancestry by combining mitochondrial DNA sequence analysis of faeces (scat) and geographic information system (GIS) technology. This method was implemented on the Alligator River National Wildlife Refuge (1000 km(2)) in northeastern North Carolina. A total of 956 scats were collected in the spring of 2000 and 2001 and global positioning system (GPS) coordinates were recorded. Seventy-five percent of the scats were assigned to species and five coyote/hybrid scats were detected. Placement of scat location coordinates on a map of the experimental population area revealed that four of the coyote/hybrid scats were detected within the home ranges of sterilized hybrids. The other coyote/hybrid scat indicated the presence of a previously unknown individual. We suggest this method be expanded to include more of the experimental population area and be optimized for use with nuclear markers to improve detection of hybrid and back-crossed individuals.

Genetic marker investigation of the source and impact of predation on a highly endangered species

In September and October 2000, the remains of a number of apparently predated northern hairy-nosed wombats (Lasiorhinus krefftii) were discovered in Epping Forest National Park, the site of the only known population of this highly endangered species. Analysis of DNA recovered from six carcasses and a section of intestine found nearby was carried out using microsatellite and Y-specific primers. This identified seven individual wombats, the identity of three of which was inferred from a genotype database prepared from animals sampled during trapping programmes. Six victims were male and one female, suggesting that female-biased predation rates are unlikely to be the cause of the current male-biased population sex ratio. DNA isolated from four canid faeces found in the vicinity revealed three distinct canid microsatellite genotypes with very high probabilities of belonging to dingoes (Canis familiaris dingo). A wombat genotype matching that of one of the dead individuals was identified from scats of two of the dingoes. In addition, two macropod microsatellites were amplified from two dingo scats. These observations provided vital information regarding predation on northern hairy-nosed wombats, and prompted the permanent exclusion of dingoes from the park by the erection of a dingo-proof fence.

Faecal genetic analysis to determine the presence and distribution of elusive carnivores: design and feasibility for the Iberian lynx

Noninvasive methods using genetic markers have been suggested as ways to overcome difficulties associated with documenting the presence of elusive species. We present and assess a novel, reliable and effective molecular genetic technique for the unequivocal genetic identification of faeces from the endangered Iberian lynx (Lynx pardinus). From mitochondrial DNA (mtDNA) cytochrome b and D-loop region sequences, we designed four species-specific primers (for products 130-161 bp long) that were considered to be likely to amplify degraded DNA. We compared two DNA extraction methods, various DNA amplification conditions and the robustness and specificity of the primer pairs with 87 lynx samples from 5 potentially different lynx populations and with 328 samples of other carnivore species. The utility of the identification technique was tested with faeces of different ages, with faeces from controlled field experiments, and with faeces collected from locales with possible lynx populations from throughout the state of Andalusia, Spain (8052 km2). Faecal mtDNA extraction was more efficient using PBS wash of the faeces instead of a faeces homogenate. Our assay increased from 92.6 to 99% efficiency with a second amplification and a reduction in template concentration to overcome polymerase chain reaction (PCR) inhibition. Our assay never produced false positives, and correctly identified all lynx faeces. Of 252 faeces samples of unknown species collected throughout Andalusia, 26.6% (from three different areas) were classified as Iberian lynx, 1.4% showed evidence of PCR inhibition and 1.2% were of uncertain origin. This method has proven to be a reliable technique that can be incorporated into large-scale surveys of Iberian lynx populations and exemplifies an approach that can easily be extended to other species.

Molecular identification of prey in predator diets

In many situations prey choice by predators in the field cannot be established or quantified using direct observation. The remains of some prey may be visually identified in the guts and faeces of predators but not all predators ingest such hard remains and even those that do consume them may also ingest soft-bodies prey that leave no recognizable remnants. The result is, at best, a biased picture of prey choice. A range of molecular techniques and applications are reviewed that allow prey remains to be identified, often to the species and even stage level. These techniques, all of which are still in use, include enzyme electrophoresis, a range of immunological approaches using polyclonal and monoclonal antibodies to detect protein epitopes, and recently developed polymerase chain reaction (PCR)-based methods for detecting prey DNA. Analyses may be postmortem, on invertebrate and vertebrate predators collected from the field, or noninvasive assays of the remains in regurgitated bird pellets or vertebrate faeces. It was concluded that although monoclonal antibodies are currently the most effective method in use today, PCR-based techniques have proved to be highly effective and versatile in recent laboratory trials and are likely to rapidly displace all other approaches.

Dietary separation of sympatric carnivores identified by molecular analysis of scats

We studied the diets of four sympatric carnivores in the flooding savannas of western Venezuela by analysing predator DNA and prey remains in faeces. DNA was isolated and a portion of the cytochrome b gene of the mitochondrial genome amplified and sequenced from 20 of 34 scats. Species were diagnosed by comparing the resulting sequences to reference sequences generated from the blood of puma (Puma concolor), jaguar (Panthera onca), ocelot (Leopardus pardalus) and crab-eating fox (Cerdocyon thous). Scat size has previously been used to identify predators, but DNA data show that puma and jaguar scats overlap in size, as do those of puma, ocelot and fox. Prey-content analysis suggests minimal prey partitioning between pumas and jaguars. In field testing this technique for large carnivores, two potential limitations emerged: locating intact faecal samples and recovering DNA sequences from samples obtained in the wet season. Nonetheless, this study illustrates the tremendous potential of DNA faecal studies. The presence of domestic dog (Canis familiaris) in one puma scat and of wild pig (Sus scrofa), set as bait, in one jaguar sample exemplifies the forensic possibilities of this noninvasive analysis. In addition to defining the dietary habits of similar size sympatric mammals, DNA identifications from faeces allow wildlife managers to detect the presence of endangered taxa and manage prey for their conservation.

Cross-species amplification, non-invasive genotyping, and non-Mendelian inheritance of human STRPs in Savannah baboons

Twenty-nine human microsatellite primer pairs were screened for their utility in the cross-species amplification of baboon DNA derived from both blood and feces as part of a larger study to identify paternal half sisters in a population of wild baboons (Papio cynocephalus). Forty-one percent (12/29) of the human primers successfully amplified baboon DNA. Of these 12 primers, six amplified fragments that were both polymorphic and heterozygous (mean number of alleles = 6, mean heterozygosity = 87%) and yielded repeatable results. However, only five of these six simple tandem repeat polymorphisms (STRPs) showed patterns of Mendelian inheritance (i.e., mothers and offspring shared at least one allele at each locus), and were therefore useful for determining relatedness between individuals. Analysis of the sixth primer revealed non-Mendelian inheritance, i.e., three of the six known mother-daughter pairs had no shared alleles. This failure was probably due to non-specific fragment amplification, and may have resulted from a different STRP locus being amplified in mother and daughter. This finding highlights the importance of sampling DNA from known parent-offspring pairs when screening microsatellite primers for genetic studies. Multiple, independent replications of genotypes and Mendelian checks are both particularly important when using cross-species amplification or when using a low-quality source of DNA.