Fewer than 3500 tigers (Panthera tigris) remain in the wild, occupying <7 % of their historical range despite intense conservation efforts. Seventy percent of these tigers occur within 42 source populations covering <0.5 % of their historic range (Walston et al. 2010). Evaluating the success of conservation interventions requires rigorous monitoring of tiger densities and connectivity among populations. DNA analysis from scats has traditionally been used for genetic monitoring of wild tigers, but scat detection rates are generally low, particularly in humid, tropical environments, which limits their utility as a genetic resource (Smith 2012; A. Johnson, pers. comm.). Scent marks by contrast, constitute an untapped genetic resource, as they are more frequently deposited and have much higher detection rates than scats (Smith et al. 1989; Yudakov and Nikolaev 2012). Surveying over 664.4 km in the Russian Far East revealed scent spray to scat deposition ratios of 319:38 for males, and 109:46 for females (Yudakov and Nikolaev 2012). In Chitwan National Park, a subtropical forest in Nepal, the detection ratio was 612:28 (Smith et al. 1989) and in Tambling Wildlife Nature Conservation, a dense lowland tropical forest reserve in southern Sumatra, the detection ratio was 53:15 over 9 months of patrol activity (unpublished data). Tigers spray on trees and overhanging leaves along territory boundaries as a means of olfactory communication (Fig. 1). To boost the effectiveness of genetic monitoring of tigers, we examined the potential for DNA amplification from tiger scent marks for individual identification and gender determination.

Fig. 1
figure 1

Scent marks are typically 10–30 cm in diameter, 42 cm–1.7 m above ground, and sprayed on trees (left, Sanjay Gubbi/NCF/Panthera) or underneath overhanging leaves such as wild ginger (right, Rob Pickles/Panthera). Fresh sprays are commonly detected in the wild due to their unique aroma

Full size image

Scent DNA samples were collected from three captive tigers (2 males, 1 female) in southern Ontario in November 2013 and June 2014. Two to four samples were collected per individual and estimated time between marking and sampling varied from 10 min to 39 h. Scent marks were swabbed using sterile cotton buds, the swabs placed in a vial containing 500 μL Buffer ASL (Qiagen) and stored at room temperature. The cotton tip was excised and placed in a sterile 2.0 mL tube with the initial Buffer ASL aliquot and 15 μL Proteinase K (Qiagen). Samples were incubated at 65 °C for 2 h, after which an additional 15 μL of Proteinase K was added prior to incubation at 37 °C for 12 h. 500 μL Buffer AL (Qiagen) was added and the sample incubated at 65 °C for 10 min, after which 500 μL of cold 100 % ethanol was added and the sample incubated at 4 °C for 1 h. The remainder of the extraction process followed the suggested QIAGEN DNeasy Blood & Tissue Kit protocol, except DNA was eluted using 50 μL Buffer AE heated to 70 °C and left to incubate on the spin column membrane for 30 min prior to centrifugation. DNA was stored at −20 °C until analyzed.

Species identification was tested by amplifying a 110 bp fragment of the cytochrome oxidase b mitochondrial gene using primers H15149 (Kocher et al. 1989) and Farrel-R (Farrell et al. 2000). Sequencing followed Caragiulo et al. (2014). Five microsatellite loci in two multiplex groups (Menotti-Raymond et al. 1999) were used for individual identification (Table S1). Gender was determined using fluorescently labeled primers for the amelogenin region of the sex chromosomes (Pilgrim et al. 2005). Each gender typing reaction consisted of 3.50 μL of QIAGEN MasterMix, 0.70 μL of Q-solution, 0.20 μL of 10 μM amelogenin primers (Pilgrim et al. 2005), 0.20 μL of DNase-free water, and 2.0 μL of DNA template. Both microsatellite and gender-typing PCRs were done in triplicate using the multiple tubes approach (Taberlet et al. 1996). All PCRs were prepared and analyzed as per Caragiulo et al. (2015). Genotyping error rates were estimated using GIMLET version 1.3.2 (Valière 2002).

All cytochrome oxidase b sequences were successfully identified as tiger. All samples yielded reliable consensus genotypes, except DNA from a single swab that failed completely in all three replicates. Gender was correctly confirmed for all individuals. The PCR success rate (Table 1) and genotyping error rates (Table 2) are comparable to tiger genetic studies using scat and fall below the thresholds described by Smith and Wang (2014) for effective estimation of genetic variation and population subdivision. Comparable error rates are expected in tiger habitat with similar temperature ranges to those experienced during collection. Although error rates in the tropics may be higher, overall collection of scent DNA samples in tiger genetic studies in addition to scats, would significantly increase overall sample sizes, facilitating resolving individual genotypes and enabling more powerful genetic studies to take place. Since genetic connectivity is key to the long-term viability of the remaining 42 source populations, increased and high quality monitoring, using novel techniques such as the one presented here, are critical for the effective conservation management of wild tigers.

Table 1 Percentage of successful independent PCR, and allelic dropout and false allele rates using DNA from captive tiger scent sprays
Full size table
Table 2 Comparison of PCR success and error rates (mean across loci) for tiger sprays versus tiger scats
Full size table