Molecular tracking of mountain lions in the Yosemite Valley region in California: genetic analysis using microsatellites and faecal DNA

1. What they did

After two people were killed by mountain lions in California in 1994, reports of mountain lions rose sharply, but nobody had a reliable estimate of how many there were. Mountain lions range widely and are hard to detect and count. The authors asked whether DNA from scats could help.

They tested twelve microsatellite markers (short, highly variable stretches of DNA), first developed for domestic cats, and asked four questions:

• Prey. Can mountain lion DNA be told apart from the DNA of their prey?
• Look-alikes. Can it be told apart from bobcats and other animals with similar scats?
• Scats. Does scat DNA give the same reliable genetic profile (genotype) as tissue?
• Individuals. Can individual mountain lions be told apart with these twelve markers?

To answer them, they genotyped tissue from 62 mountain lions from northern California, and scats taken at necropsy or capture from 15 of these animals. They also tested bobcats, dogs, coyotes, mule deer, bighorn sheep, domestic sheep, cattle, a horse and a human.

They then used the method in the field. United States Geological Survey staff collected 32 scats thought to be from mountain lions on hiking trails in the Yosemite Valley region, from March 1997 to August 1998. The scats were picked up opportunistically on set transects and while searching for lions, and ranged in estimated age from under 12 hours to over 2 weeks. Seven mountain lions had already been captured, sampled and released in the same area. The authors calculated “match probabilities” (the chance that two different animals share the same genotype) to decide whether identical scat genotypes came from the same animal.

2. What they found

Lab tests

• Every one of the 62 mountain lions had a unique genotype.
• The markers did not work on DNA from the non-cat species, so prey DNA did not confuse the results.
• Bobcats gave genotypes clearly different from mountain lions. One marker never worked in bobcats, and several had alleles (gene variants) of non-overlapping sizes. Of 103 alleles seen in the two species, only 11 were shared. Bobcats had more variation (77 alleles) than mountain lions (37).
• Genotypes from scats matched those from muscle for all 15 animals at all 12 markers.
• Scat DNA is harder to work with: an allele failed to show (dropout) in about 8% of scat reactions, against under 1% for blood or muscle, so each scat was run at least four times.
• Tests to tell the sex of the animal from scat DNA were unreliable: they sexed all 35 muscle samples correctly, but three of four female scats showed a male fragment, possibly from prey DNA.

Yosemite scats

• 15 of the 32 scats gave DNA: 12 were mountain lion and 3 were bobcat (three different bobcats). The other 17 gave no DNA, either because they were not from a cat, the DNA had broken down, or something in the scat blocked the test. Scats of all ages gave DNA.
• One scat matched a mountain lion already captured (lion no. 518). The authors describe this lion as sampled at two locations along Bridalveil Creek.
• Three scats, from near Mono Meadow, Badger Pass and Cathedral Rock, came from one other animal. The remaining eight scats were eight different animals.
• So scats identified 10 mountain lions, 9 of them never captured. With the 7 captured lions, that gives a minimum of 16 mountain lions living in or passing through the Yosemite Valley region during the study.
• Match probabilities for the matching samples were very low (about 1 in 114,000 and 1 in 135,000 with the authors’ preferred θ method), and still low even with an extreme setting.

3. What they concluded

• Scat DNA analysis is a practical and useful way to detect and identify individual mountain lions, and to locate their travel routes. Mountain lion DNA could be separated from bobcat and prey DNA, scats gave reliable genotypes, and twelve markers were enough to tell individuals apart.
• Most of the lions were found through scats rather than capture. This matters because capture is costly (often hundreds of US dollars per animal) and risky for people and lions.
• Sixteen is a minimum; collecting more scats would improve the estimate, and mark–recapture methods could be used with scat DNA to estimate numbers.
• Scat DNA needs more care, time and cost than tissue: repeated runs, steps against contamination, and checks against closely related species living in the same area. Scats of the right species can also be hard to find.
• Future work: methods to tell sex and estimate population size from scat DNA, for mountain lions and other species.

4. How tracking was or was not used

Field staff found and collected scats they judged to be from mountain lions along trails and set transects, and while searching for lions, and estimated each scat’s age. DNA then confirmed which were mountain lion (12 of 32), which were bobcat (3), and gave no answer for the rest. The paper does not describe how the scats were identified in the field or how their age was estimated. No trailing was reported. The introduction mentions earlier track-based work: track surveys used to follow mountain lion population trends, and a study identifying individual mountain lions by their tracks.

5. Tracker or observer certification

Not reported. The field scats were collected by United States Geological Survey staff (L. Chow, J. Ellingson and S. Pettit are thanked), with no description of their skill. Because only the scats that gave DNA could be checked, the paper cannot show how many of the 17 scats with no DNA were misidentified in the field.

6. Other technologies, skills or innovations

• Scat DNA extraction method adapted for mountain lions, with four to six repeat extractions per scat and bovine serum albumin added to help overcome substances in scat that block the test.
• Polymerase chain reaction (PCR) with fluorescently labelled domestic cat microsatellite markers, read on an automated DNA sequencer.
• Match probability calculations taken from human forensic DNA analysis, including the θ method recommended by the United States National Research Council for crime-scene evidence.

7. Implications for conservation (from the paper)

• California had no state-wide estimate of mountain lion numbers based on data, while conflicts with people were rising. Scat DNA offers extra data for counting and monitoring mountain lions.
• Scat sampling can find animals that capture misses, without the cost and safety risks of capture.