On the origin of faeces: morphological versus molecular methods for surveying rare carnivores from their scats

1. What they did

In Britain, pine martens, foxes, otters, mink and water voles are monitored from their scats, and scats of martens and foxes have long been relied on to show where martens live. After a run of conflicting scat surveys, pine martens were argued to be effectively extinct in England and Wales, and a reintroduction was being considered. The authors used DNA to test whether experienced marten naturalists can really tell marten scats from fox scats in the field.

• Collecting scats (1999 to 2000). Three experienced marten field workers, kept anonymous, collected fresh “marten” scats along the same transects at different times in Galloway, Scotland, using standard methods. They noted size, shape and odour and gave each scat a confidence score: “possible”, “probable” or “certain” marten. One surveyor also collected more widely in south-west Scotland and in 10 x 10 km squares in England and Wales where martens were probably rare. Other naturalists sent in scats they believed might be from martens.
• Identifying the scats by DNA. Scats were air-dried and frozen. The team extracted DNA from a small scraping of each scat’s surface and sequenced a short piece of mitochondrial DNA to identify the species, with checks against contamination.
• Testing for counter-marking. To rule out foxes marking over marten scats and confusing identification, they placed 30 fresh scats from captive martens outside where foxes are common, 10 at a time, and collected them again at two-day intervals.

2. What they found

• DNA was extracted and amplified from 53% of the 163 scats.
• Scotland: 18% of the 56 scats identified as “marten” were actually from foxes. The three surveyors’ error rates were 9%, 29% and 19%.
• Confidence did not help: of eight fox scats from Galloway, three had been scored “certain” marten and two “probable”.
• England and Wales: only one scat was from a marten (from Caernarfonshire, Wales). All the other “marten” scats (97%) were from foxes (28) or a polecat (1).
• One surveyor correctly identified every marten scat in Galloway, yet was wrong every time (12 scats) in the low-density areas of England and Wales.
• All but one of the possible marten scats sent in by other naturalists in England and Wales were from foxes (18).
• Exposure trial: DNA identified the captive marten scats in 70%, 100% and 100% of cases in the three trials, after up to 20 days in the field (mean rainfall 40 mm). All amplified DNA was from martens, so fox counter-marking is unlikely to confuse identification.

3. What they concluded

• Pine marten scat surveys based on how scats look are unreliable: surveyors were wrongly confident that they could identify marten scats.
• The surveys failed completely where martens are endangered. This is partly because scats are harder to find when animals are scarce and stop marking territories, but also because surveyors’ judgement was much worse there.
• Because many studies of rare mammals rely on identifying scats, scat identification needs to be more stringent, especially for endangered species.
• Elusive mammals should be monitored with several kinds of evidence: DNA from scats, identification of cast hairs, camera traps and non-leading “sighting” questionnaires. For national surveys, screening many samples at once with microarrays may soon be cost-effective.
• Mitochondrial DNA can work even in fairly old, rain-washed scats, but only some scats ever yield usable DNA. Some scats in this study were more than a month old when collected, rain matters, and slugs have been seen eating nearly a whole scat within 24 hours.
• Field workers, or better, dogs, could perhaps be trained to identify scats, ideally using scats from wild animals identified by DNA, because captive animals’ diet may change the look and smell of their scats.

4. How tracking was or was not used

The study tested tracks and signs identification directly: experienced naturalists identified scats as pine marten from size, shape and odour, following standard field guides, and gave each a confidence score. DNA was then used to check every identification. No trailing was reported.

5. Tracker or observer certification

The three collectors are described as “experienced marten field workers”, and the other naturalists only as people who believed their scats might be from martens. No certification is reported. Their scat identification skill was tested in this study, by DNA, and was found unreliable, especially in low-density areas.

6. Other technologies, skills or innovations

• DNA extraction from scats with Nucleon phytopure kits (a plant DNA kit), scraping the scat’s outer surface to collect gut-lining cells.
• Sequencing of a roughly 200 base pair section of mitochondrial DNA with carnivore-specific primers designed for the study, compared with marten, fox and polecat sequences in the GenBank database.
• An exposure trial testing how long scats left in the field still give usable DNA.

7. Implications for conservation (from the paper)

• Decisions such as reintroducing pine martens to England and Wales, which are crucial and expensive, must rest on entirely reliable field data. International guidelines oppose reintroductions while remnant populations survive.
• Surveying from scat appearance is especially common in Britain but may spread elsewhere, because a carnivore that becomes rare is often protected by law and harder to survey by other means.
• Accurate scat identification also matters for diet studies.