When Thomas Gilbert found the Annamite striped rabbit, he wasn’t traipsing through the jungles of Vietnam where the exceedingly rare creature lives. He wasn’t inspecting a trap, or peering through binoculars. He wasn’t even flicking through photographs captured by a camera-trap. He was, instead, looking at the rabbit’s DNA.

Which he had pulled out of a leech.

There are some 700 species of leech and many of them suck the blood of mammals. In doing so, they achieve with ease what scientists find difficult: They sneak through thick, tropical rainforest and collect DNA samples from rare and elusive species. By collecting these mini-vampires in turn, and sequencing the DNA in their bodies, scientists can get a cheap and surprisingly comprehensive snapshot of a jungle’s fauna.

The idea behind this unorthodox census technique started with Mads Bertelsen, a vet from Copenhagen Zoo. Bertelsen was doing fieldwork in Malaysia when he saw a leech fastened to the side of a tapir, and thought: Hmmm. He approached Gilbert with his crazy idea, and the duo tried to test it.

First, they fed leeches with goat blood, and showed that goat DNA survives in their bodies for at least four months. Next, they asked a colleague to collect wild leeches from the Central Annamite region of Vietnam, a lushly forested area where five new mammal species had been recently discovered.

From just 25 leeches, the team found DNA from six mammals, including pigs, cows, the small-toothed ferret-badger, and the threatened goat-like serow. They also found DNA from the Annamite striped rabbit, which was discovered in 1999 and had (at the time) never been seen in over 2,000 hours of camera-trap recordings. And they even detected DNA from the Truong Son muntjac, a small deer that was discovered in 1997 and has never been seen in the flesh. “That suggests that these animals aren’t as rare as we think or that the leeches are very good at finding them,” says Gilbert.

All conservation efforts are predicated upon knowing what lives where, and that knowledge is hard to come by in dense, rugged forests, whose residents are often rare, reclusive, and wary of humans. In such places, direct observation is almost impossible. You could interview villagers, but that’s challenging and often imprecise. You could pepper an area with camera-traps, but you’d have to lug heavy and expensive batteries to and fro. And all of these methods might miss small, tree-dwelling, or burrowing species.

Leeches, however, seem to feed on everything, and so miss little. Collecting them is also cheap, fast, and requires no special skills; as Gilbert says, “The leech collector simply offers his/herself up for bait.” Just by standing around in a rainforest for a day, a lucky collector (or unlucky, depending on how you look at it) can attract hundreds of leeches.

Attracting funding is a bigger problem. “The funny thing about leeches is that I’ve spectacularly failed to get anyone in Denmark to give me any money to study them,” says Gilbert. Fortunately, others have had better luck. Douglas Yu from the University of East Anglia and China’s Kunming Institute of Zoology, has enthusiastically leapt onto the leech-sequencing bandwagon, bringing money and extensive sequencing facilities with him. His involvement allowed the team to scale up.

They are now working with the WWF in Vietnam to find traces of the saola—a recently discovered, rarely seen, and critically endangered antelope that’s also known as the Asian unicorn. Yu has also been working with the Forestry Department of Yunnan Province to train an army of leech collectors. Some two hundred of the departments’ rangers have been walking around, plucking leeches from their bodies, stashing them in rubber pouches, and noting their GPS locations along the way.

Through their efforts, Yu ended up with some 20,000 leeches. Once preserved, the worms turn into dense, rubbery pellets that are too painful to pulverise by hand—the team resorted to small blenders. They then analysed the resulting slurry using a technique called metabarcoding, which amplifies small sections of DNA that vary distinctively between species, and compared these sequences against existing databases. They discovered sequences from martens, bears, tree shrews, mice, mongooses, monkeys, deer, cats, and more.

For now, the biggest remaining obstacle is developing a reference database to compare the results against. The gaps in the existing databases become clear when the team gets unexpectedly hilarious results, like matches  for the bearded seal. “Either it migrates down from the Arctic to South-East Asia, or something else is happening,” jokes Yu. (There’s precedent for such problems: one study of microbial DNA found that the platypus has been surreptitiously conquering the globe.)

The bearded seal: not found in Vietnam. Credit: Smudge9000, Flickr

Seals are closely related to dogs, bears, and weasels, so the team’s algorithms might be misidentifying a member of those other groups that hasn’t itself been sequenced yet. “Often, we’ll get a sequence and we won’t know what it is,” Gilbert says. “We had a 90 percent match to some kind of cat, and we don’t know if it’s a known one that hasn’t been sequenced, or one that hasn’t been described. So, we beg people for samples,” he adds. “We’ve sequenced mitochondrial genomes from tissue biopsies, bones, pickled bats, and soala skins. It’s very unglamorous but someone has to do it.”

Inevitably, the leech data will also have biases, depending on how easy different leeches are to catch, how they react to changes in climate, and which animals they prefer to feed from. But every census method suffers from bias, and few are as cost-effective or informative. “You can get everything from tiny mammals to large arboreal ones,” says Yu. “You just can’t do the kind of censuses that we can do in other ways.”

“This is really exciting,” says Holly Bik, a computational biologist at the University of Birmingham. “Accurate data is especially crucial for conservation initiatives and policy directives, and I think this work makes a strong case for why molecular tools could now start replacing expensive and laborious morphological taxonomy.”

Indeed, other scientists are already planning to use these methods in creative ways. For example, the Forest Stewardship Council certifies wood products that come from allegedly sustainable forests. Does this help the forests? Andreas Wilting from the Leibniz Institute for Zoo and Wildlife Research wants to find out by collecting leeches from certified and non-certified forests and working out how many mammals live in each. Similarly, Rosie Drinkwater at Queen Mary’s University of London is using leeches to study the effects of deforestation on wildlife in Borneo.

Leeches have long been banes of tropical hikers, and tools for doctors. Now, they are finding new gainful employment, as auditors of conservation projects, collectors of DNA, and census-takers of jungle menageries.