DNA gathered from remote waterholes in northern Australia has been used to detect an endangered bird in the wild¹ for the first time. The result is the latest milestone in the rapidly maturing science of environmental DNA, in which traces of genetic material from soil, water or ice are used to identify the presence of plants and animals.

In a study published on 14 November in Endangered Species Research, a team in Australia reports that genetic material collected from waterholes showed that Gouldian finches (Erythrura gouldiae) had visited them in the previous 48 hours. Rangers also confirmed the species’ presence at the locations.

Scientists have been using environmental DNA (eDNA) analysis for about 15 years, including for tracking rare or elusive aquatic species, such as the great crested newt (Triturus cristatus) in the United Kingdom². And in the past few years, researchers have increasingly been using the technique to identify mammals, insects — and now birds — that live on land.

Testing for eDNA is often safer — for both animals and researchers — more cost-effective and, in some cases, more accurate and sensitive than conventional methods for pinpointing rare and endangered species, such as electrofishing surveys or tracking, scientists say. This is prompting regulatory agencies in a number of countries to adopt the technology to locate creatures, such as the endangered Canada lynx (Lynx canadensis) in the United States, or to monitor for invasive species.

But the technique is yet to convince some scientists, who say eDNA results aren’t robust enough to be used as the sole basis for making environment-management decisions that can have legal implications for governments and land owners.

Early studies that used eDNA to pinpoint specific species were criticized because of the potential for improper handling of samples to cause cross-contamination, leading to false-positive results. Scientists using the method are detecting only trace amounts of genetic material, so even minute amounts of contamination from gloves or equipment can taint the results. But proponents of the field say that the recent adoption of rigorous protocols that avoid or detect contamination have largely addressed such issues. And efforts to establish whether eDNA measures up to conventional survey methods are underway, they say.

Meteoric rise

The first study to show that large-bodied animals and plants drop enough DNA into their environment — through defecation and shedding cells — to be detected was published in 2003³. Five years later, researchers in France demonstrated that DNA in pond water could be used to detect a secretive frog, the invasive American bullfrog (Rana catesbeiana)⁴. Most such studies gather genetic material from aquatic environments because DNA disperses and remains free-floating in water, and can be detected in trace amounts.

Around 2014, Michael Schwartz, who heads up the US Forest Service’s National Genomics Center for Wildlife and Fish Conservation, and his team used eDNA to detect the endangered and hard-to-monitor bull trout (Salvelinus confluentus). The researchers initially analysed 124 water samples from waterways across Montana⁵, amassing a volume of data equivalent to that collected over the previous 15 years through conventional surveys using electrofishing, a method that is risky for people and fish, in which a current is run through the water to attract and then net fish. “We were able to do that in eight days,” Schwartz says. “We have estimated that it is about two to ten times faster and two to five times more cost-effective to use eDNA compared to electrofishing.”

The genomics centre now develops eDNA tools specifically for wildlife managers. Earlier this year, Schwartz’s team published results showing that DNA left in snow tracks or in snow near camera traps could be used to identify the presence of Canada lynx and wolverine (Gulo gulo) in Montana, and a small carnivorous mammal called the fisher (Pekania pennanti) in Idaho⁶. Conventional methods to detect the presence of land animals would typically involve time-consuming ground surveys to identify an animal by its tracks alone, or scat. “We have to be efficient with our conservation dollars in this day and age,” says Schwartz.

In another case, eDNA was more sensitive than conventional methods. When a camera trap image was unable to clearly identify what looked to be a Canada lynx in an area where its presence was unknown to rangers, eDNA extracted from the snow confirmed that the creature was indeed a lynxlynx, says Schwartz.

The US Fish and Wildlife Service has been using eDNA to detect incursions of invasive silver carp (Hypophthalmichthys molitrix) and bighead carp (Hypophthalmichthys nobilis) into the Great Lakes system since 2013. If genetic material is detected, rangers follow up the result using conventional monitoring.

In some cases, eDNA analyses are being used to enforce policy. For example, in 2014, the UK government approved the use of eDNA analysis for detecting the endangered great crested newt in land-use surveys that are required by law.

With a burgeoning market for eDNA analyses, dozens of companies now offer genetic tests for detecting rare species.

To reduce problems such as false positives that plagued the field in its early days, there are now standard methods for handling samples and detecting contamination when it occurs, says Florian Leese, an aquatic ecologist at the University of Duisburg–Essen in Germany. Adequate sampling, sterile equipment and experimental controls can all help to guard against contamination, for instance. In the case of the great crested newt, blind samples are sent to several laboratories to ensure the results are robust, says Leese. DNAqua-Net, a European-based international network of researchers who work with industry bodies and regulatory agencies, is developing best-practice guidelines on how to design and validate genetic tests for individual species and to define the amount of DNA needed to be sure a test returns a genuine positive result.

Ongoing challenges

But some ecologists are reluctant to abandon conventional methods. Jean-Marc Roussel, an aquatic ecologist at the French National Institute for Agricultural Research in Rennes, says that more studies comparing the cost and accuracy of eDNA analysis to conventional monitoring methods are needed before environment-management decisions are made on the basis of eDNA results. “I think it is still a tool for research, not for management,” he says.

Molecular ecologist Cecilia Villacorta Rath at James Cook University in Townsville, Australia, thinks researchers also need to demonstrate that genetic tests are sensitive and specific enough to avoid false negatives — the failure to detect a target species that is actually there.

Robust results are essential because the discovery of an endangered species can have weighty legal ramifications. In the United States, for instance, such species need to be protected under the Endangered Species Act, so an area of land could be designated a critical habitat as a result. Equally, the identification of an invasive species can spark the execution of eradication requirements enshrined in law. “You have to be sure that your measurement is reliable,” says Leese.

As the chair of DNAqua-Net, Leese is leading the charge to develop standards that ensure genetic tests are accurate and provide agencies with confidence in their results.

The next step could be to certify companies and government laboratories conducting eDNA studies, he says, similarly to the way in which medical laboratories are certified.