In this post Matthew Dougherty discusses his recent paper ‘Environmental DNA (eDNA) detects the invasive rusty crayfish Orconectes rusticus at low abundances

oneTangled buoy strings, lost traps, pinched fingers, sweaty brows, and boats smeared with beef liver: these images define the experiences of countless managers and scientists who use baited trapping to monitor crayfish invasions, especially in lakes of the upper Midwest, USA. While these trapping methods are still widely used for crayfish monitoring and management (and with much success), they have some weaknesses. Traditional sampling methods aren’t often very effective when the target species is present at very low abundances. For example, these traditional methods may not be sensitive enough to allow managers to successfully and feasibly detect and monitor invaders at the onset of a new invasion, when abundances of the invader are still low. An inability to detect these invaders can cause serious problems, since early detection at the onset of an invasion is critical for the cost-effective control and eradication of the invader (Vander Zanden et al. 2010). Along a similar line, species of conservation concern are also often found at low relative abundances and can be equally difficult to monitor and manage, particularly in ways that aren’t potentially harmful to organisms or their habitats. For example, some imperiled crayfish species spend the majority of their lives in terrestrial burrows in wetland habitats, which are difficult to sample (e.g. excavate) without potentially harming the crayfish or destroying the burrow.

However, in the last couple of years, scientists have been testing a promising new tool that may be able to complement traditional sampling methods by providing more sensitive detections of organisms at low abundances. This new tool, which is experiencing rapid growth in applications across regions, taxa, and habitats, is called environmentwotal DNA (eDNA). eDNA is the small amount of DNA left behind by a target organism within an environmental sample (e.g. water, soil, air). eDNA has shown to be a promising and sensitive tool for early detection of some invaders like the bighead carp (e.g. Jerde et al. 2011) and the American bullfrog (Dejean et al. 2012). However, the ability of eDNA to successfully detect benthic arthropods, like crayfish, has been less studied and with equivocal results (Tréguier et al. 2014). In our paper, we ask whether eDNA is an effective tool for detecting crayfish specifically, and arthropods more generally, at relatively low abundances in an attempt to provide managers with a tool to more effectively monitor both arthropod invaders and species of conservation concern.

Our Study

We tested the ability of eDNA to detect the invasive rusty crayfish, Orconectes rusticus, and to reflect patterns of its relative abundance in temperate lakes in the upper Midwest, USA. We used traditional baited trapping as a measure of relative crayfish abundance and compared it with water samples for eDNA at each lake. Each water sample was filtered in the field and then analyzed in the laboratory with a qPCR assay. We then used hierarchical occupancy estimation models to investigate detection probabilities for O. rusticus eDNA using a handful of covariates. We found that probability of detection went up with relative abundance of the organism, but also that eDNA was able to detect our target crayfish down to very low relative abundances (only one crayfish collected in twelve traps), including in two lakes where the organism was not detected by trapping. We conclude that eDNA may be more effective than traditional sampling methods for detecting arthropod species, like O. rusticus, at low abundances, and that this methodological advance could have wide-ranging management implications.


Management Opportunities

Writing recently in Science, Emily DeMarco reports that out of the 590 species of crayfish found around the globe, about one-third are thought to be at risk of extinction. In the United States, that percentage rises to nearly 50%. At the same time, in North America alone crayfish have diversified into almost 400 species with new species being discovered every year, and experts estimate that as many as 60% of those species lack updated distributional information (DeMarco 2015). The number of crayfish species at risk of extinction, paired with the relatively little we know about already discovered crayfish species (not to mention those still yet to be discovered), leave many questions for managers looking to conserve and better understand these species. Each of these categories—species at risk of extinction and those without well understood distributional boundaries—provide managers with the challenge of monitoring these species, which often occur at low abundances on the landscape.

fourAnother problem facing managers (and connected with declines of many native crayfish species across the world) is that invasive species like our target organism, O. rusticus, continue to invade new territories. In 2009, the authors Olden, Adams and Larson reported that O. rusticus—a native of the Ohio River Basin in eastern North America—had been discovered west of the Continental Divide for the first time. The species was discovered in the John Day River of the Columbia River Basin in Oregon in 2005, and it has since spread in this system with a number of sensitive freshwater species, like Endangered Species Act-listed Chinook salmon (Olden, Adams & Larson 2009). For managers in this region, early detection of new O. rusticus invasions or secondary spread within the Columbia River Basin could facilitate more rapid containment or eradication, management responses that were missed in the John Day River until after the species was well-established and widespread. As previously mentioned, early detection of an invasion is invaluable for cost-effective management, and we suggest that eDNA is a promising tool for monitoring new invasions while population abundances are still low.

 eDNA as a complementary tool in field biology

Monitoring imperiled species, updating distributional information of lesser known species, keeping an eye on already invaded waters, and providing warnings of new species invasions: these are a handful of needs where scientists and resource managers may find use for eDNA. Importantly, our study demonstrates that eDNA is seemingly highly effective and feasible for freshwater crayfish, and we suspect that this may be true for other benthic arthropods like freshwater crabs and shrimp. The eDNA technique is young and deserves more testing and refinement across taxonomic groups and environments, but our authorship team included at least one initial skeptic at the onset of our study, who was more than won over by our results.

However, we do stress that there are many data needs and questions in field biology where eDNA is unlikely to ever replace more conventional field sampling. For example, our method can’t tell us anything about the life-history stage of our study crayfish (e.g. are females carrying eggs?), what specific habitats within a lake they may be using, or how they’re interacting with other organisms like fish. Tools like baited trapping, snorkeling or SCUBA diving, and hand collecting will never be abandoned for these organisms. And while eDNA seemingly excels at detecting occupancy (presence/absence) of organisms at very low population abundances, the method occasionally fails to accurately reflect overall abundance of organisms, although methodological refinements may improve on this over time.

Furthermore, our study was greatly facilitated both by access to long-term datasets on O. rusticus in our study lakes that let us identify an a priori abundance gradient to target for eDNA sampling (Peters & Lodge 2013), as well as museum collections (like the Illinois Natural History Survey) that provided us with tissues for designing and testing our qPCR assay. Accordingly, our eDNA method was built in part on decades of natural history observations in the field, as well as facilitated by the under-valued work of taxonomists and systematists at museums, universities, and state and federal agencies. These dependences underscore the inter-relationships between eDNA and field biology, and we emphasize that this tool is a complement to, rather than replacement for, some of our more conventional surveillance and monitoring approaches.

But with those caveats in place, we note that the implications of eDNA go beyond the science. After all, no one likes to go home smelling like beef liver after a long day in the field. Our eDNA method may make accurate detection and monitoring of some benthic arthropod populations less unpleasant than they have been in the past.



Dejean, T., Valentini, A., Miquel, C., Taberlet, P., Bellemain, E. & Miaud, C. (2012) Improved detection of an alien invasive species through environmental DNAbarcoding: the example of the American bullfrog Lithobates catesbeianus. Journal of Applied Ecology, 49, 953-959.

DeMarco, E. (2015) The missing mudbug. Science, 349, 915-917.

Jerde, C.L., Mahon, A.R., Chadderton, W.L. & Lodge, D.M. (2011) Sight-unseen: detection of rare aquatic species using environmental DNA. Conservation Letters, 530 4, 150–157.

Olden, J.D., Adams, J.W. & Larson E.R. (2009) First record of Orconectes rusticus (Girard, 1852) (Decapoda, Cambaridae) west of the Great Continental Divide in North America. Crustaceana, 82, 1347-1351.

Peters, J.A., & Lodge, D.M. (2013) Habitat, predation, and coexistence between invasive and native crayfishes: prioritizing lakes for invasion prevention. Biological Invasions, 15, 2489-2502.

Tréguier, A., Paillisson, J.M., Dejean, T., Valentini, A., Schlaepfer, M.A. & Roussel, J.M. (2014) Environmental DNA surveillance for invertebrate species: advantages and technical limitations to detect invasive crayfish Procambarus clarkii in freshwater ponds. Journal of Applied Ecology, 51, 871-879.

Vander Zanden, M.J., Hansen, G.J.A., Higgins, S.N. & Kornis, M.S. (2010) A pound of prevention, plus a pound of cure: Early detection and eradication of invasive species in the Laurentian Great Lakes. Journal of Great Lakes Research, 36, 199-205.