Scientists are using dragonfly larvae as natural indicators to identify areas where toxic mercury is accumulating in freshwater ecosystems. The approach, which traces back to a research project launched in 2012, combines years of mercury measurements from dragonfly larvae with information on water quality, soil and surrounding land cover.
Mercury can travel long distances from its original source before entering rivers, lakes and other freshwater environments. Once present in an ecosystem, it can move through the food chain and accumulate in wildlife, including fish that may later be consumed by people.
Dragonfly larvae are particularly useful for monitoring because they spend a long part of their lives in freshwater and generally remain close to the places where they develop. They also eat other insects and can accumulate mercury in their bodies, making them useful biological indicators of contamination.
The Dragonfly Mercury Project initially involved 12 US national parks in a pilot effort in 2012. It later expanded, with samples eventually collected from more than 450 sites across 100 national parks and other protected areas. Citizen scientists have also contributed to the sampling work alongside trained researchers and National Park Service staff.
Researchers found that mercury levels in dragonfly larvae were positively associated with mercury concentrations in fish and amphibians living in the same aquatic environments. This suggests that dragonfly measurements can provide valuable information about contamination within freshwater food webs.
The latest modelling approach is designed to identify potential mercury risks even in areas where direct testing has not been carried out. Scientists combine the dragonfly dataset with environmental information to understand conditions that influence mercury contamination.
One important factor is mercury methylation. Microorganisms can convert inorganic mercury into methylmercury, a form that is more easily absorbed by organisms and can accumulate through aquatic food chains. The model therefore considers environmental conditions that influence this process rather than simply measuring the presence of mercury.
The researchers say the model could help authorities determine where further water and fish testing is most urgently needed. It could also support fish consumption advisories in areas where people rely on locally caught fish.
The approach may be particularly useful around protected lands and communities where monitoring is limited, including Indigenous communities that depend on local fish for food.
Scientists stress that modelling will not replace physical sampling. Instead, it can help identify areas where detailed testing should be prioritised and allow researchers to track potential changes in mercury risk over time.
Researchers are now working towards a publicly accessible dashboard that could make the model and its data easier to use. Such a tool could help scientists and land managers monitor mercury pollution across large freshwater regions without having to test every individual lake, pond or stream.