With the continuing growth of community-driven, digital nature observations on platforms like iNaturalist, entomologists and other scientists are tapping into that store of knowledge to make new discoveries. A new study shows the potential for photo observations to answer species-specific behavior and ecological questions, such as the flower preferences among two soldier-beetle species, Chauliognathus marginatus and Chauliognathus pensylvanicus (shown here). (Photo by Curtis Eckerman via iNaturalist, CC BY-NC 4.0)

By Carolyn Bernhardt

By replacing physical samples with multimedia assets from iNaturalist to explore the unique flower use of two beetle species, a team of researchers has unveiled an innovative, efficient, and cost-effective approach to ecological investigation. By the research team’s estimation, their study is the first to use a dataset derived from citizen-science photo observations to address species-specific behavior and ecological questions.

In the study published in July in Annals of the Entomological Society of America, the researchers analyzed digital samples—photos, in other words, plus accompanying metadata—from community-collected observations that offered valuable insights into the ecological behaviors of two soldier-beetle species, Chauliognathus marginatus and Chauliognathus pensylvanicus. The scientists chose these beetle species for their distinct appearance, daytime activity, and close association with flowers—all qualities that make them popular subject matter among gardeners and naturalists, says Katlyn Catron, Ph.D., who led the project while at Virginia Tech (and is now in the Department of Entomology at Washington State University). The study analyzed digital specimens of beetles on flowers and, separately, of flowering plants potentially available to the beetles to find differences in flower colors, families, and genera utilized by each species.

Entering a Digital Landscape

iNaturalist was created by University of California, Berkeley, students as a final master’s project in 2008, established as an LLC in 2011, and joined with the California Academy of Sciences in 2014 and National Geographic in 2017 before becoming an independent nonprofit organization just last month. The platform allows users to upload nature photos, video, and audio and offers users identification and other helpful information, applying a “research grade” status to media assets that have met certain criteria, including achieving two-thirds agreement among users.

To Catron and her collaborators, iNaturalist represents a deep reservoir of community-collected digital specimens suitable for entomological data mining. This process allows scientists to observe predator-prey interactions, species variations, and plant-pollinator relationships without encountering accessibility barriers, such as traveling or mobility limitations. But, she says, “The appropriateness of the research questions needs to be evaluated, just like it would for any other study.”

The researchers often saw C. marginatus with white flowers and C. pensylvanicus with yellow flowers, which they had expected based on field observations. However, this was not a result of the flowers available—both species were with flowers of their preferred color more than twice as often as predicted by flower availability. And preferences for other flower colors, like red, pink, or purple, were not significantly different between the two species. Both beetles strongly preferred flowers in the Asteraceae family (composites), but C. marginatus used a wider variety of flower families compared to C. pensylvanicus.

However, the study didn’t include smaller-scale geographic factors like specific beetle habitats or local climates, which might affect real-world flower availability. Other factors like nectar structure and predators could also influence flower use. Catron says larger studies with matched flowers for each species would refine and clarify the paper’s findings.

Of course, humans are naturally drawn to large, unique, or flashy specimens, so the inherent bias in the pictures people choose to take and upload on platforms like iNaturalist affects which questions researchers can ask. But this study intentionally utilized this bias.

Catron also notes that previous research on digital specimens found they were not necessarily more biased than museum or university collection specimens. “Community-gathered digital specimens are not a one-size-fits-all dataset,” says Catron. “It’s continually growing and evolving, but it’s still only going to be able to answer certain questions.”

Catron says the study’s highly collaborative nature helped prevent implicit biases and internal methodological issues, expand the range of questions she could explore with the data, and enrich the analysis. Entomologists, a plant pathologist, and a quantitative ecologist helped enhance the methodology by using their expertise to ensure the team used logical and robust approaches to address the research questions. Catron says that working with a quantitative ecologist helped her identify the potential of the data found in iNaturalist’s assets and how to analyze it effectively. This collaboration also allowed the research team to develop a broadly applicable workflow for the study, which could be valuable for similar research on different organisms.

Researchers behind a new study evaluating the potential for photo observations from iNaturalist to answer species-specific behavior and ecological questions say the associated metadata offers a rich understanding of the subject matter at hand. “Each photo or video documents a subject’s interaction with its environment, habitat, conspecifics, predators, prey, etc., occurring at the moment of the recording,” says Katlyn Catron, Ph.D. “Instead of a dead specimen in a box with a (potentially very informative) label, you get a literal snapshot of that organism’s life and what was happening around it, which can be equally as informative as a traditional specimen.” (Photo by Lisa Benjamin via iNaturalist, CC BY-NC 4.0)

A Picture is Worth 1,000 Observations

Catron emphasizes the importance of analyzing metadata from digital specimens and encourages more studies like this one. Looking at the image or video responsible for certain observations offers a richer understanding of the subject matter at hand, not only because each viewer can extract unique insights but also because of how many contextual clues each photo or clip can offer.

“Each photo or video documents a subject’s interaction with its environment, habitat, conspecifics, predators, prey, etc., occurring at the moment of the recording,” Catron says. “Instead of a dead specimen in a box with a (potentially very informative) label, you get a literal snapshot of that organism’s life and what was happening around it, which can be equally as informative as a traditional specimen.”

So, Catron says it is crucial to revisit the original pictures and understand the process of that observation’s genesis. In doing so, scientists can uncover fresh and distinct layers of observations that build upon initial understanding.

As such, Catron feels this method allows for broader data analysis and could be valuable for understanding species interactions, tracking wildlife disease outbreaks, or studying various plant and animal species. Using such data can uncover useful information that might otherwise be overlooked. In the future, with appropriate study design and analytical methods, Catron says these types of data could be immensely valuable for investigating a wide range of entomological and ecological inquiries.

Catron hopes to replicate this study with endangered pollinators one day to see if they use the nectar sources researchers suspect or frequent any unexpected plants. That way, local-level resources could go toward more strategically planting nectar. “That would be a full-circle example of community science,” she says. “We would be using pictures uploaded by the community to do scientific research that informs public policy to benefit the community.”

Now, Catron is a pest biologist at the Tree Fruit Research and Extension Center at Washington State University, studying leafhopper vectors for cherry tree X-disease. She and her team have noticed many leafhopper observations on BugGuide.net that still need to be identified and could be valuable for understanding the timing of leafhopper generations. She says that early pictures of adult leafhoppers could affect sampling and modeling efforts, but having more photographic observations can be “like having extra boots on the ground to monitor and address the issue.”

Carolyn Bernhardt, M.A., is a freelance science writer and editor based in Portland, Oregon. Email: carolynbernhardt11@gmail.com.