For years, entomologists have worried about what appears to be a global decline in insect populations. But data on insect trends can be hard to come by. Scientists have studied relatively few of the some 900,000 living insect species they have named so far, and have yet to name millions more.
Now, researchers say one of the bug world’s deadliest enemies—birds—could offer much-needed help in tracking insect numbers. That’s because studies of birds often contain substantial information about the insects they eat.
Those bird-related data are pretty much missing from current studies of insect declines, says Chris Elphick, a conservation biologist at the University of Connecticut (UConn), Storrs, and a co-organizer of EntoGEM, a research effort launched in 2019 that is combing the scientific literature for insect data. The project’s initial review of bird studies, for example, has unearthed some three dozen that tracked insect populations for 10 years or longer. (The researchers presented those findings at a recent Entomological Society of America meeting and are now preparing to submit a paper to a journal.)
Elphick and Danielle Schwartz, a conservation biologist at UConn who is also involved in EntoGEM, recently spoke to Science about the search for more insect information. The interview was edited for clarity and brevity.
Q: Why should we be looking for more data on insect populations?
Chris Elphick: People talk about global insect decline, but we don’t really know what’s going on globally, because we don’t have data sets from most parts of the world. We have good evidence that insects are declining in lots of places. But it’s also clear that insects are not declining everywhere and that not all insects are declining. The data we have are patchy and biased toward Western Europe and parts of North America. Our hope is that even if [new information] doesn’t change the story, it will give us more confidence in the current story because it will be more comprehensive. We’re losing biodiversity at such a rate that we really need to find ways to use the information that we have already and get it all in one place so that we can use it more effectively.
Q: What made you think to look for insect data in bird studies?
C.E.: It kind of started because we’re not actually entomologists! I’ve studied birds my entire career. Eliza [Grames, an ecologist at the University of Nevada, Reno, who was then a Ph.D. student studying birds], and another graduate student—entomologist Graham Montgomery, now at the University of California, Los Angeles—got thinking about how we do a better job of finding and using data we’ve already collected. Eliza initiated EntoGEM and started developing [software] tools for trying to search the literature more effectively. We started finding papers with data sets that were not mentioned in any of the analyses looking at insect declines. A lot of those papers were on birds because ornithologists were interested in what the birds were eating.
Q: Why were those data overlooked?
Danielle Schwartz: The first thing is the terminology. There’s a lot of different terminology that we’ve seen that wouldn’t necessarily imply insects right off the bat—[such as references to] “protein food.”
C.E.: It may often just be as simple as not putting the right keywords into the front end of the paper, so the search engines just don’t find it. As an ornithologist you might not use the word “caterpillar” in the title, abstract, or the keywords—and that’s the information that people search. [In addition,] if you’re collecting caterpillars because you want to know how much food there is for forest songbirds, you might mention caterpillars but you’re not going to list all of the species. Those data sets don’t have the level of detail that an entomologist might be interested in. But if you’re just trying to get a gross sense of whether forest caterpillars are declining, that might still be a useful data set.
Q: What has EntoGEM done so far with ornithology papers?
D.S.: We did an initial search and came up with 35,018 papers. Before reading through the abstracts, we use another program that Eliza wrote to [filter out those] that wouldn’t be relevant. And then once we start seeing a trend—for example, when I started seeing a lot of papers about snails—I can search by keyword and filter those out.
C.E.: We’re always trying to find ways to streamline [the process]. You can use [initial reviews] to build a statistical model to predict which of the remaining papers are going to be relevant. For every paper, we have a prediction of how likely it is to be relevant.
Q: How many papers have you identified data in so far?
D.S.: Somewhere around 150. … We’ve found something like 40 that have data that span at least a 10-year period. The studies are mostly clustered in Western Europe and North America. But we are starting to find a few examples from parts of the world [with] less information.
Q: What’s an example of useful data you’ve found?
C.E.: One of the first ones that [Grimes] found is this study of Harlequin ducks in Iceland. Harlequin ducks are these pretty fancy-looking ducks that nest on mountain streams. There’s this group [of researchers] in Iceland that has been studying them since the 1970s. They measured the number of chironomids—little midges—every year, because these ducks eat chironomids. They have almost a 30-year data set.
What I like about this duck example is that people don’t generally think about ducks eating insects. And it’s not the kind of place where you would think to go looking for a longtime series on insect populations. It’s an example of the kind of information that’s out there if you just go looking for it.
Q: Besides ornithology, are there other fields that might have overlooked insect data?
C.E.: Absolutely. There’s probably an endless number of places to look. Herpetologists, mammalogists … I’m sure there are botanists who collect a lot of data on insects because insects eat plants all the time. People who do forensics work collect data on insects. Whether they do it in a way that could contribute to understanding population change, I don’t know.