Putting Bees in a Blender to Save Them
Posted by Richard Conniff on July 11, 2015
If you like food, you had better like pollinators, because you eat their work. Bees, hoverflies, butterflies, and other pollinators are essential to the production of 60 percent of crop species and 35 percent of total crop production. Apart from putting food on our tables, their services are worth about $200 billion a year worldwide. And the problem for farmers, conservationists, and food lovers alike is that pollinator populations are collapsing everywhere. They’re under assault from pesticides, pollution, habitat loss, and climate change, among other factors.
To fix the pollinator crisis, researchers need to know which species are declining, and under what circumstances. But that’s generally a slow, costly, cumbersome process, requiring highly trained taxonomists to prepare a species and identify it under a microscope. It can take years to get the results—and there aren’t enough taxonomists to do the job, in any case. So instead of studying them in minute detail, some researchers now think mashing pollinators into a soup may be
a better way to save them.
Writing in the journal Methods in Ecology and Evolution, a team of researchers from China and the United Kingdom described their pilot study to rapidly identify bee species from samples collected around southern England. The collecting itself was a low-tech affair, using blue, yellow, and red plastic bowls filled with soapy water. (They’re called pan traps.) The colors attract bees and other insects, which then become caught in the soapy water.
For their analysis, the scientists blended the samples into DNA soup and ran it through a sequencer, generating millions of short DNA sequences. Computers then flipped through a reference database of mitochondrial genomes (mitochondria are a cell’s energy factory and carry their own genomes) to match the DNA-soup sequences to telltale sequences in the mitochondria of particular bee species. In the proof-of-concept test, this method sorted 204 samples into 33 species. Instead of years, said coauthor Douglas Yu, the process can take just a few months.
Both the United States and the United Kingdom are now debating how and whether to establish a monitoring program to track pollinator decline. According to a 2012 analysis, a realistic regional or national program would require sampling 200 sites every other week in the first year, and again in year five. That would yield an estimated 1.3 million insect specimens, and including the slow business of getting them identified, it would cost upwards of $2 million. But this assumes a “work ethic and an environment like what you’d see in a Foxconn electronics factory,” said Yu, an ecologist at the University of East Anglia and China’s Kunming Institute of Zoology. Moreover, the five-year interval would be useful “if a species is dropping like a stone” but not much good at detecting a more gradual decline—or at least not in time to do much about it
Yu already relies on a variant of the “bee soup” technique—using “leech soup” instead—for a study in Southeast Asia. He sequences blood from leeches as a low-cost way to monitor the mammal populations on which they have been feeding.
He believes the bee counterpart could become a standard tool in the not-too-distant future for farmers and park managers to monitor pollinator populations on a regular basis, much as they they now monitor rainfall.
They might even skip the pan traps, he said, and simply drive around the property with sticky tape on the grill. The driver would then simply remove the day’s insect collection, put it in preservative, and ship it off for analysis. The data, shipped back by email, could verify that a farmer is meeting the requirements in the environmental incentive programs now available in the United Kingdom. They could also clue farmers in when they need to change their regimen to get better results.
The proposed “metogenomic monitoring” method doesn’t rely on the conventional PCR (polymerase chain reaction) technique, which takes a small section of DNA and amplifies it for analysis. Instead, it sequences the raw DNA in a batch. That makes it possible to determine how often a particular species appears in the batch, a good indicator of how large a population survives in the wild. It also leaves open the possibility of going back to the same results later to identify other species—hoverflies, for instance, or beetles. According to Yu, you could even detect species of mites—a factor in honeybee decline—or bacteria types. “You could go back and say, ‘We think a virus entered the population at this point.’ So the data gets more and more valuable.”
Asked to comment on the new study, Sam Droege, a bee taxonomist and coauthor of the new book Bees: An Up-Close Look at Pollinators Around the World, predicted that “our ability to process, monitor, and accurately ID bees (and other insects) will soon be orders of magnitude greater than we have now, allowing surveys and monitoring to generate insights completely opaque to us now. In such samples are not only bees but wasps, beetles, skippers, and flies that all play important but lesser roles in overall pollination but are almost uniformly ignored, primarily due to identification problems.”
So far, the only objection to the “bee soup” technique is that it entails killing bees. But so does every other technique of properly identifying them—and identifying them is the only way to monitor their well-being. The ultimate result should be to increase pollinator populations and get them back to work on farms and in gardens everywhere.