Pollen Sleuths: Tracking Pesticides in Honey Bee Pollen to Their Source Plant

Researchers studying pesticide exposure routes in honey bees were able to combine chemical analysis of pollen and the keen eye of a palynologist—an expert in identifying pollen microscopically—to track pesticide in bee-collected pollen to a source plant genus. A key step in the process involved sorting pollen pellets by color and testing them by group. (Photo credit: Kimberly Stoner, Ph.D.)
By Kimberly Stoner, Richard Cowles, and Brian Eitzer
More than 10 years after the appearance of Colony Collapse Disorder, people are still worried about honey bees, and for good reason. Beekeepers still lose a substantial portion of their honey bee colonies each year. A nationwide bee survey estimated annual loss at 40 percent for 2017-2018. Many factors are likely to be involved, but the one that we study as a team is exposure to pesticides.
We have been measuring honey bee (Apis mellifera) exposure to pesticides in pollen since 2007 by putting honey bee hives in different habitats, collecting pollen using a pollen trap, and then measuring pesticide residues in the pollen using high-pressure liquid chromatography and mass spectrometry. A pollen trap is a clever device, invented by beekeepers, which forces each foraging worker bee returning to the hive to travel through a screen. The screen removes the pollen pellets from the bee’s pollen baskets on her hind legs, whereupon the pellets are collected onto a smaller-mesh screen. Ideally, the result is a collection of pollen pellets, of fairly uniform size, each representing the pollen collected on a single foraging trip by a worker bee.

In a study on pesticide exposure routes in honey bees, researchers used chemical analysis of pollen and the keen eye of a palynologist—an expert in identifying pollen microscopically—to track pesticide in bee-collected pollen to a source plant genus. A sample with a particularly high level of pesticide traces in the study was identified as coming from plants in the genus Spiraea (shown here). The researchers were able to confirm that the staff of the plant nursery where the test hives were located had applied pesticide to Spiraea just before and during the pollen-sampling period. (Photo credit: Alejandro Chiriboga)
In 2015, the Connecticut Department of Energy and Environmental Protection funded us to put hives at three large commercial ornamental plant nurseries for one growing season. At the time, consumers and environmental groups across the country were concerned that ornamental plants treated with neonicotinoid insecticides during production could pose risks by providing insecticidal nectar and pollen to bees visiting these plants later placed in landscapes. In general we found neonicotinoids in the range of 1–4 parts per billion (ppb) in the pollen collected at the nurseries: not much different from those we found in other environments around Connecticut. However, the exception was a series of pollen samples collected at one nursery during the month of August. One sample contained 305 ppb of thiamethoxam and 31 ppb of clothianidin, and several other samples had 10–20 times the average found for these neonicotinoids. The results of our search for the source of this sample, using a novel combination of techniques, were reported Tuesday in Environmental Entomology.
What was the source of these exceptionally high pesticide residues? To investigate, we sorted the pollen pellets by color, and then tested the different color categories for pesticide residues. We sorted out 11 colors of pellets, discriminating the colors and naming them using Pantone standard colors used by interior designers. Three of the colors—”mahogany rose,” “warm sand,” and “almond buff”—had 1.5–2 times the concentrations of thiamethoxam and clothianidin compared to the bulk pollen, while the other colors—shades of yellow, green, and dark brown—had 5 percent or less of the neonicotinoid concentration found in the bulk pollen.
These samples were then analyzed by Andrea Nurse at the University of Maine, a palynologist—an expert in identifying pollen microscopically. She found a higher incidence of pollen from Spiraea in the color categories containing higher concentrations of neonicotinoids. We found that this association held up to statistical testing, and we also did the same analysis on another sample of bulk pollen, from the same hive a week later, and found that the “mahogany rose” pollen in this sample—only 4 percent of the bulk sample—was 94 percent Spiraea, and had 11 times the thiamethoxam and 15 times the clothianidin as the bulk sample.
Other researchers have tried palynology combined with pesticide residue testing to track pesticides to their plant sources but have not been successful in identifying a major plant source. Sorting the pollen pellets by color helped us narrow the potential sources, and working with an excellent palynologist and preparing the pollen to her specifications allowed us to be the first to identify a single plant genus as the source. We are now working again with Andrea Nurse and with a team at Penn State University to combine palynology with molecular methods to improve identification of pollen from ornamental plant nurseries.
We were able to confirm with the nursery staff that they had applied thiamethoxam to Spiraea (the clothianidin is a metabolite of thiamethoxam) on July 29 and August 12, just before and during the series of samples with high concentrations of these neonicotinoids. But the nursery used the same pesticide throughout the growing season, so why did it just turn up in these Spiraea samples from August? We don’t know the answer to that question. Some possibilities are that there might be differences in how Spiraea moves these systemic pesticides into pollen, compared to the many other genera of plants at the nursery, or there might have been an accidental over-application. As usual, solving one mystery leaves more questions to answer.
Kimberly Stoner, Ph.D., is an associate scientist at the Connecticut Agricultural Experiment Station (CAES) in New Haven, Connecticut. Email: kimberly.stoner@ct.gov. Richard Cowles, Ph.D., is an agricultural scientist at the CAES Valley Laboratory in Windsor, Connecticut. Email: richard.cowles@ct.gov. Brian Eitzer, Ph.D., is an agricultural scientist at CAES in New Haven. Email: brian.eitzer@ct.gov.