5,000 Canadian Students Mapped Wild Bee Communities Using DNA from Bee Hotels
Wild bees -- the roughly 20,000 species that are not managed European honeybees -- are the primary pollinators of most flowering plant species on Earth. They pollinate wild plants that underpin ecosystems, as well as agricultural crops that would otherwise require expensive hand pollination. Tracking how wild bee communities change over time, and where specific species are found across landscapes, is a prerequisite for conservation action. But gathering that data across a country as geographically vast as Canada presents a serious logistical challenge.
One approach is to enlist non-scientists. A study published in the journal Metabarcoding and Metagenomics reports on a national experiment that did exactly this: deploying approximately 5,000 students at schools across Canada to build and monitor artificial nesting structures called bee hotels, then using DNA analysis to identify the species that used them.
How Bee Hotels Work as Data Collection Tools
Bee hotels contain bundles of hollow tubes or drilled wooden blocks that mimic the natural cavities -- hollow plant stems, holes in dead wood -- that cavity-nesting bees use for egg-laying. A female bee enters a cavity, deposits an egg along with a pollen ball to feed the larva, and seals the chamber with mud or leaf material. The nest contents -- pollen, larval material, and sometimes the eggs themselves -- contain DNA that can be extracted and sequenced.
DNA barcoding uses short, standardized genetic sequences to identify species. When the DNA extracted from a bee hotel is matched against a reference database of known bee species sequences, it identifies which species visited and nested at a given site without requiring the collector to have any taxonomic expertise. This is the key feature that makes the approach compatible with student participation: species identification does not depend on the ability to distinguish closely related bee species visually, which requires years of specialist training.
Students built the bee hotels according to a standardized design, placed them at their schools or in nearby habitats, and collected nest contents at the end of the season for DNA analysis at a central laboratory. The standardization of the data collection protocol -- same hotel design, same collection timing, same analysis method -- was essential for making the data comparable across sites.
What the Data Revealed
The study identified multiple wild bee species at sites across Canada, with community composition varying by geographic region in patterns consistent with the known distributions of cavity-nesting bee species in North America. Certain species appeared only in specific regions; others were more broadly distributed. The resolution of the data was sufficient to detect geographic structure in bee communities -- a finding that requires reliable species-level identification across many sites, not just presence/absence at individual locations.
The researchers concluded that the student-generated data met the quality threshold for legitimate conservation science use. That conclusion is significant because citizen science data has historically been viewed with some skepticism, particularly for species identification tasks where non-expert observers make errors that can bias community-level conclusions. The use of DNA barcoding largely removes observer identification error from the equation -- one of the key technical advantages of the approach for large-scale citizen science programs.
What the Study Cannot Determine
Bee hotels preferentially attract cavity-nesting bee species -- a specific guild that represents only a portion of wild bee diversity. Ground-nesting bees, which include many of the most ecologically important species in North America, are not captured by this method. The community composition data generated reflects cavity-nesters specifically, not the full spectrum of wild bee diversity.
The study also represents a snapshot in time rather than a trend analysis. Changes in bee community composition over years or decades require repeated surveys at the same sites -- something the program's future depends on maintaining consistent methodology across years. Geographic coverage, while national in scope, was not uniform; schools in remote areas and regions with sparse school density were underrepresented, which can affect interpretation of geographic patterns.
The Educational Dimension
Beyond the conservation data produced, the project engaged students in authentic scientific research through a meaningful task: building something, monitoring it, and contributing to a dataset used in real analysis. This type of authentic participation -- as opposed to activities that simulate science without producing genuinely useful data -- has documented educational benefits for science engagement and environmental stewardship. Whether students showed measurable changes in knowledge or attitudes was not analyzed in this study, but the conservation value of the dataset is clear from the results.