Department of Biology

An analysis of a global dataset of collected and observed bees shows that fewer species are being found in most recent decades.

Although the word “bee” evokes in most people the image of the domesticated honeybee, Apis mellifera, more than 20,000 species of bees have been described from all over the world. Bees are the most important group of pollinating insects and are key not only to the reproduction of hundreds of thousands of wild flowering plant species but also to the yield of about 85 percent of all cultivated crops.

Bees are also strongly attuned to their environment, and their subsistence is dependent on the availability of flowering plants and adequate nesting sites. Consequently, they are vulnerable to transformation of their native habitat into large-scale agricultural and urban zones. Indeed, an increasing number of studies looking at the local or regional status of specific groups of bees are reporting decreases in their abundance and diversity.

Now, a newly published analysis of public occurrence records suggests that these reports might be local manifestations of a global phenomenon. In the study, published in the journal One Earth, researchers looked at publicly available data of bee specimens collected or observed in the period between 1946-2015 and counted how many different species were found all over the world each year. They found that although the number of yearly records has been steadily growing since the middle of the century, the average number of species in the dataset has been decreasing since the 1990s.

“The 1990s mark the beginning of a globalization era that saw a large acceleration in land-use transformation worldwide, often resulting in a replacement of diverse habitats by monocultures, along with increased used of agrochemicals; this change, compounded with climate change and international bee trade and consequent pathogen introduction, represents a tough challenge to wild bee populations” said Eduardo Zattara, a research associate in the Indiana University Bloomington College of Arts and Sciences' Department of Biology and lead author of the study. “If there is an ongoing global decline of bees, so that uncommon species become rare and rare species become rarer or even locally extinct, we could predict that the likelihood of finding at least one individual of such species in a given year would decrease over time. If we had a sample of bees collected all over the world over a long period of time, we could test this prediction.”

Finding such a dataset was a goal of Marcelo Aizen, researcher at Argentina’s CONICET and co-author of the study.

“Inspiration finally came while participating in a workshop for the SURPASS project, an international collaboration aimed at safeguarding pollinators and their services,” said Aizen. “During the workshop, we realized that the Global Biodiversity Information Facility (GBIF), which gathers specimen occurrence data ranging from historic museum collections to recent citizen science initiatives, could hold the data we were looking for.”

Zattara and Aizen teamed to analyze the data and found there the declining trend in the number of reported species that would be expected if bee populations were dwindling.

“At face value, the results present a gloomy situation for bees everywhere; however, it is important to remember that the very heterogeneous nature of the GBIF dataset is expected to introduce numerous sources of bias that could generate trends not reflective of the actual biological reality,” warned Zattara. “For example, shifts in museum collection strategies away from rare species, increased density of observations in well-populated areas away from biodiversity hotspots, or a loss of taxonomic expertise leading to fewer specimens being identified to species level, could all contribute to inflate artificially the perceived decline. We found, however, that none of these sources of bias could on their own explain the observed trends better than a real loss of wild bee diversity.”

Checking for the effect of the number of records and potential sources of bias in the dataset led to discovery of another pattern: as more records were available for a given continent or country, confidence in the temporal trends of bee diversity increased. Furthermore, results of the analyses showed better agreement to existing reports of regional bee diversity trends when larger numbers of records were available for the studied region.

“This prompted us to emphasize a second take-home message about data sharing and mobility,” said Zattara. “There are a lot of historic data out there, sitting in offline computers, filing cabinets, field notebooks, and museum collection racks; those data are key to filling the temporal and geographic gaps of the GBIF dataset. We need to make a concerted effort to fund further digitization of worldwide collections and encourage data holders to mobilize and share their data with public databases; this is true not only for bees, but for data on all living organisms.”

"This is a novel analysis that takes advantage of the growing availability of online data on biodiversity, including bees,” commented David Inouye, a pollination biologist and professor emeritus at the University of Maryland. “Given the paucity of programs monitoring native bee diversity, distribution, and abundance, this is one way to look for trends. It's unfortunate that the analysis shows such a dramatic decline, and this reinforces the need to establish more monitoring programs.”

“Although our conclusions paint a stark scenario for bees, we do not want people fixating on specific values (e.g., of percent decline),” concluded Zattara. “Instead, we hope this study will bring the issue of bee diversity decline to the spotlight, and have scientists, policymakers, and many other stakeholders begin discussing how to take action to get better data and to address the drivers of this decline.”

Eduardo Zattara is also a research associate at the Department of Invertebrate Zoology at the National Museum of Natural History (Smithsonian Institution, Washington, D.C., U.S.A.) and Marcelo Aizen, a current fellow at the Wissenschaftskolleg zu Berlin (Germany). Eduardo Zattara and Marcelo Aizen are both researchers in the Pollination Ecology Group at the Institute for Research on Biodiversity and the Environment (CONICET and National University of Comahue, Bariloche, Argentina). The latest version of the manuscript prior to peer review is available at bioRxiv.