Boxes and boxes of bees—a Sevilleta LTER dataset highlight

Boxes and boxes of bees

Humble beginnings

The magnitude of an extensive specimen collection doesn’t really hit until you see it. At the University of Mexico’s Museum of Southwestern Biology, research scientist Jade McLellan pulls out one box of bee after another. There are shelves stacked to the ceiling. Some boxes are labeled with things like “big bees,” others with individual species names like Agapostemon angelicus. Opening one reveals a dazzling pattern of the insects, collected at the Sevilleta LTER, sorted, pinned out, labeled, and stored. 

Bee collection at Sevilleta started in 2002, and, like many long-term datasets, began as a passion project for then-technician Karen Wright. Wright, who now runs the Washington Bee Atlas, is still an essential part of the process. “I’m the only one that can identify all those species,” she says. “They’re one of the most difficult groups of insects to identify at the species level.”

After a few false starts—the first collection method Wright tried failed because water in her traps evaporated too quickly in the desert—Wright began sampling bees in three different ecosystems once a month between March and October. Now, the dataset has records of over 350 different species of bees—a staggering diversity. The dataset’s twenty-plus years of data make it the longest continuous record of bee diversity anywhere.

Wright fought hard for her bee identification skills. Shortly after starting the collection, she met Terry Griswold, an entomologist and bee expert with the USDA. “I would fill my Subaru up with boxes of bees and go sleep in Terry’s basement for a week every year,” she says. “Every time I would go, I would learn a different genus of bees to the species level. So little by little we identified all of those bee species. And then I started building a reference collection at the Museum of Southwestern Biology.”

Core to the site

The bee dataset is now part of the Sevilleta LTER’s core sampling. It’s a key catalogue of biodiversity at Sevilleta. Plus, those bees pollinate a staggering diversity of plant life at the LTER; paired with Sevilleta’s plant species core dataset, bee data can reveal plant-pollinator interactions and might show how things like drought, a persistent problem in the southwest, changes the bee community.

Researchers have started to probe the dataset depths to track long-term changes in bee diversity or to study bee responses to climate change. A recent study, for example, showed that larger bees that were more resistant to heat and drying out fared better as the Sevilleta got warmer, drier, and the climate became more variable. Still, though, 46% of bee species showed declines over the same period, and while some bees might thrive in the new climate, the bee community became less diverse. 

Another student is trying to use DNA remnants on flowers to show which species of bees visit which plants. The work harkens back to the reason Wright started the dataset in the first place. “How can a desert be so diverse and have so many different beautiful flowers?,” she remembers thinking. “It’s got to be the bees, right?”

Wright stresses that there is still immense potential for publication using just the dataset itself, from simple studies cataloguing bee diversity to complex studies that relate bee community dynamics to climate change. That potential is part of why Jenn Rudgers, lead PI of the site, brought this dataset into Sevilleta’s core sampling nearly a decade ago. Now, “all the elements are in place to dive into the stories the bee data can tell us,” she says.