For National Pollinator Week, Learn Why the National Museum of Natural History is Using a State-of-the-Art Conveyor Belt System to Digitize Thousands of Insect Specimens

The ongoing effort captures high-quality images and collection data in seconds, creating an online database of bees, beetles, butterflies and flies

In the rolling hills of the Succulent Karoo — a large, dry biome located in South Africa and Namibia — a large fly as long as a guitar pick hovers from one pink-tipped flower to the next. Scattered shrubby vegetation is interspersed with patches of colorful flowering succulents, a stark contrast against the reddish-brown rocks. As more flies join in on the feast, they use their extremely long proboscis, or mouthparts, to sip floral nectar — while simultaneously covering themselves in pollen. As they move from flower to flower, they spread their dusting of pollen across the landscape.

These pollinating machines are known as the megalong tangle-vein fly (Moegistorhynchus longirostris) and are a keystone pollinator in South Africa. The flies are responsible for pollinating several endemic plant species that are only found in the Succulent Karoo, which is considered one of Earth’s 25 global biodiversity hotspots.

Yet despite their outsized ecological significance, these flies were long overlooked. It was not even known that megalong tangle-vein flies even lived in the northern parts of the Succulent Karoo until research entomologist Torsten Dikow, a curator of flies and aquatic insects at the National Museum of Natural History, observed and collected  one of the flies flitting about during a field excursion in the region in 2009 and uploaded it onto an open-access database.

“I had scientists reach out to me and say the species just didn’t exist there,” Dikow recalled. “It wasn’t until I showed them the photo that it became fact.”

To Dikow, the discovery emphasized that accessible digital records of insect species were crucial to understanding where these tiny yet irreplaceable animals live. “I think that's the greatest thing technology allows us to do: share information across the world and come together.”

The museum is now pursuing this goal on a massive scale with its new pollinator digitization project. Spearheaded by Dikow, with support from the Entomology Department and several other units within the museum and Smithsonian, including the Digitization Program Office, the project currently has funding to digitize more than 325,000 pollinating insect specimens from four taxonomic groups — butterflies, bees, flies, and beetles. A custom conveyor belt system will guide a pinned insect to a set of cameras, where high-quality images are taken from several angles of the specimen and its accompanying labels which contain valuable information on where and when the specimen was collected. These images will then be stored on an open-access database in perpetuity for anyone around the world to use.

The work began last June and by the end of this past May, 64,000 specimens of pollinating bees, beetles and flies have been imaged and are publicly available online. More than 100,000 additional specimens have also been photographed and are in the process of being uploaded.

As the museum approaches the one-year anniversary of this monumental project and celebrates National Pollinator Week, Smithsonian Voices learned more about the importance of creating a digital database of pollinating insects.

Why digitize pollinators?

When most people think of pollinators, honeybees usually come to mind. But in reality, a wide variety of insect species play a vital role in pollinating the plants and crops humans rely on every day. Just among bees alone there are 20,000 species worldwide and many are critical pollinators in agricultural and natural habitats. From tiny megalong tangle-veined flies to British soldier beetles and even cicada killer wasps, each pollinator helps keep ecosystems thriving. Without them, entire natural systems — including agricultural industries — could collapse.

The museum currently houses roughly 2.5 million specimens of insect pollinators, making it one of the largest collections of pollinators in the world. Scientists from all over the globe request access to the collection to aid in their research. There’s only one problem: to see the collection you need to come in person or request that the museum ship highly fragile, sometimes very old, specimens in the mail.

“We have specimens that are centuries old from all around the globe,” said Dikow. “Sometimes there are species we haven’t even identified, waiting to be found in the collections.”

According to Dikow, the digitization project will transform specimen data and high-quality images into a searchable, structured database — open to researchers and the public alike. The goal is to make biodiversity information more accessible, whether for scientific study or public discovery.

When running at full capacity, the new conveyor belt can digitize up to 1,500 pinned pollinators a day. These digital records allow scientists worldwide to study specimens remotely, track changes in species habitat distribution over time and uncover patterns that were previously hidden in drawers.

New Insights from Old Specimens

Once digitized, the specimens can help researchers address a variety of questions.  

These include differentiating between identical looking species. For example, in the 1980s, entomologist Annette Aiello from the Smithsonian Tropical Research Institute was interested in Adelpha butterflies — commonly called Sisters. In the tropics, the butterflies were classified by similar wing patterns. When Aiello raised the butterflies from caterpillars herself, she discovered a striking mismatch: caterpillars that looked alike and ate the same plants matured into adults with differing wing patterns.

“In other words, adult wing patterns appear to have evolved as a kind of deception,” said Aiello. “The museum’s department of Entomology is now digitizing high-resolution images of the 4,107 Adelpha specimens in its collection, opening the door for scientists worldwide to trace patterns, spot imposters and uncover just how deceptive these ‘sisters’ can be.”

Digitization can also help researchers track invasive species. For example, Osmia taurus — a species of mason bee originally hailing from Asia — was introduced to the eastern United States a few decades ago and may now have a negative impact on native bee populations. Scientists speculate they are rapidly spreading and displacing native orchard mason bees, which are essential pollinators for several crops. Research entomologist Seán Brady, the museum’s curator of bees and stinging wasps, said the digitization project will provide the necessary data to track the spread of this invasive bee over time. The data can also aid in identifying other negative ecological impacts resulting from population declines.

“We’re building a historical record with this project,” said Brady, noting that the data will help identify declining populations and pinpoint species absences. "If we have a 100-year-old bee specimen found in a particular place, we can go there now to see if it’s still thriving. If not, it could indicate a problem.”

Looking to the Future

While the conveyor belt efficiently speeds up the digitization process, the information on a specimen’s data-rich label still needs to be entered manually by human workers — a process that takes considerable time and energy. Dikow, along with entomology informatics manager Jessica Bird, are collaborating with computer scientists and entomologists at the University of Maryland to use machine learning and artificial intelligence tools to extract the data from specimen labels and efficiently pinpoint names, locations, collection dates, and other data for upload into the museum's database.

Automating this process will make it easier to identify broader insights. “Another benefit of using AI when deciphering the labels is finding patterns in the data that you just can’t anticipate by entering it manually,” Dikow said.

These patterns can come in the form of distribution maps, which can pinpoint exact locations of where a specimen was collected. Some labels are from decades ago and vary from vague descriptions of local landmarks to precise longitude and latitude GPS coordinates.

Dikow’s ultimate goal for the digitized specimens and automated collection data is creating a more complete picture of pollinator diversity. This in turn will shed light on the health of plants and ecosystems around the world.

“Every living thing is intertwined with each other,” he said "Discovering a new species adds another piece to the puzzle of our Earth. Digitizing part of the pollinator collection makes these puzzle pieces accessible to all and encourages others to discover more about our natural world.”

Erin Wunderlich | Read More

Erin Wunderlich is a science writing intern at the Smithsonian's National Museum of Natural History. She covers natural history research and collections news for the museum’s blog, Smithsonian Voices. Erin is currently a second year science and technology journalism master’s student at Texas A&M University, where she also writes science articles for the university’s newspaperThe Battalion. In her free time, she loves riding horses and exploring nature with a film camera. See more of Erin’s work at https://erinrw2.web.illinois.edu/portfolio/ and https://thebatt.com/staff_name/erin-wunderlich/.