Grad student curatorial project: U-M Herbarium begins digitizing its wood collection

What rhymes with herbarium? Xylarium! What’s a xylarium, you may ask. The University of Michigan Herbarium’s wood collection (aka xylarium) contains more than 3,300 specimens representing more than 130 tree families, primarily tropical woods. Most specimens are small branch cuttings, rectangular blocks, and large stem cross-sections.

Jordan Bemmels, a graduate student in the Department of Ecology and Evolutionary Biology, recently undertook a curatorial project to begin digitizing the herbarium's wood collection as part of his graduate student curatorial assistantship. In addition, he established protocols so that other GSCAs will be able to continue the effort at MICH, as the U-M herbarium is commonly known in herbarium circles. Bemmels’ project is part of the ongoing efforts of the herbarium to digitize its more than 1.7 million specimens. Bemmels' advisor is Professor Christopher Dick.  

“The herbarium had a bunch of old boxes of wood sitting around that hadn't been touched for a long time, and we had no digital record of what was in the collection,” explained Bemmels. Fortunately, however, there were typewritten index cards several decades old listing the boxes’ contents. Bemmels transcribed this information to create a wood specimens’ database. Next, he learned how to use the museum's database, Specify, used to keep and publish specimen records. Then, he figured out a digitization workflow for attaching barcodes to the wood specimens so each will have a unique identifier, associating the wood specimens with pressed voucher material, taking photos of voucher material, and uploading information into the database.

Vouchers are pressed and dried samples of leaves and/or flowers that were collected from exactly the same plant from which the wood was collected. “This is important because it is difficult to determine what species the sample represents just from looking at the wood,” Bemmels explained. “The pressed voucher material (especially the flowers, but sometimes also the leaves or fruit) is the physical evidence that experts use to determine the species’ identity.”

Some of the herbarium material serves as vouchers for external wood collections, including at the U.S. National Herbarium. The MICH collection is in the process of being fully digitally curated.

Much of the wood in the xylarium is from Sumatra, Belize and Guatemala, Mexico, Australia, the West Indies and Paraguay. The largest family in the wood collection is Euphorbiaceae (spurge), followed by Rubiaceae (coffee), Fabaceae (legume), Myrtaceae, Lauraceae, and Meliaceae.

Bemmels received training from several people, including herbarium project managers. He primarily worked with Beverly Walters, collection manager of vascular plants, and John Torgersen, database administrator for the LSA museums. From September to December 2016, Bemmels assessed the scope of the collection, established workflows, and digitized about seven percent of the collection.

“The project made me really aware of the connections between biological research and world history,” said Bemmels. “Many of the collections are from the end of the colonial era and I learned about how Western scientists would go to colonies in tropical countries on collecting expeditions to document and describe biodiversity there. This often involved interesting cultural exchange as well.”

For example, collector Harley H. Bartlett was an amateur ethnographer who collected anthropological artifacts from the Batak civilizations in northern Sumatra (Indonesia today, then part of the Dutch East Indies), which are part of an exhibit at the U-M Museum of Anthropology.

Some of the collections bear information about species’ names in native languages. The collection from Sumatra by Rahmat Si Boeea has Batak text written on the wood samples and herbarium voucher sheets. While Batak is still spoken in Sumatra, the alphabet has gone extinct in daily life. Therefore, these writings are not translatable. The Paraguay collection has common names of species written in Guaraní and common names are written in Mayan on samples from British Honduras (Belize today).

Bemmels discovered information in a collector's notebook about how Percy H. Gentle's wood samples were lost in the Great Belize Hurricane of 1931, demonstrating how scientific research can be vulnerable to historical and natural events.

On the other hand, history, he points out, can be an impetus for scientific research. Samples by Randolph Taylor from Bikini Atoll in the Marshall Islands were collected in 1946 just a few months before the U.S. conducted atomic bomb testing on the islands. Marshall was employed as the botanist on Operation Crossroads (the codename for the testing) and conducted surveys of the flora of the islands before they were destroyed by the atomic bombs.

Much of Bemmels’ own research requires information about the geographic distribution of tree species. “Carefully digitized museum records indicating the collection locality are one of the best ways to obtain this information in general, especially for tropical trees which are relatively rare and have not been extensively studied,” he said.

The oldest sample Bemmels digitized was of poison ivy collected in 1910 by Harley H. Bartlett in South Carolina. “I didn't touch it with my hands to test if the toxic compounds were still active.”  Most of the collection is from the first half of the 20th century, especially from the 1920-1950s and so there could be older specimens that have yet to be digitized.

Read more about the collectors and view the Xylarium database