Symposium looks at new frontiers in collecting & collections

On day two of the Botany 2020 conference, I attended the symposium “Biodiversity Research Collecting is More Important Than Ever – Ushering in a Collecting Renaissance,” put on by the Society of Herbarium Curators (SHC) and iDigBio. The eight talks given explored four different themes: smart collecting, new goals for collecting, new collecting tools, and new species discovery. They also asked the question, ‘How do we best leverage new advances in this area?’ Historically, collections were primarily used for floristic and taxonomic purposes, but in the last few decades, these uses have expanded greatly. “Recent advances in data resources, technologies, public engagement strategies, research coordination, and finding opportunities position the community well for a renaissance of biodiversity specimen collecting to address big societal and scientific challenges,” said Patrick Sweeney, president of the SHC, in his introduction. “The symposium is designed to embolden the collecting and collections communities to new ambitions.”

The first talk, given by Sweeney, looked at how well collections represent nature for a well-collected and long-studied area. Sweeney focussed on New England, which has a long history of collection, exploration, and regional scientific societies, as well as a robust recent digitization effort. Comparison of herbarium records with satellite imagery for a dozen different vegetation classes revealed that the two don’t match up well, either for vegetation type or density. Herbarium records showed a sampling bias toward collections made close to roads or herbaria themselves, at lower altitudes, and with herbs and trees over-represented, to name a few. Sweeney recommended orienting future collecting around known gaps and collecting with trait biases in mind, such as growth form, colouration (or lack thereof), very small plants, and spiny taxa that are more difficult to collect.

Thematically similar was Mary Ann Feist’s talk on botanical forays. Feist’s group at the University of Wisconsin herbarium hosts an annual, volunteer-based, weekend collecting expedition – a foray – to a strategically targeted under-collected area of the state. In addition to addressing bias in the herbarium’s collections, the forays help to build a community around the facility, protecting it and educating the public on what it does. The forays draw large numbers of volunteers and are complimented by equally well-attended winter voucher processing workshops. Feist notes the concept has now spread to at least four states.

Three talks, by Barbara Thiers, Caleb Powell, and Bonnie Isaac, focussed on using technology to create “born extended” and “born digital” specimens. Thiers advocated collecting and linking additional types of data that extend the utility of a specimen, such as its associated DNA, fungus, and seeds. Though this would likely slow collections and require new data portals and infrastructure, it would serve to diversify the users of herbarium specimens and give collections greater visibility. Powell and Isaac discussed using the apps CollNotes and iNaturalist, respectively, while working in the field. The open source CollNotes allows data to be entered and immediately turned into labels without a transcription step, saving time and effort. The popular app iNaturalist can be used to link photos of a specimen before collection to its eventual voucher using QR codes that can be attached to the voucher and scanned. Having linked photos helps preserve information on ephemeral features like flower colour that are often lost in the drying process.

Austin Mast’s talk centred around the fascinating idea of anomalies in collected specimens being an important source of information on change in nature. “Anomalies are our bread and butter,” he said to collectors, because they aren’t excluded from collections as they would be from an experimental data set. At present, discovery of anomalies noted in field books and on collection labels tends to be ad hoc and disorganized. To combat this, Mast identified anomaly terms and used a citizen science initiative called “How Weird Is That?” to assemble a training set for machine learning to automatically detect these terms. The resulting algorithm is now able to do so with high accuracy.

Finally, two talks by Lucas Majure and Kelsey Yule, focussed on on-the-ground data and specimen collection. Majure discussed his work with the biodiversity of the Greater Antilles and their “missing taxa.” He has studied the specimens of a number of earlier collectors in the region and found new species hidden in the herbarium, as well as using these to inform his own collecting. Majure emphasized the importance of floristic work in cataloguing species at risk before they’re lost, and of using herbarium specimens to inform taxonomic placement of new collections. Kelsey Yule spoke about the National Ecological Observatory Network (NEON) biorepository as a case study in the use of specimen collection to monitor environmental change. NEON is a 30 year project with 181 data types being collected at 81 different sites across the United States. Yule drew a distinction between natural history and biorepository collections, explaining that the latter is focussed on monitoring change, tied to fine-scale environmental data, and meant to be destructively used, as opposed to being preserved for perpetuity. NEON data is already being used in a number of research studies and efforts are being made to meaningfully integrate the data with natural history collections.

Overall, the symposium made a strong case for the ongoing and ever-expanding utility of natural history collections and explored some of the ways field scientists can make them even more useful and relevant in the digital age.