How’s your fern and bird coverage, BHL?

“Every one knows what a bird is,” asserts an early 20th century book that I found while browsing the Biodiversity Heritage Library (BHL).

As I’ve learned during my Professional Development Internship with Jacqueline Chapman at Smithsonian Libraries this summer, it’s not always that simple. Taxonomy is ever-changing, especially at the granular level needed by subject specialists around the world who use BHL to conduct research on organisms ranging from mosses to turtles to fungi.

BHL is a consortial digital library whose member libraries digitize works in natural history and botany based on both user requests and subject librarians’ selections. My project for this summer was to refine a collection assessment methodology for BHL using both taxonomic and bibliographic analyses. Along the way, I’ve learned valuable lessons in using library tools, troubleshooting in Python (a computer programming language), and understanding the thought processes of 19th century ornithologists and pteridologists.

Becca Greenstein

Last year, Jacqueline worked with Robin Everly, the Smithsonian’s Botany and Horticulture Librarian, to conduct a taxonomic and bibliographic analysis to assess the depth of the BHL’s fern and lycophyte literature. They presented their results at an international conference on ferns, Next Generation Pteridology, and had the unique ability to talk with many subject-specialist users from around the world. Jacqueline later shared this proof-of-concept with researchers at TDWG in Nairobi, Kenya.

For the bibliographic portion of the project, Fern Books and Related Items in English before 1900 was used to create a list that could be referenced to determine whether a book was available on BHL, and if not, if we had access to it. A year later, I furthered this analysis by seeing what has changed in the past year and making requests for partner libraries to scan items to add to the collection. I enjoyed gathering data for books with titles such as Greenhouse Ferns and the Romance of Plant Life, Rambles in Search of Ferns, and The Fern Paradise: A Plea for the Culture of Ferns (2nd edition in BHL).

As the bibliography used included all editions of a particular work, regardless of whether the content had changed, I decided to not digitize the 53 works on the list whose content was already in BHL in another edition of the same work. As you can see in the graphs below, the number of fern books on BHL from this list has increased by 36% over the past year. The 112 titles from the list that are not yet in BHL but that we have access to via partner libraries will be in BHL after they are digitized. We lack access to only 37 of the titles on the list that would add content to BHL, and it will be interesting to follow up with this study to see if current partners acquire new resources or if new partners that possess these materials join the BHL Consortium.

2015 Bibliographic Analysis: Graph presents percentage of books from the list generated using Fern Books and Related Items in English before 1900 that are in BHL, are not in BHL but are held by a BHL partner, and are neither in BHL nor held by a BHL partner.

2016 Bibliographic Analysis, showing that the BHL collection of fern books has increased from 2015 to 2016.

For the taxonomic portion of the project, BHL’s coverage of a particular taxonomic grouping using scientific names was analyzed. The digitized material on BHL is in the form of images, which the computer does not recognize as text. Using Optical Character Recognition (OCR), the images are converted to machine-readable text. Taxonomic Name Recognition (TNR) then searches the OCR to find scientific names using multiple recognized lists of scientific names.

To use this powerful analytical tool to analyze BHL’s literature on birds, I upgraded the Python 2 code used for last year’s analysis to Python 3, the newest version of the programming language. Using my code, I counted the number of mentions in BHL of each genus of birds that appear in Catalogue of Life, as determined by TNR, to identify potential gaps in the BHL collection.

Of the 2234 genera analyzed, 99.6% of them are mentioned in the BHL corpus, 131 individual genera had more than 10,000 mentions in BHL, and 88% of them had more than 100 mentions.

I conducted an in-depth analysis of the 37 genera with fewer than ten mentions in BHL to figure out possible reasons for the paucity of literature. I determined that this lack of literature could be attributed to such things as the more-recent description of some of the genera, such as within the past 20 years, to the locality of some genera, as in some birds being endemic to far-away (to 19th century European ornithologists) places like New Guinea and Mozambique, and to taxonomic changes to the genera over the years. I then looked for the first mention of each of the 37 genera in books and journal articles online and in print, in addition to submitting scan requests for the books we have access to that weren’t already in BHL. There was something surreal about trekking up to the Birds Library, which is tucked away on the sixth floor of the National Museum of Natural History, finding Ornithologische Berichte on the shelf (and no, I don’t speak German), and opening to page 118 to find Wilhelm Meise’s initial description of Stresemannia bougainvillea.

Meise’s initial description of Stresemannia bougainvillea is next to my thumb.

My internship lasted six weeks, but it did not feel like that long. I hope that BHL will use my code to analyze larger sets of data and/or data at a higher level (for example, how is BHL doing at collecting literature on Kingdom Animalia?).

Through conducting my project, I’ve learned that things you learn in library school really do apply to the real world, how an academic library at an institution without students functions, and the workflow behind digitizing materials that appear in BHL and on the Smithsonian Digital Library. I’ve learned that library tools we take for granted can be unreliable, but aren’t usually, and that getting help from people who do research on ferns and those who do speak German can be very beneficial. I hope to bring the things I’ve learned back to my final two semesters of library school, as well as into my hoped-for career as a science librarian after I graduate.

________________________

About the Author

Becca Greenstein is getting her Master’s in Library Science at UNC-Chapel Hill. For her Bachelor’s degree, she went to Carleton College, where she majored in Biology and minored in Chinese. After graduating from Carleton, she worked as a lab technician at the University of Minnesota before starting library school. After she graduates, she hopes to continue honing these skills while working in an academic or special library as a science librarian.