From Plant Press, Vol. 10, No. 3 from July 2007.
More than 35 years after the United States National Herbarium designed the first digital plant specimen database, we are continuing to make improvements to this unique resource and a significant contribution to the electronic distribution of museum information. With funding from the Mellon Foundation through a grant to W. John Kress and Rusty Russell, we are participating in a multi-institutional effort to produce digital plant species information for Latin America and consolidate these data in an online presentation under the auspices of Aluka <http://www.aluka.org>, an international non-profit organization dedicated to constructing digital libraries of academic resources.
Dubbed the Latin American Plant Initiative (LAPI), the initial task supported by Mellon is to database and image all type specimens from Latin America. For many of the participating institutions, this project represents their first attempt to identify and organize the type specimens in their custody. For others, it allows them to fully digitize their type holdings. For the U.S. National Herbarium, whose pioneering efforts at digitization began in 1970 and has resulted in the largest, fully verified and completely imaged type specimen collection (supported in part by a grant from the National Science Foundation), our participation allows Aluka to take advantage of the considerable resources we have amassed over many years. It also permits us to significantly improve upon what is already the most complete type specimen resource in the botanical community.
In the middle of the last century, when the types were segregated from the general herbarium, the decision was made to leave duplicate types (either isotypes or syntypes) in the main collection. This was viewed as both a convenience to researchers and a form of protection by hedging against loss. But ever since the Type Register project began in 1970, this decision has been debated. A few years ago the Collections Advisory Committee recommended that all verified type specimens should be segregated from the general collection. This segregation included all the left-behind specimens, which had been stamped “Additional Material”. The reasoning was that taxonomic studies required the information for all existing types, especially all syntypes for which lectotypification was eventually needed. Our ability to produce and make available the digital versions of these specimens made the convenience argument moot, and supplanted the protection argument. However, because the numbers were expected to be quite large, and the personnel to process them was lacking, there was no concerted effort made to separate them at that time.
Our participation in the Latin American Plants Initiative allows us to accomplish two primary objectives. The first is to address our need to consolidate all Latin American type specimens by extracting, databasing and imaging the specimens which are currently located in the general collection. This includes the above-mentioned duplicate types, as well as any other types located during the extraction process. The second goal of the project is to improve the resolution and image quality of the original digital images by rescanning them with newer technology. At the completion of this project, the data and high-resolution images of more than 50,000 Latin American type specimens will be made available to Aluka for the benefit of the world’s academicians.
In April of this year, Christine Allocca and Michael Butts were hired by the Department to spearhead the effort to improve our Latin American type specimen resources. Allocca comes to us with extensive experience in the information technology field and a more recent history of assisting with various herbarium collections projects. Butts received his undergraduate training at DukeUniversity and recently completed a contract working with Kress and Ida Lopez on the NSP-sponsored project to build an Instant Image Identification System (IIIS) (see the Plant Press, 10(1): 10; 2007). Long-time contract photographer Ingrid Lin, who has been personally responsible for the production of more than 150,000 high-resolution specimen images, has been brought aboard to provide her expertise at digital imaging of types.
In consultation with different Department staff members, various strategies have been developed for locating and extracting types from the general collection. We want to locate as many types as possible, and those that are clearly marked are easier to find than those for which the type status is more cryptic or even missing. Employing a combination of techniques, including the compilation of type specimen records from numerous other herbaria, the initial results in the Annonaceae, Anacardiaceae and Vitaceae have been promising. It is expected that 14,000 to 16,000 additional types will be located.
In his role as the Type Register specialist, John Boggan will be dedicated almost exclusively to the task of verifying the type status of these specimens, and will be the primary point of contact for LAPI staff relative to typification and data entry issues. As specimens are added to the database, they will immediately contribute to the Department’s online Type Register <http://ravenel.si.edu/botany/types>. Once a type has been entered, the decision will be made whether to scan or photograph the specimen. This decision is based primarily on the thickness or bulkiness of the specimen.
During the first three years of the type imaging project, begun in 2000, we used a PhaseOne LightPhase digital camera back that produced a maximum file size of 18MB. Our current digital setup, using a PhaseOne H20, produces a maximum unprocessed (RAW) file size of 48MB. For the LAPI project, we have installed two specially designed flatbed scanners in the former dark room. These scanners, designed at the Royal Botanic Gardens, Kew, are oriented upside down so that individual specimens can be digitized from above to avoid flipping them over. This setup produces an average processed TIFF file of 200MB. All the Latin American type specimens that had been digitized with the LightPhase will be re-scanned or re-photographed, therefore, at a significantly higher resolution.
The organization behind this effort is Aluka, an international, collaborative initiative building an online digital library of scholarly resources. Their name, ‘Aluka,’ is derived from a Zulu word meaning ‘to weave,’ reflecting Aluka’s mission to connect resources and scholars from around the world.
Aluka seeks to attract high-quality scholarly content from institutions and individuals across the globe. Their website includes a wide variety of scholarly materials contributed by Aluka’s partners, ranging from archival documents, periodicals, books, reports, manuscripts, and reference works, to three-dimensional models, maps, oral histories, plant specimens, photographs, and slides. By aggregating these materials online, they link materials that are widely dispersed and difficult to access, opening up new opportunities for research, teaching, and broader public discussion. An effort to assemble data and images for all African plant types, the African Plant Initiative (API) is near completion and can be seen on the Aluka website at <http://www.aluka.org/page/content/plants.jsp>.
When the Type Register project was first developed in 1970, it was hard to imagine what the informatics world would look like in the 21st century. Working with keypunch cards and mainframe computers, the work was tedious. The task of cataloging the specimen, collection, and typification data for more than 80,000 covered 13 years. When it was completed in 1983, the U.S. National Herbarium’s Type Register was not simply the largest electronic dataset of plant specimens, it was the largest electronic specimen dataset anywhere. And, most importantly for researchers, each record had been verified against the published botanical literature in order to validate its type status.
Fast forward another 20 years and the electronic landscape has changed dramatically. The U.S. National Herbarium has become a major online presence with a content-rich website that features outstanding sources of information for botanists and the general public. Today, scientists and museum professionals do not simply take advantage of the data and images on our website; they count on it to provide them with the tools and resources necessary to conduct their research. Upon completion of this project, scholars working in Latin American botany will recognize our Type Register as the most significant online resource available.
Comments