Consortium News

The Consortium of Pacific Northwest Herbaria has a new web site. Included is an update to the Consortium’s specimen database, with approximately 1.7 million specimen records and 118,000 images of specimen sheets. The database comprises the single most comprehensive source of herbarium specimen data for the Pacific Northwest.  Nearly all of the major herbaria in the region are represented, as are a growing proportion of smaller herbaria.

Additional means of data access are provided.  These include county-level species checklists for Washington, Oregon, Idaho, and Montana, and downloadable datasets for each herbarium as well as the entire Consortium database.  Specimen images can be viewed at full resolution, providing a virtual resource for those learning about or identifying plants in the PNW.

Even more web site features are under development and will become available over the coming months.  A mobile version of the web site will allow efficient access to specimen data, distribution maps, and images in the field.  Further, downloadable content for mobile devices will allow access to the data even without a network connection.  A checklist of the vascular plant flora of the Pacific Northwest is also in progress.

Questions or comments about the new web site can be sent to Ben Legler.

We are now over one year into our collaborative imaging and databasing grant. To date, we have acquired about 217,000 images from seven herbaria. New images are accruing at a rate of about 15,000 per month.

Imaging has been completed at the following herbaria:

• Western Washington University: completed Nov. 2010 (22,559 images)
• H.J. Andrews Experimental Forest, Oregon: completed Feb. 2011 (1,199 images)
• Reed College, Oregon: completed Apr. 2011 (6,520 images)
• Whitman College, Washington: completed May 2011 (16,765 images)
• Portland State University, Oregon: completed Aug. 2011 (15,141 images)

Imaging is in progress at the following herbaria:

• Montana State University: 52,680 images acquired
• University of Idaho: 102,110 images acquired
• Pacific Lutheran College, Washington: ____ images acquired
• Linfield College, Oregon: ____ images acquired

Images become available through the CPNWH search page after their label data is digitized.  Currently, about 111,000 of the 217,000 images are available online. These can be viewed online as thumbnails or through an image viewer that allows zooming and panning of the full-resolution image. The image viewer includes a measuring ruler that can be positioned at any point on the image to take precise measurements.  Images can also be downloaded as high-resolution JPEGs.

CPNWH is currently funded by a collaborative, $1.3 million grant from the National Science Foundation to image and database specimens from multiple herbaria across the Pacific Northwest. Grant partners are the University of Washington, Oregon State University, the University of Idaho, and Montana State University. Work on the grant began in June, 2010.

By the end of the three-year grant we will have imaged nearly 300,000 specimens from at least 12 herbaria, including many smaller collections, and databased these plus another 200,000 bryophyte, lichen, and fungi specimens from WTU and OSC. Imaging stations are deployed to herbaria in each of the four states (WA, OR, ID, MT). The high-throughput imaging workflow we developed uses a high-resolution digital SLR and lightbox to capture 100-150 images per hour. Images are stored centrally on the Consortium’s web server and are batch-processed using custom software scripts.

Another facet of the grant is the development of web-based software and workflows to capture label text data from the specimen images using a combination of keystroking and optical character recognition (OCR). Student help is used extensively for this process, as well as the imaging.

The database software and imaging scripts we are developing will be released as open source for use by similar projects in other regions, or as a source of ideas for those developing similar solutions. Our infrastructure leverages the technical expertise and staff at the larger partner institutions to facilitate data capture from small institutions who may lack the means to develop their own in-house databases.