Participants: Lesley Deem, Ken Robertson, Steve Seiberling, Alan Weakley
Ken agreed with the general approach to selecting and defining characters
and character states contained in the document Steve prepared, Draft Criteria
for Selection of Characters, Character States, and Character State Definitions
for Use in the OpenKey Database and Website. Along those lines we
agreed it’s desireable to:
· avoid overly detailed terms and definitions that are not useful for practical identification;
· not use characters that require more than 10x magnification to be seen;
· make the definitions understandable to non-botanist users.
O.K. to use the term “pistil” instead of “carpal”, even though carpal is now more widely used by taxonomists, since pistil is less technical and perhaps more familiar to the lay public.
Would be good to add a pubescence term that means “cobwebby” (appropriate on Halloween!) since certain prairie plants possess that character state. [The term Steve has found for that is “arachnoid”.]
Good definitions for types of bark may be difficult to come up with. Ken suggests looking in horticulture books, especially Michael Dirr, Manual of Woody Landscape Plants.
Illinois people haven’t looked systematically at what terms may be needed to distinguish the ten or so grasses on their list. May need to add some characters for them later. Likewise for plants in the Asteraceae due to their composite floral structure.
Assuming we will not be using modifying terms (e.g. broadly, narrowly, sparsely, densely, etc.) in the database, will be important to show the range of variation applicable to given character states in the illustrations or photos. We may wish to clarify with the database people whether its necessary to omit modifying terms in the database. Steve has been choosing character states and definitions with the idea that we will not be using modifying terms.
An important question to be addressed is how the database search functions will “translate” technical terms for the website/polyclave users. Bryan’s thesaurus idea, permitting the users to choose general, more familiar terms that are associated with the technical terms in the database and matched when searching in some way, seems useful. Knowing how this will work has implications for what data is recorded in the database and would be good to clarify as soon as we can.
Recording color character states for petals, sepals, etc. requires some attention. This is complicated by the fact the monitors of PCs are not calibrated to a set standard, and therefore users may see quite different colors from each other in images on the web. Will be important for us to keep the color choices simple, limited to about 15-20 colors. Ken will look into creating a list of colors with associated examples that users will be able to see on the website.
Lesley suggested the possibility of having a single general illustration for a character state, accompanied by several photographic images. Steve had been thinking of trying to find several illustrations that show the range of variation we have in mind for a given character state, along with additional photographic images.
Ken has been producing digital images for prairie plant characters. His test images can be found at: http://www.inhs.uiuc.edu/~kenr/imls.test/imlstest.html