Some time ago I created a very simple reference database, my Butterfly references database. My idea is to have it for toying around with web services as a way to harvest data from the internet relevant to Nymphalidae butterflies.
As Rod Page has been blogging about, it would be really useful to have a way of linking species names (butterflies in our case) to their original descriptions.
This would speed up taxonomic work since lack of access to primary literature is one of the issues that is crippling taxonomic practice around the globe. This is particularly true in countries where access to primary literature is unthinkable... and coincidentally those countries are the ones that host most of the world's biodiversity! Because I am from Peru, I have been there... 've done that.
So, my new toy, the Butterfly references database only has a few bibliographic references for testing purposes. However it has a web service already. It is able to provide data of bibliographic references in XML format. Thus, now our voucher database will ask the reference database whether it holds references containing a particular species' voucher that any human user might be looking at at the voucher's page. Currently this is done using the RESTful protocol.
For example if you are looking for the butterfly Morpho aurora, you might stumble upon one of our pages for specimens of that species:
If you look a the center bottom you will see a "Relevant literature" field containing a full reference of paper by Patrick Blandin (2006). No bibliographic references data is contained in the voucher database, all that info is being queried and processed from the reference database "on the fly" and "on demand" according to user's input.
Unfortunately most of the hardcore taxonomic literature is old and published in obscure journals. So most likely there will be no "Digital object identifier" (DOI) or PDF files on the web for most of the references. I am currently trying to gather more info for each reference and put it available on the reference database which already points to some DOIs and web addresses of sites hosting PDF files.