Database-backed decision trees with application to biological informatics

We describe a mechanism for the identification of biological organisms through the use of enhanced taxonomic keys-decision trees with nodes augmented by property lists that can serve as arguments to web or local services that access databases or other resources about species, specimens, and ecosystems. Authors of these identification schemes can use simple spreadsheet tools to structure the identification abstractions, and middleware renders the resulting trees into many different forms, with the databases possibly discovered and queried at the time an identification is proposed.