Lists of species are the basis of biodiversity science.  Typically after an outing, naturalists and scientists alike make a checklist of the species they saw. Checklists can be used for basic science such as calculating diversity indices and making range maps or for more applied purposes such as planning reserves for rare and endangered species and managing invasive species. While preparing for field excursions, it is common practice among scientists and others to make another kind of species list, a Predicted Local Species Lists (PLSL), comprised of what might be found during a survey. By and large, only experts with local knowledge of specific taxa have the ability to make such predictions.

To examine the feasibility of formalizing the production of PLSL, we developed a prototype for mammals. NatureServe.org has publically accessible ArcGIS shape file data of range maps for every vertebrate species (5,620 species) in North America. Using ArcGIS 10 we found these files to be accurate to within 1 km.  To test the idea we selected northern California as a geographic location. A python script was written using the ArcGIS arcpy library that took the mammal shape files and populated an SQLite database of (x, y) grid point locations at 0.1 degrees resolution (about 10.7 km) with species names when the grid point fell within the species’ range.  A website (LocalList.pythonanywhere.com) allowed a user to select any point in Northern California after which the software would returned a page with a list of the mammalian species predicted to be found within 10 km and hyperlinked to the species description on the NatureServe Explorer page.

The challenge for biodiversity scientists and informaticists is to formalize the process of computing PLSL and to provide these predictions for others at open-source web sites.  Besides using published range maps, predictions can be made from existing checklists or from species distribution models which now come in many favors.  Some sites already perform this function: eBird provide list of birds by date at local sites and Map of Life provides lists of birds within a 50 km radius of the point the user picks. LSPL data will have many uses in pure and applied biodiversity science. Testing the predictions will start a feedback process the refines models to include factors such as life stage, sex, time of day, and method of observation.