Robert P. Anderson, Department of Biology, City College of New York

Course philosophy: This graduate seminar will cover the conceptual basis and practical application of GIS-based techniques for modeling species distributions based on occurrence records and digital environmental data. Specifically, we will focus on the problem of modeling species geographic distributions using presence-only data (where occurrence records are available, but no data exist regarding absences). Despite the title of the course, these techniques are general with regard to taxon (i.e., applicable both plants and animals). Furthermore, they can be applied at any spatial scale, using a grain and extent relevant to the problem at hand. Enrollment is by permission of instructor/department, and the course is designed for students with a specific research interest in using presence-only distributional modeling.

The format will be a lecture/discussion followed by exercises on individual computers. Readings from the primary literature will be assigned, and students will conduct a research project using their own data. A few leaders in the field who work in the New York metropolitan region will give guest presentations.

Grading: Students will be evaluated based on class participation (including short presentations) and the research project. Plagiarism will be dealt with subject to CCNY/CUNY policies.

Absences: Attendance is required. Excused absences must be documented (e.g., death in the family, extreme sickness, etc.). Absence from more than 2 class periods can result in your being dropped from the course for excessive absences (WU).

Tentative schedule (may be modified if necessary):

Date	Topic
Mon. 29 Jan.	The relationship between niche and distribution
Mon. 5 Feb.	Presence-only occurrence data
Thurs. 15 Feb.	Environmental data for distributional modeling
Wed. 21 Feb.	Introduction to presence-only modeling techniques
Mon. 26 Feb.	DIVA/BIOCLIM modeling
Mon. 5 Mar.	Introduction to Maxent modeling
Mon. 12 Mar.	Model evaluation 1: threshold-dependent evaluations (e.g., binomial)
Mon. 19 Mar.	Model evaluation 2: threshold-independent evaluations (e.g., AUC/ROC)
Mon. 26 Mar.	Advanced Maxent modeling (selecting feature classes and tuning regularization)
Mon. 16 Apr.	Postprocessing 1: considering sampling adequacy
Mon. 23 Apr.	Postprocessing 2: considering biotic and historical restrictions
Mon. 30 Apr.	Applications (e.g., climate change, conservation/reserve design, invasive species, niche evolution)
Mon. 7 May	Presentations 1
Mon. 14 May	Presentations 2

Robert P. Anderson, Ph.D.