A project is required of all students. The project must have all of the following elements:
1. Clear definition of a management objective. The objective must be stated quantitatively. Note that by "management" I mean "any conservation/ natural resource decision-making problem." The key elements that make it "management" are 1) quantifiable objective(s) (e.g., maximize long term harvest; minimize extinction risk), and 2) a range of candidate decisions (e.g., harvest regulations, land management policy).
2. A dynamic model describing the system under management. The model must include system states that relate to the management object; decision variables (inputs or controls) under the control of a manager; and consideration of important environmental (uncontrollable) influences.
3. Statistical analyses to estimate important system states (e.g., population size), demographic parameters (e.g., birth rates, survival rates), and system relationships (e.g., the influence of possible influence of decision variables (e.g., harvest rates) on system dynamics. These analyses must involve at least one (1) of the major statistical methods we have employed in the course.
4. An assessment of sources of uncertainty in the dynamic model, including: (1) uncertainty over system relationships, (2) statistical error in parameter estimates, and (3) intrinsic system variability (e.g., due to environmental influences).
5. Assemble all four of the preceding elements into a formal, working "decision" model that allows decision making under uncertainty, using either program ASDP, a spreadsheet analysis, or some other approach.
Optional:
6. Provide a mechanism for incorporating adaptive feedback (learning) from management.
The material for the project may come from:
1. Original data (e.g., your thesis projects)
2. The published literature.
3. Real or contrived examples supplied by the instructor.
In many cases real, raw data will not be available. In these cases, with the instructor's assistance, you will (1) specify a parameter to be estimated, (2) select appropriate data structures, (3) specify parameter values or ranges of values and ranges in statistical error, and (4) simulate the data for use in the analyses.
A schedule of tasks to be completed and milestone dates for each will be handed out in class. Each student will make an oral presentation of the model results, and submit a written final report, near the end of the semester.
Last updated 30 Mar 2009