Mid-term Exam Answers!

Introduction
 

Rationale:

Success in the practice of ecology and resource management depends on clear thinking.  In any endeavor, clarity of thought can be measurably enhanced by developing formal abstractions of problems.  Such abstractions are known as models.  The overarching goal of NR575 is to teach students how to think clearly about ideas in ecology using models as thinking tools. A secondary goal is to teach students how to evaluate models with data.

Target Audience: Seniors and graduate students.

Course Objectives:

1)     Teach techniques for formulating and analyzing dynamic models of ecological systems.

2)     Teach modern methods for estimating model parameters and for evaluating alternative models based on data.

3)     Give students the quantitative confidence needed to use models in their research and to support future self-teaching.

Content and teaching approach:  

The course is faithful to its title - we will study methods for building and evaluating models. There will be a firm emphasis on learning by doing.  Students will be taught to use R and WinBUGS as a "modeling workbench." The open-source computing language R is becoming the standard for scientific computation worldwide. Without doubt, it is the preeminent system for bringing models together with data. R is somewhat challenging to learn and a fundamental goal of this course is to make that challenge manageable even for students with minimal programming background. In addition, we will learn to use WinBUGS, a stunningly powerful program for parameter estimation using contemporary Bayesian methods.

Weekly laboratory sessions will challenge students to build and document simple models.  It is imperative that students keep up with laboratory work.  This is the foundation of the course. 

General Information

Prerequisites:
Ideally, students should have a semester of calculus, a basic ecology course, and an introduction to statistics (ST 301 or ST 307/EH 307). None of these are absolute requirements, but if you do not have at least two of these background courses, you should be prepared to do some remedial work on your own.

Class Info:

Lectures:  Tuesday and Thursday  9:30-10:50

Labs:  Wednesday  1:00-3:00   

Text:

There will be no single text. Instead, lectures will be developed directly from the journal literature and from several influential texts including:

Bolker, B. In press. Ecological models and data in R. Princeton University Press, Princeton, N.J.

Burnham, K. P., and D. R. Anderson. 2002. Model Selection and Multi-Model Inference: A Practical Information-Theoretic Approach. Springer-Verlag. New York.

Clark, J. M. 2007. Models for ecological data. Princeton University Press, Princeton, N.J.

Haefner, J. W. 1996. Modeling Biological Systems: Principles and Applications. London., UK, Chapman and Hall. 473 pp.

Royall, R. 1997. Statistical Evidence: A Likelihood Paradigm. Chapman and Hall/CRC.

Woodworth, G. G. 2004. Biostatistics: A Bayesian introduction. Wiley, Hoboken, N.J.

Purchase of a manual introducing programming in R is required for the laboratory. These are available for the cost of copying and binding ($11 ish) at the bookstore.

Working in groups:

Grading:

Your grade will be based on 1 exam (100 points) and 7 lab write-ups, each worth 50 points. If you complete the work with attention and care, you will get an A in this course. I am far more interested in your mastery of the material than I am in making academic comparisons among you. The material in this course can appear intimidating at first, but the last thing I want is for you to be anxious about it. Everyone who has taken this course has emerged with a sturdy understanding of the key concepts and methods. It may seem daunting at first. Relax. We will get through it.

Lab work will require programming in R and WinBUGS. For each assignment, each lab group will turn in a hard copy write-up that includes text and figures communicating your results. You should also turn in a hard copy of documented code. Grading will be based on the following:

1) Quality of approach to problem: Did you use a logical, thoughtful process for solving the problem?
2) Quality of presentation: Did you present your findings in a literate document? Did you clearly communicate how you solved the problem, showing mathematical steps or a computer algorithm? Was your document attractive and well organized?
3) Quality of technique: Did you demonstrate mastery of the appropriate mathematical or computer technique?

Your lab reports should describe model results and discuss them as appropriate. Include figures and/or tables, embedded in the text, not appended at the end. All figures must have captions. All tables must have headings. Provide an appendix documenting your code.

First Assignments:

Get NR575 course manual at the bookstore.

Prepare to introduce yourself to your colleagues during the first laboratory session. Include a 1 page CV to hand out to the class.

Schedule

Note: Links for lecture notes and papers require Adobe Acrobat Reader freeware, http://www.adobe.com/products/acrobat/readstep2.html

Left click opens PDF link in frame, right click to open in new window or save.

Lab Exercises:

Lab 1

RMNP elk data update
RMNP elk time series
Lynx data
Winter severity data

Lab 2

      odesolve in R
      Lotka-Voltera iii.R
      Dobson et al.pdf
      Lab 2 supplement - example model problems
      SIR modifications

Lab 3

Lab 4
Lab 5

      Lab 5.doc
      Data for hemlock growth increments

Lab 6

      Lab 6.doc
      A primer on maximum likelihood in R using bbmle
      Serengeti data
      bbmle example 1.R
      bbmle example 2.R
      LIDET lab data.csv
      LIDET lab data.xls
      
Lab 7
Lab 8w

 

Readings
 

Week 1:  i. Hobbs, N.T. A Modeler's Primer on R
ii. Hastings, A., P. Arzberger, B. Bolker, S. Collins, A. R. Ives, N. A. Johnson, and M. A. Palmer. 2005. Quantitative bioscience for the 21st century. Bioscience 55:511-517.

Week 2:  Hobbs, N. T. The Mathematical Basis for Dynamic Models

Week 3:  i. Hobbs, N. T. A Modeler’s Primer on R
                ii. Haefner, J. W. 1996. Chapter 3. Modeling Biological Systems : Principles and Applications. Chapman and Hall. London., U.K. 473 pp.
                iii. (Browse) Parton and Innis, Some Graphs and Their Functional Forms.

Week 4:  Jackson, L. J., A. S. Trebitz, and K. L. Cottingham. 2000. An introduction to the practice of ecological modeling. Bioscience 50:694-706.

Week 5: TBA

Week 6:  Hilborn and Mangel. 1997. The ecological detective. Chapters 2 and 3. Pages 12-93.

Week 7: Johnson, J. B. and K. S. Omland. 2004. Model selection in ecology and evolution. Trends in Ecology and Evolution 19:101-108.

Week 8:  Hobbs, N. T. and Hilborn, R. 2006. Alternatives to statistical hypothesis testing in ecology. Ecological Applications 16:5-19.

Week 10: i. Hobbs, N. T., J. E. Gross, L. A. Shipley, D. E. Spalinger, and B. A. Wunder. 2003. Herbivore functional response in heterogeneous environments: a contest among models. Ecology 84:666-681.
ii. Miller, M.W., NT Hobbs, and S.J. Tavener. 2006. Dynamics of prion disease transmission in mule deer. Ecological Applications 16:2208-2214.

Week 11: Cooper, A.B., R. Hilborn, and J.W. Unsworth. 2003. An approach for population assessment in the absence of abundance indices. Ecological Applications 13:814-828.

Week 12: i. Hobbs, N. T. 2006. A WinBUGS primer.
ii. Lavine, M. What is Bayesian statistics and why everything else is wrong.

Week 13: i. Hobbs, N. T. 2006. A WinBUGS primer.
ii. Clark, J. S. 2005. Why environmental scientists are becoming Bayesians. Ecology Letters 8:2-14.

Week 14: i. Hobbs, N. T. 2006. A WinBUGS Primer.
ii. Ogle, K. and J.J. Barber. 2008. Bayesian data-model integration in plant physiological and ecosystem ecology. Progress in Botany 69: In press.

Additional Information

Computer resources needed:  A large amount of computer programming will be necessary to successfully complete the course, so students will need easy access to workstations running R and WinBUGS, both of which are free, open-source software. We will learn how to load these in the lab. You must have an equation editor installed in your word processing package, and you should know how to use it. (It is often not installed because it is not checked in a basic installation of Word. If you can't find it, do a custom install and check the Equation Editor box). Add-ins for editing equations are very useful (I use Math Type, http://www.mathtype.com/). The TA's and I will be very grumpy about equations that are not properly typeset with an equation editor. 

Accommodation of individual learning needs: If you have learning needs that may affect your performance (sight, hearing, language, or any other reason), please let me know at the beginning of the course.  We will work out ways to accommodate your needs.

Interaction outside of class: My office hours will be 3:00-5:00 on Wednesdays. If you can't make that hour, or if you have a pressing problem, be in touch by email. The TA's and I also will be available via email to answer questions on your R programming. When you have R questions, be sure to include the script causing you problems in your email. Your learning will be enhanced if you struggle with a problem before you ask a question, but we don't want you to struggle excessively. If you truly aren't getting anywhere on your own, don't spin your wheels. Contact us with a well-framed question.

My contact information is above; the TA's emails are:

Dylan George (dgeorge@lamar.colostate.edu)
Liz Harp (eharp@lamar.colostate.edu)
Paul Duffy (paul.duffy@neptuneinc.org)

Readings:  Readings are offered to supplement the lectures.  Weekly assignments will be available in PDF format on the class web site http://www.warnercnr.colostate.edu/class_info/nr575/ . We will not discuss the readings in class.  They are intended to offer a different slant on the material I present, thereby complimenting the lectures rather than duplicating them.  However, some of the particularly technical material will be covered in lecture and directly reinforced in the readings.  With the notable exception of the material that I provide as primers and handouts, it is entirely possible to do well in this course (in terms of your grade) without doing the reading.  So, if all you want is a reasonable grade, then neglecting the reading probably won’t hurt you.  Alternatively, if you want to master the material, the readings will contribute meaningfully to meeting your goal.
 

Class notes:   Notes for each lecture will be available as PDF files on the class web site.  You should not print these out until 8:00 am the morning of the lecture.

My travel: Although I avoid travel during the spring semester, there will be occasions when I must be away. I will provide video lectures or problem sets on the rare occasions when I must miss class. Student teaching assistants will conduct laboratories when I am traveling.

Related Links

http://www.zoo.ufl.edu/bolker/emd/

A course like this one with terrific notes on several topics of interest to students in NR575.

http://www.fish.washington.edu/classes/fish458/index.htm

A course like this one with terrific notes on several topics of interest to students in NR575.