Skip to main content

Table of Contents

  1. Introduction: AIACC: Climate Change and Conservation Planning
    1. Chapter1: Evidence for climate change
      1. Chapter 2: Global circulation models
        1. Chapter 4: Biodiversity responses to past changes in climate
          1. Chapter 5: Adaptation of biodiversity to climate change
            1. Chapter 6: Approaches to niche-based modelling
              1. Slide 1: Approaches to niche-based modelling - theory and practice
              2. Slide 2: Lecture Structure
              3. Slide 3: Why model species ranges?
              4. Slide 4: Used in response to
              5. Slide 5: Distribution models have been used to predict
              6. Slide 6: They have also been used to...
              7. Slide 7: Principles: Fundamental niche
              8. Slide 8: Principles: Realised niche
              9. Slide 9: Principles: Range edges
              10. Slide 10: Principles: Response curves
              11. Slide 11: Response curves estimation of different models
              12. Slide 12: Specifics: Niche-based modelling
              13. Slide 13: Niche-based modelling - assumptions
              14. Slide 14: Cautionary note on modelling in general
              15. Slide 15: Specifics: variable selection
              16. Slide 16: Example of how direct/indirect variables may affect a plant species
              17. Slide 17: Variables and their selection
              18. Slide 18: Variables determine specificity of model
              19. Slide 19: Environmental Variables
              20. Slide 20: Derived Variables
              21. Slide 21: Recommendations for variable selection
              22. Slide 22: Species distribution datasets
              23. Slide 23: Species distribution datasets...2
              24. Slide 24: Species distribution datasets...3
              25. Slide 25: How do we choose a model type?
              26. Slide 26: Different types of models
              27. Slide 27: Principles
              28. Slide 28: Various decision trees from the literature
              29. Slide 29: Decision trees from the literature (2)
              30. Slide 30: In conclusion
              31. Slide 31: Model calibration and evaluation
              32. Slide 32: Models and their selection - BioClimatic Envelope
              33. Slide 33: Models and their selection - GAM modeling
              34. Slide 34: Models and their selection - GARP
              35. Slide 35: How good are the predictions?
              36. Slide 36: Kappa statistic
              37. Slide 37: Receiver operating characteristic analysis (ROC)
              38. Slide 38: How good are the predictions?
              39. Slide 39: Test yourself
              40. Slide 40 Links to other chapters
            2. Chapter 7: Ecosystem function modelling
              1. Chapter 8: Climate change implications for conservation planning
                1. Chapter 9: The economic costs of conservation response options for climate change
                  1. Course Resources
                    1. Practical: Conservation for Climate Change
                      1. Tests to Assess your Understanding
                        1. How to run a GAM model in R

                          Slide 30: In conclusion

                          Duration: 00:01:01


                          In general, it has been shown that neural networks (which have the ability to "learn" the important variables in a system) and GAMs (possibly with an autocorrelation coefficient) are the most robust methods using minimal data. The problem with neural networks is that they are "black boxes" - that is, the decision processes within the networks are generally unavailable for examination, making it a challenge to derive which are the governing determinant factors.

                          Thus, there are two real options for deciding on the best model - firstly, to use an expert system such as BIOMOD, to try several different models and derive the one that best decribes the observed data, or to use a model that is generally robust. Secondly, a model can be selected that is specifically suited to the questions being asked, such as Ecological Niche Factor Analysis when there is no absence data to use.

                          However, generalized additive models with pseudo-absence are generally robust, and may often out-perform presence-only techniques (Brotons et al 2004).