Abstract
Species distribution models have become a commonplace exercise over the last 10 years, however, analyses vary due to different traditions, aims of applications and statistical backgrounds. In this chapter, I lay out what I consider to be the most crucial steps in a species distribution analysis: data pre-processing and visualisation, dimensional reduction (including collinearity), model formulation, model simplification, model type, assessment of model performance (incl. spatial autocorrelation) and model interpretation. For each step, the most relevant considerations are discussed, mainly illustrated with Generalised Linear Models and Boosted Regression Trees as the two most contrasting methods. In the second section, I draw attention to the three most challenging problems in species distribution modelling: identifying (and incorporating into the model) the factors that limit a species range; separating the fundamental, realised and potential niche; and niche evolution.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
Notes
- 1.
boxcox in MASS (typewriter and bold are used to refer to a function and its R-package)
- 2.
Phillips et al. (2006b): http://www.cs.princeton.edu/~schapire/maxent/
- 3.
transcan and aregImpute in Hmisc
- 4.
scale
- 5.
prcomp
- 6.
isoMDS in MASS or, more conveniently, metaMDS in vegan
- 7.
pairs
- 8.
This can be done either manually (X1.2 <- X1^2) or as part of the model formula (y ~ X1 + I(X1^2)); higher-order polynomials should be specified using poly (y ~ poly(X1, degree=3)), which calculates orthogonal polynomials.
- 9.
As proposed for the function gam in package gam: see ?gam::step.gam.
- 10.
As proposed for the function gam in package mgcv: see ?mgcv::step.gam.
- 11.
http://www.machinelearning.org/ is a good place to start exploring this field.
- 12.
Most of these “complications” can be handled by standard extensions of GLMs (see, e.g. Bolker 2008, and various dedicated R-packages). They will, however, make the model less stable, require larger run-times and still rely on getting the distribution right. There is, of course, the alternative of Bayesian implementations. Since these are also fundamentally maximum likelihood approaches, they are similar to sophisticated GLMs. In any case, there is no Bayesian Boosted Regression Tree (not to speak of a combination with spatial terms and mixed effects). It runs against the Bayesian philosophy to use boosting or bagging, and there is no efficient implementation either.
- 13.
Diagnostics for GLMs fitted in R are given by plotting the model object.
- 14.
Actually, the term “equilibrium” is a bit misleading. What is meant is that the entire width of its niche is filled. Within this niche, there may well be unoccupied sites, e.g. due to metapopulation dynamics. A problem arises, when a species does not occupy say the dry end of its soil moisture niche for historic reasons. Then the estimate of this end of the niche will be biased.
Acknowledgments
Over the years, many colleagues helped develop the above recipe. I am particularly grateful to Boris Schröder, Björn Reineking and Jane Elith, as well as the many participants of statistical workshops on this topic. I am also grateful to Fred Jopp, Hauke Reuter and Dietmar Kraft for improving a previous version. Funding by the Helmholtz Association is acknowledged (VH-NG-247).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2011 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Dormann, C.F. (2011). Modelling Species’ Distributions. In: Jopp, F., Reuter, H., Breckling, B. (eds) Modelling Complex Ecological Dynamics. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-05029-9_13
Download citation
DOI: https://doi.org/10.1007/978-3-642-05029-9_13
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-05028-2
Online ISBN: 978-3-642-05029-9
eBook Packages: Biomedical and Life SciencesBiomedical and Life Sciences (R0)