Spatio-Temporal Heterogeneity: Concepts and Analyses (Ecology, Biodiversity and Conservation)
Cambridge University Press
Edition: Pap/Cdr edition, 5/26/2011
EAN 9781107400351, ISBN10: 110740035X
Paperback, 416 pages, 22.8 X 15.4 X 2.2 cm
Language: English
Attempts by ecologists to establish models for predicting the growth of a population or the fluctuations of a natural resource can be confounded by environmental heterogeneity. Spatio-Temporal Heterogeneity explores a range of available statistical methods to help ecologists in the attempt to unravel complexities, demonstrating how to place these changes into an understandable statistical framework. It addresses several key questions, including how to interpret the parameters of statistical models in relation to the biological and environmental realities, how to design a study to collect the best sample data and how to avoid pitfalls in modelling, design, statistical assessment and interpretation. Dutilleul uses a variety of examples to facilitate understanding, from plant ecology, earth and atmospheric sciences, animal biology, forestry and limnology. The accompanying CD-ROM contains MATLAB and SAS codes to aid analyses.
Foreword
Preface
1. Conceptual introduction
2. Spatio-temporal stochastic processes
definitions and properties
3. Heterogeneity analysis of spatial point patterns
4. Heterogeneity analysis of temporal and spatio-temporal point patterns
5. Heterogeneity analysis of point patterns
process modeling and summary
6. Heterogeneity analysis of time series
7. Heterogeneity analysis of spatial surface patterns
8. Heterogeneity analysis of spatio-temporal surface patterns
9. Sampling and study design aspects in heterogeneity analysis of surface patterns
10. Conclusions
Index.
'Pierre Dutilleul has written a book that has a friendly style, includes pertinent examples as well as explanations and motivations … I know of no other book quite like it. I recommend the book highly to students, teachers and researchers in fields including biology, ecology and environmental science as well as in mainstream and applied statistics.' David Brillinger, University of California, Berkeley