Data Analysis in Vegetation Ecology Book

Preface

List of Figures

List of Tables

1 Introduction 1

2 Patterns in Vegetation Ecology 5

2.1 Pattern recognition 5

2.2 Interpretation of patterns 9

2.3 Sampling for pattern recognition 11

2.3.1 Getting a sample 11

2.3.2 Organizing the data 14

3 Transformation 17

3.1 Data types 17

3.2 Scalar transformation and the species engima 19

3.3 Vector transformation 21

3.4 Example: Transformation of plant cover data 23

4 Multivariate Comparison 25

4.1 Resemblance in multivariate space 25

4.2 Geometric approach 27

4.3 Contingency testing 29

4.4 Product moments 30

4.5 The resemblance matrix 32

4.6 Assessing the quality of classifications 33

5 Ordination 35

5.1 Why ordination? 35

5.2 Principal component analysis (PCA) 37

5.3 Principal coordinates analysis (PCOA) 41

5.4 Correspondence analysis (CA) 43

5.5 The horseshoe or arch effect 47

5.5.1 Origin and remedies 47

5.5.2 Comparing DCA, FSPA and NMDS 49

5.6 Ranking by orthogonal components 51

5.6.1 Method 51

5.6.2 A numerical example 53

5.6.3 A sampling design based on RANK (example) 55

6 Classification 59

6.1 Group structures 59

6.2 Linkage clustering 62

6.3 Minimum-variance clustering 64

6.4 Average-linkage clustering: UPGMA, WPGMA, UPGMC and WPGMC 66

6.5 Forming groups 67

6.6 Structured synoptic tables 69

6.6.1 The aim of ordering tables 69

6.6.2 Steps involved 70

6.6.3 Example: Ordering Ellenberg's data 72

7 Joining Ecological Patterns 75

7.1 Pattern and ecological response 75

7.2 Analysis of variance 77

7.2.1 Variance testing 77

7.2.2 Variance ranking 79

7.2.3 How to weight cover abundance (example) 80

7.3 Correlating resemblance matrices 84

7.3.1 The Mantel test 84

7.3.2 Correlograms: Moran's I 86

7.3.3 Spatial dependence: Schlaenggli data revisited 89

7.4 Contingency tables 92

7.5 Constrained ordination 96

8 Static Explanatory Modelling 101

8.1 Predictive or explanatory? 101

8.2 The Bayes probability model 102

8.2.1 The discrete model 104

8.2.2 The continuous model 105

8.3 Predicting wetland vegetation (example) 106

9 Assessing Vegetation Change in Time 111

9.1 Coping with time 111

9.2 Rate of change and trend 112

9.3 Markov models 115

9.4 Space-for-time substitution 122

9.4.1 Principle and method 122

9.4.2 The Swiss National Park succession (example) 125

9.5 Dynamics in pollen diagrams (example) 127

10 Dynamic Modelling 133

10.1 Simulating time processes 135

10.2 Including space processes 141

10.3 Processes in the Swiss National Park (SNP) 142

10.3.1 The temporal model 142

10.3.2 The spatial model 145

10.3.3 Simulation results 146

11 Large Data Sets: Wetland Patterns 151

11.1 Large data sets differ 151

11.2 Phytosociology revisited 153

11.3 Suppressing outliers 156

11.4 Replacing species with new attributes 158

11.5 Large synoptic tables? 162

12 Swiss Forests: A Case Study 169

12.1 Aim of the study 169

12.2 Structure of the data set 170

12.3 Methods 172

12.4 Selected questions 175

12.4.1 Is the similarity pattern discrete or continuous? 175

12.4.2 Is there a scale effect from plot size? 176

12.4.3 Does the vegetation pattern reflect the environmental conditions? 177

12.4.4 Is tree species distribution man-made? 178

12.4.5 Is the tree species pattern expected to change? 184

12.5 Conclusions 184

Appendix A On Using Software 189

A.1 Spreadsheets 189

A.2 Databases 190

A.3 Software for multivariate analysis 191

Appendix B Data Sets Used 193

References 195

Index 205