Population Ecology in Practice : Underused, Misused and Abused Methods.

Contributors xvii Preface xxi About the Companion Website xxiii Part I Tools for Population Biology 1 1 How to Ask Meaningful Ecological Questions 3Charles J. Krebs 1.1 What Problems Do Population Ecologists Try to Solve? 3 1.2 What Approaches Do Population Ecologists Use? 6 1.2.1 Generating and Testing Hypotheses in Population Ecology 10 1.3 Generality in Population Ecology 11 1.4 Final Thoughts 12 References 13 2 From Research Hypothesis to Model Selection: A Strategy for Robust Inference in Population Ecology 17Dennis L. Murray, Guillaume Bastille-Rousseau, Lynne E. Beaty, Megan L. Hornseth, Jeffrey R. Row and Daniel H. Thornton 2.1 Introduction 17 2.1.1 Inductive Methods 18 2.1.2 Hypothetico-deductive Methods 19 2.1.3 Multimodel Inference 20 2.1.4 Bayesian Methods 22 2.2 What Constitutes a Good Research Hypothesis? 22 2.3 Multiple Hypotheses and Information Theoretics 24 2.3.1 How Many are Too Many Hypotheses? 25 2.4 From Research Hypothesis to Statistical Model 26 2.4.1 Functional Relationships Between Variables 26 2.4.2 Interactions Between Predictor Variables 26 2.4.3 Number and Structure of Predictor Variables 27 2.5 Exploratory Analysis and Helpful Remedies 28 2.5.1 Exploratory Analysis and Diagnostic Tests 28 2.5.2 Missing Data 28 2.5.3 Inter-relationships Between Predictors 30 2.5.4 Interpretability of Model Output 31 2.6 Model Ranking and Evaluation 32 2.6.1 Model Selection 32 2.6.2 Multimodel Inference 36 2.7 Model Validation 39 2.8 Software Tools 41 2.9 Online Exercises 41 2.10 Future Directions 41 References 42 Part II Population Demography 47 3 Estimating Abundance or Occupancy from Unmarked Populations 49Brett T. McClintock and Len Thomas 3.1 Introduction 49 3.1.1 Why Collect Data from Unmarked Populations? 49 3.1.2 Relative Indices and Detection Probability 50 3.1.2.1 Population Abundance 50 3.1.2.2 Species Occurrence 51 3.1.3 Hierarchy of Sampling Methods for Unmarked Individuals 52 3.2 Estimating Abundance (or Density) from Unmarked Individuals 53 3.2.1 Distance Sampling 53 3.2.1.1 Classical Distance Sampling 54 3.2.1.2 Model-Based Distance Sampling 57 3.2.2 Replicated Counts of Unmarked Individuals 61 3.2.2.1 Spatially Replicated Counts 61 3.2.2.2 Removal Sampling 63 3.3 Estimating Species Occurrence under Imperfect Detection 64 3.3.1 Single-Season Occupancy Models 65 3.3.2 Multiple-Season Occupancy Models 66 3.3.3 Other Developments in Occupancy Estimation 68 3.3.3.1 Site Heterogeneity in Detection Probability 68 3.3.3.2 Occupancy and Abundance Relationships 68 3.3.3.3 Multistate and Multiscale Occupancy Models 68 3.3.3.4 Metapopulation Occupancy Models 69 3.3.3.5 False Positive Occupancy Models 70 3.4 Software Tools 70 3.5 Online Exercises 71 3.6 Future Directions 71 References 73 4 Analyzing Time Series Data: Single-Species Abundance Modeling 79Steven Delean, Thomas A.A. Prowse, Joshua V. Ross and Jonathan Tuke 4.1 Introduction 79 4.1.1 Principal Approaches to Time Series Analysis in Ecology 80 4.1.2 Challenges to Time Series Analysis in Ecology 82 4.2 Time Series (ARMA) Modeling 83 4.2.1 Time Series Models 83 4.2.2 Autoregressive Moving Average Models 83 4.3 Regression Models with Correlated Errors 87 4.4 Phenomenological Models of Population Dynamics 88 4.4.1 Deterministic Models 89 4.4.1.1 Exponential Growth 89 4.4.1.2 Classic ODE Single-Species Population Models that Incorporate Density Dependence 90 4.4.2 Discrete-Time Population Growth Models with Stochasticity 92 4.5 State-space Modeling 93 4.5.1 Gompertz State-space Population Model 94 4.5.2 Nonlinear and Non-Gaussian State-space Population Models 96 4.6 Software Tools 96 4.7 Online Exercises 97 4.8 Future Directions 97 References 98 5 Estimating Abundance from Capture-Recapture Data 103J. Andrew Royle and Sarah J. Converse 5.1 Introduction 103 5.2 Genesis of Capture-Recapture Data 104 5.3 The Basic Closed Population Models: M0, Mt, Mb104 5.4 Inference Strategies 105 5.4.1 Likelihood Inference 105 5.4.2 Bayesian Analysis 107 5.4.3 Other Inference Strategies 108 5.5 Models with Individual Heterogeneity in Detection 108 5.5.1 Model Mh 108 5.5.2 Individual Covariate Models 109 5.5.2.1 The Full Likelihood 109 5.5.2.2 Horvitz-Thompson Estimation 110 5.5.3 Distance Sampling 110 5.5.4 Spatial Capture-Recapture Models 110 5.5.4.1 The State-space 112 5.5.4.2 Inference in SCR Models 112 5.6 Stratified Populations or Multisession Models 112 5.6.1 Nonparametric Estimation 112 5.6.2 Hierarchical Capture-Recapture Models 113 5.7 Model Selection and Model Fit 113 5.7.1 Model Selection 113 5.7.2 Goodness-of-Fit 114 5.7.3 What to Do When Your Model Does Not Fit 115 5.8 Open Population Models 115 5.9 Software Tools 116 5.10 Online Exercises 117 5.11 Future Directions 118 References 119 6 Estimating Survival and Cause-specific Mortality from Continuous Time Observations 123Dennis L. Murray and Guillaume Bastille-Rousseau 6.1 Introduction 123 6.1.1 Assumption of No Handling, Marking or Monitoring Effects 125 6.1.2 Cause of Death Assessment 125 6.1.3 Historical Origins of Survival Estimation 126 6.2 Survival and Hazard Functions in Theory 127 6.3 Developing Continuous Time Survival Datasets 130 6.3.1 Dataset Structure 131 6.3.2 Right-censoring 133 6.3.3 Delayed Entry and Other Time Considerations 133 6.3.4 Sampling Heterogeneity 134 6.3.5 Time-dependent Covariates 135 6.4 Survival and Hazard Functions in Practice 135 6.4.1 Mayfield and Heisey-Fuller Survival Estimation 135 6.4.2 Kaplan-Meier Estimator 136 6.4.3 Nelson-Aalen Estimator 138 6.5 Statistical Analysis of Survival 138 6.5.1 Simple Hypothesis Tests 138 6.5.2 Cox Proportional Hazards 139 6.5.3 Proportionality of Hazards 140 6.5.4 Extended CPH 142 6.5.5 Further Extensions 143 6.5.6 Parametric Models 143 6.6 Cause-specific Survival Analysis 144 6.6.1 The Case for Cause-specific Mortality Data 144 6.6.2 Cause-specific Hazards and Mortality Rates 145 6.6.3 Competing Risks Analysis 146 6.6.4 Additive Versus Compensatory Mortality 147 6.7 Software Tools 149 6.8 Online Exercises 149 6.9 Future Directions 149 References 151 7 Mark-Recapture Models for Estimation of Demographic Parameters 157Brett K. Sandercock 7.1 Introduction 157 7.2 Live Encounter Data 158 7.3 Encounter Histories and Model Selection 159 7.4 Return Rates 163 7.5 Cormack-Jolly-Seber Models 164 7.6 The Challenge of Emigration 164 7.7 Extending the CJS Model 167 7.8 Time-since-marking and Transient Models 167 7.9 Temporal Symmetry Models 168 7.10 Jolly-Seber Model 169 7.11 Multilevel Models 169 7.12 Spatially Explicit Models 170 7.13 Robust Design Models 170 7.14 Mark-resight Models 171 7.15 Young Survival Model 172 7.16 Multistate Models 173 7.17 Multistate Models with Unobservable States 175 7.18 Multievent Models with Uncertain States 176 7.19 Joint Models 177 7.20 Integrated Population Models 178 7.21 Frequentist vs. Bayesian Methods 178 7.22 Software Tools 179 7.23 Online Exercises 180 7.24 Future Directions 180 References 180 Part III Population Models 191 8 Projecting Populations 193Stephane Legendre 8.1 Introduction 193 8.2 The Life Cycle Graph 194 8.2.1 Description 194 8.2.2 Construction 194 8.3 Matrix Models 198 8.3.1 The Projection Equation 198 8.3.2 Demographic Descriptors 200 8.3.3 Sensitivities 200 8.4 Accounting for the Environment 202 8.5 Density Dependence 203 8.5.1 Density-dependent Scalar Models 203 8.5.2 Density-dependent Matrix Models 203 8.5.3 Parameterizing Density Dependence 204 8.5.4 Density-dependent Sensitivities 204 8.6 Environmental Stochasticity 204 8.6.1 Models of the Environment 204 8.6.2 Stochastic Dynamics 205 8.6.3 Parameterizing Environmental Stochasticity 208 8.7 Spatial Structure 208 8.8 Demographic Stochasticity 209 8.8.1 Branching Processes 209 8.8.2 Two-sex Models 210 8.9 Demographic Heterogeneity 210 8.9.1 Integral Projection Models 211 8.10 Software Tools 212 8.11 Online Exercises 212 8.12 Future Directions 212 References 212 9 Combining Counts of Unmarked Individuals and Demographic Data Using Integrated Population Models 215Michael Schaub 9.1 Introduction 215 9.2 Construction of Integrated Population Models 216 9.2.1 Development of a Population Model 216 9.2.2 Construction of the Likelihood for Different Datasets 218 9.2.3 The Joint Likelihood 220 9.2.4 Fitting an Integrated Population Model 221 9.3 Model Extensions 223 9.3.1 Environmental Stochasticity 223 9.3.2 Direct Density Dependence 224 9.3.3 Open Population Models and Other Extensions 226 9.3.4 Alternative Observation Models 226 9.4 Inference About Population Dynamics 227 9.4.1 Retrospective Population Analyses 227 9.4.2 Population Viability Analyses 227 9.5 Missing Data 229 9.6 Goodness-of-fit and Model Selection 230 9.7 Software Tools 230 9.8 Online Exercises 231 9.9 Future Directions 231 References 232 10 Individual and Agent-based Models in Population Ecology and Conservation Biology 237Eloy Revilla 10.1 Individual and Agent-based Models 237 10.1.1 What an IBM is and What it is Not 238 10.1.2 When to Use an Individual-based Model 238 10.1.3 Criticisms on the Use of IBMs: Advantages or Disadvantages 239 10.2 Building the Core Model 239 10.2.1 Design Phase: The Question and the Conceptual Model 239 10.2.2 Implementation of the Core Model 240 10.2.3 Individuals and Their Traits 240 10.2.4 Functional Relationships 244 10.2.5 The Environment and Its Relevant Properties 244 10.2.6 Time and Space: Domains, Resolutions, Boundary Conditions, and Scheduling 244 10.2.7 Single Model Run, Data Input, Model Output 246 10.3 Protocols for Model Documentation 247 10.3.1 The Overview, Design Concepts, and Details Protocol 249 10.4 Model Analysis and Inference 249 10.4.1 Model Debugging and Checking the Consistency of Model Behavior 249 10.4.2 Model Structural Uncertainty and Sensitivity Analyses 252 10.4.3 Model Selection, Validation, and Calibration 254 10.4.4 Answering your Questions 256 10.5 Software Tools 257 10.6 Online Exercises 257 10.7 Future Directions 257 References 258 Part IV Population Genetics and Spatial Ecology 261 11 Genetic Insights into Population Ecology 263Jeffrey R. Row and Stephen C. Lougheed 11.1 Introduction 263 11.2 Types of Genetic Markers 264 11.2.1 Mitochondrial DNA 264 11.2.2 Nuclear Introns 265 11.2.3 Microsatellites 265 11.2.4 Single Nucleotide Polymorphisms 265 11.2.5 Next-generation Sequencing 265 11.3 Quantifying Population Structure with Individual-based Analyses 266 11.3.1 Bayesian Clustering 267 11.3.2 Multivariate Analysis of Genetic Data Through Ordinations 269 11.3.3 Spatial Autocorrelation Analysis 271 11.3.4 Population-level Considerations 273 11.4 Estimating Population Size and Trends 273 11.4.1 Estimating Census Population Size 277 11.4.2 Estimating Contemporary Effective Population Size with One Sample Methods 277 11.4.3 Estimating Contemporary Effective Population Size with Temporal Sampling 279 11.4.4 Diagnosing Recent Population Bottlenecks 280 11.5 Estimating Dispersal and Gene Flow 281 11.5.1 Estimating Dispersal and Recent Gene Flow 282 11.5.2 Estimating Sustained Levels of Gene Flow 282 11.5.3 Network Analysis of Genetic Connectivity 283 11.6 Software Tools 284 11.6.1 Individual-based Analysis 284 11.6.2 Population-based Population Size 285 11.6.3 Dispersal and Gene Flow 286 11.7 Online Exercises 286 11.8 Future Directions 286 Glossary 287 References 289 12 Spatial Structure in Population Data 299Marie-Josee Fortin 12.1 Introduction 299 12.2 Data Acquisition and Spatial Scales 302 12.3 Point Data Analysis 302 12.4 Abundance Data Analysis 304 12.5 Spatial Interpolation 306 12.6 Spatial Density Mapping 308 12.7 Multiple Scale Analysis 308 12.8 Spatial Regression 311 12.9 Software Tools 312 12.10 Online Exercises 312 12.11 Future Directions 312 Glossary 312 References 313 13 Animal Home Ranges: Concepts, Uses, and Estimation 315Jon S. Horne, John Fieberg, Luca Boerger, Janet L. Rachlow, Justin M. Calabrese and Chris H. Fleming 13.1 What is a Home Range? 315 13.1.1 Quantifying Animal Home Ranges as a Probability Density Function 316 13.1.2 Why Estimate Animal Home Ranges? 318 13.2 Estimating Home Ranges: Preliminary Considerations 319 13.3 Estimating Home Ranges: The Occurrence Distribution 321 13.3.1 Minimum Convex Polygon 321 13.3.2 Kernel Smoothing 322 13.3.3 Models Based on Animal Movements 323 13.3.4 Estimation from a One-dimensional Path 324 13.4 Estimating Home Ranges: The Range Distribution 324 13.4.1 Bivariate Normal Models 324 13.4.2 The Synoptic Model 324 13.4.3 Mechanistic Models 325 13.4.4 Kernel Smoothing 326 13.5 Software Tools 326 13.6 Online Exercises 327 13.7 Future Directions 327 13.7.1 Choosing a Home Range Model 327 13.7.2 The Future of Home Range Modeling 327 References 328 14 Analysis of Resource Selection by Animals 333Joshua J. Millspaugh, Christopher T. Rota, Robert A. Gitzen, Robert A. Montgomery, Thomas W. Bonnot, Jerrold L. Belant, Christopher R. Ayers, Dylan C. Kesler, David A. Eads and Catherine M. Bodinof Jachowski 14.1 Introduction 333 14.2 Definitions 335 14.2.1 Terminology and Currencies of Use and Availability 335 14.2.2 Use-availability, Paired Use-availability, Use and Non-use (Case-control), and Use-only Designs 336 14.2.3 Differences Between RSF, RSPF, and RUF 336 14.3 Considerations in Studies of Resource Selection 338 14.3.1 Two Important Sampling Considerations: Selecting Sample Units and Time of Day 338 14.3.2 Estimating the Number of Animals and Locations Needed 338 14.3.3 Location Error and Fix Rate Bias Resource Selection Studies 339 14.3.4 Consideration of Animal Behavior in Resource Selection Studies 339 14.3.5 Biological Seasons in Resource Selection Studies 340 14.3.6 Scaling in Resource Selection Studies 340 14.3.7 Linking Resource Selection to Fitness 341 14.4 Methods of Analysis and Examples 342 14.4.1 Compositional Analysis 342 14.4.2 Logistic Regression 343 14.4.3 Sampling Designs for Logistic Regression Modeling 344 14.4.3.1 Random Sampling of Units within the Study Area 344 14.4.3.2 Random Sampling of Used and Unused Units 344 14.4.3.3 Random Sample of Used and Available Sampling Units 345 14.4.4 Discrete Choice Models 346 14.4.5 Poisson Regression 347 14.4.6 Resource Utilization Functions 348 14.4.7 Ecological Niche Factor Analysis 348 14.4.8 Mixed Models 349 14.5 Software Tools 349 14.6 Online Exercises 350 14.7 Future Directions 350 References 351 15 Species Distribution Modeling 359Daniel H. Thornton and Michael J.L. Peers 15.1 Introduction 359 15.1.1 Relationship of Distribution to Other Population Parameters 362 15.1.2 Species Distribution Models and the Niche Concept 363 15.2 Building a Species Distribution Model 366 15.2.1 Species Data 366 15.2.2 Environmental Data 368 15.2.3 Model Fitting 368 15.2.4 Interpretation of Model Output 371 15.2.5 Model Accuracy 372 15.3 Common Problems when Fitting Species Distribution Models 374 15.3.1 Overfitting 374 15.3.2 Sample Selection Bias 375 15.3.3 Background Selection 376 15.3.4 Extrapolation 377 15.3.5 Violation of Assumptions 378 15.4 Recent Advances 378 15.4.1 Incorporating Dispersal 378 15.4.2 Incorporating Population Dynamics 379 15.4.3 Incorporating Biotic Interactions 379 15.5 Software Tools 381 15.5.1 Fitting and Evaluation of Models 381 15.5.2 Incorporating Dispersal or Population Dynamics 381 15.6 Online Exercises 381 15.7 Future Directions 381 References 383 Part V Software Tools 389 16 The R Software for Data Analysis and Modeling 391Clement Calenge 391 16.1 An Introduction to R 391 16.1.1 The Nature of the R Language 391 16.1.2 Qualities and Limits 392 16.1.3 R for Ecologists 392 16.1.4 R is an Environment 393 16.2 Basics of R 393 16.2.1 Several Basic Modes of Data 394 16.2.2 Several Basic Types of Objects 395 16.2.3 Finding Help and Installing New Packages 398 16.2.4 How to Write a Function 400 16.2.5 The for loop 401 16.2.6 The Concept of Attributes and S3 Data Classes 402 16.2.7 Two Important Classes: The Class factor and the Class data.frame 404 16.2.8 Drawing Graphics 406 16.2.9 S4 Classes: Why It is Useful to Understand Them 407 16.3 Online Exercises 410 16.4 Final Directions 410 References 411 Index 413