Population Ecology in Practice

DENNIS L. MURRAY, PHD, is Professor of Biology at Trent University and holds the position of Canada Research Chair in Integrative Wildlife Conservation, Bioinformatics, and Ecological Modeling.BRETT K. SANDERCOCK, PHD, is a Senior Research Scientist in the Department of Terrestrial Ecology at the Norwegian Institute for Nature Research.

• Contributors xviiPreface xxiAbout the Companion Website xxiiiPart I Tools for Population Biology 11 How to Ask Meaningful Ecological Questions 3Charles J. Krebs1.1 What Problems Do Population Ecologists Try to Solve? 31.2 What Approaches Do Population Ecologists Use? 61.2.1 Generating and Testing Hypotheses in Population Ecology 101.3 Generality in Population Ecology 111.4 Final Thoughts 12References 132 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. Thornton2.1 Introduction 172.1.1 Inductive Methods 182.1.2 Hypothetico-deductive Methods 192.1.3 Multimodel Inference 202.1.4 Bayesian Methods 222.2 What Constitutes a Good Research Hypothesis? 222.3 Multiple Hypotheses and Information Theoretics 242.3.1 How Many are Too Many Hypotheses? 252.4 From Research Hypothesis to Statistical Model 262.4.1 Functional Relationships Between Variables 262.4.2 Interactions Between Predictor Variables 262.4.3 Number and Structure of Predictor Variables 272.5 Exploratory Analysis and Helpful Remedies 282.5.1 Exploratory Analysis and Diagnostic Tests 282.5.2 Missing Data 282.5.3 Inter-relationships Between Predictors 302.5.4 Interpretability of Model Output 312.6 Model Ranking and Evaluation 322.6.1 Model Selection 322.6.2 Multimodel Inference 362.7 Model Validation 392.8 Software Tools 412.9 Online Exercises 412.10 Future Directions 41References 42Part II Population Demography 473 Estimating Abundance or Occupancy from Unmarked Populations 49Brett T. McClintock and Len Thomas3.1 Introduction 493.1.1 Why Collect Data from Unmarked Populations? 493.1.2 Relative Indices and Detection Probability 503.1.2.1 Population Abundance 503.1.2.2 Species Occurrence 513.1.3 Hierarchy of Sampling Methods for Unmarked Individuals 523.2 Estimating Abundance (or Density) from Unmarked Individuals 533.2.1 Distance Sampling 533.2.1.1 Classical Distance Sampling 543.2.1.2 Model-Based Distance Sampling 573.2.2 Replicated Counts of Unmarked Individuals 613.2.2.1 Spatially Replicated Counts 613.2.2.2 Removal Sampling 633.3 Estimating Species Occurrence under Imperfect Detection 643.3.1 Single-Season Occupancy Models 653.3.2 Multiple-Season Occupancy Models 663.3.3 Other Developments in Occupancy Estimation 683.3.3.1 Site Heterogeneity in Detection Probability 683.3.3.2 Occupancy and Abundance Relationships 683.3.3.3 Multistate and Multiscale Occupancy Models 683.3.3.4 Metapopulation Occupancy Models 693.3.3.5 False Positive Occupancy Models 703.4 Software Tools 703.5 Online Exercises 713.6 Future Directions 71References 734 Analyzing Time Series Data: Single-Species Abundance Modeling 79Steven Delean, Thomas A.A. Prowse, Joshua V. Ross and Jonathan Tuke4.1 Introduction 794.1.1 Principal Approaches to Time Series Analysis in Ecology 804.1.2 Challenges to Time Series Analysis in Ecology 824.2 Time Series (ARMA) Modeling 834.2.1 Time Series Models 834.2.2 Autoregressive Moving Average Models 834.3 Regression Models with Correlated Errors 874.4 Phenomenological Models of Population Dynamics 884.4.1 Deterministic Models 894.4.1.1 Exponential Growth 894.4.1.2 Classic ODE Single-Species Population Models that Incorporate Density Dependence 904.4.2 Discrete-Time Population Growth Models with Stochasticity 924.5 State-space Modeling 934.5.1 Gompertz State-space Population Model 944.5.2 Nonlinear and Non-Gaussian State-space Population Models 964.6 Software Tools 964.7 Online Exercises 974.8 Future Directions 97References 985 Estimating Abundance from Capture-Recapture Data 103J. Andrew Royle and Sarah J. Converse5.1 Introduction 1035.2 Genesis of Capture-Recapture Data 1045.3 The Basic Closed Population Models: M0, Mt, Mb1045.4 Inference Strategies 1055.4.1 Likelihood Inference 1055.4.2 Bayesian Analysis 1075.4.3 Other Inference Strategies 1085.5 Models with Individual Heterogeneity in Detection 1085.5.1 Model Mh 1085.5.2 Individual Covariate Models 1095.5.2.1 The Full Likelihood 1095.5.2.2 Horvitz-Thompson Estimation 1105.5.3 Distance Sampling 1105.5.4 Spatial Capture-Recapture Models 1105.5.4.1 The State-space 1125.5.4.2 Inference in SCR Models 1125.6 Stratified Populations or Multisession Models 1125.6.1 Nonparametric Estimation 1125.6.2 Hierarchical Capture-Recapture Models 1135.7 Model Selection and Model Fit 1135.7.1 Model Selection 1135.7.2 Goodness-of-Fit 1145.7.3 What to Do When Your Model Does Not Fit 1155.8 Open Population Models 1155.9 Software Tools 1165.10 Online Exercises 1175.11 Future Directions 118References 1196 Estimating Survival and Cause-specific Mortality from Continuous Time Observations 123Dennis L. Murray and Guillaume Bastille-Rousseau6.1 Introduction 1236.1.1 Assumption of No Handling, Marking or Monitoring Effects 1256.1.2 Cause of Death Assessment 1256.1.3 Historical Origins of Survival Estimation 1266.2 Survival and Hazard Functions in Theory 1276.3 Developing Continuous Time Survival Datasets 1306.3.1 Dataset Structure 1316.3.2 Right-censoring 1336.3.3 Delayed Entry and Other Time Considerations 1336.3.4 Sampling Heterogeneity 1346.3.5 Time-dependent Covariates 1356.4 Survival and Hazard Functions in Practice 1356.4.1 Mayfield and Heisey–Fuller Survival Estimation 1356.4.2 Kaplan–Meier Estimator 1366.4.3 Nelson–Aalen Estimator 1386.5 Statistical Analysis of Survival 1386.5.1 Simple Hypothesis Tests 1386.5.2 Cox Proportional Hazards 1396.5.3 Proportionality of Hazards 1406.5.4 Extended CPH 1426.5.5 Further Extensions 1436.5.6 Parametric Models 1436.6 Cause-specific Survival Analysis 1446.6.1 The Case for Cause-specific Mortality Data 1446.6.2 Cause-specific Hazards and Mortality Rates 1456.6.3 Competing Risks Analysis 1466.6.4 Additive Versus Compensatory Mortality 1476.7 Software Tools 1496.8 Online Exercises 1496.9 Future Directions 149References 1517 Mark-Recapture Models for Estimation of Demographic Parameters 157Brett K. Sandercock7.1 Introduction 1577.2 Live Encounter Data 1587.3 Encounter Histories and Model Selection 1597.4 Return Rates 1637.5 Cormack–Jolly–Seber Models 1647.6 The Challenge of Emigration 1647.7 Extending the CJS Model 1677.8 Time-since-marking and Transient Models 1677.9 Temporal Symmetry Models 1687.10 Jolly–Seber Model 1697.11 Multilevel Models 1697.12 Spatially Explicit Models 1707.13 Robust Design Models 1707.14 Mark-resight Models 1717.15 Young Survival Model 1727.16 Multistate Models 1737.17 Multistate Models with Unobservable States 1757.18 Multievent Models with Uncertain States 1767.19 Joint Models 1777.20 Integrated Population Models 1787.21 Frequentist vs. Bayesian Methods 1787.22 Software Tools 1797.23 Online Exercises 1807.24 Future Directions 180References 180Part III Population Models 1918 Projecting Populations 193Stéphane Legendre8.1 Introduction 1938.2 The Life Cycle Graph 1948.2.1 Description 1948.2.2 Construction 1948.3 Matrix Models 1988.3.1 The Projection Equation 1988.3.2 Demographic Descriptors 2008.3.3 Sensitivities 2008.4 Accounting for the Environment 2028.5 Density Dependence 2038.5.1 Density-dependent Scalar Models 2038.5.2 Density-dependent Matrix Models 2038.5.3 Parameterizing Density Dependence 2048.5.4 Density-dependent Sensitivities 2048.6 Environmental Stochasticity 2048.6.1 Models of the Environment 2048.6.2 Stochastic Dynamics 2058.6.3 Parameterizing Environmental Stochasticity 2088.7 Spatial Structure 2088.8 Demographic Stochasticity 2098.8.1 Branching Processes 2098.8.2 Two-sex Models 2108.9 Demographic Heterogeneity 2108.9.1 Integral Projection Models 2118.10 Software Tools 2128.11 Online Exercises 2128.12 Future Directions 212References 2129 Combining Counts of Unmarked Individuals and Demographic Data Using Integrated Population Models 215Michael Schaub9.1 Introduction 2159.2 Construction of Integrated Population Models 2169.2.1 Development of a Population Model 2169.2.2 Construction of the Likelihood for Different Datasets 2189.2.3 The Joint Likelihood 2209.2.4 Fitting an Integrated Population Model 2219.3 Model Extensions 2239.3.1 Environmental Stochasticity 2239.3.2 Direct Density Dependence 2249.3.3 Open Population Models and Other Extensions 2269.3.4 Alternative Observation Models 2269.4 Inference About Population Dynamics 2279.4.1 Retrospective Population Analyses 2279.4.2 Population Viability Analyses 2279.5 Missing Data 2299.6 Goodness-of-fit and Model Selection 2309.7 Software Tools 2309.8 Online Exercises 2319.9 Future Directions 231References 23210 Individual and Agent-based Models in Population Ecology and Conservation Biology 237Eloy Revilla10.1 Individual and Agent-based Models 23710.1.1 What an IBM is and What it is Not 23810.1.2 When to Use an Individual-based Model 23810.1.3 Criticisms on the Use of IBMs: Advantages or Disadvantages 23910.2 Building the Core Model 23910.2.1 Design Phase: The Question and the Conceptual Model 23910.2.2 Implementation of the Core Model 24010.2.3 Individuals and Their Traits 24010.2.4 Functional Relationships 24410.2.5 The Environment and Its Relevant Properties 24410.2.6 Time and Space: Domains, Resolutions, Boundary Conditions, and Scheduling 24410.2.7 Single Model Run, Data Input, Model Output 24610.3 Protocols for Model Documentation 24710.3.1 The Overview, Design Concepts, and Details Protocol 24910.4 Model Analysis and Inference 24910.4.1 Model Debugging and Checking the Consistency of Model Behavior 24910.4.2 Model Structural Uncertainty and Sensitivity Analyses 25210.4.3 Model Selection, Validation, and Calibration 25410.4.4 Answering your Questions 25610.5 Software Tools 25710.6 Online Exercises 25710.7 Future Directions 257References 258Part IV Population Genetics and Spatial Ecology 26111 Genetic Insights into Population Ecology 263Jeffrey R. Row and Stephen C. Lougheed11.1 Introduction 26311.2 Types of Genetic Markers 26411.2.1 Mitochondrial DNA 26411.2.2 Nuclear Introns 26511.2.3 Microsatellites 26511.2.4 Single Nucleotide Polymorphisms 26511.2.5 Next-generation Sequencing 26511.3 Quantifying Population Structure with Individual-based Analyses 26611.3.1 Bayesian Clustering 26711.3.2 Multivariate Analysis of Genetic Data Through Ordinations 26911.3.3 Spatial Autocorrelation Analysis 27111.3.4 Population-level Considerations 27311.4 Estimating Population Size and Trends 27311.4.1 Estimating Census Population Size 27711.4.2 Estimating Contemporary Effective Population Size with One Sample Methods 27711.4.3 Estimating Contemporary Effective Population Size with Temporal Sampling 27911.4.4 Diagnosing Recent Population Bottlenecks 28011.5 Estimating Dispersal and Gene Flow 28111.5.1 Estimating Dispersal and Recent Gene Flow 28211.5.2 Estimating Sustained Levels of Gene Flow 28211.5.3 Network Analysis of Genetic Connectivity 28311.6 Software Tools 28411.6.1 Individual-based Analysis 28411.6.2 Population-based Population Size 28511.6.3 Dispersal and Gene Flow 28611.7 Online Exercises 28611.8 Future Directions 286Glossary 287References 28912 Spatial Structure in Population Data 299Marie-Josée Fortin12.1 Introduction 29912.2 Data Acquisition and Spatial Scales 30212.3 Point Data Analysis 30212.4 Abundance Data Analysis 30412.5 Spatial Interpolation 30612.6 Spatial Density Mapping 30812.7 Multiple Scale Analysis 30812.8 Spatial Regression 31112.9 Software Tools 31212.10 Online Exercises 31212.11 Future Directions 312Glossary 312References 31313 Animal Home Ranges: Concepts, Uses, and Estimation 315Jon S. Horne, John Fieberg, Luca Börger, Janet L. Rachlow, Justin M. Calabrese and Chris H. Fleming13.1 What is a Home Range? 31513.1.1 Quantifying Animal Home Ranges as a Probability Density Function 31613.1.2 Why Estimate Animal Home Ranges? 31813.2 Estimating Home Ranges: Preliminary Considerations 31913.3 Estimating Home Ranges: The Occurrence Distribution 32113.3.1 Minimum Convex Polygon 32113.3.2 Kernel Smoothing 32213.3.3 Models Based on Animal Movements 32313.3.4 Estimation from a One-dimensional Path 32413.4 Estimating Home Ranges: The Range Distribution 32413.4.1 Bivariate Normal Models 32413.4.2 The Synoptic Model 32413.4.3 Mechanistic Models 32513.4.4 Kernel Smoothing 32613.5 Software Tools 32613.6 Online Exercises 32713.7 Future Directions 32713.7.1 Choosing a Home Range Model 32713.7.2 The Future of Home Range Modeling 327References 32814 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 Jachowski14.1 Introduction 33314.2 Definitions 33514.2.1 Terminology and Currencies of Use and Availability 33514.2.2 Use-availability, Paired Use-availability, Use and Non-use (Case-control), and Use-only Designs 33614.2.3 Differences Between RSF, RSPF, and RUF 33614.3 Considerations in Studies of Resource Selection 33814.3.1 Two Important Sampling Considerations: Selecting Sample Units and Time of Day 33814.3.2 Estimating the Number of Animals and Locations Needed 33814.3.3 Location Error and Fix Rate Bias Resource Selection Studies 33914.3.4 Consideration of Animal Behavior in Resource Selection Studies 33914.3.5 Biological Seasons in Resource Selection Studies 34014.3.6 Scaling in Resource Selection Studies 34014.3.7 Linking Resource Selection to Fitness 34114.4 Methods of Analysis and Examples 34214.4.1 Compositional Analysis 34214.4.2 Logistic Regression 34314.4.3 Sampling Designs for Logistic Regression Modeling 34414.4.3.1 Random Sampling of Units within the Study Area 34414.4.3.2 Random Sampling of Used and Unused Units 34414.4.3.3 Random Sample of Used and Available Sampling Units 34514.4.4 Discrete Choice Models 34614.4.5 Poisson Regression 34714.4.6 Resource Utilization Functions 34814.4.7 Ecological Niche Factor Analysis 34814.4.8 Mixed Models 34914.5 Software Tools 34914.6 Online Exercises 35014.7 Future Directions 350References 35115 Species Distribution Modeling 359Daniel H. Thornton and Michael J.L. Peers15.1 Introduction 35915.1.1 Relationship of Distribution to Other Population Parameters 36215.1.2 Species Distribution Models and the Niche Concept 36315.2 Building a Species Distribution Model 36615.2.1 Species Data 36615.2.2 Environmental Data 36815.2.3 Model Fitting 36815.2.4 Interpretation of Model Output 37115.2.5 Model Accuracy 37215.3 Common Problems when Fitting Species Distribution Models 37415.3.1 Overfitting 37415.3.2 Sample Selection Bias 37515.3.3 Background Selection 37615.3.4 Extrapolation 37715.3.5 Violation of Assumptions 37815.4 Recent Advances 37815.4.1 Incorporating Dispersal 37815.4.2 Incorporating Population Dynamics 37915.4.3 Incorporating Biotic Interactions 37915.5 Software Tools 38115.5.1 Fitting and Evaluation of Models 38115.5.2 Incorporating Dispersal or Population Dynamics 38115.6 Online Exercises 38115.7 Future Directions 381References 383Part V Software Tools 38916 The R Software for Data Analysis and Modeling 391Clément Calenge 39116.1 An Introduction to R 39116.1.1 The Nature of the R Language 39116.1.2 Qualities and Limits 39216.1.3 R for Ecologists 39216.1.4 R is an Environment 39316.2 Basics of R 39316.2.1 Several Basic Modes of Data 39416.2.2 Several Basic Types of Objects 39516.2.3 Finding Help and Installing New Packages 39816.2.4 How to Write a Function 40016.2.5 The for loop 40116.2.6 The Concept of Attributes and S3 Data Classes 40216.2.7 Two Important Classes: The Class factor and the Class data.frame 40416.2.8 Drawing Graphics 40616.2.9 S4 Classes: Why It is Useful to Understand Them 40716.3 Online Exercises 41016.4 Final Directions 410References 411Index 413