Course (
Venue: Architecture Computer Lab (Arch 116),
Dates: March 10, 12 -17, Timing: 8.00 am – 5.00 pm with a one hour lunch break and 15 minute morning and afternoon breaks
Instructors:
James E. Hines (
William L. Kendall (
James D. Nichols (
John R. Sauer (
Madan Oli
and H. Franklin Percival (
Course Objective:
To present a unified and science-based approach to the conservation and management of natural animal populations, and to provide participants with information and resources for implementation of this approach. This approach involves three major methodological components: modeling, estimation and decision making.
Specific Objectives:
(1) To provide a conceptual framework for the use of models in the conduct of science
and management.
(2) To briefly review frequently-used populations models, with emphasis on tailoring models to their intended use in conservation and/or science.
(3) To present a general conceptual framework for animal population/community
estimation methods.
(4) To show how this framework can be used to develop estimation methods applicable to various sampling and logistic situations.
(5) To present the specific rationale and logic underlying the more commonly used
approaches to estimating population and community-level attributes, with emphasis on tailoring these methods to meet objectives under logistical constraints.
(6) To present a general rationale and approach for the development of an animal
monitoring program, with emphasis on the use of resulting inferences for conservation and management.
(7) To present a logical framework for making management decisions and to identify the
major components of uncertainty typically encountered in the management process.
(8) To outline the implementation of a formal adaptive management process for making
informed management decisions in the face of uncertainty.
Outline:
Day 1:
1.1. Introduction to workshop (Nichols) (0.25 hr)
1.2. Introduction of instructors/participants/students and their backgrounds and objectives
(Group) (0.5 hr)
2.1. Conceptual framework for population ecology & management (Nichols)
(0.25 hr)
Include roles of modeling, estimation and decision theory
BIDE model
3. Statistical Inference
3.1.
Statistical distributions (e.g., normal, multinomial) (
3.2.
Parameter estimation (
Estimator properties (bias, precision, accuracy)
Estimation methods
Confidence intervals
BREAK
3.3.
Hypothesis testing (
Type I and II errors
Power
Likelihood ratio tests
Goodness-of-fit tests
3.4. Model
selection (information theoretic approaches) (
3.5.
Bayesian model updating (
LUNCH
3.6. Hierarchical modeling:
Bayesian approach (Sauer) (0.5 hr)
3. Statistical Inference (Continued) (Sauer) (1.00 hr)
3.7. Survey sampling (sources of variation)
3.8. Sampling design features
Replication
Randomization
Control of variation
3.9. Some designs
Simple random sampling
Stratified random sampling
Other (cluster, systematic, double, dual frame, adaptive)
4. Models
4.1. Role of models in science and management (Nichols) (0.5 hr)
BREAK
4.2. Population modeling review: basic principles (Sauer) (1.5 hr)
Discrete time matrix modeling (age/stage)
Projection matrix asymptotics (l, sensitivity, reproductive value, stable stage distribution)
Stochasticity (demographic, environmental), PVAs
Models for management
4.3. Population modeling exercise (Sauer, Hines) (0.75 hr)
Day 2:
4.3. Population modeling exercise cont. (Sauer, Hines) (0.75 hr)
5. Estimation of Animal Abundance and Density
5.1. Overview (Nichols) (0.5 hr)
Why estimate abundance? Role of monitoring in science and management.
How to estimate abundance: a canonical estimator
Indices
5.2. Observation-based methods: miscellaneous (Nichols) (0.5 hr)
Marked subpopulation
Temporal removal modeling
BREAK
5.2. Observation-based methods:
miscellaneous cont. (Nichols) (0.75 hr)
Sighting probability modeling
Multiple independent observers
Multiple dependent observers
5.3. Implementing observation-based methods
Introduction to MARK (Hines) (0.5 hr)
Computer exercises with DOBSERV and/or MARK (Hines) (0.5 hr)
LUNCH
5.4. Observation-based methods: distance sampling
Introduction to Distance Sampling (Sauer) (0.5 hr)
Introductory Concepts
Assumptions Underlying the Sampling Technique
Estimating the proportion of animals detected & counted (Sauer)(0.5 hr)
Line Transects
Point transect
Contrasting Line Transect & Point Transect Sampling
Survey Design & Field Protocol (Sauer) (0.5 hr)
Precision
Bias
BREAK
DISTANCE 4 Software (Sauer-Hines) (1.0 hr)
Brief overview
Automated Survey Design (Distance 4 exercises)
Distance Sampling Analysis (Sauer) (1.5 hr)
Basic Analysis
Analysis for Clustered Populations
Introducing Covariates into the Analysis
Distance 4 CDS/MCDS analysis exercise
Day 3:
5.5. Capture-based methods: closed CR models
2-sample model (Nichols) (0.50 hr)
Data structure
Models and estimators
Study design
2-sample model exercises (SURVIV, MARK) (Hines) (1.0 hr)
K-sample
closed models (
Data structure
Models
BREAK
K-sample
closed models cont. (
Models
Model testing and selection
Confidence interval estimation
Study design
K-sample closed model exercises, CAPTURE, MARK (Hines) (1.0 hr)
LUNCH
K-sample closed model exercises cont. (Hines) (0.5 hr)
5.6. Density estimation with closed CR models (Nichols) (0.5 hr)
Ad hoc boundary strip approach
Nested grids
Gradient designs (e.g., trapping webs)
5.7. Other capture-based methods (
Removal methods
Change-in-ratio methods
6. Estimation of Animal Vital Rates (survival,
reproduction, movement)
6.1. Introduction, relevance of detection probability (Nichols) (0.25 hr)
BREAK
6.2. All marked animals detected (Sauer) (1.0 hr)
Binomial survival model
Nest success
Radiotelemetry data
Study design
Computer exercises (SURVIV, MARK) (Hines) (1.0 hr)
Day 4:
6.3. Tag recovery models (Sauer) (1.0 hr)
6.4. Open population CR models
Single-age models (Nichols) (0.5 hr)
Data structure
Modeling
Single-models continued (Nichols) (0.5 hr)
Time-specific covariates
Multiple groups
Capture history effects
Individual covariates
Model selection
Model assumptions
Estimator robustness
BREAK
MARK: PIMs and design matrices (Hines) (0.75 hr)
MARK exercises: Single-age models, band recovery models (Hines) (1.0 hr)
LUNCH
Single-age models (Nichols) (1.0 hr)
Estimation of abundance
Estimation of 8 and components of 8
Multiple-age models (Nichols) (0.5 hr)
Data structure
Modeling
Multiple-age model exercise (Hines (0.5 hr)
BREAK
Multiple-age model exercise cont. (Hines (0.5 hr)
Multistate models (
Data structure
Modeling
Multistate model exercise (Hines) (0.5 hr)
Day 5:
Multistate model exercise cont. (Hines) (0.5 hr)
Multistate
models: special uses (
Unobservable states
Band loss
Multistate models: state misclassification (
BREAK
Multiple-age multisdtate models: variable age at recruitment (Nichols) (0.5 hr)
6.5. Open models with extra information (Nichols 0.5 hr)
Capture-recapture + band recoveries
Capture-recapture + radio telemetry
Capture-recapture + auxiliary sightings (Barker models)
6.6 Pollock’s robust design
Introduction
(
Data structure
Ad hoc approach
Recruitment components
Model-based approach
LUNCH
Model
extensions (
Temporary emigration
Open robust design
Robust design with band recoveries
Multistate robust design
“Mother of all Models”
Robust design computer exercises (Hines) (1.0 hr)
BREAK
7. Estimation of
species richness and community dynamics
7.1. Population-community analogy (Sauer) (0.25 hr)
7.2. Species richness estimation (Sauer) (0.5 hr)
Data structure and designs
Modeling and estimation
7.3. Multiple-season community dynamics (Sauer) (0.5 hr)
Data structure
Modeling and estimation
7.4. Community dynamics exercises with SPECRICH, COMDYN (Hines)(0.5 hr)
DAY 6:
7.4. Community dynamics exercises cont. (Hines) (0.5 hr)
8. Estimation of site occupancy and occupancy dynamics
8.1. Single-season, single species occupancy (Nichols) (0.5 hr)
Data structure and designs
Modeling
Assumptions and their relaxation
Computer exercise (PRESENCE) (Hines) (0.75)
BREAK
8.2. Multiple-season occupancy dynamics (Nichols) (0.75 hr)
Data structure
Modeling
Example(s)
Computer exercise with PRESENCE (Hines) (0.5 hr)
8.3 Occupancy extensions (Nichols) (0.5 hr)
2-species occupancy
Multi-state occupancy
Joint occupancy-habitat modeling
Community level occupancy
LUNCH
9. Conservation/Management in the face of uncertainty
9.1. Elements of an informed
decision (
Objectives
Management alternatives
Model(s) of system response to management
Model weights (for multiple models)
Monitoring program
9.2.
Sources of uncertainty (
Environmental variation
Partial controllability
Structural uncertainty
Partial observability
9.3. Decision analysis under uncertainty (Nichols 0.5 hr)
General approach
Example
9.4. Adaptive management (Kendall)
The process (0.5 hr)
BREAK
Examples (0.5 hr)
10. Exam review/questions (Hines-Kendall-Nichols-Sauer)
(1.5 hr +)
11. Exam (4.0 hr)
LUNCH
12. Discussion/Evaluation/Consultation
(Hines-Kendall-Nichols-Sauer) (4 hr)