Workshop on Spatial Capture-Recapture Modeling, December 10-12 2013, Patuxent Wildlife Research Center

[Chris Merkord]

Workshop on Spatial Capture-Recapture Modeling, December 10-12 2013, Patuxent Wildlife Research Center

Instructors:    Andy Royle (Patuxent Wildlife Research Center), Chris Sutherland (Cornell Univ.), Richard Chandler (Univ. Georgia) and Rahel Sollmann (North Carolina State Univ.).
Date:               10–12 December 2013 (3 days)
Venue:            USGS Patuxent Wildlife Research Center, Laurel/Maryland, USA
Computers:    People must bring their own laptops with a recent version of R (> 3.0), WinBUGS and JAGS installed and functional
Costs:              USD$450 (normal rate), USD$300 (student rate).
 
This course introduces some key models used in the analysis of spatially-explicit capture-recapture data using Bayesian and classical (likelihood) methods. It closely follows material from the recent book “Spatial Capture-Recapture” (Elsevier, 2013).
 
The workshop will cover data analysis and basic modeling concepts in R and the BUGS language (using WinBUGS or JAGS), analysis of classical capture-recapture models (models M0, Mb, Mh, etc..), concepts and methods of spatial capture-recapture models implemented in BUGS and also using the R package “secr” written by M. Efford.
 
This is an intermediate-level workshop with topical lectures followed by work sessions involving data analysis. Course requirements:

1. Good working knowledge of modern regression methods (regression, ANOVA, generalised linear models, random effects).
2. Experience with the R programming language at a level higher than “beginner”. You should be a regular R user. Manipulating data in R should be routine for you.
3. Familiarity with using WinBUGS or JAGS. You should be able to fit a basic regression model in WinBUGS.
4. Knowledge of basic capture-recapture concepts and methods. E.g., the Otis et al. monograph, Pollock’s Robust Design, distance sampling.

If you are interested in attending the workshop please submit your application no later than 08 November 2013 by filling out the following webform:
 
https://docs.google.com/forms/d/1sCuADYaNtotgWxu4DtU-Y4p4RW8l2rEXWVXz__K7otc/viewform
 
Direct inquiries can be made to Andy Royle (aroyle@usgs.gov). Workshop invitations will be sent out beginning 31 OCT until all seats are filled.
 
An approximation to the workshop content is as follows:
 
Day 1: Introductory material (R/WinBUGS and closed population models)

AM: Workshop introduction. Fitting basic statistical models (regression, GLMs) in WinBUGS/JAGS, from R, using the R2jags or R2WinBUGS packages.
PM: Introduction to ordinary capture-recapture models: Model M0, Mb, Mh, individual covariate models, distance sampling, etc. Bayesian analysis of models in WinBUGS/JAGS using data augmentation. The spatial context of capture-recapture: Heterogeneity, estimating density, what is the effective sample area? Buffering trap arrays.

 

Day 2: Introduction to Spatial Capture-Recapture

AM: General formulation of SCR models: basic models and their analysis in WinBUGS. Poisson and Bernoulli models for trapping arrays (camera traps, hair snares, etc..), multinomial encounter models, area-search models.
PM: Likelihood analysis concepts. Using the R package secr.

 

Day 3: Advanced SCR models

SCR for stratified populations (multi-session models in secr)

Models with density covariates
Modelling landscape connectivity
Space-usage and resource selection (SCR + telemetry data)
Open population models