Speaker: Bobby Cope, Graduate Assistant-Research, Natural Resource Ecology and Management

Title: Covariates in Mark-Recapture Modeling using program RMark: Inclusion, Analysis, and Prediction.

Abstract: Mark-recapture techniques are widely employed in fisheries and wildlife studies to assess populations. The umbrella of Mark-recapture modeling is wide-ranging, covering topics such as apparent survival, physical movement transition probabilities, population growth rates, and angler/hunter harvest rates. Biotic and abiotic factors often influence these dynamics, and as such, environmental or individual level covariates are often added to these models to help explain variation. The creation of program MARK allowed for streamlined mark-recapture modeling, including complex models with many covariates. Program RMark further streamlines modeling processes and allows for easier model creation and customization with covariates outside the regular MARK interface. Here, I will discuss and walk-through the basics of RMark modeling with covariates including 1) when to include covariates and where they belong in the modeling structure, 2) how to add covariates to models in program RMark, 3) how to analyze covariate effects from RMark output, and 4) how to predict and plot partial effects of covariates in complex RMark models. Hopefully, this demonstration will provide insight on covariates in mark-recapture modeling and provide baseline knowledge for modeling, analyzing, and predicting these effects.

Advance prep:  If you want to follow along / run Bobby’s code, you will need to have program MARK http://www.phidot.org/software/mark/downloads/index.html) and the RMark package (available through RStudio CRAN) downloaded on their machines.  Install Mark first.