The oSCR package (pronounced “Oscar”) implements spatial capture-recapture (SCR) models for wildlife population analysis. Following a workflow modeled after the unmarked package, users format encounter and trap-deployment data into a scrFrame, define the state space, fit likelihood-based models, and post-process results to obtain density and detection estimates.
What’s covered#
- Getting the package – GitHub installation with
install_github("jaroyle/oSCR") - Format sampling data – encounter data file (edf), trap deployment file (tdf),
data2oscr(),scrFramestructure and summary - Create the state space –
make.ssDF()with varying buffer and spatial resolution - Modelling framework – single-session, multi-session, sex-structured, and multi-session sex-structured models
- Analyze the data – likelihood-based model fitting with AIC model selection
- Post-processing output –
get.real()for back-transforming density, detection, and sigma estimates