multistate-owl-models

Multi-state occupancy models for northern spotted owls

Cara Appel

---

CONTENTS

bugs

• folder contains text files with models for JAGS (see model descriptions, below)

data2018

• nso_dh_2018.csv: spotted owl detection history
• covariates folder: contains raw and standardized versions of covariate data

scripts

• step01_models: folder with R scripts to run each model (see descriptions, below)
• step02_submodel_selection.R: evaluating sub-models using sequential-by-submodel selection
• step03_model_evaluation.R: evaluating most-supported model from sub-model selection (includes derived parameters for reporting in-text and tables, creating marginal plots, etc.)
• step04_inference_model.R: evaluating model for interpreting covariate effects on psi and R
• step05_inference_without_area.R: creating marginal plots from model without the "study area" factor covariate

figures

• location for figure outputs

MODEL DESCRIPTIONS

01a: p submodel (initial) p(NOISE + EFFORT + SEASON + BO_WEEKLY + NR200) delta(.) psi(.) R(.)
01b: p submodel (slab-spike) p(NOISE + EFFORT + SEASON + BO_WEEKLY + NR200) delta(.) psi(.) R(.)

02a: psi submodel (initial) p(NOISE + EFFORT) delta(.) psi(STUDY_AREA + NR500 + BO_TOTAL) R(.)
02b: psi submodel (slab-spike) p(NOISE + EFFORT) delta(.) psi(STUDY_AREA + NR500 + BO_TOTAL) R(.)

03a: R submodel (initial) p(NOISE + EFFORT) delta(.) psi(.) R(STUDY_AREA + NR500 + BO_TOTAL)
03b: R submodel (slab-spike) p(NOISE + EFFORT) delta(.) psi(.) R(STUDY_AREA + NR500 + BO_TOTAL)

04: most-supported model from sub-model selection process p(NOISE + EFFORT) delta(.) psi(.) R(.)

05a: model for inference, "normal1" priors: dnorm(0,0.368)
p(NOISE + EFFORT) delta(.) psi(STUDY_AREA + NR500 + BO_TOTAL) R(STUDY_AREA + NR500 + BO_TOTAL)

05b: model for inference, "normal2" priors: dnorm(0,0.5)
p(NOISE + EFFORT) delta(.) psi(STUDY_AREA + NR500 + BO_TOTAL) R(STUDY_AREA + NR500 + BO_TOTAL)

05c: model for inference, "uniform" priors: dunif(-5,5)
p(NOISE + EFFORT) delta(.) psi(STUDY_AREA + NR500 + BO_TOTAL) R(STUDY_AREA + NR500 + BO_TOTAL)

06: model for inference, "normal1" priors: dnorm(0,0.368)
p(NOISE + EFFORT) delta(.) psi(NR_SUIT500 + BO_TOTAL) R(NR_SUIT500 + BO_TOTAL)
- STUDY_AREA covariate omitted from psi and R to make marginal plots for other covariates