Simulation to evaluate response of population models to annual trends in detectability

Publication Date

2019-07-23

Start Date

2012-04-23

End Date

2017-06-30

Monroe, A.P., Wann, G., Aldridge, C., and Coates, P.S., 2019, Simulation to evaluate response of population models to annual trends in detectability: U.S. Geological Survey data release, https://doi.org/10.5066/P91L28PG.

Type: Related Primary Publication

These simulations are used to evaluate effects of annual trends in detectability on estimates of population trends from different hierarchical population models. This analysis first simulates hypothetical populations of greater sage-grouse (Centrocercus urophasianus) attending leks and then uses posterior samples of model coefficients characterizing lek attendance by yearling and adult greater sage-grouse from 'lek_attend_age.csv' to inform simulations of the detection process during lek counts. The observed counts are then analyzed with population models that either use the maximum number of grouse observed from repeated counts at lek (peak counts) or uses repeated counts to estimate detectability (N-mixture model). Posterior samples of coefficients from a lek attendance model that averaged attendance rates across age classes ('lek_attend_noage.csv') are used to correct peak counts for variation in attendance or to provide an informative prior for annual attendance in N-mixture models. The files 'lek_attend_age.csv' and 'lek_attend_noage.csv' are required for analyses with R scripts 'run.sim.R' and 'model.run.fun.R'

• Fort Collins Science Center (FORT)
• USGS Data Release Products