I conducted a bird survey in which we used three consecutive observation periods (3 min pre-broadcast, 38 second broadcast, 5 min post-broadcast period) to model occupancy. The current single-season occupancy modeling approach I'm using with UNMARKED (originally written by Ian Fiske) does not seem to account for those different lengths. I know from watching the birds' behavior that they respond very aggressively to playback (thus increasing detectability), but when I include samplingMethod as a variable (indicating that detectability varies among observation periods), the parameter estimates are nonsensical. In the best-fit model, the beta estimates for detectability during different periods are:
βbeforeBroadcast = [no estimate - reference category],
βduringBroadcast = -0.61 ± 0.39 SE,
βafterBroadcast = 0.75 ± 0.35 SE) and an
intercept (β = 7.8 ± 1.9 SE).
The estimates indicate that detectability is lowest during the broadcast, which is clearly not the case as indicated by my observations and by an independent analysis of these data. I'm pretty sure what's happening here is that we're not accounting for the difference in survey length - the post-broadcast period is showing the highest detectability because it is the longest, and the during-broadcast is showing the lowest detectability b/c its the shortest. In the independent analysis, we statistically accounted for the difference in observation period lengths in a GLM framework (using poisson errors, as explained here: http://stackoverflow.com/questions/12208346/in-r-how-do-i-calculate-expected-values-in-a-chi-square-test-when-survey-length), but I'm not sure how to apply this technique to a single-season occupancy framework.
FYI, the best fit model for detectability looks like this:
fm.best <- occu(~ wind + jday + samplingMethod ~ choice, rublUMF, knownOcc = which(rubl$WetChoice == "driveby"))
Any advice on how to account for the different length of survey occasions would be greatly appreciated.
Best,
Luke L. Powell
PhD Candidate
School of Renewable Natural Resources
Louisiana State University
--
You received this message because you are subscribed to the Google Groups "unmarked" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unmarked+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.