1) For the duck harvest data example, focusing on age ratios, describe what happens to the naïve and ML estimates when adult number of bands is 5000, and 500 for young and harvest per year is 500 ducks for the same sequence of true age ratios we simulated in class: age.ratio<-c(2,2,2,1,1,1,1.5,1.5,0.5,0.5). Now consider 2 scenarios for harvest rate variation:
Harvest rates of the two ages are different harvest young=(0.6, 0.55, 0.43, 0.41, 0.35, 0.35, 0.3, 0.3, 0.2, 0.18) , harvest adults = 0.02, 0.07, 0.08, 0.09, 0.11, 0.11, 0.22, 0.23, 0.28,0.32).
Harvest rates of young same as above, adult and young harvest rates the same.
Contrast "naive" and ML estimates of age ratio under each scenario.
2) Using the ‘known fates’ model structure in RMark, and the black duck telemetry data set, create some additional models to explain survival
Condition index and bird age have an additive effect on survival
There is an interaction between condition index and bird age in their effect on survivalT
There is an interaction between minimum temperatures and weight on their effect on survival
Compute model averaged predictions of survival and plot over the mass range using covariate. predictions()
3) Using the ‘nest survival’ model structure in RMark (Chapter 17 and Appendix C.18, MARK book), and the mallard nest survival set, build some additional models predicting duck nest survival. Run the code and interpret the results:
Which one is the best model?
How is nest survival affected by nest age? Hint: look at the sign of the beta coefficient for NestAge under the top model
How is nest survival affected by surrounding vegetation? Hint: look at the sign of the beta coefficient for PpnGrass under the top model
Does nest survival vary with the habitat type where the nest was located? Hint: check the AIC values for models including habitat type effects