Populations of giant Canada geese (Branta canadensis maxima) have rebounded substantially since the early 1900's, when this species was thought to be extinct in the Mississippi Flyway. The success of reintroductions, along with the species' high survival and reproductive rates and affinity for relatively safe urban environments, has allowed them to become abundant to point that they have in many cases become a nuisance. The primary method used to manage these increasing populations has been the use of harvest regulations. Geese are also banded annually, allowing for estimation of survival and recovery rates. The analysis of band recovery data offers the ability to identify the affects of harvest regulations and provide information on demographics of goose populations.

I used band recovery data to investigate three issues pertaining to survival and recovery rates of giant Canada geese within Iowa and the Mississippi Flyway. I first investigated the issue of banding sample sizes throughout the Mississippi Flyway by deriving sample size estimation procedures to achieve explicit statistical objectives of precision and power. Second, I investigated the effects on survival and recovery rates of special early and metropolitan hunting seasons in Iowa. I also examined the effects of these seasons on spatial and temporal distribution of recoveries. Lastly, I investigated the phenomenon known as molt migration and the potential effects of molt migrant geese on the estimation of survival rates using standard band recovery models. I also developed an alternative model incorporating molt migrants and compared the performance of this model with that of standard band recovery models.

Results from sample size exercises demonstrated that mean annual survival rate CV's of 4% can be achieved through banding approximately 1,000-1,500 adults and 1,800-3,200 juveniles annually per state/province. An annual CV of 9% can be achieved through banding approximately 1,200-2,100 adults and 1,400-2,400 juveniles annually per state/province. Results also showed that in order to detect a change in pre- and post-regulation change recovery rates with significant power, studies should use α ≥ 0.05 and should be ≥ 3 years pre- and post; i.e., 6 years total. Target effect sizes should be ≥ 1.25.

Analyses of special harvest seasons did not reveal conclusive evidence that special 2-day early season harvests directly affected recovery rates of geese in Iowa. However, results suggested that these early season harvests did affect the temporal distribution of recoveries in September. Analyses also suggested that special metropolitan hunting seasons selectively affected survival and recovery rates of Iowa's geese. Juvenile geese banded within the Iowa City/Cedar Rapids metropolitan hunting zone were found to have significantly higher recovery rates (0.178-0.246) than geese from any other cohort (0.064-0.132). Additionally, estimates of survival were found to be much lower for ICCR birds (0.444-0.481) than for DM or Rural birds (0.657-0.860). Results suggested that the presence of molt migrant geese in the banded population may be affecting model performance.

The presence of molt migrants in simulated data was found to affect model selection and estimation of survival rates. Survival estimates from standard models were found to be slightly negatively biased for adults and moderately to extremely positively biased for juveniles. Further examinations suggested that increased numbers of recoveries in band recovery data produced by molt migrants resulted in biased maximum likelihood estimates of juvenile survival. I developed an alternative model to incorporate the presence of molt migrants and compared the performance of this model relative to standard band recovery models. Simulations showed that this model was selected a large majority of the time in the presence of molt migrants. However, estimates of survival from this model, while not unreasonable, were difficult to interpret.