Integrated statistical analysis based on likelihood and confidence: applications to the hare-lynx cycles and the status of bowhead whales

The proposed project aims at further developing statistical methodology for handling a great variety (and kinds) of data (e.g., a combination of ecological and genetic data). Fisherian likelihood methods will be coupled with numerical techniques and simul ation/bootstrapping to obtain confidence distributions capturing the uncertainty in conclusions. This frequentist methodology will be compared to Bayesian methods in the two cases of application (the hare-lynx cycle and the bowhead whale population status ). In both applications we aim at obtaining substantive results of broad interest; these two application-components will be carried out in close cooperation with biologists working within these field. The two biological systems are of substantial general interest – one (the hare-lynx dynamics) having fascinated ecologists for more than 70 years, the other (the bowhead whales in the Bering-Chukchi-Beaufort Seas) being of great interest to the management and conservation of harvested biological resources. T he project aspires to be of interest to both the statistical and the biological/ecological communities. The latter will benefit both with respect to methodology and substantive knowledge within the field of biology.