Understanding region-specific survival, productivity and movement among regions is essential for holistic understanding of demography, but often requires substantial live encounter data. Data integration allows estimation of demographic parameters that are otherwise inestimable with sparse data. In this study, we aimed to integrate population survey, live encounter and dead encounter data to quantify Northern Pintail (Anas acuta) metapopulation dynamics in several ecological regions of North America. These migratory birds are of particular conservation concern because they have not positively responded to abundant ponds on breeding areas. Pintail population counts were obtained through the North American Waterfowl Breeding Population and Habitat Survey for 1961 to 2014. Pintails were also regularly banded during the same period, and bands returned to federal agencies when birds were harvested and bands reported, generating a set of age-structure at banding and dead encounter data that can be used to inform productivity and survival. The banding effort also resulted in a relatively small amount of live encounter data, which have not been efficiently used in previous analyses. Using these data types, we developed an integrated population model (IPM) to estimate pintail movement among three primary breeding regions including Prairie (PR), Northern Unsurveyed (NU), and Alaska (AK), in addition to region-specific productivity, survival and population size.
Results/Conclusions
Estimated population sizes were consistent with population survey data, and showed an increasing trend in AK and decreasing trends during the 1970s and 1980s in PR and NU. Productivity was on average higher, but also less variable in PR than in AK and NU. Productivity showed decreasing trends in PR and NU without obvious shifts in AK. Survival was generally stable among years and similar among regions. Survival was higher for males than females, but similar for adults and juveniles of the same sex. We found high site fidelity (i.e., a large probability for AK birds returning to AK, and PR birds returning to PR). However, we also found that the probability of overflight (i.e., birds reaching PR in spring subsequently moving to NU and AK) was highly variable among years, which may be driven by pond abundance in PR and density dependent processes. Estimation of movement-related parameters by jointly analyzing limited live encounter data with rich count and dead encounter data is encouraging, particularly for species of great conservation importance such as pintails, and illustrates the importance of data integration to tackle pressing ecological problems.