Developing population models with data from marked individuals

Background/Question/Methods

Despite recent progress in the analysis of data from marked individuals, developing a complete population model based on the results of such analyses remains complicated and problematic for a variety of reasons, including lack of estimates for fecundity, density-dependence, true survival, and temporal variability of survival and fecundity.  We present an integrated method for creating comprehensive and stochastic matrix-based population models based on data from the Mapping Avian Productivity and Survivorship (MAPS) program. The method allows estimating fecundity, adult and juvenile survival rates, their natural temporal variability (removing sampling variability), their dependence on density, and if combined with weather data, their functional relationship with climatic variables. This makes it possible to construct a population model that can be used for long-term projection of population dynamics. A main advantage of our method is that its key outputs may be realized with data from a single, widely available database (the MAPS program), and information about the basic biology of the species, without the need for additional data on fecundity, variability, or density-dependence.  We apply this approach to 9 bird species and demonstrate the feasibility of using data from the MAPS program to fully parameterize stochastic, density-dependent, stage-structured population models for avian species.  

Results/Conclusions

Recapture probability of adults and juveniles were estimated as a positive function of effort.  The confidence intervals for adult recapture probability were substantially narrower than for juveniles. For all but one species, fecundity was a negative function of density, as would be expected in populations fluctuating around carrying capacity. Average fecundity at mean density ranged from 0.7 to 4.8 juvenile per adult for the 9 species, and temporal variability ranged from 6% to 46%. For four species, survival was a negative function of density, as would be expected again in populations fluctuating around carrying capacity. Average apparent survival rate at mean density ranged from 0.38 to 0.54 for adults, and 0.05 to 0.31 for juveniles. For many species, the relationship of vital rates with weather suitability was unexpected, with lower survival and/or fecundity at higher weather suitability.  For all species weather explained a very small proportion of the variability in survival and fecundity. The results demonstrated the feasibility of using data from the MAPS program to fully parameterize stochastic, density-dependent, stage-structured population models for avian species. Based on a single data source, we expect this method to ease the process of modeling population dynamics for avian species with mark-recapture data.