We are interested in developing new methods for modeling population dynamics, estimating model parameters, analyzing population declines, and validating extinction risk estimates. Research in this areas has involved
- A new approach to develop fully specified population models (complete with stage structure, stochasticity, and density dependence) from mark-recapture data (the method, including the source code, the data we analyzed, and sample results for one species, can be freely downloaded via github.com/Akcakaya/MAPS-to-Models);
- A new method of global sensitivity analysis;
- Linking models at different scales to develop "metamodels" for simulating complex dynamics of species interactions, diseases, and dispersal beavior (ongoing Research Collaboration Network project supported by NSF);
- Estimating variability in survival rates and fecundities due to environmental fluctuations, by removing variance due to demographic stochastricity from total observed variance;
- Showing that the theta-logistic model (used to describe how the growth rate of a population slows as abundance increases) is unreliable for modeling most census data;
- Establishing that scalar population models (also known as count-based models) overestimate variability and risks of decline and extinction;
- Demonstrating that the reddened spectra commonly seen in time series of population abundances can be due to measurement error and/or moderate density dependence;
- Determining the predictive accuracy of the probabilistic predictions of population viability analysis, with a retrospective analysis;
- Cataloging common mistakes in population modeling and suggesting ways to avoid them.
