Data from bird counts may provide insight into big picture trends, but only individual-level data allows for understanding of individual and group variation in migration locations and paths. We are working towards the development of novel methods to combine individual bird tracking data from Smithsonian collections with spatio-temporal eBird abundance data to understand population variation in migratory patterns. Delineating and understanding population variation in migratory patterns are the first steps in determining why some populations throughout a bird’s range may be increasing in abundance while others are declining. We are developing analytical approaches to combine individual tracking data of migrating birds tracked using the ARGOS system with population-level relative abundance obtained from observations in eBird. With this data fusion approach researchers may subdivide bird populations with distinct migratory patterns.
Collaboration with Limnologists: Joint work developed with members of the Cont-Limno research team using LAGOS-NE data. This collaboration has lead to additional opportunities to contribute to various projects lead by different members of the research team. LAGOS is a multi-scaled database system and a set of tools to study lake water quality at macroscales.
Extrapolation is defined as making predictions beyond the range of the data used to estimate a statistical model. In ecological studies, it is not always obvious when and where extrapolation occurs because of the multivariate nature of the data. Previous work on identifying extrapolation has focused on univariate response data, but these methods are not directly applicable to multivariate response data, which are more and more common in ecological investigations. In this work, we extended previous work that identified extrapolation by applying the predictive variance from the univariate setting to the multivariate case. We illustrated our approach through an analysis of jointly modeled lake nutrients and indicators of algal biomass and water clarity in over 7000 inland lakes from across the Northeast and Mid-west US. In addition, we illustrated novel exploratory approaches for identifying regions of covariate space where extrapolation is more likely to occur using classification and regression trees.
Collaboration with Entomologists: Data and ecological input for this research was provided by members of the David Hughes Ant Behavior Lab at Penn State University.
Interactions between social animals provide insights into the exchange and flow of nutrients, disease, and social contacts. We consider a chamber level analysis of trophallaxis interactions between carpenter ants (Camponotus pennsylvanicus) over 4 hours of second-by-second observations. The data show clear switches between fast and slow modes of trophallaxis. However, fitting a standard hidden Markov model (HMM) results in an estimated hidden state process that is overfit to this high resolution data, as the state process fluctuates an order of magnitude more quickly than is biologically reasonable. We proposed a novel approach for penalized estimation of HMMs through a Bayesian ridge prior on the state transition rates while also incorporating biologically motivated covariates. This penalty induces smoothing, limiting the rate of state switching that combines with appropriate covariates within the colony to ensure more biologically feasible results. We developed a Markov chain Monte Carlo algorithm to perform Bayesian inference based on discretized observations of the contact network.