Projects

This page lists a selection of current and recent projects in which I and/or people in the group have been involved.

Statistical modeling of species distribution and abundance

A problem of fundamental importance in ecology is how to relate the raw, messy data of ecological surveys, citizen science programs, and field experiments to models of how species use habitats and change through time. Examples of collaborations in this area include papers on dynamics of bee communities in restored agricultural habitats (led by Lauren Ponisio), changes of bird communities over a century from the Grinnell Resurvey Project (with Steve Beissinger’s lab), and methods for multi-species statistical modeling (led by Marti Anderson).

Abundance and demography of large carnivores in Scandinavia

We work with Richard Bischof’s group at the Norwegian University of Life Sciences on large-scale spatial capture-recapture (SCR) analysis of wolves, brown bears, and wolverines from many thousands of scat samples labeled to individual animals using DNA. Together we have developed computational methods using our NIMBLE software to scale up SCR methods (led by Daniel Turek) and made them available via R package nimbleSCR. Using these methods, Bischof et al. estimated the largest-scale spatial understanding of these populations available.

Statistical modeling of agricultural ecosystems

We have been involved in several kinds of agricultural ecology questions. Recently these have included meta-analysis of conventional vs organic cropping systems (led by Lauren Ponisio) and large-scale analysis of crop rotational complexity (led by Yvonne Socolar).

Population dynamics

Recent (and not so recent) projects on population dynamics include joining Ken Newman et al. in reviewing methods for state-space (time-series) models for population dynamics, theoretical analysis of the role of individual heterogeneity in stage-structured population dynamics and statistical analysis of the roles of weather and density-dependence (led by Jonas Knape) in a large population dynamics database.

Statistical and computational methods

Many of the above projects open up needs to extend statistical methods and the computational tools to make them work. Recent projects include comparisons of different computational approaches for hierarchical Bayesian models in ecology (led by Lauren Ponisio), development of the nimbleEcology package (led by Ben Goldstein) with extensions to NIMBLE for ecological models, generalization of adaptive MCMC theory to select among alternative sampling methods (led by Dao Nguyen), and development of the nimbleSMC package for sequential Monte Carlo (SMC, aka particle filtering) in NIMBLE (led by Nick Michaud).

Other topics

A famous quote attributed to John Tukey is that statisticians “get to play in everyone’s backyard.” I’ve never quite related to that, but looking at some other recent projects, maybe there’s something to it. I’ve been involved with analysis of tropical tree chemical diversity, patterns of wildfires in relation to land management in California, analysis of eBird and iNaturalist data to see what engages people with birds, analysis of water quality trends in San Francisco Bay, and the question of rarefaction in microbiome analysis, among other interesting topics.