Quantitative Fish Biologist

The USGS Western Fisheries Research Center (WFRC) is seeking quantitatively minded individuals to fill vacancies on WFRC’s Quantitative Ecology Team. The Quantitative Ecology Team is a small group of statisticians and fish biologists working on a diverse array of quantitative ecological research questions. WFRC has a rich history in pioneering animal telemetry techniques with a particular focus on migrating juvenile salmon in regulated rivers. So naturally, much of our work has centered on developing and applying statistical models for telemetry data to estimate migration rates, migration routes, survival, and factors affecting these demographic parameters.

Desired expertise and experience include developing and applying frequentist and Bayesian statistical models such as mark-recapture models and integrated population models to fish or wildlife populations; animal movement models such as hidden Markov models or state-space models familiarity with time-series and/or spatial statistics; mechanistic simulation models such as agent based and matrix population models; demonstrated ability to work with large or complex datasets including familiarity with relational databases; and/or developing R packages and Shiny Apps.

Specifically, we are looking for someone with interest and capabilities to develop an R package for a class of space-for-time Bayesian capture-recapture models that jointly estimate migration, survival, and other demographic parameters of migrating animals which will likely use Stan for estimation.

The vacancy announcement is open until Nov. 8 and is posted on USAJOBS:
USAJOBS - Job Announcement Note: there are credit/degree requirements to qualify as a Fish Biologist under this announcement. If you are potentially interested in working with us on these types of problems, but do not think you would qualify as a Fish Biologist, please feel free to send me a direct message as we may have other opportunities in the near future.

More about the USGS Western Fisheries Research Center can be found here:
https://www.usgs.gov/centers/western-fi … rch-center

3 Likes

Cool, thanks for sharing. Hopefully this will also get some more eyes on your post:

2 Likes

These are super-fun models! Feel free to ping our forums if you get stuck on the package or need help somewhere with Stan. The Cormack-Jolly-Seber model is one of the first things Daniel Lee and I fit with Stan and I included it as an example in our User’s Guide chapter on latent discrete parameters. We were working with Fränzi Korner-Nievergelt from the Swiss Ornithological Institute.

I also spent a week hanging out with the ADMB group, which was largely about Pacific Northwest fishery models for NOAA. I’d suggest looping in someone from the ADMB side like Cole Monnahan (@monnahc), who has a lot of experience coding these models. Here’s a paper he wrote about marginalization for HMC:

Monnahan, Thorson, Branch. 2016. Faster estimation of Bayesian models in ecology using Hamiltonian Monte Carlo.

1 Like