Friday, September 26, 2008

Lab book club Fall 08/Spr 09

Fall Qtr reading group
Data Analysis Using Regression and Multilevel/Hierarchical Models

Sqr Qtr reading group
Ben Bolker's book

Yasmin gives talk on population consequences of maternal effects

At American Fisheries Society meeting in Ottawa in September:

Population Outcomes of Maternal Effects in Rockfish Depends on Life-history

http://yasmin.lucero.googlepages.com/talkAFS2008.pdf

Monday, August 4, 2008

Multivariate State-Space Models workshop now online

We did the workshop on Sunday, Aug 3, in Milwaukee and now have the course material, lectures and R case studies, posted online.

MSSM Workshop

During the workshop, we discovered a few typos in the case study write-ups which have not been corrected yet (EH: Aug 4).

Tuesday, July 29, 2008

Eric posted data-cloning MSSMs

A fair bit of Eric and my work over the past 6-months has involved using multivariate state-space models (MSSMs) to make inferences about population spatial structure. As part of this work, Eric worked on using data-cloning (Lele et al. 2007) for obtaining MLEs from MSSMs. Eric has posted for fitting MSSMs using data-cloning to FishBox.

Data-cloning MSSMs

This is "research" code, but should be useful to others trying to use data-cloning.

Brice has posted Bayesian MAR on FishBox

As part of our multi-pronged work on estimation methods for multivariate state-space models, Brice has posted a Bayesian version for fitting MARs to FishBox. This was motivated by a desire to put informative priors on the interaction matrix.

Bayesian MAR

ESA workshop on multivariate state-space models

On Sunday, Aug 10, Eric, Brice, Yasmin and myself are teaching a 1-day workshop on fitting multivariate state-space models (MSSMs) to data. This workshop uses a R routine that we wrote for fitting MSSM to multivariate time series data. A online version of the workshop will be posted after Sunday.

Friday, May 30, 2008

Kevin See joining the lab to work on spatial MARs

Kevin received a SeaGrant fellowship in Population Dynamics for his Ph.D. work starting in the fall. His proposal title was "Estimating population growth and interspecies interaction parameters with spatial replication" and will build on research on using MAR models to analyze fisheries monitoring data.

Friday, May 23, 2008

Brice leading reef survey in Monterey Bay

Next week, Brice leads a reef fish and invertebrate field survey as part of long-term monitoring program of Monterey Bay through REEF.org and National Marine Sanctuary Program.

Paper on CA sea lion population structure

Submitted to Journal of Applied Ecology, Applying multivariate-state-space models to detect spatial clustering of California sea lions in the Gulf of California, Mexico.
Chirakkal, Haridas; Ward, Eric; Gonzalez-Suarez, Manuela; Aurioles-Gamboa, David;
Holmes, Elizabeth; Gerber, Leah;

Wednesday, May 21, 2008

LAMBDA used in MAR-1 analysis of Lake Baikal data

Stephanie's paper using MAR-1 to analyze a 60-year time series of data from Lake Baikal just came out. It cites LAMBDA (the MAR-1) toolkit that Steve Viscido wrote (LAMBDA website).
  • STEPHANIE E. HAMPTON, LYUBOV R. IZMEST'EVA, MARIANNE V. MOORE, STEPHEN L. KATZ, BRIAN DENNIS, EUGENE A. SILOW (2008) Sixty years of environmental change in the world's largest freshwater lake - Lake Baikal, Siberia Global Change article
The study was also covered in the New York Times: the article

Fine-scale habitat selection of juvenile chinook salmon

Brice's paper on fine-scale habitat selection of juvenile chinook salmon is in press. This used as state-space model to analyze tracking data. Although state-space models have been used like this before, Brice's study was much bigger than previous ones and this required some numerical advances.
  • Semmens, B.X. 2008. Acoustically derived fine-scale behaviors of juvenile Chinook salmon associated with intertidal benthic habitats in an estuary. (In press, Can. J. Fish. Aquat. Sci.).

MixSIR Bayesian isotope mixing model

Brice's paper on a Bayesian Isotope mixing model just got published. Code for the model has been released on FishBox: http://fishbox.iugo-cafe.org/
  • Moore, J.W. and B.X. Semmens. 2008. Incorporating uncertainty and prior information into stable isotope mixing models. Ecology Letters.

Review of model selection approaches

Eric's paper reviewing and comparing different model-selection approaches is in press.
  • Ward, E.J. 2008. A review and comparison of four commonly used Bayesian and maximum likelihood model selection tools. Ecol. Model.

Killer whale senescence and prey limitation

Eric got his killer whale senescence and prey limitation paper provisionally accepted in Journal of Applied Ecology. This paper documents not only reproductive senescence but also that chinook abundance is correlated with lower calving probability, hence the prey limitation bit.
  • Ward, E. J., E. E. Holmes, and K. C. Balcomb. 2008. Evidence for reproductive senescence and prey limitation in killer whales.

Age-dependent maternal effects

Yasmin's paper on the age-dependent maternal effects is in press: Lucero, Y. Impact of an age-dependent maternal effect on time to recovery of an overfished population in Bulletin of Marine Sciences.

Methods for analyzing multi-site data

Hinrichsen, R. A. and E. E. Holmes. 2008. Using multivariate state-space models to study spatial structure and dynamics. Book chapter in Spatial Ecology coming out .... sometime.

Resolving the debate about when extinction is predictable

I recently collaborated with Steve Ellner (Cornell) and got some analytical results that show when -- what forecast lengths and quasi-extinction thresholds --
extinction estimates are unavoidably bad versus most definitely good. This hopefully, although maybe that is naive, will help clarify that it isn't that extinction is or isn't predictable. It is that certain types of predictions are unavoidably bad and you should avoid that type of prediction.

Accepted in Ecology Letters: Ellner, S. P. and E. E. Holmes. 2008.
Commentary on Holmes et al. (2007): resolving the debate about when extinction is predictable.