A program of the Southwest Fisheries Science Center’s Fisheries Ecology Division.
The Ecosystem Forecasting Team develops robust tools for managing ecosystems, complex dynamical systems which are incompletely observed. There are always species that are not counted, differences among individuals within species that are not tracked, or changes in gene frequencies which are not measured, all of which contribute to ecosystem function.
Instead of filtering observations through assumptions, we try to let the observations of the ecosystem and how it has responded to previous management actions indicate where the system is going next. We use this information to help develop sustainable policies for conservation and management.
This involves developing new approaches to modeling complex systems, as well as making use of recent developments in other fields like nonlinear dynamics, physics, and machine learning. These tools often outperform more traditional approaches to ecological modeling. We have applied them to predicting recruitment in harvested fish populations, estimating state-dependent species interactions, and understanding synchrony and asynchrony in marine metapopulations.
Current research areas include Bayesian nonlinear forecasting, spatio-temporal delay embedding, and approximate dynamic programming.
Therkildsen, Nina O., Aryn P. Wilder, David O. Conover, Stephan B. Munch, Hannes Baumann, and Stephen R. Palumbi.
2019. Contrasting genomic shifts underlie parallel phenotypic evolution in response to fishing. Science 365(6452):487-490.
Munch, Stephan B., Alfredo Giron-Nava, and George Sugihara.
2018. Nonlinear dynamics and noise in fisheries recruitment: A global meta-analysis. Fish and Fisheries 19(6):964-973.
Deyle, Ethan R., Robert M. May, Stephan B. Munch, and George Sugihara.
2016. Tracking and forecasting ecosystem interactions in real time. Proceedings of the Royal Society B 283(1822):20152258 (9 p.).
Boettiger, Carl, Marc Mangel, and Stephan Munch.
2015. Avoiding tipping points in fisheries management through Gaussian process dynamic programming. Proceedings of the Royal Society B 282(1801): art. 20141631 (9 p.).
Perretti, Charles T., Stephan B. Munch, and George Sugihara.
2013. Model-free forecasting outperforms the correct mechanistic model for simulated and experimental data. Proceedings of the National Academy of Sciences of the United States of America 110(13):5253-5257.
Salinas, Santiago, and Stephan B. Munch.
2012. Thermal legacies: transgenerational effects of temperature on growth in a vertebrate. Ecology Letters 15(2):159-163.
Team Leader: Steve Munch