Click image to enlargeAtlantic cod and corals in the Gulf of Maine. Credit: NOAA/Stellwagen Bank National Marine Sanctuary.
Lucey is using Rpath to model how fish biomass migrates between four ecological production units: Scotian shelf, Gulf of Maine, Georges Bank, and mid-Atlantic bight.
Visual representation of four harvest strategy (HS) simulation results using Rpath. Colors range from red, where a harvest strategy caused fish stock to crash, to green, where a harvest strategy caused fish stock to flourish
August 24, 2017
Updated September 7, 2017
Contact: Heather Soulen
Leveling Up: Improving EBFM with Rpath, an R Implementation of Ecopath with Ecosim (EwE)
Scientists from the Northeast Fisheries Science Center (NEFSC) and others helped organize and facilitate a 2017 American Fisheries Society (AFS) symposium on advancing fisheries management strategy evaluation (MSE). Sean Lucey, fisheries biologist at NEFSC, presented his collaborative work on linking multiple Ecopath models using Rpath at the symposium. This feature story is an extension of his presentation for AFS participants and those involved in fisheries science and natural resource management.
Reproducibility, the ability to repeat another scientist’s research and get the same results, is at the core of science. It gives scientists confidence in research findings and allows them to build off of each other’s work, advancing both science and society. Scientists at the Northeast Fisheries Science Center (NEFSC) and Alaska Fisheries Science Center (AFSC) are making reproducibility, transparency, and full reporting easier for ecosystem scientists with Rpath, an R implementation of Ecopath with Ecosim (EwE).
Building a new open source tool
About 20 years ago Sarah Gaichas and Kerim Aydin, fisheries scientists at NOAA’s Alaska Fisheries Science Center began working on ecosystem model simulations to address management questions. They started with the Windows-based ecosystem simulation software EwE, but found it limiting saying, “the problem was that the Ecosim graphical user interface is a kind of black box where you can’t effectively control what simulations were going on.” To gain that control, Aydin first developed an in-house version in Visual Basic, and from there they moved it into C++, a powerful and fairly technical computer programming language. There were complications with the C++ version. “The way it was designed worked for us, but was kind of awful to try to share with other scientists. It had all these other scripts [commands] to do graphics and it was just gigantic and unwieldy.” Desiring a more flexible and familiar computing framework, the team saw the potential of R and soon began working with Sean Lucey, fisheries biologist at NEFSC to develop Rpath. Rpath builds on the open source nature of Ecopath by allowing users more customization, as well as cross-platform support (Windows, Linux, Macs). Rpath helps solve the reproducibility problem with ecosystem models by containing all steps in a single file or script. Lucey and the team re-coded the source code from EwE using a combination of R and C++, allowing users to build and run their models, analyze outputs, and create publication quality graphics without switching programs. “R is much stronger for reproducibility because everything is captured in a script so you can see what changes you’ve made and you can annotate why you made those changes. Those aren’t things you can necessarily do in the EwE graphic user interface (GUI) where you’re just clicking buttons,” says Lucey. This re-coding and development of Rpath was partially funded by NOAA's Integrated Ecosystem Assessment, a program supporting Ecosystem-Based Management by providing the tools that help transfer scientific knowledge to management.
Connecting the dots
With Gaichas now at NEFSC, the team is currently customizing ecosystem simulation models to address management questions by improving capabilities like modeling different fisheries simultaneously. Rpath allows the team to see how a scallop management action might affect the yellowtail flounder fishery. “You can see these other connections that you might not see if you’re focused on a single species approach,” says Lucey. Another new area the team is exploring with Rpath is testing place-based management strategies. “What ecosystem-based fisheries management (EBFM) looks to do is be more place-based, meaning we’re managing Georges Bank, not Georges Bank cod. We’re managing Georges Bank based on many of the characteristics of Georges Bank like predator-prey interactions and the production of the area - primary production or prey fields - to make a more holistic view of what’s going on in that area rather than by species,” says Lucey.
Fish and many other marine species move around and with their movement go a whole host of factors that can affect the function and sustainability of a region. “One of the things this particular project does is linking these ecological production units – regions -- together. Fish don’t care about where you draw lines. So the point is to build a model where you can have biomass leave Georges Bank and go to the Gulf of Maine and vice versa. You need to be able to account for that in the model, especially if you want to test single-species strategies against ecosystem strategies because if you just test for a single-species strategy, you might be doing it for its whole range rather than just one region. You need to account for that biomass at all times,” says Lucey. “That’s the big trick that I’m doing right now, to allow the biomass to flow using migration terms in this model.”
Biomass transfers between regions will be relatively small given the annual time step of the model, but it allows the team to see creeps related to global warming over longer periods of time. Distribution shifts result when species move and inhabit areas that they either historically haven’t been found in or haven’t been as abundant in. Recent research by NEFSC scientists and collaborators highlight the projected northward migration of lobster and other species related to thermal habitat. While Lucey can’t directly input environmental factors like water temperature at this time, he can input an attractiveness rating for how “attractive” (e.g., lots of food, great habitat, cooler water temperatures, etc.) or “repulsive” (e.g., little food, little habitat, warmer water temperatures, etc.) a place would be for marine species. This rating will allow the team to see the immigration (movement in) and emigration (movement out) of biomass not related to the production and mortality of marine species. “I can’t think of a single Ecopath model that uses immigration and emigration terms because it’s hard,” Gaichas says.
The future looks bright
This use of linking multiple Ecopath models may not be limited to just a few discrete oceanic regions. Lucey says, “You could actually theoretically link say a mid-Atlantic Ecopath model to a Chesapeake Bay Ecopath model and now you can answer questions about connectivity from freshwater all the way to the marine system and that’s one thing that people are interested in doing.”
The future for the team and for fisheries science is looking bright with a research tool like Rpath. While exploring new territory is difficult, it’s also very exciting. “What Sean is doing is really new. People have built models at different spatial scales before, but using those terms to actually move the biomass between them in a rigorous way using what we know about species migration and then seeing what difference that makes, I don’t think it’s been done,” says Gaichas. “It’s really cool stuff.”
Rpath is not only being used by Lucey in the Northeast; it's also being used by originator Kerim Aydin in Alaska. In fact, other scientists across the U.S. and internationally are using it for their EBFM needs. Try it today! The Rpath package is publicly available on Lucey’s GitHub account.
# # #
NOAA Fisheries Service is dedicated to protecting and preserving our nation's living marine resources and their habitat through scientific research, management and enforcement. NOAA Fisheries Service provides effective stewardship of these resources for the benefit of the nation, supporting coastal communities that depend upon them, and helping to provide safe and healthy seafood to consumers and recreational opportunities for the American public. Join NOAA Fisheries on Facebook, Twitter and our other social media channels.
NOAA's mission is to understand and predict changes in the Earth's environment, from the depths of the ocean to the surface of the sun, and to conserve and manage our coastal and marine resources. Join us on Twitter, Facebook, Instagram, and our other social media channels.