Working Group: Ecological Implications of Alternative Fishing

Strategies for Apex Predators


First Meeting - November 20-21, 1998



Meeting Participants: Robert Olson, IATTC, La Jolla CA (Convenor)

Kerim Aydin, University of Washington, Seattle WA

Chris Boggs, NMFS, SWFSC, Honolulu HI (absent due to illness)

Marco García, IATTC, La Jolla CA

James Kitchell, University of Wisconsin, Madison WI

George Watters, IATTC, La Jolla CA



Many of the fisheries that harvest marine productivity target apex predators, potentially imparting a top-down effect on marine ecosystems. Unintended catches of non-target species, often discarded at sea, can be substantial. Calls have been issued for a more holistic, ecological approach to fisheries management, taking greater account of species interactions and their dependence on underlying ecosystem dynamics. A broad-based public consensus, reflected in worldwide legislation and agreements, has emerged that fisheries bycatch should be minimized or eliminated.


In the epipelagic zone of the eastern Pacific Ocean (EPO), the top levels of the food web are occupied by large tunas, dolphins, sharks, billfishes, and other large animals. The purse-seine fishery in the EPO is directed primarily at yellowfin tuna, but substantial catches of other apex predators and of juvenile tunas also occur. The species composition and magnitude of the bycatches and discards generated by three alternative fishing strategies employed by the fishery vary substantially. The ecological costs of fishing are complex, due to the size-dependence of predator-prey interactions and the changing dynamics of the fishery. The EPO surface fishery has enjoyed high rates of production and recruitment of yellowfin tuna over the past decade, and, unlike many pelagic systems, the predator-prey interactions are increasingly well documented. Yet, food web responses to removing tunas and associated predators are not understood. The member governments of the Inter-American Tropical Tuna Commission (IATTC) have issued a call to foster an ecological approach to assess the impact of bycatch and to provide an ecological basis for limiting or reducing it.



The general goal of the "Fishing Strategies" Working Group is to develop modeling tools intended to set alternative fishing practices in an ecosystem context for the tuna fishery in the tropical EPO. We are employing a version of the Ecopath modeling approach developed by Jeffrey Polovina (NMFS, Honolulu) in 1984 and now widely applied through the efforts of Villy Christensen (DANIDA) and Daniel Pauly (University of British Columbia). Ecopath creates a mass balance of trophic exchanges for entire ecosystems based on estimates of biomass for many species and trophic levels, and uses principles of energetics and trophic transfer. The steady-state Ecopath model for the tropical EPO will provide input to a dynamic version, Ecosim, created by Carl Walters (University of British Columbia). Ecosim incorporates the equivalent of population dynamics into the higher trophic levels and the consequent feedback for all trophic levels. Walters continues to improve the utility of Ecosim. Density-dependent compensatory responses and spatially-explicit effects have been incorporated in recent versions, motivated in a large part by Jim Kitchell’s NCEAS-sponsored workshops, Apex Predators in Marine Systems, held in Santa Barbara during 1997.


Meeting I.

The nominal goal for the first meeting of the Working Group was the culmination of the construction process of the first version of the tropical EPO model. Kerim Aydin set the stage by providing a thorough review of the Ecopath formulation. Bob Olson and George Watters presented first-cut estimates for any three of the four parameters, biomass, production/biomass (P/B), consumption/biomass (Q/B), and ecotrophic efficiency (EE), for each component of the model. The model is organized around 34 state variables with high precision at the top trophic levels and low precision at the bottom. The parameters were categorized by degrees of accuracy; uncertainty analysis will follow during a later phase of the project. Most of the remainder of the meeting was occupied by iteratively adjusting the model components and parameter values to achieve mass balance. Only parameters that had been estimated with a low degree of accuracy were changed. The problems encountered were related primarily to aggregation of components within trophic levels. Specifically, intra-guild predation, i.e. cannibalism, created problems in balancing the model. This was resolved by disaggregating where appropriate. The balanced Ecopath model was imported into Ecosim, and several preliminary simulations were run satisfactorily. Thus, we are ahead of schedule for the project (see below).


Marco García provided an overview of the spatial heterogeneity in the distribution and species composition of the bycatch in EPO purse-seine fishery. The data are collected by observers aboard the fishing vessels. Patchiness exists in the spatial distribution of the observations and the animals. Future consideration will be given to the spatial scale of the tropical EPO model.


The Future

The activities of this Working Group will move through three phases over two years. We are in the first phase, focused on the development, construction, and evaluation of an Ecopath model for the tropical EPO. The timetable calls for completing the tropical EPO Ecosim version by the second meeting (February 25-27, 1999), but this was accomplished ahead of schedule during Meeting I. Meeting II will comprise a larger, more diverse group focused on evaluating the tropical EPO model via simulations of Ecosim. Several of the invited participants participated in previous NCEAS-sponsored workshops on apex predators in marine systems. Others are knowledgeable about the systems involved, familiar with the fisheries issues, or experienced in working with the models. A manuscript will be drafted after Meeting II.


The second phase of the project involves rigorous analysis of the tropical EPO model. Following model revisions based on the February 1999 meeting, we will develop a means for bootstrap computer runs of the model and Monte Carlo simulations to provide estimates of uncertainty about the model outputs. Meeting III, which is scheduled for November 1999, will involve a few members of the core group working on NCEAS computers. The second meeting of Phase 2, Meeting IV, will be held to conduct a rigorous evaluation of uncertainty in the Ecosim model through collaboration with importantly-placed scientists in key agencies.


Phase 3 involves using the tropical EPO Ecosim model for discussions in the policy realm. In November 2000 the IATTC will sponsor a workshop that will bring together representatives from conservation groups and agency representatives from IATTC member countries where concerns about bycatch and sustainability are central to management policy.


Submitted 21 January 1999

Robert J. Olson