Faculty Projects and Publications

John R. Sibert

Manager, Pelagic Fisheries Research Program

Lab: (808) 956-4109
Fax: (808) 956-4104
Email: sibert@hawaii.edu
Pelagic Fisheries Research Program, JIMAR
University of Hawai‘i at Manoa
1000 Pope Road, MSB 313
Honolulu, HI 96822

Education

Ph.D. 1968, Columbia University

Research Interests

Although fish are mobile, the explicit effects of mobility on fisheries management policies are usually neglected. In cases where fishing is distributed uniformly over the range of the exploited fish species, movement may not be relevant in models of population dynamics. For tunas, however, fishing is certainly not uniform and is often restricted to a fraction of the range of the exploited species. Therefore, neglecting movement may lead to errors in estimates of exploitation rates.
Tuna movements have been the subject of many field studies that depend on a variety of tagging and tracking techniques, and extensive collections of tag release and recapture data are maintained by various organizations dedicated to the scientific analysis of tuna populations. These data have yielded important insights into the exploitation and population dynamics of tunas.
Quantitative analyses of tuna movement have been slower to develop, with treatment often confined to drawing arrows on maps to represent the long-range movement of a few tagged individuals. In fact, several general classes of models can be applied to the quantitative analysis of fish movement. Bulk transfer models, where exchange rates between large regions are characterized by transfer coefficients, were first developed in the 1950s. Models of this type have been applied to yellowfin tuna (Thunnus albacares) in the eastern Pacific, to skipjack tuna (Katsuwonus pelamis) in the western Pacific, and to southern bluefin tuna (T. maccoyii ). In some cases, these models have closed form solutions, are easily implemented in computer code, and have parameters that can be estimated statistically from tagging data. Although bulk transfer models can predict changes in population size in arbitrary regions, they cannot be used to predict the changes in population density at an arbitrary point within a region because they are not continuous in space.
Models based on diffusion concepts are continuous in both space and time. These models have a long history in animal ecology and their potential application to fisheries population modeling also dates back to 1950s. Although the diffusion concept suggests that the population moves at random, it does not require that individual fish move randomly. The small-scale movements of individuals, which are surely non-random, may, in a large population, produce a net distribution that gives the appearance of being the result of random movements. Because purely diffusive movement will ultimately produce a uniform distribution of the population at equilibrium, other factors such as directional movement or spatial variability in recruitment, mortality, or population growth rate are required to maintain persistent gradients in population density.
Directional movements are easily incorporated into a diffusion model by introducing advective terms. The advection-diffusion equation can be viewed as the limiting form of a biased random walk and equivalent to the individual-based modeling approach in the same sense that Eulerian and Lagrangian approaches are complementary views of fluid movement. The application of advection-diffusion models to fish population dynamics has grown recently. Current model development extends previous work by developing flexible parameterization movement and the use maximum likelihood to estimate model parameters from tagging data. These models are currently being applied to skipjack, yellowfin, and bigeye (T. obesus) tagging data from the central and western Pacific and from the waters around Hawaii.

Publications

Recent publications: 1994 to 1999 (Web Page)
All publications (MS Word document)

[ Top of the page ]