AbstractsBusiness Management & Administration

The spatial dimensions of fisheries: improved use of spatial information into fisheries management and information for assessments

by Kotaro Ono

Institution: University of Washington
Degree: PhD
Year: 2015
Keywords: habitat; index of population abundance; multispecies fishery; spatio-temporal model; species distribution model; Fisheries and aquatic sciences
Record ID: 2058002
Full text PDF: http://hdl.handle.net/1773/27477


Until relatively recently, fisheries resources were managed as single homogeneous units and fisheries management conveniently ignored the presence of spatial heterogeneity in stocks. In nature, species are neither distributed at random nor uniformly, but follow some pattern created by a combination of habitat preference, movement, interspecific species interactions, and fishery exploitation. Ignoring such heterogeneity and managing stocks as single homogeneous units increases the risk of serial depletion and population collapse. It is therefore important to understand and consider the spatial dynamics of a fishery and fish populations to help establish proper management plans which aid the conservation of marine ecosystems. In this dissertation, I developed and tested novel fishery analysis methods which better integrate the spatial structure of fish and fishermen to help improve the quality of information used for fisheries management. In the first chapter, I developed a multispecies fishery model and examined how the spatial overlap of species could affect the biological and economic performance of commonly used fishery management methods such as the individual quota systems or marine protected areas. In Chapter 2, I examined how the use of spatial closures affected the analysis of fishery catch per unit effort (CPUE) data, which can be used to derive indices of population abundance. From this work, I proposed a new method based on data imputation to reduce bias in the indices of abundance. In Chapter 3, I presented an intuitive but objective way to define spatial strata (using a clustering approach) for analyzing CPUE data, and showed how this method can improve the accuracy of population abundance estimates. Finally, in the last chapter, I applied a novel spatiotemporal statistical model to three large geo-referenced time-series (habitat maps, survey-based fish density estimates and fishery catch data) to study the seasonal dynamics (between Summer and Fall) of a commercially important flatfish species, Pacific Dover sole (<italic>Microstomus pacificus</italic>), off the U.S. West Coast. The multispecies modeling confirmed that the spatial overlap of species affects both the biological and economic performance of a fishery management system in a complex manner. The presence of a spatial closure increased the amount of bias in the derived index of abundance, however the imputation based method was able to reduce bias for most cases. Furthermore, clustering based area stratification eliminated bias in the derived index of abundance due to selective fishing but was not able to reduce bias when fishing grounds were shifting over time. Finally, the spatio-temporal model applied to Pacific Dover sole revealed that seasonal dynamics (between Summer and Fall) were dominated by movement. In summary, the inclusion of spatial information in fishery models improved the accuracy of species distribution models, the accuracy of abundance indices used in stock assessment models, and provide insight to managers regarding what to…