Thesis (M.S.)--University of Wisconsin--Madison. / Typescript. Includes bibliographical references (leaves 77-78).

Thesis (M.S.)--University of Wisconsin--Madison. / Typescript. Title from title screen (viewed Sept. 24, 2007). Includes bibliographical references (leaves 77-78). Online version of the print original.

Thesis (M.Eng.)--Memorial University of Newfoundland, 2003. / Includes bibliographical references.

Thesis (Ph. D.)--University of Washington, 2000. / Vita. Includes bibliographical references.

This thesis explored the influence of fish behavior and distribution on the accuracy and precision of quantitative estimates of fish biomass and abundance as assessed by hydroacoustic techniques. Rainbow smelt (Osmerus mordax), a pelagic fish known to undertake diel vertical migrations and to exhibit changes in aggregation intensity associated with these movements was used as the model species for this study. The diel vertical migrations of smelt resulted in their potential inaccessibility to the acoustic gear. To obviate this problem a model of the diel migration of smelt based on observations of their behavioral responses to ambient light and water temperatures, was developed and used to time acoustic estimates of fish biomass and abundance to periods when they were unbiased by inaccessibility. Acoustic surveys were then conducted to evaluate the direct influence of fish aggregation on estimates of fish abundance, and to examine the effect of changes in fish distributions on the statistical validity of acoustic analyses. Comparative acoustic surveys, conducted when fish were schooled and dispersed, showed abundance was underestimated by up to 50% when schooling prevailed. The influence of changes in the level of fish patchiness, induced by diel schooling, on the statistical precision of acoustic estimates of abundance and biomass was found to be insignificant. Cluster sampling, a robust approach to the inherent problems of transect sampling created by autocorrelated data series was applied to acoustic data for the first time and its effectiveness was assessed. Cluster sampling yielded estimates of biomass and of abundance that were more precise than were estimates based on the traditional approach of analyzing complete transects.

No description available.