A STOCK SURVEY TECHNIQUE USING THE TIME TO CAPTURE INDIVIDUAL FISH ONLONGLINES

Longline catches per unit of effort (CPUE) from research surveys are o ften assumed to vary in proportion to fish abundance. This assumption, however, may be invalid if the abundance of the target species is hig h enough to saturate the gear or if the abundance of nontarget species is high enough to exclude the target species from capture. We examine a new approach to surveying fish populations with longlines that is b ased on time-to-capture data measured with small, fish-activated timin g devices attached to every hook. A new measure of relative abundance, lambda, is developed that is immune to the effects of gear saturation and interspecific competition for hooks. Two estimators of lambda are compared by using capture-time data collected during summer 1987-1991 for pelagic armorhead (Pseudopentaceros wheeleri). Both estimators ar e shown to be unbiased if the underlying assumptions are true and fair ly robust to the observed departures from these assumptions.