U.S. Dept. of Commerce / NOAA / OAR / PMEL / Publications
One of the most challenging problems in marine research is the estimation of mortality rates of organisms. This is particularly the case with planktonic early life stages of marine fishes due to the broad spatial and temporal scales of their distribution, their movement with currents, other sampling difficulties, and nonconstant birth rates (i.e. often peaked spawning cycles). It is frequently inappropriate to assume a stable age distribution and constant mortality rates to justify use of catch curve analysis (Ricker 1975, Hewitt & Methot 1982). Knowledge of mortality rates of marine fish eggs and larvae is a key to understanding the recruitment process of marine fishes because, for many species, the magnitude of recruitment may be established during early life stages (Houde 1987, Sundby et al. 1989, van der Veer et al. 1990). Accurate measurement of the magnitude and variability of mortality rates in the field is required to assess the causes and processes of mortality, such as predation and nutrition. Data on interannual variation in mortality rates can also be instrumental in forecasting year-class success.
We describe a method used to estimate the daily mortality rates for larval fishes from a population characterized by highly aggregated spawning that occurs over a short period of time. Walleye pollock Theragra chalcogramma spawning in Shelikof Strait (Alaska) occurs largely within a 1 mo period (Kim & Kendall 1989, Kendall & Picquelle 1990, Picquelle & Megrey in press). Larvae live in the upper 100 m of the water column and are advected to the southwest for several weeks in the Alaska Coastal Current (ACC) (Hinckley et al. 1990, Kendall & Picquelle 1990) and larval distributions are relatively predictable in time and space.
We have identified cohorts of larval walleye pollock from their hatch-date distributions as determined from daily growth increments on otoliths and have sampled the population twice, ca 12 d apart, to estimate mortality rates for each cohort. Larval transport was estimated from movement of several main patches of cohorts, from the displacement of centroids of the entire larval distribution, and from a satellite-tracked drift buoy (drogued at 40 m, the depth of highest larval abundance) released in the center of one of the main larval concentrations. Advection of larvae across the boundaries of the sampling grid was estimated from a model of advection and diffusion.
Return to Abstract or go to next section