Use of the Daily Egg Production Method for Stock Assessment of Sardine, Sardinops sagax; Lessons Learned over a Decade of Application off Southern Australia

Abstract

Analyses of data collected over a decade off southern Australia confirm that estimates of spawning biomass of Sardinops sagax obtained using the daily egg production method are imprecise and suggest that if inappropriate analytical methods are used, estimates may also be biased. Spawning biomass estimates are most affected by variation in mean daily egg production (P0), spawning area, and spawning fraction. The log-linear mortality model (with one egg added to each day class of eggs at each positive site) should be used to estimate P0 because it fits strongly over-dispersed sardine egg density data better and provides more logically consistent and precautionary estimates of P0 than the exponential mortality model or generalized linear models. Most generalized linear models produced inflated estimates of P0 when egg data were strongly over-dispersed. The area surveyed should be sub-divided into a large number (e.g., 300) of similar sized grids. The Voronoi natural neighbor method should be used to calculate grid size for estimation of the spawning area because it reduces subjectivity in sub-division of the sampling area. Potential biases in spawning fraction can be minimized by calculating this parameter using data from all three stages of post-ovulatory follicles (combined). Priorities for future research are identified.