A stage-based estimator of daily egg production

Snapper (Chrysophrys auratus) in South Australia spawn over a wide range of daily times and hatch in similar to 1.5 d. Traditional estimates of daily egg production (i) divide sampled eggs into daily age cohorts, e.g. days 1 and 2, (ii) compute cohort ages by assuming that all spawning occurs at a fixed daily peak hour, and (iii) link the estimation of spawning egg density with egg mortality rate by regressing cohort egg densities against cohort age. We propose a method for estimating daily egg production, the number of eggs spawned per unit area per day (P-0), without these assumptions. We use a range of estimates of snapper egg mortality rate obtained from prior studies to backcorrect for egg mortality to the age 0 spawning egg density for each stage of eggs individually rather than aggregating into daily cohorts. P-0 is estimated as a mean of sample tow densities rather than as a regression intercept. This stage-based P0 estimator avoids errors associated with assuming a fixed daily spawning hour and classifying eggs into day 1 or 2. It requires no regressions of sampled daily cohort egg densities against age, which often provide imprecise estimates of mortality rates. Simulation testing showed 1-2% accuracy for this estimator of P-0. The uncertainty of assuming egg mortality rates is moderated by the insensitivity of estimates of P-0 to the input value of egg mortality rate.