Rapid inference of larval fish recruitment potential from size spectrum models

Abstract

The ratio of mortality to growth of larval fish provides a metric of a cohort’s ‘recruitment potential’. Estimating recruitment potential is arduous, requiring growth and mortality to be estimated independently. Here, we propose using the exponent of size spectrum models to indicate recruitment potential from body measurement data alone. This approach has several advantages including 1) reducing data collection times, 2) removing uncertainty in estimates generated from age estimation, and 3) allowing re-analysis of larval fish databases or archived collections. To test the validity of this approach, we conduct simulations comparing estimates of recruitment potential from an abundance spectrum model with other common methods. By varying larval flux rates, growth, mortality, and measurement error, we show the abundance spectrum model is more accurate and precise at smaller sample sizes, and more robust to variance in individual rates and measurement error of ages, but more susceptible to measurement error of size. We highlight that a size-based approach to estimating recruitment potential provides another useful tool for understanding larval survival, reducing resource demands on research compared to traditional methods.