Recent estimates of the size at first maturity (L50) of Sardinella brasiliensis showed contradictory results with a decreasing in the fish stock biomass encompassed by increasing values of L50. The methodological approach used hereby allowed to separate sardines classified in the virginal maturity stage from those categorized in the recovery stage, and ready for one next spawning event. This study evaluated the hypothesis of the existence of separated stocks experiencing distinct environmental conditions and fishing pressures which may have altered L50 estimates using a robust dataset based on biological samples collected along the entire species distribution area in the southeast-south Brazilian coast [Rio de Janeiro (RJ), São Paulo (SP), Paraná (PR), Santa Catarina (SC) and Rio Grande do Sul (RS)] between 2000 and 2018. A reclassification of the gonadal maturity stages provided a more realistic estimate of L50. Combining biological, reproductive, fishing data and the mean temperature of the catch (MTC), the leave-one-out classification correctly re-assigned individuals with an overall accuracy of 85% [100% (RJ), 45% (SP), 99% (PR), 99% (SC) and 82% (RS)]. The connectivity between the local populations of S. brasiliensis off RJ (23°S) and the southern populations is limited, contrasting to spatial structured semi-discrete population-units found between SP and RS (24°S-30°S). The northern extreme population-unit (RJ, 22°S-23°S) showed an expressive reduction of L50, and a negative correlation was detected between the increasing MTC values and the abundance of early maturing individuals and recruits of the species. Stock specific L50 estimates seemed to act as indicators of long term environmental fluctuations.
Read full abstract