Abstract
An experimental grow-out protocol aimed at optimising Food Conversion Rate (FCR) was tested utilising juvenile mangrove snapper (Lutjanus argentimaculatus) and golden pompano (Trachinotus blochii) in an aquaculture facility in Mahe, Seychelles. The theoretical basis for using this protocol instead of feeding to satiation, the most common commercially used approach, is presented. The results of a modelling study with both species as subjects revealed that sacrificing some growth is a superior strategy to achieving a low FCR, rather than attempting to maximise growth, and overfeeding as a consequence. In a long-term growth trial fish were fed with a formulated feed suitable for culturing warm-water marine finfish (Biomar). Survival for mangrove snapper was 100 % during the trial and fish were successfully cultured from approximately 100 g to 1 kg during an 11-month period with an average FCR of 1.31 (SD 0.07). For golden pompano, survival was 90.3 % during the trial and fish were successfully cultured from approximately 50 g to 1 kg in 10 months with an average FCR of 1.52 (SD 0.03). These FCR values are lower than previously recorded in the literature. Growth models for both species were also constructed with the grow-out data using the machine learning module Scikit-learn.