Contents lists available at ScienceDirect Aquaculture journal homepage: www.elsevier.com/locate/aquaculture Genotype by environment interaction across time for Nile tilapia, from juvenile to finishing stages, reared in different production systems Arthur F.A. Fernandes a , Érika R. Alvarenga b , Gabriel F.O. Alves b , Ludson G. Manduca b , Fábio L.B. Toral b , Bruno D. Valente c , Martinho A. Silva b , Guilherme J.M. Rosa a , Eduardo M. Turra b, ⁎ a University of Wisconsin-Madison, 1675, Observatory Dr., Madison, WI 53706, USA b Escola de Veterinária da Universidade Federal de Minas Gerais, Av. Antônio Carlos, no 6627, Caixa Postal 567, Campus da UFMG, CEP 30123-970. Belo Horizonte, MG, Brazil c Genus PLC, 100 Bluegrass Commons Blvd., Ste. 2200, Hendersonville, TN 37075, USA ARTICLE INFO Keywords: Random regression Selection optimization Breeding strategies Biofloc technology Oreochromis niloticus ABSTRACT Nile tilapia is one of the most important fish for aquaculture worldwide and it is produced under many different environmental and system conditions. Even though genotype by environment interaction (GEI) can arise due to many factors, most studies with Nile tilapia have focused more on the effects of water quality parameters and pound or cage systems. There is also lack of knowledge on GEI between fish at different ages, as before and after reaching sexual maturity. Therefore, the objective of the current work was to evaluate the magnitude of GEI on body weight in Nile tilapia raised in biofloc technology (BFT), cage and recirculation aquaculture systems (RAS) from 100 to 350 days of age. To evaluate the temporal trend of GEI a multi-trait random regression model (MTRRM) with age as random regression covariable was employed. Higher values of heritability estimates were found around 225 days, with a maximum of 0.4 for BW at RAS. Estimated genetic correlations between BFT and RAS were above 0.7 for almost any combination of ages evaluated. On the other hand, genetic correlations between Cage at ages below 150 days and either BFT or RAS at any other age were positive but lower than 0.6. Results for the estimated ratio of the indirect by direct response to selection were lower for selection performed at very younger ages whereas the desired response is for older ages. However, if the selection is performed around 225 days for either BTF or RAS this ratio is above 0.8 for most of the situations considered. In conclusion, MTRRM is shown to be a powerful statistical tool to assess changes across the time for genetic parameters of interest, such as covariance, correlations, and heritability. Also, the GEI across the three production systems considered was found to be dynamic across the ages evaluated, with stronger effects between Cage and both BFT and RAS if the selection is performed prior to 150 days of age. 1. Introduction Nile tilapia is one of the most widely farmed fish worldwide, with a production of 3.67 million tons in 2014 (FAO, 2016). It is produced in more than 100 countries under many different environmental condi- tions ranging from tropical to temperate climates. Therefore, many different production systems are used for tilapia production, with the most common being pond and cage systems. These systems, are highly affected by changes in climate and quality of water source but tend to be the most cost-efficient. Other systems of importance are recirculation aquaculture systems (RAS) and biofloc technology (BFT). RAS is known for its effective water reuse, control of water quality parameters, high biosecurity, and high stocking densities. Therefore, this technology is used worldwide in breeding programs, reproduction facilities or production of animals that have higher commercial value and higher requirements of water quality. Nonetheless, RAS has a high operation cost. Despite the eco- nomic restrictions, the interest in RAS has grown in recent years, due to the appeal of minimal environmental impact and system integration with other crops (Dalsgaard et al., 2013). BFT follows the same prin- ciple of a minimal environmental impact and higher control of water quality parameters and biosecurity as RAS but it has some advantages over the later. The BFT system is based on the mutual production of microbial community with the aquatic species of interest, without the https://doi.org/10.1016/j.aquaculture.2019.734429 Received 15 April 2019; Received in revised form 9 August 2019 ⁎ Corresponding author. E-mail address: eduardoturra@yahoo.com.br (E.M. Turra). Aquaculture 513 (2019) 734429 Available online 30 August 2019 0044-8486/ © 2019 Published by Elsevier B.V. T