Triploid Atlantic salmon growth is negatively affected by communal ploidy rearing during seawater grow-out in tanks

• Triploid salmon growth is negatively affected by communal ploidy rearing at sea
• Malformation in both ploidy appears to be a function of faster growth rate
• Genotype influences prevalence of deformity

Using sterile triploid salmon is of interest to fish farmers as a means to mitigate interbreeding between farmed and wild fish, prevent pre-harvest maturation and offer potential for faster growth. We investigated within the same experiment whether growth, deformity and cataract prevalence in diploid and triploid Atlantic salmon post-smolts (8 full-sib families) were influenced by isolated- or mixed-ploidy rearing during 12 months of sea water grow-out in tanks until harvest. Diploids attained significantly higher harvest weights than triploids in both treatments (+ 6.7% isolated, + 26.3% mixed). Triploid harvest weight was 29.6% lower under mixed ploidy rearing than their isolated triploid siblings. A similar effect was also observed in diploids (12% lower harvest weight). Family had a significant effect on harvest weight in both ploidy. However, family harvest weight in triploids showed a linear relationship between isolated and mixed rearing (r2 = 0.76), while a non-linear relationship was found between diploids reared in isolation or mixed populations (r2 = 0.88). Furthermore diploid family weight correlated positively with triploid weight in isolation (r2 = 0.65), but a non-significant relationship was observed under mixed rearing (r2 = 0.49). Diploids had a significantly higher condition (K) factor at harvest than triploids in both treatments, while triploid K factor was significantly reduced by mixed rearing with diploids. Triploids had a significantly higher incidence of cataract than diploids in both treatments (60.9–77.3% vs. 21.7–47.2%, severity ≥ 1). However, severity of cataract was greater in both triploids (2.43) and diploids (1.44) grown in isolation than those grown in a mixed ploidy population (triploid 2.17; diploid 1.09). Triploids had a significantly higher prevalence of visible deformity than diploids (18.1 vs. 7.2%), however, triploids grown in isolation had significantly higher prevalence than triploids under mixed rearing (30.4 vs. 5.8%). Collectively, the results provide clear evidence that triploids perform very differently when reared in the presence or absence of diploid conspecifics. Furthermore, observations of reduced deformity and cataract in the slower growing mixed populations support the hypothesis that such malformations are in part a function of higher growth rate. The results of this study also indicate that care should be taken when designing conventional common garden experiments to evaluate family performance between ploidy as interactions between ploidy may influence growth traits, although family effects appear to remain consistent.