Should only females of giant freshwater prawn Macrobrachium rosenbergii be selected in genetic improvement programmes?

Sexual dimorphism is widely observed in almost all farmed aquatic animal species but giant freshwater pawn (GFP) is unique, with males characterized by three main morphotypes (blue claw, orange claw and small males) and females by different reproduction status (ovary, berried egg and already-spawned females). There has been reported evidence that the effect of male morphotype may have masked genetic variation in growth-related traits, as a result the heritability for male body weight was lower than that estimated in female. A pending question has arisen whether selection should be made in female only. To answer this question, we used an 8-year data set from a long-term selection programme (2008–2015) for high growth in this species comprising 106,756 individuals that were offspring of 515 sires and 810 dams. The body weight data of female and male GFP or of each morphotype was treated as a separate trait and a multi-trait approach was used to estimate genetic correlations for homologous traits between sexes and between morphotypes. Our analysis showed that there were little differences in the heritability estimates between female and male. In female, mature ovary individual displayed higher heritability than berried egg and already-spawned females. For male, the heritability for blue claw, orange claw and small males were 0.11, 0.06 and 0.00 respectively. Between-sex genetic correlation was moderate (0.55 ± 0.11) for body weight, suggesting that the trait expressions in female and male may be genetically different. In female, the genetic correlations for body weight among three female types were close to one (0.91–0.94). In contrast, the genetic correlations for body weight between male morphotypes especially between blue claw or orange claw and small males were low (0.15–0.25). Furthermore, we estimated genetic gain as the difference in least square means (LSM) or estimated breeding values (EBV) between the selection line and control group. The genetic gain in body weight was smaller in females than in males. It is concluded that there is no need to run separate breeding programme for female and male GFP. A combined selection using both female and male data can achieve selection response for body weight as demonstrated in the present study.