Due to the high demand for cleanerfish in salmonid ocean farming operations, increasing lumpfish hatchery production and rearing efficiency are of great importance. Juvenile lumpfish are cannibalistic which is controlled, to some extent, though frequent size grading of the fish, however, cannibalism still occurs. Understanding and mitigating for factors that exacerbate aggressive behaviors in juvenile lumpfish, which can lead to cannibalism, would help achieve the goal of increasing juvenile production in the hatchery. We hypothesize that lumpfish cannibalism is linked to a specific ontogenetic period related to fish size and can be exacerbated by various stressors such as stocking density and photoperiod.
To test this hypothesis, we subjected two different size classes of juvenile lumpfish (5g and 11g) to varying densities (40 g/L, 65 g/L, or 90 g/L) under different photoperiod regimes (ambient, constant low light, or constant bright light) for an 8-week duration in winter 2022. Fish growth, survival, and aggression were measured biweekly, and stocking densities adjusted to baseline levels biweekly by removing any necessary fish.
For 5g lumpfish, overall percent growth ranged from 92.60 ± 8.51% to 170.65 ± 24.24%. Both stocking density (two-way ANOVA, P <0.01) and light (two-way ANOVA, P <0.001) significantly affected overall percent growth of 5g fish. Fish subjected to the 65 g/L treatment grew 31.5% faster than fish in the 40 g/L treatment (P < 0.01) and 22.9% faster than fish in the 90 g/L treatment (P < 0.05). 5g fish exposed to ambient lighting grew 41.8% faster than fish in the constant low light treatment (P < 0.01) and 31.3% faster than fish in the constant bright light treatment (P < 0.05). Unlike the 5g lumpfish, neither stocking density (two-way ANOVA, P = 0.53) nor light (two-way ANOVA, P = 0.154) significantly affected the overall percent growth of the 11g fish. For 11g lumpfish, overall percent growth ranged from 31.58 ± 0.73% to 43.92 ± 4.86%. Contrary to our hypothesis, neither variable significantly affected aggressive behaviors in either the 5g (two-way ANOVA, stocking density: P = 0.460, light: P = 0.065) or 11g fish (two-way ANOVA, stocking density: P = 0.412, light: P = 0.988). However, a trend towards higher fin nipping in smaller fish was observed (~45% fin nipping occurrence in 5g fish vs ~20% in 11g fish), indicating that cannibalism may be even greater when fish are < 5g but decreases as the fish grow.
Manipulating lighting and stocking density (up to 90 g/L) can be used to suppress or increase growth rates in small lumpfish, depending on a hatchery’s desired outcome, without resulting in an increase in fish aggression. However, these variables are less effective tools for controlling growth in larger juveniles. Future studies should focus on how these variables affect lumpfish
< 5g as there are indications that aggression is most severe at this size class.