Behavioural and genomic characteristics of selected farmed salmon families related to robustness, welfare and performance

The main aim of the project is to advance our knowledge and methods towards breeding of farmed salmon for robustness and performance traits. So far, selective breeding has been mainly based on production performance characteristics collected at the end of the production cycle and results from disease challenge tests (e.g. sea lice, PD). In BEHAVEGENES, individually PIT-tagged salmon were measured for weight and length five times during the production phase. Individual behavior and environmental preferences were scrutinized as potential breeding traits by using PIT-antennas at strategic positions in both tanks and sea cages. The largest and most comprehensive experiment included 6000 fish and 60 family groups from two major commercial breeding strains in a common garden sea cage. Heritability estimate of harvest size was high (as expected), while the correlation between growth from 4 to 11 months at sea and 11 months to harvest (16 months) was low and not significant. This finding is intriguing as it strongly suggests genetic variation for growth curves, which can be utilized in selection for specific production strategies, for example fast growth and lower feed utilization in defined periods of the life cycle. The registered swimming depth in the cages at sea show family differences, and that the preference for swimming depth per se was stronger than utilizing environmental gradients of temperature; the same families positioned themselves relatively deep/shallow both during the spring and autumn. As estimated over periods of stable environmental conditions, moderate to high estimates of heritability for the novel trait swimming depth was found. Fish size was correlated with the time of ascending towards surface during the spring (half a year after sea transfer), and growth towards harvest was at the family level correlated with swimming depth. The fact that family differences for swimming depth is detected, highlights the potential for selective breeding towards novel farming technologies and can aid in the selection of deeper swimming fish to lower the encounter rate with the infective stage of the salmon sea lice. Analysis of full genome sequencing of 360 individuals with growth and/or swimming depth contrasts, will be finished in 2018, and are expected to provide new insight into the genetic basis for the complex traits of growth and behaviour, and reveal the potential for use as markers in breeding programs. The second work package in BEHAVEGENES investigates epigenetic effects of variable environmental conditions during the freshwater stage on growth, behavior and stress responsiveness over the full production cycle. For the first time, cloned salmon were used, which made it possible to avoid effects of individual genetic variance and to study only epigenetic effects. From 3g size and throughout the freshwater production phase we exposed groups of salmon to four different treatments: Normal rearing environment (Control), daily draining of water (Consecutive stress), temperature fluctuations between 9 and 15°C in a 12:12h regime (Temperature Stress), or variation in water current velocity (Physical Exercise). Growth of the Consecutive stress group was reduced over the freshwater phase (16% lower than the Control), while a mild positive effect was found for the Exercise group, but without significant effects compared to the Control. At smoltification, all groups were transferred to a common garden sea cage for ongrowth until harvest. After 5 months at sea, all treatment groups were of similar average size, and all treatment groups were able to mount an adequate acute stress response to a standardized acute stressor, suggesting good welfare. Surprisingly, fish given Exercise showed a significant increase in mortality rate and variation of individual growth over the last year at sea. Compared with the Control, showing the highest average harvest weight (5.7 kg), the harvest weight of the groups exposed to Temperature stress, Consecutive stress or Exercise over the freshwater stage was respectively 3, 8 and 13% lower. Individual swimming depth, as recorded by PIT-antennas, was linked with treatment over the first half year at sea, where the Temperature stress and to a lesser degree, the Exercise group differed from the Control. In sum, the results of performance and behaviour highlight that the early life environment can impose long-term effects, and a need for increased focus on investigating treatment effects over extended periods in a full production perspective in studies aiming to reveal either environmental or molecular-genetic control of important production traits. Furthermore, high individual variation between genetically identical (cloned) individuals, in fact comparable to that seen in pedigreed family material, indicate high phenotypic plasticity clear epigenetic effects within treatments. Results on epigenetics effects on methylation of the genome will be finished during 2018.

The main aim of the project is to advance our understanding of the concept of robustness in farmed fish so that robustness traits can be implemented in on-going breeding regimes. We want to determine why some farmed salmon are more robust and perform bett er than others, and to determine if these performance related traits are heritable, and if we can find genomic markers for these. So far, selective breeding has been mainly based on production performance characteristics collected at the end of the produc tion cycle. We do, however, not know why the best performing families are successful. Both the loci and the molecular pathways responding to such positive selection are largely unknown. Until now, technologies to observe and identify a large number of i ndividuals under full scale farming conditions have been lacking, thus little is known regarding why, how, and when the best performing fish acquire their performance traits. This project is a collaboration between experts in the principles of animal bree ding and genomics, integrative (whole animal) biology, and fish behavior monitoring methods. The consortium holds excellent technical facilities and a large number of selected salmon families, as well an ongoing production of double haploids and cloned f ish groups. We will utilize a set of novel methods developed by the project consortium, including newly developed technology to collect frequent estimates of individual size and spatial position in tanks and cages and an automatic profiling system that co llects and processes water quality data. Heritability of a range of new traits will be calculated, and massive parallel sequencing to search for genomic markers for the performance-related phenotypic traits and for possible epigenetic effects will be perf ormed.