Experiments examining potential impacts of growth hormone (GH) transgenesis in fish typically use a single source strain, and do not address potential differential impacts in strains of different genetic backgrounds. Here, we examine the effects of differing genetic backgrounds on the growth of transgenic and non-transgenic coho salmon produced by mating sires from different rivers with transgenic dams from a single origin. We found a significant difference in size between offspring of sires originating from various river systems in British Columbia. This difference was independent of differences between transgenotypes (i.e., transgenic vs. non-transgenic offspring). However, the effects of strain or sire were relatively small compared to the effects of the transgene, which were consistent regardless of sire origin. Thus, results derived from studies of GH transgenic fish from a single source pouplation could provide useful information for assessments of GH transgenic salmon from other systems. This has important implications for examining potential risks from introgression of a transgene into different populations.

Three experimental groups of coho salmon were established: Experiment 1 (Exp 1: sire effects) examined the effects of a wild-type sire on size of transgenic offspring.  Experiment 2 (Exp 2: strain effects) was used to examine the effect of sire length and river of origin (strain) on lengths and weights of transgenic offspring over a 1.5 year period. Experiment 3 (Exp 3: transgenotype) compared the effects of sire length and strain over a shorter timeframe on weights and lengths of both transgenic and non-transgenic offspring. Experiments were carried out under permitting from the Pacific Region Animal Care Committee (AUP#11-020 and 15-001) and following Canadian Council on Animal Care guidelines.

Exp 1: sire effects: Twenty –four wild-type males were collected from the Chehalis River, British Columbia, Canada in 2004; all males were crossed with a single homozygous growth hormone (GH) transgenic female from experimental line strain M77 (Devlin RH, Biagi CA, Yesaki TY. 2004. Growth, viability and genetic characteristics of GH transgenic coho salmon strains. Aquaculture 236:607–632. doi: 10.1016/j.aquaculture.2004.02.026) maintained at the Department of Fisheries and Oceans Canada’s Pacific Science Enterprise Centre in West Vancouver, BC. This transgenic line was derived from the Chehalis River, BC strain. All progeny resulting from this cross carried the transgene. Fertilized eggs were divided into 2 rearing replicates of approximately 50 eggs each in order to account for tank effects (replicates are termed A and B in the dataset). Replicates were randomly assigned to hatchery trays, and after hatching they were maintained in separate 5L buckets randomly assigned to the larger tanks; buckets were re-assigned at each sampling date. For all experiments, fish were fed to satiation 5 times per day using stage-appropriate commercial diets (Skretting Canada, Vancouver, BC); feeding was reduced to 3 times a day after 3.5 months. Weights and fork lengths of 25 randomly collected fish from each replicate were measured at ponding (April 6, 2004) and all fish from each cross were measured 1.5 months after ponding (June 1, 2004) and 3 months after ponding (July 19, 2004).

For Exp 2 and Exp 3, sires included wild male coho collected in 2015 from five river systems within British Columbia, Canada (Chehalis River (Ch), Kitimat River (K), Robertson Creek (R), Capilano River (Ca) and Big Qualicum River (BQ)). Fork length of sire was recorded. Eggs were collected from GH transgenic females from experimental line strain M77. Exp 2: strain effects involved crossing 115 individual wild-type males (N per strain: BQ =25; Ca=18; Ch=24; K=23; R=24) with females that were homozygous for the GH transgene. A random mix of eggs from 25 females was subdivided for fertilization by each male, resulting in 115 families in which all offspring were transgenic. In Exp 3: transgenotype the same 115 males were crossed with a mix of eggs from 9 females that were hemizygous for the transgene, resulting in 115 families that were a mix of transgenic and non-transgenic offspring; approximately 50% of the offspring from each family were transgenic. For all families, eggs and hatched-out alevin were incubated in Heath trays until yolk sacs were absorbed at which time fish were ponded into individual 5L floating buckets within 200L tanks (5 buckets per tank) with densities of up to 50 fish for Exp 2 and up to 100 fish for Exp 3. Fish were fed to satiation 5 times per day using stage-appropriate commercial diets (Skretting Canada, Vancouver, BC); feeding was reduced to 3 times a day after 3.5 months.

For Exp 2: strain effects, weights and fork lengths of 15 fish from each family, selected at random, were measured at ponding (March 30, 2015) and 30 fish from each family, selected at random, were measured 1.5 months after ponding (May 21, 2015). All fish from each family were measured 3.5 months after ponding (July 14, 2015) at which time 20 fish from each family were tagged using passive integrated transponder (PIT) tags for individual identification. PIT tagged fish were combined and moved to larger tanks (19,900L) where they were maintained at densities below 16g/m³. All fish were measured again at 6.5 months post ponding (October 17, 2015) at which time they were transferred to a saltwater mesocosm tank (350,000L) within the same facility. Subsequent samplings occurred 9.5 months (January 19, 2016), 12.5 months (April 19, 2016) and 15.5 months (July 13, 2016) after initial ponding.  At the time of PIT tagging it was noted that, due to high densities for some families in the smaller tanks up until that point, there was a relationship between the number of fish in the family and the average size of the fish in that family (i.e. fish from families that maintained close to the 50 fish that were ponded were smaller than those from families that had fewer fish due to mortalities since ponding).

For Exp 3: transgenotype, fish were measured 1.5 months after ponding (May 13, 2016) and 2.5 months after ponding (June 18, 2016); at both time points 30 fish were sampled at random from each family and weights (g) and lengths (cm) were measured. All sampled fish were tested for the presence of the transgene using a transgene-specific qPCR test (Fitzpatrick et al. 2011). At each sampling period, approximately 15 transgenic and 15 non-transgenic fish were sampled from each family.

Sire and offspring weights and length are reported here.

Table 1 contains size data for sires from each experiment. MaleID can be used to match males with offspring in other tables from this dataset.

Table 2 contains ffspring weights (g) and lengths (cm) from Exp 1: sire effects collected on 3 sampling dates. There are two replicates (Group A or B) for each sire.

Table 3 contains size data for Exp 2: strain effects. Offspring weights (g) and lengths (cm) collected over 7 sample dates. River of sire origin is given along with the sire ID for each family.

Table 4 contains ize data for Exp 3: transgenotype. Offspring weights (g) and lengths (cm) were collected on 2 sample dates. River of sire origin is given along with the sire ID for each family. Genotype indicates if offspring were transgenic (T) or non-transgenic (NT).

Table 5 contains the density of fish in tanks held as part of Exp 2: strain effects. Density refers to the number of fish in each tank prior sampling in July 2015 (max 50).