Sperm irradiation protocol
The sperm irradiation protocol was developed in 2011 from the protocol for cod sperm , using the same equipment, the recommended dilution (1:40) and the same germicidal UV lamp (254 nm, 15 W, 220 V, 50 Hz). Milt was diluted with milt fluid (milt from several males were centrifuged until clear and the clear milt fluid was frozen and stored at -20 oC and thawed before the experiment). 15 mL aliquots of the diluted milt was placed in a 9 cm petridish surrounded by ice and placed on top of a magnetic stirrer and irradiated at 0.48 mWcm-2 (VLX-3.W radiometer, Cole Parmer, USA). The source-filter to sample distance was maintained at 20 cm throughout the experiments. Optimal irradiation dose (50% activity compared to diluted unirradiated sperm [see 28], was found to be 6-8 mins and cold freshwater was used to activate the sperm.
Optimising the mitotic gynogenesis protocol
All experiments were done with eggs from the domesticated and commercially available Aquagen strain, Aqua Gen AS, Trondheim, Norway. On 14 December 2011 (Exp2011), 4 ml of milt from one male salmon was diluted with 160 ml milt fluid (1:40). Twelve 15 ml aliquots of the diluted milt were irradiated with UV light for 6 or 8 mins in a 9 cm petridish at 0.48 mWcm-2 and transferred to 25 ml polyethylene (PE) containers and stored refrigerated and in darkness until fertilization. One control group was made with the diluted unirradiated milt. Each of the thirteen sperm aliquots were used to fertilize groups of 1000 salmon eggs which were left to hydrate in 0.5 L PE bottles at 8 oC until pressure treatment. At 3798 (3800), 4000 (4000), 4193 (4200), 4403 (4400), 4605 (4600) and 4816 (4800) minutedegrees (minC) (minutes*degree Celsius) the PE bottles were transferred to the pressure chamber and the eggs were subjected to a hydrostatic pressure of 655 bar (TRC-APV, Aqua Pressure Vessel, TRC Hydraulics inc., Dieppe, Canada) for 5 mins (the simplified group name is shown in brackets). Also, one batch of eggs fertilized with unirradiated milt were sampled every 10 mins from 540 mins (4320 minC) to determine the first cleavage interval (FCI). The samples (10-15 eggs) were cleared in 10% acetic acid before inspection.
On 12 December 2012 (Exp2012), 6 aliquots of diluted (1:40) sperm were made. Five of these were irradiated (0.45 mWcm-2) and stored as described above. The last aliquot was left as an unirradiated control. Five egg groups of ~1500 (between 1239 and 1788) salmon eggs were fertilized with UV irradiated sperm (6 mins) and one group was fertilized with the control sperm. Hydration and pressure treatment was according to Exp2011, but treatments were done at 4713 (4700), 4800 (4800), 4899 (4900), 5011 (5000) and 5123 (5100) minC in addition to the control fertilization.
Fish management and rearing
Eggs from Exp2011 were incubated at approximately 6 oC. From 23 April 2012 surviving larvae from the different groups were fed at 12 oC in individual square grey, covered, fibreglass tanks (1×1×0.25 m) and the temperature was switched to natural temperature in June 2012. All tanks were fed a commercial salmon feed in excess (Nutra Olympic, Skretting AS, Averøy, Norway) with automatic feeders (ARVO-TEC T Drum 2000, Arvotec, Huutokoski, Finland). Feed was given in small portions through the continuous light photoperiod. For illumination, two 18 W fluorescent daylight tubes (OSRAM L 18W/840 LUMILUX, OSRAM GmbH, Ausburg, Germany) were used to produce 960 lx under water in the centre of the tank. Photoperiod and feeding were controlled automatically by a PC operated system (Normatic AS, Norfjordeid, Norway). On 7 August 2012 the surviving fish were PIT tagged and a small fin-clip was taken for genotyping. After tagging and sampling the experimental fish (including 45 control fish) were transferred to three tanks (1.5x1.5x0.7m) for further ongrowing in common garden conditions.
Body mass and fork length were collected at eight time points; Aug 2012, Nov 2012, Jun 2013 (transfer to SW), Nov 2013, Jul 2014, Jun 2015 and Nov 2016. At each sampling time, fish were anaesthetized in 100 mgL-1 Finquel® (MS 222). In Nov 2012, some stunted individuals were euthanized together with deformed fish (mainly curvatures and shortening of the vertebral column; see reduction in numbers in Additional file 1: Table S1; Figure 1). Thirty-four control fish were also euthanized. In June 2013, 30 fish from the treatment groups and the remaining 11 controls were radiographed and checked for vertebral deformities. The controls were euthanized during the sampling. When the fish were measured in November 2013 and July 2014 fish were also euthanized (30 fish each time) to get samples for several other studies. Euthanasia was always done in 500 mgL-1 Finquel® (MS 222) followed by exsanguination.
Fish were sexed by visual examination of the gonads at the time of terminal sampling. The condition factor (K) was calculated as K = 100 x weight (g) x length-3 (cm). Specific growth rate (SGR, % per day) was calculated from the formula: SGR = (eq − 1) × 100 , where q = [In(W2) − In(W1)] (t2-t1)-1  and where W2 and W1 were the live body weights at times t2 and t1, respectively. Levels of sexual maturation, based on external morphology, were assessed throughout.
Rearing of the fish produced in Exp2012 was the same as in Exp2011. Surviving larvae were first fed from 22 March. On 18 September 2013 the surviving progeny (240 fish from treatment groups and 30 controls) were tagged with passive integrated transponder (PIT) tags and tissue sampled. From Exp2012 only the survival data and genotyping are presented here.
Production of clonal lines
In 2014, six females from Exp2011 were sexually mature. On 4 December 2014 they were all ovulated and their eggs were hand stripped and fertilized with three aliquots of diluted and UV irradiated sperm (same protocol as described above with 6 mins irradiation at 454 mW). Eggs were left to hydrate in 2L PE bottles at 8 °C until pressure treatment. At 300 minC (second meiotic division ) the PE bottles containing eggs were transferred to the pressure chamber and they were thereafter pressurized for 5 mins at 655 bar. The weight of the females after stripping, the weight of the drained egg mass and the weight of 100 eggs, were recorded.
Rearing conditions for the fish in the clonal lines were the same as for the production of the double haploids described above. Survival to the eyed stage was registered on 10 February 2015 and surviving larvae were counted and first fed from 9 April 2015. On 8 June 2015 (day 60 of feeding), surviving fish were counted, distributed to other experiments and a small number were reared on for sampling and phenotypic description. On 16 November 2015, 8-10 fish from each clonal line were euthanized (500 mgL-1 Finquel® (MS 222) and fin tissue samples were taken for microsatellite DNA analysis. On 30 November 2015, the remaining fish were euthanized, a blood sample taken for ploidy determination, length, weight, sex and deformities in the vertebral column were recorded. Deformities recorded included curvatures (lordosis, scoliosis and kyphosis) and/or shortening (short trunk and short tail . Deformities in the head skeleton (upper and lower jaw and opercula) were recorded and as the pelvic fin was found to be missing in some individuals, this was also recorded.
DNA was extracted from fin-clips. This was performed in 96-well plates using a commercially available extraction kit (Qiagen DNeasy®96 Blood & Tissue Kit). Each 96-well plate included two blank wells as negative controls. The samples were subject to genotyping with a set of 18 microsatellites that are routinely used in the molecular genetics laboratory at the Institute of Marine Research for Atlantic salmon genetics projects including ploidy determination [e.g. 33-35]. The samples taken from Exp2011 and 2012 were analyzed with all 18 of the microsatellites and the clonal fish produced in 2014 were analyzed with 16 of these. These loci were amplified in three multiplexes, using standard protocols for fresh tissues (Additional file 3); SSsp3016 (Genbank no. AY372820), SSsp2210, SSspG7, SSsp2201, SSsp1605, SSsp2216 , Ssa197, Ssa171, Ssa202 , SsaD157, SsaD486, SsaD144 , Ssa289, Ssa14 , SsaF43 , SsaOsl85 , MHC I  and MHC II . Polymerase chain reaction (PCR) products were analysed on an ABI 3730 Genetic Analyser and sized by a 500LIZ™ size-standard. The raw data was checked manually twice.
Data were analysed in GraphPad Prism, version 6.0. Significance was assigned at p ≤ 0.05. Length, weight, and K factor data were first checked for normality within group using the Shapiro-Wilk test. Subsequently, parametric data were analysed using one-way ANOVA whereas non-parametric data were analysed using the Kruskal-Wallis test with treatment (3-6 levels depending on the timepoint) as a categorical variable. When main effects were significant, we used Tukey’s or Dunn’s multiple comparisons tests for parametric and non-parametric data, respectively. Each time point was analysed separately.