ere routinely dehydrated in ethanol, equilibrated in xylene and embedded in paraffin according to standard histological procedures. Sections of 35 mm were stained with haematoxylin and eosin and blindly evaluated under a light microscope by a trained veterinarian. Morphology of PI and DI was evaluated according to the criteria previously described in Atlantic salmon: length of mucosal folds; the degree of vacuolization in the absorptive cells; width and cellularity of lamina propria and submucosa; frequency of goblet cells. The degree of histological changes was graded as normal, mild, moderate or severe. Statistics The data were statistically evaluated using JMP version 9.0 Statistical DiscoveryTM. Most of the results were subjected to two-way ANOVA with GM and SBM inclusion as the class variables and performed on calculated tank means or results from pooled samples based on a variable number of sampled individuals per tank depending on sampling time and parameter. However, the histological scores were compared using non-parametric contingency BIX-01294 custom synthesis analysis and performed on data from 12 fish per dietary treatment. All reported p-values are two-sided and significance was set at p,0.05. Trends/tendencies towards significance are discussed when pvalues were in the range of 0.050.10. Skeletal development examination Radiography of frozen fish from the day 99 sampling, 50 to 58 individuals per tank, was conducted in a semi-digital system, with a Giotto Image mammography X-ray source combined with an FCR Profect image plate reader and FCR Console. The system comprises automated 13130132 image enhancement-procedures for contrast enhancement and edge visualization. The skeletal morphology was evaluated by trained personnel according to the following criteria: fusions and fusion-associated changes; compressed vertebrae in neck; complex deformities, consisting of a mixture of fusions, compressions and distorted shapes; head deformities. Results Analyzed proximate compositions of the experimental diets were similar to expected compositions. The protein:energy ratios of the four diets were close to the predicted 25 g/MJ. SBM-containing diets were slightly lower in crude protein and lipid levels, and somewhat higher in carbohydrate, also as expected. Very low mortality was observed among the fish throughout the experiment with a cumulative mortality of #2.5%, indicating high quality feed and husbandry conditions. Interestingly and contrary to studies in more mature salmon, survival and performance of the juveniles was enhanced by addition of SBM to the experimental diets and no SBMinduced inflammatory changes in the distal intestine were Quantitative real-time PCR Real-time quantitative PCR analyses were limited to the DI, where possible interactions in SBM-fed Gene Reverse Effects of GM Bt-Maize in Diets for Juvenile Atlantic Salmon 17032903 Means and pooled standard errors were calculated from the means of three replicate tanks per treatment group, with measurements performed on 10 fish per tank. Initial body weight: 0.1760.01; initial body length: 2.560.1; initial condition factor: 1.0460.08. The p values are given for the main variables non-GM/GM and non-SBM/SBM inclusion, respectively, as well as p values for interactions between the variables by two-way ANOVA analysis. doi:10.1371/journal.pone.0099932.t003 Effects of GM Bt-Maize in Diets for Juvenile Atlantic Salmon observed. As this was an unexpected finding with implications for the interpretation of 
the results
