Table 3

The effect of treatment, type and their interaction on fish weight and gene expressions
Salmo salar L. n Effects
Fixed Random DFn DFd Sum Sq F P
Weight [log] 180 Treatment Treatment:Tank 1 28 0.62 23.19 <0.0001***
Type 2 27 6.06 113.11 <0.0001***
Gene n DFn DFd F P
EF1A A 177 Type Treatment:Tank 2 26 0.0002 1.17 0.3
HSP70 173 Treatment Treatment:Tank 1 26 0.0017 0.72 0.3
GR 173 Type Treatment:Tank 2 25 0.08 7.92 0.001***
Cu/Zn SOD 166 Type Treatment:Tank 2 24 0.06 3.49 0.04*
Mn SOD 170 Type Weight [log] 2 24 0.03 5.42 0.01**
GP 172 Type Weight [log] 2 24 0.07 6.79 0.003**
IGF-1 168 Treatment Weight [log] 1 25 0.09 13.57 <0.001***
Type 2 24 0.15 10.97 <0.001***
IGF-1R 173 Type Treatment:Tank 2 23 0.01 1.37 0.24

Summary of the best linear mixed effect models testing for difference in fish weight, and mRNA levels (raw Cq values of EF1AA, ΔΔCq values of target genes) between treatments and genetic origins (type). Including the interaction between treatment and type did not improve the fit of any of the models, thus all groups were affected by treatment in the same matter. The random effects were included by forward selection. Treatment:Tank; tank nested within treatment, Weight [log]; log transformed weight measurement of the fish. For more information on model selection based upon AIC values, see Additional file 1. The statistical significance is marked with asterisks, where * = P ≤ 0.05, ** = P ≤ 0.01 and *** = P ≤ 0.001.

Solberg et al.

Solberg et al. BMC Research Notes 2012 5:672   doi:10.1186/1756-0500-5-672

Open Data