Table 2 Effect of both “species” (dwarf vs. normal) and “lake” on fish shape
Effect ofDFApprox FP-value
All lakesa
 Species1, 215161.25<0.001*
 Lake4, 43211.97<0.001*
 Sex1, 2152.610.076**
 Species × lake4, 4322.750.006*
 Species × sex1, 2150.370.691
 Lake × sex4, 4322.140.031*
 Species × lake × sex4, 4320.990.444
Within lakeb
 Cliff
  Species1, 2948.10<0.001*
  Sex1, 290.670.519
  Species × sex1, 290.790.465
 East
  Species1, 5540.13<0.001*
  Sex1, 551.830.171
   Species × sex1, 550.280.756
 Indian
  Species1, 3590.05<0.001*
  Sex1, 352.260.119
   Species × sex1, 351.860.171
 Témiscouata
  Species1, 4817.56<0.001*
  Sex1, 483.150.051+
   Species × sex1, 481.130.332
 Webster
  Species1, 4427.41<0.001*
  Sex1, 441.330.275
   Species × sex1, 441.180.317
  • MANOVA, multivariate analysis of variance.

  • a Effect of species identity (Dwarf vs. Normal), lake, sex, and their interaction on the two first axes of a PCA on 15 landmarks representing fish shape tested with a (MANOVA).

  • b Five independent MANOVAs (one for each lake) testing for the effect of species identity, sex and their interaction on the same response variables

  • * Significant result (P-value < 0.05).

  • ** Marginally significant result (P-value < 0.1)