Biology Reference
In-Depth Information
FIGURE 12.2
Norms of reaction for two genotypes reared in two environments. The environmental factor is
plotted on the
x
-axis, the phenotype on the
y
-axis. (A) The phenotypes of the two genotypes reared in the same
environment differ; both are invariant across the range of environments. (B) The phenotypes of the two genotypes
reared in the same environment differ; both respond to the environment to the same degree and in the same
direction; the norms of reaction are parallel. (C) The phenotypes of the two genotypes reared in the same environ-
ment differ and one responds to the environment of rearing whereas the other does not. (D) The phenotypes of
the two genotypes reared in the same environment differ; both respond to the environment but in contrasting
directions so that their norms of reaction cross.
case, both factors and the interaction term would be statistically significant as well because
the impact of the environment depends on the genotype. We would also see a significant
interaction term if both genotypes exhibit plasticity but to varying degrees or in different
directions. A fourth possibility is that both genotypes exhibit plasticity but in contrasting
directions
their norms of reaction cross each other (
Figure 12.2D
). In this instance, the
statistical analysis would reveal no impact of either genotype or environment because the
lines here cross at the mean, but the interaction term would be significant. When there are
more than two genotypes or environments, the analysis of all groups is usually supple-
mented by pairwise comparisons to determine which means differ from which (and by
how much).
Norms of reaction are more difficult to depict for multivariate data because those norms
of reaction no longer describe a change in a single dimension. When the phenotype has
only one dimension, its value on that dimension can only increase or decrease so every
phenotype can be plotted on the same phenotype axis. But when the phenotype is
