The Norm of Reaction is a curve
that relates, for a given genotype,
the contribution of environmental variation to observed phenotypic variation. The
shape of the curve may be essentially flat across environments [Left],
such that the expected
phenotype of any given genotype is highly predictable, independent of
the environment. In this model, the more narrow the distribution, the
less variable the trait. This is what is typically meant when a trait
is said
to be "genetic." In the first example, Genotype A always has a higher trait value than Genotype B, and this trait value is quite constant across environments.
Alternatively, the Norms of Reaction for two genotypes may have parallel slopes [Middle], such that relative rankings of trait values of Genotypes A & B are always the same (A is superior to B), but the absolute trait value is dependent on environment. In the second example, B has the higher trait value in either environment, but the trait value of B in the "X" environment exceeds that of A in the "Y" environment. The heritability of the trait would be close to 1.0 in either environment, but this does not predict the behavior in the other environment.
<> Finally, the shape of the Norm of Reaction for two different genotypes may be quite
different, such that it is not possible to predict from analysis of one
genotype the response of the another. Consider two breeds of cattle (genotypes
A & B).
Among cattle raised in a free-range, nutrient-poor environment
(environmental "Y"), those with genotype A always produce more milk fat (narrow red curve)
than do those with genotype B (broader
blue curve). The range of phenotypes within each
breed is relatively constant over a wide range of free-range
environments. A conventional study of the heritability of milkfat
production conducted in this environment will conclude that genes have
a strong influence on milkfat production (H ~ 1.0),
that is, the difference between breeds is "mostly genetic," as in the middle example.
On the other hand, when the same two
breeds are moved to a feed-lot environment with abundant nutrients
(environment "X"), both breeds show a marked improvement
in average milkfat production. Further, cattle with genotype B now typically produce richer milk
than those with genotype A (mean
of blue curve slightly higher), the reverse of the previous situation.
Both breeds also show a wide range of milk fat production (broad blue and red
curves), depending on the exact environmental conditions [e.g., feed
types]. A study of heritability in this environment will conclude that
genes have relatively low influence on milk fat production (low
heritability), which is mostly a consequence of environmental
variation.
Thus the relative importance of "genes"
and "environment" is not a unitary value, and may vary greatly
depending on exactly which environments the genes are expressed in.
Studies of heritability carried out in a single environment cannot
accurately estimate the Norm of Reaction, and often may not predict
phenotypic response in a different environment.