Norms of Reaction
Prof Dan
I think my explanation of Norms of Reaction using Figure 18-7 was a
bit confusing so let's see if I can clarify. This is a somewhat
complicated figure but here's the basic idea. What its trying to show
is that for a fixed genotype, a given environment will result in a
specific phenotype. Importantly, this is ignoring developmental
noise. Because there is no developmental noise and genotype is fixed
the only free variable left to explain phenotype is environment. This
enables us to draw a line relating environment and phenotype. This
much is hopefully intuitive. The reason to look at these norm of
reaction graphs is to get a feeling for how phenotypes vary across
environments with no other factors affecting phenotype. Check out
Figures 18-6 and 18-9 for norms of reaction from real data. Ideally
you would have more environments so the norms of reaction would look
more like curves but even the three data point graphs give you a feel
for what these things can look like.
Why they made Figure 18-7 more complicated with the distributions is
not clear to me. I think its not really worth worrying about but if
you are curious, here's my interpretation. My understanding is that
if you attempted to grow equal numbers of genetically identical plants
over a range of temperatures, you might get a distribution like the
blue bell-shaped curve over the x-axis, where plant survival was
highest at 20 degrees and dropped off to either side. So if you had
such a distribution, their point is that the distribution itself could
be converted into a phenotype (height) distribution (the yellow curve)
using the norm of reaction. And further they make the point that
because the norm of reaction is a curved line, it can convert a normal
looking bell-curve environmental distribution into a skewed phenotype
distribution (skewed meaning it is not symmetric). I sort of see why
this is interesting but its not presented very clearly and is somewhat
tangental to the main point that environment directly translates into
phenotype when genotype is fixed and developmental noise is absent.
Hope this helps.
Dan