I am reading Evolution in Changing Environments by R. Levin but have a hard time.
At page 25, Levin says as a conclusion
The more efficient a caste at meeting the contingency for which it is specialized, the lower its abundance. This is seen in Figure 2.5b
Here is figure 2.5b It is actually a reprodcution I did in R but ressemble closely the original
W1 is the fitness in environment 1 and W2 is the fitness in environment 2. The figures comes with the following legend:
The optimal population consists of two castes in the proportions of the line segments by the Adaptive Function (dotted line). Thus castes C1 and C2 are more or less equally common, but a population of C1 and the more specialized C3 will be mostly C1.
The terms "Adaptive Function" is not really defined in the book but it basically mean the overall fitness given the fitness components W1 and W2. The terms "castes" and "contingencies" aren't accurately defined either. In his model, the population experiences only these two environments, where $p$ is the probability to be found in one specific environment. As I understand little from this book, it is hard to know what kind of a prior information I could/should give you.
Can you help me to understand the above statement and why it makes sense?