Epistemologically, fitness is not a property of individuals, but of hereditary types. Individuals do not have a probability to die before maturity; they just die or survive. Moreover, the phrase that fitness is a property of types only makes sense under an implicit ceteris paribus condition, as the fitness of a type necessarily depends on the environment in which it lives. Hence, except in very restricted cases, fitness should be treated not as a single number, but as a function of two variables, type and environment.
What objects should be considered individuals, as well as the attribution of types, depends on the evolutionary question. About the only restriction is that adaptive evolution crucially depends on the almost faithfulness of reproduction, that is, offspring should nearly always have the same type as their parents or at least a very similar type. If this were not the case, the smearing out over type space would overwhelm any evolutionary dynamics due to selection. This also means that the appropriate reference frame for any definition of fitness is clonal reproduction. The connection with the Mendelian world is made by observing that alleles reproduce clonally.
Was this article helpful?