Longevity: Selection

It is one thing to select on coat color in mammals or bristle number in insects, but another to select on a functional character that depends on many distinct biochemical pathways. As dog breeding shows, it is possible to select stringently on relatively arbitrary features, like the color or oiliness of fur. But dog breeding also shows that selection programs that use a small number of breeders run into problems of inbreeding depression: infertility, structural defects, blindness, deafness, and so on. This is not a special problem to the breeding of dogs or even mammals. Breeding insects for increased fecundity using small selected groups usually runs into problems of inbreeding depression, such as inviability, reduced male fertility, and reduced longevity.

A further problem arises when selection breeds for reductions in functional characters. Sexual outbreeding populations carry numerous recessive deleterious alleles at very low frequencies. When there is no deliberate breeding, natural selection keeps these alleles rare, and they play little role in the physiology or development of most organisms in the population. Human populations are good examples of this. There are many human genetic diseases caused by recessive deleterious alleles: cystic fibrosis, Tay-Sach's disease, and so on. But most people do not develop ailments due to such genes. They die instead of commonplace diseases, both contagious and degenerative, that do not depend on such genes. But populations that undergo selection for reductions in functional characters will show rises in the frequencies of alleles that are generally deleterious, alleles that do not play an important role in shaping these characters for most individuals in the population.

For this reason, it is unlikely that selection for reduced longevity or selection using small population sizes will reveal the genetics of longevity in the vast majority of individuals within a population. Thus, for example, the 1920s studies of longevity by Raymond Pearl using Drosophila mutants are of no value for uncovering the genetics of longevity in fruit flies that have not been inbred or mutated. Longevity is not a well-defined character like eye color, which can be analyzed using mutations of large effect or selective screening for pathological extremes. Longevity is not unique in this respect. The same problem applies to such characters as fecundity and developmental speed.

This makes the appropriate experimental design for selection on longevity some type of mass selection for increased longevity. Unfortunately, this experimental design will usually be very difficult to sustain, because it takes a long time to collect longevity data and it has to be done every generation in a selection experiment. Furthermore, if longevity is defined as the age at death, selection requires the collection of progeny from the selected group before that group has actually died. Many of these problems can be overcome by the use of laborious experimental designs, but these are rarely used in selection studies.

Fortunately, there are experimental designs for selection on longevity that evade many of these problems, yet produce healthy animals with greater mean and maximum longevities. We turn to these designs now.