Two questions were addressed: (1) What is the genetic variance–covariance structure of a suite of
four female life history traits in D. melanogaster? and (2) Does the genetic architecture of these
traits differ among populations? Three populations of D. melanogaster were studied. Genetic
variances and covariances were estimated by sib analysis three times for each population:
immediately upon establishment of populations in the laboratory, and subsequently after
approximately 6 months and 2 years of laboratory culture. Entire genetic variance–covariance
matrices, as well as their individual components, were compared between populations by means of
likelihood ratio tests. All traits studied were significantly heritable in at least one-half of estimates.
Despite large sample sizes, additive genetic covariances were for the most part not statistically
significant, and only two significant negative covariance estimates were obtained throughout the
experiments. Therefore, these experiments provide little support for evolutionary life history
theories that are based on negative genetic correlations among life history components. Neither do
they support the idea that genetic variance for fitness components is maintained by trade-offs.
Evidence suggests that the G matrix of one population was initially different from those of the
other two populations. Those differences disappeared after 2 years of laboratory culture. At the
level of individual (co)variance components, there were relatively few differences among
populations, and the overall impression was that the three populations had generally similar
genetic architectures for the traits studied.