9409369 Wilkinson A fundamental question of population biology is how does selection alter suites of multiple traits. It has long been know that selection on one trait can cause correlated changes in traits not under selection. These correlated changes can be predicted over the short term from the among trait genetic variances and covariances statistics estimated from a breeding selection study. However, to make predictions beyond a few generations ones has to determine how the genetic covariances themselves change. In this study the investigators propose to investigate the influences of selection on the genetic covariances among a suite of 11 morphological and life history traits in eight species and eight selected lines of stalk-eyed flies. These flies provide a model system for studying this problem because eye span dimorphism has arisen independently in at least four genera. In monomorphic species males and females of equivalent body size have similar eye spans while in dimorphic species male eye span exceeds body length. By estimating genetic covariances from breeding studies and comparing they with respect to phylogenetic relationships obtained from DNA sequencing information that will be able to determine the rate of direction of genetic covariance change when a trait is under strong directional selection. These correlational studies will be corroborated by comparing genetic covariances among lines that have experienced directional selection for over 20 generations in a lab. This work will produce three major results: 1) an improved understanding of the phylogenetic relationships among intriguing group of flies from Africa and south-east Asia, some of which are agricultural pests; 2) new statistical methods for comparing covariance matrices which are more robust, make fewer assumptions and are less computationally demanding than the current methods; 3) the most comprehensive analysis yet performed relating to whether or not genetic correlations act as a constraint on or a con sequence of natural selection.
