Does size really matter? Selection, constraints and allometry

The appearance of a species depends on how body parts scale with body weight. Despite much variation in scaling across species (e.g. wallaby and wombat limbs), within each species scaling relationships are precise: body size accurately predicts trait size. Is this due to natural selection against deviant growth patterns or developmental and genetic processes that canalization trait growth and constrain the evolution of scaling relationships? To quantify the importance of selection I will artificially select for new scaling relationships and then compare the fitness of males with relatively smaller and larger genitalia than occur naturally. Genitalia are sexually selected, making them strong predictors of evolutionary fitness.