Indirect models for the evolution of female preference assume that pre
ference does not evolve as a result of direct selection but rather thr
ough genetic correlations with positively selected male traits. An ess
ential assumption of Fisherian models and many recent good genes model
s is that assortative mating causes a genetic correlation to develop b
etween alleles conferring high preference and alleles conferring high
levels of the male attractive phenotype. In a direct test of indirect
selection models, mass selection on attractive male coloration did not
result in a correlated response in female preference in replicate tre
atments from a high predation population of the Trinidad guppy. The re
sponse was consistently low and a power analysis showed that genetic d
rift was not likely to explain the low response across all replicates.
These results contrast with studies demonstrating such a correlation
in another population of guppies and in three-spined sticklebacks.