Courtship bouts of six founder-flush populations (two-pair founder-flush) a
nd two nonbottlenecked controls of the housefly were videotaped over the co
urse of 26 generations in order to evaluate the stability of mating behavio
ur. Limited-choice mate preference tests were conducted periodically to ass
ess levels of homogamic preference. Both founder-flush and control treatmen
ts showed significant evolutionary potential in courtship, along with homog
amic and heterogamic preferences. The founder-flush populations were signif
icantly differentiated from the controls in courtship repertoire, but all o
f the populations pursued convergent evolutionary trajectories in adapting
to the laboratory, resulting in dissolution of homogamic preferences. The p
henotypic shifts in courtship and mate preferences were unrelated to evolut
ionary trends in overall mating vigour; therefore, the convergence in court
ship could not be attributed to either a fitness meltdown due to inbreeding
or the purge of deleterious alleles. Only one founder-flush population sho
wed some independence from the selectional pressures for convergence; there
fore, the single two-pair founder-flush event was generally inadequate to s
timulate stable incipient speciation. This study thus demonstrates how conv
ergent evolution can dissolve founder-flush effects.