A study I conducted suggested that female mallards chose males based o
n natural variation in individual ornaments, especially the bill. Orna
ment manipulations were used to test experimentally whether females we
re choosing males based on ornament appearance. Four experiments were
conducted: anterior shaving (removed the outer colored part of the hea
d, neck and breast feathers), wing patch clipping, bill blackening and
bill yellowing. Anterior shaving caused 15 of 16 males to have lower
pairing success. Wing patch clipping did not affect pairing success. B
ill blackening caused a significant decrease in rank pairing success,
and bill yellowing caused a weaker, but not significant, decrease in s
uccess. Furthermore, the latter three manipulations had no significant
effect on male dominance rank. Taken together with the findings of th
e natural variation study, this work shows that female preferences dir
ectly select for male ornamentation.