Figure 2
Changes in the strength of male preferences during the five audience treatments [(1) no audience (control), (2) audience male was presented throughout 2nd test period, (3) audience was presented only during habituation phase preceding the 2nd test period, (4) like (3) but audience male could interact sexually with previously preferred female, (5) like (3) but audience male could interact sexually with previously non-preferred female]. Shown are preference scores (percent time spent near initially preferred female during 2nd part – time spent near the same female during 1st part), such that negative values indicate that male preferences decreased during the 2nd part of the tests. The difference between the control treatment (1) and treatment (2) can be interpreted as audience effect (AE). P-values refer to post hoc pair-wise Fisher's PLSD tests; only significant values are given.