Figure 1.

Schematic of the experimental aquaria. All fish were tested at their home aquaria. The divider in the aquarium helped to separate the male and female cleaner during trials. It created a holding (1) and testing (2) compartment. In this example, the female cleaner (i.e. the smaller fish) is in the testing compartment, while the male cleaner (i.e. larger fish) is in the holding compartment. The paradigm in (a) shows an example of two plates, like in the reversal-learning and ‘client quality’ tasks, as well as in the treatment condition of the audience effect task (i.e. the control was a single plate). The paradigm in (b) shows an example of the detour task. The opaque and transparent separations are shown in (a), and also used for the detour task, helped to isolate the focal individual and prepare the test set-up before each trial.