Figure 7. Effect of spider size on honeybee behaviour (Experiment 2).

Proportion of honeybee visits to spider inflorescences vs spider size (A) considering all the trials conducted in the experiment and (B) considering only those trials that received more than six honeybee visits to the patch. Black circles represent spiders with the dorsal part of the abdomen painted on blue, grey triangles represent spiders with the forelimbs painted on blue and white circles represent control spiders. Solid lines represent fitted probability of landing on spider harbouring inflorescences. Although the relationship between probability of landing on spider inflorescences and spider size is not apparent in panel (A), the statistical model gives relatively little weight to trials with few honeybee visits removed to produce (B). Dotted lines represent the expected value if honeybees treated all inflorescences alike (p = 1/3).