Several animals, including bees, use visual search to distinguish targets of interest and ignore distractors. While bee flower choice is well studied, we know relatively little about how they choose between multiple rewarding flowers in complex floral environments. Two important factors that could influence bee visual search for multiple flowers are the physical saliency (colour contrast against the background) of flowers and the reward value associated with them. We here investigated how these two different factors contribute to bee visual search. We trained bees to independently recognize two rewarding colours that in different experiments differed in either physical saliency, reward value or both. We then measured their choices and attention to these colours in the presence of distractors in a test without reinforcement. We found that bees preferred more salient or higher rewarding flowers and ignored distractors. When the high-reward flowers were less salient than the low-reward flowers, bees were nonetheless equally likely to choose high-reward flowers. Bees were more also more likely to attend to these high-reward flowers, with higher inspection times around them and faster search times when choosing them. When flowers differed in reward, we also found an effect of the training order with low-reward targets being more likely to be chosen if they had been encountered during the more immediate training session prior to the test. Our results parallel recent findings from humans demonstrating that reward value can attract attention even when targets are less salient and irrelevant to the current task.

中文翻译：

---

在大黄蜂视觉搜索多个奖励目标期间，奖励价值比身体显着性更重要

包括蜜蜂在内的几种动物都使用视觉搜索来区分感兴趣的目标并忽略干扰物。尽管对蜂花的选择进行了充分的研究，但我们对蜂花如何在复杂的花卉环境中选择多个奖励花之间的选择知之甚少。可能会影响蜜蜂对多朵花的视觉搜索的两个重要因素是花的物理显着性（与背景的颜色对比）以及与之相关的奖励价值。我们在这里调查了这两个不同的因素如何促进蜜蜂的视觉搜索。我们训练蜜蜂独立识别两种奖励颜色，这些颜色在不同实验中的显着性，奖励价值或两者均不同。然后，在没有加固的测试中，我们在存在干扰物的情况下测量了他们对这些颜色的选择和关注度。我们发现蜜蜂偏爱更显眼或更高奖励的花朵，而忽略了干扰因素。当高奖励花不如低奖励花显着时，蜜蜂同样有可能选择高奖励花。蜜蜂也更有可能参加这些高奖励的花朵活动，它们周围的检查时间更长，而选择它们时的搜索时间更快。当花朵的奖励有所不同时，我们还发现，如果在测试前更紧迫的训练过程中遇到低奖励目标，则更有可能选择低奖励目标。我们的结果与人类最近的发现相似，表明即使目标不那么突出且与当前任务无关，奖励价值也可以引起人们的注意。