Effect of flower perceptibility on spatial-reward associative learning by bumble bees

The associative learning of spatial and reward information is generally considered an adaptive behavior of foraging animals that collect food from renewable resources. However, learning may not always maximize foraging efficiency if choosing a high-reward location accurately based on learned information imposes a time cost on the forager (speed-accuracy trade-off). To examine the effect of speed-accuracy trade-offs on associative learning, we observed bumble bees, Bombus ignitus (Smith), foraging in mixed arrays of high-and lowrewarding artificial flowers under two conditions, i.e., arrays of small flowers where bees could not easily detect the next nearest flowers on leaving a flower and arrays of large flowers where bees could easily recognize the next nearest flower. When flowers were small, bees created foraging routes by selectively incorporating the locations of high-rewarding flowers with their experience. When flowers were large, bees flew between flowers more quickly than when flowers were small, creating foraging routes without accounting for the locations of high-rewarding flowers. Estimated foraging efficiency was higher when flowers were large than when they were small, at least until flower visitation number reached 3000, suggesting that rapid foraging might be a better choice than accurate foraging when individuals are able to locate flowers easily. These results suggest that associative learning of spatial and reward information might be a choice that foragers can apply according to the cost-benefit balance of learning. Significance statement Previous studies focusing on spatial-reward associative learning in foraging animals assumed that foraging efficiency increased as the forager learned the locations of greater rewards. However, this study suggests that bumble bees learn the location of high-rewarding flowers depending on the cost-benefit balance of learning, irrespective of the opportunity or their aptitude for learning. Thus, learning might be a choice that foragers can apply according to the circumstances.