Free flying honeybees were tested outdoors on blue–white and blue–yellow dimorphic artificial flower patches to examine the influence of reward difference, flower handling‐time difference and flower colour choice on foraging decisions. We employed different flower‐well depths to vary handling times (costs), and differences in sucrose molarity to vary reward quality. Tests were performed with 2 and 6 μl rewards to vary quantity. We show that when handling time is correlated with flower‐colour morphs on a pedicellate artificial flower patch, a honeybee's foraging behaviour is dependent on the flower colours used in the choice tests. This supports a honeybee foraging model where constraints are a significant factor in decision making. Bees visiting blue–yellow flower patches exhibited flower constancy to colour, where they restricted most visits to a single flower colour, some bees to blue and others to yellow, irrespective of handing time differences. When offered a choice of equally rewarding blue or white flowers, bees were not constrained by flower colour and chose to visit flowers with a lower handling time. When reward molarity varied with well depth between blue and white flowers, foragers chose shallow‐well flowers (short‐handling time) with a smaller net harvest rate over deep‐well flowers (long‐handling time) with a greater net harvest rate. Results using the blue–white dimorphic flower patch suggest that when foraging options simultaneously involve reward and handling‐time choices, honeybee forager behaviour is inconsistent with an absolute method of evaluating profit.

In exploring how foragers perceive rewards, we often find that well-motivated individuals are not too choosy and unmotivated individuals are unreliable and inconsistent. Nevertheless, when given a choice we see that individuals can clearly distinguish between rewards. Here we develop the logic of using responses to two-choice problems as a derivative function of perceived reward, and utilize this model to examine honey bee perception of nectar quality. Measuring the derivative allows us to deduce the perceived reward function. The derivative function of the perceived reward equation gives the rate of change of the reward perceived for each reward value. This approach depends on presenting free-flying foragers with a series of two different rewards presented simultaneously (i.e., two-choice, binomial tests). We also examine how honey bees integrate information from a range of reward qualities to formulate a functional response. Results suggest that honey bees overestimate higher quality rewards and that direct comparison is an important step in the integration of information from a range of rewards.