Animals are constantly faced with having to make behavioral decisions. An example is when they forage — searching for food. Bee behavior when questing for pollen and nectar has been a favorite subject for biologists to study. One reason is that rather than searching randomly for flowers, bees exhibit clear tendencies and preferences. Research has determined that bees use information about flower characteristics to make their foraging decisions.
Various aspects of bee foraging have been the subject of experiments. There are a number of factors that affect their decisions on where and when to forage. Bees typically have to choose between “Several dozen flower species which all differ in reward and signal, and they may encounter several flowers with different signals per second of flight.”1 Bees have been shown to be flexible, adjusting their decision-making behavior based on the conditions they encounter. One piece of information that bees use in foraging is the depth of the flower, which affects the time required to forage (called handling time). Flowers of shallow depth require less handling time, which can be viewed as a trade-off cost, and thus are more efficient. Other things being equal, bees generally prefer flowers of shallow depth with, therefore, reduced handling time. It has also been shown that bees prefer flowers that provide a greater reward (i.e., more nectar) if handling time is not a factor. Some experiments have shown that bees have a distinct preference for flowers with optimum concentrations of nectar.2,3 Bees also show a preference for continuing to forage on flowers of the same color, sometimes even when flowers of a different color offer superior rewards.
New Study from the University of Sheffield
A recent paper from the University of Sheffield in the UK studied foraging by honeybees (Apis mellifera), analyzing their decision-making process, including an assessment of their accuracy and efficiency. In documenting the experiment, the scientists describe some of the elements of bee decision-making:
We could suppose that decision to be influenced by what the bee can sense about the flower, her past experiences with that flower type, the context (is a predator nearby?), the state of the bee (does she already carry a full load of nectar and pollen?) and the state of her colony (what does the colony need?). Even this simple decision is a whole-brain activity involving sensory systems, memory systems, motor systems, and the bee’s subjective state.4
The experiment and analysis described in the paper applied signal detection theory, which originated in the development of radar systems during World War II. It has since been applied in many aspects of engineering, including communication systems. It has more recently been applied in the analysis of human decision-making psychology. The basic principle is related to the common condition that signals include the presence of noise. Radar and communication systems always have to account for the signal-to-noise ratio. That directly determines the probability of detection of the desired signal. In the case of human psychology, noise is represented by the amount of uncertainty in the information used in decision-making, and the probability of making a correct decision. It is also applicable to various aspects of animal visual and audible detection. Similarly, as it relates to animal behavior, signal detection theory “[p]rovides a framework for understanding and predicting how animals make decisions under uncertainty by modelling the relationship between the sensory information they received and their ability to accurately discriminate between stimuli.”5
Reward, Punishment, Probabilities
In the experiment, honeybees were trained to associate different colors with either a reward or punishment, and with a range of probabilities. The reward was a sugary liquid, while the punishment was a bitter quinine. For each trial the bees were offered one pair of colors with one color in the pair rewarded more often than the other during training. The “signal” in this case was the color associated with reward or punishment. Following the training, the bees were tested three times under different conditions. The decision for the bee was whether to “accept” the associated liquid, or simply reject it. The correct decision would be to accept rewards and reject punishments. The goal was to assess the ability of the bees to modify their decisions based on the likelihood of reward and punishment provided by the signals. Metrics that were assessed included the percentage of correct and incorrect decisions, and the time taken by the bees to make decisions.
Key findings from the experiment include that for tests which were intended to provide easy discrimination between reward and punishment, the bees made the correct choice significantly more often than mere chance would lead us to expect. For tests with more difficult discrimination, the bees made the same number of correct and incorrect decisions. Another finding was that the bee’s decision was influenced by both the likelihood of reward and the available evidence. Bees made correct acceptances faster than incorrect acceptances. Conversely, decisions based on reduced evidence were slower and less accurate. This is similar to experimental results with primates. This behavior is counterintuitive because theoretically, greater accuracy is achieved with longer decision times, as has been found in most experiments with humans and different animals: “For a given task difficulty, decisions are typically faster and less accurate when conditions favor speed, and are slower and more accurate when conditions favor accuracy.”6
The Risk of Predation
There likely is a logical reason why bees minimize their exposure time when foraging on flowers. Landing on the flower exposes bees to far greater predation risk. As explained in the paper, “Many bee predators, particularly mantids and spiders, have evolved as flower mimics and/or hide in vegetation close to flowers. A foraging bee feeding on a flower is therefore exposed to greater risks than a bee in flight.”7 For this reason, the more optimal behavior is to make a decision to land on the flower quickly when there is high confidence, but also to reject landing more often when there is uncertainty. Previous research with bumblebees supported this concept: it was found that the bees modified their search strategy by rejecting flowers at a higher rate when there was the potential of a lurking predator.8
Several elements make a bee’s decision-making behavior complex. The bee must 1) identify the flower’s color; 2) decide if there is a reward or punishment; 3) make an association between color and outcome; 4) store this association in memory; and, 5) use the stored information to make decisions about a subsequent condition encountered (i.e., is this a color that will reward or punish?). One might add that distinguishing a real flower from, say, a flower print on a woman’s dress can come into play, possibly requiring some experimental probing — attempting to land on the printed flower — to make the distinction. Beside the complexity of the information processing involved, such behavior illustrates the ability of bees to adjust in response to changing external conditions.
“Sophistication and Subtlety”
The authors indicate that the study, “Unveils the remarkable sophistication and subtlety of honeybee decision-making.” They also comment that the sophistication of honeybee decision-making has features in common with primates. That is all the more remarkable given the small size of their brains (less than 1 million neurons). These behaviors are largely controlled by a segment of the brain called mushroom bodies, which contain multisensory integration, learning and memory formation, and comprise about 40 percent of the brain neurons.9 In comparison, the brains of goldfish and hummingbirds are roughly 100 times larger. Despite the small size of the bee’s brain, expert Lars Chittka has documented the significant repertoire of honeybee behaviors, many that involve decision-making.10 Chittka comments, “These elegantly miniaturized brains are much more than input-output devices; they are biological prediction machines, exploring possibilities.”11 Despite a significant amount of research concerning these behaviors, including the study of bee brain neural networks, “The underlying mechanisms which drive these remarkable decision-making capabilities remain unclear.”12 The origin of the mechanisms also remains a mystery, that is, unless the hypothesis of intelligent design is considered.
Notes
- Peter Skorupski et al., “Visual Search and Decision Making in Bees: Time, Speed, and Accuracy,” International Journal of Comparative Psychology, 2006, 19, 342-357.
- Jonathan Cnaani, et al., “Flower Choice and Learning in Foraging Bumblebees: Variation in Nectar Volume and Concentration,” Ethology, 112 (2006), 278-285.
- DW Roubik, et al., “On optimal nectar foraging by some tropical bees (Hymenoptera: Apidae),” Apidologie, Springer Verlag (Germany), 1995, 26 (3), 197-211.
- HaDi MaBouDi, et al., “How honey bees make fast and accurate decisions,” eLife, 2023;12:e86176.
- MaBouDi, et al., “How honeybees make fast and accurate decisions.”
- Dominic Standage, et al., “On the neural implementation of the speed-accuracy trade-off,” Frontiers in Neuroscience, 13 August 2014.
- MaBouDi, et al., “How honeybees make fast and accurate decisions.”
- Thomas C. Ings and Lars Chittka, “Speed-Accuracy Tradeoffs and False Alarms in Bee Responses to Cryptic Predators,” Current Biology 18, October 14, 2008, 1520-1524.
- Claudia Groh and Wolfgang Rossler, “Analysis of Synaptic Microcircuits in the Mushroom Bodies of the Honeybee,” Insects, 7 January 2020.
- Lars Chittka and Jeremy Niven, “Are Bigger Brains Better?”, Current Biology 19, R995–R1008, November 17, 2009.
- Lars Chittka. The Mind of a Bee (Princeton, Princeton U. Press: 2022), 3.
- MaBouDi, et al., “How honey bees make fast and accurate decisions.”