During the last 30 years, animal behaviorists have become serious players in the quest to understand the interaction between plants and their flower-visiting, foraging pollinators (Waddington 1983, 1997; Barth 1985). Flower-visiting bees, flies, butterflies, and beetles are the sole agents for reproduction in many species of plants. Through the larder of pollen and nectar they provide, plants also affect the foraging success and reproductive output of these insects. The pollinator and the plant, each of separate evolutionary lineages, are in a long-term game where each is dependent on the other and each affects the evolution of the other (Selten & Shmida 1991).
On a local scale, in a field of flowers, a forager such as a nectar-collecting bee makes thousands of sequential decisions during a foraging trip. These decisions are reflected in the choice of flowers visited. These decisions determine: which flowers receive visits and which do not; who mates with whom; the distance between mating plants; the transfer of intra- or interspecific pollen; and the amount of self-pollination and out-crossing. The decisions also affect the bee's success on its foraging foray. Through experience, the bee makes associations between different kinds of flowers (e.g., species) and the rewards they provide, and it seeks out the flowers with the greatest net rewards. Animal behaviorists have played an important role in learning how pollinators make these choices among flowers.
Although general patterns of pollinator foraging behavior have been found, variation among individual foragers has not been well studied. Individuals observed in foraging experiments often differ in their behavior even when given the same problem (e.g., choosing among flowers in the same patch; Waddington & Holden 1979). Inter-individual variation might be due to differences in experience, such that each forager has a slightly different experience in the same patch and makes decisions based on different information (see Thomson & Chittka, this volume). Sampling error could produce important differences in experience among individuals, especially if the sample of previous flower visits used to make future decisions is small. However, genetic differences among individuals may also contribute to behavioral variation. We need studies designed specifically to examine the underlying causes of variation among individuals in their assessment of information and choice behavior.
Most published work on the genetics of foraging behavior describes honeybees. Genotypic variability has been shown for many important traits (from Page et al. 1995): (1) the decision to forage for pollen or nectar (Calderone & Page 1992; Guzman-Novoa & Gary 1993); (2) behavioral plasticity associated with switching foraging resources (Fewell & Page 1993); (3) nectar and pollen load sizes (Milne et al. 1986; Deng 1996); (4) round-trip time and foraging activity (Guzman-Novoa & Gary 1993); (5) rate of foraging trip initiations (Deng 1996); (6) the age at which individuals initiate foraging (Calderone & Page 1988; Deng 1996), and the duration of the foraging career (Deng 1996). These constitute most components of individual foraging behavior, and some could affect the effectiveness, quantity, and quality of pollination.
Equipped with an understanding of how bees choose among flowers and an understanding of causes of individual variability in choice, we can better evaluate how plants may evolve to "manipulate" the bee's behavior so as to enhance their reproductive success. On this larger scale, both in time and space, the evolution of floral morphology and color, flowering phenology, and patterns of nectar and pollen presentation are likely influenced in no small way by the choice behavior of their floral visitors (Grant 1949; Macior 1970; Faegri & van der Pijl 1979). The salience of flowers in the pollinator's olfactory and visual fields, the spatial patterning of flowers, and the quality and quantity of the food are in part the result of selection for the reliable services of pollinators and the manipulation of their behavior.
In this chapter, I focus on the decision-making process of foraging bees with the goal of better understanding their choice behavior. This decision-making process includes the evaluation of relevant information used by the bees to make decisions which result in choice. Particularly relevant to their success on each foraging trip - and ultimately their fitness -
is information on the quality and quantity of their food. I will especially consider bees' evaluations of nectar concentration and pollen quality in making choices. How are nectar concentration and pollen quality evaluated, and what choices are made based on these evaluations.? I will also explore some studies of genetic variation for these evaluations and the choice of food. Because most work has been conducted on bumble bees and honeybees, genera of the same family Apidae, this chapter will reflect that bias.
Individuals just beginning a foraging career must learn which flowers are likely to provide the most pollen or nectar, and which flowers are not profitable to visit. During this early stage, colors, shapes, and odors of flowers are associated with profits; then this information is used to make decisions in the future. Foragers must decide where and when to look for food, and which food to search out, pollen or nectar. Perhaps before each takeoff from a flower, but certainly before each landing, a bee decides which flower to visit. Should it visit another flower on the same plant, or a more distant flower of the same color and odor (same species), or a flower of a different species? The outcome of these innumerable decisions, the amount of food collected, likely affects the bee's reproductive fitness. Recent studies have aimed at better understanding this decision-making process in order to understand the parameters that affect bees' choices. Both proximate (e.g., von Frisch 1967; Waddington & Holden 1979; Waddington 1983; Waddington & Gottlieb 1990; Greggers & Menzel 1993; Raveret-Richter & Waddington 1993; Shafir 1994) and ultimate (e.g., Possingham et al. 1990) causation of choice have been investigated.
Most studies of pollinators' flower choices address the relationship between gains and costs associated with foraging, and the pattern of choice behavior (Pyke 1984). The aim of these studies is to understand the direct relationship between objective information (the actual volume of nectar, the actual quality of pollen, the actual time to access the nectar inside the flower, etc.) and choice behavior. The implicit assumption made in these studies is that objective information is the direct guide to choice behavior. That is, choice behavior is guided directly by the actual (absolute) concentration or volume of nectar, the actual time to fly between flowers, the actual time to handle flowers, and so on. In fact, to do otherwise is assumed by most behavioral ecologists to be irrational behavior (Shafir 1994).
However, the cognitive psychologist will note that the direct guide to behavior is the subjective evaluation of the objective information (Rachlin 1989). Ideally, one would like to study subjective evaluations of objective information and their relationship to flower choice behavior. Studies done in my laboratory examine this relationship in the honeybee.
Subjective evaluations by honeybees, and genotypic variation - the dance
It is possible to study independently various stages of human beings' decision-making behavior. Choice behavior in the marketplace can be directly observed. The kind of investment instruments traded under various conditions can be determined. Furthermore, a person's subjective evaluation of the objective alternatives and other available information can be accessed independently. Questionnaires are used in the laboratory to determine these relationships. Thus, the relationships between the subjective evaluations of the options and the person's choice behavior can be understood. We do not have a well-developed protocol for tapping into non-human animals' evaluation of objective information. Recently, we developed a protocol for examining the relationships between objective costs of foraging, objective rates of energy intake, and intakes, and the subjective evaluation of these variables by honeybees.
Aspects of honeybees' dance are well known to be correlated with the distance and direction of food from the hive (von Frisch 1967). The bees use their perception of distance and direction to perform the dance in a certain way. I conducted experiments designed to find whether the honeybees' in-hive dance could be used to measure their evaluation of recently experienced foraging costs and gains (Fig. 3.1).
Von Frisch (1967) found that the "vigor" of the dance changed with the concentration of sugar solution imbibed. When bees had collected highly concentrated solution from a feeder, their dance was more excited than after collecting a weak solution. I expanded his studies to quantify "vigor" (Waddington 1982). Bees foraged back and forth between two artificial flowers that contained sugar solution (nectar). I manipulated gains by varying concentration and costs by varying the distance between the two flowers. The rate of the "round dance" increased as concentration
Decision making process
Was this article helpful?