It should not be surprising that beetles, flies, and lepidopterans can learn, as learning ability of one sort or another has been found in virtually all animals tested (Alloway 1973). However, the kind of learning (e.g., habituation, sensitization, associative learning), the extent to which an insect depends on learned rather than innate responses, and the behavioral context or sensory modality in which learning is expressed, vary between taxa, probably as a result of selection (Papaj & Prokopy 1989). Stephens (1993) argues that learning ability should evolve under environmental conditions of intermediate predictability - that is, in situations in which the environment is too unpredictable within one or a few generations for fixed behavior patterns to be favored, but not so unpredictable that the individual cannot behaviorally track changes. Floral resources do vary in this way, and indeed, most flower-visiting insects that have been investigated have been shown to be capable learners.
I know of no investigation of beetle learning in relation to flower-visitation behavior. It is not clear whether this truly reflects beetles' inability to learn in the context of flower visiting, or is due to a lack of experimental investigation or non-reporting of negative results. Plotkin (1979) reports that a predaceous ground beetle could learn to reduce a strong thigmotactic response in an open field and hence locate a centrally placed water source. Alloway (1973) summarized other reports of spatial maze learning in larval and adult grain beetles (Tenebrionidae).
Various fly taxa possess efficient capacities for associative learning of visual and olfactory stimuli. Sheep blowflies rapidly learn to associate color with reward (Fukushi 1989), and although hoverflies could be trained to land on colored artificial flowers (Kugler 1950), they could not be trained to extend their proboscides towards a rewarded color, responding instead to their innately preferred yellow (Lunau 1992). Olfactory conditioning has been reported for drosophilid, muscid, and tephritid flies (Fukushi 1973; Spatz et al. 1974; Prokopy et al. 1982). Additionally, some flies can, like honeybees, learn and make use of spatial landmarks. Male hoverflies hover stably in mid-air and, using visual cues, return to approximately the same position in space after chases (Collett & Land
Lepidopterans can associatively learn a number of different stimuli in a range of behavioral contexts, including oviposition, nectar foraging, and perhaps navigation. In the context of oviposition, moths and butterflies can learn to associate odors or tastes with leaf shape (Papaj 1986) or with color (Traynier 1986). Ovipositing females preferentially select host-plants with which they have had experience, and consequently make fewer landing mistakes on non-hosts in the field (Stanton 1984). Increased oviposition experience on a novel host does not, however, result in female Euphydryas editha (Nymphalidae) butterflies becoming more efficient at locating that novel host in the field (Parmesan et al. 1995).
Foraging moths and butterflies can rapidly learn to associate a sugar reward with odor (Hartlieb 1996; Fan et al. 1997) or color (Swihart & Swihart 1970; Swihart 1971; Goulson & Cory 1993; Weiss 1995, 1997; Kelber 1996; Kelber & Henique 1999; Kinoshita et al. 1999), sometimes after only a single rewarded trial. Insects trained to one color can rapidly learn to reverse their preferences when the reward is switched to a previously unrewarding color (Goulson & Cory 1993; Kelber 1996; Weiss 1997; Kelber & Henique 1999; Kinoshita et al. 1999). Macroglossum stellatarum (Sphingidae) moths trained in a dual-choice situation learned not only to visit the rewarding color but also to avoid the unrewarding color when given a choice of three; such avoidance of unrewarding stimuli has also been found in honeybees (Kelber 1996). Both moths and butterflies improve at finding nectar in real and artificial flowers with increased experience, and can also learn to access nectar in a new floral location after learning an initial pattern (Lewis 1986; Kandori & Ohsaki 1996; Cunningham et al. 1998).
Some moths and butterflies seem to be able to use landmarks to return to a given spatial location. Heliconius (Nymphalidae) butterflies return to nocturnal roost sites, "trapline" from flower to flower, and avoid areas where they have been captured and released, all of which may involve learned use of visual landmarks (Turner 1981; Waller & Gilbert 1982; Mallet et al. 1987). Macroglossum stellatarum moths return to a location where they were previously fed (Kelber & Pfaff 1997), again suggesting spatial learning ability.
As the research reviewed above demonstrates, non-hymenopteran insects can readily learn a range of floral parameters, including color, odor, morphology, and perhaps even location. Varying experimental designs, training, and testing protocols make it difficult to compare learning abilities across taxa. However, Fig. 9.1 shows that flies and butterflies can, like bees, associate a color with a sugar reward after only a single exposure.
Was this article helpful?