Fig. 6.3. The Amax values of photoreceptors of 11 species of bumble bees superimposed on their phylogenetic tree (according to Scholl etal. 1990; Williams 1994), with Apis mdlifera for outgroup comparison. #-UV receptors; (B-blue receptors; A-green receptors. Note that the absence of some receptor types in some cases does not actually mean that the species is lacking that receptor: in some cases, the authors did not seek to find all the receptors present. References: 1, Mazokhin-Porshnyakov (1969); 2, Bernard & Stavenga (1978); 3, Meyer-Rochow (1980); 4, Peitsch etal. (1992).
performance, such as a receptor with an altered spectral sensitivity. In a stable population, only one (or very few) of these queens will survive and again produce fertile offspring. Many will succumb to frost in the winter or bird predators, or their newly founded colonies may be attacked by cuckoo bumble bees or parasites; the question of whether a queen survives all these hardships is entirely unrelated to its foraging ability. Even if the blessed queen successfully starts a colony, and if its worker offspring forage slightly more efficiently, the next generation of queens will be subject to the same unpredictable hazards. Now imagine that the mutation in question is rare and occurs only once in several generations. It is clear, then, that its chances of spreading are very slim. Generally, the probability of the mutation spreading to fixation is correlated with the frequency of the mutation and its relative adaptive advantage and inversely correlated with population size (Ohta 1993). The influence of genetic drift will be further enhanced when reproductive success varies strongly among individuals (Adkison 1995), as is the case in bumble bees (Imhoof & Schmid-Hempel 1999), or when (repeated) bottlenecks occur, such as in Canary Island bumble bees (Widmer et al. 1998). The influence of these parameters on the goodness of fit in biological signal-receiver systems has generally been ignored, but should be extremely worthwhile to explore in the future.
The conservation of Amax values within the bumble bees need not necessarily reflect any kind of constraint, however. If flower-color detection and identification in all these habitats require similar receiver systems, then we would expect conservation even in a world without phylogenetic constraint. Indeed, the estimated optimal color coding systems for Brazilian, Israeli, and German flowers from several habitats were almost identical (Chittka 1996). Be that as it may, the search for sensory adaptations is predictably frustrating if several related species display the same trait. Ideally, we need to study a trait that is variable both within and between closely related species (Chittka & Briscoe 2001). The only striking variation in receptors among the Hymenoptera is the occurrence of red receptors in very few species (Peitsch et al. 1992). Why most bee species lack such receptors defies a simple adaptive explanation. Although pure red flowers are rare in many habitats, many flowers do present information in the red part of the spectrum. Bee color-vision systems would, in theory, gain substantially if they had red receptors in addition to UV, blue, and green receptors (Chittka & Menzel 1992).
The evolution of flower-color preference in bumble bees
In an attempt to identify a visual trait that might reveal a more interesting pattern of adaptation to the visual environment, we evaluated the innate floral color preferences of bumble bees. We hypothesized that evolutionary changes of such preferences require changes only in the synaptic efficiency between neurons coding information from the color receptors. Therefore, color preferences might adapt more readily to environmental requirements than do the wavelength sensitivities of color receptors.
In one study, a good correlation was found between the color preferences of naive honeybees and the nectar offerings of different flowers in a nature reserve near Berlin (Giurfa et al. 1995). In brief, honeybees preferred the colors violet (bee UV-blue) and blue (bee blue), which were also the colors most associated with high nectar rewards. This pattern is not unique to the German flora: a similar association of flower color with reward was found in Israel (Menzel & Shmida 1993). However, a correlation never indicates causality. To show that color preferences evolved to match floral offerings, we need to compare a set of closely related bee species (or populations of the same species) that live in habitats in which the association of floral colors with reward is different.
We tested seven species of bumble bees from three subgenera: four from central Europe (Bombus terrestris terrestris [229; 8; 4698], B. lucorum [39; 2; 547], B. pratorum [14; 1; 395], and B. lapidarius [83 ;2; 1446]); two from Japan (B. ignitus [89; 3; 2782] and B. hypocrita [54; 2; 1691]); and one from North America (B. occidentalis [122; 4; 3405]). Numbers in brackets give the number of individuals tested, the number of colonies, and the number of choices evaluated. All species preferred the violet-blue range, presumably a phylogenetically ancient preference (Fig. 6.4). In addition, however, B. occidentalis had the strongest preference for red of all mainland bumble bee populations examined. This is provocative because this species frequently robs nectar and forages heavily from red flowers apparently adapted for pollination by hummingbirds (Chittka & Waser 1997; Irwin & Brody 1999). Clearly, this preference is derived and therefore might be an adaptation unique to B. occidentalis.
We also tested Bombus terrestris terrestris from Holland [85; 3; 1670], B. terrestris terrestris from Germany [144; 5; 3028], B. terrestris dalmatinus from Israel [156; 5; 5731], B. terrestris dalmatinus from Rhodes; [150; 5; 5335];
Was this article helpful?