The Occasional Hymenoptera: Physiological adaptation necessary to digest pollen?
The rapid radiation of flowering plants (angiosperms) since their first appearance in the Lower Cretaceous is attributable to the innovations toward something like mutualism with herbivores–perhaps “armed détente” would be a better description. Rather than exclusively relying on defensive adaptations to protect themselves from herbivores, they developed mechanisms to take advantage herbivory, such as for seed dispersal. Instead of protecting seeds in unappetizing wood-like cones, angiosperms developed nutritious fruits to encase seeds. When herbivores ate the fruit, they also ate the seeds and carried them away and deposited them with a supply of fecal matter rich in essential elements for plant growth. At the same time, plants also developed features to prevent wholesale destruction of leaves and branches by herbivores–such as thorns, thick bark, toxic or distasteful chemicals and so forth.
The innovations of angiosperms had significant impact on the nature of large herbivores during the Mesozoic and afterwards. In the Northern Hemisphere (Laurasia) large sauropods (which were designed to accommodate large guts which acted as fermenting vats for processing the low nutrition of the Jurassic plants) were replaced in the Cretaceous by hadrosaurs and ceratopsians with specialized beaks and dental arrays for chewing the more delicate and nutritious angiosperms. After the extinction of dinosaurs, a similar relationship between angiosperms and larger herbivores, this time mammals, developed in the Cenozoic.
Sexual reproduction by active pollination (by vectors such as insects and birds) was another method by which the diversificaiton of angiosperms accomplished. This innovation led to great adaptations among insects. Although beetles were the first insects to pollinate flowers, bees descended from predatory wasps (possibly through pollen wasps) to become the first specialized pollen agents. They arose in the Albian Epoch of the Cretaceous about 100 million years ago, not very long after the widespread appearance of angiosperms. See G. O. Poinar Jr. & B. N. Danforth, “A Fossil Bee from Early Cretaceous Burmese Amber,” 314 Science 614 (October 27, 2006) (on-line access).
Bees rapidly radiated throughout the world because they became much more efficient at obtaining the nectar and pollen from flowers than moths, butterflies, beetles and pollen wasps. In turn they drove the radiation and modification of angiosperms.
The 20,000-30,000 bee species alive now are by far the most important pollen vectors. The challenge they present to flowering plants, however, stems from their highly efficient ability to remove pollen. They can often remove 70-90% of pollen in a flower in one visit. Many bees carry the pollen in scopae (hair brushes on the leg or abdomen) or in the crop, which prevents loss to the bee and deposit (for fertilization) on other plants. The more efficient a bee becomes in removing pollen from a plant, the less useful it is to the plants who depend on insects for pollination. Unless bees deposit at least some pollen on the pistils of other plants, the removal of pollen is not useful to the plant and becomes parasitism.
The selective pressure on plants to make available pollen to pollinators but to ensure that an optimal amount of it is used for pollination is the key feature of the relationship between insects (and birds) and flowering plants. When an accommodation is obtained it is something like mutualism between the plant and pollinator. The view of the relationship as a form of mutualism between insects and flowering plants has long been regarded as one of the iconic examples of co-evolution. In 1792 Christian Konrad Sprengel first systematically observed the role insects played in the sexual reproduction of plants and published his observations in his Das entdeckte Geheimniss der Naturim Bau und in der Befruchtung der Blumen (Berlin: F. Vieweg: 1793) (translated as Discovery of the secret of nature in the structure and fertilization of ﬂowers by P. Haase in D.G. Lloyd & S.C.H. Barrett (eds), Floral biology: studies on ﬂoral evolution in animal-pollinated plants (NY: Chapman & Hall: 1996), pp 3-43). He proposed that plants had characteristics which attracted particular kinds of pollinators in order to to perform cross-fertilization between individual plants; the plant characteristics in this relationship would later be called “pollination syndrome.” (“Pollination syndrome,” as pioneered by Federico Delpino, is technically the suite of characteristics that covary with particular pollen vectors. Thus a flower with bright purple color, a nectar guide and deeply hidden but ample nectar–are among the characters in the pollination syndrome for butterflies. See a table of pollinator syndrome traits at a wildflower page maintained by the US Forest Service.)
Sprengel’s observations were largely ignored in Germany at the time and were not given due regard until Darwin considered them in the decade right after he published his Voyage of the Beagle. In the 1840s Darwin made extensive observations of plant-insect interactions which confirmed much of what Sprengel observed but which in some respects also contradicted Sprengel’s conclusions about the “design” and “purpose” of bees. (In his August 16, 1841 letter to the Gardener’s Chronicle, Darwin writes that “by boring a hole into the flower instead of brushing over the stamens and pistils,” the humble bees he observed cheated the “final cause” of their existence as Sprengler imagined it.) In 1862 Darwin’s first work after Origin of Species was On the various contrivances by which British and foreign orchids are fertilised by insects, and on the good effects of intercrossing (London: John Murray: 1862), which was designed as a collection of observations and experimental proof of natural selection and the co-evolution of insects and orchids. The work was highly influential and went through a number of editions. Hermann Müller, German botanist and supporter of Darwin, extended the research by examining the entire ecology to show how both plants and insects adapted. Herman Müller & Federico Delpino, “Application of the Darwinian theory to ﬂowers and the insects which visit them” (1869), translated by R.L. Packard in 5 American Naturalist 271–297 (1871). Delpino, as we noted, originated the concept of pollination syndrome, which became a standard way of viewing flower design.
Recently, the neatness of this view has been questioned. In a presidential address delivered to the American Society of Naturalists at a joint meeting with the Society for the Study of Evolution and the Society of Systematic Biologists in 2001 (reprinted as “When is it Mutualism?,” 162 American Naturalist (Supplement) S1-S9 (October 2003) (pdf file)), James Thomson, for example, questioned whether mutualism is the correct description of the relationship between a particular pollinator and a particular plant:
“[A] pollinating animal may function as a mutualist to a plant in some ecological circumstances but as a parasite in others. That is to say, the sign of the net interaction depends not only on the intrinsic properties of the species but also on the biotic background in which they interact. Because of this, we should not say that insect species A is a mutualist for plant species B. The furthest we can go is to say that A helps B at some particular place and time but might hurt it in some other situation.” (p S2.)
His argument was that the selective pressures on plants are to present pollen to the most efficient pollinators. In an environment where only bees exist, the bee-plant relationship may be mutualistic. But when hummingbirds are present (hummingbirds “lose” much less pollen than bees), the same bee becomes a parasite on the plant. The selective pressure on the plant in that environment is to favor features that discourage bees without discouraging hummingbirds–such as a narrow, deep nectar tube, diluted nectar, etc. So mutualism has to be viewed within the context of the entire environment, rather than simply by viewing the interactions of a particular pollinator and particular plant.
Bees, unlike, for example, hummingbirds are particularly inefficient pollinators because they use pollen to feed larva. (Foragers feed on nectar; queens and brood attendants depend on honey and occasionally pollen.) The pollen used to feed larva obviously is not available to fertilize a plant. And bee larva consume a great deal of pollen. One estimate is that the pollen of up to 3000 flowers is necessary to feed one offspring. A. Müller, “Unusual host plant of Hoplitis pici, a bee with hooked bristles on its mouthparts (Hymenoptera: Megachilidae: Osmiini), 103 European Journal of Entomology 497–500 (2006). Their pollen needs, combined with their efficiency at stripping a flower of its pollen and the relatively small amount of pollen bees deposit on other flowers put pressure on plants to restrict access to bees when there are more efficient pollinators available.
Flowers have acquired numerous morphological features to discourage pollen removal by less efficient (from the plant’s perspective) pollinators. Almost all flowers that depend on bees for pollination are dorsiventral (i.e., they are flattened with a distinct top and bottom). This is because bees themselves (like almost all terrestrial animals) are dorsiventral (owing to effect of gravity on bilaterally symmetrical organisms). This flower plan allows for the bee to land, move toward the food source by which means the plant attaches the pollen to the underside of the bee. Changes from this design usually make it more difficult or impossible for a bee (or all but specialized ones) to obtain the pollen. One strategy is to “hide” the pollen from the “bait” (such as the nectar) or have part of it in places where it can attach to the bee other than in the scopae. Bilabiate plants, for example, have “roofs” and “floors” with a nectary deep inside. Often when bees seek out the nectar pollen from the “roof” attaches to the bee’s dorsal side. Other designs and mechanisms are involved in other bilabiate flowers, which arose independently in 38 different angiosperm families. See Christian Westerkamp & Regine Claßen-Bockhoff, “Bilabiate Flowers: The Ultimate Response to Bees?,” 100 Annals of Botany 361–374 (2007) (full text). Other morphological “defenses” include keel blossoms (Christian Westerkamp, “Keel blossoms: bee flowers with adaptations against bees,” 192 Flora 125–132 (1997)), narrow tubes, (A. Müller, “Morphological specializations in Central European bees for the uptake of pollen from flowers with anthers hidden in narrow corolla tubes (Hymenoptera: Apoidea),” 20 Entomologia Generalis 43–57 (1995)), and other structures that prevent access to those not possessing specialized morphological or behavioural adaptations (see R.W. Thorpe, “The collection of pollen by bees,” 222 Plant Systematics and Evolution 211-223 (2000) (behind pay wall)).
In addition to morphological features of flowers plants have another potential defense to inefficient pollinators–the chemistry of the pollen itself. Just as there are examples of plants that use thorns or other morphological defenses to herbivory, there are examples of chemicals that make leaves distasteful or toxic to herbivores. The question is whether the chemistry of pollen can be used as a defense against inefficient pollinators. Pollen is highly variable in its composition and mix of amino acids, lipids, starch, sterols, vitamins or secondary metabolites. Some have low nutritional value, and some contains secondary compounds that are toxic. (See, e.g., T.H. Roulston & J.H. Cane, “Pollen nutritional content and digestibility for animals. 222 Plant Systematics and Evolution 187–209 (2000)). Even so, herbivorous insects have been known to adapt physiologically to deal with plant toxins. See Sebastian E.W. Opitz & Caroline Müller, “Plant chemistry and insect sequestration,” 19 Chemoecology 117-154 (2009).
In Claudio Sedivy, Andreas Müller & Silvia Dorn, “Closely related pollen generalist bees differ in their ability to develop on the same pollen diet: evidence for physiological adaptations to digest pollen,” Functional Ecology (January 31, 2011–online in advance of print) (full access), the authors, from ETH Zurich, compared the larval performance of the two very closely related and highly pollen generalist solitary bee species Osmia bicornis and Osmia cornuta on four different pollen diets. The results were as follows: O. bicornis developed well on Ranunculus pollen but failed to do so on Echium pollen, whereas the reverse held true for O. cornuta (with the exception of two dwarf adults). Both bee species performed well on Sinapis pollen, while neither developed on Tanacetum pollen. This was the ﬁrst evidence that larvae of two closely related generalist bee species differ in their physiological ability to digest pollen from the same host plant. They conclude that this difference resulted from adaptation.
The authors say that the need for physiological adaptation to digest different pollen raises the question “whether unfavourable pollen properties have evolved as protection against pollen-collecting bees, whether they are by-products of the plants’ physiology serving other primary goals or whether they are a pleiotropic consequence of chemical defence against herbivores in other tissues.” They hypothesize that the high pollen demand of bees puts pressure on plants to select for protective properties in the pollen that would limit and select pollen consumers. They reason as follows. First, selective pressures have been responsible for morphological adaptation of flowers for the same reason, so it is not unlikely that pressure from the same source would act on chemical composition of pollen. Second, recent studies have shown that pollen of certain plants contain insecticidal compounds in higher concentration in the pollen than in leaves and stems. This suggests that the purpose is to protect against pollen parasitism. Finally, those cases where pollen has been found to contain compounds that adversely affect larval development have all been from flowers of a type with pollen freely accessible to any bee (i.e., they had no morphological adaptation to filter out or discourage any bee or other insect). This suggests that the presumed selective pressure operated in those cases on pollen since it did not operate on morphology.
All this seems plausible. The question that seems open, however, is how having unfavorable chemical composition prevents pollen extraction from the unfavored bee. Obviously, it does limited good (for the plant) if the adaptation by the plant operates only after the bees have deprived it (and all its cospecies in the neighborhood) of all its pollen. It is possible that it operates on the principle that the toxin kills off all colonies in the area and therefore the plant is able thereafter to thrive. While the plants from which the pollen in the experiment came are all perennials and therefore could survive a season without producing seeds, usually a toxin which has long-term effects works by aposematism (such as warning coloration), discouraging the animal from preying on the protected organism or at least by warning the predator of the toxin. This of course depends on the memory of the predatory animal, either hardwired or by individual experience. It’s easy to see how morphological adaptations prevent pollen taking; it’s not so easy to see how pollen whose toxicity only operates long-term on larva can do so.