Have you ever considered the diversity of flowers? Why do they come in so many different shapes, sizes, and colors? And why do they produce so many different odors – or none at all? Flowering plants evolved around 140 million years ago, a fairly recent emergence evolutionarily speaking. Along with them evolved numerous species of insects, birds, and mammals. In his book, The Triumph of Seeds, Thor Hanson describes the event this way: “In nature, the flowering plants put sex, seeds, and dispersal on full display, spurring not only their own evolution but also that of the animals and insects with which they became so entwined. In most cases, the diversity of dispersers, consumers, parasites – and, most especially, pollinators – rose right alongside that of the plants they depended upon.”

Speaking of dependence, most flowering plants depend upon pollinators for successful reproduction – it is, for the most part, a mutually beneficial relationship. Even the casual observer of flowers will note that a large portion of the creatures that visit them appear to be pollinators. Thus, it is no wonder that pollination biologists have given pollinators so much credit in shaping the flowers that we see today.

Consider G. Ledyard Stebbins and his Most Effective Pollinator Principle which he defined in a paper published in 1970: “the characteristics of the flower will be molded by those pollinators that visit it most frequently and effectively in the region where it is evolving.” He then goes on to reference pollination syndromes, a phenomenon that describes how the traits of flowers are best suited for their “predominant and most effective vector[s].” In my post about pollination syndromes a few months ago, I discussed how a strict adherence to this concept has waned. In the next two posts, I discuss how the Most Effective Pollinator Principle (MEPP) may not be the best way to explain why flowers look the way they do.

 

To make this argument I am drawing mainly from two chapters in the book Plant-Pollinator Interactions: From Specialization to Generalization. The first is “Ecological Factors That Promote the Evolution of Generalization in Pollination Systems” by Jose M. Gomez and Regino Zamora, and the second is “The Evolution of Specialized Floral Phenotypes in a Fine-grained Pollination Environment” by Paul A. Aigner.

According to Aigner the MEPP “states that a plant should evolve specializations to its most effective pollinators at the expense of less effective ones.” And according to Gomez and Zamora it “states that natural selection should modify plant phenotypes [observable characteristics derived from interactions between a plant’s genes and its surrounding environment] to increase the frequency of interaction [between] plants and the pollinators that confer the best services,” and so “we would expect the flowers of most plants to be visited predominantly by a reduced group of highly effective pollinators.” This is otherwise known as adaptive specialization.

Specialization is something that, in theory, plants are generally expected to evolve towards, particularly in regards to plant-pollinator relationships. Observations, on the other hand, demonstrate the opposite – that specialization is rare and most flowering plants are generalists. However, the authors of both chapters advise that specialization and generalization are extreme ends to a continuum, and that they are comparative terms. One species may be more specialized than another simply because it is visited by a smaller “assemblage” of pollinators. The diversity of pollinators in that assemblage and the pollinator availability in the environment should also be taken into consideration when deciding whether a relationship is specialized or generalized.

That pollinators can be agents in shaping floral forms and that flowering plant species can become specialized in their interactions with pollinators is not the question. There is evidence enough to say that it occurs. However, that the most abundant and/or effective pollinators are the main agents of selection and that specialization is a sort of climax state in the evolutionary process (as the MEPP seems to suggest) is up for debate. Generalization is more common than specialization, despite observations demonstrating that pollinators are drawn to certain floral phenotypes. So, could generalization be seen as an adaptive strategy?

In exploring this question, Gomez and Zamora first consider what it takes for pollinators to act as selective agents. They determine that “pollinators must first benefit plant fitness,” and that when calculating this benefit, the entire life cycle of the plant should be considered, including seed germination rate, seedling survival, fecundity, etc. The ability of a pollinator species to benefit plant fitness depends on its visitation rate and its per-visit effectiveness (how efficiently pollen is transferred) – put simply, a pollinator’s quantity and quality during pollination. There should also be “among-pollinator differences in the evolutionary effect on the plant,” meaning that one species or group of pollinators – through being more abundant, effective, or both – contributes more to plant fitness compared to others. “Natural selection will favor those plant traits that attract the most efficient or abundant pollinators and will also favor the evolution of the phenotypes that cause the most abundant pollinators to also be the most effective.” This process implies possible “trade-offs,” which will be discussed in part two.

When pollinators act as selective agents in this way, the MEPP is supported; however, Gomez and Zamora argue that this scenario “only takes place when some restrictive ecological conditions are met” and that while specialization can be seen as the “outcome of strong pollinator-mediated selection,” generalization can also be “mediated by selection exerted by pollinators…in some ecological scenarios.” This is termed adaptive generalization. In situations where ecological forces constrain the development of specialization and pollinators are not seen as active selection agents, nonadaptive generalization may be occurring.

Gomez and Zamora spend much of their chapter exploring “several causes that would fuel the evolution of generalization” both adaptive and nonadaptive, which are outlined briefly below.

• Spatiotemporal Variability: Temporal variability describes differences in pollinator assemblages over time, both throughout a single year and over several years. Spatial variability describes differences in pollinator assemblages both among populations where gene flow occurs and within populations. Taken together, such variability can have a measurable effect on the ability of a particular pollinator or group of pollinators to act as a selective agent.
• Similarity among Pollinators: Different pollinator species can have “equivalent abundance and above all comparable effectiveness” making them “functional equivalents from the plant perspective.” This may be the case with both closely and distantly related species. Additionally, a highly effective pollinator can select for floral traits that attract less effective pollinators.
• The Real Effects on Plant Fitness: An abundant and efficient pollinator may select for one “fitness component” of a plant, but may “lead to a low overall effect on total fitness.” An example being that “a pollinator may benefit seed production by fertilizing many ovules but reduce seedling survival because it causes the ripening of many low-quality seeds.” This is why “as much of the life cycle as possible” should be considered “in assessing pollinator effectiveness.”
• Other Flower Visitors: Pollinators are not the only visitors of flowers. Herbivores, nectar robbers, seed predators, etc. may be drawn in by the same floral traits as pollinators, and pollinators may be less attracted to flowers that have been visited by such creatures. “Several plant traits are currently thought to be the evolutionary result of conflicting selection exerted by these two kinds of organisms,” and “adaptations to avoid herbivory can constrain the evolution of plant-pollinator interactions.”

This, of course, only scratches the surface of the argument laid out by Gomez and Zamora. If this sort of thing interests you, I highly encourage you to read their chapter. Next week I will summarize Aigner’s chapter. If you have thoughts on this subject or arguments to make please don’t hesitate to comment or contact me directly. This is a dialogue, dudes.