Floral isolating mechanisms consist of barriers to interspecific pollination in angiosperms imposed by structural contrivances . . . [and] by the constancy of the pollinators to one kind of flower. . .
Ist die Pollen-übertragung durch Insekten geeignet, die zur Artbildung nötige (mechanische) Isolierung zu fördern? (Is pollen transfer by insects suitable for promoting the mechanical isolation needed for speciation?)
Another very obvious deficiency of observations indispensable to be made on the subject . . . resulted . . .[from] . . .the fertilisation of flowers by insects being studied by botanists but little acquainted with insects.
It often is claimed that Darwin had little to say about the evolution of species, in spite of the title of his 1859 book. This is not strictly true: a close reading of the Origin of Species reveals that Darwin envisioned speciation for the most part as the eventual extension of a process of divergence beginning at a much smaller scale within a single species, and driven for the most part by natural selection. What is true, however, is that a detailed understanding of speciation in its many forms remains an elusive and desirable prize: speciation is, so to speak, the holy grail of evolutionary biology. Many questions confront us still. How often is speciation a simple extension of microevolution as Darwin proposed.? When it is not, what new mechanisms of evolution come into play? What role is played by processes other than natural selection? Do different traits (e.g., floral phenotype, vegetative ecology, reproductive isolation) evolve together during speciation? What are the spatial elements of the process;
in particular, how common is speciation in near or complete sympatry? What are the genetics of the process.? What is the time course; is speciation slow or fast? These questions exemplify those that have been discussed vigorously since the Evolutionary Synthesis of the 1930s and 1940s (see e.g., Endler 1977; Felsenstein 1981; Templeton 1981; Maynard Smith 1983; Otte & Endler 1989; Harrison 1990; Howard & Berlocher 1998; Dieckmann & Doebeli 1999).
Since the Synthesis, the flowering plants have been put forth as a group in which speciation and macroevolution are straightforward. The reason for this optimism is easy to see. The remarkable evolutionary radiation of the angiosperms over the last 100 million years, which makes them the dominant land plants, was matched in broad outline by radiations in several animal groups, including birds and those insect taxa comprising most pollinators (Crepet 1983; Grimaldi 1999). Furthermore, radiation appears to have been much more rapid in families of animal-pollinated plants than in those whose members are pollinated abiotically (Dodd et al. 1999). The traits by which taxonomists recognize angiosperm species with complex flowers are disproportionately reproductive ones, suggesting that these traits have been centrally involved in speciation (Grant 1949). These observations, combined with apparent specialization of many plants to a specific type of pollinator (e.g., a specific insect order or vertebrate class; Pijl 1961), suggest a cohesive and attractive view of the mechanics of angiosperm speciation, and ofthe role ofanimal pollinators.
This view is as follows: plants tend to evolve toward specialization in their attraction of specific pollinators, and the pollinators themselves exhibit "fidelity", i.e., are specialized to visit only (or mostly) the plant species in question; or, if not specialized as species, exhibit individual "constancy" in flower visits, meaning that an individual pollinator exhibits fidelity at least over the shorter term ofa single foraging bout, a single day, or several days. Because of these behaviors, pollinators contribute to two linked processes that comprise angiosperm speciation (Grant 1949,1952,1994; Straw 1956; Pijl 1960; Baker 1963; Grant & Grant 1964,1965,1967; Free 1966; Stebbins 1970; Macior 1971; Jones 1978; Levin 1978; Crepet 1983, 1984; Wells et al. 1983; Hodges & Arnold 1994; Bradshaw et al. 1995; Schemske & Bradshaw 1999). As agents of selection, pollinators foster divergence in floral traits, because different pollinators select in different directions. Simultaneously, as agents of gene flow, their fidelity causes a great reduction or complete cessation of gene exchange between incipient species, as a pleiotropic side-product of floral divergence.
With reference to Grant's (1949) concept of pollinator-mediated reproductive isolation (see the quotation above), I refer to this as the "ethologi-cal isolation paradigm." Some evidence has accumulated to support the paradigm, even in its boldest claims of cospeciation of plants and pollinators (e.g., Powell 1992) and of speciation in sympatry (Hiesey et al. 1971; Vickery 1995; Ippolito & Holtsford 1999). But this positive evidence is surprisingly limited so far. In part this proves how difficult it is to accumulate empirical evidence that will satisfy a skeptical inquirer (as opposed to accumulating "plausibility arguments" for the paradigm). In addition there are reasons to suppose that the paradigm will be far from universal, stemming from recent empirical studies of the aforementioned skeptical enquirers which suggest that the situations fostering disruptive selection by pollinators in sympatry, and allowing cospeciation, will be special ones (e.g., Patel et al. 1993; Herrera 1996; Wcislo & Cane 1996; Wilson & Thomson 1996). Further doubts arise from contemplating how pollinators are expected to behave in choosing flowers.
It is my purpose in this chapter to argue that closer ties between botanists and zoologists, in particular those studying the behavior of foraging animals, will enrich our understanding of pollinator-mediated speciation in flowering plants. In so doing I will develop a different scenario for the role of pollinators in plant speciation. I contend that faster progress will be made if we search for the grail armed with a range of scenarios, in particular ones that include the perspective of foraging animals.
Was this article helpful?