This post was written for The Giant’s Shoulders, a monthly blog carnival focusing on classic research.
Since Darwin, evolutionary biologists have thought that interactions between species cause diversification. However, it wasn’t until the second half of the Twentieth Century that scientists began to draw a connection between species interactions and speciation. One of the earliest of these studies was Verne Grant’s 1949 discovery of cleverly indirect evidence that pollinator isolation shapes the evolution of flowers [$-a].
Pollinator isolation is reproductive isolation created when animal pollinators don’t transfer pollen between plants of two different species. This could be because of pollinator behavior – say, because pollinator species tend to prefer a single plant. Or it could be because of the mechanics of pollen presentation by a flower, with each plant species applying pollen to a different part of a pollinator’s body so that foreign pollen is less likely to come into contact with the female floral parts. In either case, flowers are the key to the isolation – either to guide pollinators to their preferred target, or to make sure that the wrong pollen isn’t delivered.
Grant reasoned that pollinator isolation should have a real effect on how plant species are classified. Pollinator-isolated species probably have very different flowers; taxonomists, who look for characteristics that easily differentiate between related organisms, might therefore be more likely to use floral characteristics to tell pollinator-isolated plant species apart. To test this, Grant collected published classifications of plants pollinated by specialized animals (birds, bees, and long-tongued flies) and plants pollinated either by non-specialized animals, by water, or by wind.
Figure 1 from Grant (1959), showing the effect of pollinator isolation.
The result is presented in the tidy graph seen here. In plants pollinated by birds (A) and bees or long-tongued flies (CD), a much larger of the characteristics used by taxonomists to identify species were floral traits, compared to plants with non-specialized pollinators (E) or wind- and water-pollinated plants (FG). To follow up this result, Grant took systematic observations of pollinators’ movements through an experimental garden planted with three subspecies of Gilia capitata, each of which had differently colored flowers. The bees seemed to forage mainly among plants with similar flowers, and when Grant raised seeds from the experimental plants in the greenhouse, he found that there were fewer hybrids between subspecies than would be expected from random pollinator movement.
Generally, today, we wouldn’t assume that species classifications are an unbiased proxy for biological diversity – to some degree, they’re human constructs. But the basic idea that Grant develops, that speciation is an accidental consequence of plants’ interactions with pollinators, is still very important to how we understand the history of life. Together, flowering plants and insects make up the majority of the diversity of life on Earth, and it seems reasonable to think that this may be because the two groups interact so intimately.
More than fifty years after Grant’s study, pollinator isolation is a well-established mechanism for speciation. And the principle that Grant proposed, that increased divergence in floral traits is a sign of pollinator isolation, is still very useful. My lab, for instance, recently found that two forms of Joshua trees pollinated by different moth species are more different in certain floral dimensions than in non-floral traits [PDF]. That’s only the first step in what promises to be a long program of research (including my dissertation), seeking answers to some of the same questions that motivated Grant’s study.
W. Godsoe, J.B. Yoder, C.I. Smith, & O. Pellmyr (2008). Coevolution and divergence in the Joshua tree/yucca moth mutualism. The American Naturalist, 171 (6), 816-23 DOI: 10.1086/587757
V. Grant (1949). Pollination systems as isolating mechanisms in angiosperms. Evolution, 3, 82-97