Yuccas and yucca moths have one of the most peculiar pollination relationships known to science. The moths are the only pollinators of yuccas, carrying pollen from flower to flower in specialized mouthparts and actively tamping it into the tip of the pistil. Before she pollinates, though, each moth lays eggs in the flower—the developing yucca seeds will be the only thing her offspring eat. How does such a specialized, co-adapted interaction evolve in the first place? My coauthors and I attempted to answer this question in a paper just published in the Biological Journal of the Linnean Society, by reconstructing the ecology of yucca moths before they were yucca moths [PDF].
Using the present to reconstruct the past
Before I describe our study’s results, let me explain a little about how biologists can reconstruct the characteristics of extinct species using what we know about living ones. First, we use DNA data to reconstruct evolutionary relationships between our favorite living species—this gives us an evolutionary tree, or phylogeny, like the ones in the illustration below. A phylogeny diagrams the branching evolutionary history that led to the living species at the tips of the tree. If we map the different states of some character that all those species have—say, the color of their feathers, onto the tips, we can infer what the ancestors at each of the inner branch points might have been like.
For instance, consider the possible scenarios for species A, B, C, and D in the illustration below. In the first case, if A and B are both red, then their common ancestor was probably red, too. However, C is blue—what does that mean for the common ancestor of C, A, and B? Because D is blue, we infer that the common ancestor of C, A, and B was also blue, as was the common ancestor of all four species. This is the most parsimonious reconstruction—it minimizes the number of times that color changes in the evolutionary history of the four species.
In the second scenario, species D is red, so the same logic infers that the common ancestor of A, B, and C was red. In the third scenario, adding another red species (E) to the tree might also alter the most likely character states for the ancestral species on the tree—but this depends on where the DNA suggests that the new species fits on the tree. Most modern reconstructions of ancestral character states are more statistically complex than what I’ve just described, but the underlying logic is the same.
What did the ancestors of yucca moths do for a living?
So in order to reconstruct what yucca moths were like before they became yucca moths, we need to know the evolutionary relationships between yucca moths and their close relatives, the other members of the moth family Prodoxidae. This is a diverse group, including
- The yucca-pollinating genera Tegeticula and Parategeticula;
- The genus Prodoxus, moths that lay eggs on various parts of yuccas (and other related plants) without pollinating them;
- The genus Mesepiola, which lay eggs on the flowers of plants similar to yuccas—woody desert monocots;
- The genus Greya, which lay eggs in the flowers of a number of different plants, and pollinates some in the process [PDF]; and
- The genus Lampronia, which lay eggs in another wide assortment of plants.
This diversity offers some intriguing possibilities—depending on how these genera are related to each other, the moths that would colonize yuccas and evolve obligate pollination mutualism might have lived on anything from roses to saxifrages, and their larvae might have eaten leaf tissue, woody twigs, fruit, or flowers. However, the last study to reconstruct the evolutionary relationships among these groups included only one species of Lampronia, leaving a number of current host plant associations and larval feeding habits unrepresented.
So we collected new DNA sequences from another dozen species in the genus Lampronia, reconstructed their relationships to the rest of the Prodoxidae, and used the resulting phylogeny to estimate the host plant association and larval feeding habit of the ancestral species that gave rise to the yucca moths. The results are presented in the large, color-coded figure below. Interpretation of this figure is similar to the example I gave above, except that the reconstruction method we used allows us to estimate the relative probability of each character state at the ancestral nodes, which we present in color-coded pie charts.
This gives us a better picture of the evolutionary changes in the lineage that would become yucca moths. The ancestral moths probably fed inside floral ovaries all the way back to the origin of the Prodoxidae. Before colonizing woody monocots (the Agavaceae, the family including yuccas, and possibly the Ruscaceae, the family fed on by Mesepiola), they most likely fed on plants in the rose family.
This reconstruction gives us the best picture we’ve had to date of the conditions under which yucca moths evolved obligate mutualism—before they were active pollinators, the moths were already feeding inside developing flowers. This suggests that active pollination evolved to help ensure a larval food supply. We might imagine, then, that plants used by these pollinating seed parasites would evolve greater dependence on their highly efficient pollen delivery, moving toward the yucca-yucca moth mutualism we see today.
Update, 1017: Just realized I scheduled this post without adding a link to my interview with Chris Clarke, which covers the results of this paper from a rather different angle.
Brown, J., Leebens-Mack, J., Thompson, J., Pellmyr, O., & Harrison, R. (1997). Phylogeography and host association in a pollinating seed parasite Greya politella (Lepidoptera: Prodoxidae). Molecular Ecology, 6 (3), 215-24 DOI: 10.1046/j.1365-294X.1997.t01-1-00171.x
Pellmyr, O. (1999). Forty million years of mutualism: Evidence for Eocene origin of the yucca-yucca moth association Proc. Nat. Acad. Sci. USA, 96 (16), 9178-83 DOI: 10.1073/pnas.96.16.9178
Yoder, J.B., Smith, C.I., & Pellmyr, O. (2010). How to become a yucca moth: minimal trait evolution needed to establish the obligate pollination mutualism Biol. J. Linnean Soc., 100 (4), 847-55 DOI: 10.1111/j.1095-8312.2010.01478.x