One of these moths is not like the other … but does that matter to Joshua trees?

A Joshua tree flower, up close.

A Joshua tree flower, up close.

A huge diversity of flowering plants rely on animals to carry pollen from one flower to another, ensuring healthy, more genetically diverse offpsring. These animal-pollinated species are in a somewhat unique position, from an evolutionary perspective: they can become reproductively isolated, and to form new species, as a result of evolutionary or ecological change in an entirely different species.

Evolutionary biologists have had good reason to think that pollinators often play a role in the formation of new plant species since at least the middle of the 20th century, when Verne Grant observed that animal-pollinated plant species are more likely to differ in their floral characteristics than plants that move pollen around via wind. More recently, biologists have gone as far as to dissect the genetic basis of traits that determine which pollinator species are attracted to a flower—and thus, which flowers can trade pollen.

However, while it’s very well established that pollinators can maintain isolation between plant populations, we have much less evidence that interactions with pollinators help to create that isolation in the first place. One likely candidate for such pollinator-mediated speciation is Joshua tree, the iconic plant of the Mojave Desert.

2009.03.20 - Tikaboo Valley (closeup)

Joshua trees.

Joshua trees are pollinated by yucca moths, which are unusually focused, as pollinators go. Your average honeybee will blunder around in a flower, scooping up pollen and drinking nectar, and maybe accidentally pollinate the flower in the process. A yucca moth, on the other hand, gathers up a nice, tidy bundle of pollen in specialized mouthparts, carries it to another Joshua tree flower, and deliberately packs it into place. She does that because the fertilized flower provides more than a little nectar for her—she’s laid her eggs inside the fertilized flower, and when they hatch her offspring will eat some of the seeds developing inside it.

That’s pretty cool in its own right. But what’s especially interesting about Joshua trees, from an evolutionary perspective, is that they’re pollinated by two different moth species. And it turns out that the flowers of Joshua trees associated with the different moth species also look pretty different. The most dramatically different feature is in the length of the stylar canal in the pistil, the part of the flower that determines how the moths lay their eggs.


Joshua tree pistils (left) seem to be matched to two different species of moths that pollinate the trees. Based on Yoder et al. (2013), figure 1.

Trees pollinated by the larger moth species tend to have longer styles (the blue circles in the figure above), and trees pollinated by the smaller moth species tend to have shorter styles (the green triangles). Moreover, the rest of the trees structure differs by moth association as well. In the figure above with the two Joshua trees side by side, the tall, tree-like one is typical of the populations pollinated by the large moth; the short, bushy one is what we usually find in populations pollinated by the small moth. And, through most of the Mojave Desert, the moth-tree association is exceptionally tight: except for a narrow contact zone between the two tree types (where that photo was taken), you only find one tree type, and one moth species, in most Joshua tree populations.

All of this strongly suggests that the two types of Joshua tree must be reproductively isolated by their association with different pollinators, right? Well, maybe. Part of my doctoral dissertation work focused testing this question using genetic data, and in the past couple months, my collaborators and I have published two papers reporting the results of that work, and a related study at the contact zone.

It turns out that, at the contact zone, anyway, the two moths move between the tree types often enough to create quite a lot of hybrids [PDF]. And when we examined the population genetics of Joshua trees across the entire Mojave Desert, we found that they’re not very strongly differentiated at all [PDF].

Genetic structure analysis of Joshua tree populations across the Mojave Desert. Based on Yoder et al (2013), figure 2.

Genetic structure analysis of Joshua tree populations across the Mojave Desert. Based on Yoder et al (2013), figure 2.

Here’s a graph illustrating that second point: in this analysis, we used a clustering algorithm to determine whether the genetics of each tree in our analysis suggested it belonged to one of two possible groups. In the graph, each bar is colored according to the relative probability that a single tree is in each group belongs to each group. The bars are grouped according to the location of the trees they represent, and which moth species is present at each site. As you can see, there are a lot of bars that aren’t clearly one color or the other—and the ambiguous trees they represent aren’t confined to the sites where both moths are present.

So on the one hand, the different pollinators aren’t doing a very good job of isolating the two forms of Joshua tree. But on the other, that result leaves the very interesting question of why there are two types of Joshua tree at all. An obvious hypothesis is that the two different pollinators are exerting divergent natural selection on the trees they pollinate—and my collaborator Chris Smith will be presenting data to test that hypothesis in the not-too-distant future.


Schemske D.W. & Bradshaw H.D. (1999). Pollinator preference and the evolution of floral traits in monkeyflowers (Mimulus). Proceedings of the National Academy of Sciences, 96 (21) 11910-11915. DOI:

Grant V. (1949). Pollination systems as isolating mechanisms in angiosperms. Evolution, 3 (1) 82. DOI:

Starr T.N., Gadek K.E., Yoder J.B., Flatz R. & Smith C.I. (2013). Asymmetric hybridization and gene flow between Joshua trees (Agavaceae: Yucca brevifolia) reflect differences in pollinator host specificity. Molecular Ecology, 22 (2) 437-449. DOI:

Yoder J.B., Smith C.I., Rowley D.J., Flatz R., Godsoe W., Drummond C. & Pellmyr O. (2013). Effects of gene flow on phenotype matching between two varieties of Joshua tree (Yucca brevifolia; Agavaceae) and their pollinators. Journal of Evolutionary Biology. DOI:

9 comments on “One of these moths is not like the other … but does that matter to Joshua trees?

  1. jeffollerton says:

    Really interesting study, thanks for summarising it. One question: how closely related are the moths? Any evidence that they also hybridise?

    • Yoder says:

      Excellent questions! The moths are sister species (each other’s closest evolutionary relative), but they don’t hybridize at all. The best estimate we have is that they’ve been reproductively isolated for more than a million years. See Smith et al (2008) [PDF] for details on that point.

  2. Sam Steffen says:

    I never would have thought there would be one type of animal or insect that would work within a single species of flower. I wonder if these traits could be utilized to help repopulate a species of flora that is on the verge of extinction. I am also curious about the why. Why does these two moths stay with the Joshua Tree? Is there a specific need they have met that is only in the Joshua Tree?

    • Yoder says:

      To answer your last question first, Joshua trees provide the moths with their only significant source of food—they eat developing Joshua tree seeds while they’re caterpillars, then they burrow into the soil to pupate for at least a year, and then they spend a week or so as adults, too busy mating to do much about eating.

      As to why they’re dependent on Joshua trees specifically … that’s hard to say exactly. There are other yucca moth species using other species of yucca, and some moth species can use mutliple host species. Something about adapting to use Joshua tree—adapting to oviposit in Joshua tree fruit, or to match the timing of Joshua tree flowering, maybe—may mean the Joshua tree moths are poorly suited to using other yuccas.

      The ancestors of the yucca moths laid eggs in different host plants entirely, and they probably didn’t pollinate. The thinking is, pollination evolved to ensure a larval food source, and maybe that required greater specialization, too. One thing I’m more sure about: any pollinator this specialized is going to make a pretty poor candidate for helping to “repopulate a species of flora.”

  3. […] Homo sapiens. Goldenrod and goldenrod gallflies. Pseudomonas aeruginosa. Pine trees and crossbills. Joshua trees. Caenorhabditis elegans. Whole assemblages of tropical forest trees. And, of course, […]

  4. […] genetics is about to get a lot more powerful than the 10 microsatellite loci I used for my dissertation research. And it will involve fieldwork in the Mojave Desert, which is objectively one of the most beautiful […]

  5. […] genetics is about to get a lot more powerful than the 10 microsatellite loci I used for my dissertation research. And it will involve fieldwork in the Mojave Desert, which is objectively one of the most beautiful […]

  6. […] fascinated by all the weird, baroque ways that living things <a href="influence and coevolve with each other—so Ed Yong’s new TED talk about mind-controlling parasites is right up our […]

  7. […] my benefit. Showy flowers are an adaptation to attract animal pollinators. Some flowers are quite precisely matched to a single species of pollinator, but most flowers have lots of visitors. These less specialized […]

Comments are closed.