SCRP 2009

My research examines the pollination biology and population genetics of Joshua trees (Yucca brevifolia). If you’re not familiar with Joshua trees, they are large, woody monocots (the group of plants that includes grasses, orchids, and palm trees) that are endemic to the Mojave Desert region of California, Arizona and Nevada. Joshua trees are known for their comical (or ugly, depending on your outlook) appearance, and for their remarkable pollination biology (see below). John C. Fremont, an early American explorer described them as “The most repulsive tree in the vegetable Kingdom”. Later, Mormon settlers saw in their silhouette the figure of the prophet Joshua, pointing the way to the Promised Land.

Students working in my lab will work on projects related to one of two aspects of the biology of Joshua trees. One project, which is currently fairly well developed, looks at the pollination biology of these remarkable plants. The second project, which is still in early stages of development, looks at the effects of global warming on population growth rates. Below, I’ve provided detailed information about these two projects, along with relevant references and citations, and some ideas for questions that could be answered in student projects. Students interested in applying to work in my lab should read the project descriptions below, and the descriptions of potential student projects.

When submitting their SCRP application, students should explain in a 3-5 page essay (as described on the SCRP webpage) their ideas for how they might answer some of questions identified in the project description (below). This essay should do two things: 1. It should present a coherent plan for how student research could address the question, and 2. It should show that the applicant has done background reading and research. Background reading could include reading some of the references listed here, but should also include other relevant research articles from the primary literature. These articles might be examples of how other scientists have tackled similar problems, or general articles about the natural history of the system, or the theory behind the statistics and experimental design. Of course, it will be challenging to put together a clear research proposal for a system that students have only recently read about. The main things that I will be looking for in reviewing student application are the ability to think clearly about the problems, and that the applicants have taken some initiative to learn about the systems and relevant concepts. I also encourage prospective SCRP students to meet with me before submitting their application to discuss their ideas.

Joshua Tree Pollination Biology

Joshua trees (Yucca brevifolia), like all yuccas, are pollinated exclusively by highly specialized yucca moths. The moths, in turn, lay their eggs exclusively on the flowers of the Joshua tree, and as the flower develops into a ripening fruit, the moth’s larvae feed on the developing seeds. To ensure that the flower will develop and provide food for her offspring, the female yucca moth deliberately deposits a load of pollen directly onto the flower’s stigmatic surface using specialized, uniquely derived, tentacle-like mouthparts (Pellmyr 2003). The interaction between yuccas and yucca moths is therefore the archetypical example of an obligate pollination mutualism, one which Darwin called “the most remarkable pollination system ever described” (Darwin 1874).

Within this system, there is thought to be strong reciprocal natural selection (coevolution) promoting phenotype matching between Joshua trees and their pollinators. Across species, there is a strong, statistically significant correlation between the length of the female moth’s ovipositor (a blade-like appendage used to cut into flowers when laying eggs) and the thickness of the flower’s ovary wall at the site of oviposition. Likewise, there are strong functional constraints acting on both moth and pollinator features: the moth’s ovipositor must be long enough to reach ovules within the flower, but an ovipositor that is too long may injure the developing flower, which can prompt flower to die along with the moth’s larvae (Marr and Pellmyr 2003). (See Figure 1)

In the Joshua tree system specifically, recent work revealed that across its range Y. brevifolia is pollinated by two distinct, parapatrically distributed moths, Tegeticula synthetica Riley in the west, and T. antithetica Pellmyr in the east (Pellmyr and Segraves 2003). These two moths differ significantly in overall body size, and in the size of the female ovipositor (Pellmyr and Segraves 2003). Recent work has shown that Joshua trees pollinated by the different moth species differ significantly in overall floral and vegetative morphology, and differ more than any other feature in the length of the floral stylar canal – the path through which the female yucca moth inserts her ovipositor during oviposition (Godsoe et al. 2008). These patterns suggest that floral style length and moth ovipositor length may experience reciprocal natural selection (coevolution in the strict sense) favoring phenotype matching.

One way to test his hypothesis would be to compare the performance of moths of a given ovipositor length on flowers that differ in style length using a reciprocal transplant experiment, but to do so would require moving moths between different populations, keeping the moths alive during transport, and forcing them to interact with foreign trees. These factors potentially make testing this hypothesis very challenging. However, the recent discovery of a ~4 km-wide contact zone where the two species of yucca moth co-occur in Tikaboo Valley, Nevada (Smith et al. 2008) now makes it possible to address this question using a ‘natural experiment’. Contact zones like that in Tikaboo Valley have been used in other pollination systems to study the role of natural selection and host preference in maintaining species boundaries (Fulton and Hodges 1999; Grant 1952; Hodges and Arnold 1994), but such an approach has not previously been used in an obligate pollination mutualism.

Within Tikaboo Valley, trees with different pollinator-associated morphotypes occur side-by-side, along with some trees that appear to be intermediate between the two types, and there is greater variation in both style length and ovipositor length than in any other population. Although moths of both species have been collected on each of the tree morphotypes in the zone of sympatry, it is unclear whether female moths show a preference in oviposition for trees whose floral morphology matches their ovipositor. If reciprocal natural selection favors phenotype matching in this system, then moths ovipositing on the “wrong” tree type should have reduced fitness in terms of offspring that survive to adulthood.

We can begin to address some of the questions by measuring how often moths move between the two tree types in the contact zone, and how often the moths successfully lay eggs on each of the two trees. By attaching plastic cards coated in glue to receptive flowers we can trap moths as they visit the trees, and then determine how often each species visit the different tree types. (See Figure 2). Likewise, we can get a rough idea about how often the two moths oviposit into the “wrong flowers” by collecting larvae as they emerge from fruits. Although the larvae of the two species are essentially indistinguishable in appearance, we can compare mitochondrial DNA, as well as regions of highly variable DNA called microsatellites, to identify the larvae to species.

My colleagues at the University of Idaho and I completed an experiment like this in 2007, with the following results:

A. Moth visitation rates
	Moth species (forewing length)
Tree Type	T. antithetica	T. synthetica	∑
Eastern	10	3	13
Western	22	23	45
∑	32	26	58
B. Larval emergence rates
	Larval mitotype
Tree Type	T. antithetica	T. synthetica	∑
Eastern	296	3	299
Western	15	45	60
∑	311	48	359

Table 1: Estimates of host specificity and oviposition success of each moth species on each tree type. Intermediate (hybrid) trees were not included in this experiment. Visitation rates are based on passive sampling using sticky-cards. Larval emergence rates reflect larvae collected from fruits of each tree type, identified to species using mtDNA sequencing.

Whereas eastern trees are visited primarily by T. antithetica (the eastern moth), western trees receive almost exactly equal visitation by the two moth species. A similar asymmetry is seen in the larvae emerging from the two tree types. Whereas 99% of the larvae emerging from eastern trees are of the eastern moth species (T. antithetica), a full quarter of the larvae on western trees are of the eastern moth species. These preliminary results suggest that western moths (T. synthetica) are either much more ‘choosy’ about which trees they will visit, or have a much more difficult time laying eggs when they visit their non-native tree (or possibly both). However, we still don’t know exactly what mediates host specificity and ovipositon success in this system, so there are still a lot of questions we would like to answer. Many of these could be the subjects of student research projects.

1. Currently, we differentiate between the two moth species caught in the sticky traps based on the size of the insects forewings. However, the moths definitely vary in body size within species, so it’s possible that we are sometimes mistaking an individual of one species, for an individual of the other. It would therefore be desirable to know how reliable body size is in telling the two moth species apart. Is there a way that we could ‘double check’ species identifications?

2. We are currently using DNA sequencing to distinguish larvae of the two moth species that emerge from fruits. There are some drawbacks to this, however. DNA sequencing is very expensive, and takes a lot of time. We would like to develop a less expensive way to distinguish larvae. One way would be to see if there are reliable morphological differences that could allow us to tell the two species apart at the larval stage. Another approach would be to use some less expensive means of genotyping the larvae. Some possibilities include using microsatellite DNA, or methods such as ‘PCR RFLP’s’, or regular old RFLP’s to distinguish larvae.

3. As I mentioned above, at the Tikaboo valley site, there are not only “western” and “eastern” trees, but also some trees that are morphologically intermediate, that might be hybrids. We’d like to know how often the moths of each species visit these hybrid trees. If it really is matching between the flower’s style and the moth’s ovipositor that makes a difference, you would expect the western moths to do a little bit better on flowers with medium-sized styles.

4. Obviously, one key step in the experiment above is being able to reliably distinguish between “eastern” and “western” trees in the hybrid zone. So, it would be desirable to develop quantitative measures of how reliable different features are for distinguishing different tree types (see the Godsoe et al paper for details about different morphological features), and how often morphological measures of tree type agree with genetic measures.

5. Lastly, we’d like to know whether the geographic location of trees within the contact zone affects how many moths of each species visit the trees. The contact zone is fairly small (about 4 KM across), and on either side of the contact zone are populations of “pure” eastern or western trees. You might expect that trees that are closer to –say- the western edge of the contact zone would receive more western moths, and this could mess up our estimates of host specificity. How could we test this hypothesis?

Consequences of Global Warming for Joshua Tree

Based on the environments where Joshua trees currently grow, we can use climate data to infer the range of environments that they are theoretically able to occupy (their fundamental niche). By then combining these “niche models” with predictions of the climate is expected to change over the next century, we can develop predictions about how the Joshua tree may respond to global warming. Based on these models, Joshua tree is predicted to be severely impacted by global warming over the next century (Dole et al. 2003; Shafer et al. 2001); local population extinctions are projected over much of its current range. Populations at low elevations, and in the southern portion of the range, including much of Joshua Tree National Park and the Mojave National Preserve are expected to go extinct.

The time scale on which these extinctions may occur is unclear, however. If we compare populations in areas expected to experience local extinctions, with those that are predicted to retain Joshua trees, there is a striking difference in the character of the populations. In areas expected to experience population declines and extinctions, population density is extremely low, and most trees are very large and (apparently) very old. (See Figure 3). Conversely, populations at higher elevations are composed primarily of young, pre-reproductive trees (See Figure 4). This pattern is even more noticeable if we compare the proportion of the population in different “size classes” (Table 2).

Size Class	% of Population
	4000’	2000’
Seedlings	75	35
2	8	3
3	6	12
4	7	16
5	3	15
6	1	11
7	0	5
Oldest	0	4

Table 2: Size class data for Joshua trees near Palmdale, CA.

At the high elevation site in Table 2, almost all of the trees are in the smallest size class, and only a small minority of trees is beyond the seedling stage. On the other hand, at the low elevation site, the majority of trees are past the seedling stage, and many are very old indeed. These differences in demography are typical of what we see in human populations experiencing either population growth (majority or the population is in the youngest age class), or population decline (majority of the population in older age classes) (Gotelli 2001).

If populations in the warmest climates are already on their way to extinction, it suggests that the effects of climate change may be much more dire for Joshua tree than the current models would predict. The underlying models assume that every place that Joshua tree currently exists represents an area where the trees can flourish. However, if many of these areas are actually not suitable, because the current climate is actually too warm or too dry, the future range of Joshua tree may be even smaller than we currently expect.

This project is still in its infancy, however, so there are lots of questions we’d like to know the answer to.

1. The data presented here represent a comparison across an elevational gradient in one part of the range. Would we find similar patterns in other parts of the range?

2. The data here are based on size classes. However, size classes are not really a very good way to estimate demography. Just for one thing, maybe trees actually grow at different rates in different elevations or different climates. So, we’d like to develop some independent estimate of tree age. We can’t use tree rings, because Joshua trees are monocots, and don’t have tree rings. One approach that has been useful for other plants that lack tree rings is to use radiocarbon dating to estimate the age of trees (Vieira et al. 2005). Could such an approach be used here?

3. Could the differences in size class data be used to estimate the rates of population growth and decline in different populations? What kind of additional information would we need?

4. Another common approach to studying population growth and population expansion is to use genetic data, such as microsatellites, to try to measure growth rates (Cornuet and Luikart 1996). Would such an approach be possible here?

Some Useful References:

Cornuet, J. M., and G. Luikart. 1996. Description and Power Analysis of Two Tests for Detecting Recent Population Bottlenecks From Allele Frequency Data. Genetics 144:2001-2014.

Darwin, C. 1874. Letter to J D Hooker, April 7, 1874 in F. Burkhardt, and S. Smith, eds. A Calendar of the Correspondence of Charles Darwin, 1821-1882 Cambridge, The Press Syndicate of the University of Cambridge.

Dole, K. P., M. E. Loik, and L. C. Sloan. 2003. The relative importance of climate change and the physiological effects of CO2 on freezing tolerance for the future distribution of Yucca brevifolia. Global and Planetary Change 36:137-146.

Fulton, M., and S. A. Hodges. 1999. Floral isolation between Aquilegia formosa and Aquilegia pubescens. Proceedings of the Royal Society of London Series B 266:2247-2252.

Godsoe, W. K. W., J. B. Yoder, C. I. Smith, and O. Pellmyr. 2008. Coevolution and divergence in the Joshua tree/yucca moth mutualism. American Naturalist 171:816-823.

Gotelli, N. J. 2001, A Primer of Ecology. Sunderland, MA, Sinauer.

Grant, V. 1952. Isolation and hybridization between Aquilegia formosa and A. pubescens. Aliso 2:341-360.

Hodges, S. A., and M. L. Arnold. 1994. Floral and ecological isolation between Aquilegia formosa and Aquilegia pubescens. Proceedings of the National Academy of Sciences 91:2493-2496.

Marr, D. L., and O. Pellmyr. 2003. Effect of pollinator-inflicted ovule damage on floral abscission in the yucca-yucca moth mutualism: the role of mechanical and chemical factors. Oecologia 136:236-243.

Pellmyr, O. 2003. Yuccas, yucca moths and coevolution: a review. Annals of the Missouri Botanical Garden 90:35-55.

Pellmyr, O., and K. A. Segraves. 2003. Pollinator divergence within an obligate mutualism: two yucca moth species (Lepidoptera; Prodoxidae: Tegeticula) on the Joshua Tree (Yucca brevifolia; Agavaceae). Annals of the Entomological Society of America 96:716-722.

Shafer, S. L., P. Bartlein, and R. S. Thompson. 2001. Potential changes in the distribution of western North American tree and shrub taxa under future climate scenarios. Ecosystems 4:200-215.

Smith, C. I., W. K. W. Godsoe, S. Tank, J. B. Yoder, and O. Pellmyr. 2008. Distinguishing coevolution from covicariance in an obligate pollination mutualism: Asynchronous divergence in Joshua tree and its pollinators. Evolution 62:2676-2687.

Vieira, S., S. Trumbore, P. B. Camargo, D. Selhorst, J. Q. Chambers, N. Higuchi, and L. A. Martinelli. 2005. Slow growth rates of Amazonian trees: Consequences for carbon cycling. Proceedings of the National Academy of Sciences of the United States of America 102:18502-18507.