Re-evaluating the role of selective abscission in moth/yucca mutualisms

Conflicts of interest are common to mutualisms, particularly those derived from exploitative interactions. Conflicts of interest are particularly pronounced in pollination/seed predation mutualisms, such as moth/yucca interactions, where consumption of seeds by larvae of a plant's pollinator will raise the fitness of the pollinator but lower the fitness of the plant. A central question in these mutualisms is, therefore, “what limits seed predation?” If plants with excess flowers selectively abscise flowers containing many eggs, they may reduce seed predation and overall increase their fecundity. If eggs in abscised flowers die, selective abscission may additionally contribute to the limitation or regulation of pollinator populations, thereby decreasing the probability of future overexploitation. We examined the effect of selective abscission in the mutualism between Yucca kanabensis and one of its pollinating moths, Tegeticula altiplanella. Per capita mortality of moth eggs due to abscission was high (95.5%), but did not increase on inflorescences with more ovipositions per flower. Overall mortality was partitioned into two components based upon the proportion of visited flowers abscised (i.e. resource‐limitation) and additional mortality (=selective abscission). Resource‐limitation per se inflicted 93.9% egg mortality, or most of the mortality due to abscission. But, the average number of eggs in fruit was lower than the average number of eggs in flowers, indicating that there was some selectivity of abscission. However, neither source of mortality increased on inflorescences with more ovipositions per visited flower. Egg mortality resulting from selective abscission was not as high as possible, because the yuccas appeared to use oviposition‐damaged ovules as a cue for selective abscission, and there was considerable variation in the relationship between oviposition number and damaged ovules. However, even if yuccas had retained the flowers containing the fewest eggs, selective abscission still would not have been higher on inflorescences with more ovipositions per flower. Considering also that, 1) number of ovipositions is a poor predictor of the number of larvae that hatch and feed on the developing seeds in a fruit and that, 2) there are several moth/yucca interactions in which selective abscission does not occur, we conclude that abscission, and particular selective abscission, may have density‐limiting effects on moth populations, but will fail as general explanations for regulating the dynamics of moth populations.     

Geographic isolation trumps coevolution as a driver of yucca and yucca moth diversification

Coevolution is thought to be especially important in diversification of obligate mutualistic interactions such as the one between yuccas and pollinating yucca moths. We took a three-step approach to examine if plant and pollinator speciation events were likely driven by coevolution. First, we tested whether there has been co-speciation between yuccas and pollinator yucca moths in the genus Tegeticula (Prodoxidae). Second, we tested whether co-speciation also occurred between yuccas and commensalistic yucca moths in the genus Prodoxus (Prodoxidae) in which reciprocal evolutionary change is unlikely. Finally, we examined the current range distributions of yuccas in relationship to pollinator speciation events to determine if plant and moth speciation events likely occurred in sympatry or allopatry. Co-speciation analyses of yuccas with their coexisting Tegeticula pollinator and commensalistic Prodoxus lineages demonstrated phylogenetic congruence between both groups of moths and yuccas, even though moth lineages differ in the type of interaction with yuccas. Furthermore, Yucca species within a lineage occur primarily in allopatry rather than sympatry. We conclude that biogeographic factors are the overriding force in plant and pollinator moth speciation and significant phylogenetic congruence between the moth and plant lineages is likely due to shared biogeography rather than coevolution.     

Rapid evolution and specialization following host colonization in a yucca moth

Host shifts followed by specialization can result in sympatric genetic differentiation, and may have fuelled the diversification of phytophagous insects. This study examines a recent colonization of a non‐native host by Prodoxus quinquepunctellus (Lepidoptera: Prodoxidae). Allozyme differentiation was detected among different host feeding populations, yet was nearly absent among similar host feeding populations in sympatry. Geographical patterns of allozyme variation showed a much higher level of population structure among populations feeding on the derived host. Conversely, mtDNA haplotype frequencies were nearly homogeneous in the derived populations compared to the ancestral populations, suggesting a bottleneck and/or rapid fixation of haplotypes following host colonization. Moth emergence coincided with host plant flowering, and phenological differences between host species translated into allochronic isolation between populations feeding on different hosts. Derived moth populations also differed significantly in three ovipositor characters from ancestral populations. These findings suggest rapid host‐specific genetic differentiation, and specialization of moth emergence time and ovipositor morphology following host colonization.     

Cheating in mutualism: defection of Yucca baccata against its yucca moths

Yucca baccata cheats in its obligate pollination/seed predation mutualism with yucca moths. Although all individuals use the pollination services of yucca moths, many individuals do not reciprocate in sustaining yucca moth larvae. Cheating is associated with the morphology of Y. baccata pistils. In Y. baccata , the apex of the ovary contains only inviable ovules, and there are two distinct flower types, one of which has twice as many potentially viable ovules as the other. Because yucca moths oviposit at the apex of Y. baccata ovaries, larvae in flowers with few viable ovules fail to encounter viable ovules and therefore perish. Inflorescences generally have just one flower type, implying that some individuals cheat whereas others maintain the yucca moth population. Our most surprising observation, however, is that although the proportion of cheaters should be low, over 70% of Y. baccata individuals cheat. We hypothesize that both density- and frequency-dependent processes maintain a balance of cheaters and noncheaters in this system.     

Phylogeographic structure in the bogus yucca moth Prodoxus quinquepunctellus (Prodoxidae): comparisons with coexisting pollinator yucca moths

The pollination mutualism between yucca moths and yuccas highlights the potential importance of host plant specificity in insect diversification. Historically, one pollinator moth species, Tegeticula yuccasella, was believed to pollinate most yuccas. Recent phylogenetic studies have revealed that it is a complex of at least 13 distinct species, eight of which are specific to one yucca species. Moths in the closely related genus Prodoxus also specialize on yuccas, but they do not pollinate and their larvae feed on different plant parts. Previous research demonstrated that the geographically widespread Prodoxus quinquepunctellus can rapidly specialize to its host plants and may harbor hidden species diversity. We examined the phylogeographic structure of P. quinquepunctellus across its range to compare patterns of diversification with six coexisting pollinator yucca moth species. Morphometric and mtDNA cytochrome oxidase I sequence data indicated that P. quinquepunctellus as currently described contains two species. There was a deep division between moth populations in the eastern and the western United States, with limited sympatry in central Texas; these clades are considered separate species and are redescribed as P. decipiens and P. quinquepunctellus (sensu stricto), respectively. Sequence data also showed a lesser division within P. quinquepunctellus s.s. between the western populations on the Colorado Plateau and those elsewhere. The divergence among the three emerging lineages corresponded with major biogeographic provinces, whereas AMOVA indicated that host plant specialization has been relatively unimportant in diversification. In comparison, the six pollinator species comprise three lineages, one eastern and two western. A pollinator species endemic to the Colorado Plateau has evolved in both of the western lineages. The east-west division and the separate evolution of two Colorado Plateau pollinator species suggest that similar biogeographic factors have influenced diversification in both Tegeticula and Prodoxus. For the pollinators, however, each lineage has produced a monophagous species, a pattern not seen in P. quinquepunctellus.     

Phenotypic divergence and inter-specific trait correlation in a plant-pollinator/seed predator mutualism

Plant-pollinator interactions have been suggested as key drivers of morphological divergence and speciation of the involved taxa. These interactions can also promote sexual dimorphism in both the plant and pollinator, particularly if the pollinator is also a seed-eater and/or exerts different selection pressures on male and female plants. Here we tested the hypotheses that plant-pollinator interactions can be reflected in trait variation and sexual dimorphism in both organisms within and across populations. Across nine European populations, we examined intraspecific variation and sexual dimorphism in phenotypic traits potentially involved in the plant–insect interaction of the dioecious white campion Silene latifolia (Caryophyllaceae) and its specialist pollinator Hadena bicruris (Noctuidae). This interaction is expected to entail sex-specific selective pressures, as female moths lay eggs on female plants and the larvae predate on the seeds during their development. We compared divergence in phenotypic traits among populations and between sexes within populations, examined correlations between plant and pollinator traits, and between phenotypic distances and genetic distances among co-occurring populations for both plants and insects. We found key differences in phenotypic traits across populations of both the plant and moth, though only in the moth were these differences correlated with geographic distances. We also found evidence for sexual dimorphism in the plant but not in the pollinator. Evolution of floral sexual dimorphism in S. latifolia most likely results from the joint contribution of different selective forces, including biotic interactions with H. bicruris moths.     

Genetic consequences of specialization: yucca moth behavior and self-pollination in yuccas

Reciprocal specialization in interspecific interactions, such as plant-pollinator mutualisms, increases the probability that either party can have detrimental effects on the other without the interaction being dissolved. This should be particularly apparent in obligate mutualisms, such as those that exist between yucca and yucca moths. Female moths collect pollen from yucca flowers, oviposit into floral ovaries, and then pollinate those flowers. Yucca moths, which are the sole pollinators of yuccas, impose a cost in the form of seed consumption by the moth larvae. Here we ask whether there also is a genetic cost through selfish moth behavior that may lead to high levels of self fertilization in the yuccas. Historically, it has been assumed that females leave a plant immediately after collecting pollen, but few data are available. Observations of a member of the Tegeticula yuccasella complex on Yucca filamentosa revealed that females remained on the plant and oviposited in 66% of all instances after observed pollen collections, and 51% of all moths were observed to pollinate the same plant as well. Manual cross and self pollinations showed equal development and retention of fruits. Subsequent trials to assess inbreeding depression by measuring seed weight, germination date, growth rate, and plant mass at 5 months revealed significant negative effects on seed weight and germination frequency in selfed progeny arrays. Cumulative inbreeding depression was 0.475, i.e., fitness of selfed seeds was expected to be less than half that of outcrossed seeds. Single and multilocus estimates of outcrossing rates based on allozyme analyses of open-pollinated progeny arrays did not differ from 1.0. The discrepancy between high levels of behavioral self-pollination by the moths and nearly complete outcrossing in mature seeds can be explained through selective foreign pollen use by the females, or, more likely, pollen competition or selective abortion of self-pollinated flowers during early stages of fruit development. Thus, whenever the proportion of pollinated flowers exceeds the proportion that can be matured to ripe fruit based on resource availability, the potential detrimental genetic effects imposed through geitonogamous pollinations can be avoided in the plants. Because self-pollinated flowers have a lower probability of retention, selection should act on female moths to move among plants whenever moth density is high enough to trigger abortion. Received: 18 March 1996 \Accepted: 30 July 1996     

Comparative phylogeography of a coevolved community: concerted population expansions in Joshua trees and four yucca moths

Comparative phylogeographic studies have had mixed success in identifying common phylogeographic patterns among co-distributed organisms. Whereas some have found broadly similar patterns across a diverse array of taxa, others have found that the histories of different species are more idiosyncratic than congruent. The variation in the results of comparative phylogeographic studies could indicate that the extent to which sympatrically-distributed organisms share common biogeographic histories varies depending on the strength and specificity of ecological interactions between them. To test this hypothesis, we examined demographic and phylogeographic patterns in a highly specialized, coevolved community--Joshua trees (Yucca brevifolia) and their associated yucca moths. This tightly-integrated, mutually interdependent community is known to have experienced significant range changes at the end of the last glacial period, so there is a strong a priori expectation that these organisms will show common signatures of demographic and distributional changes over time. Using a database of >5000 GPS records for Joshua trees, and multi-locus DNA sequence data from the Joshua tree and four species of yucca moth, we combined paleaodistribution modeling with coalescent-based analyses of demographic and phylgeographic history. We extensively evaluated the power of our methods to infer past population size and distributional changes by evaluating the effect of different inference procedures on our results, comparing our palaeodistribution models to Pleistocene-aged packrat midden records, and simulating DNA sequence data under a variety of alternative demographic histories. Together the results indicate that these organisms have shared a common history of population expansion, and that these expansions were broadly coincident in time. However, contrary to our expectations, none of our analyses indicated significant range or population size reductions at the end of the last glacial period, and the inferred demographic changes substantially predate Holocene climate changes.     

Steroidal sapogenins. LVIII. Steroidal sapogenins from the Joshua tree

The Joshua tree,Yucca brevifolia, is distributed over parts of Arizona, California, Nevada, and Utah. A study of its steroidal sapogenins showed that the leaves contain practically none; the wood, small amounts of smilagenin; the seeds high amounts of hecogenin and tigogenin.     

Phylogenetic origins of Lophocereus (Cactaceae) and the senita cactus-senita moth pollination mutualism

Recent ecological research has revealed that the Sonoran Desert columnar cactus Lophocereus and the pyralid moth Upiga virescens form an obligate pollination mutualism, a rare but important case of coevolution. To investigate the phylogenetic origins of this unusual pollination system, we used molecular sequence data to reconstruct the phylogeny of the four taxa within the genus Lophocereus and to determine the phylogenetic position of Lophocereus within the North American columnar cacti (tribe Pachycereeae). Our analysis included Lophocereus, six Pachycereus species, Carnegiea gigantea, and Neobuxbaumia tetetzo within the subtribe Pachycereinae, and Stenocereus thurberi as an outgroup within the Stenocereinae. Extensive screening of chloroplast and mitochondrial genomes failed to reveal sequence variation within Lophocereus. At a deeper phylogenetic level, however, we found strong support for the placement of Lophocereus within Pachycereus as sister group to the hummingbird-pollinated P. marginatus. We discuss possible hypotheses that may explain the transition from bat pollination (ancestral) to moth and hummingbird pollination in Lophocereus and P. marginatus, respectively. Additional phylogenetic analyses suggest that the genus Pachycereus should be expanded to include Lophocereus, Carnegiea, Neobuxbaumia, and perhaps other species, whereas P. hollianus may need to be excluded from this clade. Future study will be needed to test taxonomic distinctions within Lophocereus, to test for parallel cladogenesis between phylogroups within Lophocereus and Upiga, and to fully delineate the genus Pachycereus and relationships among genera in the Pachycereinae.     

Leaf domatia mediate mutualism between mites and a tropical tree

Although associations between mites and leaf domatia have been widely reported, their consequences for plants, especially for natural tree populations, particularly in the tropics, are largely unknown. In experiments with paired Cupania vernalis (Sapindaceae) saplings in a semi-deciduous forest in south-east Brazil, we blocked leaf domatia to examine their effect: (1) on mites and other arthropods, and (2) on damage caused by fungi and herbivorous arthropods. In general, plants with resin-blocked domatia had fewer predaceous mites on leaves than control plants with unaltered domatia, but the total abundances of fungivorous and of phytophagous mites remained unchanged. However, phytophagous eriophyid mites, the most numerous inhabitants of domatia, decreased on leaf surfaces with the blocking treatment. In a second experiment, treated plants lacking functional domatia developed significantly greater numbers and areas of chlorosis, apparently due to increased eriophyid attacks, whereas fungal attack, epiphyll abundance and leaf-area loss were unaffected. This seems to be the first experimental study to demonstrate that leaf domatia can benefit plants against herbivory in a natural system. The possible stabilizing effect of leaf domatia on predator-prey interactions is discussed.     

Host specificity and reproductive success of yucca moths (Tegeticula spp. Lepidoptera: Prodoxidae) mirror patterns of gene flow between host plant varieties of the Joshua tree (Yucca brevifolia: Agavaceae)

Coevolution between flowering plants and their pollinators is thought to have generated much of the diversity of life on Earth, but the population processes that may have produced these macroevolutionary patterns remain unclear. Mathematical models of coevolution in obligate pollination mutualisms suggest that phenotype matching between plants and their pollinators can generate reproductive isolation. Here, we test this hypothesis using a natural experiment that examines the role of natural selection on phenotype matching between yuccas and yucca moths (Tegeticula spp.) in mediating reproductive isolation between two varieties of Joshua tree (Yucca brevifolia var. brevifolia and Y. brevifolia var. jaegeriana). Using passive monitoring techniques, DNA barcoding, microsatellite DNA genotyping, and sibship reconstruction, we track host specificity and the fitness consequences of host choice in a zone of sympatry. We show that the two moth species differ in their degree of host specificity and that oviposition on a foreign host plant results in the production of fewer offspring. This difference in host specificity between the two moth species mirrors patterns of chloroplast introgression from west to east between host varieties, suggesting that natural selection acting on pollinator phenotypes mediates gene flow and reproductive isolation between Joshua‐tree varieties.     

Defection by plants in the yucca–yucca moth association: a test of the cheater plant hypothesis for Yucca treculeana

In mutualisms, an underlying conflict of interests may select for defection from providing benefits. In the obligate mutualism between yuccas and yucca moths, where pollination service and seeds for pollinator larvae are traded, it has been suggested that some individuals in a population of Y. baccata may defect by preventing pollinator egg or larvae from development. We tested this hypothesis in Y. treculeana , another species suggested to contain cheater plants. Five specific predictions were tested during two years of study. A prediction that a surplus of plants without pollinator larvae should be present was met. Predicted existence of two distinct fruit morphs was rejected, and none of several highly variable morphological traits were linked to presence/absence of larvae. Predicted excess of intact seeds in the fruits of plants without larvae was not found; in fact, such plants produced fewer seeds, contrary to the hypothesis. A suggestion that inverse frequency-dependent fitness could explain the pattern was rejected. Contrary to prediction, distribution of larvae of a closely related cheater yucca moth was positively associated with pollinator larvae, even though it would not be affected by the proposed killing mechanism. The results together provide strong support against the existence of cheater plants in Y. treculeana .     

The influence of interaction type and feeding location on the phylogeographic structure of the yucca moth community associated with Hesperoyucca whipplei

The interactions between herbivorous insects and their host plants have been central in generating diversification in both groups. We used a community of four yucca moth species, monophagous on the host plant Hesperoyucca whipplei (Agavaceae), to examine how the type of interaction and where insects feed within a plant influence phylogeographic structure of herbivorous insects. These four species included two fruit-feeders, one mutualistic and one commensalistic, and two commensalistic stalk-feeders. Surveys based on mtDNA cytochrome oxidase I sequence data demonstrated that the moth species differed in phylogeographic history. Populations of the mutualist pollinator, Tegeticula maculata, exhibited the most subdivision in comparison to the three commensal Prodoxus species (both genera in Lepidoptera, Prodoxidae). Feeding location was also correlated with differences in phylogeographic history through its influence on population sizes and the probability of gene flow. The results suggest that both the outcome of interactions and where insects feed may influence population structure.     

Functional divergence of spatially conserved olfactory glomeruli in two related moth species

In different moth species, the number and spatial arrangement of olfactory glomeruli in the antennal lobe (AL) vary widely, but the spatial map within a species is thought to be invariant, making it possible to identify single glomeruli across individuals. We investigated the relationship between the physiological tuning of pheromone-selective interneurons and their association with specific, identified glomeruli in the macroglomerular complex (MGC) of the noctuid moth, Heliothis subflexa. Three odorants that are required for pheromone-source location in this species were tested individually and in blends. Recordings from 27 pheromone-specific projection neurons (PNs) indicated that the majority (48%) were selectively activated by the major pheromone component of this species, Z-11-hexadecenal (Z11-16:Ald), with 33% primarily tuned to Z-9-hexadecenal and 19% to Z-11-hexadecenol. Intracellular staining revealed that the dendrites of PNs tuned to Z11-16:Ald always branched within the largest glomerulus of the MGC, the cumulus. Similarly, each of the other two classes of PN was associated with a different 'satellite' glomerulus in the MGC. The spatial configuration of the four-glomerulus H. subflexa MGC was indistinguishable from that previously reported in the closely related species, Heliothis virescens. Hence, as these two species diverged, changes in the association of satellite MGC glomeruli with particular odorants have occurred without a measurable change in the anatomical arrangement of the glomerular array.     

Calibrated tree priors for relaxed phylogenetics and divergence time estimation

The use of fossil evidence to calibrate divergence time estimation has a long history. More recently, Bayesian Markov chain Monte Carlo has become the dominant method of divergence time estimation, and fossil evidence has been reinterpreted as the specification of prior distributions on the divergence times of calibration nodes. These so-called "soft calibrations" have become widely used but the statistical properties of calibrated tree priors in a Bayesian setting hashave not been carefully investigated. Here, we clarify that calibration densities, such as those defined in BEAST 1.5, do not represent the marginal prior distribution of the calibration node. We illustrate this with a number of analytical results on small trees. We also describe an alternative construction for a calibrated Yule prior on trees that allows direct specification of the marginal prior distribution of the calibrated divergence time, with or without the restriction of monophyly. This method requires the computation of the Yule prior conditional on the height of the divergence being calibrated. Unfortunately, a practical solution for multiple calibrations remains elusive. Our results suggest that direct estimation of the prior induced by specifying multiple calibration densities should be a prerequisite of any divergence time dating analysis.     

Effect of pollinator-inflicted ovule damage on floral abscission in the yucca-yucca moth mutualism: the role of mechanical and chemical factors

The long-term persistence of obligate mutualisms (over 40 Mya in both fig/fig wasps and yucca/yucca moths) raises the question of how one species limits exploitation by the other species, even though there is selection pressure on individuals to maximize fitness. In the case of yuccas, moths serve as the plant's only pollinator, but eggs laid by the moths before pollination hatch into larvae that consume seeds. Previous studies have shown that flowers with high egg loads are more likely to abscise. This suggests that yucca flowers can select against moths that lay many eggs per flower through selective abscission of flowers; however, it is not known how yucca moths trigger floral abscission. We tested how the moth Tegeticula yuccasella triggers floral abscission during oviposition in Yucca filamentosa by examining the effects of ovipositor insertion and egg laying on ovule viability and floral abscission. Eggs are not laid at the site of ovipositor insertion: we used this separation to test whether wounded ovules were more closely associated with the ovipositor site or an egg's location. Using a tetrazolium stain to detect injured ovules, we determined whether the number of ovipositions affected the number of wounded ovules in naturally pollinated flowers. Two wounding experiments were used to test the effect of mechanical damage on the probability of floral abscission. The types of wounds in these experiments mimicked two types of oviposition-superficial oviposition in the ovary wall and oviposition into the locular cavity-that have been observed in species of Tegeticula. The effect of moth eggs on ovule viability was experimentally tested by culturing ovules in vitro, placing moth eggs on the ovules, and measuring changes in ovule viability with a tetrazolium stain. We found that ovules were physically wounded during natural oviposition. Ovules showed a visible wounding response in moth-pollinated flowers collected 7-12 h after oviposition. Exact location of wounded ovules relative to eggs and oviposition scars, as well as results from the artificial wounding experiments, showed that the moth ovipositor inflicts mechanical damage on the ovules. Significantly higher abscission rates were observed in artificially wounded flowers in which only 4-8% of the ovules were injured. Eggs did not affect ovule viability as measured by the tetrazolium stain. These results suggest that physical damage to ovules caused by ovipositing is sufficient to explain selective fruit abscission. Whether injury as a mechanism of selective abscission in yuccas is novel or a preadaptation will require further study.