Pollinator attraction to volatiles from virgin and pollinated host flowers in a yucca/moth obligate mutualism

The classic obligate pollination–seed consumption mutualism between yuccas and yucca moths has been thought to be mediated by chemical cues, but empirical data on pollinator attraction to host floral volatiles in this association have been lacking. Here we show that the scent from virgin flowers of the host Yucca glauca is sufficient to attract its obligate pollinator Tegeticula yuccasella in Y‐tube olfactometer tests. Interestingly, both sexes of moths were attracted to the scent stimulus. Because yucca moths mate inside host flowers, the attraction of both females and males to host floral volatiles is likely to increase encounter rates. In a second test, female moths did not discriminate between virgin and hand‐pollinated flowers, indicating no post‐pollination change in scent production by the host that would lead to a reduction in pollinator attraction and thereby limit exploitation of the available seeds in host flowers. However, other mechanisms that could stabilise the mutualism between T. yuccasella and its yucca hosts have already been documented, i.e. selective abortion of heavily infested flowers, and a female‐derived host‐marking pheromone. Headspace collection and GC–MS were used to identify the blend of floral volatiles emitted by Y. glauca, which was found to be very similar to those of two other allopatric capsular‐fruited species, Y. elata and Y. filamentosa, revealing strong conservation of this trait within Yucca section Chaenocarpa.     

The evolution of obligate pollination mutualisms: senita cactus and senita moth

We report a new obligate pollination mutualism involving the senita cactus, Lophocereus schottii (Cactaceae, Pachyceereae), and the senita moth, Upiga virescens (Pyralidae, Glaphyriinae) in the Sonoran Desert and discuss the evolution of specialized pollination mutualisms. L. schottii is a night-blooming, self-incompatible columnar cactus. Beginning at sunset, its flowers are visited by U. virescens females, which collect pollen on specialized abdominal scales, actively deposit pollen on flower stigmas, and oviposit a single egg on a flower petal. Larvae spend 6 days eating ovules before exiting the fruit and pupating in a cactus branch. Hand-pollination and pollinator exclusion experiments at our study site near Bahia Kino, Sonora, Mexico, revealed that fruit set in L. schottii is likely to be resource limited. About 50% of hand-outcrossed and open-pollinated senita flowers abort by day 6 after flower opening. Results of exclusion experiments indicated that senita moths accounted for 75% of open-pollinated fruit set in 1995 with two species of halictid bees accounting for the remaining fruit set. In 1996, flowers usually closed before sunrise, and senita moths accounted for at least 90% of open-pollinated fruit set. The net outcome of the senita/senita moth interaction is mutualistic, with senita larvae destroying about 30% of the seeds resulting from pollination by senita moths. Comparison of the senita system with the yucca/yucca moth mutualism reveals many similarities, including reduced nectar production, active pollination, and limited seed destruction. The independent evolution of many of the same features in the two systems suggests that a common pathway exists for the evolution of these highly specialized pollination mutualisms. Nocturnal flower opening, self-incompatible breeding systems, and resource-limited fruit production appear to be important during this evolution. Received: 19 August 1997 / Accepted: 24 November 1997     

Facultative non-mutualistic behaviour by an “Obligate” mutualist: “Cheating” by Yucca moths

The interaction between yucca moths (Tegeticula spp., Incurvariidae) and yuccas (Yucca spp., Agavaceae) is an obligate pollination/seed predation mutualism in which adult female yucca moths pollinate yuccas, and yucca moth larvae feed on yucca seeds. In this paper we document that individual yucca moths, which are capable of acting as mutualists, facultatively cheat by ovipositing in yucca pistils without attempting to transfer pollen. Additionally, a high proportion of flowers are unlikely to receive pollen even when pollination is attempted, because many yucca moths carry little or no pollen. The probability of occurrence of non-mutualistic behaviour is not affected by the amount of pollen a moth carries: moths with full pollen loads are just as likely to act non-mutualistically as moths carrying little or no pollen. We propose four hypotheses that could explain facultative non-mutualistic behaviour in yucca moths.Present address: Department of Biology, Simon Fraser University, Burnaby, British Columbia V5A 1S6, Canada     

Characterizing the interaction between the bogus yucca moth and yuccas: do bogus yucca moths impact yucca reproductive success?

Yucca moths are most well known for their obligate pollination mutualism with yuccas, where pollinator moths provide yuccas with pollen and, in exchange, the moth larvae feed on a subset of the developing yucca seeds. The pollinators, however, comprise only two of the three genera of yucca moths. Members of the third genus, Prodoxus, are the bogus yucca moths and are sister to the pollinator moths. Adult Prodoxus lack the specialized mouthparts used for pollination and the larvae feed on plant tissues other than seeds. Prodoxus larvae feed within the same plants as pollinator larvae and have the potential to influence yucca reproductive success directly by drawing resources away from flowers and fruit, or indirectly by modifying the costs of the mutualism with pollinators. We examined the interaction between the scape-feeding bogus yucca moth, Prodoxus decipiens, and one of its yucca hosts, Yucca filamentosa, by comparing female reproductive success of plants with and without moth larvae. We determined reproductive success by measuring a set of common reproductive traits such as flowering characteristics, seed set, and seed germination. In addition, we also quantified the percent total nitrogen in the seeds to determine whether the presence of larvae could potentially reduce seed quality. Flowering characteristics, seed set, and seed germination were not significantly different between plants with and without bogus yucca moth larvae. In contrast, the percent total nitrogen content of seeds was significantly lower in plants with P. decipiens larvae, and nitrogen content was negatively correlated with the number of larvae feeding within the inflorescence scape. Surveys of percent total nitrogen at three time periods during the flowering and fruiting of Y. filamentosa also showed that larval feeding decreased the amount of nitrogen in fruit tissue. Taken together, the results suggest that although P. decipiens influences nitrogen distribution in Y. filamentosa, this physiological effect does not appear to impact the female components of reproductive success.     

Pollination and seed predation by moths on Silene and allied Caryophyllaceae: evaluating a model system to study the evolution of mutualisms

Nursery pollinators, and the plants they use as hosts for offspring development, function as exemplary models of coevolutionary mutualism. The two pre-eminent examples--fig wasps and yucca moths--show little variation in the interaction: the primary pollinator is an obligate mutualist. By contrast, nursery pollination of certain Caryophyllaceae, including Silene spp., by two nocturnal moth genera, Hadena and Perizoma, ranges from antagonistic to potentially mutualistic, offering an opportunity to test hypotheses about the factors that promote or discourage the evolution of mutualism. Here, we review nursery pollination and host-plant interactions in over 30 caryophyllaceous plants, based on published studies and a survey of researchers investigating pollination, seed predation, and moth morphology and behavior. We detected little direct evidence of mutualism in these moth-plant interactions, but found traits and patterns in both that are nonetheless consistent with the evolution of mutualism and merit further attention.     

Coevolution and divergence in the Joshua tree/yucca moth mutualism

Theory suggests that coevolution drives diversification in obligate pollination mutualism, but it has been difficult to disentangle the effects of coevolution from other factors. We test the hypothesis that differential selection by two sister species of pollinating yucca moths (Tegeticula spp.) drove divergence between two varieties of the Joshua tree (Yucca brevifolia) by comparing measures of differentiation in floral and vegetative features. We show that floral features associated with pollination evolved more rapidly than vegetative features extrinsic to the interaction and that a key floral feature involved in the mutualism is more differentiated than any other and matches equivalent differences in the morphology of the pollinating moths. A phylogenetically based, ancestral states reconstruction shows that differences in moth morphology arose in the time since they first became associated with Joshua trees. These results suggest that coevolution, rather than extrinsic environmental factors, has driven divergence in this obligate pollination mutualism.     

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.     

Infertile seeds ofYucca schottii: A beneficial role for the plant in the yucca-yucca moth mutualism?

Summary The yucca-yucca moth interaction is a classic case of obligate mutualism. Female moths pollinate and oviposit in the gynoecium of the flower; however, maturing larvae eat a fraction of the developing seeds. We studied within-fruit distributions of four seed types (fertile, infertile, eaten and uneaten seeds) in order to evaluate costs and benefits in aYucca schottii population in southeastern Arizona. We focused on how the spatial arrangement of seeds affected larval behaviour and, hence, the costs of the mutualism to the yucca. Infertile seeds were distributed throughout both infested and uninfested locules. Additionally, moth larvae feeding in a single locule preferred fertile seeds and even avoided infertile seeds and left the fruit significantly more often when they encountered infertile seeds. We suggest that, regardless of the cause of infertile seeds, they function as blocking units within seed locules and therefore reduce seed predation by moth larvae. We also suggest that, together with certain other fruit traits, the presence of infertile seeds promotes the evolutionary stability of this pollination mutualism.     

Pollen dispersal in Yucca filamentosa (Agavaceae): the paradox of self-pollination behavior by Tegeticula yuccasella (Prodoxidae)

We investigated pollen dispersal in an obligate pollination mutualism between Yucca filamentosa and Tegeticula yuccasella. Yucca moths are the only documented pollinator of yuccas, and moth larvae feed solely on developing yucca seeds. The quality of pollination by a female moth affects larval survival because flowers receiving small amounts of pollen or self-pollen have a high abscission probability, and larvae die in abscised flowers. We tested the prediction that yucca moths primarily perform outcross pollinations by using fluorescent dye to track pollen dispersal in five populations of Y. filamentosa. Dye transfers within plants were common in all populations (mean ± 1 SE, 55 ± 3.0%), indicating that moths frequently deposit self-pollen. Distance of dye transfers ranged from 0 to 50 m, and the mean number of flowering plants between the pollen donor and recipient was 5 (median = 0), suggesting that most pollen was transferred among near neighbors. A multilocus genetic estimate of outcrossing based on seedlings matured from open-pollinated fruits at one site was 94 ± 6% (mean ± 1 SD). We discuss why moths frequently deposit self-pollen to the detriment of their offspring and compare the yucca-yucca moth interaction with other obligate pollinator mutualisms in which neither pollinator nor plant benefit from self-pollination.     

The evolutionary ecology of cheating: does superficial oviposition facilitate the evolution of a cheater yucca moth?

Abstract 1. A major question in the study of mutualism is to understand how mutualists may revert to antagonists that exploit the mutualism (i.e. switch to cheating). In the classic pollination mutualism between yuccas and yucca moths, the cheater moth Tegeticula intermedia is sister to the pollinator moth T. cassandra. These moth species have similar ovipositor morphology, but T. intermedia emerges later, oviposits into fruit rather than flowers, and does not pollinate. 2. We tested if the pollinator, T. cassandra, was pre‐adapted to evolve a cheater lineage by comparing its emergence and oviposition behaviour on yucca fruit to a distantly related pollinator, T. yuccasella, that differs in ovipositor morphology and oviposition behaviour. We predicted that if T. cassandra was pre‐adapted to cheat, then these pollinators would emerge later and be able to oviposit into fruit in contrast to T. yuccasella. 3. Contrary to expectations, a common garden‐rearing experiment demonstrated that emergence of T. cassandra was not significantly delayed relative to T. yuccasella. Moth emergence patterns overlapped broadly. 4. No choice oviposition experiments with female moths demonstrated that both pollinator species attempted to oviposit into fruit, but only T. cassandra was successful. Four out of 84 T. cassandra successfully oviposited into older fruit, whereas zero out of 79 T. yuccasella oviposited into older fruit. The rarity of the cheating behaviour in pollinators, however, meant that no significant difference in oviposition ability was detected. 5. The results suggest that a shift in emergence phenology is likely not a pre‐adaptation to the evolution of cheating, but that the ability to successfully lay eggs into fruit may be. The results also demonstrate that cheating attempts are rare in these pollinator species and, hence, the evolutionary transition rate from pollinator to cheater is likely to be low.     

Evolution of obligate pollination mutualism in New Caledonian Phyllanthus (Euphorbiaceae)

About half a dozen obligate pollination mutualisms between plants and their seed-consuming pollinators are currently recognized, including fig-fig wasp, yucca-yucca moth, and the recently discovered Glochidion tree-Epicephala moth mutualisms. A common principle among these interactions is that the pollinators consume only a limited amount of the seed crop within a developing fruit (or fig in the case of fig-fig wasp mutualism), thereby ensuring a net benefit to plant reproduction. A novel obligate, seed-parasitic pollination mutualism between two species of New Caledonian Phyllanthus (Euphorbiaceae), a close relative of Glochidion, and Epicephala moths (Gracillariidae) is an exception to this principle. The highly specialized flowers of Phyllanthus are actively and exclusively pollinated by species-specific Epicephala moths, whose larvae consume all six ovules of the developing fruit. Some flowers pollinated by the moths remain untouched, and thus a fraction of the fruits is left intact. Additional evidence for a similar association of Epicephala moths in other Phyllanthus species suggests that this interaction is a coevolved, species-specific pollination mutualism. Implications for the evolutionary stability of the system, as well as differences in mode of interaction with respect to the Glochidion-Epicephala mutualism, are discussed.     

Patterns of speciation in the Yucca moths: parallel species radiations within the Tegeticula yuccasella species complex

The interaction between yuccas and yucca moths has been central to understanding the origin and loss of obligate mutualism and mutualism reversal. Previous systematic research using mtDNA sequence data and characters associated with genitalic morphology revealed that a widespread pollinator species in the genus Tegeticula was in fact a complex of pollinator species that differed in host use and the placement of eggs into yucca flowers. Within this mutualistic clade two nonpollinating "cheater" species evolved. Cheaters feed on yucca seeds but lack the tentacular mouthparts necessary for yucca pollination. Previous work suggested that the species complex formed via a rapid radiation within the last several million years. In this study, we use an expanded mtDNA sequence data set and AFLP markers to examine the phylogenetic relationships among this rapidly diverging clade of moths and compare these relationships to patterns in genitalic morphology. Topologies obtained from analyses of the mtDNA and AFLP data differed significantly. Both data sets, however, corroborated the hypothesis of a rapid species radiation and suggested that there were likely two independent species radiations. Morphological analyses based on oviposition habit produced species groupings more similar to the AFLP topology than the mtDNA topology and suggested the two radiations coincided with differences in oviposition habit. The evolution of cheating was reaffirmed to have evolved twice and the closest pollinating relative for one cheater species was identified by both mtDNA and AFLP markers. For the other cheater species, however, the closest pollinating relative remains ambiguous, and mtDNA, AFLP, and morphological data suggest this cheater species may be diverged based on host use. Much of the divergence in the species complex can be explained by geographic isolation associated with the evolution of two oviposition habits.     

Patterns of genetic structure among populations of an oligophagous pollinating yucca moth (Tegeticula yuccasella)

Plant-insect associations have served as models for investigations of coevolution and the influence of biotic interactions on diversification. The pollination association between yuccas and yucca moths is a classic example of an obligate mutualism often suggested to have been affected by coevolution. Recent work has shown high host specificity in pollinating yucca moths, and here we use Tegeticula yuccasella, the species with the widest diet breadth, to ask how host specificity and isolation by distance contribute to specialization. Isolation by distance at a regional scale was observed in nucleotide variation within the mitochondrial gene cytochrome oxidase I (COI) (r =.294; P =.003). Host-related genetic structure (F(ct) = 0.08) was found to be slightly lower than the level of structure observed between eastern and western moth populations (F(ct) = 0.096). However, 56% of the COI haplotypes sampled from moths on Yucca filamentosa mapped to a host-specific clade in the haplotype network. Taken together, these results suggest that differentiation among T. yuccasella populations on alternative hosts is slight, but gene flow is influenced by both host association and geographic distance.     

HOST SPECIFICITY AND THE GENETIC STRUCTURE OF TWO YUCCA MOTH SPECIES IN A YUCCA HYBRID ZONE

Host specialization is an important mechanism of diversification among phytophagous insects, especially when they are tightly associated with their hosts. The well-known obligate pollination mutualism between yucca moths and yuccas represent such an association, but the degree of host specificity and modes of specialization in moth evolution is unclear. Here we use molecular tools to test the morphology-based hypothesis that the moths pollinating two yuccas, Yucca baccata and Y. schidigera, are distinct species. Host specificity was assessed in a zone of sympatry where the hosts are known to hybridize. Because the moths are the only pollinators, the plant hybrids are evidence that the moths occasionally perform heterospecific pollination. Nucleotide variation was assessed in a portion of the mitochondrial gene COI, and in an intron within a nuclear lysozyme gene. Moths pollinating Y. baccata and Y. schidigera were inferred to be genetically isolated because there was no overlap in alleles at either locus, and all but one of the moths was found on their native host in the hybrid zone. Moreover, genetic structure was very weak across the range of each moth species: estimates of F_ST for the lysozyme intron were 0.043 (SE = ± 0.004) and 0.021 (SE = ± 0.006) for the baccata and schidigera pollinators, respectively; estimated F_ST for COI in the baccata moths was 0.228 (± 0.012), whereas schidigera pollinators were fixed for a single allele. These results reveal a high level of migration among widely separated moth populations. We predict that pollen-mediated gene flow among conspecific yuccas is considerable and hypothesize that geographic separation is a limited barrier both for yuccas and for yucca moths.     

Variation in the costs and benefits of mutualism: the interaction between yuccas and yucca moths

Summary Yucca moths are both obligate pollinators and obligate seed predators of yuccas. I measured the costs and net benefits per fruit arising for eight species of yuccas from their interaction with the yucca moth Tegeticula yuccasella. Yucca moths decrease the production of viable seeds as a result of oviposition by adults and feeding by larvae. Oviposition through the ovary wall caused 2.3–28.6% of ovules per locule to fail to develop, leaving fruit with constrictions, and overall, 0.6–6.6% of ovules per fruit were lost to oviposition by yucca moths. Individual yucca moth larvae ate 18.0–43.6% of the ovules in a locule. However, because of the number of larvae per fruit and the proportion of viable seeds, yucca moth larvae consumed only 0.0–13.6% of potentially viable ovules per fruit. Given both oviposition and feeding effects, yucca moths decreased viable seed production by 0.6–19.5%. The ratio of costs to (gross) benefits varied from 0% to 30%, indicating that up to 30% of the benefits available to yuccas are subsequently lost to yucca moths. The costs are both lower and more variable than in a similar pollinator-seed predator mutualism involving figs and fig wasps.There were differences between species of yuccas in the costs of associating with yucca moths. Yuccas with baccate fruit experienced lower costs than species with capsular fruit. There were also differences in costs between populations within species and high variation in costs between fruit within populations. High variability was the result of no yucca moth larvae being present in over 50% of the fruit in some populations, while other fruit produced up to 24 larvae. I present hypotheses explaining both the absence and high numbers of larvae per fruit.     

RAD‐seq phylogenomics recovers a well‐resolved phylogeny of a rapid radiation of mutualistic and antagonistic yucca moths

The interaction between yucca plants and yucca moths has been one of the focal model systems investigated in the study of pollination mutualism and coevolution, especially in terms of understanding the prevention of overexploitation by mutualist partners. Yuccas have the ability to assess the number of eggs placed by pollinators into their ovaries, and can preferentially abort those flowers that would have many moth larvae consuming yucca seeds. Previous phylogenetic research identified a rapid radiation of moth species that corresponded with shifts in the interaction with their host plants. These shifts led to the evolution of moth species that circumvent the egg detection method used by yuccas to limit seed damage. In particular, some pollinator species deposit their eggs so that they are undetectable by the plants, whereas other species are ‘cheaters’ that have lost the ability to pollinate, yet deposit eggs into developing fruit rather than flowers. The evolution of these new species happened so quickly that the phylogeny of the moths has remained unresolved despite repeated attempts with standard Sanger sequencing of mtDNA loci and AFLP marker generation. Here, we use extensive analyses of RAD‐seq data to determine the phylogenetic relationships among yucca moth species. The results provide a robust phylogenetic framework that identifies the evolutionary relationships among shifts in egg‐laying strategies, as well as determining the closest pollinating relatives to the cheater species. Based on the obtained phylogeny, a shift in egg‐laying strategy that avoided the overexploitation regulatory mechanism used by yucca plants was a precursor for the evolution of two species with cheating behaviour.     

传粉细蛾与大戟科植物专性授粉的互惠共生体系

在已知的昆虫与植物所形成的专性授粉互惠共生体系中,榕树—榕小蜂、丝兰—丝兰蛾体系是经典实例,国内外学者已经从不同角度进行了大量的研究,为我们理解植物—传粉者互惠共生体系协同进化的机理和历史积累了宝贵的资料。近些年的研究发现鳞翅目细蛾科头细蛾属昆虫与大戟科植物之间也存在相似的协同进化关系。文章对国内外学者有关传粉细蛾与大戟科植物互惠共生协同进化的研究进行了整理。     

Florivore impacts on plant reproductive success and pollinator mortality in an obligate pollination mutualism

Florivores are present in many pollination systems and can have direct and indirect effects on both plants and pollinators. Although the impact of florivores are commonly examined in facultative pollination mutualisms, their effects on obligate mutualism remain relatively unstudied. Here, we used experimental manipulations and surveys of naturally occurring plants to assess the effect of florivory on the obligate pollination mutualism between yuccas and yucca moths. Yucca filamentosa (Agavaceae) is pollinated by the moth Tegeticula cassandra (Lepidoptera: Prodoxidae), and the mutualism also attracts two florivores: a generalist, the leaf-footed bug Leptoglossus phyllopus (Hemiptera: Coreidae), and a specialist, the beetle Hymenorus densus (Coleoptera: Tenebrionidae). Experimental manipulations of leaf-footed bug densities on side branches of Y. filamentosa inflorescences demonstrated that feeding causes floral abscission but does not reduce pollen or seed production in the remaining flowers. Similar to the leaf-footed bugs, experimental manipulations of beetle densities within individual flowers demonstrated that beetle feeding also causes floral abscission, but, in addition, the beetles also cause a significant reduction in pollen availability. Path analyses of phenotypic selection based on surveys of naturally occurring plants revealed temporal variation in the plant traits important to plant fitness and the effects of the florivores on fitness. Leaf-footed bugs negatively impacted fitness when fewer plants were flowering and leaf-footed bug density was high, whereas beetles had a positive effect on fitness when there were many plants flowering and their densities were low. This positive effect was likely due to adult beetles consuming yucca moth eggs while having a negligible effect on floral abscission. Together, the actions of both florivores either augmented the relationship of plant traits and fitness or slightly weakened the relationship. Overall, the results suggest that, although florivores are always present during flowering, the impact of florivores on phenotypic selection in yuccas is strongly mitigated by changes in their densities on plants from year to year. In contrast, both florivores consistently influenced pollinator larval mortality through floral abscission, and H. densus beetles additionally via the consumption of pollinator eggs.     

An interaction between a specialized seed predator moth and its dioecious host plant shifting from parasitism to mutualism

Seed predator/pollinator and host plant interactions, which may be considered as antagonistic, have the potential to provide good model systems for the study of the early stages of evolution towards mutualism. I describe a relationship between a seed predator, the geometrid moth Perizoma affinitatum , and the dioecious plant Silene dioica . The moth is an obligate seed predator on its host plant. The searching and ovipositing behaviour of the female moths, number of eggs deposited per flower, the pollinating ability of the moths and the seed consumption by the larvae are described as different parameters and studied in two Finnish coastal populations. A high pollinating ability and limited seed consumption by the predator was found and discussed in relation to fitness models of P. affinitatum and S. dioica . In a mutualistic relationship there must be a balance between the costs and benefits so that the seed production by the moths is larger than the seed consumption by the larvae, given a net seed output larger than zero. The data of the parameters included in a seed production/consumption model give a positive seed output when the proportion of S. dioica flowers pollinated by other non-predating insects is less than 60%. Accordingly, even if P. affinitatum would become the exclusive pollinator it would not endanger the survival of the host plant and both partners would benefit from this interaction. Limited seed consumption, high pollinating ability and host specificity as seen in the P. affinitatum and S. dioica interaction are considered to have been important pre-existing qualities in the evolution of the obligate mutualisms between yucca and yucca moths and fig and fig wasps. In isolated serpentine populations where the gene flow is restricted and co-pollinators are rare the interaction between P. affinitatum and S. dioica has the potential to shift from parasitism to mutualism.     

Reversal of mutualism in a leafflower–leafflower moth association: the possible driving role of a third‐party partner

A major goal in the study of mutualism is to understand how co‐operation is maintained when mutualism may potentially turn into parasitism. Although certain mechanisms facilitate the persistence of mutualism, parasitic species have repeatedly evolved from mutualistic ancestors. However, documented examples of mutualism reversals are still rare. Leafflowers (Phyllantheae; Phyllanthaceae) include approximately 500 species that engage in obligate mutualism with leafflower moths (Epicephala; Gracillariidae), which actively pollinate flowers, and whose larvae feed on the resulting seeds. We found that the Taiwanese population of the Phyllanthus reticulatus species complex was associated with six sympatric Epicephala species, of which three were derived parasites that induced gall formation on flowers/buds and produced no seeds. Notably, two parasitic species have retained mutualistic pollination behaviour, suggesting that the parasitism was likely not selected for to reduce the cost of mutualism. We propose that the galling habit evolved as an adaptation to escape parasitism by a specialized braconid wasp. The tough gall produced by one species was almost free of braconid parasitism, and the swollen gall induced by the other species probably prevents attack as a result of the larger airspace inside the gall. Our findings suggest that the presence of a third‐party partner can greatly influence the evolutionary fate of mutualisms, regardless of whether the pairwise interaction continues to favour co‐operation.