Evolution of Cooperation

Abstract. The ant-tended Australian butterfly, Jalmenus evagoras, has been a model system for studying the ecology and evolution of mutualism. A phylogeographic analysis of mitochondrial DNA cytochrome oxidase I sequences from 242 butterflies (615 bp) and 66 attendant ants (585 bp) from 22 populations was carried out to explore the relationship between ant association and butterfly population structure. This analysis revealed 12 closely related butterfly haplotypes in three distinct clades roughly corresponding to three allopatric subpopulations of the butterflies. Minimal genetic diversity and widespread haplotypes within biogeographical regions suggest high levels of matrilineal gene flow. Attendant ants are significantly more diverse than was previously thought, with at least seven well-defined clades corresponding to independent morphological determinations, distributed throughout the range of the butterflies. Nested analysis of molecular variance showed that biogeography, host plant, and ant associate all contribute significantly in explaining variation in butterfly genetic diversity, but these variables are not independent of one another. Major influences appear to come from fragmentation due to large-scale biogeographical barriers, and diversification following a shift in habitat preference. A consequence of such a shift could be codiversification of the butterfly with habitatadapted ants, resulting in apparent phylogenetic concordance between butterflies and ants. The implications of these results are discussed in terms of possible effects of ant attendance on the diversification of Lycaenidae as a whole.

We investigate the evolution of host association in a cryptic complex of mutualistic Crematogaster (Decacrema) ants that inhabits and defends Macaranga trees in Southeast Asia. Previous phylogenetic studies based on limited samplings of Decacrema present conflicting reconstructions of the evolutionary history of the association, inferring both cospeciation and the predominance of host shifts. We use cytochrome oxidase I (COI) to reconstruct phylogenetic relationships in a comprehensive sampling of the Decacrema inhabitants of Macaranga. Using a published Macaranga phylogeny, we test whether the ants and plants have cospeciated. The COI phylogeny reveals 10 well-supported lineages and an absence of cospeciation. Host shifts, however, have been constrained by stem traits that are themselves correlated with Macaranga phylogeny. Earlier lineages of Decacrema exclusively inhabit waxy stems, a basal state in the Pachystemon clade within Macaranga, whereas younger species of Pachystemon, characterized by nonwaxy stems, are inhabited only by younger lineages of Decacrema. Despite the absence of cospeciation, the correlated succession of stem texture in both phylogenies suggests that Decacrema and Pachystemon have diversified in association, or codiversified. Subsequent to the colonization of the Pachystemon clade, Decacrema expanded onto a second clade within Macaranga, inducing the development of myrmecophytism in the Pruinosae group. Confinement to the aseasonal wet climate zone of western Malesia suggests myrmecophytic Macaranga are no older than the wet forest community in Southeast Asia, estimated to be about 20 million years old (early Miocene). Our calculation of COI divergence rates from several published arthropod studies that relied on tenable calibrations indicates a generally conserved rate of approximately 1.5% per million years. Applying this rate to a rate-smoothed Bayesian chronogram of the ants, the Decacrema from Macaranga are inferred to be at least 12 million years old (mid-Miocene). However, using the extremes of rate variation in COI produces an age as recent as 6 million years. Our inferred timeline based on 1.5% per million years concurs with independent biogeographical events in the region reconstructed from palynological data, thus suggesting that the evolutionary histories of Decacrema and their Pachystemon hosts have been contemporaneous since the mid-Miocene. The evolution of myrmecophytism enabled Macaranga to radiate into enemy-free space, while the ants' diversification has been shaped by stem traits, host specialization, and geographic factors. We discuss the possibility that the ancient and exclusive association between Decacrema and Macaranga was facilitated by an impoverished diversity of myrmecophytes and phytoecious (obligately plant inhabiting) ants in the region.

Large blue (Maculinea) butterflies are highly endangered throughout the Palaearctic region, and have been the focus of intense conservation research(1-3). In addition, their extraordinary parasitic lifestyles make them ideal for studies of life history evolution. Early instars consume flower buds of specific host plants, but later instars live in ant nests where they either devour the brood (predators), or are fed mouth-to-mouth by the adult ants (cuckoos). Here we present the phylogeny for the group, which shows that it is a monophyletic clade nested within Phengaris, a rare Oriental genus whose species have similar life histories(4,5). Cuckoo species are likely to have evolved from predatory ancestors. As early as five million years ago, two Maculinea clades diverged, leading to the different parasitic strategies seen in the genus today. Contrary to current belief, the two recognized cuckoo species show little genetic divergence and are probably a single ecologically differentiated species(6-10). On the other hand, some of the predatory morphospecies exhibit considerable genetic divergence and may contain cryptic species. These findings have important implications for conservation and reintroduction efforts.

The estimated 6000 species of Lycaenidae account for about one third of all Papilionoidea. The majority of lycaenids have associations with ants that can be facultative or obligate and range from mutualism to parasitism. Lycaenid larvae and pupae employ complex chemical and acoustical signals to manipulate ants. Cost/benefit analyses have demonstrated multiple trade-offs involved in myrmecophily. Both demographic and phylogenetic evidence indicate that ant association has shaped the evolution of obligately associated groups. Parasitism typically arises from mutualism with ants, arid entomophagous species are disproportionately common in the Lycaenidae compared with other Lepidoptera. Obligate associations are more common in the Southern Hemisphere, in part because highly ant-associated lineages make up a larger proportion of the fauna in these, regions. Further research on phylogeny and natural history, particularly of the Neotropical fauna, will be necessary to understand the rote ant association has played in the evolution of the Lycaenidae.

Butterflies in the family Lycaenidae that have obligate associations with ants frequently exhibit ant-dependent egg laying behaviour. In a series of field and laboratory choice tests, we assessed oviposition preference of the Australian lycaenid Jalmenus evagoras in response to different species and populations of ants. Females discriminated between attendant and nonattendant ant species, between attendant ant species, and to some extent, between populations of a single ant species. When preferences were found, ovipositing butterflies preferred their locally predominant attendant ant species and geographically proximate attendant ant populations. A reciprocal choice test using adults from a generation of butterflies reared in the absence of ants indicated a genetic component to oviposition preference. Individual females were flexible with respect to oviposition site choice, often ovipositing on more than one treatment during a trial. Preferences arose from a hierarchical ranking of ant treatments. These results are discussed in terms of local adaptation and its possible significance in the diversification of ant-associated lycaenids.

We assessed the quality of different ant species as partners of the facultatively myrmecophilous lycaenid butterfly Glaucopsyche lygdamus. We compared disappearance and parasitism rates of G. lygdamus larvae in the field, and development of non-feeding pre-pupae in the laboratory, when individuals were untended or tended by one of four ant species. Formica podzolica was the only ant species to provide a clear benefit to G. lygdamus, in the form of reduced larval parasitism relative to untended larvae. F. 'neogagates' (F. neogagates + F. lasioides) and Tapinoma sessile were essentially neutral partners, providing no significant cost or benefit for any of the parameters measured. Relative to untended individuals, association with F. obscuripes significantly increased larval disappearance and significantly decreased pupal mass. Thus, F. obscuripes may act as a parasite of the general association between G. lygdamus and ants under certain conditions. Ant species also differed in their persistence as tenders of G. lygdamus larvae once an interaction was established. Over the lifetime of a larva, F. podzolica and F. obscuripes usually remained as the attendant ant species on plants over consecutive census dates, while F. 'neogagates' and T. sessile were frequently replaced, most commonly by F. obscuripes. It remains to be determined if disappearance and developmental outcomes reported here reflect true fitness costs (i.e. reduced survivorship and lower reproductive success) for G. lygdamus. The potential and limitations for specialization in association between G. lygdamus and high quality ant partners are discussed.

Ecological theory has long supported the idea that species coexistence in a homogeneous habitat is promoted by spatial structure, but empirical evidence for this hypothesis has lagged behind theory. Here we describe a Neotropical ant-plant symbiosis that is ideally suited for testing spatial models of coexistence. Two genera of ants, Allomerus cf. demerarae and three species of Azteca are specialized to live on a single species of ant-plant, Cordia nodosa, in a Western Amazonian tropical rain forest. Empirically, using census data from widely separated localities, we show that the relative colonization abilities of the two ant genera are a function of plant density. A parameterized model shows that this pattern alone is sufficiently robust to explain coexistence in;the system. Census and experimental data suggest that Azteca queens are better long-distance flyers, but that Allomerus colonies are more fecund. Thus, Azteca can dominate in areas where host-plant densities are low land parent colony-sapling distances are long), and Allomerus can dominate in areas where host-plant densities are high. Existing spatial heterogeneity in host-plant densities therefore can allow regional coexistence, and intersite dispersal can produce local mixing. In conclusion, a dispersal-fecundity trade-off appears to allow the two genera to treat spatial heterogeneity in patch density as a niche axis. This study further suggests that a spatially structured approach is essential in understanding the persistence of some mutualisms in the presence of parasites.

This study examines the pattern of opsin nucleotide and amino acid substitution among mimetic species 'rings' of Heliconius butterflies that are characterized by divergent wing colour patterns. A long wavelength opsin gene, OPS1, was sequenced from each of seven species of Heliconius and one species of Dryas (Lepidoptera: Nymphalidae). A parsimony analysis of OPS1 nucleotide and amino acid sequences resulted in a phylogeny that was consistent with that presented by Brewer & Egan in 1997, which was based on mitochondrial cytochrome oxidase I and II as well as nuclear wingless genes. Nodes in the OPS1 phylogeny were well supported by bootstrap analysis and decay indices. An analysis of specific sites within the gene indicates that the accumulation of amino acid substitutions has occurred independently of the morphological diversification of Heliconius wing colour patterns. Amino acid substitutions were examined with respect to their location within the opsin protein and their possible interactions with the chromophore and the G-protein. Of the 15 amino acid substitutions identified among the eight species, one nonconservative replacement (A226Q) was identified in a position that may be involved in binding with the G-protein. (C) 2001 The Linnean Society of London.

Juveniles of the Australian common imperial blue butterfly, Jalmenus evagoras, produce substrate-borne vibrational signals in the form of two kinds of pupal calls and three larval calls. Pupae stridulate in the presence of conspecific larvae, when attended by an ant guard, and as a reaction against perturbation. Using pupal pairs in which one member was experimentally muted, pupal calls were shown to be important in ant attraction and the maintenance of an ant guard. A pupa may use-calls to regulate levels of its attendant ants and to signal its potential value in these mutualistic interactions. Therefore substrate-borne vibrations play a significant role in the communication between J. evagoras and its attendant ants and pupal calls appear to be more than just signals acting as a predator deterrent. Similarly, caterpillars make more sound when attended by Iridomyrmex anceps, suggesting that larval calls may be important in mediating ant symbioses. One larval call has the same mean dominant frequency, pulse rate, bandwidth and pulse length as the primary signal of a pupa, suggesting a similarity in function. (C) 2000 The Association for the Study of Animal Behaviour.

If a mutualistic relationship entails providing services at a cost, selection will favor individuals that maximize the net benefits of the interaction and minimize the costs. Larvae of many species of lycaenid butterflies secrete nutritious food rewards to attending ants and, in return, receive protection against predators and parasitoids. Because ants typically recruit more workers to larger resources, by forming groups the larvae may ensure more reliable access to ants and thereby gain better protection. A further consequence of aggregating should be a change of the cost-benefit relationship for individual larvae. The larger the group, the smaller a single larva's influence will be on total ant density, which could lead to a smaller investment in secretion, thus reducing the per capita cost of cooperation. In this study, the influence of ant attendance, group size, and companion quality on larval investment was investigated. The interaction between the obligately ant-dependent lycaenid, Jalmenus evagoras, and its attendant Iridomyrmex ants was manipulated and the effect on larval secretion measured. As the level of ant attendance increased, the delivery of food rewards increased, both for solitary and for aggregated larvae. When aggregated, larvae provided less food rewards to ants than when solitary, and secretion rate decreased with increasing group size. Furthermore, larvae had lower secretion rates when paired with a bigger, more attractive larva than when paired with a smaller one. The considerable reduction in secretion rates for larvae in groups suggests that gaining protection at a lower secretion cost could be one factor that promotes aggregation in myrmecophilous lycaenids.

Exploring the factors governing the maintenance and breakdown of cooperation between mutualists is an intriguing and enduring problem for evolutionary ecology, and symbioses between ants and plants can provide useful experimental models for such studies. Hundreds of tropical plant species have evolved structures to house and feed ants, and these ant-plant symbioses have long been considered classic examples of mutualism. Here, we report that the primary ant symbiont, Allomerus cf. demerarae, of the most abundant ant-plant found in south-east Peru, Cordia nodosa Lam., castrates its host plant. Allomerus workers protect new leaves and their associated domatia from herbivory, but destroy flowers, reducing fruit production to zero in most host plants. Castrated plants occupied by Allomerus provide more domatia for their associated ants than plants occupied by three species of Azteca ants that do not castrate their hosts. Allomerus colonies in larger plants have higher fecundity. As a consequence, Allomerus appears to benefit from its castration behaviour, to the detriment of C. nodosa. The C. nodosa-ant system exhibits none of the retaliatory or filtering mechanisms shown to stabilize cheating in other cooperative systems, and appears to persist because some of the plants, albeit a small;minority, are inhabited by the three species of truly mutualistic Azteca ants.

Populations of the myrmecophilous lycaenid Falmenus evagoras Donovan were assessed for genetic structure at three hierarchical spatial scales: sites, geographically-defined subpopulations, and subpopulations defined by species of mutualistic ant-associate. Estimates of Wright's F-ST generated from multilocus electrophoretic data revealed low, though significant, amounts of genetic structure. Most structure was observed at the level of geographic subpopulations, suggesting that adult butterflies do not exhibit preferential mating and oviposition along the lines of ant associate. The genetic structure data, together with estimates of Nei's genetic distance (D) for pairwise site and subpopulation comparisons, suggest that F. evagoras populations are spatially and temporally dynamic. These patterns are considered in the context of extinction and recolonization models. The extreme patchiness of F. evagoras populations stems from the stringent requirements of both host plant and host ant, contributing to an extinction/ recolonization process. We discuss the key parameters influencing genetic cohesion versus differentiation under an extinction/recolonization regime, including mode of butterfly dispersal, site turnover rate, and the effects of host dispersal and phenology. This system provides a model of population-level consequences of certain mutualistic interactions as well as of a class of patterns arising from an extinction/recolonization process. (C) 1996 The Linnean Society of London

Abstract. Larvae of 17 Lycaenidae butterfly species from Europe, North America, South East Asia and Australia were observed to retain at least some of their adaptations related to myrmecophily even after parasitic braconid larvae have emerged from them. The myrmecophilous glandular organs and vibratory muscles of such larval carcasses remain functional for up to 8 days. The cuticle of lycaenid larvae contains extractable “adoption substances” which elicit antennal drumming in their tending ants. These adoption substances, as well, appear to persist in a functional state beyond parasitoid emergence, and the larval carcasses are hence tended much like healthy caterpillars. In all examples, the braconids may receive selective advantages through myrmecophily of their host larvae, instead of being suppressed by the ant guard. Interactions where parasitoids exploit the ant-mutualism of their lycaenid hosts have as yet been recorded only from the Apanteles group in the BraconidaeMicrogasterinae.

Moths and butterflies whose larvae do not feed on plants represent a decided minority slice of lepidopteran diversity, yet offer insights into the ecology and evolution of feeding habits. This paper summarizes the life histories of the known predatory and parasitic lepidopteran taxa, focusing in detail on current researchin the butterfly family Lycaenidae, a group disprotionately rich in aphytophagous feeders and myrmecophilous habits.

For several species of lepidoptera, most of the similar to 350-bp mitochondrial control-region sequences were determined. Six of these species are in one genus, Jalmenus; are closely related; and are believed to have undergone recent rapid speciation. Recent speciation was supported by the observation of low interspecific sequence divergence. Thus, no useful phylogeny could be constructed for the genus. Despite a surprising conservation of control-region length, there was little conservation of primary sequences either among the three lepidopteran genera or between lepidoptera and Drosophila. Analysis of secondary structure indicated only one possible feature in common-inferred stem loops with higher-than-random folding energies-although the positions of the structures in different species were unrelated to regions of primary sequence similarity. We suggest that the conserved, short length of control regions is related to the observed lack of heteroplasmy in lepidopteran mitochondrial genomes. In addition, determination of flanking sequences for one Jalmenus species indicated (i) only weak support for the available model of insect 12S rRNA structure and (ii) that tRNA translocation is a frequent event in the evolution of insect mitochondrial genomes.