Global stability of obligate mutualism in community modules with facultative mutualists

Mutualism is a fundamental building block of ecological communities and an important driver of biotic evolution. Classic theory suggests that a pairwise two‐species obligate mutualism is fragile, with a large perturbation potentially driving both mutualist populations into extinction. In nature, however, there are many cases of pairwise obligate mutualism. Such pairwise obligate mutualisms are occasionally associated with additional interactions with facultative mutualists. Here, we use a mathematical model to show that when a two‐species obligate mutualism has a single additional link to a third facultative mutualist, the obligate mutualism can become permanently persistent. In the model, a facultative mutualist interacts with one of two inter‐dependent obligate mutualists, and the facultative mutualist enhances the persistence not only of its directly interacting obligate mutualist, but also that of the other obligate mutualist indirectly, enabling the permanent coexistence of the three mutualist species. The effect of the facultative mutualist is strong; it can allow a three‐species permanent coexistence even when two obligate mutualists by themselves are not sustainable (i.e. not locally stable). These results suggest that facultative mutualists can play a pivotal role for the persistence of obligate mutualisms, and contribute to a better understanding on the mechanisms maintaining more complex mutualistic networks of multiple species.     

Limiting cheaters in mutualism: evidence from hybridization between mutualist and cheater yucca moths

Mutualisms are balanced antagonistic interactions where both species gain a net benefit. Because mutualisms generate resources, they can be exploited by individuals that reap the benefits of the interaction without paying any cost. The presence of such 'cheaters' may have important consequences, yet we are only beginning to understand how cheaters evolve from mutualists and how their evolution may be curtailed within mutualistic lineages. The yucca-yucca moth pollination mutualism is an excellent model in this context as there have been two origins of cheating from within the yucca moth lineage. We used nuclear and mitochondrial DNA markers to examine genetic structure in a moth population where a cheater species is parapatric with a resident pollinator. The results revealed extensive hybridization between pollinators and cheaters. Hybrids were genetically intermediate to parental populations, even though all individuals in this population had a pollinator phenotype. The results suggest that mutualisms can be stable in the face of introgression of cheater genes and that the ability of cheaters to invade a given mutualism may be more limited than previously appreciated.     

Selective feedback between dispersal distance and the stability of mutualism

The evolution and maintenance of mutually beneficial interactions has been one of the oldest problems for evolutionary theory. For cooperation to be stable, mechanisms such as spatial population structure must exist that prevent non‐cooperative individuals from invading cooperative groups. Selection for certain traits like increased dispersal can erode that structure. Here, I used a spatially explicit individual based dual lattice computer simulation to investigate how the evolution of dispersal interacts with the evolution of mutualism and how this interaction affects the stability of mutualism in the face of non‐mutualists. I ran simulations manipulating the self‐structuring phenotype, dispersal distance, over a range of environmental conditions, as well as letting both dispersal and mutualism evolve independently, with and without a cost of dispersal. I found that environmental productivity is negatively correlated with the stability of mutualism, and that the stability of mutualism relied on the ability of mutualists to evolve shorter dispersal distances than non‐mutualists. The inclusion of a dispersal cost essentially fixed the upper limit of dispersal, and therefore limits the ability of non‐mutualists to evolve higher average dispersal than mutualists, but as costs are relaxed, the differences are recovered. These results show how selection on seemingly unrelated traits can align suites of traits into holistic life history strategies.     

Advancing the missed mutualist hypothesis,the under-appreciated twin of the enemy release hypothesis

Introduced species often benefit from escaping their enemies when they are transported to a new range, an idea commonly expressed as the enemy release hypothesis. However, species might shed mutualists as well as enemies when they colonize a new range. Loss of mutualists might reduce the success of introduced populations, or even cause failure to establish. We provide the first quantitative synthesis testing this natural but often overlooked parallel of the enemy release hypothesis, which is known as the missed mutualist hypothesis. Meta-analysis showed that plants interact with 1.9 times more mutualist species, and have 2.3 times more interactions with mutualists per unit time in their native range than in their introduced range. Species may mitigate the negative effects of missed mutualists. For instance, selection arising from missed mutualists could cause introduced species to evolve either to facilitate interactions with a new suite of species or to exist without mutualisms. Just as enemy release can allow introduced populations to redirect energy from defence to growth, potentially evolving increased competitive ability, species that shift to strategies without mutualists may be able to reallocate energy from mutualism toward increased competitive ability or seed production. The missed mutualist hypothesis advances understanding of the selective forces and filters that act on plant species in the early stages of introduction and establishment and thus could inform the management of introduced species.     

Three-way coexistence in obligate mutualist-exploiter interactions: the potential role of competition

Many mutualisms host "exploiter" species that consume the benefits provided by one or both mutualists without reciprocating. Exploiters have been widely assumed to destabilize mutualisms, yet they are common. We develop models to explore conditions for local coexistence of obligate plant/pollinating seed parasite mutualisms and nonpollinating exploiters. As the larvae of both pollinators and (at a later time) exploiters consume seeds, we examine the importance of intraspecific and (asymmetric) interspecific competition among and between pollinators and exploiters for achieving three-way coexistence. With weak intra- and interspecific competition, exploiters can invade the stable mutualism and coexist with the mutualists (either stably or with oscillations), provided the exploiters' intrinsic birthrate (b(E)) slightly exceeds that of the pollinators. At higher b(E), all three species go locally extinct. When facing strong interspecific competition, exploiters cannot invade and coexist with the mutualists if intraspecific competition in pollinators and exploiters is weak. However, strong intraspecific competition in pollinators and exploiters facilitates exploiter invasion and coexistence and greatly expands the range of b(E) over which stable coexistence occurs. Our results suggest that mutualist/exploiter coexistence may be more easily achieved than previously thought, thus highlighting the need for a better understanding of competition among and between mutualists and exploiters.     

Costs of two non-mutualistic species in a yucca/yucca moth mutualism

Mutualisms often involve significant costs for participants. Costs are inflicted by mutualists themselves, as well as by associated, non-mutualistic species. These costs are rarely quantified, however, particularly the ones extrinsic to the pairwise interaction. We compare costs inflicted by an obligate mutualist pollinator and two common exploiters of an Arizona yucca over a 2-year period. The magnitude of seed damage from seed and fruit-feeding beetle larvae (Carpophilus longus, Nitidulidae) was similar to damage from the seed-eating larvae of Yucca schottii's pollinator moth Tegeticula yuccasella (Prodoxidae), averaging about 15 seeds destroyed per fruit in each case. The two seed predators usually fed within the same fruits, although rarely side by side. In contrast, the presence of fruit-galling moth larvae (Prodoxusy-inversus, Prodoxidae) appeared to benefit the yucca: individual Tegeticula destroyed only half as many seeds in galled fruits as they did in ungalled fruits. We discuss three general implications of these results. Firstly, the costs of non-mutualists to the two mutualistic partners are not necessarily parallel. Secondly, measurable costs of non-mutualists do not necessarily translate into an impact on the success of the mutualism itself, because they may be incurred after mutualistic activities take place. Finally, the costs of mutualists to each other can differ substantially depending on the presence or absence of non-mutualistic species. Received:17 July 1996 / Accepted:10 June 1997     

Dynamics of mutualist populations that are demographically open

1. Few theoretical studies have examined the impact of immigration and emigration on mutualist population dynamics, but a recent empirical study (A.R. Thompson Oecologia, 143, 61-69) on mutualistic fish and shrimp showed that immigration can prevent population collapse, and that intraspecific competition for a mutualistic partner can curb population expansion. To understand in a theoretical context the implications of these results, and to assess their generality, we present a two-species model that accounts explicitly for immigration and emigration, as well as distinguishing the impacts of mutualism on birth rates, death rates and habitat acquisition. 2. The model confirms that immigration can stabilize mutualistic populations, and predicts that high immigration, along with enhanced reproduction and/or reduced mortality through mutualism, can cause population sizes to increase until habitat availability curbs further expansion. 3. We explore in detail the effects of different forms of habitat limitation on mutualistic populations. Habitat availability commonly limits the density of both populations if mutualists acquire shelter independently. If a mutualist depends on a partner for habitat, densities of that mutualist are capped by the amount of space provided by that partner. The density of the shelter-provider is limited by the environment. 4. If a mutualism solely augments reproduction, and most locally produced individuals leave the focal patch, then the mutualism will have a minimal effect on local dynamics. If the mutualism operates by reducing rates of death or enhancing habitat availability, and there is at least some immigration, then mutualism will affect local dynamics. This finding may be particularly relevant in marine systems, where there is high variability (among species and locations) in the extent to which progeny disperse from natal locations. 5. Overall, our results demonstrate that the consequences of immigration and emigration for the dynamics of mutualists depend strongly on which demographic rate is influenced by mutualism. 6. By relating our model to a variety of terrestrial and aquatic systems, we provide a general framework to guide future empirical studies of the dynamics of mutualistic populations.     

Obligate mutualism with a predator: Stability and persistence of three-species models

We present a general model for three interacting populations, where one population, called a mutualist, benefits a predator in its interaction with the prey. Biologically, there are four different ways in which the mutualist could benefit the predator: by enhancing prey growth rate, by enhancing the rate of prey capture, by providing an alternative food supply for the predator, and by enhancing the efficiency of utilization of prey, once they are ingested. We discuss examples of each type of interaction. We restrict our model to those situations in which the predator cannot survive on the prey in the absence of the mutualist. Therefore, if mutualism exists, it is obligate for the predator. Other conditions of the model include the dynamics of the prey and the mutualist alone and together in the absence of the predator. Given additional reasonable restrictions on the model, we determine the conditions for persistence, where persistence is defined as the continued existence of all three populations without any of them going extinct. There are two ways in which survival may arise in these models. Under one set of conditions, which is equivalent to the predator being able to invade a prey-mutualist system when rare, persistence will occur for any set of positive critical population sizes. Alternatively, survival will occur if there is an asymptotically stable interior equilibrium. However, the conditions for this are complex, and survival may occur only for initial populations in a limited region around the equilibrium.     

Spatial Heterogeneity in Soil Microbes Alters Outcomes of Plant Competition

Plant species vary greatly in their responsiveness to nutritional soil mutualists, such as mycorrhizal fungi and rhizobia, and this responsiveness is associated with a trade-off in allocation to root structures for resource uptake. As a result, the outcome of plant competition can change with the density of mutualists, with microbe-responsive plant species having high competitive ability when mutualists are abundant and non-responsive plants having high competitive ability with low densities of mutualists. When responsive plant species also allow mutualists to grow to greater densities, changes in mutualist density can generate a positive feedback, reinforcing an initial advantage to either plant type. We study a model of mutualist-mediated competition to understand outcomes of plant-plant interactions within a patchy environment. We find that a microbe-responsive plant can exclude a non-responsive plant from some initial conditions, but it must do so across the landscape including in the microbe-free areas where it is a poorer competitor. Otherwise, the non-responsive plant will persist in both mutualist-free and mutualist-rich regions. We apply our general findings to two different biological scenarios: invasion of a non-responsive plant into an established microbe-responsive native population, and successional replacement of non-responders by microbe-responsive species. We find that resistance to invasion is greatest when seed dispersal by the native plant is modest and dispersal by the invader is greater. Nonetheless, a native plant that relies on microbial mutualists for competitive dominance may be particularly vulnerable to invasion because any disturbance that temporarily reduces its density or that of the mutualist creates a window for a non-responsive invader to establish dominance. We further find that the positive feedbacks from associations with beneficial soil microbes create resistance to successional turnover. Our theoretical results constitute an important first step toward developing a general understanding of the interplay between mutualism and competition in patchy landscapes, and generate qualitative predictions that may be tested in future empirical studies.     

Dual lattice model of the evolution of facultative symbiosis with continuous Prisoner's Dilemma game

Mutualism is ubiquitous in nature and is thought to have played a key role in the history of life. However, how mutualism could evolve despite being prone to unilateral exploitation is a puzzling question in evolutionary ecology. Some theoretical studies have shown that spatial structure of habitat can facilitate the emergence and maintenance of mutualism. However, they are based on the simple assumption that the trait in question is discrete: each individual is either a mutualist or a non-mutualist. In this article I develop a simple simulation model of coevolution of facultative symbiosis using a one-shot continuous Prisoner's Dilemma game to investigate the evolutionary dynamics of mutualism between two species. In this model I assume continuous traits for both species from -1 (fully deceptive) to 1 (fully cooperative). The habitat has a dual-lattice structure, each layer is inhabited by one species. Interspecific interaction is restricted between two corresponding sites of the two layers. Without limitation on the magnitude of a single mutation, I find that mutualism can arise and persist when the intrinsic reproduction rate is low (but is above a threshold) and the benefit/cost ratio of the cooperative strategy is large, which is consistent with Yamamura et al. [2004. Evolution of mutualism through spatial effects. J. Theor. Biol. 226, 421-428]. In these cases, extreme antagonism often evolves starting from a neutral population that seems nearly stable, but once mutualism arises, the cooperative individuals quickly increase and both the populations eventually become mutualistic on average, although they are polymorphic. However, when the effect of a single mutation was limited to be small, extreme antagonism is much likely to dominate unless the intrinsic reproduction rate is low. When only one species is allowed to evolve, mutualism arises when the initial strategy of the other species is cooperative. Otherwise, excessive deception evolves in the former, and the latter often becomes driven to extinction.     

Evolutionary origins and diversification of proteobacterial mutualists

Mutualistic bacteria infect most eukaryotic species in nearly every biome. Nonetheless, two dilemmas remain unresolved about bacterial–eukaryote mutualisms: how do mutualist phenotypes originate in bacterial lineages and to what degree do mutualists traits drive or hinder bacterial diversification? Here, we reconstructed the phylogeny of the hyperdiverse phylum Proteobacteria to investigate the origins and evolutionary diversification of mutualistic bacterial phenotypes. Our ancestral state reconstructions (ASRs) inferred a range of 34–39 independent origins of mutualist phenotypes in Proteobacteria, revealing the surprising frequency with which host-beneficial traits have evolved in this phylum. We found proteobacterial mutualists to be more often derived from parasitic than from free-living ancestors, consistent with the untested paradigm that bacterial mutualists most often evolve from pathogens. Strikingly, we inferred that mutualists exhibit a negative net diversification rate (speciation minus extinction), which suggests that mutualism evolves primarily via transitions from other states rather than diversification within mutualist taxa. Moreover, our ASRs infer that proteobacterial mutualist lineages exhibit a paucity of reversals to parasitism or to free-living status. This evolutionary conservatism of mutualism is contrary to long-standing theory, which predicts that selection should often favour mutants in microbial mutualist populations that exploit or abandon more slowly evolving eukaryotic hosts.     

Carbon allocation and competition maintain variation in plant root mutualisms

Plants engage in multiple root symbioses that offer varying degrees of benefit. We asked how variation in partner quality persists using a resource‐ratio model of population growth. We considered the plant's ability to preferentially allocate carbon to mutualists and competition for plant carbon between mutualist and nonmutualist symbionts. We treated carbon as two nutritionally interchangeable, but temporally separated, resources—carbon allocated indiscriminately for the construction of the symbiosis, and carbon preferentially allocated to the mutualist after symbiosis establishment and assessment. This approach demonstrated that coexistence of mutualists and nonmutualists is possible when fidelity of the plant to the mutualist and the cost of mutualism mediate resource competition. Furthermore, it allowed us to trace symbiont population dynamics given varying degrees of carbon allocation. Specifically, coexistence occurs at intermediate levels of preferential allocation. Our findings are consistent with previous empirical studies as well the application of biological market theory to plantroot symbioses.     

Defaunation leads to interaction deficits,not interaction compensation,in an island seed dispersal network

Following defaunation, the loss of interactions with mutualists such as pollinators or seed dispersers may be compensated through increased interactions with remaining mutualists, ameliorating the negative cascading impacts on biodiversity. Alternatively, remaining mutualists may respond to altered competition by reducing the breadth or intensity of their interactions, exacerbating negative impacts on biodiversity. Despite the importance of these responses for our understanding of the dynamics of mutualistic networks and their response to global change, the mechanism and magnitude of interaction compensation within real mutualistic networks remains largely unknown. We examined differences in mutualistic interactions between frugivores and fruiting plants in two island ecosystems possessing an intact or disrupted seed dispersal network. We determined how changes in the abundance and behavior of remaining seed dispersers either increased mutualistic interactions (contributing to “interaction compensation”) or decreased interactions (causing an “interaction deficit”) in the disrupted network. We found a “rich‐get‐richer” response in the disrupted network, where remaining frugivores favored the plant species with highest interaction frequency, a dynamic that worsened the interaction deficit among plant species with low interaction frequency. Only one of five plant species experienced compensation and the other four had significant interaction deficits, with interaction frequencies 56–95% lower in the disrupted network. These results do not provide support for the strong compensating mechanisms assumed in theoretical network models, suggesting that existing network models underestimate the prevalence of cascading mutualism disruption after defaunation. This work supports a mutualist biodiversity‐ecosystem functioning relationship, highlighting the importance of mutualist diversity for sustaining diverse and resilient ecosystems.     

Population dynamics and mutualism: functional responses of benefits and costs

We develop an approach for studying population dynamics resulting from mutualism by employing functional responses based on density-dependent benefits and costs. These functional responses express how the population growth rate of a mutualist is modified by the density of its partner. We present several possible dependencies of gross benefits and costs, and hence net effects, to a mutualist as functions of the density of its partner. Net effects to mutualists are likely a monotonically saturating or unimodal function of the density of their partner. We show that fundamental differences in the growth, limitation, and dynamics of a population can occur when net effects to that population change linearly, unimodally, or in a saturating fashion. We use the mutualism between senita cactus and its pollinating seed-eating moth as an example to show the influence of different benefit and cost functional responses on population dynamics and stability of mutualisms. We investigated two mechanisms that may alter this mutualism's functional responses: distribution of eggs among flowers and fruit abortion. Differences in how benefits and costs vary with density can alter the stability of this mutualism. In particular, fruit abortion may allow for a stable equilibrium where none could otherwise exist.     

Host discrimination in modular mutualisms: a theoretical framework for meta-populations of mutualists and exploiters

Plants in multiple symbioses are exploited by symbionts that consume their resources without providing services. Discriminating hosts are thought to stabilize mutualism by preferentially allocating resources into anatomical structures (modules) where services are generated, with examples of modules including the entire inflorescences of figs and the root nodules of legumes. Modules are often colonized by multiple symbiotic partners, such that exploiters that co-occur with mutualists within mixed modules can share rewards generated by their mutualist competitors. We developed a meta-population model to answer how the population dynamics of mutualists and exploiters change when they interact with hosts with different module occupancies (number of colonists per module) and functionally different patterns of allocation into mixed modules. We find that as module occupancy increases, hosts must increase the magnitude of preferentially allocated resources in order to sustain comparable populations of mutualists. Further, we find that mixed colonization can result in the coexistence of mutualist and exploiter partners, but only when preferential allocation follows a saturating function of the number of mutualists in a module. Finally, using published data from the fig–wasp mutualism as an illustrative example, we derive model predictions that approximate the proportion of exploiter, non-pollinating wasps observed in the field.     

On the structure and stability of mutualistic systems: Analysis of predator-prey and competition models as modified by the action of a slow-growing mutualist

Two basic models of mutualism are presented in which interactions among three species lead to mutualism between two of them. The models represent 2-species predator-prey or competition systems in which a third species acts as a mutualist with either the predator, the prey, or one of the competitors. The models include the assumptions that there is a cost of associating with the mutualist and that the mutualist population grows much more slowly than the other two populations. Special cases of these two models correspond to six qualitatively different types of mutualistic benefit, all of which are known to occur in nature: deterring predation, increasing prey availability, feeding on (or competing with) a predator, increasing competitive interactions, decreasing competitive interactions, and feeding on (or competing with) a competitor. These models and their special cases are subjected to a local stability analysis. The results show that mutualism based upon deterring predation, competing with a predator, or decreasing competitive interactions enhances local stability, while mutualism based upon increasing prey availability or increasing competitive interactions reduces local stability. These results clearly reject the idea that mutualism is an inherently unstable process, and reinforces the idea that each different kind of mutualism will have to be considered separately. Compared to 2-species models of mutualism, the 3-species models provide a more realistic representation of the structure of many mutualistic systems, the mechanisms by which one species benefits another, and the regulation of the interaction.     

Perturbed partners: opposite responses of plant and animal mutualist guilds to inundation disturbances

Mutualists have been suggested to play an important role in the assembly of many plant and animal communities, but it is not clear how this depends on environmental factors. Do, for instance, natural disturbances increase or decrease the role of mutualism? We focused on entire guilds of mutualists, studying seed‐dispersing ants and ant‐dispersed plants along gradients of inundation disturbances. We first studied how abundance and richness of the mutualists, relative to non‐mutualists, change along 35 small‐scale gradients of inundation disturbances. We found that at disturbed sites, mutualistic plant species, those that reproduce by seeds dispersed by ants, increased in abundance and in consequences in richness, relative to other herbaceous plants. In contrast, we found that among the epigeic arthropods the abundance of mutualists declined, even more so than other arthropods. Correspondingly, distributions of plant and animal mutualists became increasingly discordant at disturbed sites: most plant mutualists were spatially separated from most animal mutualists. We finally found that high abundances of plant mutualists did not translate into a high nutrition service rendered to ants: at disturbed sites, many of the plants of ant‐dispersed species did not produce seeds, which coincided with a decline in seed dispersal by ants and a changing searching behavior of the ants. Overall, the small‐scale natural disturbances we studied were correlated to a major change in the assembly of mutualist guilds. However, the correlation was often opposite between interacting plant and animal mutualist guilds and may thus reduce the potential interaction between them.     

HOW TO PREVENT CHEATING: A DIGESTIVE SPECIALIZATION TIES MUTUALISTIC PLANT-ANTS TO THEIR ANT-PLANT PARTNERS

Mutualisms often involve reciprocal adaptations of both partners. Acacia ant-plants defended by symbiotic Pseudomyrmex ant mutualists secrete sucrose-free extrafloral nectar, which is unattractive to generalists. We aimed to investigate whether this extrafloral nectar can also exclude exploiters, that is nondefending ant species. Mutualist workers discriminated against sucrose whereas exploiters and generalists with no affinity toward Acacia myrmecophytes preferred sucrose, because mutualist workers lacked the sucrose-cleaving enzyme invertase, which is present in workers of the other two groups. Sucrose uptake induced invertase activity in workers of parasites and generalists, but not mutualists, and in larvae of all species: the mutualists loose invertase during their ontogeny. This reduced metabolic capacity ties the mutualists to their plant hosts, but it does not completely prevent the mutualism from exploitation. We therefore investigated whether the exploiters studied here are cheaters (i.e., have evolved from former mutualists) or parasites (exploiters with no mutualistic ancestor). A molecular phylogeny demonstrates that the exploiter species did not evolve from former mutualists, and no evidence for cheaters was found. We conclude that being specialized to their partner can prevent mutualists from becoming cheaters, whereas other mechanisms are required to stabilize a mutualism against the exploitation by parasites.     

Global stability of the coexistence equilibrium for a general class of models of facultative mutualism

Many models of mutualism have been proposed and studied individually. In this paper, we develop a general class of models of facultative mutualism that covers many of such published models. Using mild assumptions on the growth and self-limiting functions, we establish necessary and sufficient conditions on the boundedness of model solutions and prove the global stability of a unique coexistence equilibrium whenever it exists. These results allow for a greater flexibility in the way each mutualist species can be modelled and avoid the need to analyse any single model of mutualism in isolation. Our generalization also allows each of the mutualists to be subject to a weak Allee effect. Moreover, we find that if one of the interacting species is subject to a strong Allee effect, then the mutualism can overcome it and cause a unique coexistence equilibrium to be globally stable.     

Parasites of mutualisms

Cooperation invites cheating, and nowhere is this more apparent than when different species cooperate, known as mutualism. In almost all mutualisms studied, specialist parasites have been identified that purloin the benefits that one mutualist provides another. Explaining how parasites are kept from driving mutualisms extinct remains an unsolved problem because existing theories explaining the maintenance of cooperation do not apply to parasites of mutualisms. Nonetheless, these theories can be summarized in such a way as to suggest how mutualisms can persist in the face of parasites. (1) For cooperation to occur, the recipient of a benefit must reciprocate, and the recriprocated benefit must be captured by the initial giver or its offspring. (2) For cooperation to persist, the mutualism must be re-assembled each generation. Because most mutualisms are of the "by-product' type, broadly defined, the first condition is normally always fulfilled. Thus, the maintenance of mutualism usually requires enforcement of the second condition: reliable re-assembly. Hence, I argue that the persistence of mutualism is best understood by using theories of species coexistence, because each mutualist can be considered a resource for the other, and species coexistence theory explains how multiple taxa (e.g. parasites and mutualists) can stably partition a resource over multiple generations. This approach connects the study of mutualism to theories of population regulation and helps to identify key factors that have promoted the evolution, maintenance and breakdown of mutualism. I discuss how these ideas might apply to and be tested in ant-plant, fig-wasp and yucca-moth mutualisms.