When are Cheaters not “Cheaters”? (Re)expanding the Concept of Exploitation in Mutualism

Judie Bronstein, University of Arizona

Virtually all mutualisms are exploited by individuals and species that obtain rewards and services but return nothing to their partners. By defining “cheaters” as individuals that experience higher fitness from cheating than by cooperating and that inflict a fitness cost on their partners when they cheat, then almost all known cases of exploitation are defined away. I define and present examples of three kinds of exploitation that fall outside this definition, then argue that each offers exciting avenues for further research. The first and best-understood are “third party exploiters;” they are incapable of cooperating, and indeed may be from lineages with no cooperation in their evolutionary pasts. Open questions include how they coexist at ecological and evolutionary timescales with mutualists; how their presence changes the evolutionary trajectory of the mutualism itself; how third-party exploitation evolves, and how such exploiters diversify. The second overlooked group is “low-cost exploiters,” which can alternatively cheat or cooperate, experience a benefit from cheating, but inflict no measurable fitness costs on their partner when they do. Their ubiquity suggests the largely untested hypothesis that selection favors traits that confer tolerance to exploiters’ damaging effects. Finally, individuals may  experience context-dependent fitness benefits from exploiting their partner, and/or may inflict context-dependent costs upon it. Selection favoring and obstructing exploitation must surely be a function of the context-dependency of exploitation’s costs and benefits, but research on this topic has barely begun. We would benefit from exploring once again the natural history of exploitation we may have forgotten how richly diverse it is, and how exciting to study in all its manifestations.

Cheating and punishment in cooperative animal societies

Christie Riehl, Princeton University

Cooperative animal societies – including group-living insects, birds, mammals, fish, and humans – may evolve as a result of kin selection, reciprocity, shared fitness benefits, or a combination of all three. Kin selection involves indirect fitness benefits that accrue to genetic relatives, so if cooperative partners are related, selection typically does not favor cheating. But how can groups composed of non-relatives be evolutionarily stable? If cheating behaviors are adaptive – in other words, if they increase the fitness of the cheater at the expense of cooperative group members – then theory predicts that cheats should easily undermine cooperation. One popular hypothesis is that cooperative individuals punish cheaters, lowering their fitness and enforcing cooperative behavior. An alternative hypothesis, however, is that in many cooperative societies, “cheating” behaviors do not actually result in higher fitness, because the fitness interests of group members are closely aligned. This is likely to be the case in interactions where a cooperative act itself increases the fitness of the actor, regardless of what its partner does (also known as “byproduct” mutualisms). Here I discuss one case study in depth: cooperative breeding in the Greater Ani, a tropical cuckoo that breeds in groups composed of unrelated, socially monogamous pairs. The direct fitness benefits of group membership – including increased territory quality and nest success – appear to outweigh the costs of competition between group members. Since all group members reproduce in the same nest, their fitness interests are tightly linked. Conspecific brood parasitism (a.k.a. “egg dumping”) is one common form of cheating – but parasites actually have lower fitness than cooperators, and females typically act as brood parasites only after their other reproductive options have failed. Anis, therefore, represent an extreme example of how direct, shared fitness benefits can favor cooperation between non-relatives. Similar mutualistic interactions are likely to stabilize cooperation in a variety of other societies, including those composed of relatives.


Disentangling fitness conflict and alignment in mutualism

Stephanie Porter, Washington State University, Vancouver

In mutualisms, discriminating mechanisms such as partner choice, sanctions, screening and partner fidelity feedback preferentially direct mutualism benefits to more-cooperative individuals. These mechanisms are key to our understanding of mutualism persistence over evolutionary time because they impose natural selection against cheaters or exploiters; without them, mutualisms are predicted to devolve into antagonistic or commensal interactions and cooperation lost. I present a series of experiments based on natural variation among legumes and rhizobia to demonstrate that: 1) discriminating mechanisms are critical to mutualism persistence because in their absence, selection for cheating would favor less cooperative rhizobia, 2) mutualist conflicts can extend beyond the rates at which resources are exchanged to include conflict over joint phenotypes co-expressed by the mutualist participants and thus mechanisms controlling cheating may need to be similarly complex, and 3) discriminating mechanisms imposed by a host species can  locally deplete less-cooperative partners over generations, increasing the fitness of conspecific hosts over successive generations, but potentially relaxing selection for the maintenance of discriminating mechanisms. Further research on the evolutionary and ecological impacts of mechanisms controlling cheating will be instrumental to understanding of how cooperation between species is maintained in natural populations.

Integrating theory and data by parameterizing a model of resource trade

Maren Friesen, Michigan State University

One of the major challenges in the study of mutualism is understanding variation in the amount of goods and services exchanged between partners. These exchanges underlie fitness variation, which in turn determines selection on the traits of each partner that mediate the interaction, in particular whether cheating strategies can destabilize the mutualism. Moving towards predictive models is essential in our quest to understand the underlying mechanisms that shape mutualism. To explore whether a model that assumes fair trade between partners is sufficient to explain variation in this interaction, we paramaterized a trading model with empirical data from a rhizobia-legume mutualism grown under controlled conditions across a resource gradient. When external resources vary, the plant should alter its investment into direct acquisition versus acquisition via investment in its mutualist partner. We find that the model does a good job of predicting growth, but fails to predict patterns of allocation to direct uptake (through roots) versus mutualist allocation (through nodules) as well as the trade ratio. Extensions to the model are thus required to explain our empirical data, but it is not clear that cheating is required. In general, trading models predict that when trade is initiated both partners benefit — at these boundaries there is thus no fitness conflict. However, cheating could occur through alteration of the trade-ratio to benefit one partner more than the other, or through coercion to force a partner to trade when it is not in its best interest. Future work will focus on linking trade to fitness and understanding the metabolic underpinnings of trade that impose physiological constraints on this important mutualism. Overall, work in this direction is aimed beyond identifying cheating at a comprehensive, quantitative framework for understanding the extrinsic and intrinsic drivers of mutualism variation.

Evolution of “cheating” as a competitive strategy

Christopher Johnson, ETH Zürich

The idea of cheating in mutualism is often framed in terms of the costs and benefits of an interaction between mutualistic partners. Cheating may also strongly affect species within the same guild as cheaters (e.g., pollinators visiting the same plants as nectar robbers) because cheaters deplete mutualistic commodities (e.g., nectar) upon which mutualists rely without contributing to the mutualism. More broadly, the relationship between species’ rewardingness, regardless of whether it is cheating, and its ability to compete for mutualistic commodities is an important, but largely unresolved, question in mutualism. I develop a theoretical framework for investigating how rewardingness relates to species’ competitive ability for commodities. The framework is mechanistic in that commodities are explicitly incorporated such that species deplete commodities and go extinct if commodities are sufficiently depleted. The model suggests three key factors determining coexistence vs. exclusion of competitors for commodities: their commodity requirements, competitive ability (efficiency at acquiring commodities), and rewardingness. The key prediction is that selection should reduce rewardingness of species that are efficient at acquiring commodities (thereby limiting commodities available to less-efficient competitors) and increase rewardingness of species that are inefficient at acquiring commodities (buoying the fitness of partner species and thus increasing commodity availability). Coexistence arises if a trade-off exists such that efficient competitors for commodities are more rewarding or if partners offer ample commodities that efficient, but less-rewarding species, such as cheaters, are strongly limited by resources external to the mutualism (space, light, etc.). Indeed, very little is currently known about whether the best mutualists are also the best competitors for commodities, pointing to a key gap in the literature. The broader significance of the model is that it explains the coexistence of competitors exhibiting a wide spectrum of rewardingness and competitive ability. Considering the spectrum of rewardingness (at one extreme, cheating) in the context of competition for commodities provides a way to resolve the paradox of why extreme variation in partner quality exists in nature.