Cooperation among species, or mutualism, is ubiquitous but perplexing because of the problem of cheaters. Cheaters receive benefits without reciprocating. What prevents them from breaking down a mutualism? Two solutions have been proposed. Under partner choice, choosy partners reward beneficial mutualists and punish cheaters. Under partner fidelity, cheating on a long-term mutualist feeds back to harm the cheater. Scant work has tested these theories empirically. In an ecologically important mutualism, root-nodulating rhizobial bacteria convert atmospheric nitrogen to a plant-available form in exchange for resources produced through photosynthesis by the legume host. However, some rhizobia infect hosts without providing benefit and some ignore hosts altogether. Little is known about the selective forces that prevent these bacteria from spreading. This project will study how selection can promote cooperation in rhizobia. Surveys and manipulative experiments will: (1) characterize rhizobial populations in soil, on roots, and in nodules, (2) document population genetic changes in nodulated rhizobia during infection, (3) test for release of cooperative rhizobia from senescing nodules, and (4) examine partner choice in several legume species.
Plants and animals, including humans and our food sources, are critically dependent on beneficial microbes. However, mutualistic microbes can become parasitic. Phenomena such as coral bleaching might be explained by environmental sensitivity of mutualistic microbes. Understanding what factors drive these shifts and how cooperation is maintained helps us understand how to limit the evolution of virulence in human pathogens, improve crop and livestock productivity, and predict and manage responses to environmental change. Research results will be used to enrich classroom learning and public lectures. Students from diverse backgrounds will learn research methods, and graduate students and postdoctoral scholars will learn to mentor undergraduate students.
Mutualisms are cooperative interactions among species, but they are expected to be vulnerable to cheats, individuals that accept more benefits from a partner while reciprocating less. The problem is especially acute when individual hosts interact with multiple symbionts acquired infectiously from the environment: less-beneficial symbionts might get a free ride from the host and other, more-beneficial symbionts. Yet, mutualistic bacterial symbionts are widespread in animal and plant hosts and are fundamentally important to their ecology and evolution. Three mechanisms might promote evolutionary maintenance of cooperation by symbionts: (1) Automatic feedback mechanisms ensure that any improvement a symbiont contributes to its host’s reproduction automatically improves its own reproduction, (2) Hosts avoid interacting with less-beneficial symbionts, (3) Hosts stop providing resources to less-beneficial symbionts. In addition to promoting cooperation, all three mechanisms could cause positive feedback between hosts and symbionts, which can cause these species to occur in patches across the landscape. Since each bacterial symbiont can interact with only one host and may not be free to escape, natural selection might also favor cheating hosts, which can extract benefits at the expense of future symbiont reproduction. Given this possibility, selection should favor symbiont traits that give some control over the interaction. What are the evolutionary implications of such traits? The funded work explored the evolutionary mechanisms that maintain cooperation between legume hosts and rhizobial symbionts. Rhizobia are soil-dwelling bacteria that can infect legume roots and convert atmospheric di-nitrogen into ammonium while they live inside root nodules. Nitrogen (N) is essential for all life; its availability helps determine the productivity of both terrestrial and aquatic environments. Legumes trade sugar for rhizobial ammonium in a mutually beneficial interaction that is crucially important to the biosphere and agriculture. Research on wild legumes and rhizobia is challenging some of the results of earlier work on agricultural species. Our work supports recent arguments that valuable symbiotic traits may have been lost during domestication of agricultural cultivars from their wild ancestors, which suggests that plant breeding should reintroduce legume traits that improve plant control of the interaction. Ten important results include: (1) A theoretical model showed that, if rhizobia have any control over resource trade, hosts that avoid helping cheating symbionts might nonetheless fail to prevent their spread in symbiont populations. (2) A greenhouse experiment found natural selection favors rhizobial strains that cheat by providing less benefit to hosts, which could cause less-beneficial strains to spread in rhizobium populations. (3) However, the natural population from which these rhizobia were sampled was dominated by highly cooperative rhizobia, which suggests that in nature, where they have access to many rhizobia genotypes, legumes can avoid providing benefits to less-cooperative rhizobia. (4) Despite strong evidence that selection favored cheating rhizobia, there was no evidence that selection favored cheating hosts. (5) A survey of several native legumes revealed variation in how much they specialize on particular rhizobia. (6) A greenhouse experiment suggested that variation in specialization may be maintained by trade-offs associated with the ability to specialize. Although a generalist host could benefit from more different symbionts, a specialist host species obtained greater fitness benefit from its beneficial symbionts. Further, each host species performed better with rhizobia isolated from its own species, suggesting that plants and rhizobia have co-adapted to each other. (7) A greenhouse experiment found that low-benefit rhizobium genotypes gained less fitness advantage from hosts, suggesting that hosts might modify the soil rhizobium community. (8) A two-generation experiment found positive feedback between legume generations: plants performed better when grown in a pot that had been previously occupied by a member of its own species. Positive intergenerational feedback occurred only when legumes could modify the soil rhizobium community, which suggests that positive plant-soil feedback could contribute to their patchy distributions in nature. (9) In the greenhouse, plants did not regulate nodule number in response to mineral nitrogen availability. Instead, when infected with rhizobia that could not fix nitrogen, this wild legume used mineral nitrogen to increase nodule number. Further, these ineffective nodules did not reduce plant growth, suggesting that initiating a nodule involves little cost for the plant. (10) Adding mineral nitrogen had no effect on nodule number when plants were inoculated with a strongly-fixing symbiont, but did influence the average size of nodules. As mineral availability increased, the plant allocated less resource to each nodule, thereby causing total nodule weight to be unaffected by mineral nitrogen availability. This pattern suggests that mineral nitrogen is no more or less costly to the plant than symbiotic nitrogen, which contrasts with what is found in agricultural legumes.
