MBI Videos

Workshop 5: Coevolution and the Ecological Structure of Plant-insect Communities

  • video photo
    Paul Hohenlohe
    The adaptive landscape, long a useful metaphor, is also a rigorous tool for understanding evolution when it is linked to empirical measurements of fitness. However, empirical estimates of fitness surfaces are often concave, implying an evolutionarily unstable situation under general conditions in the short term, and untenable extrapolations to longer-term evolution under the traditional adaptive landscape model. Incorporating the multivariate genetic and ecological context of concave selection would lead to a more robust adaptive landscape model. The several non-exclusive hypotheses for the prevalence of concave selection fall roughly into two groups: static and dynamic. Coevolution lies at the heart of dynamic solutions and is thus critical to resolving the paradox of concave selection. In this talk I will discuss the basic hypotheses and empirical evidence for the prevalence of concave selection and explore its relationship to coevolution. I will propose a revised adaptive landscape model that can account for concave selection, yet maintain the adaptive landscape's heuristic value for understanding multivariate phenotypic evolution.
  • video photo
    Jose Gomez
    The geographic mosaic theory of coevolution (GMTC) considers that populations differ in evolutionary dynamics due to spatial variation in selective regimes. According to GMTC, three components of geographic structure drive the overall coevolutionary dynamics of such interactions: selection mosaics, coevolutionary hotspots, and trait remixing. Furthermore, the GMTC suggests the occurrence of a spatial pattern of interaction-mediated local adaptation and maladaptation. Empirical support to these theoretical predictions has come mostly from specialist antagonistic interactions. Contrasting with specialist interactions, free-living generalist interactions are formed by multispecies networks of interacting organisms that vary spatially in composition. Extreme reciprocal specialization between pairs of species is rare in these interactions. Consequently, multispecific selection and diffuse coevolution are prevalent in generalist interactions. Here I explore the possibilities of selection mosaic and interaction-mediated local adaptation in multiespecific generalized systems. To overcome the inherent difficulties of studying the interactions occurring among many species, I propose the combined use of structural equation modeling and individual-based network tools. This approach has allowed to detect that geographic mosaic of selections occur in generalist systems as well, and may be even a driver of the evolution of generalization in these types of multispecific systems.
  • video photo
    Scott Nuismer
    No description available.
  • video photo
    Bruce Anderson
    Recently, several studies have used the geographic matching of morphological traits (e.g. proboscis versus corolla length) to infer that coevolution has taken place between two interacting organisms. However, geographic trait matching alone is not sound evidence for coevolution because it is not a mandatory end point for coevolutionary relationships, and nor is coevolution the only selective mechanism capable of giving rise to geographic trait matching. Here I demonstrate how both coevolved and non-coevolved relationships result in patterns of trait matching but that through selection studies and by a process of elimination, it is possible to determine what the mechanism is behind trait matching patterns. In addition, using data from several published studies, I suggest how the steepness of the slope and intercept may yield important information about the relative strength of selection acting on the morphological traits of interacting species. For example in plant-insect relationships, plants seem to consistently have more exaggerated morphological traits than insects at high trait magnitudes. This suggests that selection to exaggerate the magnitude of the plant trait is stronger than for insects to exaggerate the magnitude of their corresponding traits. Thus, when plant and insect morphological traits are coevolved, insects are the most likely to be the losers of the coevolutionary race.
  • video photo
    Ben Ridenhour
    No description available.
  • video photo
    Marc Johnson
    (Co)evolutionary ecologists have long appreciated that ecology drives evolution, and that evolution ultimately shapes the ecological processes and patterns of populations and communities over long periods of time. However, it remains unclear how these two processes interact to affects the ecology, evolution and coevolution of communities over short timescales (e.g. one to several generations). We address this problem by studying interactions between a native plant (common evening primrose, Oenothera biennis) and the diverse assemblage of arthropods that interact with this plant. Using a combination of field experiments and theory we address two related hypotheses: First, we examine if herbivory drives rapid evolution of life-history and plant defensive traits within plant populations. Second, we test whether standing genetic variation and rapid evolution within plant populations shape the structure and dynamics of arthropod communities over one to several generations. In support of the first hypothesis, we find that herbivores impose natural selection on many heritable plant traits. Using an experimental evolution approach in the field, we show that this selection causes rapid evolution within plant populations over just three generations. In support of the second hypothesis, we find that standing genetic variation in evening primrose is an important ecological factor affecting the abundance of many herbivore populations, as well as the composition and diversity of over 100 arthropod species in the community. Using quantitative genetics theory and simulations, we show that the observed natural selection and evolution within evening primrose is expected to drive rapid ecological changes in arthropod abundance and diversity. Therefore in this plant-herbivore system, ecological and evolutionary dynamics interact over very short time scales and using an eco-evolutionary approach provides greater insight into the factors that affect both the ecology and evolution of ecosystems.
  • video photo
    Claire de Mazancourt
    In this talk I will present several theories related to co-evolution between plants and insects. First I will present a model of predator-prey coevolution, showing that rapid evolution in the predator can lead to prey diversification and a decrease in the number of preys available to the predator. This correspond to a "Red King" scenario, where rapid evolution leads to an ecological disadvantage. Second I will present the evolution from an antagonistic to a mutualistic interaction. Finally I will contrast expected patterns of coevolution in mutualistic and antagonistic networks.
  • video photo
    Sharon Strauss
    Our best examples of coevolution come from simplified interactions, but communities are rarely simple. Are complex communities less coevolved, or is coevolution just harder to recognize? I discuss coevolution and community complexity in light of both natural and invaded systems.
  • video photo
    Peter Abrams
    Most theoretical work on the evolution of competing species has used models having the minimum number of species (i.e. two), and has not represented either enemies or resources of those two consumer species. Empirical studies of character displacement involve species that share multiple resources, and usually multiple predators as well. Although some prominent experimental systems involve only two competitors, the natural communities where these species occur often have more. My talk will concentrate on two types of extension of the current body of theory. The first is models of the evolution of competitors in a community or food web context. This illustrates the importance of ecological interactions on other trophic levels when determining the response of one species to changes in a putative competitor on its own trophic level. The second is the study of multi-species coevolution under such a community context. It has long been known that the impact of one competitor on a second may be positive when the community contains many competitors; this occurs via indirect effects on other mutual competitors. However, the implications of such positive population-level interactions on predictions of character displacement have not been explored. If removal of one competitor causes niche shifts of similar competitors, will large magnitude shifts be propagated throughout the competitive community? These and other aspects of the coevolution of species on a given trophic level will be considered. Where possible, examples of relevant plant-insect systems will be discussed.
  • video photo
    Martin Heil
    Ant-plants are important structural elements in many disturbed tropical ecosystems and the mutualism between plants and their defending ant symbionts is increasingly being used as a model to study general factors that stabilize a horizontally transmitted mutualisms. As these mutualisms must be established anew in every consecutive generation they are particularly prone to the destabilization by cheaters (former mutualists that ceased the service provisioning) or parasites (non-reciprocating exploiters with no evolutionary history as a mutualist). Theoretical models predicted high exploitation rates for high-reward mutualisms. We empirically tested this prediction and found that sympatric Mexican Acacia myrmecophytes differ at the species level in the amount of food rewards and nesting space that they provide to their defending Pseudomyrmex ant mutualists. Several biochemical specializations help to exclude non-adapted potential consumers from feeding on the host-plant derived food rewards: food bodies and extrafloral nectar are biochemically protected from nectar-infecting microorganisms, nectar robbers and herbivores. Interestingly, the same adaptations appear to make the symbiotic ants fully dependent on these rewards. At the phenotypic level, hosts that produced more extrafloral nectar were more aggressively defended by their inhabitants. Genetically fixed, species-specific plant traits combine with phenotypic plasticity to create 'host sanction' mechanisms, which bind fitness-relevant traits of both partners to each other. This assumption is confirmed by the observation that two high-reward producing among four investigated host species were less commonly exploited by non-defending ant species. As reward production can be costly, this allows the diversification into 'low cost-high risk' versus 'high cost-low risk' strategies, and both strategies are indeed realized by sympatric and congeneric host species.
  • video photo
    Emily Jones
    Species exist in complex biotic environments, engaging in a variety of antagonistic and cooperative interactions that contribute to their population and evolutionary dynamics. However, studies tend to concentrate on each pairwise interaction in isolation. By doing so, they may overlook significant feedbacks between the interactions. In this talk, I will focus on plant-pollinator mutualisms, which are often beset by species that reduce the benefits of the mutualism by exploiting the plant, the pollinator, or both. Through a combination of theoretical and empirical results, I will demonstrate how predictions about the ecological stability of mutualisms and the level of cooperation between mutualistic partners are changed by consideration of exploiters such as non-pollinating seed predators and predators of pollinators.
  • video photo
    Jacqui Shykoff
    One very robust result of models of host-parasite co-evolution is that under regimes of mixed infection, where different strains of parasites compete for limiting host resources, parasites should evolve higher virulence strategies. This has wide reaching ramifications for optimal parasite strategies, since parasites are seldom alone in exploiting hosts. However, not all interactions between parasite strains within hosts are equal. When closely related parasite strains share hosts one might expect different evolutionary outcomes than when distantly related strains interact, particularly if the degree of relatedness between strains varies in nature such that parasites may be confronted with more or less related competitors. In addition to variation in the relatedness between interacting or competing parasite strains within hosts, the nature of the within-host interaction can vary greatly. Three types of within-host interactions have been identified, though this may not represent the whole range of the possibilities: competition for limiting host resources, production of public goods such as extra-cellular enzymes that promote infection success or efficient exploitation of the host and that can be used by all parasites within the same host, and spiteful interference between pathogen strains. Theoretical expectations differ for how virulence will change between single and mixed infections, and particularly for mixed infections of differing degree of relatedness between the parasite strains for these different types of interactions. I will discuss the outcome of experiments designed to examine how virulence differs among mixed infections that vary in degree of relatedness in the anther smut fungus Microbotryum violaceum, a Basidiomycete fungal pathogen of plants that is transmitted by insect vectors, with a small aside to examine the idea of relatedness and the applications of kin selection to microbial interactions. Individual strains, when involved in mixed infections, drew greater resources from their host plant and produced more spores per flower than did strains that infected on their own. Furthermore virulence, this time measured in terms of the degree of sterilisation of the host plant, varied among mixed infections that differed in genetic relatedness among strains. Individual strains interfered with each other's ability to colonise the plant, thereby reducing the virulence of mixed infections for this trait, but interfered less when strains were closely related, suggesting spiteful interactions. Thus for different traits of the infection genetic relatedness among interacting parasite strains differentially influence virulence.
  • video photo
    Franck Jabot
    In many plant-pollinator systems, interactions present a high degree of generalism, so that coevolution should be studied at the community level. Indeed intraspecific trait variation in such systems may both lead to variation in the gains that individuals are drawing from their interactions, and to variation in their choice/attraction of interaction partners. In this contribution, I will study whether the structural properties of the network of interactions between plants and animals impact the magnitude of coevolution and the properties of coevolutionary outcomes through simulations. In these simulations, individuals interact following simple interaction rules based on a single trait, the interaction strength between two individuals being controlled by their trait values. The fitness of an individual is proportional to the sum of its interactions, and traits present a constant heritability. The distribution of trait values in every species as well as species abundances are monitored through time to capture the natural selection applying on each species as a function of its position in the simulated network of interactions.

View Videos By