| Symbioses, intimate interactions between two species, are often thought to exist along a continuum spanning parasitism, in which the symbiont benefits at the cost of the host, and mutualism, in which both partners benefit from the interaction. Empirical evidence indicates that mutualists can evolve from parasites and vice versa, and several factors are thought to influence these evolutionary transitions. These include symbiont transmission mode, the potential for symbiont phenotypes to vary across environments, and the extent to which hosts can tolerate the costs of interacting with the symbiont. Here, I focus on the effects of both symbiont transmission mode and of correlations in symbiont traits on evolutionary transitions between parasitism and mutualism.;In theoretical work, I first explore how transmission mode is expected to affect the evolution of the benefits provided by symbionts to their hosts when a mechanism (e.g. pleiotropy, partner choice by hosts) creates a positive covariance between these benefits and symbiont horizontal transmission. The analysis reveals that not only can greater exposure to horizontal transmission lead to stronger selection on symbionts to increase benefits, but also high rates of vertical transmission can interfere with the evolution of these benefits. These results run counter to most expectations and suggest a critical role for including system-specific mechanistic details in future models of mutualism evolution.;Next, I examine how transmission mode affects the coevolution of a recently discovered mutualism between the bacterium, Escherichia coli, and its filamentous phage, M13. I report results from a coevolution experiment in which the availability of horizontal transmission varied across treatments. While coevolution with hosts led to higher growth rates in coevolved populations, these same phage also evolved to be harmful toward the ancestral host genotype. Further, rare horizontal transmission enabled greater benefit evolution in many populations than vertical transmission alone, and restriction to only horizontal transmission did not generally lead to the evolution of parasitic phage. These data lend support to my theoretical work and suggest that the mechanism underlying phage mutualism is tied to both horizontal transmission and the potential for parasitism.;Last, I ask how readily parasitic phage can re-evolve mutualism with their hosts and to what extent this evolution depends on host coevolution and the ancestral phage genotype. To do so, I chose six phage populations from the first experiment that had evolved to harm the ancestral host and coevolved them with the parasitized host in a treatment that allowed for rare horizontal transmission. By the end of the experiment, all phage populations had re-evolved mutualism with their coevolved host. The data suggest that this was achieved through either evolution of hosts to tolerate the costs of infection or through selection on phage to both decrease the costs and increase the benefits of infection. Further, the trajectory of coevolution varied across replicates derived from different ancestral phage populations, indicating a likely role for initial phage genotype in determining the eventual outcomes of coevolution with bacteria. |
