Interaction with a phage gene underlie costs of a ß-lactamase.

The fitness cost of an antibiotic resistance gene (ARG) can differ across host strains, creating refuges that allow the maintenance of an ARG in the absence of direct selection for its resistance phenotype. Despite the importance of such ARG-host interactions for predicting ARG dynamics, the basis of ARG fitness costs and their variability between hosts are not well understood. We determined the genetic basis of a host-dependent cost of a ß-lactamase, blaTEM-116*, that conferred a significant cost in one Escherichia coli strain but was close to neutral in 11 other Escherichia spp. strains. Selection of a blaTEM-116*-encoding plasmid in the strain in which it initially had a high cost resulted in rapid and parallel compensation for that cost through mutations in a P1-like phage gene, relAP1. When the wild-type relAP1 gene was added to a strain in which it was not present and in which blaTEM-116* was neutral, it caused the ARG to become costly. Thus, relAP1 is both necessary and sufficient to explain blaTEM-116* costs in at least some host backgrounds. To our knowledge, these findings represent the first demonstrated case of the cost of an ARG being influenced by a genetic interaction with a phage gene. The interaction between a phage gene and a plasmid-borne ARG highlights the complexity of selective forces determining the maintenance and spread of ARGs and, by extension, encoding phage and plasmids in natural bacterial communities.IMPORTANCEAntibiotic resistance genes (ARGs) play a major role in the increasing problem of antibiotic resistance in clinically relevant bacteria. Selection of these genes occurs in the presence of antibiotics, but their eventual success also depends on the sometimes substantial costs they impose on host bacteria in antibiotic-free environments. We evolved an ARG that confers resistance to penicillin-type antibiotics in one host in which it did confer a cost and in one host in which it did not. We found that costs were rapidly and consistently reduced through parallel genetic changes in a gene encoded by a phage that was infecting the costly host. The unmutated version of this gene was sufficient to cause the ARG to confer a cost in a host in which it was originally neutral, demonstrating an antagonism between the two genetic elements and underlining the range and complexity of pressures determining ARG dynamics in natural populations.

From individual behaviors to collective outcomes: fruiting body formation in Dictyostelium as a group-level phenotype.

Collective phenotypes, which arise from the interactions among individuals, can be important for the evolution of higher levels of biological organization. However, how a group's composition determines its collective phenotype remains poorly understood. When starved, cells of the social amoeba Dictyostelium discoideum cooperate to build a multicellular fruiting body, and the morphology of the fruiting body is likely advantageous to the surviving spores. We assessed how the number of strains, as well as their genetic and geographic relationships to one another, impact the group's morphology and productivity. We find that some strains consistently enhance or detract from the productivity of their groups, regardless of the identity of the other group members. We also detect extensive pairwise and higher-order genotype interactions, which collectively have a large influence on the group phenotype. Whereas previous work in Dictyostelium has focused almost exclusively on whether spore production is equitable when strains cooperate to form multicellular fruiting bodies, our results suggest a previously unrecognized impact of chimeric co-development on the group phenotype. Our results demonstrate how interactions among members of a group influence collective phenotypes and how group phenotypes might in turn impact selection on the individual.

Fitness of evolving bacterial populations is contingent on deep and shallow history but only shallow history creates predictable patterns.

Long-term evolution experiments have tested the importance of genetic and environmental factors in influencing evolutionary outcomes. Differences in phylogenetic history, recent adaptation to distinct environments and chance events, all influence the fitness of a population. However, the interplay of these factors on a population's evolutionary potential remains relatively unexplored. We tracked the outcome of 2000 generations of evolution of four natural isolates of Escherichia coli bacteria that were engineered to also create differences in shallow history by adding previously identified mutations selected in a separate long-term experiment. Replicate populations started from each progenitor evolved in four environments. We found that deep and shallow phylogenetic histories both contributed significantly to differences in evolved fitness, though by different amounts in different selection environments. With one exception, chance effects were not significant. Whereas the effect of deep history did not follow any detectable pattern, effects of shallow history followed a pattern of diminishing returns whereby fitter ancestors had smaller fitness increases. These results are consistent with adaptive evolution being contingent on the interaction of several evolutionary forces but demonstrate that the nature of these interactions is not fixed and may not be predictable even when the role of chance is small.

The cost of evolved constitutive lac gene expression is usually, but not always, maintained during evolution of generalist populations.

Beneficial mutations can become costly following an environmental change. Compensatory mutations can relieve these costs, while not affecting the selected function, so that the benefits are retained if the environment shifts back to be similar to the one in which the beneficial mutation was originally selected. Compensatory mutations have been extensively studied in the context of antibiotic resistance, responses to specific genetic perturbations, and in the determination of interacting gene network components. Few studies have focused on the role of compensatory mutations during more general adaptation, especially as the result of selection in fluctuating environments where adaptations to different environment components may often involve trade-offs. We examine whether costs of a mutation in lacI, which deregulated the expression of the lac operon in evolving populations of Escherichia coli bacteria, were compensated. This mutation occurred in multiple replicate populations selected in environments that fluctuated between growth on lactose, where the mutation was beneficial, and on glucose, where it was deleterious. We found that compensation for the cost of the lacI mutation was rare, but, when it did occur, it did not negatively affect the selected benefit. Compensation was not more likely to occur in a particular evolution environment. Compensation has the potential to remove pleiotropic costs of adaptation, but its rarity indicates that the circumstances to bring about the phenomenon may be peculiar to each individual or impeded by other selected mutations.

Dynamics of bacterial adaptation.

Determining pattern in the dynamics of population evolution is a long-standing focus of evolutionary biology. Complementing the study of natural populations, microbial laboratory evolution experiments have become an important tool for addressing these dynamics because they allow detailed and replicated analysis of evolution in response to controlled environmental and genetic conditions. Key findings include a tendency for smoothly declining rates of adaptation during selection in constant environments, at least in part a reflection of antagonism between accumulating beneficial mutations, and a large number of beneficial mutations available to replicate populations leading to significant, but relatively low genetic parallelism, even as phenotypic characteristics show high similarity. Together, there is a picture of adaptation as a process with a varied and largely unpredictable genetic basis leading to much more similar phenotypic outcomes. Increasing sophistication of sequencing and genetic tools will allow insight into mechanisms behind these and other patterns.

Historical Contingency Causes Divergence in Adaptive Expression of the lac Operon.

Populations of Escherichia coli selected in constant and fluctuating environments containing lactose often adapt by substituting mutations in the lacI repressor that cause constitutive expression of the lac operon. These mutations occur at a high rate and provide a significant benefit. Despite this, eight of 24 populations evolved for 8,000 generations in environments containing lactose contained no detectable repressor mutations. We report here on the basis of this observation. We find that, given relevant mutation rates, repressor mutations are expected to have fixed in all evolved populations if they had maintained the same fitness effect they confer when introduced to the ancestor. In fact, reconstruction experiments demonstrate that repressor mutations have become neutral or deleterious in those populations in which they were not detectable. Populations not fixing repressor mutations nevertheless reached the same fitness as those that did fix them, indicating that they followed an alternative evolutionary path that made redundant the potential benefit of the repressor mutation, but involved unique mutations of equivalent benefit. We identify a mutation occurring in the promoter region of the uspB gene as a candidate for influencing the selective choice between these paths. Our results detail an example of historical contingency leading to divergent evolutionary outcomes.

Bacterial Evolution in High-Osmolarity Environments.

Bacteria must maintain a cytosolic osmolarity higher than that of their environment in order to take up water. High-osmolarity environments therefore present formidable stress to bacteria. To explore the evolutionary mechanisms by which bacteria adapt to high-osmolarity environments, we selected Escherichia coli in media with a variety of osmolytes and concentrations for 250 generations. Adaptation was osmolyte dependent, with sorbitol stress generally resulting in increased fitness under conditions with higher osmolarity, while selection in high concentrations of proline resulted in increased fitness specifically on proline. Consistent with these phenotypes, sequencing of the evolved populations showed that passaging in proline resulted in specific mutations in an associated metabolic pathway that increased the ability to utilize proline for growth, while evolution in sorbitol resulted in mutations in many different genes that generally resulted in improved growth under high-osmolarity conditions at the expense of growth at low osmolarity. High osmolarity decreased the growth rate but increased the mean cell volume compared with growth on proline as the sole carbon source, demonstrating that osmolarity-induced changes in growth rate and cell size follow an orthogonal relationship from the classical Growth Law relating cell size and nutrient quality. Isolates from a sorbitol-evolved population that captured the likely temporal sequence of mutations revealed by metagenomic sequencing demonstrated a trade-off between growth at high osmolarity and growth at low osmolarity. Our report highlights the utility of experimental evolution for dissecting complex cellular networks and environmental interactions, particularly in the case of behaviors that can involve both specific and general metabolic stressors.IMPORTANCE For bacteria, maintaining higher internal solute concentrations than those present in the environment allows cells to take up water. As a result, survival is challenging in high-osmolarity environments. To investigate how bacteria adapt to high-osmolarity environments, we maintained Escherichia coli in a variety of high-osmolarity solutions for hundreds of generations. We found that the evolved populations adopted different strategies to improve their growth rates depending on the osmotic passaging condition, either generally adapting to high-osmolarity conditions or better metabolizing the osmolyte as a carbon source. Single-cell imaging demonstrated that enhanced fitness was coupled to faster growth, and metagenomic sequencing revealed mutations that reflected growth trade-offs across osmolarities. Our study demonstrated the utility of long-term evolution experiments for probing adaptation occurring during environmental stress.

Environment-dependent costs and benefits of recombination in independently evolved populations of Escherichia coli.

Understanding of the causes by which reproductive isolation arises remains limited. We examine the role of adaptation in driving reproductive isolation among 12 Escherichia coli populations evolved in two different environments. We found that, regardless of whether parents were selected in the same or different environments, the average fitness of recombinants was lower than the expected, consistent with a prevailing influence of incompatibility between independently accumulated mutations. Exceptions to this pattern occurred among recombinants of some parents evolved in different environments. These recombinants were less fit than expected in the selective environment of one parent, but more fit than expected in the selective environment of the other parent. Our results indicate that both parallel and divergent adaptation can quickly lead to intrinsic genetic barriers contributing to the initial stages of speciation and show that these barriers can be complex, for example, depending on the environment in which recombinant offspring are tested.

Diversity in lac Operon Regulation among Diverse Escherichia coli Isolates Depends on the Broader Genetic Background but Is Not Explained by Genetic Relatedness.

Transcription of bacterial genes is controlled by the coordinated action of cis- and trans-acting regulators. The activity and mode of action of these regulators can reflect different requirements for gene products in different environments. A well-studied example is the regulatory function that integrates the environmental availability of glucose and lactose to control the Escherichia colilac operon. Most studies of lac operon regulation have focused on a few closely related strains. To determine the range of natural variation in lac regulatory function, we introduced a reporter construct into 23 diverse E. coli strains and measured expression with combinations of inducer concentrations. We found a wide range of regulatory functions. Several functions were similar to the one observed in a reference lab strain, whereas others depended weakly on the presence of cAMP. Some characteristics of the regulatory function were explained by the genetic relatedness of strains, indicating that differences varied on relatively short time scales. The regulatory characteristics explained by genetic relatedness were among those that best predicted the initial growth of strains following transition to a lactose environment, suggesting a role for selection. Finally, we transferred the lac operon, with the lacI regulatory gene, from five natural isolate strains into a reference lab strain. The regulatory function of these hybrid strains revealed the effect of local and global regulatory elements in controlling expression. Together, this work demonstrates that regulatory functions can be varied within a species and that there is variation within a species to best match a function to particular environments.IMPORTANCE The lac operon of Escherichia coli is a classic model for studying gene regulation. This study has uncovered features such as the environmental input logic controlling gene expression, as well as gene expression bistability and hysteresis. Most lac operon studies have focused on a few lab strains, and it is not known how generally those findings apply to the diversity of E. coli strains. We examined the environmental dependence of lac gene regulation in 20 natural isolates of E. coli and found a wide range of regulatory responses. By transferring lac genes from natural isolate strains into a common reference strain, we found that regulation depends on both the lac genes themselves and on the broader genetic background, indicating potential for still-greater regulatory diversity following horizontal gene transfer. Our results reveal that there is substantial natural variation in the regulation of the lac operon and indicate that this variation can be ecologically meaningful.

The fitness challenge of studying molecular adaptation.

Advances in bioinformatics and high-throughput genetic analysis increasingly allow us to predict the genetic basis of adaptive traits. These predictions can be tested and confirmed, but the molecular-level changes - i.e. the molecular adaptation - that link genetic differences to organism fitness remain generally unknown. In recent years, a series of studies have started to unpick the mechanisms of adaptation at the molecular level. In particular, this work has examined how changes in protein function, activity, and regulation cause improved organismal fitness. Key to addressing molecular adaptations is identifying systems and designing experiments that integrate changes in the genome, protein chemistry (molecular phenotype), and fitness. Knowledge of the molecular changes underpinning adaptations allow new insight into the constraints on, and repeatability of adaptations, and of the basis of non-additive interactions between adaptive mutations. Here we critically discuss a series of studies that examine the molecular-level adaptations that connect genetic changes and fitness.

Environment changes epistasis to alter trade-offs along alternative evolutionary paths.

The fitness effect of a mutation can depend on both its genetic background, known as epistasis, and the prevailing external environment. Many examples of these dependencies are known, but few studies consider both aspects in combination, especially as they affect mutations that have been selected together. We examine interactions between five coevolved mutations in eight diverse environments. We find that mutations are, on average, beneficial across environments, but that there is high variation in their fitness effects, including many examples of mutations conferring a cost in some, but not other, genetic background-environment combinations. Indeed, even when global interaction trends are accounted for, specific local mutation interactions are common and differed across environments. One consequence of this dependence is that the range of trade-offs in genotype fitness across selected and alternative environments are contingent on the particular evolutionary path followed over the mutation landscape. Finally, although specific interactions were common, there was a consistent pattern of diminishing returns epistasis whereby mutation effects were less beneficial when added to genotypes of higher fitness. Our results underline that specific mutation effects are highly dependent on the combination of genetic and external environments, and support a general relationship between a genotype's current fitness and its potential to increase in fitness.

Predicting microbial growth in a mixed culture from growth curve data.

Determining the fitness of specific microbial genotypes has extensive application in microbial genetics, evolution, and biotechnology. While estimates from growth curves are simple and allow high throughput, they are inaccurate and do not account for interactions between costs and benefits accruing over different parts of a growth cycle. For this reason, pairwise competition experiments are the current "gold standard" for accurate estimation of fitness. However, competition experiments require distinct markers, making them difficult to perform between isolates derived from a common ancestor or between isolates of nonmodel organisms. In addition, competition experiments require that competing strains be grown in the same environment, so they cannot be used to infer the fitness consequence of different environmental perturbations on the same genotype. Finally, competition experiments typically consider only the end-points of a period of competition so that they do not readily provide information on the growth differences that underlie competitive ability. Here, we describe a computational approach for predicting density-dependent microbial growth in a mixed culture utilizing data from monoculture and mixed-culture growth curves. We validate this approach using 2 different experiments with Escherichia coli and demonstrate its application for estimating relative fitness. Our approach provides an effective way to predict growth and infer relative fitness in mixed cultures.

The influence of depth and a subsea pipeline on fish assemblages and commercially fished species.

Knowledge of marine ecosystems that grow and reside on and around subsea oil and gas infrastructure is required to understand impacts of this offshore industry on the marine environment and inform decommissioning decisions. This study used baited remote underwater stereo-video systems (stereo-BRUVs) to compare species richness, fish abundance and size along 42.3 km of subsea pipeline and in adjacent areas of varying habitats. The pipeline is laid in an onshore-offshore direction enabling surveys to encompass a range of depths from 9 m nearshore out to 140 m depth offshore. Surveys off the pipeline were performed across this depth range and in an array of natural habitats (sand, macroalgae, coral reef) between 1 km and 40 km distance from the pipeline. A total of 14,953 fish were observed comprising 240 species (131 on the pipeline and 225 off-pipeline) and 59 families (39 on the pipeline and 56 off-pipeline) and the length of 8,610 fish were measured. The fish assemblage on and off the pipeline was similar in depths of <80 m. In depths beyond 80 m, the predominant habitat off-pipeline was sand and differences between fish assemblages on and off-pipeline were more pronounced. The pipeline was characterised by higher biomass and abundances of larger-bodied, commercially important species such as: Pristipomoides multidens (goldband snapper), Lutjanus malabaricus (saddletail snapper) and Lutjanus russellii (Moses' snapper) among others, and possessed a catch value 2-3 times higher per stereo-BRUV deployment than that of fish observed off-pipeline. Adjacent natural seabed habitats possessed higher abundances of Atule mate (yellowtail scad), Nemipterus spp. (threadfin bream) and Terapon jarbua (crescent grunter), species of no or low commercial value. This is the first published study to use stereo-BRUVs to report on the importance of subsea infrastructure to commercially important fishes over a depth gradient and increases our knowledge of the fish assemblage associated with subsea infrastructure off north-west Australia. These results provide a greater understanding of ecological and fisheries implications of decommissioning subsea infrastructure on the north-west shelf, and will help better inform decision-making on the fate of infrastructure at different depths.

Effects of Beneficial Mutations in pykF Gene Vary over Time and across Replicate Populations in a Long-Term Experiment with Bacteria.

The fitness effects of mutations can depend on the genetic backgrounds in which they occur and thereby influence future opportunities for evolving populations. In particular, mutations that fix in a population might change the selective benefit of subsequent mutations, giving rise to historical contingency. We examine these effects by focusing on mutations in a key metabolic gene, pykF, that arose independently early in the history of 12 Escherichia coli populations during a long-term evolution experiment. Eight different evolved nonsynonymous mutations conferred similar fitness benefits of ∼10% when transferred into the ancestor, and these benefits were greater than the one conferred by a deletion mutation. In contrast, the same mutations had highly variable fitness effects, ranging from ∼0% to 25%, in evolved clones isolated from the populations at 20,000 generations. Two mutations that were moved into these evolved clones conferred similar fitness effects in a given clone, but different effects between the clones, indicating epistatic interactions between the evolved pykF alleles and the other mutations that had accumulated in each evolved clone. We also measured the fitness effects of six evolved pykF alleles in the same populations in which they had fixed, but at seven time points between 0 and 50,000 generations. Variation in fitness effects was high at intermediate time points, and declined to a low level at 50,000 generations, when the mean fitness effect was lowest. Our results demonstrate the importance of genetic context in determining the fitness effects of different beneficial mutations even within the same gene.

Diminishing-returns epistasis decreases adaptability along an evolutionary trajectory.

Populations evolving in constant environments exhibit declining adaptability. Understanding the basis of this pattern could reveal underlying processes determining the repeatability of evolutionary outcomes. In principle, declining adaptability can be due to a decrease in the effect size of beneficial mutations, a decrease in the rate at which they occur, or some combination of both. By evolving Escherichia coli populations started from different steps along a single evolutionary trajectory, we show that declining adaptability is best explained by a decrease in the size of available beneficial mutations. This pattern reflected the dominant influence of negative genetic interactions that caused new beneficial mutations to confer smaller benefits in fitter genotypes. Genome sequencing revealed that starting genotypes that were more similar to one another did not exhibit greater similarity in terms of new beneficial mutations, supporting the view that epistasis acts globally, having a greater influence on the effect than on the identity of available mutations along an adaptive trajectory. Our findings provide support for a general mechanism that leads to predictable phenotypic evolutionary trajectories.

Benefit of transferred mutations is better predicted by the fitness of recipients than by their ecological or genetic relatedness.

The effect of a mutation depends on its interaction with the genetic background in which it is assessed. Studies in experimental systems have demonstrated that such interactions are common among beneficial mutations and often follow a pattern consistent with declining evolvability of more fit genotypes. However, these studies generally examine the consequences of interactions between a small number of focal mutations. It is not clear, therefore, that findings can be extrapolated to natural populations, where new mutations may be transferred between genetically divergent backgrounds. We build on work that examined interactions between four beneficial mutations selected in a laboratory-evolved population of Escherichia coli to test how they interact with the genomes of diverse natural isolates of the same species. We find that the fitness effect of transferred mutations depends weakly on the genetic and ecological similarity of recipient strains relative to the donor strain in which the mutations were selected. By contrast, mutation effects were strongly inversely correlated to the initial fitness of the recipient strain. That is, there was a pattern of diminishing returns whereby fit strains benefited proportionally less from an added mutation. Our results strengthen the view that the fitness of a strain can be a major determinant of its ability to adapt. They also support a role for barriers of transmission, rather than differential selection of transferred DNA, as an explanation of observed phylogenetically determined patterns of restricted recombination among E. coli strains.

Grappling with anisotropic data, pseudo-merohedral twinning and pseudo-translational noncrystallographic symmetry: a case study involving pyruvate kinase.

Pyruvate kinase is a key regulatory enzyme involved in the glycolytic pathway. The crystal structure of Escherichia coli type I pyruvate kinase was first solved in 1995 at 2.5 Šresolution. However, the space group was ambiguous, being either primitive orthorhombic (P2(1)2(1)2(1)) or C-centred orthorhombic (C222(1)). Here, the structure determination and refinement of E. coli type I pyruvate kinase to 2.28 Šresolution are presented. Using the same crystallization conditions as reported previously, the enzyme was found to crystallize in space group P2(1). Determination of the space group was complicated owing to anisotropic data, pseudo-translational noncrystallographic symmetry and the pseudo-merohedrally twinned nature of the crystal, which was found to have very close to 50% twinning, leading to apparent orthorhombic symmetry and absences that were not inconsistent with P2(1)2(1)2(1). The unit cell contained two tetramers in the asymmetric unit (3720 residues) and, when compared with the orthorhombic structure, virtually all of the residues could be easily modelled into the density. Averaging of reflections into the lower symmetry space group with twinning provided tidier electron density that allowed ∼30 missing residues of the lid domain to be modelled for the first time. Moreover, residues in a flexible loop could be modelled and sulfate molecules are found in the allosteric binding domain, identifying the pocket that binds the allosteric activator fructose 1,6-bisphosphate in this isozyme for the first time. Lastly, we note the pedagogical benefits of difficult structures to emerging crystallographers.

Cellular Growth Arrest and Persistence from Enzyme Saturation.

Metabolic efficiency depends on the balance between supply and demand of metabolites, which is sensitive to environmental and physiological fluctuations, or noise, causing shortages or surpluses in the metabolic pipeline. How cells can reliably optimize biomass production in the presence of metabolic fluctuations is a fundamental question that has not been fully answered. Here we use mathematical models to predict that enzyme saturation creates distinct regimes of cellular growth, including a phase of growth arrest resulting from toxicity of the metabolic process. Noise can drive entry of single cells into growth arrest while a fast-growing majority sustains the population. We confirmed these predictions by measuring the growth dynamics of Escherichia coli utilizing lactose as a sole carbon source. The predicted heterogeneous growth emerged at high lactose concentrations, and was associated with cell death and production of antibiotic-tolerant persister cells. These results suggest how metabolic networks may balance costs and benefits, with important implications for drug tolerance.

Adaptation of Escherichia coli to glucose promotes evolvability in lactose.

The selective history of a population can influence its subsequent evolution, an effect known as historical contingency. We previously observed that five of six replicate populations that were evolved in a glucose-limited environment for 2000 generations, then switched to lactose for 1000 generations, had higher fitness increases in lactose than populations started directly from the ancestor. To test if selection in glucose systematically increased lactose evolvability, we started 12 replay populations--six from a population subsample and six from a single randomly selected clone--from each of the six glucose-evolved founder populations. These replay populations and 18 ancestral populations were evolved for 1000 generations in a lactose-limited environment. We found that replay populations were initially slightly less fit in lactose than the ancestor, but were more evolvable, in that they increased in fitness at a faster rate and to higher levels. This result indicates that evolution in the glucose environment resulted in genetic changes that increased the potential of genotypes to adapt to lactose. Genome sequencing identified four genes--iclR, nadR, spoT, and rbs--that were mutated in most glucose-evolved clones and are candidates for mediating increased evolvability. Our results demonstrate that short-term selective costs during selection in one environment can lead to changes in evolvability that confer longer term benefits.

Constraints on adaptation of Escherichia coli to mixed-resource environments increase over time.

Can a population evolved in two resources reach the same fitness in both as specialist populations evolved in each of the individual resources? This question is central to theories of ecological specialization, the maintenance of genetic variation, and sympatric speciation, yet relatively few experiments have examined costs of generalism over long-term adaptation. We tested whether selection in environments containing two resources limits a population's ability to adapt to the individual resources by comparing the fitness of replicate Escherichia coli populations evolved for 6000 generations in the presence of glucose or lactose alone (specialists), or in varying presentations of glucose and lactose together (generalists). We found that all populations had significant fitness increases in both resources, though the magnitude and rate of these increases differed. For the first 4000 generations, most generalist populations increased in fitness as quickly in the individual resources as the corresponding specialist populations. From 5000 generations, however, a widespread cost of adaptation affected all generalists, indicating a growing constraint on their abilities to adapt to two resources simultaneously. Our results indicate that costs of generalism are prevalent, but may influence evolutionary trajectories only after a period of cost-free adaptation.