Genetic structure and host–parasite co‐divergence: evidence for trait‐specific local adaptation

Host–parasite co‐evolution can lead to genetic differentiation among isolated host–parasite populations and local adaptation between parasites and their hosts. However, tests of local adaptation rarely consider multiple fitness‐related traits although focus on a single component of fitness can be misleading. Here, we concomitantly examined genetic structure and co‐divergence patterns of the trematode Coitocaecum parvum and its crustacean host Paracalliope fluviatilis among isolated populations using the mitochondrial cytochrome oxidase I gene (COI). We then performed experimental cross‐infections between two genetically divergent host–parasite populations. Both hosts and parasites displayed genetic differentiation among populations, although genetic structure was less pronounced in the parasite. Data also supported a co‐divergence scenario between C. parvum and P. fluviatilis potentially related to local co‐adaptation. Results from cross‐infections indicated that some parasite lineages seemed to be locally adapted to their sympatric (home) hosts in which they achieved higher infection and survival rates than in allopatric (away) amphipods. However, local, intrinsic host and parasite characteristics (host behavioural or immunological resistance to infections, parasite infectivity or growth rate) also influenced patterns of host–parasite interactions. For example, overall host vulnerability to C. parvum varied between populations, regardless of parasite origin (local vs. foreign), potentially swamping apparent local co‐adaptation effects. Furthermore, local adaptation effects seemed trait specific; different components of parasite fitness (infection and survival rates, growth) responded differently to cross‐infections. Overall, data show that genetic differentiation is not inevitably coupled with local adaptation, and that the latter must be interpreted with caution in a multi‐trait context.     

Comparative population genetics of a parasitic nematode and its host community: the trichostrongylid Neoheligmonella granjoni and Mastomys rodents in southeastern Senegal

Contrasting host and parasite population genetic structures can provide information about the population ecology of each species and the potential for local adaptation. Here, we examined the population genetic structure of the nematode Neoheligmonella granjoni at a regional scale in southeastern Senegal, using 11 microsatellite markers. Using the results previously obtained for the two main rodent species of the host community, Mastomys natalensis and Mastomys erythroleucus, we tested the hypothesis that the parasite population structure was mediated by dispersal levels of the most vagile host. The results showed similar genetic diversity levels between host and parasite populations, and consistently lower levels of genetic differentiation in N. granjoni, with the exception of one outlying locus with a high F_ST. The aberrant pattern at this locus was primarily due to two alleles occurring at markedly different frequencies in one locality, suggesting selection at this locus, or a closely linked one. Genetic differentiation levels and isolation by distance analyses suggested that gene flow was high and random in N. granjoni at the spatial scale examined. The correlation between pair-wise genetic differentiation levels in the parasite and its main host was consistent with the hypothesis tested. Models of local adaptation as a function of the dispersal rates of hosts and parasites suggest that opportunities for local adaptation would be low in this biological system.     

Fine-scale genetically based differentiation of life-history traits in the perennial shrub Lupinus arboreus

Across large spatial scales, plants often exhibit genetically based differentiation in traits that allow adaptation to local sites. At smaller spatial scales, sharp boundaries between edaphic conditions also can create strong gradients in selection that counteract gene flow and result in local adaptation. Few studies, however, have examined the degree to which continuous populations of perennial plants exhibit genetically based differentiation in life-history traits over small spatial scales. We quantified the degree of genetically based differentiation in adaptive traits among bush lupine (Lupinus arboreus) from nearby dune and grassland sites (sites separated by < 0.75 km) that formed part of a larger continuous population of L. arboreus. We also investigated the spatial genetic structure of bush lupine by examining how genetic structure differed between seeds and juvenile plants that were less than two years old. We calculated F-statistics from gel electrophoresis of 10 polymorphic loci. We then used these values to infer levels of gene flow. To examine differentiation in adaptive traits, we created full-sibling/half-sibling families of lupine within each area and established reciprocal common gardens at each site. Across two years, we measured canopy volume, flowering time, seed set, and mortality of progeny planted in each garden. Spatial genetic structure among seeds was virtually nonexistent (F(ST) = 0.002), suggesting that gene flow between the three areas could be quite high. However, genetic structure increased 20-fold among juvenile plants (F(ST) = 0.041). We found strong evidence for fine-scale genetically based differentiation and local adaptation in adaptive traits such as plant size, flowering phenology, fecundity, and mortality. Thus, it is likely that strong but differing selection regimes within each area drive spatial differentiation in lupine life-history traits.     

Comparative host-parasite population structures: disentangling prospecting and dispersal in the black-legged kittiwake Rissa tridactyla

Although much insight is to be gained through the comparison of the population genetic structures of parasites and hosts, there are, at present, few studies that take advantage of the information on vertebrate life histories available through the consideration of their parasites. Here, we examined the genetic structure of a colonial seabird, the black-legged kittiwake (Rissa tridactyla) using seven polymorphic microsatellite markers to make inferences about population functioning and intercolony dispersal. We sampled kittiwakes from 22 colonies across the species' range and, at the same time, collected individuals of one of its common ectoparasites, the tick Ixodes uriae. Parasites were genotyped at eight microsatellite markers and the population genetic structure of host and parasite were compared. Kittiwake populations are only genetically structured at large spatial scales and show weak patterns of isolation by distance. This may be due to long-distance dispersal events that erase local patterns of population subdivision. However, important additional information is gained by comparing results with those of the parasite. In particular, tick populations are strongly structured at regional scales and show a stepping-stone pattern of gene flow. Due to the parasite's life history, its population structure is directly linked to the frequency and spatial extent of within-breeding season movements of kittiwakes. The comparison of host and parasite gene flow therefore helps us to disentangle the intercolony movements of birds from that of true dispersal events (movement followed by reproduction). In addition, such data can provide essential elements for predicting the outcome of local co-evolutionary interactions.     

Hosts are ahead in a marine host-parasite coevolutionary arms race: innate immune system adaptation in pipefish syngnathus typhle against Vibrio phylotypes

Microparasites have a higher evolutionary potential than their hosts due to an increased mutation rate and a shorter generation time that usually results in parasites being locally adapted to their sympatric hosts. This pattern may not apply to generalist pathogens as adaptation to sympatric host genotypes is disadvantageous due to a narrowing of the host range, in particular under strong gene flow among host populations. Under this scenario, we predict that the immune defense of hosts reveals adaptation to locally common pathogen phylotypes. This was tested in four host populations of the pipefish Syngnathus typhle and associated bacteria of the genus Vibrio. We investigated the population divergence among host and bacteria populations and verified that gene flow is higher among host populations than among parasite populations. Next, we experimentally assessed the strength of innate immune defense of pipefish hosts using in vitro assays that measured antimicrobial activity of blood plasma against sympatric and allopatric Vibrio phylotypes. Pipefish plasma displays stronger antimicrobial activity against sympatric Vibrio phylotypes compared to allopatric ones. This suggests that host defense is genetically adapted against local bacteria with a broad and unspecialized host spectrum, a situation that is typical for marine systems with weak host population structure.     

Tests for Parasite-mediated Frequency-dependent Selection in Natural Populations of an Asexual Plant Species

Genetic variation in plant populations for resistance to pathogens and herbivores might be maintained by parasite-mediated negative frequency-dependent selection (FDS). But it is difficult to observe the time-lagged oscillations between host and parasite genotypes that should result from FDS. To evaluate the potential for FDS, we tested for local adaptation of parasites to common clones, the role of host genetic diversity in resistance to parasites, and genetic correlations among fitness, parasitism, and the frequency of host clones. We studied three populations of Arabis holboellii, a short-lived apomictic (asexual by seed) plant attacked by rust fungi and insect herbivores. To estimate clone frequency, we used polymorphic allozyme markers on 200 individuals in each population in 1990 and in 2000. We also recorded levels of parasitism and host fitness (fruit production). Only the rust fungi showed evidence for local host adaptation; they usually increased in incidence as a function of clone frequency, and they tracked temporal change in clone frequency. In further support of FDS, parasitism was lower in populations with higher genetic diversity. However, total parasitism (herbivory and disease combined) decreased as host clone frequency and fitness increased. Thus, although the highly virulent rust pathogen showed potential for driving the cycles that result from FDS, this apparently does not occur in the populations studied because the host clones were also attacked by herbivores.Co-ordinating editor: J.F. Stuefer     

Dispersal in a parasitic worm and its two hosts: consequence for local adaptation

Characterizing host and parasite population genetic structure and estimating gene flow among populations is essential for understanding coevolutionary interactions between hosts and parasites. We examined the population genetic structure of the trematode Schistosoma mansoni and its two host species (the definitive host Rattus rattus and the intermediate host Biomphalaria glabrata) using microsatellite markers. Parasites were sampled from rats. The study was conducted in five sites of the Guadeloupe Island, Lesser Antilles. Mollusks display a pattern of isolation by distance whereas such a pattern is not found neither in schistosomes nor in rats. The comparison of the distribution of genetic variability in S. mansoni and its two host species strongly suggests that migration of parasites is principally determined by that of the vertebrate host in the marshy focus of Guadeloupe. However, the comparison between genetic differentiation values in schistosomes and rats suggests that the efficacy of the schistosome rat-mediated dispersal between transmission sites is lower than expected given the prevalence, parasitic load and migration rate of rats among sites. This could notably suggest that rat migration rate could be negatively correlated to the age or the infection status of individuals. Models made about the evolution of local adaptation in function of the dispersal rates of hosts and parasites suggest that rats and mollusks should be locally adapted to their parasites.     

Parasites promote host gene flow in a metapopulation

Local adaptation is a powerful mechanism to maintain genetic diversity in subdivided populations. It counteracts the homogenizing effect of gene flow because immigrants have an inferior fitness in the new habitat. This picture may be reversed in host populations where parasites influence the success of immigrating hosts. Here we report two experiments testing whether parasite abundance and genetic background influences the success of host migration among pools in a Daphnia magna metapopulation. In 22 natural populations of D. magna, immigrant hosts were found to be on average more successful when the resident populations experienced high prevalences of a local microsporidian parasite. We then determined whether this success is due to parasitism per se, or the genetic background of the parasites. In a common garden competition experiment, we found that parasites reduced the fitness of their local hosts relatively more than the fitness of allopatric host genotypes. Our experiments are consistent with theoretical predictions based on coevolutionary host-parasite models in metapopulations. A direct consequence of the observed mechanism is an elevated effective migration rate for the host in the metapopulation.     

Comparative population structure and genetic diversity of Arceuthobium americanum (Viscaceae) and its Pinus host species: insight into host-parasite evolution in parasitic angiosperms

In a recent study we revealed that the parasitic angiosperm Arceuthobium americanum is comprised of three distinct genetic races, each associated with a different host in regions of allopatry. In order to assess the role of host identity and geographical isolation on race formation in A. americanum, we compared the genetic population structure of this parasite with that of its three principal hosts, Pinus banksiana, Pinus contorta var. latifolia and Pinus contorta var. murrayana. Despite the fact that A. americanum was divided into three genetic races, hosts were divided into only two genetic groups: (i) Pinus banksiana and hybrids, and (ii) P. contorta var. latifolia and var. murrayana. These findings suggest that factors such as geographical isolation and adaptation to different environmental conditions are important for race formation in the absence of host-driven selection pressures. To assess factors impacting population structure at the fine-scale, genetic and geographical distance matrices of host and parasite were compared within A. americanum races. The lack of a relationship between genetic and geographical distance matrices suggests that isolation-by-distance plays a negligible role at this level. The effect of geographical isolation may have been diminished because of the influence of factors such as random seed dispersal by animal vectors or adaptation to nongeographically patterned environmental conditions. Host-parasite interactions might also have impacted the fine-scale structure of A. americanum because the parasite and host were found to have similar patterns of gene flow.     

Host-dependent genetic structure of parasite populations: differential dispersal of seabird tick host races

Abstract Despite the fact that parasite dispersal is likely to be one of the most important processes influencing the dynamics and coevolution of host-parasite interactions, little information is available on the factors that affect it. In most cases, opportunities for parasite dispersal should be closely linked to host biology. Here we use microsatellite genetic markers to compare the population structure and dispersal of two host races of the seabird tick Ixodes uriae at the scale of the North Atlantic. Interestingly, tick populations showed high within-population genetic variation and relatively low population differentiation. However, gene flow at different spatial scales seemed to depend on the host species exploited. The black-legged kittiwake ( Rissa tridactyla ) had structured tick populations showing patterns of isolation by distance, whereas tick populations of the Atlantic puffin ( Fratercula arctica ) were only weakly structured at the largest scale considered. Host-dependent rates of tick dispersal between colonies will alter infestation probabilities and local dynamics and may thus modify the adaptation potential of ticks to local hosts. Moreover, as I. uriae is a vector of the Lyme disease agent Borrelia burgdorferi sensu lato in both hemispheres, the large-scale movements of birds and the subsequent dispersal of ticks will have important consequences for the dynamics and coevolutionary interactions of this microparasite with its different vertebrate and invertebrate hosts.     

Population structure and the co-evolution between social parasites and their hosts

Co-evolutionary trajectories of host-parasite interactions are strongly affected by the antagonists' evolutionary potential, which in turn depends on population sizes as well as levels of recombination, mutation, and gene flow. Under similar selection pressures, the opponent with the higher evolutionary rate is expected to lead the co-evolutionary arms race and to develop local adaptations. Here, we use mitochondrial DNA sequence data and microsatellite markers to assess the amount of genetic variability and levels of gene flow in two host-parasite systems, each consisting of an ant social parasite--the European slavemaker Harpagoxenus sublaevis and the North American slavemaker Protomognathus americanus--and its two main host species. Our population genetic analyses revealed limited gene flow between individual populations of both host and parasite species, allowing for a geographic mosaic of co-evolution. In a between-system comparison, we found less genetic variability and more pronounced structure in Europe, where previous behavioural studies demonstrated strong local adaptation. Within the European host-parasite system, the larger host species Leptothorax acervorum exhibited higher levels of both genetic variability and gene flow, and previous field data showed that it is less affected by the social parasite H. sublaevis than the smaller host Leptothorax muscorum, which has genetically depleted and isolated populations. In North America, the parasite P. americanus showed higher levels of gene flow between sites, but overall less genetic diversity than its hyper-variable main host species, Temnothorax longispinosus. Interestingly, recent ecological and chemical studies demonstrated adaptation of P. americanus to local host populations, indicating the importance of migration in co-evolutionary interactions.     

Maintaining genetic diversity and population panmixia through dispersal and not gene flow in a holocyclic heteroecious aphid species

Heteroecious holocyclic aphids exhibit both sexual and asexual reproduction and alternate among primary and secondary hosts. Most of these aphids can feed on several related hosts, and invasions to new habitats may limit the number of suitable hosts. For example, the aphid specialist Aphis glycines survives only on the primary host buckthorn (Rhamnus spp.) and the secondary host soybean (Glycine max) in North America where it is invasive. Owing to this specialization and sparse primary host distribution, host colonization events could be localized and involve founder effects, impacting genetic diversity, population structure and adaptation. We characterized changes in the genetic diversity and structure across time among A. glycines populations. Populations were sampled from secondary hosts twice in the same geographical location: once after secondary colonization (early season), and again immediately before primary host colonization (late season). We tested for evidence of founder effects and genetic isolation in early season populations, and whether or not late-season dispersal restored genetic diversity and reduced fragmentation. A total of 24 single-nucleotide polymorphisms and 6 microsatellites were used for population genetic statistics. We found significantly lower levels of genotypic diversity and more genetic isolation among early season collections, indicating secondary host colonization occurred locally and involved founder effects. Pairwise F(ST) decreased from 0.046 to 0.017 in early and late collections, respectively, and while genetic relatedness significantly decreased with geographical distance in early season collections, no spatial structure was observed in late-season collections. Thus, late-season dispersal counteracts the secondary host colonization through homogenization and increases genetic diversity before primary host colonization.     

Influence of anti-filarial chemotherapy strategies on the genetic structure of Wuchereria bancrofti populations

Lymphatic filarial (LF) parasites have been under anti-filarial drug pressure for more than half a century. Currently, annual mass drug administration (MDA) of diethylcarbamazine (DEC) or ivermectin in combination with albendazole (ALB) have been used globally to eliminate LF. Long-term chemotherapies exert significant pressure on the genetic structure of parasitic populations. We investigated the genetic variation among 210 Wuchereria bancrofti populations that were under three different chemotherapy strategies, namely MDA with DEC alone (group I, n = 74), MDA with DEC and ALB (group II, n = 60) and selective therapy (ST) with DEC (group III, n = 34) to understand the impact of these three drug regimens on the parasite genetic structure. Randomly amplified polymorphic DNA profiles were generated for the three groups of parasite populations; the gene diversity, gene flow and genetic distance values were determined and phylogenetic trees were constructed. Analysis of these parameters indicated that parasite populations under ST with a standard dose of DEC (group III) were genetically more diverse (0.2660) than parasite populations under MDA with DEC alone (group I, H = 0.2197) or with DEC + ALB (group II, H = 0.2317). These results indicate that the MDA may reduce the genetic diversity of W. bancrofti populations when compared to the genetic diversity of parasite populations under ST.     

Molecular and quantitative genetic differentiation across Europe in yellow dung flies

Relating geographic variation in quantitative traits to underlying population structure is crucial for understanding processes driving population differentiation, isolation and ultimately speciation. Our study represents a comprehensive population genetic survey of the yellow dung fly Scathophaga stercoraria, an important model organism for evolutionary and ecological studies, over a broad geographic scale across Europe (10 populations from the Swiss Alps to Iceland). We simultaneously assessed differentiation in five quantitative traits (body size, development time, growth rate, proportion of diapausing individuals and duration of diapause), to compare differentiation in neutral marker loci (F(ST)) to that of quantitative traits (Q(ST)). Despite long distances and uninhabitable areas between sampled populations, population structuring was very low but significant (F(ST) = 0.007, 13 microsatellite markers; F(ST) = 0.012, three allozyme markers; F(ST) = 0.007, markers combined). However, only two populations (Iceland and Sweden) showed significant allelic differentiation to all other populations. We estimated high levels of gene flow [effective number of migrants (Nm) = 6.2], there was no isolation by distance, and no indication of past genetic bottlenecks (i.e. founder events) and associated loss of genetic diversity in any northern or island population. In contrast to the low population structure, quantitative traits were strongly genetically differentiated among populations, following latitudinal clines, suggesting that selection is responsible for life history differentiation in yellow dung flies across Europe.     

Hybrid fitness in a locally adapted parasite

The parasite (Red Queen) hypothesis for the maintenance of sexual reproduction and genetic diversity assumes that host-parasite interactions result from tight genetic specificity. Hence, hybridization between divergent parasite populations would be expected to disrupt adaptive gene combinations, leading to reduced infectivity on exposure to parental sympatric hosts, as long as gene effects are nonadditive. In contrast, hybridization would not cause reduced infectivity on allopatric hosts unless the divergent parasite populations possess alleles that are intrinsically incompatible when they are combined. In three different experiments, we compared the infectivity of locally adapted parasite (trematode) populations with that of F(1) hybrid parasites when exposed to host (snail) populations that were sympatric to one of the two parasite populations. We tested for intrinsic genetic incompatibilities in two experiments by including one host population that was allopatric to both parasite populations. As predicted, when the target host populations were sympatric to the parasite populations, the hybrids were significantly less infective than the parental average, while hybrid parasites on allopatric hosts were not, thereby ruling out intrinsic genetic incompatibilities. The results are consistent with nonadditive gene effects and tightly specific host-driven selection underlying parasite divergence, as envisioned by coevolutionary theory and the Red Queen hypothesis.     

Comparative population genetic structures and local adaptation of two mutualists

Similar patterns of dispersal and gene flow between closely associated organisms may promote local adaptation and coevolutionary processes. We compare the genetic structures of the two species of a plant genus (Roridula gorgonias and R. dentata) and their respective obligately associated hemipteran mutualists (Pameridea roridulae and P. marlothi) using allozymes. In addition, we determine whether genetic structure is related to differences in host choice by Pameridea. Allozyme variation was found to be very structured among plant populations but less so among hemipteran populations. Strong genetic structuring among hemipteran populations was only evident when large distances isolated the plant populations on which they live. Although genetic distances among plant populations were correlated with genetic distances among hemipteran populations, genetic distances of both plants and hemipterans were better correlated with geographic distance. Because Roridula and Pameridea have different scales of gene flow, adaptation at the local population level is unlikely. However, the restricted gene flow of both plants and hemipterans could enable adaptation to occur at a regional level. In choice experiments, the hemipteran (Pameridea) has a strong preference for its carnivorous host plant (Roridula) above unrelated host plants. Pameridea also prefers its host species to its closely related sister species. Specialization at the specific level is likely to reinforce cospeciation processes in this mutualism. However, Pameridea does not exhibit intraspecific preferences toward plants from their natal populations above plants from isolated, non-natal populations.     

Contrasting patterns of pollen and seed flow influence the spatial genetic structure of sweet vernal grass (Anthoxanthum odoratum) populations

The spatial genetic structure of plant populations is determined by a combination of gene flow, genetic drift, and natural selection. Gene flow in most plants can result from either seed or pollen dispersal, but detailed investigations of pollen and seed flow among populations that have diverged following local adaptation are lacking. In this study, we compared pollen and seed flow among 10 populations of sweet vernal grass (Anthoxanthum odoratum) on the Park Grass Experiment. Overall, estimates of genetic differentiation that were based on chloroplast DNA (cpDNA) and, which therefore resulted primarily from seed flow, were lower (average F(ST) = 0.058) than previously published estimates that were based on nuclear DNA (average F(ST) = 0.095). Unlike nuclear DNA, cpDNA showed no pattern of isolation by adaptation; cpDNA differentiation was, however, inversely correlated with the number of additions (nutrients and lime) that each plot had received. We suggest that natural selection is restricting pollen flow among plots, whereas nutrient additions are increasing seed flow and genetic diversity by facilitating the successful germination and growth of immigrant seeds. This study highlights the importance of considering all potential gene flow mechanisms when investigating determinants of spatial genetic structure, and cautions against the widespread assumption that pollen flow is more important than seed flow for population connectivity in wind-pollinated species.     

Swimming against the current: genetic structure, host mobility and the drift paradox in trematode parasites

Life-cycle characteristics and habitat processes can potentially interact to determine gene flow and genetic structuring of parasitic species. In this comparative study, we analysed the genetic structure of two freshwater trematode species with different life histories using cytochrome c oxidase I gene (COI) sequences and examined the effect of a unidirectional river current on their genetic diversity at 10 sites along the river. We found moderate genetic structure consistent with an isolation-by-distance pattern among subpopulations of Coitocaecum parvum but not in Stegodexamene anguillae. These contrasting parasite population structures were consistent with the relative dispersal abilities of their most mobile hosts (i.e. their definitive hosts). Genetic diversity decreased, as a likely consequence of unidirectional river flow, with increasing distance upstream in C. parvum, which utilizes a definitive host with only restricted mobility. The absence of such a pattern in S. anguillae suggests that unidirectional river flow affects parasite species differently depending on the dispersal abilities of their most mobile host. In conclusion, genetic structure, genetic diversity loss and drift are stronger in parasites whose most mobile hosts have low dispersal abilities and small home ranges. An additional prediction can be made for parasites under unidirectional drift: those parasites that stay longer in their benthic intermediate host or have more than one benthic intermediate hosts would have relatively high local recruitment and hence increased retention of upstream genetic diversity.     

Temporal patterns of genetic variation for resistance and infectivity in a Daphnia-microparasite system

Theoretical studies have indicated that the population genetics of host-parasite interactions may be highly dynamic. with parasites perpetually adapting to common host genotypes and hosts evolving resistance to common parasite genotypes. The present study examined temporal variation in resistance of hosts and infectivity of parasites within three populations of Daphnia magna infected with the sterilizing bacterium Pasteuria ramosa. Parasite isolates and host clones were collected in each of two years (1997, 1998) from one population; in two other populations, hosts were collected from both years, but parasites from only the first year. We then performed infection experiments (separately for each population) that exposed hosts to parasites from the same year or made combinations involving hosts and parasites from different years. In two populations, patterns were consistent with the evolution of host resistance: either infectivity or the speed with which parasites sterilized hosts declined from 1997 to 1998. In another population, infectivity, virulence, and parasite spore production did not vary among host-year or parasite-year. For this population, we also detected strong within-population genetic variation for resistance. Thus, in this case, genetic variability for fitness-related traits apparently did not translate into evolutionary change. We discuss a number of reasons why genetic change may not occur as expected in parasite-host systems, including negative correlations between resistance and other traits, gene flow, or that the dynamic process itself may obscure the detection of gene frequency changes.     

Population differentiation and genetic variation in host choice among pea aphids from eight host plant genera

Habitat choice plays a critical role in the processes of host range evolution, specialization, and ecological speciation. Pea aphid, Acyrthosiphon pisum, populations from alfalfa and red clover in eastern North America are known to be genetically differentiated and show genetic preferences for the appropriate host plant. This species feeds on many more hosts, and here we report a study of the genetic variation in host plant preference within and between pea aphid populations collected from eight genera of host plants in southeastern England. Most host-associated populations show a strong, genetically based preference for the host plant from which they were collected. Only in one case (populations from Vicia and Trifolium) was there little difference in the plant preference spectrum between populations. All populations showed a significant secondary preference for the plant on which all the aphid lines were reared: broad bean, Vicia faba, previously suggested to be a "universal host" for pea aphids. Of the total genetic variance in host preference within our sample, 61% could be attributed to preference for the collection host plant and a further 9% to systematic differences in secondary preferences with the residual representing within-population genetic variation between clones. We discuss how a combination of host plant preference and mating on the host plant may promote local adaptation and possibly ecological speciation, and whether a widely accepted host could oppose speciation by mediating gene flow between different populations.