Conservation implications of genetic variation between spatially and temporally distinct colonies of the endangered damselfly Coenagrion mercuriale

Abstract.  1. Good conservation management is underpinned by a thorough understanding of species' historical and contemporary dispersal capabilities along with the possible adaptive or neutral processes behind any spatio-temporal genetic structuring. These issues are investigated with respect to the rare damselfly Coenagrion mercuriale (Charpentier) – the only odonate species currently listed in the U.K.'s Biodiversity Action Plan – in east Devon where its distribution has become fragmented.
2. The two east Devon C. mercuriale populations, only 3.5 km apart, have accumulated strong differences in frequencies of alleles at 14 microsatellite loci as a consequence of poor adult dispersal and drift. There is no contemporary migration between sites.
3. A genetic signature of population decline at both sites corresponds with known demographic reductions. Coenagrion mercuriale in east Devon are now significantly genetically less diverse than those from a population stronghold in the Itchen Valley.
4. Colonies would benefit from improved connectivity between areas and possibly by a transfer of individuals from other ecologically similar areas.
5. Because C. mercuriale has a semivoltine life cycle throughout the U.K., the possibility that alternate-year cohorts are reproductively isolated is explored. Genetic differentiation among cohorts is an order of magnitude less than between sites, suggesting that some larvae delay their development into adults for a year and recruit to a different cohort.
6. To our knowledge, this is the first study to document migration and gene flow between alternate-year cohorts in a species of odonate. From a conservation standpoint, the cohorts do not require separate management.     

Gene flow, dispersal, and nested clade analysis among populations of the stonefly Peltoperla tarteri in the southern Appalachians

We examined the genetic structure and phylogeography of populations of the stonefly Peltoperla tarteri in the Southern Appalachians to determine the extent and likely mechanism for dispersal of this stream insect. A 454-base-pair (bp) portion of the mitochondrial control region was sequenced from a minimum of 20 individuals from eight populations. Pairwise FST and exact tests showed high levels of differentiation among almost all populations except those on the same stream. amova analysis detected significant genetic differentiation between streams within drainages (phi(SD) = 0.14, P < 0.001), and there was a slight positive correlation between aquatic distance and genetic distance (r = 0.295, P = 0.03). According to nested clade analysis, the present day pattern of genetic variation in P. tarteri is the result of a historical range expansion coupled with restricted gene flow with isolation by distance. Together, these analyses suggest that adult dispersal is limited and that movement by larvae is the primary dispersal mechanism for P. tarteri.     

Allozyme diversity and population genetic structure of the caddisfly Orthopsyche fimbriata and the mayfly Acanthophlebia cruentata in New Zealand streams

1. Allozymes were used to measure genetic variation within and among regional populations of the caddisfly Orthopsyche fimbriata and the mayfly Acanthophlebia cruentata in North Island New Zealand streams.
2. High levels of genetic differentiation were recorded in populations of O. fimbriata within and among catchments separated by more than 100 km, but little or no differentiation in populations separated by around 10 km. The Auckland isthmus appears to be a major barrier to north–south gene flow, with nearly fixed allelic differences at one locus. Genotype frequencies conformed to Hardy–Weinberg equilibrium.
3.  Acanthophlebia cruentata had low levels of genetic variation; the results are unexpected given that O. fimbriata apparently has greater potential for dispersal. The limited genetic data for A. cruentata provided evidence for genetic differentiation among populations separated by around 100 km, or more, within catchments and some differentiation between catchment populations separated by only 25 km.     

Genetic diversity and migration patterns of the aquatic macrophyte Potamogeton malaianus in a potamo-lacustrine system

1. Previously, the Yangtze River connected thousands of shallow lakes which together formed a potamo-lacustrine system capable of sustaining a rich variety of submerged macrophytes.
2.  Potamogeton malaianus is one of the dominant submerged macrophytes in many lakes of this area. Genetic variation and population structure of P. malaianus populations from ten lakes in the potamo-lacustrine system were assessed using inter-simple sequence repeat markers.
3. Twelve primer combinations produced a total of 166 unambiguous bands of which 117 (70.5%) were polymorphic. Potamogeton malaianus exhibited a moderate level of population genetic diversity ( P _P = 70.5%, H _E = 0.163 and I =  0.255), as compared with that of plants in the same habitat and range. The main factors responsible for this moderate value were the plant's mixed breeding system (both sexual and asexual) and the hydrological connectivity among habitats.
4.  F statistics, calculated using different approaches, consistently revealed a moderate genetic differentiation among populations, contributing about 20% of total genetic diversity. An estimate of gene flow (using F _ST) suggested that gene flow played a more important role than genetic drift in the current population genetic structure of P. malaianus ( Nm  = 1.131).
5. The genetic diversity of P. malaianus did not increase downstream. A high level of linkage–disequilibrium at the whole population level suggested that metapopulation processes may affect genetic structure. The migration pattern of P. malaianus was best explained by a two-dimensional stepping stone model, indicating that bird-mediated dispersal could greatly influence gene movements among lakes.     

Gene flow and genetic structure of the aquatic macrophyte Sparganium emersum in a linear unidirectional river

1. River systems offer special environments for the dispersal of aquatic plants because of the unidirectional (downstream) flow and linear arrangement of suitable habitats.
2. To examine the effect of this flow on microevolutionary processes in the unbranched bur-reed ( Sparganium emersum ) we studied the genetic variation within and among nine (sub)populations along a 103 km stretch of the Niers River (Germany–The Netherlands), using amplified fragment length polymorphisms.
3. Genetic diversity in S. emersum populations increased significantly downstream, suggesting an effect of flow on the pattern of intrapopulation genetic diversity.
4. Gene flow in the Niers River is asymmetrically bidirectional, with gene flow being approximately 3.5 times higher in a downstream direction. The observed asymmetry is probably caused by frequent hydrochoric dispersal towards downstream locations on the one hand, and sporadic zoochoric dispersal in an upstream direction on the other. The spread of vegetative propagules (leaf and stem fragments) is probably not an important mode of dispersal for S. emersum , suggesting that gene flow is mainly via seed dispersal. Realized dispersal distances exceeded 60 km, revealing a potential for long-distance dispersal in S. emersum .
5. There was no correlation between geographical and genetic distances among the nine S. emersum populations (i.e. no isolation by distance), which may be due to the occurrence of long-distance dispersal and/or colonization and extinction dynamics in the Niers River.
6. Overall, the genetic population structure and regional dispersal patterns of S. emersum in the Niers River are best explained by a linear metapopulation model. Our study shows that flow can exert a strong influence on population genetic processes of plants inhabiting stream systems.     

DEPENDENCE OF GENE FLOW ON GEOGRAPHIC DISTANCE IN TWO SOLITARY CORALS WITH DIFFERENT LARVAL DISPERSAL CAPABILITIES

When the level of gene flow among populations depends upon the geographic distance separating them, genetic differentiation is relatively enhanced. Although the larval dispersal capabilities of marine organisms generally correlate with inferred levels of average gene flow, the effect of different modes of larval development on the association between gene flow and geographic distance remains unknown. In this paper, I examined the relationship between gene flow and distance in two co-occurring solitary corals. Balanophyllia elegans broods large, nonfeeding planulae that generally crawl only short distances from their place of birth before settling. In contrast, Paracyathus stearnsii free-spawns and produces small planktonic larvae presumably capable of broad dispersal by oceanic currents. I calculated F-statistics using genetic variation at six (P. stearnsii) or seven (B. elegans) polymorphic allozyme loci revealed by starch gel electrophoresis, and used these F-statistics to infer levels of gene flow. Average levels of gene flow among twelve Californian localities agreed with previous studies: the species with planktonic, feeding larvae was less genetically subdivided than the brooding species. In addition, geographic isolation between populations appeared to affect gene flow between populations in very different ways in the two species. In the brooding B. elegans, gene flow declined with increasing separation, and distance explained 31% of the variation in gene flow. In the planktonically dispersed P. stearnsii distance of separation between populations at the scale studied (10–1000 km) explained only 1% of the variation in gene flow between populations. The mechanisms generating geographic genetic differentiation in species with different modes of larval development should vary fundamentally as a result of these qualitative differences in the dependence of gene flow on distance.     

Developmental plasticity as a cohesive evolutionary process between sympatric alternate-year insect cohorts

Many species, particularly insects, pass through a series of distinct phases during their life history, with the developmental timing directed towards appropriate resources. Any factor that creates variation in developmental timing may partition a population into discrete populations-or 'cohorts'. Where there is continued failure to recruit outside the natal cohort then alternate cohorts will have their own internal dynamics, eventually leading to independent demographic and evolutionary trajectories. By contrast, continued variation in development rates within a cohort-cohort splitting-may homogenise otherwise independent demographic units. Using a panel of 14 microsatellite loci, we quantify the genetic signature of apparent demographic isolation between coexisting, but alternate, semivoltine cohorts of the damselfly Coenagrion mercuriale at locations that span its distribution in the UK. We find consistently low levels of genetic divergence between sympatric cohorts of C. mercuriale, indicative of developmental plasticity during the larval stage (unregulated development) whereby some individuals complete their development outside the predominant 2-year (semivoltine) period. Thus, individuals that alter their developmental rate successfully recruit to a different cohort. Despite maintaining contrasting population sizes, gene flow between alternate cohorts broadly is sufficient to place them on a similar evolutionary trajectory and also buffers against loss of genetic diversity. Such flexible larval development permits a response to local conditions and may facilitate response to environmental change.     

Phylogeography of five Polytrichum species within Europe

Using allozymes and microsatellites we have analysed the genetic structure among European populations for several Polytrichum species to infer relevant factors, such as historical events or gene flow, that have shaped their genetic structure. As we observed low levels of genetic differentiation among populations, and no decreasing levels of genetic variation with increasing latitude within most of the examined species, no genetic evidence was obtained for a step-wise recolonization of Europe from southern refugia after the latest glacial period for P. commune , P. uliginosum , P. formosum and P. piliferum . The near absence of population substructuring within these species does indicate that extensive spore dispersal is the most important factor determining the genetic structure among European Polytrichum populations. Gene flow levels have apparently been sufficient to prevent genetic differentiation among populations caused by genetic drift, and to wipe out any genetic structure caused by the postglacial recolonization process. On the other hand, increased genetic differentiation of alpine P. formosum populations suggests that mountain ranges might restrict gene flow significantly among Polytrichum populations. In contrast to most examined Polytrichum species, P. juniperinum showed high levels of genetic differentiation and a profound genetic structure. Assuming that gene flow is not more restricted in P. juniperinum , these findings suggest that this species has recolonized Europe after the latest glacial period from two different refugia, one possibly being the British Isles.  © 2003 The Linnean Society of London, Biological Journal of the Linnean Society 2003, 78, 203–213.     

Genetic population structure of the endemic fourline wrasse (Larabicus quadrilineatus) suggests limited larval dispersal distances in the Red Sea

The connectivity among marine populations is determined by the dispersal capabilities of adults as well as their eggs and larvae. Dispersal distances and directions have a profound effect on gene flow and genetic differentiation within species. Genetic homogeneity over large areas is a common feature of coral reef fishes and can reflect high dispersal capability resulting in high levels of gene flow. If fish larvae return to their parental reef, gene flow would be restricted and genetic differentiation could occur. Larabicus quadrilineatus (Labridae) is considered as an endemic fish species of the Red Sea and Gulf of Aden. The juveniles of this species are cleaner fish that feed on ectoparasites of other fishes. Here, we investigated the genetic population structure and gene flow in L. quadrilineatus among five locations in the Red Sea to infer connectivity among them. To estimate genetic diversity, we analysed 369 bp of 237 mitochondrial DNA control region sequences. Haplotype and nucleotide diversities were higher in the southern than in the northern Red Sea. Analysis of molecular variance (amova) detected the highest significant genetic variation between northern and central/southern populations (Phi(CT) = 0.01; P < 0.001). Migration analysis revealed a several fold higher northward than southward migration, which could be explained by oceanographic conditions and spawning season. Even though the Phi(ST) value of 0.01 is rather low and implies a long larval dispersal distance, estimates based on the isolation-by-distance model show a very low mean larval dispersal distance (0.44-5.1 km) compared to other studies. In order to enable a sustainable ornamental fishery on the fourline wrasse, the results of this study suggest that populations in the northern and southern Red Sea should be managed separately as two different stocks. The rather low larval dispersal distance of about 5 km needs to be considered in the design of marine protected areas to enable connectivity and self-seeding.     

Differential defoliation of Eucalyptus grandis arises from indiscriminant oviposition and differential larval survival

1 The influence of six open-pollinated families (OPFs) of Eucalyptus grandis on both the growth and development of larvae and the oviposition preference of a paropsine chrysomelid ( Paropsis atomaria ) was investigated. The OPFs had previously been identified as differing in their susceptibility to defoliation by P. atomaria in forestry progeny trials.
2 Oviposition preference for resistant and susceptible foliage was tested using binary choice tests. These tests did not demonstrate any significant preference for either resistant or susceptible open-pollinated material indicating that adult host preference for susceptible trees was not a likely cause of differential defoliation.
3 Quantification and analysis of growth and development parameters for all larval stages of P. atomaria showed that feeding on genetic material identified as resistant resulted in a significant reduction of relative growth rate of first instar larvae and an alteration to normal feeding behaviour. There was also a trend towards increased larval mortality on resistant E. grandis .
4 We argue that although the magnitude of these effects was minor, interactions with additional biotic and abiotic sources of mortality in the field have the potential, when magnified over successive generations, to result in significant variation in defoliation of host genotypes in the field.     

Dispersal and recruitment of Tasiagma ciliata (Trichoptera: Tasimiidae) in rainforest streams, south-eastern Australia

1. This study examined genetic variation within and among populations of the caddis fly Tasiagma ciliata (Tasimiidae: Trichoptera) from rainforest streams in south-east Queensland, Australia.
2. Very low levels of genetic differentiation at large spatial scales, between subcatchments and between catchments, indicated that dispersal by the winged adults is widespread. However, significant genetic differentiation at the smallest spatial scale examined, within reaches in a single stream, suggested limited movement by larvae within streams.
3. A patchy distribution of deviations from Hardy–Weinberg equilibrium and differences in patterns among allozyme loci suggested that populations in particular reaches were the result of only a few matings.
4. These results are surprising, given the large numbers of larvae present within a single reach. We suggest that stochastic effects of recruitment may underlie much of the spatial and temporal variation in population numbers in these rainforest streams.     

Correlations between observed dispersal capabilities and patterns of genetic differentiation in populations of four aquatic insect species from the Arizona White Mountains, U.S.A.

SUMMARY 1. Dispersal ability is an important ecological factor that can influence population structure. In an attempt to determine the extent that the pattern of genetic differentiation is correlated with dispersal ability in stream-dwelling aquatic insects, we used the amplified fragment length polymorphism (AFLP) technique to characterise genetic variation in four aquatic insect species: Gumaga griseola (Trichoptera: Sericostomatidae), Helicopsyche mexicana (Trichoptera: Helicopsychidae), Psephenus montanus (Coleoptera: Psephenidae) and Ambrysus thermarum (Hemiptera: Naucoridae). Individuals were sampled from several sites within two adjacent catchments in the Arizona White Mountains. In addition to the genetic analyses, a 20-week-long trapping study was used to determine the relative dispersal ability of adults of the four species examined.
2. We obtained hierarchical indicators of genetic differentiation for catchments, sites within catchments and sites across the region examined. Overall, average estimators of genetic differentiation ( F -statistics) were consistent with direct observations of organismal movement, although it was our direct observations on adult insect flight that permitted us to interpret our results correctly. This was because of the fact that a lack of genetic differentiation across watersheds can be interpreted in two ways.
3. In contrast to F -statistics, patterns of genetic isolation by distance for each species more clearly reflected dispersal ability, suggesting that such analytical approaches provide less ambiguous information about the importance of gene flow in the hierarchical partitioning of genetic variation in stream organisms.     

Population genetic structure of two water strider species in the Ecuadorian Amazon

1. We used mtDNA sequence variation to estimate population genetic structure between and among water strider populations of Potamobates williamsi and P . sumaco in the Ecuadorian Amazon.
2. Sequencing of the COI mitochondrial gene revealed 16 haplotypes, which were summarised into four haplotype groups. P . williamsi and P . sumaco shared two common and widespread haplotypes.
3. Population structure was moderate and gene flow was low. Both air and river distances were significantly correlated with gene flow and, thus, indicative of isolation by distance.
4. The genetic structure within and among populations of P. williamsi and P. sumaco was probably not influenced by the dynamic tropical lotic system.     

Local adaptation despite high gene flow in the waterfall‐climbing Hawaiian goby,Sicyopterus stimpsoni

Environmental heterogeneity can promote the emergence of locally adapted phenotypes among subpopulations of a species, whereas gene flow can result in phenotypic and genotypic homogenization. For organisms like amphidromous fishes that change habitats during their life history, the balance between selection and migration can shift through ontogeny, making the likelihood of local adaptation difficult to predict. In Hawaiian waterfall‐climbing gobies, it has been hypothesized that larval mixing during oceanic dispersal counters local adaptation to contrasting topographic features of streams, like slope gradient, that can select for predator avoidance or climbing ability in juvenile recruits. To test this hypothesis, we used morphological traits and neutral genetic markers to compare phenotypic and genotypic distributions in recruiting juveniles and adult subpopulations of the waterfall‐climbing amphidromous goby, Sicyopterus stimpsoni, from the islands of Hawai'i and Kaua'i. We found that body shape is significantly different between adult subpopulations from streams with contrasting slopes and that trait divergence in recruiting juveniles tracked stream topography more so than morphological measures of adult subpopulation differentiation. Although no evidence of population genetic differentiation was observed among adult subpopulations, we observed low but significant levels of spatially and temporally variable genetic differentiation among juvenile cohorts, which correlated with morphological divergence. Such a pattern of genetic differentiation is consistent with chaotic genetic patchiness arising from variable sources of recruits to different streams. Thus, at least in S. stimpsoni, the combination of variation in settlement cohorts in space and time coupled with strong postsettlement selection on juveniles as they migrate upstream to adult habitats provides the opportunity for morphological adaptation to local stream environments despite high gene flow.     

Gene flow of Acanthaster planci (L.) in relation to ocean currents revealed by microsatellite analysis

Population outbreaks of the coral-eating starfish, Acanthaster planci , are hypothesized to spread to many localities in the Indo-Pacific Ocean through dispersal of planktonic larvae. To elucidate the gene flow of A. planci across the Indo-Pacific in relation to ocean currents and to test the larval dispersal hypothesis, the genetic structure among 23 samples over the Indo-Pacific was analysed using seven highly polymorphic microsatellite loci. The F -statistics and genetic admixture analysis detected genetically distinct groups in accordance with ocean current systems, that is, the Southeast African group (Kenya and Mayotte), the Northwestern Pacific group (the Philippines and Japan), Palau, the North Central Pacific group (Majuro and Pohnpei), the Great Barrier Reef, Fiji, and French Polynesia, with a large genetic break between the Indian and Pacific Oceans. A pattern of significant isolation by distance was observed among all samples ( P =  0.001, r  = 0.88, n  = 253, Mantel test), indicating restricted gene flow among the samples in accordance with geographical distances. The data also indicated strong gene flow within the Southeast African, Northwestern Pacific, and Great Barrier Reef groups. These results suggest that the western boundary currents have strong influence on gene flow of this species and may trigger secondary outbreaks.     

Tug of war between continental gene flow and rearing site philopatry in a migratory bird: the sex-biased dispersal paradigm reconsidered

Nonrandom dispersal has been recently advanced as a mechanism promoting fine-scale genetic differentiation in resident populations, yet how this applies to species with high rates of dispersal is still unclear. Using a migratory species considered a classical example of male-biased dispersal (the greater snow goose, Chen caerulescens atlantica ), we documented a temporally stable fine-scale genetic clustering between spatially distinct rearing sites (5–30 km apart), where family aggregates shortly after hatching. Such genetic differentiation can only arise if, in both sexes, dispersal is restricted and nonrandom, a surprising result considering that pairing occurs among mixed flocks of birds more than 3000 km away from the breeding grounds. Fine-scale genetic structure may thus occur even in migratory species with high gene flow. We further show that looking for genetic structure based on nesting sites only may be misleading. Genetically distinct individuals that segregated into different rearing sites were in fact spatially mixed during nesting. These findings provide new, scale-dependent links between genetic structure, pairing, and dispersal and show the importance of sampling different stages of the breeding cycle in order to detect a spatial genetic structure.     

Genetic diversity and population structure of spottedtail goby (Synechogobius ommaturus) based on AFLP analysis

Larval dispersal may have an important impact on genetic structure of benthic fishes. To examine population genetic structure of spottedtail goby Synechogobius ommaturus, samples from five different locations of China and Kunsan population in Korea were analyzed by using amplified fragment length polymorphism (AFLP) technology. A total of 253 bands were identified from 91 individuals by 5 primer combinations and the percentage of polymorphic bands was 43.87%. The average gene diversity was 0.0794 ± 0.1470 and Shannon’s information index was 0.1279 ± 0.2138. The pairwise F_st values ranged from 0.022 to 0.201. The results of AMOVA analysis indicated that 90.54% of the genetic variation contained within populations and 9.46% occurred among populations. The gene flow estimates (Nm) demonstrated that different gene flow existed among populations from 0.994 to 11.114. No significant genealogical branches or clusters were recognized on the UPGMA tree. The results support the hypothesis that larval dispersal ability can be responsible for the genetic diversity and population structuring.     

Spatial genetic structuring in a vagile species, the European wood mouse

We examined the genetic structure of natural populations of the European wood mouse Apodemus sylvaticus at the microgeographic (<3 km) and macrogeographic (>30 km) scales. Ecological and behavioural studies indicate that this species exhibits considerable dispersal relative to its home-range size. Thus, there is potential for high gene flow over larger geographic areas. As levels of population genetic structure are related to gene flow, we hypothesized that population genetic structuring at the microgeographic level should be negligible, increasing only with geographic distance. To test this, four sites were sampled within a microgeographic scale with two additional samples at the macrogeographic level. Individuals ( n =415) were screened and analysed for seven polymorphic microsatellite loci. Contrary to our hypothesis, significant levels of population structuring were detected at both scales. Comparing genetic differentiation with geographic distance suggests increasing genetic isolation with distance. However, this distance effect was non-significant being confounded by surprisingly high levels of differentiation among microgeographic samples. We attribute this pattern of genetic differentiation to the effect of habitat fragmentation, splitting large populations into components with small effective population sizes resulting in enhanced genetic drift. Our results indicate that it is incorrect to assume genetic homogeneity among populations even where there is no evidence of physical barriers and dispersal can occur freely. In the case of A. sylvaticus , it is not clear whether dispersal does not occur across habitat barriers or behavioural dispersal occurs without consequent gene flow.     

Gene flow by larval dispersal in the Antarctic notothenioid fish Gobionotothen gibberifrons

The diversification of the teleost suborder Notothenioidei (Perciformes) in Antarctic waters provides one of the most striking examples of a marine adaptive radiation. Along with a number of adaptations to the cold environment, such as the evolution of antifreeze glycoproteins, notothenioids diversified into eight families and at least 130 species. Here, we investigate the genetic population structure of the humped rockcod ( Gobionotothen gibberifrons ), a benthic notothenioid fish. Six populations were sampled at different locations around the Scotia Sea, comprising a large part of the species' distribution range ( N  = 165). Our analyses based on mitochondrial DNA sequence data (352 bp) and eight microsatellite markers reveal a lack of genetic structuring over large geographic distances (Φ_ST ≤ 0.058, F _ST ≤ 0.005, P values nonsignificant). In order to test whether this was due to passive larval dispersal, we used GPS-tracked drifter trajectories, which approximate movement of passive surface particles with ocean currents. The drifter data indicate that the Antarctic Circumpolar Current (ACC) connects the sampling locations in one direction only (west–east), and that passive transport is possible within the 4-month larval period of G. gibberifrons . Indeed, when applying the isolation-with-migration model in IMA, strong unidirectional west-east migration rates are detected in the humped rockcod. This leads us to conclude that, in G. gibberifrons , genetic differentiation is prevented by gene flow via larval dispersal with the ACC.     

MAINTENANCE OF ECOLOGICALLY SIGNIFICANT GENETIC VARIATION IN THE TIGER SWALLOWTAIL BUTTERFLY THROUGH DIFFERENTIAL SELECTION AND GENE FLOW

Differential selection in a heterogeneous environment is thought to promote the maintenance of ecologically significant genetic variation. Variation is maintained when selection is counterbalanced by the homogenizing effects of gene flow and random mating. In this study, we examine the relative importance of differential selection and gene flow in maintaining genetic variation in Papilio glaucus. Differential selection on traits contributing to successful use of host plants (oviposition preference and larval performance) was assessed by comparing the responses of southern Ohio, north central Georgia, and southern Florida populations of P. glaucus to three hosts: Liriodendron tulipifera, Magnolia virginiana, and Prunus serotina. Gene flow among populations was estimated using allozyme frequencies from nine polymorphic loci. Significant genetic differentiation was observed among populations for both oviposition preference and larval performance. This differentiation was interpreted to be the result of selection acting on Florida P. glaucus for enhanced use of Magnolia, the prevalent host in Florida. In contrast, no evidence of population differentiation was revealed by allozyme frequencies. F_ST-values were very small and Nm, an estimate of the relative strengths of gene flow and genetic drift, was large, indicating that genetic exchange among P. glaucus populations is relatively unrestricted. The contrasting patterns of spatial differentiation for host-use traits and lack of differentiation for electrophoretically detectable variation implies that differential selection among populations will be counterbalanced by gene flow, thereby maintaining genetic variation for host-use traits.