Genetic differentiation of the marbled white butterfly, Melanargia galathea, accounts for glacial distribution patterns and postglacial range expansion in southeastern Europe

Isolation of Mediterranean species in the southern European peninsulas during the cold glacial phases often resulted in differentiation of several genetic lineages confined to the respective peninsulas. However, whilst there is good genetic evidence for multiple refugia in Iberia, there are only limited data available for the Balkans. Therefore, we wish to examine the hypothesis of a strong genetic structuring within southeastern Europe for the existence of multiple Balkan differentiation centres and/or several leading edges. As a model we use the marbled white butterfly, Melanargia galathea. We studied 18 allozyme loci of 564 individuals from 16 populations distributed over a large part of southeastern Europe. The single populations showed moderately high genetic diversity and no northward decline of genetic diversity was detected. The overall genetic differentiation between populations was considerable (F(ST) 7.0%). Cluster analysis discriminated three genetic groups: (i) a western flank in the former Yugoslavia, parts of eastern Austria and Hungary; (ii) an eastern flank with populations from Bulgaria and Romania (south of the southern Carpathians and eastern Carpathians); and (iii) the eastern Carpathian Basin. Hierarchical variance analysis distributed 53% of the variance among populations between these three groups. One sample from the Greek-Bulgarian border clustered within the eastern flank, but showed some tendency towards the eastern Carpathian Basin populations. Two populations from Carinthia clustered together with the eastern Carpathian Basin ones and a population from Styria showed an intermediate genetic composition between the three groups. Most probably, the eastern and the western flank groups are due to postglacial range expansion from the northeastern and the northwestern edges of the glacial differentiation centre (so-called leading edges). The eastern Carpathian Basin group may have resulted from postglacial expansion from northern Greece through valley systems of the central Balkan peninsula, maybe even expanding westwards north of the Balkan mountains reaching some parts of eastern Austria (e.g. Carinthia). Therefore, the Balkanic refugium of M. galathea may or may not have been continuous along the coastal areas of the Mediterranean, but must have been strongly genetically structured.     

Molecular biogeography of the arctic-alpine disjunct burnet moth species Zygaena exulans (Zygaenidae, Lepidoptera) in the Pyrenees and Alps

Aim The phylogeography of ‘southern’ species is relatively well studied in Europe. However, there are few data about ‘northern’ species, and so we studied the population genetic structure of the arctic‐alpine distributed burnet moth Zygaena exulans as an exemplar. Location and methods The allozymes of 209 individuals from seven populations (two from the Pyrenees, five from the Alps) were studied by electrophoresis. Results All 15 analysed loci were polymorphic. The mean genetic diversities were moderately high (A: 1.99; H_e: 11.5; P: 65%). Mean genetic diversities were significantly higher in the Alps than in the Pyrenees in all cases. F_ST was 5.4% and F_IS was 10%. Genetic distances were generally low with a mean of 0.022 between large populations. About 62% of the variance between populations was between the Alps and the Pyrenees. The two samples from the Pyrenees had no significant differentiation, whereas significant differentiation was detected between the populations from the Alps (F_ST = 2.8%, P = 0.02). Main conclusion Zygaena exulans had a continuous distribution between the Alps and the Pyrenees during the last ice age. Most probably, the species was not present in Iberia, and the samples from the Pyrenees are derived from the southern edge of the glacial distribution area and thus became genetically impoverished. Post‐glacial isolation in Alps and Pyrenees has resulted in a weak genetic differentiation between these two disjunct high mountain systems.     

Parapatric diversification after post-glacial range expansion in the gall fly Urophora cardui (Tephritidae)

Aim Primary and secondary genetic clines in post-glacial colonized regions have different implications for biogeographic distributions and the origin of species. Primary clines arise in situ after colonization as adaptive responses to environmental gradients, while secondary clines are caused by contact between vicariant lineages. Here we analyse primary versus secondary origin of a genetic cline in the tephritid fly Urophora cardui in Jutland, Denmark, in a post-glacial landscape. Location Western Palaearctic. Methods Phylogeographic and demographic analyses of U. cardui based on mitochondrial DNA (mtDNA) genealogies, hierarchical genetic variance tests based on allozymes and distribution analysis of a rare allele from the Jutland cline. Results There was no phylogeographic divergence between the Jutland population of U. cardui north of the cline and neighbouring western European regional populations, which all shared the common western European mtDNA haplotype H1. At nuclear loci, by contrast, the North Jutland population was diverged above the mean level of divergence among regional populations and had no loss of genetic variation. A rare allozyme allele that was frequent in the cline area (up to 45%) and was missing north of the cline also occurred at low frequency (0–14%) elsewhere in the sampling range. Shallow phylogeographic divergence was observed between Russian and western European populations and between English and continental populations. Main conclusions The genetic variation patterns support primary cline evolution and parapatric divergence in Jutland following a demographic expansion of a western European ancestral source population of U. cardui, and suggest cryptic refugia and/or selection in other European population assemblages. The patterns of intra-specific regional divergence are discussed with respect to the interpretation of cryptic refugia in Europe after the most recent ice age.     

Limited hybridization along a large contact zone between two genetic lineages of the butterfly Erebia medusa (Satyrinae, Lepidoptera) in Central Europe

Genetic lineages evolving during glacial isolation frequently come into contact as the result of postglacial range expansions. Hybridization often occurs along these contact zones. In Europe, the high mountain systems of the Alps and Pyrenees are well known for their hybrid belts. This article studies the contact zone of the Woodland Ringlet Erebia medusa in a Hercynian mountain area in the Czech-German border region not exceeding 1500 m a.s.l. Hybrid populations between an eastern and a western genetic lineage were detected by amova -based tests, principal component analysis and neighbour joining analysis. Over most of the range of the Czech-German Border Mts, the ridges separate the western and the eastern genetic lineage from each other. However, two important hybrid areas were detected: (1) the watershed of the Ohře river in the north-west of that area, a major valley system passing through these mountains and (2) the high plateaux of the Šumava Mts in the south-east, an extended area of high elevation. The location of hybrid populations in geographical vicinity to non-hybrid populations and the generally low F _IS (2.1%) make reduced fitness of hybrid individuals little likely. The hybrid populations have intermediate genetic diversity between the genetically poor western and the genetically rich eastern lineage populations.     

Intraspecific allozymatic differentiation reveals the glacial refugia and the postglacial expansions of European Erebia medusa (Lepidoptera: Nymphalidae)

Allozyme analysis of Erebia medusa over large regions of Europe revealed a significant population differentiation (F_ST: 0.149 ± 0.016). A UPGMA-analysis showed a division into four major lineages with mean inter-group genetic distances ranging from 0.051 (±0.010) to 0.117 (±0.024). An AMOVA revealed that rather more than two-thirds of the variance between samples was being between these lineages and less than one-third within lineages. An eastern group included the samples from the Czech Republic, Slovakia and north-eastern Hungary. This genetic lineage expressed significantly higher genetic diversity than the other three. A second lineage was formed by the samples from France and Germany. The two samples from western Hungary represent a third delimited lineage and the sample from northern Italy a fourth. We suppose that this genetic differentiation took place during the last ice-age in four disjunct refugia. The genetically more diverse eastern genetic lineage might have evolved in a relatively large refugium in south-eastern Europe. We assume that the other three lineages developed in relatively small relict areas around the Alps. It is likely for the western lineage that its ice-age distribution showed at least one disjunction in late Würm with the consequence of further genetic differentiation. Most probably, the eastern lineage colonized postglacial Central Europe using two alternative routes: one north and one south of the Carpathians. Up to now, neither similar glacial refugia, nor comparable secondary disjunctions in late Würm, are reported for any other animal or plant species.     

The evolutionary significance of bimodal emergence in the butterfly, Maniola jurtina (Lepidoptera: Satyrinae) (L.).

Most temperate butterflies exhibit a tightly synchronized unimodal adult emergence to facilitate mate location. Exceptions are presumably subject to unusual selection pressure. This study examines the pattern of emergence in Maniola jurtina , which was found to exhibit both unimodal and bimodal emergence patterns at different sites in south-east England. The bimodal pattern was found on chalk grassland; elsewhere the emergence was unimodal. Adults from each emergence peak rarely meet, suggesting that there may be some degree of reproductive isolation. Morphological measurements and electrophoretic analysis of allozyme frequencies are carried out to quantify differentiation between emergence peaks. Captive stock was reared to examine differences in the immature stages. Butterflies from each emergence differ significantly in most morphological variables measured, those from the second peak tending to be smaller. The immature stages differ in morphology and longevity of the egg stage. Allozyme frequencies did not differ between peaks, suggesting that they are not reproductively isolated. Explanations for the maintenance of differences between emergence peaks despite gene flow are discussed. I propose that division of offspring between two emergence times may have evolved to avoid the risk inherent in placing all offspring in one peak which may be rendered inviable by temporal fluctuations in habitat quality.     

Reconstruction of the colonization route from glacial refugium to the northern distribution range of the European butterfly Polyommatus coridon (Lepidoptera: Lycaenidae)

Genetic analyses are useful tools for reconstructing glacial distribution patterns and postglacial expansion corridors. However, little information is available about the latter. We reconstruct the expansion corridors of the butterfly Polyommatus coridon from its glacial refugium to the northern edge of its current distribution by comparing populations from southern Lower Saxony (central Germany) to other existing genetic data sets. The populations from Lower Saxony clearly belonged to a western lineage that expanded postglacially from the Adriatic‐Mediterranean region. They form part of a southern German group passing through the Burgundian Gap. In the southern German group, populations belong to a western subgroup. Therefore, expansion followed the Rhine valley and through Hesse, finally reaching southern Lower Saxony and western Thuringia in central Germany. Thus, we present a complete colonization route from the glacial refugium to the northern distribution range of P. coridon. Such data are useful for understanding the biogeographic structure and migration corridors for other mobile Mediterranean species.     

A multivariate approach to the determination of faunal structures among European butterfly species (Lepidoptera: Rhopalocera)

Multivariate analyses of 393 butterfly species over 85 geographical areas (R- and Q-data matrices) in Europe and North Africa have produced a consistent pattern of faunal structures (units and regions). Prominent features to emerge are the latitudinal zonation of geographical units and the division of the Mediterranean into western and eastern components; southwards in Europe, endemicity increases whereas faunal structures decrease in spatial dimensions. Central Europe–from the Urals to the British Isles–forms a single large faunal structure (extent unit and region). A model has been constructed to account for Pleistocene evolutionary changes and endemism in European butterflies and for the east-west taxonomic divisions in the extent faunal structure which dominates central Europe. Periodic Pleistocene climatic changes have resulted in cycles of population extinction, isolation, evolution and migration, but the nature and timing of events has depended on the environmental tolerances of species belonging to different faunal units. During Pleistocene glaciations, southern species have been relatively static and more isolated and have evolved independently. By comparison, northern species have been more mobile and have migrated over large distances. Contact and hybrid zones among cosmopolitan species in northern Europe are probably of some antiquity. They result from persistent survival and isolation of refuge populations in the west and east Mediterranean during glacial phases; dispersal from these refuges leads to their regeneration during each interglacial.     

Is the last glaciation the only relevant event for the present genetic population structure of the meadow brown butterfly Maniola jurtina (Lepidoptera: Nymphalidae)?

Phylogeographical studies are available for a considerable number of European species, but few analyses exist for temperate species with very large and fairly continuous populations that are also absent from Northern Europe. Therefore, we studied the butterfly Maniola jurtina as a model for this group. The species has two major genetic lineages (mean genetic distance between lineages: 0.033; F _CT: 0.052), most probably evolving in glacial differentiation centres in the western and eastern Mediterranean. The onset of this differentiation might have been the beginning of the last glacial stage maximum some 40 kyr bp . A hybrid zone between these two lineages exists in western Central Europe. No genetic substructures have been found within the two lineages ( F _SC: 0.017) and average genetic distances are very small. Therefore, it is highly probable that postglacial expansion was of the phalanx type. There is, at most, very limited differentiation at regional and local scales. However, the genetic diversity within populations is high (means: A : 2.68; H _E: 17.2%; P : 78%), as would be predicted for such a common species. Comparison of these results with a published allozyme analysis revealed a similar phylogeographical pattern, but lower genetic diversity in the latter. Morphological patterns of wings and genitalia show similar geographical patterns as allozyme data.   © 2005 The Linnean Society of London, Biological Journal of the Linnean Society , 2005, 85 , 419–431.     

Africa goes Europe: The complete phylogeography of the marbled white butterfly species complex Melanargia galathea/M. lachesis (Lepidoptera: Satyridae)

Climatic oscillations influence the distribution of species in time. Thermophilic species survived the ice ages in refugia around the Mediterranean. Northern Africa is one of the possibly important refugia. In this study we test the genetic differentiation between northern African and European populations, using the marbled white butterfly species complex, Melanargia galathea/M. lachesis, as a model. We studied 18 allozyme loci in 876 individuals from 23 populations representing a major part of Europe (northern Spain to Romania) and the western part of northern Africa (Atlas Mountains). The African populations resemble the European ones in allelic richness; their genetic diversity is higher than in Europe. Cluster analysis discriminated five European genetic groups: M. lachesis, a western European lineage, and three eastern European lineages. However, the African samples did not form a separate cluster within this phenogram, but clustered randomly within the Balkan/southeastern European groups. The genetic differentiation among the African populations (F_ST 8.8%) was higher than that within any of the European lineages (F_ST 2.6–5.5%). The high genetic diversity and the relatively strong differentiation of the four African populations sampled in a comparatively limited area of the Atlas Mountains indicate that the most probable origin of M. galathea is northern Africa, with its sibling species, M. lachesis, evolving in parallel in Iberia. Most probably, M. galathea colonised Europe first during the Eem interglacial, some 130 ky ago. Since M. lachesis must have existed on the Iberian peninsula during that period already, M. galathea should have reached Europe via Italy. The genetic differentiation to distinct groups in Europe most probably evolved during the following Würm glacial period.     

The species of the genus Zethera Felder (Lepidoptera: Nymphalidae,Satyrinae)

The butterfly genus Zethera Felder, endemic to the Philippines and northern Sulawesi, is shown to include six species, four of which comprise the Z. pimplea (Erichson) superspecies. Type data is given for all nominal taxa, and a key to the species recognised is presented, together with data on their variation and distribution. Attention is drawn to the zoogeographic subregions of the Philippine islands, as originally described by Semper, and illustrated by the Zethera species. Their mimicry is also noted, with particular reference to its sex-limitation in the Z. pimplea superspecies, and non-limitation in Z. incerta (Hewitson) and Z. hestioides C. & R. Felder. The need to demonstrate whether or not Zethera represents a true monophyletic unit is briefly discussed.     

From Asia to Europe: The first record of the blue tiger Tirumala limniace (Cramer, [1775]) (Lepidoptera: Nymphalidae: Danainae) in the European continent

The introduction of species by human action is one of the main causes of rarefaction and extinctions of species. This fact is especially relevant in insular ecosystems. Here, we reported for the first time the arrival of Tirumala limniace (Cramer, [1775]) from the Asian to European continent. A single specimen was recorded in the Balearic Islands, which a large list of introductions around their history. Some reasons could explain the occurrence of this species into the islands, but the most likely are the cargo and nursery trade as well as the release of specimens for weddings and other social events as well as escaped from a private butterfly’s garden. This record highlights the importance of implementing early detection of introduced species in insular ecosystems.     

Dispersal,fragmentation, and isolation shape the phylogeography of the European lineages of Polyommatus (Agrodiaetus) ripartii (Lepidoptera: Lycaenidae)

Polyommatus ripartii is a biogeographically and taxonomically poorly understood species of butterfly with a scattered distribution in Europe. Recently, it has been shown that this species includes several European endemic and localized taxa (galloi, exuberans, agenjoi) that were previously considered species and even protected, a result that poses further questions about the processes that led to its current distribution. We analysed mitochondrial DNA and the morphology of P. ripartii specimens to study the phylogeography of European populations. Three genetically differentiated but apparently synmorphic lineages occur in Europe that could be considered evolutionarily significant units for conservation. Their strongly fragmented and counterintuitive distribution seems to be the result of multiple range expansions and contractions along Pleistocene climatic oscillations. Remarkably, based on the 79 specimens studied, these genetic lineages do not seem to extensively coexist in the distributional mosaic, a phenomenon most evident in the Iberian Peninsula. One of the important gaps in the European distribution of P. ripartii is reduced by the discovery of new Croatian populations, which also facilitate a better understanding of the biogeography of the species. © 2013 The Linnean Society of London, Biological Journal of the Linnean Society, 2013, 109 , 817‐829.     

Effects of recent and past climatic shifts on the genetic structure of the high mountain Yellow‐spotted ringlet butterfly Erebia manto (Lepidoptera,Satyrinae): a conservation problem

Mountain species have evolved important genetic differentiation due to past climatic fluctuations. The genetic uniqueness of many of these lineages is now at risk due to global warming. Here, we analyse allozyme polymorphisms of 1306 individuals (36 populations) of the mountain butterfly Erebia manto and perform Species Distribution Models (SDMs). As a consensus of analyses, we obtained six most likely genetic clusters: (i) Pyrenees with Massif Central; (ii) Vosges; (iii–v) Alps including the Slovakian Carpathians; (vi) southern Carpathians. The Vosges population showed the strongest genetic split from all other populations, being almost as strong as the split between E. manto and its sister species Erebia eriphyle. The distinctiveness of the Pyrenees‐Massif Central group and of the southern Carpathians group from all other groups is also quite high. All three groups are assumed to have survived more than one full glacial–interglacial cycle close to their current distributions with up‐hill and down‐slope shifts conforming climatic conditions. In contrast with these well‐differentiated groups, the three groups present in the Alps and the Slovakian Carpathians show a much shallower genetic structure and thus also should be of a more recent origin. As predicted by our SDM projections, rising temperatures will strongly impact the distribution of E. manto. While the populations in the Alps are predicted to shrink, the survival of the three lineages present here should not be at risk. The situation of the three other lineages is quite different. All models predict the extinction of the Vosges lineage in the wake of global warming, and also the southern Carpathians and Pyrenees‐Massif Central lineages might be at high risk to disappear. Thus, albeit global warming will therefore be unlikely to threaten E. manto as a species, an important proportion of the species’ intraspecific differentiation and thus uniqueness might be lost.     

The effect of ant association on the population genetics of the Australian butterfly Jalmenus evagoras (Lepidoptera: Lycaenidae)

Populations of the myrmecophilous lycaenid Jalmenus evagoras Donovan were assessed for genetic structure at three hierarchical spatial scales: sites, geographically-defined subpopulations, and subpopulations defined by species of mutualistic ant-associate. Estimates of Wright's FST generated from multilocus electrophoretic data revealed low, though significant, amounts of genetic structure. Most structure was observed at the level of geographic subpopulations, suggesting that adult butterflies do not exhibit preferential mating and oviposition along the lines of ant associate. The genetic structure data, together with estimates of Nei's genetic distance (D) for pairwise site and subpopulation comparisons, suggest that J. evagoras populations are spatially and temporally dynamic. These patterns are considered in the context of extinction and recolonization models. The extreme patchiness of J. evagoras populations stems from the stringent requirements of both host plant and host ant, contributing to an extinction/ recolonization process. We discuss the key parameters influencing genetic cohesion versus differentiation under an extinction/recolonization regime, including mode of butterfly dispersal, site turnover rate, and the effects of host dispersal and phenology. This system provides a model of population-level consequences of certain mutualistic interactions as well as of a class of patterns arising from an extinction/recolonization process.     

Origins of recently re‐established and newly discovered populations of the endangered butterfly Shijimiaeoides divinus (Lepidoptera: Lycaenidae) in Oita Prefecture,Japan

The endangered butterfly Shijimiaeoides divinus was believed to have been extirpated from Oita Prefecture, Kyushu, Japan, but was rediscovered in Taketa in recent years. This population is considered to have re‐established as a result of natural dispersal from Kumamoto, a neighboring prefecture located to the west of Oita. Furthermore, another population was recently found in Yufu, Oita Prefecture, which is an area where the species had never been recorded. To elucidate the origins of these two populations newly found from Oita Prefecture, their DNA sequences of the mitochondrial cytochrome c oxidase subunit I (COI) gene were compared with those of other S. divinus populations from Kumamoto Prefecture, Honshu and Korea. The results supported the hypothesis that the Taketa population originated from Kumamoto Prefecture. However, it was not clear whether this population originated from the natural dispersal or deliberate release of individuals. It was also found that the Yufu population was not established by the deliberate release of individuals from Honshu or Korea; however, it remained unclear whether the population of S. divinus was native to Yufu, or originated from other localities in Kyushu.     

The noncosmopolitanism paradigm of freshwater zooplankton: insights from the global phylogeography of the predatory cladoceran Polyphemus pediculus (Linnaeus, 1761) (Crustacea,Onychopoda)

A major question in our understanding of eukaryotic biodiversity is whether small bodied taxa have cosmopolitan distributions or consist of geographically localized cryptic taxa. Here, we explore the global phylogeography of the freshwater cladoceran Polyphemus pediculus (Linnaeus, 1761) (Crustacea, Onychopoda) using two mitochondrial genes, cytochrome c oxidase subunit I and 16s ribosomal RNA, and one nuclear marker, 18s ribosomal RNA. The results of neighbour‐joining and Bayesian phylogenetic analyses reveal an exceptionally pronounced genetic structure at both inter‐ and intra‐continental scales. The presence of well‐supported, deeply divergent phylogroups across the Holarctic suggests that P. pediculus represents an assemblage of at least nine, largely allopatric cryptic species. Interestingly, all phylogenetic analyses support the reciprocal paraphyly of Nearctic and Palaearctic clades. Bayesian inference of ancestral distributions suggests that P. pediculus originated in North America or East Asia and that European lineages of Polyphemus were established by subsequent intercontinental dispersal events from North America. Japan and the Russian Far East harbour exceptionally high levels of genetic diversity at both regional and local scales. In contrast, little genetic subdivision is apparent across the formerly glaciated regions of Europe and North America, areas that historical demographic analyses suggest that were recolonized just 5500–24 000 years ago.     

Diversity and genetic structure of three species of Dioon Lindl. (Zamiaceae, Cycadales) from the Pacific seaboard of Mexico

We have estimated levels of genetic diversity and partitioning in the Mexican endemic cycad species Dioon sonorense, Dioon tomasellii, and Dioon holmgrenii, whose populations are exclusively distributed along the Pacific seaboard. For the three species, the patterns of variation at 19 allozyme loci in a total of 11 populations were evaluated. The average number of alleles per locus was in the range 2.05–1.68, corresponding to the northernmost population of D. sonorense (Mazatán), and the southernmost population of Dioon holmgrenii (Loxicha), respectively. In turn, the percentage of polymorphic loci peaked (94.73) in the El Higueral and Altamirano populations of Dioon tomasellii, and was estimated to be lowest (57.89) in the Loxicha population of D. holmgrenii. The mean expected heterozygosis varied markedly between taxa, with relatively high indices for D. sonorense and D. tomasellii (H_E = 0.314 and 0.295, respectively) and substantially lower values for D. holmgrenii (H_E = 0.170). Comparison of the inferred genetic structure based on F‐statistics for the three species also indicated differences along the north‐south Pacific seaboard axis. For D. sonorense and D. tomasellii, local inbreeding (F_IS) was zero but global inbreeding (F_IT) values were positive and significantly different from zero (0.130 and 0.116, respectively). By contrast, values of both F_IT and F_IS were negative and significantly different from zero (?0.116 and ?0.201, respectively) for D. holmgrenii. The genetic differentiation between populations (F_ST) had positive values in all taxa and corresponded with their geographic location along the north‐south axis: according to this statistic, D. sonorense was the most differentiated species (F_ST = 0.151), D. tomasellii had intermediate values (F_ST = 0.145), and D. holmgrenii was the less differentiated taxon (F_ST = 0.069). Finally, a phenogram representing Nei's genetic distances among populations displayed three major groups, each one corresponding to each of the studied species. Within D. tomasellii (of intermediate geographic distribution), a further division into two clusters corresponded precisely to the pair of populations that are geographically divided by the Trans Mexican Neovolcanic Mountains. © 2008 The Linnean Society of London, Biological Journal of the Linnean Society, 2008, 94 , 765–776.     

Adaptation at specific loci. VII. Natural selection,dispersal and the diversity of molecular-functional variation patterns among butterfly species complexes (Colias: Lepidoptera,Pieridae)

Natural genetic variants at the phosphoglucose isomerase, PGI, gene differ in spatial patterning of their polymorphism among species complexes of Colias butterflies in North America. In both lowland and alpine complexes, molecular-functional properties of the polymorphic genotypes can be used to predict genotype-specific adult flight performances and resulting large genotypic differences in adult fitness components. In the lowland species complex, there is striking uniformity of PGI polymorph frequencies at a number of sites across the American West; this fits with earlier findings of strong, similar differences in fitness components over this range. In an alpine complex, Colias meadii shows similar uniformity of PGI frequencies within habitat types, either montane steppe or alpine tundra, over several hundred kilometres in the absence of dispersal. At the same time, large shifts (10-20%) in frequency of the most common alleles occur between steppe and tundra populations, whether these are isolated or, as in some cases, are in contact and exchange many dispersing adults each generation. Data on male mating success of common C. meadii PGI genotypes in steppe and tundra show heterozygote advantage in both habitat types, with shifts in relative homozygote disadvantage between habitats which are consistent with observed frequency differences. Nonadaptive explanations for this situation are rejected, and alternative, thermal-ecology-based adaptive hypotheses are proposed for later experimental test. These findings show that strong local selection may dominate dispersal as an evolutionary agent, whether or not dispersal is present, and that selection may often be the major force promoting 'cohesion' of species over long distances. This case offers new opportunities for integrating studies of molecular structure and function with ecological aspects of natural selection in the wild, both within and among species.