The range of hosts exploited by a parasite is determined by several factors, including host availability, infectivity and exploitability. Each of these can be the target of natural selection on both host and parasite, which will determine the local outcome of interactions, and potentially lead to coevolution. However, geographical variation in host use and specificity has rarely been investigated. Maculinea (= Phengaris ) butterflies are brood parasites of Myrmica ants that are patchily distributed across the Palæarctic and have been studied extensively in Europe. Here, we review the published records of ant host use by the European Maculinea species, as well as providing new host ant records for more than 100 sites across Europe. This comprehensive survey demonstrates that while all but one of the Myrmica species found on Maculinea sites have been recorded as hosts, the most common is often disproportionately highly exploited. Host sharing and host switching are both relatively common, but there is evidence of specialization at many sites, which varies among Maculinea species. We show that most Maculinea display the features expected for coevolution to occur in a geographic mosaic, which has probably allowed these rare butterflies to persist in Europe. This article is part of the theme issue ‘The coevolutionary biology of brood parasitism: from mechanism to pattern’.

Differences in genetic variability between two ecotypes of the endangered myrmecophilous butterfly Phengaris (=Maculinea) alcon -the setting of conservation priorities Abstract. 1. The endangered butterfly Phengaris alcon exists in two ecotypes (P. 'alcon' and P. 'rebeli'), which inhabit contrasting biotopes (wet and warm ⁄ dry grasslands respectively) and use different larval food plants. The initially flower-budfeeding caterpillars complete their development as social parasites of Myrmica ants, and the specificity of these relationships shows geographical variation.2. We studied the genetic structure of 16 populations (365 individuals) of both ecotypes in eastern Europe, sampling P. 'rebeli' in two disjunct areas in Lithuania and southern Poland, and P. 'alcon' on Polish localities between them. We analysed the cytochrome c oxidase subunit I (COI) mitochondrial gene, the EF1-a nuclear gene and five polymorphic microsatellite loci.3. All individuals shared an identical COI haplotype, which we hypothesise may be linked to a selective sweep associated with the presence of the Wolbachia B strain in all populations. 4. For nuclear markers, we did not find a clear pattern reflecting division into two putative ecotypes. However, ecotypes differed significantly in their genetic variability, i.e., the P. 'rebeli' ecotype was less polymorphic, and its populations were much more differentiated (F ST : 0.632 for EF1-a and 0.504 for microsatellites) than the P. 'alcon' ecotype (0.177 and 0.082, respectively). 5. Our microsatellite data suggest that all populations of P. 'alcon' form a single clade but that P 'rebeli' can be split into either six or two clades. The former model would indicate many independent origins, especially in the mountainous areas of southern Poland. The latter, not mutually exclusive, grouping clearly reflects the use of different host ants.

No abstract

No abstract

Abstract.Plebejus argus is a model species for studying the biology, population ecology and genetics of butterflies. It is patchily distributed throughout most of its European range and considered to be sedentary. Habitats of the butterfly have to encompass two vital larval-resources, i.e. specific food plants and ants, since caterpillars are obligatorily myrmecophilous. The genetic structure of nine P. argus populations (85 individuals) was studied at an intermediate geographical scale (Eastern Poland, diameter of about 400 km) using two kinds of molecular markers i.e. COI (mtDNA) and EF-1 (nuclear gene). Both markers were highly variable with as many as 16 haplotypes and 39 alleles, respectively. Great genetic differentiation in the COI gene was detected (overall FST = 0.411, P < 0.001) but little genetic differentiation in the EF-1 gene (FST = 0.021, P < 0.001). The number of COI haplotypes (ranging from one to seven) and their distribution varied considerably among P. argus populations. The possibility that this heterogeneity was related to Wolbachia was excluded as this endoparasitic bacterium was not detected in samples from any of the populations studied. PCA and SAMOVA analyses divided the sampled populations into two or three groups, which could indicate different colonization routes. Moreover, the differences in genetic differentiation with respect to mtDNA and nuclear markers may suggest male-biased dispersal of P. argus at a larger scale. The hypothesis that females are philopatric is consistent with direct observations of the restricted colonization abilities of the butterfly, while the relatively homogeneous genetic structure revealed by previous allozyme studies in some areas might be explained by the possible higher mobility of males.

Small and isolated populations are prone to future extinctions and thus perceived as 'living dead'. Although generally considered to be of low conservation value, their existence can still enhance species survival at the landscape scale through improving the connectivity of other populations and facilitating some (even if little) gene flow. We investigated the demography and genetic status of a tiny and highly isolated local population of Maculinea (= Phengaris) alcon near its distribution margin with the aim of identifying the features that allow it to persist. The study comprised intensive mark-recapture, surveys of Gentiana pneumonanthe foodplants and butterfly eggs laid on them, as well as genetic analyses. The population has been found to be characterised by low genetic diversity and estimated at only a few tens of individuals. The foodplant availability turned out to be the most obvious factor limiting M. alcon abundance. Nevertheless, the life expectancy of adult butterflies is fairly long, and their flight period very short, implying that most individuals occur within the same time window. Together with the relatively little protandry observed, i.e. almost synchronous emergence of males and females, this increases the chances of random mating among the individuals. Moreover, the butterflies move freely across the core habitat fragment. All things concerned, the effective population size is presumably not much lower than the recorded population size. Our findings provide guidelines for pinpointing those among 'living dead' populations that are likely to be the most persistent and thus worth conservation efforts aimed at preserving them.

Colias myrmidone has suffered a dramatic decline in Europe, and now its occurrence is restricted to just a few countries. We investigated one of the last viable metapopulations (Knyszyn Forest, NE Poland), where the butterfly is almost completely dependent on forestry, inhabiting some clearings and young tree plantations rich in larval food plants (Chamaecytisus ruthenicus) and nectar resources. Intensive mark-recapture studies were performed in 2017 on eight occupied patches separated by 0.5-5 km. The overall population size of imagoes in the second brood was calculated at about 750 individuals. Sex ratio was shown to be well-balanced and average residency was estimated at 5.6 days. Dispersal occurred mostly between neighbouring patches, and isolation of patches contributed to the high mortality of emigrants. The average distance covered during dispersal was significantly higher for males. However, females left small patches much more readily than males. These patches were probably used only as 'stepping stone' sites in dispersal. Restricted dispersal could be related to barriers created by forest stands but it is also not excluded that individuals living in an isolated metapopulation become increasingly sedentary and our results are an early warning sign. The most distant local population was clearly isolated, with hardly any immigration from the other populations. To maintain a network of more stable habitat patches some clearings should be left deforested and appropriately managed. However this goal is difficult to achieve under the current forestry rules and therefore (re)creation of habitats on other available open areas should be considered.

No abstract