Evolutionary units and their history
The patterns of genetic differentiation among C. elatus populations are quite similar for mitochondrial (COI) and nuclear (Bog_C3) markers. All examined populations are grouped in a monophyletic clade and distances among subclades are low or moderate (up to 3% for COI and 2% for Bog_C3). This and the lack of clear morphological differences in other characters between C. elatus populations, suggest that all the studied populations belong to single taxon (species), however, divided into subunits. A distinct nature of these units is partially confirmed by other characters like slight differences in morphology (Asian populations designated previously as subspecies C. e. repletus), geographic isolation of populations from different part of species range (mostly around the Black Sea) and ecological (habitat) preferences (‘marshland’ populations). Although these features have not been studied detailed in this work, no strong taxonomic conclusions can be drawn. Probably, more detailed studies on environmental niche divergence between genetically distant lineages (Kozak et al., 2008) could show additional support for their designation as separate taxonomic units within C. elatus, similarly like it has been done, for example, for Ochthebius glaber water beetle (Sánchez-Fernández et al., 2011).
General patterns of mitochondrial and nuclear differentiation in C. elatus populations are similar, although some minor differences between markers are observed. Two geographic regions show concordant differentiation in both markers: the Caucasus and the Azov Sea coast. It is also likely that a distinct clade inhabits western Europe; however, western part of the species range was only represented by a single specimen from France. Some discrepancies between markers were found in the central part of the C. elatus range. Two mitochondrial and two nuclear lineages are present there; however, their distributions do not overlap as almost all populations from this area belong to a single mtDNA clade, with the exception of two populations from Kujawy (north-central Poland), while in respect of the nuclear marker populations in this area are divided into central European and Balkan clades. The distinctiveness of C. e. repletus from Asia has been confirmed. On the other hand, there are only slight genetic indications that the ecologically distinct ‘marshland’ populations from the wet grasslands of Poland and Belarus could be considered a distinct genetic unit. Actually, these beetles may differ only in respect of their habitat preferences and their larger size may be caused by ecological factors. But, one example suggests that something more than mere ecological differences may be at play here. Highly similar mitochondrial haplotypes were found not only in the marshland populations of north-central Poland and Belarus but also in the 80-year-old specimen from the vicinity of Cracow in southern Poland (~300–400 km away in a straight line). It is interesting to note that also this population from Cracow (unfortunately now extinct) inhabited wet meadows. This phenomenon can be explained in three ways: intentional man-made translocation of beetles (rather implausible); natural migration in the past (theoretically possible); or a simultaneous persistence of two ecological forms of C. elatus in north-central Europe: one typical of steppe-like habitats and another one of wet grasslands. The mitochondrial similarities between marshland populations and the fact that they differ slightly from steppic populations in the same area suggest that these forms may be isolated, but this is currently too weak a premise for unit designation. On the other hand, significant genetic differences between populations from central Europe (including the Balkans) and the Azov Sea coast observed in both markers may be a strong indicator for designating these two groups of populations as evolutionary units.
The large-scale genetic structure of C. elatus may result from a fragmentation of the ancestral range and isolation in four or five areas which may have formed refugia of this species: the Caucasus, Crimea, the Balkans and/or central Europe and probably western Europe. This partially fits the major refugia identified in Europe – the Mediterranean and Asia Minor/Caucasus (e.g. Taberlet et al., 1998; Hewitt, 1999; Avise, 2000). Other refugia, located more to the north, were also postulated to exist in western and central Europe, for example, in the Pannonian Basin and in the Carpathians (Stewart & Lister, 2001; Adams, 2002; Schmitt, 2007), eastern Europe (Sommer & Nadachowski, 2006; Markova et al., 2009), and for some European species, also in Asia (Flanders et al., 2009). A similar pattern of genetic diversity is quite common in Eurasian insects (Rozas et al., 2003; Wahlberg & Saccheri, 2007; Maryanska-Nadachowska et al., 2012). It is worth to notice that for steppic insects there is substantial evidence for late Tertiary connection between Asian-eastern European main steppe zone and steppes of Iberian Peninsula (Ribera & Blasco-Zumeta, 1998). Long-term survival and largely independent evolution of steppic insects in these two areas during Pleistocene glaciations could lead to high level of endemism as demonstrated in south-eastern Spain and Anatolia (e.g. Mico et al., 2008; Avgin & Emre, 2009; Piñero et al., 2011).
The most surprising and difficult to explain observation is the distinctiveness of two populations from Kujawy (north-central Poland), which are located among other populations belonging to mitochondrial and nuclear central European clades. The Kujawy clade is highly distinct, with a mitochondrial genetic distance of about 3% mtDNA and approximately 1% nuclear DNA from the other clades. The Kujawy populations could not have persisted in situ for longer than 10 000–12 000 years (Lindner et al., 2006) or 17 000–18 000 years (Marks, 2002; Wysota et al., 2009), as earlier this area had been covered by an ice sheet. These populations are most similar in terms of their nuclear and mitochondrial DNA to the population from the Dniester valley (Ukraine), which suggests that the former had originated somewhere in eastern Europe. The most probable explanation for the presence of these distinct jewel beetles in Kujawy is their recent expansion from the east (probably along the Volhyn Upland migration route; Pawlowski, 1991) followed by settlement in a C. elatus -free area or by replacement of beetles from the central European evolutionary lineage. The presence of a unique and distinct population in Kujawy is not limited only to C. elatus. A similar pattern was found in a steppic weevil Centricnemus leucogrammus (Kajtoch et al., 2009; Kajtoch, 2011). This strongly suggests that relatively recently some steppic beetle populations colonised the Kujawy area from the east. It is difficult to explain why these lineages are absent in other areas of central Europe.
No fossil (from the Pleistocene or earlier times) remains of these beetles are known from the past, so strict dating for this species is impossible. Nevertheless, assuming the mutation rate to be approximately 2% per Myr, as calculated for the mtDNA of other beetles (Farrell, 2001; Ribera et al., 2001; Barraclough & Vogler, 2002), 2–3% of COI distances among C. elatus lineages (so per-lineage divergence 1–1.5%) suggest that they diverged approx. 0.5–0.75 Ma, during the Middle Pleistocene (Donian/Ilynian/Pokrovian stages according to the Russian Plain Stages or Cromerian complex stages according to the North-West European Stages) (Cohen & Gibbard, 2011). Pattern of C. elatus genetic diversity suggests that this species is presently divided into several evolutionary units whose distribution is limited to particular areas. Some of them (these from western Asia and eastern Europe) fit to main ‘warm-stage’ refugia of continental species suggested by Stewart et al. (2010) in Pontic region. Distinct populations from the central Europe (mitochondrial clade extending from Hungary to the Baltic) could be connected to Pannonian ‘warm-stage’ refugia of steppic species proposed by Stewart et al. (2010). Other populations, especially north of Carpathians and in the Balkans, which also may be characterized as distinct evolutionary units (especially Kujawy mtDNA clade and Balkan nuclear clade), strongly suggest that the steppic refugia described by Stewart et al. (2010) do not represent all possible refugial areas in Europe. Such cryptic ‘warm-stage’ refugia are probably located also in western Ukraine – southern Poland, northern Poland – Belarus, the Balkans and probably also in other areas in western Europe.
The highest mitochondrial diversity was found in populations from eastern regions (the Caucasus, the Azov Sea coast and the Dniester in Ukraine) and also in some central European populations (from Slovakia and southern Poland). The lowest mtDNA variation was found in the northernmost populations (from the Baltic region) but also central and southern populations (from the Balkans and Hungary, but samples available for these populations were very small). With respect to the nuclear marker, all populations appeared to be monomorphic, but moderate differentiation was found between most regions with the exception of the Balkans (without Croatia) and the Caucasus, which were surprisingly monomorphic. Female philopatry and only male dispersal is probably an insufficient explanation for this pattern. Some other factors may be involved in the discrepancies between nuclear and mitochondrial diversities, especially in the Caucasus.
The isolation of populations, especially between regions, is corroborated by a high number of private haplotypes, high values of FST indices, positive and high percentages of genetic variation found among these regional groups according to analyses of molecular variance, and significant correlations among genetic and geographic distances for both markers. On the other hand, sharing of mitochondrial haplotypes and the low values of FST indices indicate that gene flow occurs among adjacent populations, for example, between north-eastern Poland and Belarus, within north-central Poland, within southern Poland and within the Balkans. Also, the presence of the same nuclear haplotypes among populations in central Europe, within Balkans and within Caucasus suggests gene flows in these regions; however, due to the lower rates of evolution in the nuclear marker, this may also result from the retention of ancestral alleles.
The genetic distinctiveness of C. elatus populations and various levels of genetic diversity within regional groups of populations should be explained in the light of dispersal abilities of this species. C. elatus is winged; however, in stable environmental conditions (e.g. strawberry plantations) it appears sedentary. It can disperse in natural environments, where conditions are much more unpredictable and suitable patches are more fragmented. Its mobility is probably dependent mostly on availability of host plants and suitable environments. It seems that in areas further south (e.g. around Mediterranean and Black Seas) several distinct evolutionary units occur. These units may have evolved as a result of ancient isolation and low dispersal abilities. On the other hand, populations in central Europe appeared more recently and their relative genetic uniformity may result from recent common ancestry or present possibly migrations (gene flow). The Kujawy population may constitute the exception in this respect. It was showed that beetles living in more stable environmental conditions are worse flyers and have an increased phylogeographical structure than species occupying short-term geologically persistent habitats where they show higher dispersal abilities and increased gene flow among populations (e.g. Abellán et al., 2009; Arribas et al., 2012).
Highly similar patterns – genetic similarities of neighbouring populations and substantial differences between populations from different regions were also found for other steppic beetles from central-east Europe: weevils (Kajtoch, 2011; Kajtoch et al., 2012) and leaf-beetles (Kajtoch et al., 2013; Kubisz et al., 2012a), for steppic insects of Iberian Peninsula: the tiger beetle Cicindela deserticoloides (Diogo et al., 1999) and orthopteran Mioscirtus wagneri (Ortego et al., 2009) and for species distributed in these two steppe areas: butterfly Melitaea cinxia (Wahlberg & Saccheri, 2007) and fly Wohlfahrtia magnifica (Hall et al., 2009). This suggests that most steppic insects (including Coleoptera) are structured geographically and populations of a particular species are highly isolated from other regions, but migration is still possible within regions. This may have important implications for conservation and management strategies for steppic assemblages.
The existence of reciprocally monophyletic groups within C. elatus is a valuable clue for the identification of evolutionary significant units. ESU status should be assigned to all the major regional groups, that is, populations from the Caucasus, the Azov Sea coast and central Europe (from the Baltic coast to the areas surrounding the Carpathian Mountains) with the Balkans. All these groups of populations are genetically distinct (however some of them, especially the Caucasus group, do not received substantial support in the phylogenetic analyses) and thus are genetically isolated from each other. On a more local scale populations from Kujawy in north-central Poland should be also designated as ESU. Very similar conservation units have been recently proposed for other steppic beetles (Kajtoch, 2011; Kajtoch et al., 2013; Kubisz et al., 2012a). Conservation planning and management strategies should be prepared for particular populations of C. elatus. This species is rare and endangered mainly in central European parts of its range, where it inhabits fragmented grasslands (xerothermic or wet), which are seriously threatened, mainly by anthropogenic pressure (intentional degradation through transformations into fields, forestation, or destruction). In southern and eastern Europe and Asia, this species seems to be still quite abundant, so no special protection is currently needed there. Special conservation actions should be undertaken in central Europe not only for this species, but for entire assemblages of steppe-like habitats. Under the present climatic and environmental conditions, additionally amplified by human activity, such habitats must be protected actively by grazing, bush-cutting, or even by periodically prescribed burning, which is considered a good method for keeping open lands, but is illegal, for example, in Poland. Additionally, for some species translocations of individuals or population reintroduction are probably necessary to sustain viable populations. Moreover, assessment of conservation priorities for steppic insects could be based on criteria proposed for, for example, water beetles in Spain (Abellán et al., 2005). Without active protection of grasslands (especially xerothermic areas and marshlands) in central Europe, the future of these habitats and their plant and animal assemblages remain highly uncertain.