Dicranotropis hamata (Boheman): phallotreme on the left aedeagus side (Figs 1–4), only in some exceptions on the right side, and slender, more or less straight styles without preapical tooth (Fig. 5), as figured e.g. in Ossiannilsson (1978, Fig. 469). Females display the same type of triangular genital scale as D. remaniaca with narrow and deep caudal incision (Fig. 43), and have similar gonocoxae VIII without basal mediad protruding protuberance.

We examined specimens from Finland (Fig. 55), Denmark (Figs 56, 108, 109), most parts of France (Figs 50–54, 98–107) except for the southwestern regions, Germany (Figs 58, 59, 112–115) except for small areas in the southernmost parts, Czech Republic (Figs 63, 122, 123), Switzerland north of the main Alpine chain (Figs 57, 110, 111), Austria (Figs 60–62, 116–121), a small area in northeasternmost Italy, eastern Slovenia (Figs 65, 126, 127), Hungary (Figs 64, 124, 125), Ukraine (Crimea) (Figs 72, 140, 141), Russia (Krasnodar region) (Figs 73, 142, 143) and on the Balkan Peninsula: Serbia (Figs 66, 128, 129), Montenegro, Bulgaria (Figs 68, 69, 132–135), Macedonia (Figs 67, 130, 131), Albania, Greece (Figs 70, 71, 136–139).

Specimens from northern Poland display characters as those of other Central European regions, but have a small preapical tooth on their styles (Fig. 74, see also fig. 470 in Ossiannilsson 1978).

The pygofer and aedeagus morphology (in males), e.g. width of the aedeagus, number of the aedeagal spines (Figs 98–143), and shape of pygofer protuberances (Figs 189–197, 207–215, 225–242), and the morphology of the genital scale (in females) is to some degree variable. For the variability of the genital styles see Figs 50–73.

D. hamata/D. remaniaca intermediate forms: intermediate style characters were found in specimens from southwestern France (Figs 29, 32, 75, 77–83). They display an aedeagus with phallotreme in some cases on the left, in others on the right side (Figs 27, 28, 30, 31, 144–147, 150–158). One specimen is found also in a more northwestern region (Dep. Saone-et-Loire) (Figs 76, 148, 149). 40 specimens with intermediate characters were examined on the whole: 22 had an aedeagus with phallotreme on the left side and 18 an aedeagus with phallotreme on the right side. 31 specimens were from the same locality (St. Béat), 17 of which had an aedeagus with phallotreme on the left side and 14 with phallotreme on the right side.

The pygofer morphology is illustrated in Figs 198–200, 216–218, 243–248.

Females are undistinguishable from D. remaniaca and D. hamata females.

Dicranotropis zenata Logv.: stout, curved style shape with very long preapical tooth (Fig. 36), a robust aedeagus with phallotreme on the left side (Figs 33–35), and a pygofer similar to that of D. remaniaca (Figs 37–39; see also Fig. 1 in Logvinenko 1969). Females: The genital scale is quite large and displays a proportionally short and very narrow caudal incision (Fig. 47). Gonocoxae VIII similar to those of D. hamata.

Remark: The record of D. hamata from Caucasus (Georgia: Kodžori, two males) by Dlabola (1958, Figs 43, 44) refers apparently to this taxon. The figures of the styles show a long preapical tooth as is typical for this taxon (the lack of the subbasal thorn in these figures is probably due to the fact that Dlabola apparently did not dissect the genital apparatus and therefore the base of the styles were not well visible for him).

Dicranotropis sagata Logv.: small and short styles without preapical tooth and without basal spine shaped protuberance (Fig. 42), and a small aedeagus with low number of teeth on both sides and phallotreme on the right side (Figs 40, 41, see also Figs 25–27 in Logvinenko 1976). The species lacks the large dorsocaudally protruding pygofer protuberances present in D. hamata and D. zenata, but shares with these taxa the small tooth on the inner side of the dorsocaudal pygofer margin. Females display a small elongate genital scale with very long apical incision (Fig. 48). The gonocoxae VIII are narrowed basad and basally abruptly protruding mediad (Fig. 49).

Unfortunately, our knowledge on both taxa from the Caucasus region is based only on very few specimens, thus the range of variability in these taxa is unknown.

Besides the different shape of the genital styles in D. hamata and D. remaniaca, the most distinct difference between both taxa consists in their aedeagus morphology with this structure in one taxon being the mirror image of the other (Figs 1, 18). Exceptions, i.e. aedeagi with phallotreme on the right side in D. hamata (Fig. 118, 119, 138, 139) or with phallotreme on the left side in D. remaniaca (Figs 183, 184), are found in both taxa but they are not very common. It makes approximately 1% in D. remaniaca, and ca. 4% for D. hamata, but for the latter species should be considered that most of the reverse specimens were from localities peripheric within the area of that species and not far from the area of D. remaniaca.

For the explanation of this situation we may go back to a period when the areas of the ancestors of both recent taxa were separated and speciation was in progress.

But before we have to make some considerations: The asymmetry of the aedeagus in Delphacidae (as in most of the other insect groups with asymmetric genital structures) is in most cases directional, i.e. only one of the two mirror symmetric possibilities is observed (rare exceptions are found in many taxa). There are, however, several cases in delphacids where the aedeagus asymmetry is not directional and both possible aedeagus types are present in a proportion of 50:50. This phenomenon was recorded for example for Stiroma affinis Fieber (De Jong 1985) and Chloriona vasconica Ribaut (Guglielmino and Bückle 2010). It is called chiral dimorphism, mirror image dimorphism or antisymmetry, and the two possible mirror images are termed enantiomorphs (Schilthuizen 2013). Apparently, this situation provides no disadvantage for the species and has no impact for the mating ability. In several taxa of insects, species of the same genus differ in the direction of chirality (Schilthuizen 2007, 2013; Huber et al. 2007). This implies that intermediate stages of chiral dimorphism must have existed, either during cladogenesis or during anagenesis.

The aedeagus morphology in D. hamata and D. remaniaca with one species representing the mirror image to the other may be interpreted in the same way, with a transitional stage of antisymmetry and a subsequent return to a directional asymmetry opposite to the original one. As such processes are more likely to occur in small populations, possibly this happened in the ancestor populations of one of the two taxa during a situation where their area was distinctly more limited than now due to climate constraints.

A striking parallel case is to be mentioned in another delphacid genus, Chlorionidea Löw. In central and eastern Europe and central Asia occurs C. flava Löw, on the Apennines C. apenninica Guglielmino and Bückle. Both species differ mostly by differences in the morphology of their anal tube and in their aedeagus morphology with this organ being in one species the mirror image in respect to the other (Guglielmino and Bückle 2010).

Chiral dimorphism is observed also in other groups of insects as in the mantid genus Ciulfina. Populations of four species belonging to this genus were investigated. In one of them a proportion near 50% between both enantiomorphs was observed, in a second one only one enantiomorph was present (directional asymmetry). For two species, however, the proportions of both enantiomorphs were far from 50:50 and unequal among the populations of the same species (Holwell and Herberstein 2010). A completely different situation is found in the snail Partula suturalis Pfeiffer. This species is polymorphic for the direction of coiling. Populations with directional asymmetry are prevailing. Mixed populations are generally small and unstable. As mating between snails of opposite coil is difficult there is apparently a strong selection against chirally dimorphic populations which exist only under special conditions (Johnson et al. 1990).

Preliminary remark: The existence of supposed hybrids between Dicranotropis hamata and D. remaniaca taxa north of the Pyrenees may imply to describe them on a subspecies level. However, in other contact regions between both taxa (South Germany, Slovenia, Switzerland) to date no specimens were observed that present unequivocably intermediate characters in their genital morphology or a mixture of both possible enantiomorphic aedeagus types. In those areas, the species show no signs of hybridisation.

In southwestern France Dicranotropis specimens were found with genital styles (Figs 75, 77–83) that lie in between the shapes of D. hamata and D. remaniaca. Geographically, they are situated between an area with apparently pure D. remaniaca populations on the one side (the Iberian Peninsula, Figs 84–86) and pure D. hamata populations on the other (southeastern and central France, Figs 52–54). Of course, the Pyrenees were an interface between populations of numerous species which expanded from separate glacial refugia. During postglacial expansions, the Pyrenees formed a barrier for populations on both sides of these mountains. The contact between southern D. remaniaca and northern D. hamata populations may have been hindered for a long time, until D. remaniaca populations from the South succeeded in surmounting this barrier and mixed with D. hamatapopulations from the North. Therefore, we interpret the intermediate characters in the genital morphology of the populations immediately north of the Pyrenees as due to hybridization of populations of both species.

In these supposed hybrid populations, aedeagi with phallotreme on the left side (Fig. 30, as in D. hamata) and with phallotreme on the right side (Fig. 28, as in D. remaniaca) are present. Both aedeagus types are not rare. The proportion, based on 40 specimens, is not far from 50:50 (the phallotreme on the left side, i. e. the “hamata-type”, is slightly prevailing). Two scenarios are possible: (1) the fixation of directional asymmetry is lost and the supposed hybridisation resulted in a real antisymmetry (i.e. a not fixed direction of the (asymmetric) aedeagus shape and consequently a 50:50 proportion of both aedeagus types); (2) each specimen has its individual aedeagus orientation not by chance, as in true antisymmetry, but due to special genetic constraints based on the combination of its genetic heritage as the result of hybridisation between populations each of which had their fixed aedeagus directionality. Thus, the hybrid populations consist of a mixture of specimens with different directionally asymmetric aedeagi. In this case, the proportion of the two aedeagus types may be different from the 50:50 proportion, moreover it may be varying between different areas of the hybrid area. This condition may be named “pseudo-antisymmetry”.

We may mention in this context a similar case in the Zyginidia pullula group in northwestern Italy. Mazzoglio and Arzone (1993) found hybrids including a vast spectrum of intermediate forms between Z. pullula (Boheman, 1845) and Z. ribauti Dworakowska, 1970 in Liguria, and Bocca et al. (1988) describe hybrids between Z. pullula and Z. scutellaris (Herrich-Schäffer, 1838) in the Aosta Valley. Della Giustina (1989) considers the possibility of Zyginidia ribauti (present in the whole Apennine Peninsula) being a hybrid between Z. pullula (present in north Italy and eastern Central Europe) and Z. scutellaris (present in Western Europe and Sardinia).

There is little doubt that the division of D. hamata and D. remaniaca from each other happened not long ago, probably during the last glaciation. The two taxa have a nearly identical aedeagus shape (except for the opposite orientation of one taxon in respect of the other, see above), with only some barely discernable differences in the pygofer morphology and the different shape of the central and apical parts of the genital styles, i.e. differences that certainly need a relatively short time to evolve.

We suppose that the area of the common ancestor of both taxa was restricted during a cold climate period, and finally divided in two separate areas, which was the basal situation for a speciation process towards the presently observed two taxa. During a following warmer period both groups may have extended their areas, and developed a hybrid area where they got in contact with each other.

The present disjunct distribution of D. remaniaca, occuring on parts of the Iberian Peninsula on the one and continental Italy with some adjacent areas on the other hand requires further explanation.

One scenario is the colonisation of the Iberian Peninsula directly from Italian mainland or, less probably, viceversa via drifted macropterous specimens crossing the Mediterranean Sea. Generally macropterous specimens are found quite frequently within D. hamata and D. remaniaca populations, even though brachypterous ones prevail by far. Thus, this possibility cannot be completely excluded. On the other side it is noticeable that for D. remaniaca, in spite of the flight ability of macropterous specimens, there are no records from Sicily and Sardinia, though it is present on the entire peninsular Italy until Calabria.

In our opinion another scenario is more probable: we suppose that the taxon in former times had a continuous distribution in the Westmediterranean region (and possibly not only there) including at least southern France. A following restriction of its area due to climatic changes may have resulted in the division in two separated areas on the two Peninsulas, respectively.

Finally, D. hamata populations might have extended their area in southwestern direction, filled in southeastern France the gap between D. remaniaca populations in Italy and Spain and hybridized with D. remaniaca north of the Pyrenees.

In the central part of the Alps D. remaniaca apparently passed the barrier of the main Alpine chains and established itself in a small part of south Germany (probably it is present also in the western parts of Austria: Tirol and Vorarlberg). On the other side D. hamata occurs in a small part of the southern Alps in northern Friuli-Venezia Giulia; north of this area in Carinthia D. hamata is found as well, whereas in western Slovenia D. remaniaca occurs.

It is quite evident that D. sagata differs distinctly from the other taxa treated in this study. The large protruding dorsocaudal protuberances of the pygofer are less developed, the styles (Fig. 42) are small and devoid of a basal spine shaped process, the gonocoxae VIII (Fig. 49) have a distinct basal protuberance. Nevertheless it shares with the three other taxa the general morphology of the aedeagus (even though in a smaller and more simple version, Figs 40, 41), the deep caudal incision in the genital scale of females (Fig. 48) and a small thorn near the caudolateral part of the pygofer. These features might represent a synapomorphy of all four taxa.

D. hamata, D. remaniaca and D. zenata are very closely related taxa. They share with each other (1) the general shape of their pygofer (Figs 7, 23, 38) with its dorsolateral parts distinctly protruding caudad, (2) their aedeagus shape (Figs 1, 2, 17, 18, 33, 34) including the arrangement of spines and teeth on both sides, and (3) the subbasal thorn on their genital styles (Figs 5, 21, 36).

D. zenata differs from D. remaniaca only slightly in the more robust aedeagus (Figs 33, 17), the longer tooth on the genital style (Figs 36, 21), and in the shape of the genital scale (Figs 47, 45). Both species have as a common character an aedeagus with its phallotreme on the right side. They share this aedeagus directionality with D. sagata (Fig. 40), what suggests that this is the plesiomorphic condition and the phallotreme on the left side of the aedeagus in D. hamata is apomorphic. Furthermore, they have generally a shorter pygofer (Figs 38, 23) than D. hamata (Fig. 7), but this character is quite variable. The genital styles with their distinctly bent central part and the robust preapical tooth is structurally similar in D. zenata and D. remaniaca as well (Figs 36, 21), even if this tooth is distinctly longer in D. zenata. Possibly this preapical tooth is a synapomorphic character of both taxa, and D. zenata and D. remaniaca are sister species, and together the sister group to D. hamata. Alternatively, it may represent an apomorphic character of the common ancestor of the three hamata group taxa, which is lost in D. hamata. In this case it is a plesiomorphic character of the three hamata group species and does not support monophyly of D. remaniaca + D. zenata.

The small preapical tooth in two D. hamata specimens from northern Poland (Fig. 74) can be interpreted as a residue of the preapical tooth which is generally lost in D. hamata but was possibly present in its ancestor populations, or it may be a result of hybridisation in the past. Presently these populations are apparently surrounded exclusively by areas with pure D. hamata populations.

For a better understanding of the distribution of D. hamata and D. remaniaca it would be necessary to collect more material above all from the region where the areas of both taxa are adjacent to each other, specifically in the Alpine region, Slovenia, southern Germany, western Alps and southeastern France, but also in Spain, northeastern Europe, and, of course, around the supposed hybrid area in southwestern France.

Furthermore, it would be interesting to compare morphological data, gathered in the presented paper and in future studies, with bioacoustic and molecular data, in order to get further hints on how the present disjunct area of D. remaniaca may be explained, and to assess the hypothesis of a hybrid area in southwestern France.

Crossing experiments between populations from the latter region, and the examination of the offspring of left side phallotreme and right side phallotreme males would as well furnish interesting results.