http://species.wikimedia.org/wiki/Copelatus_convexus

Neotype, by present designation (NHMB): R91 (imprint) + R74 (counter-imprint) from the locality of Kleinkems (SW Germany, ca. Eocene-Oligocene boundary): fossil of the whole beetle in dorsal view, head, pronotum and elytra almost completely preserved; appendages missing.

Body oblong-oval, broadest in basal third of elytra. Head relatively broad; compound eyes large, not exceeding body outline; clypeus rounded. Pronotum broadest between posterior angles, lateral margins regularly, moderately curved. Mesoscutellar shield well preserved, broadly triangular. Base of elytra as broad as pronotal base; lateral margins of elytra moderately curved. Only mesocoxae, part of metathoracic anepisternum, and probably part of apical abdominal ventrite perceptible from ventral part of body (Figs 1–4).

Surface sculpture. Pronotum with distinct longitudinal median impression, and poorly perceptible short longitudinal striolae on disc. Elytra with 12 moderately impressed longitudinal striae.

Measurements. TL: 6.3 mm, TL-h: 5.6 mm, TW: 3.2 mm. EL: 5.2 mm.

Theneotype corresponds well with the original description and drawings by Förster (1891) in the following characters: (i) general body shape; (ii) shape of the pronotum with projecting anterior angles; (iii) elytra with large number of longitudinal striae [preserved only in posterior portion of elytra in the holotype and their number is therefore estimated by Förster (1891) to be at least 10; 12 striae are present in the specimen from Kleinkems]; (iv) shape and proportion of the scutellar shield [much wider than long]; (v) body size [TL = 6.5 mm, EL = 4.3 mm, TW = 3.8 mm for the specimen from Brunstatt according to Förster (1891)]. The body proportions differ slightly between both specimens (the specimen from Brunstatt is relatively wider), but this may easily have been caused by deformation during the fossilization process or by the inaccuracy of the drawings by Förster (1891) which is quite usual for historic authors (e.g., compare the drawings by Théobald (1937) in Figs 6 and 12 with the actual appearance of the respective fossils). Moreover, Förster (1891) mentioned that his fossil resembles the dytiscid genus Agabus in most characters and did not assign it to the Dytiscidae merely because of its ventral morphology which was reconstructed by him as resembling that of Polyphaga (Fig. 5). It seems that Théobald (1937) examined Förster’s types as he mentioned certain details which are absent in Förster’s (1891) original publication, and his opinion about the conspecificity therefore also has to be considered as a strong argument.

The reasons provided above together with the same age, geological origin and geographical proximity of both outcrops (Brunnstatt and Kleinkems) provide strong support for the conspecificity of both specimens mentioned by Théobald (1937). As the specimen from Brunstatt (i.e., the holotype) is lost, we consider it adequate to designate the specimen from Kleinkems as the neotype.

The preserved characters of the ventral morphology, i.e. the narrow metathoracic anepisternum arising from the median coxal cavity and the anepisterno-metaventral suture directed lateroposteriad correspond closely with the ventral morphology of the Dytiscidae (see, e.g., Fig. 7.6.1 in Balke (2005)). The hydrodynamic body shape, large eyes, broad mesoscutellar shield, medium body size and distinct elytral striae enable us to classify the specimen without any doubt as belonging to the genus Copelatus Erichson, 1832 of the family Dytiscidae.

Copelatus is currently pantropical in its distribution and contains more than 400 described species (Nilsson 2001). Most species of Copelatus are characterised by longitudinal elytral striae whose number has been used to group the species into species groups (Sharp 1882); only a few species have smooth elytra (e.g., Hájek et al. 2010). Although the presence and number of elytral striae provides only limited evidence of phylogeny (Balke et al. 2004), the species groups delimited by number and position of elytral striae are frequently used as a tool for better orientation within the genus (e.g., Guignot 1961; Guéorguiev 1968; Nilsson et al. 1997). The European species previously classified in Copelatus have elytra without striae and have been transferred to the genus Liopterus Dejean, 1833 by Balke et al. (2004); they are not closely related to the fossil dealt with in this study.

Altogether five species of fossil Copelatus species are known: Copelatus aphroditae Balke, 2003 from Baltic amber (Eocene), Copelatus predaveterus Miller, 2003 from Dominican amber (Miocene) (Miller and Balke 2003), and Copelatus fossilis Říha, 1974, Copelatus ponomarenkoi Říha, 1974 and Copelatus stavropolitanus Říha, 1974 from the Miocene deposit of Stavropol (Říha 1974). The differences between all known species are summarized in Table 1.

Copelatus convexus differs from all known fossil and extant species of the genus in the presence of 12 longitudinal striae on each elytron. Sharp (1882) erected a group characterized by 12 discal striae (group 7) for a single species Copelatus interruptus Sharp, 1882 which is, however, currently classified in the genus Exocelina Broun, 1886 (Nilsson 2007). In contrast to the fossil Copelatus convexus, this recent species has elytra with numerous short striolae rather than ‘true’ striae, see, e.g. Wewalka et al. (2010). Therefore, Copelatus convexus might be provisionally classified in a separate Copelatus convexus-group. However, it is necessary to point out that the counting of the precise number of lateral elytral striae is problematic in compressed fossils as the imprint of the submarginal stria may coincide with the lateral margin of the body or with the epipleuron. Therefore, we cannot rule out that a short submarginal stria was present in Copelatus convexus although it is not perceptible in the fossil. In this case, Copelatus convexus would belong to the Copelatus simoni-group sensu Nilsson (2001).

As in the case of Copelatus convexus, Théobald (1937) transferred Escheria crassipunctata to the hydrophilid genus Hydrobius and assigned fossil specimen no. R 707 from the locality of Kleinkems (deposited in NHMB) to this species. We have examined the specimen from Kleinkems for this study (Figs 10–11) but we cannot confirm that it is conspecific with Escheria crassipunctata for the following reasons: (i) the elytra are slightly constricted sub-basally in the specimen from Kleinkems, but evenly rounded laterally in Escheria crassipunctata; (ii) the body outline is distinctly interrupted between the pronotum and the elytra, but it is uninterrupted in Escheria crassipunctata, (iii) eyes are large and globular in the specimen from Kleinkems, but relatively small in Escheria crassipunctata. A more detailed comparison is impossible as the holotype of Escheria crassipunctata is lost and was moreover preserved in dorsal view based on the drawing by Förster (1891), whereas the specimen from Kleinkems is preserved in ventral view.

Based on the original drawing by Förster (1891), Escheria crassipunctata does not bear any synapomorphy of the Hydrophiloidea. For this reason, the species is removed from the fossil record of the Hydrophiloidea and is placed in Polyphaga incertae sedis.

Specimen no. R707 from Kleinkems does not bear any synapomorphy of the Hydrophiloidea, and moreover bears a combination of characters which excludes its assignment to the Hydrophiloidea: (i) prosternal process wide, (ii) mesocoxal cavities rather wide apart, and (iii) eyes large and globular. The preserved characters of this specimen do not allow an unambiguous family assignment (see Lawrence et al. 1999).

Based on the original drawing by Förster (1891), the morphology of Escheria dimidiata agrees with that of Hydrophilidae: Hydrophilinae in many aspects: (i) mesocoxal cavities transverse, narrowly isolated from each other, (ii) metanepisternum rather narrow, (iii) epipleuron narrow but reaching elytral apex, and (iv) elytron with 10 longitudinal punctural series. None of these characters or their combination is, however, unique for the Hydrophiloidea and may be found in other beetle families as well (see e.g. Lawrence et al. 1999). Moreover, the medium body size (EL: 9 mm according to Förster (1891)) would indicate that the fossil should belong to the subtribes Hydrobiusina or Hydrophilina, whose representatives are characterized by a relatively large and well developed triangular mesoscutellar shield; in contrast, the scutellar shield is very small or reduced in Escheria dimidiata. Moreover, the re-examination of the fossil is impossible as the holotype was lost or destroyed. For all these reasons, Escheria dimidiata is removed from the fossil record of the Hydrophiloidea and is placed in Polyphaga incertae sedis.

Based on the drawing by Förster (1891), the ventral morphology of this species agrees with that of Hydrophilidae: Hydrophilinae in many characters: (i) mesocoxae transverse and very narrowly separated, (ii) mesepimeron well separated, triangular, (iii) metanepisternum rather narrow; (iv) abdomen with five ventrites. However, none of these characters or their combination is unique for the Hydrophiloidea and may be found in other beetle families as well (see Lawrence et al. 1999). Moreover, two characters illustrated on the drawing and/or mentioned in the original description exclude the placement of Escheria punctulata in the Hydrophiloidea: (i) elytra bear only 6 deeply impressed striae [9–11 striae are present in all Hydrophiloidea with striate elytra, only rarely is the number of series higher but in such cases they are never impressed to striae]; (ii) mesoventrite fused with mesepisternal (i.e. not separated from them by sutures) [in Hydrophiloidea, the mesoventrite is fused to mesepisterna only in derived groups of the Sphaeridiinae which are characterized by a highly elevated median portion of the mesoventrite; the elevated median elevation is missing from the fossil]. For these reasons, Escheria punctulata is removed from the fossil record of the Hydrophiloidea and is placed into Polyphaga incertae sedis; its family placement remains unclear.