Ghost shrimp Calliax de Saint Laurent, 1973 (Decapoda: Axiidea: Callianassidae) in the fossil record: systematics, palaeoecology and palaeobiogeography.

Ghost shrimps of the family Callianassidae are very common in the fossil record, but mostly as isolated cheliped elements only. The assignment to biologically defined genera, diagnosed on the basis of soft part morphology, is thus rather difficult. In this respect, proxy characters present on chelipeds that are the most durable ghost shrimp remains are needed to ascribe fossil material to extant genera. The genus Calliax de Saint Laurent, 1973 has been particularly obscure in this respect. Thorough comparison of extant members of the genus resulted in evaluation of characters present on chelipeds being taxonomically important on the genus level, specifically: 1) rectangular major P1 propodus with two ridges on the base of the fixed finger extending onto manus; 2) major P1 fingers relatively short; and 3) minor P1 chela with dactylus longer than fixed finger and possessing a wide gap between fingers. On this basis, Callianassa michelottii A. Milne Edwards, 1860, from the Oligocene and Miocene of Europe is herein reassigned to Calliax. Further re-examination of the ghost shrimp fossil record revealed that C. szobensis Müller, 1984, from the Middle Miocene of Hungary represents the same animal as C. michelottii and they are synomymised herein. The known geographic distribution of C. michelottii is expanded by the first confirmed occurrence of the species in Slovakia. All occurrences of C. michelottii known to date are reviewed and documented. The presence of Calliax michelottii comb. nov. may be considered an indicator of deeper marine settings. Based on the scarce fossil record known to date, Calliax has a Tethyan origin; it supposedly migrated westward to establish present day communities in the Caribbean sometime before the Middle Miocene.

The fossil record of ghost shrimps is very robust and they are present in most associations of Cenozoic decapod crustaceans described so far (Glaessner 1969;Bishop & Williams 2005); however, the generic assignment of their remains is rather difficult because their preservation is often incomplete. In general, there are several different views on the evaluation of taxonomically important characters as exemplified by works of Biffar (1971), Manning & Felder (1991), Ngoc-Ho (2003), Poore (1994Poore ( , 2008 and Sakai (1999Sakai ( , 2005Sakai ( , 2011. Palaeontological literature usually emphasizes the contribution of Manning & Felder (1991), which treated some characters present on chelipeds as taxonomically important on the genus level.
Neontological and palaeontological practice commonly handles ghost shrimps differently: whereas zoologists usually treat chelipeds as too variable for being used in taxonomy, palaeontologists have to rely on chelipeds, because only rarely they have more than isolated cheliped elements at hand. It has been, however, demonstrated that using statistically robust data can help to resolve the taxonomic problems caused with sexual dimorphism and intraspecific variation, both in neontological (Dworschak 2012) and palaeontological studies (Hyžný 2012;Hyžný & Hudáčková 2012). More comparative research incorporating views from both sides of the discipline are needed to address the generic status of numerous fossil taxa.
Neontological classification of the extant callianassid ghost shrimps relies mostly on the soft part morphology (Mxp3, male Plp1-2, uropods and telson). Although figuring P1 (chelipeds) is a common practice, only limited attention has been paid to the variation in the nature of chelipeds (Dworschak 2006(Dworschak , 2008(Dworschak , 2011(Dworschak , 2012. Moreover, as stated by Dworschak (2012: p. 54), it is important to study specimens of both sexes over a wide size range. This is not always possible to do because larger sample of these fossorial animals is not at hand.
Palaeontological descriptions often lack comparison with extant ghost shrimp genera, partly because a catchall taxon, Callianassa Leach, 1814, has been successfully embedded in the minds of palaeontologists and partly because of the rather confused state of neontological classification. Major revisions of Sakai (1999Sakai ( , 2005Sakai ( , 2011 remain questionable at the subfamilial and generic levels (cf. Dworschak 2007;Poore 2008;Dworschak et al. 2012). In fact, the basic cheliped morphology of Callianassa as commonly recognized by palaeontologists may comprise numerous distinct morphotypes that can be matched to currently recognized extant genera.
The callianassid genus Calliax de Saint Laurent, 1973, has rather complex taxonomic history and only limited attention has been paid to it when evaluating the ghost shrimp fossil remains. Little fossil material has ever been assigned to the genus (Swen et al. 2001;Schweitzer et al. 2003;Charbonnier et al. 2013), although the present study shows that its fossil record is not as poor as previously thought (Schweitzer et al. 2003). This, of course, owes much to the preservational bias toward isolated cheliped elements, whereas features of the carapace, maxillipeds, eyes, pleopods, uropods and telson, which are used to assign extant species to the genus, are very rarely if at all preserved in the fossil record. It has been argued that at least for some genera the morphology of the major propodus can be diagnostic on the genus level (Hyžný 2012;Hyžný & Müller 2012). In this respect, the revision of the ghost shrimp fossil record is ongoing (Hyžný & Müller 2010Hyžný 2012;Hyžný & Hudáčková 2012;Hyžný & Karasawa 2012;Hyžný et al. 2013). This paper is intended to be a part of the revision and its goal is to discuss in detail assignment of fossil taxa to Calliax based on the case study of Callianassa michelottii A. Milne Edwards, 1860, from the Oligocene and Miocene of Europe.

Localities and geological settings
Personally studied fossil material consists of specimens from two different areas ( Fig. 1): 1) NW part of Mecklenburg-Vorpommern (Germany). At several places the calcareous and siderite sandstone of Oligocene (Chattian) age is exposed. These strata are known as "Sternberger Gestein" and are extremely fossiliferous (e.g. Janke 1993). Specimens of Callianassa michelottii studied herein originate from exposures at Sternberg and Pinnow bei Schwerin.
2) Neogene basins of the former Central Paratethys Sea. The presence of C. michelottii has been confirmed at several Oligo-Miocene (Chattian-Serravallian) localities of Austria, Germany, Slovakia, Hungary and Slovenia, namely: a. North Alpine Foreland Basin (Austria)-Pucking locality was a temporary outcrop in the vicinity of Pucking in Upper Austria. Laminated sandy and silty clays of the Ebelsberg Formation (Krenmayr & Schnabel 2006) were exposed here. They are part of the upper Egerian stage (=Aquitanian) and correspond to the nannoplankton zone NN2 (Gregorová et al. 2009;Grunert et al. 2010). Various aspects of the fossil fauna of the locality were discussed by Gregorová et al. (2009), Grunert et al. (2010, Harzhauser et al. (2011) and Harzhauser & Schlögl (2012).
b. North Alpine Foreland Basin (Germany)-At Neuhofen bei Tettenweis the type locality of the Neuhof Beds (Neuhofener Schichten) is situated. They are composed of clayey to fine sandy marls (Doppler et al. 2005 Remarks on the taxonomy-Calliax has a complex taxonomic history. The genus was erected by de Saint Laurent (1973) with Callianassa lobata de Gaillande & Lagardère, 1966, as the type species. Since then the concept of the genus has been changed several times (cf. Manning & Felder 1991;Sakai 1999Sakai , 2005Sakai , 2011Ngoc-Ho 2003;Hyžný 2012; see also Dworschak 2007). Here the view of Ngoc-Ho (2003) and Sakai (2011) is adopted, and thus, only two formally described extant species are recognized. Discussion on distinguishing Calliax from related taxa based on soft-part morphology was provided by Ngoc-Ho (2003) and will not be repeated here.
When dealing with chelipeds the two known extant species of Calliax can be characterized by unequal and dissimilar chelae, from which the minor one has "fixed finger shorter than and separated from the dactylus by a wide gap, bearing a large triangular proximal tooth (Ngoc-  fig. 17D), C. doerjesti (Sakai 1999: figs. 28, 29b) and Calliax sp. (Taviani et al. 2013: fig. 8) clearly shows consistency in its general shape, i.e. propodus is rectangular and usually longer than high and on its lateral surface the fixed finger possesses two ridges accompanied by furrow (or furrows) parallel to the lower margin of propodus and extending onto manus. The ridges are visible especially when viewing under low angle light (Fig. 2C). There are several distinct setal pores (accompanied by tubercles) arranged obliquely across the lateral surface of propodus. In Calliax the major P1 carpus is always much shorter than manus, with rounded proximo-lower margin (Fig. 3). The merus is longer than high with a distinct meral keel and its lower margin is usually armed with small spines. The above mentioned combination of the characters of minor chela and major propodus, carpus and merus is unique for Calliax; thus, the genus can be identified on the basis of chelipeds alone.
The number of spines on the lower margin of P1 merus may vary between respective members of Calliax and may help in distinguishing taxa at the species level, although possible variation has not been studied in detail yet. Regarding the number of meral spines, there are discrepancies in the literature. Sakai (1999: 114) in the description of C. doerjesti mentioned that the lower margin of the merus was "armed with three interspaced denticles". One of the figures (Sakai 1999: fig. 29b) indeed shows three small spines, however, in the other one (Sakai 1999: fig. 28) depicting the same specimen (holotype) the merus is armed with seven spines (Fig. 3a). Ngoc-Ho (2003: fig. 17D; note that the published figure depicts the right major chela, whereas the caption refers to it as the left one) figured the holotype (male) of C. lobata with seven spines on the merus and de Saint Laurent & Božić (1976: fig. 23a) figured a female specimen of C. lobata also with seven spines. Calliax cf. C. lobata, examined and figured herein (Figs 2, 3E-F), possesses only four blunt spines, presumably mirroring its small size (Fig. 4).
Remarks on the fossil record-Articulated chelipeds are relatively sparse in the fossil state and often only isolated propodi are at hand. In this respect, the major P1 propodus of Calliax is distinct enough to be differentiated from all other ghost shrimp genera. It must be stressed, however, that the more cheliped elements are found, the more secure assignment at the genus level can be provided.
The best preserved and most numerous remains of Calliax in the fossil record belong to species originally described as Callianassa michelottii (Figs 5-10). It is discussed in detail below. Feldmann et al. (2005) reported several isolated cheliped elements from the Miocene of the Navidad Formation of Chile as Callianassoidea sp. 1. Although the authors stated that it does not resemble any callianassoid genus (Feldmann et al. 2005: 431), the figured material (Feldmann et al. 2005: fig. 2A) exhibits striking similarities with Calliax as discussed herein. Interestingly, Callianassa szobensis Müller, 1984, which is herein considered a junior subjective synonym of C. michelottii (see below) and hence a member of Calliax, is mentioned by Feldmann et al. (2005) as similar to their Callianassoidea sp. 1. The same locality also yielded another specimen which has been identified as Callichirus sp. The figured propodus (Feldmann et al. 2005: fig. 2A) shows a relatively short manus; however, the fixed finger with ridges and a furrow indicates its affinities to Callianassoidea sp. 1 (Feldmann et al. 2005: fig. 2D). Both specimens are treated here as Calliax sp. 1. Feldmann et al. (2011) reported fragmented material from the Miocene of Tierra del Fuego (Argentina) as "Cheliped Form B" of indeterminate callianassoid. As already noted by Hyžný & Hudáčková (2012: 13), the minor chela exhibits remarkable similarities to Calliax, (Feldmann et al. 2011: fig. 5E) showing a fixed finger shorter than the dactylus and separated from it by a wide gap with a proximal tooth (cf. Ngoc-Ho 2003: 490). The major P1 propodus (Feldmann et al. 2011: fig. 5B), however, does not possess the ridges on the fixed finger. It is fairly likely that the two specimens do not belong to the same taxon, as they were not found associated with each other. For the purposes of this contribution only the minor chela is referred to here as Calliax sp. 2. This occurrence is considered the oldest confirmed fossil record of the genus, treated here as Calliax sp. 3.
Callianassa whiteavesi Woodward, 1896 from the Campanian of Canada (Woodward 1896;Feldmann & McPherson 1980;Schweitzer et al. 2003) was assigned to Calliax by Schweitzer et al. (2003). The material is rich and sufficiently preserved to reconstruct both chelipeds (Feldmann & McPherson 1980). The species differs markedly from any Calliax species. It does not possess the typically shaped minor cheliped as discussed above, nor has it parallel ridges on the base of the fixed finger. Moreover, some specimens exhibit a rather deep dactylus, a character not observed in Calliax. As a result, the species is excluded from Calliax herein. Until the type material is restudied we suggest to keep the species under Callianassa sensu lato. Occurrence and distribution-Paleocene-Holocene. Two formally described extant species are known from West Atlantic (Florida) and Mediterranean (Sakai 2011). Based on the reports discussed above (Feldmann et al. 2005(Feldmann et al. , 2011, the geographical distribution of the genus was much wider during the Miocene than today, and the genus was apparently also present in the East Pacific (see below). All occurrences are reviewed in Table 1.  Diagnosis-Major cheliped massive; merus longer than high (L/H=2.2) with convex upper margin, carpus higher than long (L/H=0.45), about half the length of manus or shorter, upper margin straight, proximolower margin regularly rounded and smooth in outline; propodus rectangular, longer than high, exhibiting two morphotypes (L/H=1.1-1.2; 1.3-1.5); outer lateral surface of manus smooth, adorned with several tubercles; fixed finger shorter than manus, triangular, with two parallel ridges extending onto manus, lower (less developed) ridge positioned close to the lower propodal margin; occlusal margin of fixed finger serrated and adorned with a blunt tooth pointing up and forward; dactylus slender, with serrated occlusal margin and pointed tip.
Description-Major cheliped massive. Merus longer than high (L/H=2.2), with prominent keel along the midline; upper margin unarmed, convex; lower margin concave, serrated. Carpus higher than long (L/H=0.45), upper margin straight or concave, proximolower margin regularly rounded and smooth in outline. Propodus rectangular, longer than high (L/ H=1.1-1.5), distinctly longer than carpus; upper and lower margins straight, parallel to each other, forming distinct keels on the inner sides, rounded proximally; distal margin straight, parallel to proximal margin, the base of the fixed finger may form a notch. Outer lateral surface of manus smooth, adorned with several tubercles (at the bases of setal pores) lying in a row positioned obliquely. Inner lateral surface of manus concave, with distinct depression at the base of fixed finger. Fixed finger shorter than manus, triangular, with two ridges extending onto manus, lower (less developed) ridge positioned close to the lower propodal margin, upper one (strongly developed) accompanied with several tubercles proximally; occlusal margin serrated and adorned with a blunt tooth pointing up and forward. Dactylus slender, with serrated occlusal margin and pointed tip.
Variations-Major propodi of Calliax michelottii comb. nov. exhibit a certain degree of variation. When examining a large number of specimens, one can observe differences in the manus length/height ratio attaining values in two intervals, i.e. 1.1-1.2 (shorter morph) and 1.3-1.5 (longer morph). Unfortunately it is difficult to state whether it is intraspecific variation or if it mirrors sexual dimorphism. Polkowsky (2005) tentatively interpreted two morphotypes of C. michelottii as sexual dimorphs. Some propodi may possess a relatively well developed notch at the base of the fixed finger. Its development seems to be at least partly correlated with the size, but as already pointed out by Polkowsky (2005), the notch in the longer morphotype is usually better developed. Development of the ridges is a stable character, i.e. there are always two ridges present, and the lower one is rather faint. What differs in various propodi is the shape of the upper ridge in its proximalmost part. In some specimens it is straight, but in others it is curved downwards. The development of tubercles accompanying the upper ridge also varies, depending on the size of the animal. Glaessner (1928) reported strong ornamentation on the isolated propodi from Jarenina, Slovenia (Figs 8A-B). This can, however, be ascribed mainly to preservational aspects. Specimens from Želiezovce, Slovakia (Špinar et al. 1965: fig. X-184; refigured here as Fig.  8L) and Szob, Hungary (Müller 1984: pl. 7, figs. 3, 4; refigured here as Figs 8J-K) exhibit a similar pattern of ornamentation, although not so pronounced as in the material from Jarenina. Possibly it is related to calcification of the cuticle. Experimental data, which are lacking at this point of time, would help in elucidating this issue.
Major P1carpus may have a concave upper margin (Fritsch 1871: pl. 17, fig. 9; refigured here as Fig. 6E). A similarly concave upper margin was also observed in extant Calliax doerjesti (Sakai 1999: fig. 28, 29; Fig. 3A). This feature may be related to age and/or size of the individual. For resolving this issue in Calliax michelottii comb. nov. more preserved carpi must be examined.  Table 2. For details see text below.

Material examined-Only
Remarks-Callianassa michelottii was originally described based on isolated propodi (Fig. 5) from the Miocene of Superga hill near Turin, Italy (A. Milne Edwards 1860). A. Milne Edwards (1860: p. 211) noted that the presence of the keel on the fixed finger accompanied by the tubercles above it is unique not only for fossil species, but this character distinguishes the species from all extant callianassids known at that time as well. Indeed, the type species of Calliax was described more than a century later by de Gaillande & Lagardère (1966). Crema (1895) reported C. michelottii from the Middle Miocene of nearby localities in the Turin hills.
Fritsch (1871) reported C. michelottii from the "Middle" Oligocene (Rupelian) "Septarienthon" of the Mainz Basin (W Germany). Interestingly, Fritsch (1871: pl. 17, figs 5-13) reported and also figured a small chela preserved together with the major P1 propodus (Fig. 6G). He tentatively interpreted it as remains of P2, but it in fact represents a minor P1 chela. The description of Fritsch (1871: 696) is very clear in this sense ("Der bewegliche Finger steht auf einem weit nach vorn vorspringenden Theil des Ballens. Vom unbeweglichen Finger ist nu rein ganz geringer Theil sichtbar, warscheinlich war derselbe indess kurz, die hervortretenden Leisten desselben Gliedes der Vorderhand fehlen." English translation: Dactylus is located on the manus portion projecting forward. Only very small portion of the index is visible. It was probably short, although the tip is missing.) Fritsch (1871: 692) mentioned also the presence of the cephalothorax ("Wohl liegt ein Exemplar vor, an welchem man Theile des kleinen dünnschaligen Cephalothorax und der seitlich zusammengedrückten hinteren Fusspaare erkennt, doch is dasselbe für eine Charakteristik des Thieres zu ungenügend erhalten." English translation: There is a sample in which the individual shows the preserved small cephalothorax with thin cuticle and the laterally compressed "rear" pereiopod. However, the preservation of both characters is insufficient.), but he did not figure it. The material of Fritsch (1871) was reexamined by Beurlen (1931); he confirmed much of what was already done in the original work.
Noetling (1886: pl. 5, fig. 4; refigured here as Figs 6H-K) reported the species from the Late Oligocene "Sternberger Gestein"of Kobrow, Germany. The latest contribution on Callianassa michelottii from this facies is that by Polkowsky (2005) who discussed at length the variation of the species as well as its geographic distribution. Lőrenthey (1907: pl. 1, fig. 5; refigured in Fig. 6L) reported and figured undetermined isolated propodi from the Miocene of Cagliari, Sardinia. With two parallel ridges on the fixed finger, the specimen shows affinities to Calliax michelottii comb. nov., but possesses rather convex upper propodal margin, which is unusual for Calliax, although the character itself may reflect mere intraspecific variation. The material (collection of Lovisato) is considered lost (De Angeli & Garassino 2006); thus, the re-examination of the material is not possible. As a consequence, we treat the occurrence as questionable.
Wagner-Klett (1919) reported Callianassa michelottii from the Oligocene Septarienton of Wiesloch, Germany. The figure he provided (Wagner-Klett 1919: pl. 2, fig. 1), however, does not conform to the diagnosis of the species. The specimen consisting of an articulated carpus, propodus, and dactylus shows no ridges on the fixed finger. The material was either not properly figured or it represents a different species. Glaessner (1928) reported the species from the Helvetian (Early Miocene) of Jahring (today's Jarenina, Slovenia) in the Styrian Basin. He, however, did not provide figures. The re-examination of the material by the senior author confirms the identification of the material (Figs 8A-B). Beurlen (1939: 143) when discussing affinities of Callianassa brevimanus Beurlen, 1939, (currently recognized as a member of Lepidophthalmus Holmes, 1904; see Hyžný & Dulai in press) briefly mentioned C. michelottii as coming from the Oligocene of the Mainz Basin and Miocene of Italy and the Vienna Basin, however, without any further reference. It is supposed here that Beurlen (1939) mentioned the occurrence from the Vienna Basin erroneously, as there is no such published report known to the authors. He might have been referring to the published occurrence from the Styrian Basin by Glaessner (1928).
Several chelae from the Middle Miocene ("Badenian") of Želiezovce (Slovak part of the Danube Basin) figured by Houša in Špinar et al. (1965: figs. X-184-185, refigured herein as 8L) as Callianassa sp. can be clearly assigned to Callianassa michelottii. Unfortunately, the repeated search for the material was unsuccessful, thus, it is herein considered lost. Philippe & Secretan (1971) reported 10 fragmentary propodi from the Burdigalian of SE France. The figured specimens (Pl. C,figs. 13,14) show the keeled fixed finger accompanied with furrows, thus, pointing to attribution of the material to Calliax. The age and location of the specimens would speak for indentification as C. michelottii. Without personal re-examination of the material, however, we are reluctant to treat the specimens as conspecific with C. michelottii.
The description and figures of "Callianassa" szobensis Müller, 1984 from the Middle Miocene (Badenian) of Hungary fit the diagnosis and variations of C. michelottii. The propodus is longer than high, the fixed finger possesses two ridges parallel to one another, and there are tubercles on the lateral surface of the propodus (Figs 8J-K). As a consequence, "C." szobensis is considered a junior subjective synonym of C. michelottii.
From the Miocene of Spain (Catalonia), Müller (1993: fig. 3A) reported and figured an isolated propodus of a callianassid shrimp. He classified it as Callianassa cf. michelottii. The specimen, however, is rather dissimilar to C. michelottii; it does not possess a double ridge on the fixed finger and has the upper margin converging proximally, which is very atypical for the latter species.
Calliax michelottii n. comb. is morphologically very close to its extant congeners. The upper margin of the major P1 merus is, however, more convex in C. michelottii n. comb. (Fritsch 1871: pl. 17, fig. 9; see also Figs. 8I), whereas it is rather straight in both extant species, C. doerjesti and C. lobata ( Fig. 3A and Fig. 3C, respectively). The development of spines on the lower margin of the major P1 merus in Calliax michelottii comb. nov. is closer to that of C. lobata. The tooth formula of the occlusal margin of the fixed finger looks different in all three species, but this may be a matter of variation and is not considered taxonomically important here.
Reports on articulated individuals are scarce and are limited to preservation in fine siliciclastics (Figs 8I, 9E, 9H).
The scarcity of major P1 carpi and meri as well as of minor P1 elements can be ascribed to rather low fossilization potential of these elements as a consequence of their weak calcification. Another factor influencing our knowledge in a negative way may be collecting and reporting bias, as smaller pieces often are more difficult to interpret and may be neglected by collectors donating the material to scholars.
Interestingly, Calliax michelottii comb. nov. has been reported associated with Ophiomorpha Lundgren, 1891, burrows (Polkowsky 2005). Herein, another occurrence is reported, specifically from Košiše outcrop, where isolated propodi have been found in direct proximity of burrow structures. Preservation of the ghost shrimp body fossils in situ within their burrows or in direct association with them are often considered scarce (Bishop & Williams 2005;Hyžný 2011); thus, every report in this respect is of note.

Palaeoenvironmental significance
Today Calliax seems to inhabit a wide range of habitats. Calliax lobata is known from very shallow water environments up to 21 m (Ngoc-Ho 2003), whereas Calliax sp. reported by Taviani et al. (2013) has been found at ca. 800 m depth. Unfortunately, the bathymetry is not known for Calliax doerjesti (Sakai 1999(Sakai , 2011. Taviani et al. (2013) reported great densities of Calliax sp. in the pockmark mud of the strait of Sicily around a cold seepage, where an assemblage of chemosymbiotic organisms has been recovered. In the environment, carbonate concretions were formed (Taviani et al. 2013). Such sedimentological conditions can be compared to the situation in the Middle Miocene locality at Rohožník, Slovakia, where carbonate concretions contain mollusc shells together with remains of fish and decapod crustaceans, including Calliax michelottii comb. nov. No closer examination of the concretions or faunal elements has been conducted so far, thus the analogy is pure speculation at this moment. The preliminary results based on foraminifers and molluscs, however, point to deeper marine conditions with intervals of shallow littoral sea influx (Fuksi et al. 2011).
Considering Calliax michelottii comb. nov. an indicator of deep water conditions may be premature; so far, however, its remains are almost unequivocally connected with facies deposited in such conditions.

Notes on palaeobiogeography
The oldest fossil record of Calliax as recognized herein is that from the Late Paleocene of Pakistan (Charbonnier et al. 2013); thus, it seems the genus has a Tethyan origin. Unfortunately, so far no Eocene occurrences are known; therefore, it is rather difficult to interpret the dispersal pattern of the genus. Only a few remarks can be made.
Today, Calliax has been reported only from the Mediterranean and Atlantic Ocean (Sakai 2011). The fossil record shows that the genus was present in the Proto-Mediterranean and Paratethys seas during Oligocene and Miocene times. Thus, Calliax has occupied the Western Tethys at least since the Paleocene (Fig. 11). It remains to be answered when the dispersal across the Atlantic Ocean actually took place, as the genus is known from the Miocene of Argentina; at least the minor chela reported by Feldmann et al. (2011) suggests it. The confirmed record from the Late Miocene of Chile (Feldmann et al. 2005), i.e. Eastern Pacific, may support the previous occurrence. The pattern of supposed oceanic circulation during the Cenozoic as presented by Feldmann & Schweitzer (2006: fig. 3), however, suggests that Calliax dispersed into the Eastern Pacific via the Proto-Caribbean area and later reached the southernmost part of South American continent. More fossil occurrences are needed to further speculate on this issue. A westward migratory route across the Atlantic Ocean during the Oligo-Miocene time has been suggested for several marine groups, including ghost shrimps (Harzhauser et al. 2002Hyžný & Müller 2012).

1.
Based on thorough comparison between extant and fossil ghost shrimps, a set of characters present on P1 are considered of taxonomic importance at the genus level. The combination of rectangular major P1 propodus with short fixed finger exhibiting two more-or-less parallel ridges extending onto the manus and a minor P1 with the fixed finger distinctly shorter than the dactylus and with a wide gap between the fingers is unique among ghost shrimps and characterizes Calliax. Even the morphology of the major P1 propodus alone is distinct enough to be safely assigned to Calliax, yet the fossil specimens have been rarely interpreted as members of this genus.
2. Callianassa michelottii originally described from the Miocene of NW Italy (A. Milne Edwards 1860) and later widely reported from the Oligocene and Miocene of Europe (Fig. 1, Table 2) is re-assigned to Calliax. Callianassa szobensis from the Middle Miocene of Hungary (Müller 1984) is considered a junior subjective synonym of C. michelottii. New occurrences of the species are reported from the Neogene basins of the former Central Paratethys Sea, specifically from the North Alpine Foreland Basin and the Vienna Basin.

3.
The presence of Calliax michelotti comb. nov. may be considered an indicator of deeper marine settings. This hypothesis, however, must be further tested.

4.
The fossil record of Calliax is revised and the presence of the genus in the Miocene strata of South America is documented. Compared to today's occurrences, the geographic distribution of Calliax was wider in the geologic past. Based on the fossil record known so far, a Tethyan origin for the genus is postulated.    Europe PMC Funders Author Manuscripts  Calliax michelottii (A. Milne Edwards, 1860) comb. nov., occurrences published previously. A-G: Oligocene of the Mainz Basin, Flörsheim and Offenbach at Frankfurt (Germany). A-C, right major P1 propodus articulated with dactylus from inner, outer and upper view, respectively; D, cross section of major P1 propodus; E, right major P1 merus articulated with carpus; F, right major P1 propodus articulated with carpus in inner view; G, left major P1 propodus associated with right minor P1 chela (propodus articulated with dactylus). H: Miocene of Sardinia (Italy), the assignment to C. michelottii comb. nov. is questionable. I-L: Oligocene of "Sternberger Gestein", Kobrow (Germany), left major P1 propodus in inner (I), outer (J), lower (K) and upper view (L). A-G are refigured from Fritsch (1871), H is refigured from Lőrenthey (1907), I-L are refigured from Noetling (1886). A-H are to scale, scale bar equals 5 mm.  (Germany), specimen in I is from Pinnow bei Schwerin (Germany). All specimens are to scale and were covered with ammonium chloride prior to photography. Scale bar equals 5 mm.   Biogeographical evolution of Calliax. Chronostratigraphical dates are adopted from Gradstein et al. (2004). C. sp.* =sensu Taviani et al. (2013)