Pyreneplax basaensis new genus , new species ( Decapoda , Brachyura , Vultocinidae ) from the Priabonian ( Late Eocene ) of the Pyrenees of Huesca ( Aragón , Spain ) , and remarks on the genus Lobonotus A . Milne-Edwards , 1863

New fossil brachyuran finds in the Margas de Arguís Formation (Priabonian, Late Eocene) in the Central Pyrenees of Aragon (Spain), allow the description of a new genus and a new species (Pyreneplax basaensis), and the revision of some taxa previously assigned to the genus Lobonotus A. Milne-Edwards, 1863 (Carpilioidea: Tumidocarcinidae). According to the close similarity between Pyreneplax n. gen. and Lobonotus sandersi (Blow and Manning, 1997), L. granosus (Beschin, Busulini, De Angeli and Tessier, 2002) and L. sommarugai Beschin, Busulini and Tessier, 2009, these species are here assigned to the new genus. In addition, an important set of dorsal and ventral characters shared with the extant species Vultocinus anfractus Ng and Manuel-Santos, 2007, suggests the inclusion of the new genus in the family Vultocinidae Ng and Manuel-Santos, 2007 (Goneplacoidea). Remarks on the genus Lobonotus are also discussed herein.


INTRODUCTION
The recent discovery of fossil decapods in the uppermost levels the Margas de Arguís Fm. (Priabonian) enhances the knowledge of the Iberian Priabonian crustacean faunas.Usually scarce, few Priabonian taxa have been reported or described from the Iberian Peninsula (see: Artal et al., 2013;Ossó & Domínguez, 2013), in contrast with the fossil crustacean richness of the Ypresian, Lutetian and Bartonian levels of the Ebro Foreland Basin and nearby basins cropping out in Aragón and Catalonia.
The co-occurrence of other brachyuran taxa in the same outcrop such as Retrocypoda almelai Vía, 1959 (Retroplumidae), Portunus catalaunicus Vía, 1959 (Portunidae), Chasmocarcinus cf.guerini (Vía, 1959) (Chasmocarcinidae) and Stenodromia calasanctii Vía, 1959 (Calappidae), all of them known from the Bartonian of the above mentioned basins, indicates the persistence of such taxa during the progressive infilling of the basins that shift to non-marine conditions by the end of the Eocene.
The good condition of preservation of specimens of Pyreneplax basaensis n. gen.and n. sp., with wellpreserved partial ventral structures such as thoracic sternum, sterno-abdominal cavity and abdominal somites, allows an almost complete description of the characters of the new taxa, and its comparison with the extant Vultocinus anfractus Ng and Manuel-Santos, 2007.With regards to the genus Lobonotus A. Milne-Edwards, 1863, and its type species L. sculptus A. Milne-Edwards, 1863, the taxonomy has been a bit confused as many works frequently used L. mexicanus Rathbun, 1930 in their comparisons, instead of the type species (see discussion).Comparisons between the figures of the holotype of L. sculptus and the illustrations of several samples of Archaeopilumnus caelatus (a junior synonym of L. sculptus, according to Rathbun, 1920: 384) in Rathbun (1919) demonstrates that L. sculptus and L. mexicanus are not congeneric.Consequently, we remove L. mexicanus from Lobonotus and revise the composition of this genus.
Locality.The area object of this study is located in the central part of the southern Pyrenees, around the small village of Yebra de Basa, northwest of the province of Huesca (Aragón, N Spain).The locality is situated over the northern bank of the Basa River, a small contributor in the affluents network of the Ebro river basin (Fig. 1).
Geological setting.The area is located in the Basa anticline, in the so-called Jaca Basin.This basin, having an E-W orientation, developed in time along the continued S-N compression of the Pyrenees.The basin infilling started around Lutetian times, and the general depositional sequence, named Secuencia de Jaca (Remacha & Picart, 1991) comprises several different formations (see also Puigdefábregas, 1975).
The layers bearing the decapod crustaceans object of this study correspond to the Margas de Arguís Formation (Puigdefábregas, 1975), which in this area basically consist of a sequence about 700 meters thick of blue-grey marls interbedded with sandstone intervals.The base is defined by large scale sandstone deposits of offshore shelf facies, marked by several episodes of shallowingdeepening culminating in a relatively long transgressive episode that filled the basin.The general marine condition of the Margas de Arguís Fm. corresponds in fact to a tectonic transgressive-regressive cycle (Toledo, 1992).It is also worth noting that these layers began to be affected by deltaic contributions, with important terrigenous inputs, that become totally dominant in the upper levels.The distribution of fauna in the stratigraphic column with bryozoans, nummulites and bivalves in the lower levels, and corals, algae and gastropods in the upper ones, indicates a decrease of the bathymetry.These faunal associations suggest deep sediments corresponding to the aphotic zone of a platform that gradually became shallower.All the material studied here was recovered from the uppermost levels.
The sedimentological and biostratigraphic aspects of the Margas de Arguís Formation have been studied by several authors, including its planktonic foraminifers (Canudo & Molina, 1988;Canudo et al., 1988), and magnetostratigraphy (Hogan, 1991).According to these authors, the material of the Margas de Arguís Formation corresponds to a Bartonian-Middle Priabonian age.This formation displays in its lower levels marly deposits corresponding to a distal offshore, grading to mouth bars sandstones with abundant vegetal remains in the upper levels.Leaves, stems and mangrove palm fruits as Nypa (Arecaceae) and other indeterminate ones, suggest the area was under such tidal and/or storm influences (Oms & Remacha, 1992).et al., 1991).Fig. 2. Columna estratigràfica de la zona de l'aflorament (modificada de Canudo et al., 1991).
As indicated in the stratigraphic column (Fig. 2), the crab remains were recovered from the uppermost levels of the section.The sediments where such decapods occur are included in the depositional sequence SD6 of Canudo et al. (1991).This sequence, according to the confirmed planktonic foraminifera corresponds to a Priabonian age.The upper part of the formation consists of marly layers locally interbedded with thin levels of sandstones.Both marls and sandstones locally present pockets with high accumulations of fauna, represented by corals, bryozoans, gastropods, bivalves, nummulites and vegetal remains or, conversely, levels with a low accumulation of remains.In the latter, almost all the fossils are well preserved, with the brachyurans Serrablopluma diminuta Artal, Van Bakel, Fraaije and Jagt, 2013 (Retroplumidae) and Chasmocarcinus cf.guerini (Vía, 1959) (Chasmocarcinidae) predominant.Most of the brachyurans are well-articulated, suggesting a low energy environment and scarce transportation.In contrast with that, in high accumulation layers most of the occurring fossils are fragmented and brachyurans consist of disarticulated carapaces or only small remains.The new taxon has been always recovered from such latter layers, interpreted as corresponding to a high energy environment involving transportation.
Etymology.From Pyrene, related to the origin of Pyrenees in Greek mythology, and joining the suffix -plax (meaning "flat" in Greek), usual within Goneplacoidea.Gender feminine.
Discussion.Pyreneplax basaensis n. gen., n. sp.closely resembles some coeval taxa currently assigned to the genus Lobonotus: L. sandersi from the Eocene of South Carolina and L. granosus from the Lutetian (middle Eocene) of Northern Italy, both previously assigned to the genus Eohalimede Blow and Manning, 1996, and L. sommarugai from the Late Eocene of Northern Italy.On the basis of similarities described below, we propose the transfer of these four species into Pyreneplax n. gen.(see Blow & Manning, 1997;Beschin et al., 2002Beschin et al., , 2009)).
A set of dorsal features, such as possessing a suboctagonal carapace which is slightly wider than long, a uniform pattern of regions with elevated lobes and covered by granules, regions which are well delimited by large smooth grooves, the anterolateral margins armed with four rounded and granulated teeth or lobes, the slightly convex posterolateral margin ornamented with granules, the posterior margin convex with a median concave depression, characteristic inflated and elongated intestinal lobes separated by a narrow depression, and an identical fronto-orbital structure, are shared by the species here included in Pyreneplax n. gen.
Pyreneplax n. gen., differs from Lobonotus (type species L. sculptus A. Milne-Edwards, 1863 = Archaeopilumnus caelatus Rathbun, 1919) sensu stricto in: 1) possessing a different carapace outline which is distinctly longer than broad compared to Lobonotus; 2) having relatively smaller orbits, and a greater fronto-orbital ratio as in Lobonotus; 3) with medially depressed protogastric lobes (distinctly inflated in Lobonotus); 4) the absence of the characteristic small urogastric lobe present in Lobonotus; 5) the absence of a continuous intestinal ridge flanking the posterior carapace margin as in Lobonotus; 6) the anterolateral margin has granulated lobes instead of spines in Lobonotus;7) the posterior carapace margin is proportionately shorter and medially depressed compared to Lobonotus which is more convex and continuous; and 8) the more prominent grooves delimiting the regions which are not so extended as in Lobonotus.
Regarding the ventral features, the differences are even more evident.Pyreneplax has a proportionately narrower, more flattened and granulated thoracic sternum instead of the wider, concave and smooth one of Lobonotus; thoracic sternite 3 has a shallow median longitudinal groove that connects with the sterno-abdominal cavity, reaching the end of sternite 4 in Pyreneplax, whereas sternite 3 is entire, with the sterno-abdominal cavity endings at sternite 4 in Lobonotus; sternites 3 and 4 in Pyreneplax n. gen.are divided by a complete groove while in Lobonotus, this groove is absent and the suture is only visible laterally; the telson in Pyreneplax has a rounded tip instead of the more acute one of Lobonotus (cf. A. Milne-Edwards, 1863, pl. 10, figs. 1, 1a, 1b, and 4, 4a, 4b;Rathbun, 1919, pl. 6, figs. 6, 7, pl. 7, figs. 10-13, and pl. 8, figs. 4 to 7; Natural History Museum London web, Picture library: http://piclib.nhm.ac.uk/).
The partially well preserved ventral structures of some samples of Pyreneplax n. gen., allow comparisons with members of fossil and extant brachyuran families, and as a result, we are relatively confident in its familial assignment.
On the basis of the dorsal and ventral features, the extant family Vultocinidae Ng and Manuel-Santos, 2007 (Goneplacoidea), represented by the sole species Vultocinus anfractus Ng and Manuel-Santos, 2007, appears to be clearly related to Pyreneplax n. gen.
Both genera share a set of characters and ratios, including: similar carapace outline, similar fronto-orbital pattern, similar shape of swollen regions covered by granules and divided by large and smooth grooves, bilobed frontal margin with granulated edges, supraorbital margin with two fissures, four anterolateral granulated teeth or lobes, posterolateral margin ornamented with granules, posterior margin slightly convex and medially depressed, and the characteristic transversally elongate and inflate intestinal lobes (Fig. 3.1, 2; Ng & Manuel-Santos, 2007, pl. 50, figs. 1, 2, 3a).
Nevertheless, it would be important to compare the unusual narrow abdomen of Vultocinus females, with that of Pyreneplax females, once they are found to further cement this suggested relationship (see Ng & Richer de Forges, 2009, fig.1).
The two genera can essentially be differentiated by the shape of some dorsal regions, more sculptured in Vultocinus, and the size and depth of carapace grooves, larger and deeper in Vultocinus; the median depression in the posterior margin which is more evident in Pyreneplax n. gen.than in Vultocinus, and the thoracic sternum is slightly broader in the new genus.
Therefore, in view of the above discussed similarities in the carapace sculpture and ornamentation and despite the age difference, a close relationship between the two genera seems to be clear.In our opinion, Pyreneplax n. gen.should be classified in the family Vultocinidae.
Pyreneplax is also compared here with the Eocene genus Martinocarcinus Böhm, 1922 (type species, Martinocarcinus ickeae Böhm, 1922) the only genus and species of the family Martinocarcinidae Schweitzer, Feldmann and Bonadio, 2009 (Goneplacoidea).Pyreneplax n. gen. shares with Martinocarcinus a very similar ventral pattern as described before: similar flattened and granulated thoracic sternum with sternites 3 and 4 divided by an axial shallow groove, sternites 1-2 fused, suture 2/3 complete, suture 3/4 with a shallow transverse groove, and sternite 4 with prominent lateral edges.They only differ in the depth of the sternoabdominal cavity and the width of grooves in sternites 3 and 4, being apparently deeper and wider in Martinocarcinus, and the thoracic sternum also appears to be relatively broader in the latter genus.However significant differences are evident in dorsal view: Pyreneplax has the regions strongly swollen while, in Martinocarcinus, the regions even if discernible and with ornamentation covering the same areas, are practically flat.Other substantial differences are: a more ovate carapace outline and an anterolateral margin armed with five spines in Martinocarcinus versus a suboctagonal outline and an anterolateral margin with only four granulated lobes in Pyreneplax n. gen.; and a much broader frontoorbital ratio in the new genus than in Martinocarcinus.These differences are sufficient to exclude the inclusion of the new genus in Martinocarcinidae (see Schweitzer et  al., 2009).
Pyreneplax basaensis is also compared to several other Eocene brachyuran taxa with similar carapace features.
Lobonotus natchitochensis Stenzel, 1935 is close to P. basaensis n. gen., n. sp., especially because of its very similar outline and regions pattern; however it differs in its ornamentation, formed by coarse pustules, and by the lack of a deep groove between the mesogastric lobe and the urogastric lobe.
The genus Pakicarcinus Schweitzer, Feldmann and Gingerich, 2004 (Xanthoidea: Panopeidae) has superficial dorsal similarities with Pyreneplax n. gen., but differs by its more convex carapace as well as by the large and inflated protogastric lobes instead of the medially subdivided ones of the latter.While Pyreneplax displays four lobes in the anterolateral margin, Pakicarcinus has four spines and does not show the characteristic granulation covering the top of the dorsal regions in the new genus.Moreover, the much broader thoracic sternum and the clear suture 3/4 of Pakicarcinus, distinct in Pyreneplax, discard a possible relationship between both genera (see Collins & Morris, 1978;Schweitzer et al., 2004).
Some species ascribed to Titanocarcinus A. Milne-Edwards, 1864 (Carpilioidea: Tumidocarcinidae) exhibit several dorsal and ventral features similar to Pyreneplax: well-marked and granulated dorsal regions, dorsal surface moderately convex longitudinally, frontal margin bilobed, supra-orbital margin with two notches, developed lateral cardiac lobes, thoracic sternum with sternites 1 and 2 fused, sternites 2 and 3 divided by a clear suture, and sternites 3 and 4 divided by a clear groove (sensu Schweitzer et al., 2007b).It is noteworthy that only one species of this genus, T. decor Schweitzer, Artal, Van Bakel, Jagt and Karasawa, 2007, has preserved thoracic and abdominal structures and thus, the comparison based on ventral characters is limited to this taxon.
However, there are clear differences between the two genera.The carapace outline of Titanocarcinus is subtrapezoidal whereas it is suboctagonal in Pyreneplax; its anterolateral margins are proportionately longer, the posterolateral margins are more convergent and the posterior carapace margin relatively shorter than the new genus; the maximum width is placed more posteriorly than in Pyreneplax; Titanocarcinus also lacks the well extended intestinal lobes that Pyreneplax has.Ventral differences are also apparent.In particular, the strong transversally vaulted shape of thoracic sternites 3 and 4 (see Schweitzer et al., 2007b, figs. 3b, 3c, 4) of Titanocarcinus, contrasts with the flattened ones that the new genus has (as usual in Goneplacoidea, see Ng et al., 2007).In addition, the well-marked axial groove in the thoracic sternite 3 of Pyreneplax is barely distinguishable in Titanocarcinus (see Schweitzer et al., 2007b, same figures).These differences argue against a close relationship between the two genera.Pyreneplax basaensis n. sp.Etymology: referring to the locality where the material was collected.
Description.Carapace suboctagonal, slightly wider than long, flattened, somewhat convex anteriorly, widest at third anterolateral tooth level.Dorsal regions well defined, elevated, ornamented with granules dorsally, well delimited by large and smooth grooves.Frontal margin bilobed, straight, divided by U-shaped notch, faintly advanced, edge ornamented with pointed granules.Orbits oblique, relatively broad, slightly raised, separated from frontal margin by deep V-shaped fold; supraorbital margin semi-circular, ornamented with pointed granules; two shallow notches near outer margin separates outer-  Discussion.Differences between the new species and the three ones assigned to the genus are not very evident.In the case of Pyreneplax sandersi, despite being poorly figured in Blow & Manning (1997, fig.2), a thorough examination of the high resolution pictures of holotype ChM PI 15210 and paratype ChM PI 15200, kindly provided by the Charleston Museum (South Carolina, USA), allowed us a comparison with the new species.
Pyreneplax basaensis n. sp., can be differentiated from P. sandersi, mainly by the grooves separating the regions being relatively narrower than in P. sandersi; the epigastric lobes are closer to the protogastric lobes in P. basaensis than in P. sandersi; the urogastric region is proportionately large in P. basaensis whereas is narrow and acute in P. sandersi; the supra-epibranchial lobe is slightly oblique in P. basaensis but is almost horizontal in P. sandersi, visible in particular in the paratype ChM PI 15200 (see Blow & Manning, 1997, 1998).
With regard to the two Italian species, the differences with P. basaensis are small.Pyreneplax granosa (Beschin, Busulini, De Angeli and Tessier, 2002) n. comb. (Fig. 4. 5) differs from P. basaensis in having relatively less ornamented regions with the granules coarser, the supra-epibranchial lobe is more oblique than in P. basaensis and the sub-epibranchial lobe is rounded and less extended than in P. basaensis, also the fronto-and supra-orbital margin seems to be more swollen (see Beschin et al., 2002: 19-20, fig. 14, pl. 4, figs. 1, 2).
The carapace of P. sommarugai (Beschin, Busulini and Tessier, 2009) n. comb.appears to be more lengthened than in P. basaensis, the grooves delimiting regions are narrower and the ornamentation of the regions is more uniform with granules completely covering all the lobes (see Beschin et al, 2009: 12, 14-15, pl. 3, figs. 1, 2).
Lobonotus sculptus and L. mexicanus are clearly distinguishable between them.Examination of the good figures of dorsal and ventral features in Rathbun (1919) as Archaeopilumnus caelatus or even the pictures of the holotype, accessible online at the web of the Natural History Museum of London, regarding L. sculptus, and the figures of the holotype that illustrate Rathbun's (1930) work, regarding L. mexicanus, allow to make a comparison.
Lobonotus mexicanus differs from the type species because it is clearly lengthened, grooves separating regions are wider than in L. sculptus which has more tumid and lobulated regions; it lacks the well-marked urogastric lobe and the continuous intestinal inflation, parallel to the posterior margin; hepatic and epibranchial lobes are not separated from the anterolateral margin as in L. sculptus; anterolateral margin is toothed whereas it is spiny in L. sculptus; posterior margin shorter that in L. sculptus; thoracic sternum narrower, sternites 3-4, vaulted, large groove parallel to lateral margin of sternite 4, whereas in L. sculptus, sternites 3 and 4 appears flattened and the groove parallel to lateral margin of sternite 4 appears narrow.These differences show clearly that the two species are not congeneric.Some of the above mentioned differences were pointed out by Stenzel (1935: 384).
Several subsequent authors studying the genus Lobonotus, used in part L. mexicanus for comparisons instead of the type species L. sculptus.For instance, Schweitzer et al. (2004: 106) compared the thoracic sternal sutures of Pakicarcinus orientalis (Collins and Morris, 1978) (= Lobonotus orientalis Collins and Morris, 1978) with L. mexicanus.Schweitzer et al. (2007a: 8;2007b: 279), emended the diagnosis of the genus and discussed on Lobonotus based on ventral features of L. mexicanus.De Angeli & Cecchi (2012: 39) also compared the ventral features of L. beschini with L. mexicanus in their systematic discussions.
Despite the literature discussing on Lobonotus (Collins & Morris, 1978;Schweitzer et al., 2002Schweitzer et al., , 2004Schweitzer et al., , 2007aSchweitzer et al., , 2007b; and the particularly complete compilation in Beschin et al., 2009), we are convinced that only Lobonotus sculptus, the type species, should be considered as Lobonotus sensu stricto.In any case, even if the revision of the species assigned to Lobonotus is beyond the aim of this paper, we recommend the need for a complete and thorough revision of the generic and familial placement of all the taxa associated to this genus.
Moreover, the familial placement of Lobonotus must be revised as well.We remove herein the genus Lobonotus from the Tumidocarcinidae Schweitzer, 2005(Carpilioidea Ortmann, 1893) and according with its general shape, carapace transversely ovate, convex, spiny anterolateral margin, and free male somites, we retain the genus, with reservations, in Pilumnidae (Pilumnoidea) as already suggested by Rathbun (1920: 384) (see also Karasawa & Schweitzer, 2006).

CONCLUSIONS
Thanks to the discovery of Pyreneplax basaensis n. gen., n. sp., it has been possible to erect a new genus and, based in its dorsal features, to relate it to three species previously referred to Lobonotus, distributed along the Late Eocene across the western Tethys, from Italy and the Iberian Peninsula to the eastern coast of North America in the western Atlantic Ocean.Also, on the basis of the ventral features preserved in some samples of Pyreneplax basaensis, the genus is clearly related to the extant genus Vultocinus (Vultocinidae, Goneplacoidea) which occurs in the western Pacific Ocean.
As a result of the comparisons between Pyreneplax and the species formerly attributed to Lobonotus, we conclude that only the type species of the latter (L.sculptus A. Milne-Edwards, 1863) belongs to this genus and suggest a revision of all the species formerly attributed to it considering, as far as possible, the ventral characters.Furthermore, both the familial and suprafamilial assignment of these species should be revised.Currently placed into family Tumidocarcinidae, Lobonotus typically does not possess the general dorsal shape of this family members nor that of the Carpilioidea, being their ventral affinity based on features present in many brachyuran taxa.The same is true for the genus Titanocarcinus, also placed in Tumidocarcinidae.Even if the diagnosis for this family is expanded substantially (see Schweitzer et al., 2007b: 279 andSchweitzer &Feldmann, 2012: 24), we still find difficult to keep both Lobonotus and Titanocarcinus in the Tumidocarcinidae.A revision is clearly necessary.
Several Eocene genera present a similar dorsal aspect resembling to Pyreneplax n. gen., however, no close relationships can be concluded between them as explained above.Dorsal features and, in particular, the characters present in the two anterior thirds of the carapace shown by the new genus, are relatively common in numerous fossil and extant brachyuran genera, representing different families of the Xanthoidea or Pilumnoidea.In the case of fossil taxa, many of them are only known by more or less well preserved fragments of their carapace, but lacking remains of thoracic and abdominal parts.Since such parts are important for a correct systematic diagnosis, their precise familial placement becomes very difficult to deal with (see Guinot, 1979).In their absence, no definitive statements can be made.The persistence of the dorsal carapace pattern through time, would suggest either an evolutionary success or an example of convergence.This convergence is well known (see Ng et al., 2008;Lai et al., 2011).It is clear that new discoveries of fossil taxa with well-preserved ventral parts would help to better interpret the fossil record of these groups.