Morphologic analyses of a large collection of coleoid cephalopods from the Lebanese Upper Cretaceous yielded a much higher diversity than previously assumed and revealed numerous extraordinarily well-preserved, soft-part characters. An analysis of the Prototeuthidina, a gladius-bearing group with a slender torpedo-shaped body, revealed two species: Dorateuthis syriaca and Boreopeltis smithi n. sp. Previously unknown soft-part characters, such as the digestive tract, the gills, and the cephalic cartilage considerably improved our knowledge of D. syriaca. Since none of the investigated specimens show more than eight arms, similarities with modern squids are regarded as superficial. Boreopeltis smithi n. sp. is erected on the basis of its comparatively wide Paraplesioteuthis-like gladius. The latter species represents the first unambiguous record of this genus in Upper Cretaceous deposits. Phylogenetic analyses indicate that the prototeuthidid clade consists of two lineages. The plesioteuthidid lineage originates from early Jurassic Paraplesioteuthis and leads to Plesioteuthis and Dorateuthis. The other lineage is morphologically more conservative and leads to Boreopeltis.
THE PROTOTEUTHIDINA Naef, 1921 is an extinct group of coleoid cephalopods that includes taxa such as ParaplesioteuthisNaef, 1921, PlesioteuthisWagner, 1859 and DorateuthisWoodward, 1883 (Naef, 1922; Jeletzky, 1966; Fuchs et al., 2007b). Prototeuthidids are currently known from the Early Jurassic to the latest Cretaceous. Owing to their torpedo-shaped body and their slender, arrow-shaped gladius, prototeuthidids have long been considered the “root stock” of teuthids. However, since Bandel and Leich (1986) observed more similarities with vampyropod coleoids, focus has been placed on the prototeuthidid's soft-part morphology. Coleoid soft-parts are well-known from the Middle Jurassic La Voulte-sur-Rhône and the Upper Jurassic of Solnhofen and several other localities (Fischer and Riou, 1982; Mehl, 1990; Engeser and Keupp, 1999; Fischer and Riou, 2002; Haas, 2002; Fuchs, 2006a, 2006b; Fuchs et al., 2007a, 2007b; Klug et al. 2009, 2010). The Plattenkalks of Hâkel, Hâdjoula (Cenomanian) and Sâhel Aalma (Santonian) represent an extraordinary window into the soft-part morphology of Cretaceous coleoids (Fuchs, 2006a).
Although the “Vampyropoda-hypothesis” has been largely supported (Engeser and Bandel, 1988; Engeser, 1988; Haas, 2002; Donovan et al., 2003; Klug et al., 2005; Fuchs, 2006a, 2006b; Fuchs et al., 2007b), the systematic position of the Prototeuthidina remains a source of considerable debate (Donovan and Toll, 1988; Young et al., 1998; Vecchione et al., 1999; Bizikov, 2008).
On the basis of eight undescribed specimens from Hâkel and Hâdjoula, Fuchs (2006a) reinvestigated the soft-part morphology of the prototeuthidid Dorateuthis syriacaWoodward, 1883. He classified Dorateuthis as a vampyropod, but could not equivocally support the “Vampyropoda-hypothesis.” The present investigation in contrast yielded not only a second prototeuthidid species in addition to Dorateuthis syriaca, but also several previously unknown soft-part characters with phylogenetic significance.
The fossil fish from Lebanon have been known for centuries; Herodotus reported on their existence as early as 450 B.C. (Davis, 1887). As a result, there have been many publications on the fauna of the Upper Cretaceous fishes and invertebrates of Lebanon, most notably Pictet (1850), Davis (1887), Hay (1903), Roger (1946), Gaudant (1978), Gayet (1980), Cappetta (1980a, 1980b), Gayet et al. (2003) and Dalla Vecchia (2004). However, none of these authors adequately addressed the cephalopods of Lebanon.
The Sannine Limestone of Hâkel and Hâdjoula is a lithographic limestone approximately 650 m in maximum thickness (see Avnimelich  for an extensive bibliography of the region's geology). These limestones were deposited in small basins (a few 100 m across) in a warm, shallow southern Tethys ∼400 km away from the North African coastline (Hückel, 1970; Philip et al., 1993). At Hâkel and Hâdjoula, the formation has been subdivided into eight lithologic units, numbered from the top down (Hückel, 1970). The cephalopods are associated with the marine fish fauna from Units 4 and 5 from Hâkel and Hâdjoula, respectively (Forey et al., 2003). The fossil fish sites at the former are younger and 20 m stratigraphically higher than the former (Hückel, 1970).
The assigned age of the deposits of Hâkel and Hâdjoula differ depending on which author and which biostratigraphic indicators employed. Hückel (1970) concluded, based upon planktic foraminifera, that the fish fauna was late Early Cenomanian. A more recent study on the foraminifera by Forey et al. (2003) concluded the fossil fish deposits of Hâkel and Hâdjoula were late Middle Cenomanian. The co-occurrence of the ammonite AllocriocerasSpath, 1926 documented by Wippich and Lehmann (2004) places the age of the deposits at Hâkel within the Late Cenomanian; and the presence of the ammonite MetoicocerasHyatt, 1903 dates those same deposits somewhere between the late Middle Cenomanian to the middle Late Cenomanian. Because both ammonites are among the best index fossils, the quarries at Hâkel are determined to be from the lower-upper Cenomanian, or 95–96 Ma (Janke, 2008). Because Hâdjoula is slightly older than Hâkel, faunas from the two areas differ slightly as expected.
The deposits of Sâhel Aalma are substantially younger and dated as Late Santonian (85 Ma as determined by Gayet et al., 2003). Ejel and Dubertret (1966) used Foraminifera to constrain the age of the deposits and also came up with a Late Santonian age between 83 to 85 Ma. Meister (1993) reported the presence of the ammonites Muniericeras blanfordianum (Stoliczka, 1865) and Texanites texanus (Roemer, 1852) which also places them in the Santonian. The Santonian in the Western Interior of North America has been dated to 84–87 Ma based on radiometric dates taken from bentonite samples (Cobban et al., 2006).
material and methods
Since the “coleoid fossil market” is currently dominated by specimens from Hâkel and Hâdjoula, a huge number of Lebanon coleoids has been accumulated in private as well as in public collections. The largest collection is currently housed in the Black Hills Institute (Hill City, South Dakota, USA). It includes a total number of 236 specimens from various coleoid suborders. We have subdivided the work into four systematic portions; here we consider the Prototeuthidina. Part II through IV will describe the Teudopseina Starobogatov, 1983, Belemnoidea Hyatt, 1884 and Octobrachia Fioroni, 1981, respectively.
A total number of 73 prototeuthidid specimens from various collections have been studied. The material included the holotype of Dorateuthis syriacaWoodward, 1883 as well as originals of Fraas (1878), Kolbe (1888), Naef (1922), Roger (1946), Engeser and Reitner (1986), Lukeneder and Harzhauser (2004) and Fuchs (2006a). The majority of the specimens originated from the Cenomanian localities of Hâkel or Hâdjoula; only a few specimens come from the Santonian of Sâhel Aalma.
All of the cephalopods in the BHI collections are from the Hâkel and Hâdjoula quarries and were mined by numerous collectors, most notably the Abi Saad family of Bibylos (Lebanon) and Flavio Bachia and his employees of Milan (Italy). Most of the specimens are represented by only one half; the other half was either destroyed during collecting or sold to someone else.
Some paleontologists have utilized UV fluorescence photography to aid in research (e.g., Morris, 1978; Lund, 1980; Fuchs, 2006a, 2006b; and Larson et al., 2010). UV photography has produced excellent results in highlighting the phosphatic tissues in squids from Lebanon (Larson et al., 2010). As a result, UV fluorescence has aided in the discovery of cirri, internal organs, eye capsules, nerves and blood vessels (Larson et al., 2010). All UV photography for this paper was undertaken using Nikon SLR digital cameras under UV LCD, UV fluorescence, and UV mercury.
The gladius terminology used in the text follows that of Fuchs (2006a, 2006b), Fuchs et al. (2007b) and Fuchs and Weis (2008, 2010). The general morphology of the prototeuthidid gladius and its terminology is given in Figure 1.
BHI, Black Hills Institute of Geological Research Hill City; BMNH, British Museum of Natural History London; MNHN, Musée National D'Histoire Naturelle Paris; MNHNL, Musée National D'Histoire Naturelle Luxembourg; MSNM, Museo Civico di Storia naturale di Milano.
Subclass ColeoideaBather, 1888
Superorder VampyropodaBoletzky, 1992
Suborder PrototeuthidinaNaef, 1921
Original diagnosis of Naef (1921–1923, p. 132)
“Fossil Teuthoidea (Liassic to Cretaceous) in which the median plate of the strongly calcified gladius is usually sharply delimited by asymptotes, and the gladius very blunt anteriorly and at least half as wide as the mantle sac; lower side of middle plate without a median keel which is gutter-shaped ventrally (Figure 62), but instead at most with solid supporting ridges which become reduced or disappear anteriorly.”
Emended diagnosis of Fuchs et al. (2007b, p. 241)
“Comparatively slender gladius with triangular median field and ventrally closed (funnel-like) conus. Hyperbolar zones reduced or absent. Lateral field length less than the half gladius length. Median and lateral reinforcements present on the median field. Vestiges of septa and guard, as reported by Jeletzky (1966, p. 43–44), have never been observed.”
Naef (1921–1923) originally placed the four families Plesioteuthididae Naef, 1921, Leptotheuthididae Naef, 1921, Geopeltidae Regteren Altena, 1949 (his Geoteuthidae), and Loligosepiidae Regteren Altena, 1949 (his Belopeltidae) in the Prototeuthidina. Jeletzky (1965, 1966) excluded the latter two families and put them into the new suborder Loligosepiina. The Leptotheuthididae were later classified as loligosepiids by Engeser (1988) and Doyle et al. (1994). Recently, Fuchs (2006b, 2006c), Fuchs et al. (2007b) and Fuchs and Weis (2008) adapted this classification. As a result, the Prototeuthidina include only one family, the Plesioteuthididae. Owing to this substantial reorganization within the Prototeuthidina, the original diagnosis proposed by Naef (1921–1923) and emended by Jeletzky (1966, p. 43) is no longer accurate. To clearly distinguish the Prototeuthidina from the Loligosepiina, Fuchs et al. (2007b) emended the diagnosis of the Prototeuthidina.
Family PlesioteuthididaeNaef, 1921
Plesioteuthis Wagner, 1859
Original diagnosis of Naef (1922, p. 111)
“More or less slender prototeuthoids with a conus vane that is restricted to the posterior end and is bent down to form a pointed cone, grading anteriorly into the relatively broad, gradually tapering lateral plates, in which the median plate bears a simple or double median rib and narrow lateral bands clearly demarcated from the broad central part.—The fins are very short, subterminal, resting on the conus.”
PlesioteuthisWagner, 1859; DorateuthisWoodward, 1883; Paraplesioteuthis, Naef, 1921; RomaniteuthisFischer and Riou, 1982; RhomboteuthisFischer and Riou, 1982; BoreopeltisEngeser and Reitner, 1985; SenefelderiteuthisEngeser and Keupp, 1999; NesisoteuthisDoguzhaeva, 2005.
(?)Late Triassic (Rhaetian), Early Jurassic (Toarcian)–Late Cretaceous (Maastrichtian) of Europe, Central Russia, Lebanon, North America and Australia.
The generic composition of the Plesioteuthididae noticeably changed during the past (Table 1). Naef (1921–1923, p. 47; 1922, p. 111–119) originally included Paraplesioteuthis, Plesioteuthis, Dorateuthis and StyloteuthisFritsch, 1910 in the Plesioteuthididae. Jeletzky (1966, p. 45) generally accepted this classification with the exception that he placed Dorateuthis together with Leptotheuthis von Meyer, 1834 in the Leptotheuthididae. This was followed by a progressive increase in the number of plesioteuthidid genera (e.g., Fischer and Riou, 1982; Engeser and Reitner, 1985). According to Engeser (1988), the Plesioteuthididae includes Paraplesioteuthis, Plesioteuthis, Dorateuthis, Boreopeltis, MaioteuthisReitner and Engeser, 1982, Rhomboteuthis, and Romaniteuthis. Engeser (1988, p. 71) and Fuchs (2010, p. 68) regarded StyloteuthisFritsch, 1910 as a teudopseid genus. Doguzhavea (2005) recently described a new plesioteuthidid-like gladius and erected the new genus Nesisoteuthis. Fuchs et al. (2007b, p. 241) assigned Paraplesioteuthis, Plesioteuthis, Dorateuthis, Boreopeltis, Rhomboteuthis, Romaniteuthis and Nesisoteuthis to the Plesioteuthididae. In this publication, Maioteuthis was interpreted as a junior synonym of Dorateuthis. Moreover, Fuchs et al. (2007b, p. 246) regarded Senefelderiteuthis as a junior synonym of Dorateuthis. Fuchs (2006a, p. 7) and Fuchs et al. (2007b, p. 246) assumed that the gladius of the type species of Dorateuthis, Dorateuthis syriacaWoodward, 1883, possessed a funnel-like conus formed by lateral fields that occupied at least 20% of the total gladius length. The present investigation, however, has shown that Dorateuthis syriaca has a considerably reduced lateral field. Since “Dorateuthis” tricarinata (Münster, 1846) (senior synonym: Senefelderiteuthis krausiEngeser and Keupp, 1999) from the Tithonian of the Solnhofen region exhibits well-developed lateral fields (Engeser and Keupp, 1999, Fuchs et al., 2007b, p. 247), it cannot be included within Dorateuthis. Hence, we re-activate herewith Senefelderiteuthis and attribute “Dorateuthis” tricarinata (Münster, 1846) to this genus.
In their revision of Early Jurassic Loligosepiina, Fuchs and Weis (2008, p. 96) remarked that “Loligosepia” neidernachensisReitner, 1978 from the Rhaetian of southern Germany is unlike a typical loligosepiid and instead more similar to Paraplesioteuthis in having short lateral fields and well developed lateral reinforcements. Therefore, we suggest that this oldest known gladius belongs to a prototeuthidid.
Genus DorateuthisWoodward, 1883
Dorateuthis syriacaWoodward, 1883 from the Cenomanian-Santonian Plattenkalks of Lebanon.
With certainty only Dorateuthis syriaca (Woodward, 1883) from the Cenomanian-Santonian of Lebanon; poorly preserved “Neololigosepia” stahleckeri (Reitner and Engeser, 1982) and “Maioteuthis” morroensis (Reitner and Engeser, 1982) both from the Barremian of the Cape Verde Islands; “Plesioteuthis” sp. Engeser and Reitner, 1985 and “Maioteuthis” damesi (Engeser and Reitner, 1985) both from the Aptian of Heligoland (N Germany); “Plesioteuthis” arcuata (Von der Marck, 1873) from the Campanian of NW Germany; “Plesioteuthis” maestrichtensis (Binkhorst van den Binkhorst, 1861) from the Maastrichtian of the Netherlands, all of which display a narrow gladius with continuous lateral keels, are preliminarily assigned to this genus.
Diagnosis (modified from Fuchs et al. 2007b)
Gladius long and slender (ratio gladius width:gladius length ∼0.10) without a median keel. Median field triangular with a bipartite median ridge and pronounced lateral keels present continuously from anterior to posterior extremities. Lateral fields and conus strongly reduced.
With certainty from the Cenomanian through Santonian of Lebanon (and potentially from the Barremian up to Maastrichtian of the Cape Verde Islands, Northern Europe and Russia).
Dorateuthis syriaca Woodward, 1883
SepialitesFraas, 1878, p. 346.
Dorateuthis Syriaca n. sp. Woodward, 1883, p. 1, pl. 1.
Curculionites senonicusKolbe, 1888, p. 135, pl. 11, fig. 8.
SepialitesWoodward, 1896, p. 231.
“Plesioteuthis fraasi” Naef, 1922, p. 133, fig. 50.
Sepialites Sahil Almae. Klinghardt, 1943, p. 12, fig. 8.
Leptoteuthis syriaca (Woodward). Roger, 1946, p. 14, text-fig. 6, 7, pl. 4, fig. 5, 6, pl. 9, fig. 1, 2.
Sepialites Sahel-almae Fraas. Roger, 1952, p. 738.
Leptoteuthis syriaca Roger (sic). Roger, 1952, p. 739.
?Dorateuthis sp. Engeser and Reitner, 1986, p. 4, pl. 1, fig. 2.
Dorateuthis sahilalmae.Fuchs, 2006b, pl. 17H, pl. 22F.
Dorateuthis syriaca.Fuchs, 2006b, pl. 17I, pl. 22E.
Dorateuthis sp. Fuchs 2006b, pl. 17J.
Holotype, BMNH C5017, original of Woodward (1883, pl. 1)
Sâhel Aalma, Lebanon
early Late Cenomanian–Late Santonian of Lebanon
Although Dorateuthis syriaca is the most abundant coleoid in the Lebanese plattenkalks, diagnostic characters that unambiguously identify the taxon are still poorly known. Particularly, the most important character complex, the gladius, has been only scantly or inconsistently described. In his original description, Woodward (1883, p. 3) stated “…the shaft of which is marked by three equidistant ridges, one median and two lateral (as in Acanthoteuthis tricarinata), which converge together at the very acute distal extremity; two extremely delicate lateral expansions, or alæ (like the blade of a spear-head, are developed upon each side, and give rigidity to the two ellipse fins…..” Naef (1922, p. 118) presumed a strong median rib and doubted the presence of lateral fields. However, Roger (1946, p. 14) again observed long and wide lateral fields and, therefore, regarded the gladius to be close to the genus Leptotheuthis. Fuchs (2006a, p. 7), who first described Dorateuthis syriaca from the Cenomanian of Hâkel and Hâdjoula, discussed the inconsistent observations concerning lateral fields. He argued that if lateral fields were present, they must have been extremely short and narrow. Additionally, he remarked that there are no specific differences between Dorateuthis sahilalmae (Naef, 1922) and Dorateuthis syriaca as well as between Cenomanian and Santonian specimens.
Since the holotype is relatively poorly preserved and to overcome previous inconsistencies, an emended description based on a higher number of specimens is given below.
The slender gladius consists of an anteriorly diverging median field and very short and narrow lateral fields (Fig. 2.1–2.7). The central part of the median field must have been very thin, because it is usually not preserved. Because of this, the gladii are often longitudinally disrupted given the improper impression of a comparatively broad gladius (Engeser and Reitner, 1986, p. 3; Fuchs, 2006a, p. 7). A delicate bipartite median ridge is sometimes visible (BMNH C5017, BHI2210, IPR.6746, Fig. 2.1, 2.5). A strong median rib as Naef (1922, p. 118) observed is not present. The shape of the anterior median field margin is poorly known, but it was possibly curved backward (concave) with a secondary short anterior projection in the central part (BHI2210, BHI2205, Fig. 2.5, 2.6).
The flanks of the median field are strongly thickened by at least two shell layers. This thickening occurs on the dorsal as well as the ventral gladius surface. The dorsal layer is smooth, whereas the ventral shell layer exhibits forward curved growth increments (Fig. 2.6). The lateral keels diverge at an acute angle of 6–10° and extend from posterior to anterior extremities. Their width hardly increases from posterior to anterior. In contrast to the dorsal keels, the ventral thickenings are less pronounced and bear a median furrow (Fig. 2.3, 2.4). Only a single juvenile gladius (BHI5618) yielded some weak imprints of lateral fields (Fig. 2.7). According to this specimen, the lateral fields are very short (∼10% of the total gladius length) and narrow (∼6% of the total gladius length). It is still unclear whether the lateral fields formed a conus or not. However, none of the studied specimens yielded long and wide lateral fields as Woodward (1883) and Roger (1946) described. Hyperbolar zones are unknown.
Both the holotype (Fig. 2.1) and the original specimen of Lukeneder and Harzhauser (2004, fig. 2) display a mantle outline that appears to be extremely slender and arrow-shaped. However, this is most likely the result of imperfect preservation as the bulk of the specimens rather exhibit a bullet- or torpedo-shaped mantle outline (Fig. 2.2, 2.3).
Three specimens from Sâhel Aalma (among them the holotype and NHM 1998z0105/0000) show eight arms (Lukeneder and Harzhauser, 2004; Fuchs, 2006a). A fourth specimen showing eight arms (BMNH C2919, the type specimen of “Plesioteuthis fraasi” Naef, 1922) cannot be unambiguously classified because the arm crown is isolated without evidence of a gladius, but the general arm morphology is very typical of Dorateuthis syriaca. There are three different arm lengths discernible in each of these specimens (Fig. 3.1–3.6). The shortest and most delicate arms are ventral; the longest and thickest are in dorsal position. The ventrolateral and dorsolateral arm pairs are equal in length and thickness. A fifth strongly elongated ventrolateral arm pair as drawn by Woodward (1883) or assumed by Naef (1922) is not visible in the holotype nor in other specimens.
The longest arm pair of the juvenile holotype (gladius length: 38 mm) is comparatively thin and short (ratio arm length:gladius length ∼0.4); those of the possibly adolescent Vienna specimen (gladius length: 97 mm) appear slightly thicker and longer (ratio arm length:gladius length ∼0.45). Large specimens (gladius length: >150 mm) from the BMNH, MSNM and MNHN collections with more or less outspread arms suggest a ratio between arm length:gladius length ratio of at least 0.5. These analyses support the hypothesis proposed by Fuchs (2006a, p. 8), that suggested the differences in relative arm length and thickness is correlated with ontogenetic differences.
None of the specimens studied yielded evidence of hooks, suckers or cirri. Therefore, our observations do not concur with those of Woodward (1883) who mentioned the presence of hooks in the holotype, or with Lukeneder and Harzhauser (2004, p. 4) who detected sucker-like structures in the Vienna specimen.
Arms are also preserved in Cenomanian specimens from Hâkel and Hâdjoula, but are less distinct than in specimens from Sâhel Aalma.
Specimens from Hâkel or Hâdjoula often exhibit pairs of pear-shaped imprints observable in the longitudinal axis of the body and between the anterior margin of the gladius and the buccal masses (Fig. 4.1–4.3). The maximum width of both imprints corresponds to the anterior gladius width; the maximum length slightly exceeds the width. The outer margins are distinctly curved. The inner margins are weakly curved and very close to each other, but do not appear connected. However, if present, the imprints always appear in pairs indicating a median connection. Occasionally, the esophagus is preserved as a cord of yellowish staining, passing through the inner margins (Fig. 4.4–4.6). The same yellowish staining surrounds the imprints in a reticulated pattern (Fig. 4.1–4.6). It is likely that this pattern indicates the eyes' capillary system (see also Riegraf, 1987, figs. 8, 9).
In three specimens (BHI2212, 2231, 5779), there are additional pairs of boomerang-shaped structures associated with the pear-shaped imprints (Fig. 4.4–4.6). Each of the specimens is in dorsal view indicating that the “boomerangs” cover the “pears.” In contrast to the pear-shaped imprints, the whitish material of “boomerangs” is illuminating under UV-light (Fig. 4.6).
We hypothesize that the “boomerangs” and the “pears” represent the dorsal and ventral parts of the cephalic cartilage, respectively. Similar to the cephalic cartilages in Sepia or Vampyroteuthis, the lateral projections form the orbital fossa (Nixon, 1998, fig. 3). A preliminary reconstruction of the present cartilage is given in Figure 4.7–4.9.
Lateral embedded specimen BHI2213 indicates that the cephalic cartilage was more or less circular in lateral view. In specimen BHI5779 there is a pair of irregular structures discernable close to the posterior side of the cephalic cartilage (Fig. 4.5, 4.6). This yellowish staining might represent the former position of the statocysts. The presumed statocysts provide no evidence of statoliths. Statoliths cannot be preserved as aragonite is usually dissolved in the Lebanon plattenkalks. No evidence of nuchal cartilage has been observed.
The preservation of the buccal mass as a white spot anterior to the head is a common feature in Dorateuthis syriaca (Fuchs, 2006a; Lukeneder and Harzhauser, 2004; Fig. 5.1). Beaks are occasionally discernible in the centre of the buccal mass, but without any morphological information owing to distortion. Posterior to the buccal mass, a distinct cord, representing the esophagus, is visible as Lukeneder and Harzhauser (2004) correctly described. In some specimens, how the esophagus passes through the cephalic cartilage can be viewed (Fig. 4.4–4.6). A salivary gland described by Lukeneder and Harzhauser (2004) cannot be confirmed. The position of the muscular stomach is often indicated by an accumulation of chopped fish remains and situated between the ink sac and the rear mantle (specimen MSNMi24802, BHI5779; Fig. 5.2). Specimen BHI2233 even preserves the cuticular walls of the stomach (Fig. 5.3). A subdivision into a coarse anterior part and a fine posterior part as Lukeneder and Harzhauser (2004) observed is sometimes visible. Lukeneder and Harzhauser (2004) interpreted the coarse anterior part to represent the stomach and the posterior part to be equivalent to the caecum. Alternatively, the coarse anterior part with chopped food might represent a crop, whereas the posterior part with semi-digested food is equivalent to the stomach. Specimen CRE050 shows a comparatively long band of undigested fish remains between the anterior mantle and the stomach. It is possible that this band might correspond to food that had been regurgitated into the esophagus, but it can also be interpreted as a tubular crop (Fig. 5.4).
In two specimens, feather-like structures are visible in the lateral mantle that may represent gills (Fig. 6.1–6.5). They differ in color from the surrounding matrix. In specimen BHI 2230, the staining is yellowish, whereas in specimen BHI 2219 it is black (most likely stained by leaked ink). In the latter specimen, the staining reflects a longitudinal axis with a dense series of at least 18, long lappets branching off from this axis. These structures might represent remains of the cartilaginous gill skeleton (Mehl, 1990, p. 78), but it clearly reflects the primary efferent gill vessel and its outer branches ( = secondary efferent vessels, branchial laminae).
Woodward (1883) described elliptical fins for the holotype. However, what he identified as fins, are clearly the outline of the imperfectly preserved mantle (Fig. 2.1). Nevertheless, on the basis of two specimens (MSNM i25128, i25134) from Hâdjoula, Fuchs (2006a, p. 7) reported a pair of lobate fins. Weak impressions on three additional specimens (BHI 2203, 2227, 2230) confirm this observation. According to the definition proposed by Bizikov (2008, fig. 147), the fins of Dorateuthis syriaca are oar-shaped and closely resemple those of Recent Vampyroteuthis or Cirroteuthis. In contrast to the latter taxa, the fins of Dorateuthis syriaca are clearly terminal. UV-light revealed no further details. Fins remain unknown in specimens from Sâhel Aalma.
The gladius of Dorateuthis syriaca mainly differs from that in the Early Jurassic Paraplesioteuthis sagittata (Münster, 1843) through a more slender gladius, the absence of true lateral fields, and the presence of prominent lateral keels (Fig. 7.1). Dorateuthis syriaca is similar to Middle Jurassic Romaniteuthis gevreyiRoman, 1928, Late Jurassic Senefelderiteuthis tricarinata (Münster, 1846), Plesioteuthis subovata (Münster, 1846) and Plesioteuthis prisca (Rüppell, 1829) in having a slender gladius that is flanked by lateral keels (Fig. 7.2–7.5). In contrast to R. gevreyi, D. syriaca bears only a weakly developed bipartite median ridge and lacks true lateral fields (Fig. 7.2). Dorateuthis syriaca differs from S. tricarinata through the presence of a bipartite median ridge and the absence of true lateral fields (Fig. 7.3). Plesioteuthis prisca and P. subovata can be distinguished from Dorateuthis syriaca through the presence of a pronounced median keel and the absence of continuous lateral keels (Fig. 7.4). Cretaceous Nesisoteuthis simbirskensisDoguzhaeva, 2005, “Neololigosepia” stahleckeri (Reitner and Engeser, 1982), “M.” morroensis (Reitner and Engeser, 1982), P. sp. (Engeser and Reitner, 1985), “Maioteuthis” damesi (Engeser & Reitner, 1985), “P.” arcuata (Von der Marck, 1873) as well as “P.” maestrichtensis (Binkhorst van den Binkhorst, 1861) seem to be similar in having lateral keels; however, a reliable differential diagnosis is problematic owing to an incomplete morphological knowledge of these taxa.
A dorsal arm pair that is longer than the other arms is also a well-known feature of P. prisca (Fuchs, 2006b). However, the arms are considerably longer in D. syriaca than in P. prisca. Although imperfectly preserved, the holotypes of R. gevreyi and Rhomboteuthis lehmani suggest a similar arm length as D. syriaca. As compared to S. tricarinata, the arms are shorter in D. syriaca (Engeser and Keupp, 1999; Fuchs et al., 2007b).
Apart from Dorateuthis syriaca, a cephalic cartilage is known only from Glyphiteuthis abisaadiorumFuchs and Weis, 2009 from Hâdjoula (Fuchs and Weis, 2009). In this trachyteuthidid, the cephalic cartilage simply consists of a pair of eye capsules similar to the living octopod EledoneLeach, 1817 (Nixon, 1998, fig. 3d). The cephalic cartilage of D. syriaca resembles VampyroteuthisChun, 1903 in having lateral projections.
A continuous digestive system consisting of buccal mass, esophagus, crop and stomach as described above is unique and therefore cannot be compared with other fossil cephalopods.
The gills of Dorateuthis syriaca resemble those of Plesioteuthis prisca (Bandel and Leich, 1986; Mehl, 1990; Haas, 2002) in having a comparatively large number of branchial laminae. Incirrate octopods Keuppia levanteFuchs et al., 2009 and K. hyperbolarisFuchs et al., 2009 both from Hâdjoula are described as having only 6–7 branchial laminae (Fuchs et al., 2009). Branchial laminae have been also reported from the belemnoid “Phragmoteuthis” conocaudaQuenstedt, 1849, but their total number is still unclear (Reitner and Mehl, 1989; Mehl, 1990; Reitner, 2009).
A pair of terminal lobate fins appears to be a common feature within the Plesioteuthididae, because similar fins are also described from Rhomboteuthis lehmani and Plesioteuthis prisca (Fischer and Riou, 1982; Fuchs, 2006a, 2006b; Fuchs et al., 2007b). Compared to the latter taxa, the fins of Dorateuthis syriaca are slightly longer.
Genus BoreopeltisEngeser and Reitner, 1985
Boreopeltis helgolandiaeEngeser and Reitner, 1985.
Boreopeltis sagittata (Naef, 1921) from the early Tithonian of southern Germany, B. helgolandiaeEngeser and Reitner, 1985 from the early Aptian of Heligoland (German North Sea), and B. smithi n. sp. from the early late Cenomanian of Lebanon.
Original diagnosis of Engeser and Reitner (1985, p. 252; translated from German)
“Moderately slender plesioteuthidids with a broad median line, wide subdivided lateral fields and wide lateral plates, which are restricted to the posterior two thirds of the total gladius length. The conus vanes are presumably small.”
Late Jurassic–early Late Cenomanian of Germany and Lebanon.
“Boreopeltis” soniaeWade, 1993 from the Albian of Australia possess a peculiar gladius probably of the prototeuthidid type, but significantly differs from Boreopeltis through a median thickening instead of lateral reinforcements. Placement of “B.” soniae within Boreopeltis is therefore regarded as doubtful.
Boreopeltis smithi new species
Gladius comparatively wide without a median keel. Median field triangular with lateral plate-like reinforcements present continuously from anterior to posterior extremities diverge at an angle of 12°. Lateral fields anteriorly narrow and posteriorly wide forming a pointed conus.
The holotype preserves a comparatively small gladius in dorsal view (Fig. 8.1–8.3). Although the anterior median field is heavily distorted, the total length of the gladius is measurable through reinforcements preserved on the lateral flanks of the median field. According to these plate-like lateral reinforcements ( = “lateral fields” of Engeser and Reitner, 1985), the gladius is 50 mm in total length and 10 mm in maximum width. The lateral reinforcements are subdivided into a narrow outer part and a wider inner part. The reinforcements are visible throughout the entire gladius length, and they diverge at an angle of ∼12°. Anteriorly, they converge again, but this is regarded to be a diagenetic artifact. The width of the reinforcements increases continuously from posterior to anterior. Their maximum width is 1 mm.
The shape of anterior median field margin is indistinct, but appears to be retracted behind the lateral reinforcements. Between the lateral reinforcements one can see mantle musculature.
Posteriorly, narrow lateral fields ( = “lateral plates” of Engeser and Reitner, 1985) form the pointed gladius end. Weak growth increments on the posterior lateral fields seem to be somewhat bended outward. This observation indicates a conus that is ventrally closed. The maximum width of the lateral fields is 8 mm. The width rapidly decreases anteriorly, but then remains constant for a comparatively long distance until they intersect the lateral reinforcements. The total length of the lateral fields is 29 mm; hence, the lateral field length:gladius length ratio is 0.58. Hyperbolar zones between the lateral fields and the median field are not evident, i.e., inner asymptotes correspond to the outer part of the lateral reinforcements. A median ridge is absent, but there is a slightly diverging central median field ( = “median line” of Engeser and Reitner, 1985) recognizable behind the ink sac.
Except for some spots of preserved mantle musculature and the ink sac, there are no other soft tissues preserved.
Dedicated to R. Smith (Muscat, Oman). He recognized the peculiarity of the specimen and kindly donated it for scientific research.
Holotype, MNHNL CRE11; deposited in the Musée National D'Histoire Naturelle, Luxembourg.
Metoicoceras geslinanum Zone (international standard; corresponds to Sciponoceras gracile Zone in Western Interior); mid-Upper Cenomanian.
Late Cenomanian-?Santonian of Lebanon.
Boreopeltis smithi n. sp., the first Late Cretaceous representative of its genus, is can only be compared with members of the Prototeuthidina. Affiliation with the Teudopseina is excluded because this group is characterized through a pointed or rounded anterior median field (Fuchs et al., 2007b; Fuchs and Weis, 2010). Representatives of the Loligosepiina are also typified by a triangular median field, but attribution to this group is unlikely since B. smithi lacks a hyperbolar zone which is usually well-developed and notch-like as in Loligosepiina (Fuchs and Weis, 2008). Lateral reinforcements are furthermore unknown in the Loligosepiina.
Boreopeltis smithi n. sp. is close to Early Jurassic Paraplesioteuthis sagittata (Münster, 1843) in having a similar apical angle, conus shape, comparatively long lateral fields, and plate-like lateral reinforcements (P. hastata (Münster, 1843) is here regarded as conspecific with P. sagittata). In contrast to B. smithi, P. sagittata is known to have a bipartite median ridge (Fig. 7.1). Lateral reinforcements are wider and lateral fields are ∼20% shorter in P. sagittata than in B. smithi.
The slender gladius with lateral reinforcements of Boreopeltis smithi is similar to those of Rhomboteuthis, Romaniteuthis, Plesioteuthis, Senefelderiteuthis, Nesisoteuthis and Dorateuthis. Middle Jurassic R. gevreyi (Fig. 7.2), Late Jurassic S. tricarinata (Fig. 7.3) and P. prisca (Fig. 7.4) mainly differ from B. smithi based on the conus (Naef, 1922; Fischer and Riou, 1982; Engeser and Keupp, 1999; Fuchs, 2006b; Fuchs et al., 2007b), which does not seem to be funnel-like in B. smithi. The angle of diverging reinforcements is wider in B. smithi than in R. gevreyi, P. prisca, S. tricarinata and co-existing D. syriaca. In these taxa, lateral reinforcements are moreover developed as rounded and prominent keels. Finally, the lateral fields are markedly longer in B. smithi than in R. gevreyi, S. tricarinata, P. prisca, and D. syriaca.
The gladius of Boreopeltis smithi is most similar to that of Late Jurassic B. sagittata (Naef, 1921; note that neither Naef nor any other subsequent workers realized the close phylogenetic relationship between B. sagittata (Fig. 9.1) and Paraplesioteuthis sagittata (Fig. 7.1)) and Early Cretaceous B. helgolandiaeEngeser and Reitner, 1985. These three taxa share a similar apical angle, conus shape, comparatively long and narrow lateral fields and lateral subdivided reinforcements. B. smithi mainly differs from B. sagittata and B. helgolandiae through the length of the lateral fields (Fig. 9.1–9.3). Lateral field length is considerably longer in B. sagittata (80% of the total gladius length) and B. helgolandiae than in B. smithi.
Boreopeltis smithi n. sp. is possibly conspecific with a single fragmentary specimen from Sâhel Aalma vaguely described and sketched by Roger (1946, p. 16, fig. 9), “Geoteuthis sahel-almae (Fraas) (Naef).” As seen in Figure 9.4, the posterior gladius is very similar to B. smithi. Unfortunately, this specimen was not found in the collection of the MNHN, where Roger's other types have been investigated. A detailed comparison is thus impossible. In any case, Roger's specimen is definitely not identical with “Sepialites sahil-almae (Fraas, 1878)” of Naef (1922, p. 134, Fig. 49c). “S. sahil-almae (Fraas, 1878)” exhibits prominent lateral keels and is, therefore, considered to be conspecific with Dorateuthis syriaca (see above).
Donovan (1977) reconsidered Roger's (1946) specimen and compared it with “Geoteuthinus” Kretzoi, 1942. “Geoteuthinus” is however a younger synonym of DoryanthesMünster, 1846 and has nothing to do with our specimen (Fuchs, 2006c).
Later, Engeser and Reitner (1986) and again Engeser (1988) revised to taxonomic position of Roger's (1946) specimen. They concluded that it belongs to “Geopeltis sp. 1,” but the present specimen has nothing in common with the loligosepiid genus Geopeltis Van Regteren Altena, 1949.
Fuchs et al. (2007b) recently reconsidered the phylogenetic relationships of the Prototeuthidina. These authors supported the initial hypothesis of Bandel and Leich (1986) and later elaborated by Engeser and Keupp (1999), Haas (2002), Fuchs (2006a, 2006b; 2007) in which the general gladius outline of the Prototeuthidina is only superficially similar to living teuthids and that the soft-part morphology known from Plesioteuthis, Senefelderiteuthis and Dorateuthis (eight arms, dorsal arm pair elongated, cirri, radial suckers, interbrachial web) instead suggest affinities with the eight-armed vampyropods. Present investigation on numerous arm crowns of D. syriaca again failed to find a preserved fifth arm pair. Instead, the presumed existence of a crop in D. syriaca corroborates the “Vampyropoda-hypothesis” as a crop is unknown in decabrachiate coleoids (Nesis, 1987; Mangold and Young, 1998). Although the phylogenetic value is still poorly known (even in living taxa), the cephalic cartilage morphology of Dorateuthis also indicates closer affinities with vampyropods than with teuthids (compare Nixon, 1998, fig. 3). The gill apparatus of both D. syriaca and P. prisca is similar to modern teuthids, but this can be seen as a homoplasy. Reitner (2009) recently stated that gill type is most probably due to a similar life style and has no phylogenetic significance.
The phylogenetic origin of the Prototeuthidina within the vampyropods is still unsolved. Naef (1922) did not discuss the possible origin. Jeletzky (1966) considered the prototeuthidids to be descendents of the Loligosepiina Jeletzky, 1965. Donovan (1977) also suggested that prototeuthidids became differentiated from loligosepiids, although he illustrated in his text-fig. 11 a direct origin from phragmoteuthidids. Similarly, Doyle et al. (1994, fig. 1) indicated that the prototeuthidid lineage and the loligosepiid/teudopseid lineage originated independently within the Phragmoteuthida Jeletzky in Sweet, 1964. According to Fuchs (2006b, fig. 4.1–4.3), prototeuthidids evolved from a hypothetic ancestor with a demineralized phragmocone and a phragmoteuthidid-like tripartite proostracum. He regarded the prototeuthidids to be a stem-group of the Vampyropoda. In fact, the assumption of a direct origin from a loligosepiid root-stock appears unlikely since the ventrally closed (funnel-like) prototeuthidid conus is certainly a plesiomorphic character inherited form a belemnoid (?phragmoteuthidid) phragmocone (Fuchs, 2006b; Fuchs et al., 2007b). The cup-like loligosepiid conus, in contrast, clearly represent a reduced conus (Fuchs, 2006b; Fuchs and Weis, 2008). Therefore, it is plausible that the Prototeuthidina diverged from the vampyropod stem-lineage earlier than the Loligosepiina did. This assumption is supported by the oldest known gladius. The gladius of “Loligosepia” neidernachensis from the Rhaetian (Reitner, 1978) is, as mentioned above, rather a prototeuthidid and closer related to Paraplesioteuthis than to any other loligosepiid.
Despite this incomplete knowledge related to the direct origin of the Prototeuthidina, polarization of some character states can be ascertained. The existence of a funnel-like conus is certainly plesiomorphic within the Prototeuthidina. In case the derivation from a phragmoteuthidid “root-stock” is correct, the absence of hyperbolar zones must be an apomorphy (Fuchs, 2006b; Fuchs and Weis, 2008). Lateral reinforcements are a unique diagnostic character of the Prototeuthidina and therefore another apomorphy. Similarly, the presence of very short and narrow lateral fields is unknown from the Phragmoteuthida, Loligosepiina or Teudopseina and therefore a further apomorphy of the Prototeuthidina. The character state of a more or less straight anterior median field end ( = triangular median field) is equivocal. Since phragmoteuthids exhibit an anteriorly convex median field, Fuchs and Weis (2008, p. 108) assumed that a more or less straight (or slightly concave) anterior median field end developed independently in Prototeuthidina and Loligosepiina.
In summary, we are far from presenting a robust cladogram owing to an erratic fossil record of prototeuthidid gladii; nevertheless, we are able to detect evolutionary transformations. Based on the new evidence provided here, the evolutionary pathways and phylogenetic relationships within the Prototeuthidina as suggested by Fuchs et al. (2007b, fig. 5) must be accordingly revised (Fig. 10).
Starting from the Middle Jurassic, there are two main lineages observable within the Prototeuthidina. Both lineages most probably originated from Paraplesioteuthis-like ancestors. One lineage retained many characters of Paraplesioteuthis (e.g., comparatively wide median field, plate-like reinforcements), but slightly reduced the conus length and elongated the lateral fields. This lineage is leading to the genus Boreopeltis (Fig. 10). A reconstruction of the intrageneric relationships is problematic owing to limited differences and different stratigraphic occurrences. The lineage B. sagittata, B. helgolandiae, and B. smithi n. sp. might demonstrate gradual gladius transformations of a single species through time.
The other lineage portrays a reduction in the median field width and modification of the plate-like lateral reinforcements into keels. This lineage leads to Romaniteuthis, Rhomboteuthis, Senefelderiteuthis, Plesioteuthis and Dorateuthis (Fig. 10). Within this “plesioteuthidid” lineage, Middle Jurassic Romaniteuthis is primitive because this genus retains a bipartite median ridge and comparatively wide lateral fields as seen in Paraplesioteuthis. Whether Rhomboteuthis or Nesisoteuthis are close to the Senefelderiteuthis-Plesioteuthis-Dorateuthis clade is still unclear, owing to an incomplete knowledge of their gladius. This clade is characterized by a gladius with narrow lateral fields (as a result of a lateral field reduction, the conus is also reduced). In contrast to Senefelderiteuthis, both Plesioteuthis and Dorateuthis share a gladius with a clearly reduced lateral field length. As Senefelderiteuthis contains mostly plesiomorphies (i.e., funnel-like conus, slender gladius, comparatively long lateral fields, lateral keels, bipartite median ridge), this taxon might be the source of the more advanced Plesioteuthis/Dorateuthis clade. Plesioteuthis displays a remarkably modified gladius in which the median keel developed, whereas Dorateuthis further reduced the lateral fields.
The authors sincerely thank the collectors: P., J., and A. Abi Saad along with F. Bachia and his employees. We also thank M. Freiji who acquired many of the coleoid specimens from various other collectors and brought them to us for purchase. R. Farrar assisted with finding publications. Furthermore, we are grateful to R. Young and S. von Boletzky for fruitful discussions about modern squid morphology. Finally, we want to thank C. Klug and G. Schweigert for their thorough revisions.
- Accepted 21 October 2010.