cf. PAN-CRYPTODIRA Joyce, Parham and Gauthier, 2004

cf. Pan-Cryptodira indet.

Material. MCNA 15055, a partial peripheral plate lacking the medial and latero-posterior regions (Fig. 2A–C).

Description. The plate lacks its medial region. For this reason, the width/length cannot be estimated. It is a peripheral that does not contact the plastron, but is close to the plastral bridge. This plate is crossed by a sulcus located between two marginal scutes. Considering that the most distal region of this kind of sulci is always anteriorly directed, this plate is identified as a left peripheral. The angle between the dorsal and the ventral surfaces is greater in the anterior region than in the posterior one (Fig. 2B and C). Therefore, the anterior margin corresponds to that located nearest to the plastral bridge, this plate being a posterior peripheral. Although the width of this plate is not known (i.e. the length between the medial and distal margins), the ventral surface covered by the marginal scutes is relatively narrow, substantially less than half of the width of the plate (Fig. 2B). A well-developed change of level, which does not match the ventro-medial margin of the marginal scutes, is present. It is parallel to that margin, being more medially located. The outer surface is partially altered. However, several discontinuous dichotomous sulci are preserved. Therefore, this plate was ornamented. Because the medial region is not preserved, it is not possible to know if the pleural scutes overlapped the peripheral plates. It is interpreted that, if such overlap occurs, it would be short, because the marginal scutes cover all the preserved dorsal region of the peripheral plate.

Discussion. The limited information provided by this element prevents its accurate systematic allocation. Characters, such as the absence of the autapomorphic ornamental pattern of Dortokidae, a clade of basal Pan-Pleurodira recognized in the continental deposits of Laño, as well as its larger size, allow excluding its assignment to that group. The absence of tubercles confirms that this specimen cannot be assigned to Solemydidae, another clade of turtles recognized in Laño. The presence of an ornamental pattern composed by discontinuous and dichotomous sulci is shared with Bothremydidae (Pleurodira), the third clade recognized in Laño. However, characters, such as the relatively narrow ventral development of the marginal scutes allow excluding its assignment to this group. This ornamental pattern has also been identified in several undetermined pan-cryptodiran taxa recognized both in the Early and Late Cretaceous of Western Europe (see Pereda-Suberbiola et al., 2009a and Pérez-García and Murelaga, 2014), generally referred as “chelydroids” in aspect, and probably belonging to different clades. Although all or at least some of these poorly known taxa have fontanelles between the costal and the peripheral plates, the absence of information on the medial region of MCNA 15055 does not allow comparing this character.

PAN-PLEURODIRA Joyce, Parham and Gauthier, 2004

PLEURODIRA Cope, 1864

PELOMEDUSOIDES Cope, 1868

BOTHREMYDIDAE Baur, 1891

BOTHREMYDINAE Baur, 1891

BOTHREMYDODDA Baur, 1891 sensu Gaffney, Tong and Meylan, 2006

BOTHREMYDINI Baur, 1891 sensu Gaffney, Tong and Meylan, 2006

Foxemydina Gaffney, Tong and Meylan, 2006

cf. Polysternon atlanticum

Material. MCNA 10863, the almost complete medial half of a sixth left costal plate (Fig. 2D–F).

Description. The only broken margin corresponds to the distal one. Therefore, the preserved region of this costal plate documents the morphology of the margins, which contacted with the neural series and with other costals. The antero-medial contact with a neural is observed (Fig. 2E, F). The posterior region of this neural plate was relatively wide in relation to its length. An approximately parallel margin to the axial axis suture, which is interpreted as the medial contact with another costal plate, is developed in the medial edge of the preserved costal. The postero-medial edge of this plate corresponds to the suture with another costal, the medial area of contact of the last costals of the carapace being developed in zigzag.

The medial region of MCNA 10863 is covered by a portion of a vertebral scute. The sulci separating this scute and two pleural ones are observed (Fig. 2E, F). Therefore, this plate is an even-numbered costal. Given this, plus its curvature, and the described contact with other plates, this element can be attributed to a sixth left costal. Taken this into account, this plate antero-medially contacts the sixth neural, medially the sixth right costal, and postero-medially the seventh right costal plates. Consequently, the neural series is interpreted as discontinuous (i.e. not continuous between the nuchal and the suprapygal plates), being composed of six plates. The best preserved areas show that the plate was ornamented by discontinuous and dichotomous sulci (Fig. 2E).

Discussion. Despite its limited information, it is possible to observe that MCNA 10863 shares several characters with Bothremydidae: a discontinuous neural series, the posterior region of the last neural being noticeably wider than long; the morphology and arrangement of the contact between the vertebral and the pleural scutes; and finally, the ornamental pattern.

The only bothremydid from the European Late Cretaceous record hitherto characterized by the presence of six neurals is Iberoccitanemys Pérez-García, Ortega, and Murelaga, 2012. However, variability in this character is known in some bothremydid taxa, as Chedighaii hutchisoni Gaffney, Tong and Meylan, 2006, from the Late Cretaceous of United States, in which the number of neurals varies from six to seven. In Late Cretaceous taxa from southwestern Europe represented by several specimens, such as Iberoccitanemys, Rosasia Carrington da Costa, 1940 and Foxemys Tong et al., 1998, the so far known number of these plates is constant (being seven in the latter two taxa). However, unpublished material of Polysternon provinciale (Matheron, 1869) reveals variability in the neural series (six or seven plates). Therefore, some specimens of Polysternon provinciale with the same number of neurals as in Iberoccitanemys are known. The Laño bothremydid Polysternon atlanticum (the only bothremydid species currently identified here sensu Pérez-García et al., 2010) has been recognized by relatively scarce and disarticulated material. Lapparent de Broin and Murelaga (1999) indicated that its neural series includes at least seven neurals, the costals 8 medially meeting. However, the morphology of the medial region of the sixth costal MCNA 7051 from Laño (see plate 5.12 in Lapparent de Broin and Murelaga, 1999) is similar to that of the specimen MCNA 10863 from Albaina. Therefore, we recognize that a neural series composed of six plates may be present in some Polysternon atlanticum specimens. MCNA 10863 cannot be referred to Polysternon provinciale because thin striations composed of parallel streaks are absent on the carapace. Therefore, the plate from Albaina probably belongs to Polysternon atlanticum, the only bothremydid recognized in Laño. However, given the limited availability of characters and the described compatibility of this element with other European representatives of Foxemydina, i.e. Iberoccitanemys (taking into account its morphology, but also the arrangement of the neural and the costals with which it is in contact), we opted for its identification as cf. Polysternon atlanticum.

ORNITHISCHIA Seeley, 1887

ORNITHOPODA Marsh, 1881

HADROSAUROIDEA Cope, 1869 sensu Sereno, 1986

Hadrosauroidea indet.

Material. LU-JL-LAÑ001, a fragmentary right femur (Fig. 3).

Description. LU-JL-LAÑ001 consists of the medial and distal part of a right femur, which is broken just below the fourth trochanter (not preserved in the specimen). The bone shows evidence of erosion, especially on the distal condyles (Fig. 3A and B). As preserved, the femur is 155 mm long for a maximum distal width of 60 mm (see measurements below); it probably corresponds to a femur that did not exceed 380 mm in length when complete. The shaft is straight in lateral and medial views (Fig. 3C and D). The cross-section at the preserved proximal end of the shaft is ovoid, wider than long; the medial side is longer anteroposteriorly than the lateral one. The anterior surface of the shaft is convex whereas the posterior surface is slightly concave. Although eroded, the distal condyles seem to have been well-developed. They are expanded anteriorly and posteriorly to form an H-shaped outline in distal view (Fig. 3E). The anterior intercondylar groove for the passage of the extensor tendons is deep and U-shaped; it is partially enclosed by expansions of both medial and distal condyles. In anterior view, the medial condyle is wider than the lateral one and is located distally below it (Fig. 3A). In distal view, the intercondylar flexor groove is deep, U-shaped, and narrower than the anterior groove. The medial posterior condyle is wider and more projected than the lateral one, with a slightly medial orientation (Fig. 3B). The lateral condyle is projected perpendicular to the long axis of the femur. The posterolateral surface of the lateral condyle shows a shallow vertical groove for the M. ilio-fibularis.

Measurements of femur LU-JL-LAÑ001: –

length (as preserved) = 155 mm;

Discussion. The combination of femoral characters observed in LU-JL-LAÑ001, i.e. straight distal shaft, anteroposterior expansion of the distal condyles, and deep, nearly closed intercondylar extensor groove on the distal end, allow us to assign the specimen to the clade Hadrosauroidea (see Horner et al., 2004, Prieto-Márquez, 2008 and Xing et al., 2012: supplement). We follow here the stem-based definition of Hadrosauroidea by Sereno (1998): all hadrosauriforms closer to Parasaurolophus Parks, 1922 than to Iguanodon Mantell, 1825. It should be noted that the development of a deep intercondylar extensor groove, with the edges of the groove meeting or nearly meeting anteriorly to enclose an extensor tunnel, has been regarded by Wu and Godefroit (2012) as a synapomorphy of Hadrosauridae, a clade defined by these authors as the most recent common ancestor of Bactrosaurus Gilmore, 1933 and Parasaurolophus, plus all the descendants of this common ancestor (see Sereno, 2005 and Prieto-Márquez, 2010 for different definitions). According to Wu and Godefroit (2012), a deep intercondylar extensor groove is absent in non-hadrosaurid ornithopods, including in basal iguanodontians, such as rhabdodontids, which subsisted in Europe until the Latest Cretaceous (Ősi et al., 2012 and Weishampel et al., 2003).

In Europe, hadrosauroid femora have been found in several Maastrichtian localities, from the Iberian Peninsula to Crimea passing through northern and central Europe (see Cruzado-Caballero, 2012 for a list). Some of these femora have been referred in the literature to Hadrosauridae indet. or Euhadrosauria indet. (see Cruzado-Caballero, 2012 and Dalla Vecchia, 2006), but others belong to taxa like Telmatosaurus Nopcsa, 1903 (Weishampel et al., 1993), Tethyshadros Dalla Vecchia, 2009 (Late Campanian to Early Maastrichtian according to Dalla Vecchia, 2009b), “Orthomerus” (now regarded as a nomen dubium; see Brinkmann, 1988 and Horner et al., 2004) and the lambeosaurines Arenysaurus Pereda-Suberbiola et al., 2009a and Blasisaurus Cruzado-Caballero et al., 2010a.

The general morphology of the hadrosauroid femora is quite conservative, but some individual variation has been observed (Brett-Surman and Wagner, 2007). Few ontogenetic changes have been documented in the limb bones (i.e., femora) and at least immature specimens of basal hadrosauroids can provide reliable information for taxonomic and phylogenetic inferences (Prieto-Márquez, 2011). LU-JL-LAÑ001 looks roughly like two hadrosauroid femora from Els Nerets, Lleida (Casanovas et al., 1985) and Peguera, Barcelona (Pereda-Suberbiola et al., 2003b), but they show signs of erosion and/or deformation. The Albaina specimen apparently differs from the femora of Telmatosaurus (Weishampel et al., 1993) and Arenysaurus (Cruzado-Caballero et al., 2013) in the absence of a closed extensor tunnel (but an artefact due to erosion cannot be totally discarded for the Albaina femur). Moreover, the Albaina specimen lacks the elliptical foramen located just above the anterior distal condyles in several femora from Blasi 1 in Arén (Huesca) that could belong to Blasisaurus (Cruzado-Caballero, 2012). LU-JL-LAÑ001 differs from the femora of “Orthomerus dolloi” Seeley, 1883 from Limburg (Mulder, 1984 and Mulder et al., 2005) in having a U-shaped flexor groove (instead of V-shaped), and from the femur of an indeterminate hadrosauroid from Bavaria in that the flexor groove is not so wide (Wellnhofer, 1994). Comparisons with Tethyshadros are difficult because in the latter the femur is only available in lateral view (Dalla Vecchia, 2009b). The femur of both Pararhabdodon Casanovas, Santafé & Isidro, 1993 (Casanovas et al., 1999 and Prieto-Márquez et al., 2006a) and Canardia Prieto-Márquez et al., 2013 from Midi-Pyrénées is presently not known. At last, the femora of “Orthomerus weberi” Riabinin, 1945 from the Crimea Peninsula and of an indeterminate hadrosauroid from Bulgaria (Godefroit and Motchurova-Dekova, 2010) are very fragmentary to make helpful comparisons. Pending the discovery of new material, the Albaina femur is here referred to Hadrosauroidea indet.

As noted above, the estimated total length of the Albaina femur would be around 35 cm. It is possible to estimate the body length of the Albaina hadrosauroid on the basis of more complete hadrosauroid specimens. The femur of the complete and articulated specimen of Tethyshadros insularis Dalla Vecchia, 2009 from the latest Cretaceous of Italy is 420 mm long, representing 11.6% of the total skeleton length (3620 mm from the tip of the snout to the distal end of the tail; see Dalla Vecchia, 2009b: table). In specimens of the hadrosaurine Maiasaura peeblesorum Horner and Makela, 1979 from North America, the femur length ranges from 13.3% to 15.5% of the body length (including different growth stages, from nestlings to adults; Horner et al., 2000: table 1). This ratio is similar in other hadrosauroid taxa, including basal and derived forms (see Brett-Surman, 1989). Thus, the Albaina individual was approximately 2.5–3 m in body length at death. Using the equation of Alexander (1989), the estimated body mass of this individual was approximately 95 kg [body mass in kg = a ^b , where a (circumference of femur in mm) = 0.00016 and b = 2.73]. Histological studies by Horner et al. (2000) suggested that the late juvenile stage represented by individuals of about 3 m may reach an adult size of about twice, but we prefer to be cautious about this type of estimates until we have a better knowledge of both the mechanisms and growth rates in hadrosauroids (see Brinkman, 2011).

The presence of three clades of turtles, all of them represented by taxa recognized as new forms, was notified in the continental levels of Laño (see Lapparent de Broin and Murelaga, 1996). A turtle currently identified as a terrestrial form (Scheyer et al., 2012), Solemys vermiculata Lapparent de Broin and Murelaga, 1996, was erected there. It belongs to a clade of basal turtles (Solemydidae). Solemydidae is known from the Late Jurassic to the Latest Cretaceous of North America and Europe. The other two taxa are members of Pan-Pleurodira, and both are aquatic forms. One of them, Dortoka vasconica Lapparent de Broin and Murelaga, 1996, is of small size (being the estimated maximum length of the adults less than 20 cm), which belongs to a primitive clade of Pan-Pleurodira exclusively recorded in Europe from the Early Cretaceous to the Paleocene (Lapparent de Broin et al., 2004 and Pérez-García et al., 2014). The third clade recognized is Bothremydidae (Pleurodira), represented by Polysternon atlanticum, and also defined in Laño (see Lapparent de Broin and Murelaga, 1999 and Pérez-García et al., 2012). Although Bothremydidae is a clade of Gondwanan origin, several lineages of this group diachronically migrated to Europe (Pérez-García and Lapparent de Broin, 2014). Thus, bothremydids are the most abundant and diverse clade of turtles currently recognized in the European latest Cretaceous record. Representatives of these three groups of turtles have been found in numerous outcrops of the uppermost Cretaceous of southwestern Europe.

Some taxa of Bothremydidae have been interpreted as marine littoral forms, or with the ability to use the coastlines to migrate (Lapparent de Broin and Werner, 1998). The bothremydids found in the Late Cretaceous of Europe are considered to have been freshwater taxa. However, their identification in some deposits interpreted as located near the coastline suggests that they could live in environments with higher salinity than those in which the representatives of Dortokidae lived (Guede et al., 2013 and Pérez-García, 2012). Therefore, and taking into account the environment in which the three clades of turtles recognized in the continental deposits of Laño lived, the occurrence of Bothremydidae in Albaina appears more probable than that of the representatives of the other two clades.

The presence of an indeterminate Pan-Cryptodira has also been reported in the Spanish uppermost Cretaceous site of Lo Hueco (Cuenca) (Pérez-García et al., 2009). The presence of unpublished “chelydroid”-like taxa has been also recognized in other Late Cretaceous sites of Western Europe. These taxa, recognized as aquatic forms, could correspond to freshwater or marine turtles (Guede et al., 2013 and Pérez-García et al., 2009). No sea turtle has been identified in the uppermost Cretaceous of the Basque–Cantabrian Region, to the exception of scapular remains of an Allopleuron-like chelonioid from the Santonian of San Pantaleón de Losa (Burgos) studied by Bardet et al. (1993). Only another indeterminate Pan-Cryptodira has been mentioned in Mesozoic levels of the Basque–Cantabrian Region, more precisely from the Hauterivian-Barremian of Vega de Pas (Cantabria) (Pérez-García and Murelaga, 2014).

Therefore, the peripheral plate found in Albaina allows the identification of a fourth taxon in the Laño quarry. Unlike taxa so far described there, this form probably belongs to Pan-Cryptodira.

LU-JL-LAÑ001 is the second hadrosauroid remain found in the Laño quarry, the first one being an isolated tooth (MCNA 10510) from the fluvial deposits of the Laño 1 site, ascribed to the Late Campanian–Early Maastrichtian (Pereda-Suberbiola et al., 2003a). Buffetaut (2005) questioned the provenance of this tooth and suggested an accidental pollution from the Late Maastrichtian marine beds that overlie the older sandy and silty layers which have yielded the terrestrial and freshwater vertebrate remains. However, this hypothesis is rejected here because the hadrosauroid tooth MCNA 10510 was obtained by screen washing of the fossiliferous fluvial sands and silts of Laño 1, and there is no clear evidence of re-elaboration processes in these beds (Pereda-Suberbiola et al., 2000). Consequently, the Laño quarry provides information of the occurrence of hadrosauroids in Iberia as early as the Late Campanian: Laño is currently the only Iberian site where hadrosauroid material has been found together with rhabdodontid and titanosaurian remains, the locality of Els Nerets in the South-central Pyrenees of Lleida attesting of a similar dinosaur association is currently regarded as Late Maastrichtian in age (B. Vila, pers. comm.). The association of hadrosauroids with Rhabdodon is unusual in the latest Cretaceous of southern France: it has been mentioned only in the Late Maastrichtian Vitrolles–La Plaine locality of Provence (although an artefact due to reworking cannot be excluded; see Valentin et al., 2012).

It is interesting to note that the Albaina femur was found in marine deposits that have yielded benthic foraminifera, molluscs, arthropods, selachians, actinopterygians and marine reptiles, including turtles (Cappetta and Corral, 1999). In addition to the Laño quarry, hadrosauroid fossils have been reported from various Late Maastrichtian marine sites of Europe: the Maastrichtian type area (Buffetaut, 2009 and Jagt et al., 2003; Limburg and Liège (see Mulder et al., 2005 and Weishampel et al., 1999), Midi-Pyrénées (Bilotte et al., 2010, Laurent, 2003, Laurent et al., 2002, Laurent et al., 1999 and Paris and Taquet, 1973), southern Bavaria (Wellnhofer, 1994), northwestern Bulgaria (Godefroit and Motchurova-Dekova, 2010), and the Crimean Peninsula (Riabinin, 1945). The relative abundance of hadrosauroid remains in Late Maastrichtian marine deposits of the European archipelago probably reflects the dominance of these herbivorous dinosaurs on nearby landmasses, as suggested by the large amount of fossils found in contemporaneous continental sites of Europe and especially in the Ibero-Armorican Domain (see Dalla Vecchia, 2009a and Prieto-Márquez et al., 2013). The occurrence of hadrosauroid fossils in shallow marine environments can be interpreted as the result of the passive transport of floating carcasses over fairly long distances from the mainland (Buffetaut, 1994 and Godefroit and Motchurova-Dekova, 2010).

Horner (1979) listed the dinosaur specimens from Upper Cretaceous marine deposits of North America and noted the relative abundance of hadrosaurines (unadorned or solid crested hadrosaurids), primarily in the Western Interior, suggesting that may have inhabited coastal environments. Horner (1979) also noticed that approximately one-half of the specimens belong to young or juvenile individuals the size or smaller than the type of Claosaurus agilis (Marsh, 1872) (femur length 670 to 676 mm; Lull and Wright, 1942), and over three-quarters of the specimens are from individuals smaller than the type of Hadrosaurus foulkii Leidy, 1858 (femur length 1055 mm; Prieto-Márquez et al., 2006b). According to Godefroit and Motchurova-Dekova (2010), the overrepresentation of immature specimens in marine deposits can be explained taphonomically as the result of an attritional death profile of the fossil accumulation (instead of a local catastrophic event), with a selective mortality of younger and smaller individuals (see Lauters et al., 2008).

Lambeosaurines and non-hadrosaurid hadrosauroids have been reported in Europe (Dalla Vecchia, 2009a, Pereda-Suberbiola et al., 2009b and Prieto-Márquez et al., 2013). The possible presence of hadrosaurines (Cruzado-Caballero et al., 2010b) has been questioned by Prieto-Márquez et al. (2013). With regard to the hadrosauroid records from the Late Maastrichtian marine deposits of Europe, the specimens are mostly fragmentary and not diagnostic beyond Hadrosauroidea gen. et sp. indet. All these specimens, with the exception of “Orthomerus weberi” from Crimea (estimated femur length 750–800 mm; Riabinin, 1945), belong to small individuals (i.e., femur length of “Orthomerus dolloi”: 495 mm, Seeley, 1883; femur length of the Bavarian taxon: 340 mm, Wellnhofer, 1994). We can wonder if the small size of the specimens reflects the presence of immature individuals or of small-bodied adults. The Albaina femur is presently the only hadrosauroid specimen found in the marine facies of Europe that has been studied from a histological point of view. The analysis indicates that the specimen belongs to a not fully-grown individual of a moderate-size species.