A tiny lizard (Lepidosauria, Squamata) from the Lower Cretaceous of Spain

Authors


Corresponding author.

Abstract

Abstract:  The smallest living amniotes are all lizards, but the fossil history of this size trait in Squamata is difficult to follow because small skeletons have low preservation potential and are often hard to detect in the field. A new squamate taxon, Jucaraseps grandipes gen. et sp. nov., is here described on the basis of an articulated skeleton from the Early Cretaceous Spanish lagerstätten of Las Hoyas. It differs from other known Mesozoic lizards in combining very small body size with a short rostrum, low maxillary tooth count, a relatively slender and elongated body, and short limbs with large hind feet. Phylogenetic analysis using TNT places it on the stem of a clade encompassing scincomorphs, gekkotans, snakes, amphisbaenians and anguimorphs. Comparison with modern lizards suggests it was probably a cryptic surface or subsurface ground dweller but not a burrower.

For terrestrial animals, small body size confers a number of advantages, including shorter maturation times, increased agility and ease of concealment, and lower energy requirements (Blanckenhorn 2000). These, in turn, can allow for a higher population density in a given area and thereby a larger gene pool (Rodda et al. 2001). Amongst living amniotes, the smallest recorded taxa are all lizards, most notably species of the gekkotan Sphaerodactylus (e.g. S. parthenopion, mean snout-vent length (SVL) 16 mm, Hedges and Thomas 2001; Daza et al. 2008) and of the Malagasy Leaf Chamaeleon, Brookesia (e.g. B. minima, mean snout-vent length (SVL) 17 mm; Blanckenhorn 2000). Very small lizards are rarely preserved as fossils, but here we describe a tiny articulated lizard skeleton from the Lower Cretaceous of Spain.

Geological background

The Las Hoyas fossil site is located in the Serranía de Cuenca (south-western Iberian Ranges, Spain). It is a lagerstätte, the fine-grained laminated limestone of which belong to the late Barremian La Huérguina Formation (Buscalioni and Fregenal-Martinez 2010) and have produced a diverse assemblage of plants (e.g. charophytes, ferns, gymnosperms, angiosperms), invertebrates (crustaceans, diverse insects) and vertebrates (sarcopterygian and actinopterygian fish; anuran, caudate and albanerpetontid amphibians; turtles, lizards, crocodiles, pterosaurs, dinosaurs, birds: Ortega et al. 1999; Vullo et al. 2009) that lived in a seasonal subtropical wetland ecosystem (Buscalioni and Fregenal-Martinez 2010). Las Hoyas is one of only two Mesozoic localities on the Iberian Peninsula yielding fully articulated lizard remains, the other being La Pedrera de Meià (Late Berriasian–Early Valanginian, Catalonia, e.g. Bolet and Evans 2010). The Las Hoyas fossil site is the more productive of the two, although lizard material is still relatively rare. The locality has yielded six described (and further undescribed) specimens of Meyasaurus diazromerali (Evans and Barbadillo 1997); one of Hoyalacerta sanzi (Evans and Barbadillo 1999); two of Scandensia ciervensis (Evans and Barbadillo 1998; Bolet and Evans 2011); a partial paramacellodid (Evans and Bolet in press); and the new genus and species described herein.

Material

The new specimen, Museo de Cuenca, LH 18505, consists of a main block bearing most of the skull and skeleton of a small lizard and a small partial counterpart bearing impressions and fragments of the skull and neck.

Institutional abbreviations.  IPPS, Instituto Provincial de Palaeontologia, Sabadell, Spain; LH, Museo de Cuenca, Las Hoyas Collection, Cuenca, Spain.

Anatomical abbreviations. ac, astragalocalcaneum; acf, anterior coracoid fenestra; an, angular; ar, articular; br, braincase; c, carpals; cl, clavicle; co, coronoid; co, coracoid; d, dentary; dt3,4, distal tarsals 3,4; f, frontal; fi, fibula; h, humerus; j, jugal; lf, left femur; lil, left ilium; lpu, left pubis; m, maxilla; mt5, fifth metatarsal; n, nasal; o, olecranon epiphysis; p, parietal; pa, palatine; pf, postfrontal; pi, pisiform; pm, premaxilla; po, postorbital; pra, prearticular; prf, prefrontal; pt, pterygoid; q, quadrate; r, radius; rap, retroarticular process; rf, femur; ril, right ilium; rpu, right pubis; s1, s2, sacrals 1–2; sa, surangular; sc, scapula; scf, scapulocoracoid fenestra; sco, scleral ossicles; sp, splenial; sq, squamosal; st, supratemporal; ti, tibia; u, ulna; v, vomer; 1–5, digits 1–5.

Systematic palaeontology

LEPIDOSAURIA Haeckel, 1866
SQUAMATA Oppel, 1811
Genus JUCARASEPS gen. nov.

Type and only species. Jucaraseps grandipes sp. nov.

Derivation of name.  From the Jucar River of Cuenca, and (Latin) seps, a lizard or snake.

Diagnosis.  As for species.

Jucaraseps grandipes sp. nov.
Figures 1–3

Figure 1.

Jucaraseps grandipes LH 18505, holotype. A, dorsal view. B, interpretive drawing. Scale bar represents 5 mm.

Figure 2.

 Skull of Jucaraseps grandipes, LH 18505. A, as preserved, grey shaded areas represent bone on the counterpart block. B, reconstruction. Scale bar represents 1 mm.

Figure 3.

Jucaraseps grandipes, LH 18505, limb skeleton. A, left forelimb and B, right hind limb. Scale bars represent 1 mm

Derivation of name.  Latin, grandis, large; pes, foot, in reference to the proportionally large feet of this lizard.

Holotype.  LH 18505 (Fig. 1A), Museo de Cuenca, Cuenca, Spain.

Locality and horizon.  Las Hoyas (Cuenca, Spain). Calizas de la Huérguina Formation (Limestone Unit III, Ortega et al. 1999); Early Cretaceous (Late Barremian).

Differential diagnosis.  A small lizard characterised by the combination of small adult size, a low tooth count, an expanded retroarticular process on the lower jaw, an increased number of presacral vertebrae (31), deep cristae cranii (subolfactory crests) on the frontals and short limbs with large feet (pes 52 per cent of total limb length, 20 per cent of snout-vent length (SVL)).

It differs from other Las Hoyas lizards except Hoyalacerta in the combination of an extended presacral vertebral column and relatively short limbs; differs from Hoyalacerta in having a smaller number (8–10) of relatively large maxillary teeth (40 +  small teeth in Hoyalacerta), 31 presacral vertebrae (27–28 in Hoyalacerta) and frontal morphology (shallow cristae cranii in Hoyalacerta); differs from all other described skeletons of Jurassic–Early Cretaceous lizards in the combination of small size, a short rostrum, a shallow lower jaw with unicuspid teeth, a low coronoid and a large expanded retroarticular process, a long body and tail, the absence of osteoderms or cranial sculpture, and the relatively short limbs (forelimb 26.6 per cent SVL; hind limb 39 per cent SVL) with large hind feet (pes 52 per cent of total hind limb length, 20 per cent SVL).

Remarks.  The Japanese Kaganaias (Evans et al. 2006) resembles Jucaraseps in being long-bodied, but differs in being much larger, having more presacral vertebrae (36–46 dorsals, neck unknown) and having hind limbs that are much shorter in relation to the body.

Jucaraseps resembles the Upper Jurassic Guimarota genera SaurillodonSeiffert, 1973 and ChalcidosaurusKosma, 2004 in having a relatively short dentary, although the latter taxon has a higher tooth count (21–23) than the former (9–16, Kosma 2004). However, with the exception of one partially associated skeleton of Saurillodon (Broschinski 2000), these taxa are represented only by isolated jaws and their body proportions and skull morphology remain unknown, precluding further comparison with Jucaraseps. A similar problem exists for other Jurassic and Early Cretaceous lizard taxa known only from jaw material (e.g. Seiffert 1973; Evans 1998; Nydam and Cifelli, 2002).

Comparative description

The type and only specimen of Jucaraseps is an almost complete skeleton preserved in dorsal view. It has an estimated snout-vent length (SVL) of 27 mm and is interpreted as a subadult individual based on skeletal development (see Discussion: Ontogenetic status). The skull is compressed and some bones are poorly preserved making their boundaries difficult to reconstruct. Postcranially, the specimen is well preserved except for the right forelimb and includes a long but incompletely preserved tail (Fig. 1).

Skull.  The skull of LH 18505 is small (total length, premaxilla to posterior edge of skull, 5.8 mm) and narrow (3.4 mm width posteriorly), with a short snout and large orbits that contain remnants of the scleral ossicles (Fig. 2). None of the skull bones shows any trace of sculpture. The premaxillae are fused, but the ascending nasal process is broken. Only three small unicuspid premaxillary teeth are visible. The left maxilla bears about eight small, slightly recurved and relatively well-spaced unicuspid teeth. The tooth bases are visible through the bone of the maxilla and show that the implantation was pleurodont. A triangular maxillary facial process is visible on the left side, and a narrow posterior process extends below the orbit. The paired nasals border small nares anteriorly and meet the maxillae and the frontals posteriorly. They are separated from the prefrontals by a maxilla-frontal contact. The frontals are split between the part and counterpart blocks, but they can be reconstructed as long and paired, almost parallel-sided anteriorly, but increasing in width posterolaterally (Fig. 2B). Ventrally, the cristae cranii (subolfactory processes) are deep and partially enclose the olfactory tracts. The single parietal is crushed and some of the more ventral bones of the skull have been pushed through it. There is no obvious parietal foramen but given the damage, we cannot be certain it was absent. The lateral border of the parietal is obscured on the left by a displaced quadrate and other elements and on the right by damage, but there is no indication that the adductor muscles extended onto the dorsal surface. Short divergent postparietal (supratemporal) processes are separated by a gently curved posterior border with shallow nuchal fossae. The prefrontals are small, and the presence or absence of a free lacrimal is uncertain. The maxillary ramus of the jugal is preserved on the left. Its slender postorbital ramus is broken but bone fragments and impressions suggest it formed a complete orbital rim. An elongate, weakly concave right postfrontal straddles the frontoparietal suture. Lateral to it are narrow bars of bone that may be parts of the right postorbital and squamosal. Further posteriorly, a slender bone lateral to the parietal is probably the right supratemporal. On the left side, these elements may be represented by bone fragments between the quadrate and parietal. The left quadrate lies beside the parietal and has a rather ovoid outline. The right quadrate lies under the parietal but has broken through it posteromedial to the postfrontal.

The palate is mostly obscured by the dorsal skull elements. Paired vomers can be seen through the narial openings, and parts of the palatines and left pterygoid are visible through the orbits.

Poor preservation precludes description of the braincase, but it appears to be relatively large, a trait common amongst small lizards (Rieppel 1984).

Lower jaw.  The posterior part of the left lower jaw is partially exposed in lingual view, with its postcoronoid region preserved in impression. Overall, the jaw is long, straight and slender. The articular-prearticular, surangular, angular, coronoid and splenial are all separate. The dentary is almost completely hidden, but its posterior end did not extend beyond the posterior margin of the coronoid process. The coronoid is rather elongated with a concave ventral margin and a low dorsal process. The dentition is largely hidden by the overlying skull bones, but those teeth visible through the jugal are similar in size and shape to those of the maxilla. The right mandible is again mostly hidden below the skull but the surangular, the short wide articular surface for the quadrate, and a large, medially deflected, retroarticular process are exposed posteriorly.

Axial skeleton.  There are at least 31 presacral vertebrae, two sacrals and more than 40 caudals (the end of the tail is not preserved). The vertebrae are procoelous and bear low neural spines. In general, the vertebrae are rather large in proportion to the corresponding ribs. The cervical vertebrae are shorter than the dorsals, and although some short cervical ribs are in articulation, it is difficult to determine at what level they began. The first vertebra with long, probably sternal, ribs preserved is the ninth, but the cervical vertebral count could be between six and eight. Ribs are visible on all dorsal vertebrae except the last. They are relatively long, thin and posteriorly directed, giving the body a slender outline (trunk width 2.4 mm). Their proximal heads are mostly obscured by the adjacent vertebrae but where visible, are somewhat expanded and grooved posterolaterally. The tail is long (143 per cent SVL as preserved, c. 180 per cent SVL in life allowing for the missing part) and autotomous. The first few caudal vertebrae bear single transverse processes that quickly decrease in size. The anterior fracture planes pass through the transverse process, but the precise level at which they begin is difficult to identify.

Pectoral girdle and forelimbs.  The left pectoral girdle is partially preserved, with the scapula and coracoid in articulation but not completely co-ossified (Fig. 3A). There are two emarginations: one between the scapula and coracoid and one (anterior coracoid fenestra) within the coracoid. As preserved, the clavicle is narrow proximally and expands distally but this region is preserved only in impression and is incomplete. The left forelimb is more complete than the right. It is short (7.2 mm, 26 per cent SVL), slender and delicate, with a short humerus (2.4 mm) bearing well-ossified proximal and distal epiphyses. The radius and ulna are even shorter (1.8 mm), but the olecranon process of the latter is not completely fused to the shaft. Because of the very small size, it is difficult to differentiate between the carpals and the detached proximal epiphyses of the metacarpals, but several ossified carpals, including a large pisiform, are visible. The rest of the hand is irregularly preserved, one of the digits being lost (or hidden inside the matrix), and only digit four is complete. This complete digit is longer than the humerus (estimated manus length 3 mm).

Pelvic girdle and hindlimbs.  The pelvic bones are not co-ossified (Fig. 3B). The ilium has a long, tapering blade without an anterior tubercle. The pubis is slender with a long, anteriorly directed, symphysial component and a large proximal obturator foramen but no pectineal process. The ischium is hidden by the sacrum except for a small fragment close to the left pubis. The width of the body across the pelvis is approximately 2.8 mm. The hind limbs are gracile (10.55 mm, 39 per cent SVL) and are comprised of a short straight femur (3.2 mm), a shorter tibia and fibula (1.9 mm), and a long foot (5.45 mm, 52 per cent of total hind limb length; 20 per cent SVL). The left tibia has a clearly visible proximal epiphysis that is attached but not fully co-ossified with the shaft and is at least partially ossified. The astragalocalcaneum shows no remnant of a suture and bears articular surfaces for the tibia and the fibula. There are two distal tarsals (dt4 and dt3). The pes is almost complete, but the fifth metatarsal is partly hidden below the fourth. The phalangeal formula is 2:3:4:5:4, the fourth digit being the longest. In total, the pes is almost three times the length of the tibia.

Phylogenetic relationships

We performed a cladistic analysis using the matrix of Conrad (2008), modified by Norell et al. (2008) and Bolet and Evans (2010, 2011). The matrix is too large (254 taxa, 372 characters) to be run with PAUP (Swofford 2002), and we therefore used the programme TNT (Goloboff et al. 2003, 2008). Character ordering (additive characters) followed Conrad (2008). Jucaraseps can be scored for 124 of 372 characters (Table 1), but this is comparable with many fossil taxa because soft-tissue characters cannot be scored. Kuehneosauridae was the designated outgroup taxon.

Table 1.   Coding of 372 characters for Jucaraseps in the matrix of Conrad (2008) as extended in Norell et al. (2008).
000000-0001???00??000??0??0000???00000??????00??0?0??1000101
100-0??0????1?0?3??1?01???00100010???????????00?????????????
?0????????????????????????????????????????00?0?????????????0
???????0???0??00?????0?000110??000000????00000???1200??2?0??
0???0?000000?0??0??0?10?00????????00?00????1211?00?1????????
????0-0-00????????????????????????????????????????????????
????????????    

The analysis was first performed using the Traditional search facility of TNT (1000 replicates, TBR branch swapping). This yielded 10 equally parsimonious trees of Length (L) 3952 (consistency index (CI) = 0.136; rescaled CI (RC) = 0.096; retention index (RI) = 0.704, as reported after import into PAUP, Swofford 2002). In all trees, Jucaraseps was placed on the stem of Scleroglossa (scincomorphs, gekkotans, snakes, amphisbaenians and anguimorphs, but see Townsend et al. 2004; Vidal and Hedges 2005 for an alternative molecular squamate phylogeny), nested within a subset of Jurassic–Cretaceous fossil taxa (Fig. 4): the Upper Jurassic Eichstaettisaurus, Ardeosaurus and Bavarisaurus (Solnhofen, Germany; Hoffstetter 1964; Evans 1994a); the Middle Jurassic to Early Cretaceous Parviraptor (Britain, Portugal; Evans 1994b); and the Early Cretaceous Yabeinosaurus (China, Evans et al. 2005) and Sakurasaurus (Japan, Evans and Manabe 2009). The matrix is too large for calculation of Bootstrap or Bremer support so we reran the analysis using the TNT New Technology search with Ratchet (100 iterations; 1000 Random Addition (RAM) sequences). A second Traditional Search (1000 repetitions, TBR branch swapping) was then run using the three ratchet trees (L, 3946; CI, RC, RI as above) as starting trees. In this Traditional search, the saved trees exceeded the available computer memory (at 90 000), but the Strict, Adams (Combinable components) and Majority Rule consensus of these 90 000 trees were each well resolved with respect to the position of Jucaraseps. Although the full search found some shorter overall trees (L, 3943), the position of Jucaraseps was unchanged from that in the Traditional search.

Figure 4.

 Condensed cladogram showing the position of Jucaraseps and several other Jurassic–Early Cretaceous squamate taxa, based on an analysis of the matrix of Conrad (2008) run using TNT (Goloboff et al. 2003). The analysis included 254 taxa, the remainder of which fall into either Iguania or Scleroglossa.

Discussion

Ontogenetic status

Jucaraseps is the smallest of the Las Hoyas squamates (Table 2). Its very small size raises the question of its ontogenetic age. Age estimation for a fossil lepidosaur has to be based on indicators of skeletal maturity, notably terminal fusions of the braincase, vertebrae, limb girdles, and limb epiphyses (Maisano 2001, 2002). These fusions (particularly of the limb bone epiphyses) are generally indicative of the end of growth, but there is considerable variation in the timing and sequence of these events (Maisano 2002). In the Jucaraseps specimen, the pectoral and pelvic components are sutured but not fused, as in immature lizards. Against this, the co-ossification of the astragalus and calcaneum, the triangular rather than rounded shape of the fourth distal tarsal, the attachment (but not complete fusion) and mineralisation (at least partial ossification) of the limb and digital epiphyses and, to a lesser degree, the closure of the vertebral sutures suggest an animal approaching maturity, and this is supported by the complete formation and articulation of the skull bones (although these are thin). The open girdle sutures do not necessarily conflict with this. Body size reduction can be correlated with apparent skeletal immaturity, so an adult may show features that appear juvenile (e.g. small adults of the scincid Eumeces, Griffiths 1990). LH18505 is therefore interpreted as a subadult animal at perhaps 70–80 per cent of adult size (c. 30–33 mm estimated adult SVL, 7–9 mm estimated adult skull length). This is at the lower end of the size range of living lizards, especially given the presacral elongation, although not at the extreme represented by the miniaturised gecko Sphaerodactylus parthenopion (Hedges and Thomas 2001; Daza et al. 2008).

Table 2.   Las Hoyas lizard taxa.
GenusMax SVL mmSkull mmPS No.FL/SVL per centHL/SVL per centFoot/HL per cent
  1. SVL, snout-vent length; PS, presacral vertebral number; FL, forelimb length; HL, hind limb length. *Based on the largest available specimen of Meyasaurus diazromerali, but M. faurai from Montsec (IPPS 10) is larger (SVL c. 139 mm). †Estimate (incomplete skull in Scandensia; for Jucaraseps up to possible adult size).

Meyasaurus*521825415850
Scandensia3512†25–26475345
Hoyalacerta341027–2823.532.535
Jucaraseps24–33†7–9†31263454

Comparison

Although Mesozoic terrestrial/freshwater squamates with long bodies and reduced limbs are rare (e.g. Kaganaias, from the Early Cretaceous of Japan; Evans et al. 2006), there are many fossil taxa for which the postcranial skeleton is unknown or incomplete. A short-jawed, long-bodied lizard was described from the Middle Jurassic of England (Evans 1998), but this reptile differs markedly from Jucaraseps in having very elongate vertebrae with fused osteoderms. The genus Saurillodon from the Late Jurassic of Portugal (Seiffert 1973) is also short-jawed but a partial association (jaws, frontals, parietal) included humeri of relatively normal length (Broschinski 2000), with no indication of limb reduction. In a later study, Kosma (2004) described several additional species of Saurillodon from Guimarota, all based on isolated dentaries, as well as a second short-jawed taxon, Chalcidosaurus. He reconstructed these lizards (Kosma 2004, p. 156, fig. 138) with varying degrees of elongation and limb reduction, but this is conjectural pending recovery of postcranial material. Tarratosaurus (Broschinski and Sigogneau-Russell 1996) from the Early Cretaceous of Morocco is equally problematic. It was interpreted as a burrower on the basis of the short lower jaw (and hence short rostrum), but again the postcranial skeleton is unknown. Perhaps most relevant to Jucaraseps, the Las Hoyas genus Hoyalacerta (Evans and Barbadillo 1999) is also somewhat elongated (28–29 presacrals) and short-limbed but it differs cranially and dentally (frontal morphology, tooth number and size) from Jucaraseps, and its feet are relatively smaller (Table 2).

Phylogenetic relationships

The results of the phylogenetic analysis place Jucaraseps on the stem of Scleroglossa (sensuEstes et al. 1988; Conrad 2008) with several other Jurassic/Cretaceous taxa (Fig. 4). It thus falls within Conrad’s (2008) more inclusive Scincogekkonomorpha. However, this phylogenetic placement must be considered provisional for two reasons. The first is that the skull of Jucaraseps is incompletely known, limiting the number of characters that can be coded, most notably those of the palate, braincase and jaws. As skull characters make up a large proportion of the data matrix, future finds with more complete cranial material may change the topology of the tree.

The second reason is that the phylogenetic relationships of even extant squamates remain incompletely resolved. The first comprehensive phylogenetic analysis of Squamata was that of Estes et al. (1988). They recovered a tree with a primary dichotomy into Iguania (then Iguanidae, Agamidae and Chamaeleontidae) and Scleroglossa (all other squamates). Subsequent morphological analyses, including most of those involving fossil taxa, have obtained broadly similar results (reviewed inConrad 2008). However, phylogenetic analyses using molecular characters (e.g. Townsend et al. 2004; Vidal and Hedges 2005) and those using a combined molecular/morphological data set (e.g. Wiens et al. 2010; Müller et al. 2011) consistently obtain a markedly different topology in which Iguania lies closer to anguimorphs and snakes. In these trees, Gekkota (Townsend et al. 2004), Dibamidae (Vidal and Hedges 2005; Wiens et al. 2010; Müller et al. 2011) or a combination of the two (Bayesian trees of Wiens et al. 2010) form the sister group to the remaining squamates rendering Scleroglossa paraphyletic. There are therefore major issues to be resolved with respect to squamate relationships but these are outside the scope of the current work.

Lifestyle

The Las Hoyas wetland supported both an aquatic community (water plants, arthropods, fish, frogs, salamanders, turtles and crocodiles) and a terrestrial one (Buscalioni and Fregenal-Martinez 2010). The latter comprised a diversity of plants and insects, spiders and terrestrial vertebrates including albanerpetontid amphibians, terrestrial crocodiles, pterosaurs, dinosaurs, birds and lizards (Buscalioni and Fregenal-Martinez 2010).

Compared to the other Las Hoyas lizards (Meyasaurus, Scandensia, Hoyalacerta, and a paramacellodid, Evans and Bolet in press), Jucaraseps is characterised by small size, a narrow somewhat elongated body, relatively short forelimbs, longer hind limbs (but with short pro- and epipodials and a long pes) and a long tail (145–180 per cent SVL allowing for the missing part) (Fig. 5). On the basis of living lizards of similar size and proportions (e.g. the scincid Lerista elegans, SVL 21–37 mm, 0.597 g; Benesch and Withers 2002), Jucaraseps is estimated to have weighed less than one gram.

Figure 5.

 Las Hoyas lizards. A–D, scaled silhouettes of: A, B, Meyasaurus, A, holotype of Meyasaurus diazromerali (LH 370, subadult); B, possible maximum size of Meyasaurus based on adult of Meyasaurus fauraiVidal 1915 (IPPS 10, Montsec). C, Scandensia ciervensis (LH 11001, adult). D, E, Jucaraseps grandipes; D, subadult holotype specimen (LH 18505); E, estimate of adult size. F, G, Hoyalacerta sanzi; F, holotype specimen (LH 11000, subadult); G, estimate of adult size. H, life reconstruction of Jucaraseps grandipes. Scale bars represent 20 mm (A–G) and 10 mm (H).

The smallest living lizards are Brookesia chamaeleons and sphaerodactylid geckos, but both have relatively short bodies and tails. Presacral elongation is found primarily in modern Anguidae, Gymnophthalmidae and Scincidae (e.g. Wiens and Slingluff 2001; Brandley et al. 2008), but only in the latter group is body elongation commonly combined with small size (e.g. Greer 1989, 2001; Melville and Swain 2000; Greer and Wadsworth 2003). Greer (2001) examined representatives of 1206 living scincid species and recorded a modal SVL of 55 mm, with 24 species having a SVL of 34 mm or less (mostly members of the lygosomine genus Menetia). Small scincids seem therefore to provide the most appropriate eco-analogue for Jucaraseps, and studies examining the relationship between body proportions and lifestyle provide potentially useful insights into its lifestyle.

Given its size and shape, Jucaraseps seems to best fit the ecological model of small cryptic surface to subsurface dwelling scincids with moderately elongated bodies and short limbs (Greer and Wadsworth 2003), but with long feet that may have been used to aid balance and grip in a very small animal moving on uneven surfaces (Melville and Swain 2000). Jucaraseps is unlikely to have been a burrower, as lizards weighing less than one gram reportedly lack the power to force their heads through the substrate (Benesch and Withers 2002). Moreover, burrowers generally do not have the long tails of surface dwellers (Brandley et al. 2008). Many small scincids are associated with leaf litter or the uppermost layers of soil, and Greer (1989) links this to body elongation and limb reduction. There are also small gymnophthalmids (e.g. Gymnophthalmus pleii, Turk et al. 2010) with a similar morphology and lifestyle.

A long tail coupled with an increase in presacral numbers enhances body flexibility especially if combined with shortened limbs. Elongated lizards that retain limbs tend to use a mixed locomotor pattern – lateral undulation when moving through vegetation or soft substrate and limbs when moving quickly over the ground (e.g. Eremiascincus, some Lerista, some Niveoscincus: Greer 1989; Greer et al. 1998; Melville and Swain 2000). Ground cover at Las Hoyas would have varied between the wetter and drier areas of the ecosystem, but fossil remains of fern fragments (pinnae and pinnulae), and both gymnosperm and angiosperm leaf types (Buscalioni and Fregenal-Martinez 2010) suggest there could have been a carpet of loose plant material (equivalent to modern leaf litter) in which Jucaraseps lived and moved. However, its slender body would also have been well adapted to sliding into crevices or under rocks to escape heat, rain, fire or predators – all of which are likely to have been a threat in the Las Hoyas ecosystem (Buscalioni and Fregenal-Martinez 2010). Lengthening the body may have had additional advantages. Within the scincid genus Eumeces, individuals of larger species (e.g. E. fasciatus) have fewer presacral vertebrae (26–27) than those of the smallest species (30–32, E. brevirostris, E. lynxe; Griffiths 1990). A similar pattern is found in other scincids (e.g. Chalcides, Scelotes, Lerista; Griffiths 1990) and it has been suggested that by increasing the relative volume of the abdomen, presacral elongation may compensate for the overall reduction in size, and thus clutch space, in small females.

On the basis of the gut contents, the most common Las Hoyas lizard, Meyasaurus, apparently fed in and around the water and this may explain its relative abundance at the locality (Evans and Barbadillo 1997, Evans and Bolet in press). The other Las Hoyas squamates are rarer and probably lived on drier ground, further from the water. Scandensia was a specialised climber (Evans and Barbadillo 1998; Bolet and Evans 2011), either on rocks or vegetation, whereas Hoyalacerta and the paramacellodid, like Jucaraseps, are interpreted as ground-living (Evans and Bolet in press). Jucaraseps had short jaws (c. 3–4 mm) with 8–10 relatively large teeth. Given its small size, its gape, and therefore prey size, would have been limited. Most of the adult insects and spiders recovered from Las Hoyas would have been too large for Jucaraseps to eat (and may even have posed a threat; Greer 1989), but small larvae and spiderlings would have been within its range.

Conclusions

Jucaraseps lies at the lower end of the modern lizard size range and demonstrates that very small-bodied lizards existed in the Early Cretaceous. Its slender, elongated body is reminiscent of many small living scincid lizards and some gymnophthalmids, and Jucaraseps is likely to have had a similar lifestyle. By comparison with living ecomorphs, its size, shape and large feet are suggestive of a secretive ground-living lizard that is likely to have hunted for small insects in and around loose vegetation but would itself have been a target for other small vertebrates and, possibly, large invertebrates. Jucaraseps thus increases our understanding of the Las Hoyas assemblage as a whole.

Phylogenetic analysis places Jucaraseps on the stem of a traditional monophyletic Scleroglossa (sensuEstes et al. 1988 and Conrad 2008; contraTownsend et al. 2004 and Vidal and Hedges 2005), which would be consistent with the frequency of body elongation/limb reduction within this group. However, further material (especially of the skull) may lead to a refinement of this position, as may more global molecular/morphological analyses of Squamata as a whole.

Acknowledgments

Acknowledgements.  AB was funded by FPI grant (BES-2009-026731) associated with the project CGL2008-06533-C03-01/BTE (Ministerio de Ciencia E Innovación, Spain). Funding for a 2-month stay in London in 2010 came from the EEBB programme of the same grant. SE thanks Dr Ángela Buscalioni and colleagues at the Universidad Autónoma de Madrid for the invitation to work on Las Hoyas material, funded through grants CGL2005-001121 BTE and CGL2009-11838 BTE (Ministerio de Ciencia E Innovación, Spain). We thank the Willi Hennig Society for the free access to TNT (Goloboff et al. 2008).

Editor. Kenneth Angielczyk

Ancillary