Report

Viviparity and K-Selected Life History in a Mesozoic Marine Plesiosaur (Reptilia, Sauropterygia)

See allHide authors and affiliations

Science  12 Aug 2011:
Vol. 333, Issue 6044, pp. 870-873
DOI: 10.1126/science.1205689

Abstract

Viviparity is known in several clades of Mesozoic aquatic reptiles, but evidence for it is lacking in the Plesiosauria. Here, we report a Late Cretaceous plesiosaur fossil consisting of a fetus preserved within an adult of the same taxon. We interpret this occurrence as a gravid female and unborn young and hence as definitive evidence for plesiosaur viviparity. Quantitative analysis indicates that plesiosaurs gave birth to large, probably single progeny. The combination of viviparity, large offspring size, and small brood number differs markedly from the pattern seen in other marine reptiles but does resemble the K-selected strategy of all extant marine mammals and a few extant lizards. Plesiosaurs may have shared other life history traits with these clades, such as sociality and maternal care.

Viviparity, or birthing live young, is common among reptiles, having evolved over 80 times among extant clades (1). Live birth has also been documented in several groups of Mesozoic aquatic reptiles, including ichthyosaurs (2), mosasauroids (3), choristoderans (4), and nothosaur-grade sauropterygians (5). However, to date no evidence exists for viviparity in the sauropterygian clade Plesiosauria, despite an excellent fossil record and a collection history spanning almost 200 years. This lack of evidence for viviparity is enigmatic given that plesiosaurs were large animals whose bodies lacked a firm connection between the appendicular and axial skeleton, a condition impeding the terrestrial nesting required in an oviparous amniote (57). Viviparity was documented recently in basal, amphibious eosauropterygian clades (5), even though these clades have shorter and less intense collecting histories. However, the single published claim of plesiosaur embryonic material (8) was shown to be misinterpreted shrimp burrows (9), and until now no pregnant plesiosaur fossil has been reported. In this paper, we present fossil evidence of a gravid plesiosaur; furthermore, analysis of the fetus demonstrates that plesiosaurs exhibited a reproductive pattern unique among marine reptiles.

LACM 129639 (Natural History Museum of Los Angeles County) was discovered in 1987 by Charles Bonner on the Bonner Ranch in Logan County (Kansas, USA) at the base of the Pierre Shale [Sharon Springs Member, Campanian (10)]. The fossil consists of the largely articulated skeletal remains of two plesiosaur individuals, an adult and a juvenile displaying features at very early stages of development (Fig. 1). The adult is a large, short-necked plesiosaur referable to Polycotylus latippinus Cope, complete save for the head and anterior 16 cervical vertebrae. Measurement of the adult vertebral column allows a confident length estimate of 470 cm (11). The juvenile consists of a mass of poorly ossified and largely disarticulated bones spilled from the body cavity of the adult, intermingled with phalanges of the adult right fore-paddle (Fig. 2).

Fig. 1

(A) Photograph and (B) interpretive drawing of LACM 129639, as mounted. Adult elements are light brown, embryonic material is dark brown, and reconstructed bones are white. lc indicates left coracoid; lf, left femur; lh, left humerus; li, left ischium; lp, left pubis; rc, right coracoid; rf, right femur; rh, right humerus; ri, right ischium; and rp, right pubis. [Photo credit: R. Cripps]

Fig. 2

Maps of the LACM 129639 jackets containing fetal material, depicted before mounting. Fetal bones are dark brown; adult bones are light brown. c, centra; f, femur; g, gastralia; lh, left humerus; li, left ischium; lp, left pubis; rep, right epipodial; rh, right humerus; ri, right ischium; rp, right pubis; rs, right scapula; and v, vertebrae.

Some ambiguity exists concerning the original orientation of the fossil because no quarry map was kept of the excavation. However, a confident reassembly was possible on the basis of the position of elements within each jacket, with subsequent evaluation and validation by the original excavator, Charles Bonner (Fig. 1) (11). Of the 12 jackets, three contain a mixture of adult and juvenile elements (Fig. 2). The elements of the juvenile pelvis are near life position, and the juvenile right pubis and left ischium are adhered to the visceral (internal) surface of the adult right coracoid by gypsum and matrix. The presence of phalanges from the right forelimb in the two jackets containing the balance of the juvenile material indicates that their original position was in the vicinity of the right forelimb. The position of this limb relative to the thorax was established on the basis of the location of the pectoral girdle and an associated paddle bone in the thorax jacket. The rest of the mount was laid out in approximate articulation, and the resulting mount was certified as accurate by the original excavator (11).

Multiple lines of evidence indicate that the juvenile specimen is an in situ fetus. First, the partially articulated fetal pelvis adhered to the visceral surface of the right adult coracoid establishes that the juvenile was both articulated and within the adult body cavity at the time of deposition, before burial (Fig. 2). Second, both specimens are referable to the same species, P. latippinus, on the basis of their possession of that taxon’s distinctive humerus. This humerus has a shaft that is narrow and strongly curved relative to other plesiosaurs and also possesses four discrete facets on its distal edge for articulation with the radius, ulna, and two additional ossifications (1113). Third, the juvenile skeleton is poorly ossified and displays numerous embryonic features, and its overall state of development is comparable to known prepartum eosauropterygian embryos (14). Fourth, the juvenile shows no indication of having been consumed by the adult: There is no maceration of cartilaginous surfaces by stomach acid as seen in other consumed plesiosaurs (10, 15), and no gastroliths (stomach stones) or other evidence of a gastric mass (16). The above taphonomic, taxonomic, and ontogenetic evidence establishes that the adult was a gravid female containing a fetus.

The fetal skeleton belongs to a single individual considering the lack of duplicate elements, the presence of paired elements of equal size, and the partial articulation of the pelvis. The fetal remains comprise a skull fragment; 20 vertebral centra, many with associated but unfused neural arches; numerous ribs and gastralia (belly ribs); both pubes and ischia; one ilium; both scapulae; both humeri, one with associated paddle bones; and the left femur (Fig. 2). All vertebral centra have deep notochordal pits, and perichondral ossification is restricted to a ring encircling the center third of each centrum. Identifiable vertebral centra are all thoracic except for one cervical, as expected of a reptilian embryo in which thoracic and posterior cervical centra ossify earlier than other vertebrae (17). The neural arches are very immature, consisting of unfused bilateral arch elements with open neurocentral sutures, rudimentary zygapophyses, and a complete lack of ossified neural spines. Lack of midline fusion of the neural arch is characteristic of eosauropterygian embryos [e.g., Lariosaurus (14) and Keichousaurus (18)]. The belly ribs are well developed, also in concordance with the precocial ossification of these in eosauropterygians (18, 19). Both humeri are short, with poorly ossified proximal ends but wide distal ends. These and the femur are small relative to the rest of the body (11), a condition consistent with the small humerus and femur proportions in prenatal eosauropterygian embryos [e.g., Lariosaurus (17), Neusticosaurus (19), and Keichosaurus (20)]. The skull fragment, consisting of a rounded and poorly defined left quadrate encased in a thin sheet of squamosal, demonstrates that the skull was proportionally large. It was poorly ossified, however, and meaningful quantitative measurements are impossible.

The fetus is notably large given its ontogenetic stage. By considering axial-propodial proportions and their known scaling relationships in eosauropterygian embryos (11), we estimate that the fetus was at most two-thirds mature. By using the preserved vertebral centra, which generally scale isometrically with postnatal body size (21), we calculate a conservative minimum body length estimate of 1.5 m at time of death (11) (32% of the 470-cm maternal length). Length at full term would have been significantly longer, given the poor ossification of the fetal skeleton. A minimum estimate of fetus length at birth is 35% of maternal length, with an upper bound at the 50% estimated length of a young juvenile of the short-necked plesiosaur Leptocleidus (22) (Fig. 3).

Fig. 3

Reconstructions of female P. latippinus and newborn young. Gastralia were present in both animals but have been omitted for clarity.

A tabulation of hatchling length relative to maternal length in eosauropterygians, mosasauroids, choristoderans, and ichthyosaurs (Table 1) shows that parturition length in these clades generally conforms to lengths predicted from empirical analyses of maternal-neonate scaling in extant reptiles derived by Andrews (23) and is always less than 30%. This pattern indicates an r-selected reproductive pattern in which females gave birth to several relatively small young, as documented in a number of gravid fossils from these clades (25). Calculation of the lizard regression equation of Andrews (23) for an adult length of 470 cm in LACM 129639 yields a predicted neonatal length of 103 cm, or 22% of maternal length, for Polycotylus. This length is 10% shorter than the minimum LACM 129639 fetal length at time of death and 15% shorter than the estimated length at full term. A 15% increase in neonate length embodies an ~50% increase in mass; therefore Polycotylus young at birth were at least one and a half times as massive as those in other, more typical terrestrial and marine reptile clades. The presence of a large, single fetus in LACM 129639 therefore documents a K-selected reproductive strategy unique among marine reptiles.

Table 1

Brood size and parturition length, stated as percentage of maternal length, for viviparous amniote taxa. Length is body length or snout vent length, depending on the study. Asterisks indicate extinct clades of Mesozoic marine reptiles. Cetacean taxa are extant marine mammals, and Egernia species are extant terrestrial skinks.

View this table:

The closest extant ecological analogs to short-neck plesiosaurs such as Polycotylus are odontocete cetaceans. Despite radically different locomotor adaptations, both clades are secondarily aquatic tetrapods with similar body size ranges and feeding morphologies. A survey of odontocetes with body lengths comparable to that of Polycotylus shows they are also K-selected, giving birth to large, single progeny (24) (Table 1). Odontocetes are highly social and engage in maternal care, and one may argue that plesiosaurs displayed similar behaviors on the basis of ecological and reproductive similarity. However, this comparison is complicated because all whales, and all marine mammals, invariably give birth to single, large, and precocious young (25). A K-selected reproductive strategy may be an adaptive constraint operating on all marine mammals (26) that apparently did not apply to marine reptiles (Table 1), except for plesiosaurs.

Among modern reptiles, the plesiosaur-like trait combination of viviparity, small brood size, and large birth size is rare, but it does occur in the scincid Egernia species group. The members of this clade with the largest parturition sizes also exhibit mammal-like social behaviors, including stable, kin-related group structures and parental care (2729) (Table 1). Although this example is from a terrestrial clade, it demonstrates another instance where K-selected reproduction is associated with social behavior and maternal care. This example is pertinent because it involves squamate reptiles, which are arguably the closest living relatives of plesiosaurs (30). Post-hatching parental care is also rare in squamates generally (31).

Because both cetaceans and Egernia-group lizards are highly social and engage in substantial maternal care, plesiosaurs may have behaved similarly. We hypothesize that large plesiosaur fetus size may indicate that plesiosaurs lived in gregarious social groups and engaged in parental care. More evidence is needed to test this hypothesis; however, it is certain that plesiosaur life history differed markedly from that of other Mesozoic marine reptiles and resembled that of highly social marine mammals and scincid lizards.

Supporting Online Material

www.sciencemag.org/cgi/content/full/333/6044/870/DC1

Materials and Methods

Fig. S1

Tables S1 and S2

References (3234)

References and Notes

  1. Materials and methods are available as supporting material on Science Online.
  2. Acknowledgments: The fossil described here (LACM 129639) is reposited in the vertebrate paleontology collections of the Los Angeles County Museum and is currently on display in the public gallery. Metric data from the specimen is listed in tables S1 and S2. We thank Phil Fraley Productions for the skilled preparation and mounting of the fossil that made this study possible. B. Kear provided access to unpublished Leptocleidus data. Thanks to R. Cripps for the mount photograph in Fig. 1 and to D. Trankina for Fig. 1 line art. Many thanks to S. Abramowicz for the life reconstruction art, Fig. 3, and work on other figures. J. Long provided valuable feedback on an early version of this manuscript. This material is based on work supported by the NSF under award no. EPS-1003907.
View Abstract

Navigate This Article