Triadobatrachus massinoti ( PIVETEAU , 1936a)

Gardner, James D., 2016, The Fossil Record Of Tadpoles, Fossil Imprint 72 (1 - 2), pp. 17-44 : 29-31

Triadobatrachus massinoti ( PIVETEAU, 1936a) from the Early Triassic of Madagascar

The most iconic of the purported tadpole fossils is the holotype and only known specimen of the basal salientian Triadobatrachus massinoti ; note the original generic name “ Protobatrachus ” was amended by Kuhn (1962). Triadobatrachus is from the Early Triassic of Madagascar and it is the geologically oldest lissamphibian fossil. The holotype is challenging to interpret because (1) it consists of an incomplete skeleton, missing the anterior end of the snout and most of the hands and feet, and is of moderate size (preserved midline length of about 9 cm), (2) the skeleton is preserved as natural molds or impressions exposed in dorsal and ventral aspect on two halves of a split nodule ( Text-fig. 2 a–d), and (3) it exhibits a mixture of primitive and derived characters. Although its relevance to the evolutionary history of anurans once was a source of debate (e.g., see historical summaries and discussions by Hecht 1962, Griffiths 1963, Estes and Reig 1973, Roček and Rage 2000b, Rage 2006, and references therein), Triadobatrachus now is widely regarded as the sister to all other salientians (e.g., Ford and Cannatella 1993, Báez and Basso 1996, Evans and Borsuk-Białynicka 1998, Rage 2006, Anderson 2007, Marjanović and Laurin 2014, Ascarrunz et al. 2016). The ontogenetic age of the holotype of Triadobatrachus also has been debated.

In his treatments of Triadobatrachus (as “ Protobatrachus ”), Pivetaeau (1936a, b, 1937, 1955) regarded the holotype skeleton as being from an adult animal. That remained the conventional wisdom until challenged by Griffiths (1956, 1963), who proposed that the holotype was “a tadpole in the later stages of metamorphosis” ( Griffiths 1956: 343). In support of that larval interpretation, Griffiths (1963: 275–276) listed the following postcranial and cranial features: tail present, consisting of small vertebrae; no anuran-style sacrum (i.e., lacks diapophyses that are laterally directed, expanded, and incorporate fused sacral ribs); iliac shaft only moderately elongate anteriorly; radius and ulna in forelimbs and tibia and fibula in hindlimbs unfused; femur considerably longer than tibia and fibula; upper jaw elements absent; lower jaw bones weakly sutured; and parietal recess present along midline of the paired frontoparietals. In the latter paper, Griffiths (1963: 276–277) also argued for an aquatic origin for salientians and, under that scenario, softened his stance on the ontogenetic age of the holotype of T. massinoti by admitting it could be either a larva or an adult. Orton (1957: 80) quickly embraced Griffiths’ (1956) larval interpretation, calling it “an illuminating idea which can account for virtually all of the peculiarities of the specimen”. Hecht (1962: 43) disagreed by saying “the interpretation of Protobatrachus as a tadpole, is considered unlikely due to the well-ossified nature of the fossil.” Perplexingly, but perhaps understandable given his staunch view that Triadobatrachus had nothing to do with the evolution of anurans, Tihen (1965: 309–310) dismissed the debate about the ontogenetic age of the holotype by stating “Whether it is a late stage larval or metamorphic individual … or an adult … is not of major import”. Despite their detailed examination and reconsideration of the holotype of T. massinoti, Estes and Reig (1973) could offer only vague comments about its ontogenetic age, such as “[it was] probably a young stage” (p. 42) and “It may represent a young stage but is probably not far from adult” (p. 49). Their statement that its “interpretation as adult or larva … has not been satisfactorily resolved” ( Estes and Reig 1973: 49) was a fair assessment of the situation at the time and remained so for over a decade to come.

Rage and Roček (1986: 257, 1989: 13) argued that the holotype of T. massinoti was from a postmetamorphic individual because it has fully developed dermal skull bones, a columella of adult size, ossified carpal and tarsal elements, and presumably both an ossified parahyoid and thyrohyals; note the presence of those hyoid bones recently was corroborated by a micro-CT study of the holotype by Ascarrunz et al. (2016: fig. 4). Rage and Roček (1986: 257, 1989: 13) also disputed the cranial features listed by Griffiths (1956, 1963) by (1) dismissing the weakly sutured lower jaw bones as equivocal evidence for Triadobatrachus being a larva, (2) identifying the posterior ends of both maxillae, thereby showing that at least some upper jaw bones were present ( Rage and Roček 1989: fig. 2; see also Ascarrunz et al. 2016: fig. 4), and (3) re-interpreting the “parietal recess” of Griffiths (1963: 276) simply as part of the indistinct, dorsal ornament on the frontoparietals. As for the suite of postcranial features listed by Griffiths (1956, 1963), Rage and Roček (1989: 13) countered that those only need be considered larval features if one assumes – as Griffiths implicitly appeared to have done – that Triadobatrachus was a primitive anuran. In that interpretation, had the holotype individual lived longer it would have undergone metamorphosis, during which its “larval” features would have transformed into features characteristic of adult or fully metamorphosed anurans. For example, the tail would be lost as the caudal vertebrae fused to form a urostyle, the iliac shaft would grow farther anteriorly, and the radius and ulna would fuse to form a composite radioulna. The alternate interpretation, favored by Rage and Roček (1986, 1989) and tacitly accepted since, is that the postcranial features listed by Griffiths (1956, 1963) are not larval features, but instead are plesiomorphies indicative of Triadobatrachus being a basal salientian. Rage and Roček (1989: 13) concluded that the holotype of T. massinoti was a postmetamorphic individual, although they noted that the unossified epiphyses of its long bones indicated it likely died before reaching full maturity. The recent micro-CT study by Ascarrunz et al. (2016) revealed no features at odds with the interpretation that the holotype skeleton T. massinoti is from a postmetamorphic individual.

Figures

Text-fig. 2. Examples of putative tadpole body fossils. Photographs are at different magnifications; see corresponding scale bars. a–d – Triadobatrachus massinoti (PIVETEAU, 1936a) (Salientia); basal Triassic, Betsieka, northern Madagascar. Two copies of original holotype specimen, MNHN (Museum National d’Historie Naturelle, Paris, France) F.MAE.126, a split nodule exposing dorsal and ventral surfaces of a nearly complete skeleton preserved as a natural mold; all copies are oriented with head towards top of figure: first pair is replica (epoxy cast) of original fossil showing bones as impressions, in dorsal (a) and ventral (b) views; second pair is negative cast (also epoxy) showing bones in positive relief, in dorsal (c) and ventral (d) views. Interpreted as a metamorphic tadpole by Griffiths (1956, 1963), the unique holotype of T. massinoti now is regarded as being from a young post metamorphic individual (e.g., Rage and Roček 1986, 1989). See Rage and Roček (1989), Roček and Rage (2000), and Ascarrunz et al. (2016) for labelled interpretive drawings and other images. Replicas provided by MNHN and deposited in collections of the Royal Tyrrell Museum of Palaeontology (Drumheller, Canada) as TMP 2008.003.0009 (first pair) and TMP 2008.003.0008 (second pair). Photographs courtesy of G. Housego. e – Palaeospondylus gunni TRAQUAIR, 1890, NHMUK (Natural History Museum, London, England) PV P61846; Middle Devonian,Achanarras Quarry, northern Scotland. Complete skeleton preserved in typical pose and orientation, in this case with body curved, head (directed towards top of figure) and body exposed in dorsal view, and tail twisted such that its left side is exposed and the deeper, ventral portion of tail faces towards top of figure. Note continuous sequence of ring-like vertebrae along central axis of body and tail, and elongate and asymmetric tail fin supported internally by well-developed, bifurcating radials. Suggested to be a tadpole by Dawson (1893) and Jarvik (1980), Palaeospondylus is widely regarded as a larval fish, although its affinities remain controversial (e.g., Johanson et al. 2010 and references therein). See Moy-Thomas (1940) and Forey and Gardiner (1981) for labelled interpretive drawings of other specimens. Photograph courtesy of Z. Johanson. f, g – The Daohugou “tadpole”, CAGS (Chinese Academy of Geological Sciences, Beijing, China) IG01-705; late Middle or early Late Jurassic, Daohugou, Inner Mongolia. Entire slab (f) and close up (g) of fossil; differences in matrix color are because photographs were taken using different lighting. This soft-bodied fossil lacks any indication of internal skeletal tissue. Originally interpreted by Yuan et al. (2004) as a tadpole body fossil (in dorsal or ventral view, with anterior towards top of figure), consisting of a head and body with prominent hind limbs and a thin, straight tail. Reinterpreted by Huang (2013: see his interpretive line drawing on p. 143) as an incomplete, cicadomorph insect (also in dorsal or ventral view, but with anterior towards bottom of figure), consisting of anteriorly projecting rostrum (= “tail” of tadpole), pro- and mid-thorax portion of body (= “head and body” of tadpole), and forelegs (= “hind legs” of tadpole). Photographs courtesy of Wang X. R and Yuan C. X., and their inclusion here arranged by Wang Y.