In the 21st century, Antarctica is an icy landscape, largely known for its harrowing survival stories of early adventurers, charming birds, and the harsh winters. It's not until Antarctica permanently separated from South America, forming the circum-Antarctic current, that it reached the glaciated state that persisted to today (e.g., Kennett, 1977). But for much of geologic time, Antarctica was in fact much more habitable, at least for some parts of the year, with extensive forests and a much more diverse set of tetrapods. And that includes temnospondyls!
The cold snap this weekend (driven by Winter Storm Harvey and the polar vortex), inspired me (sitting in my house without functional internet but with functional heat) to write this week’s blog post about temnospondyls from the South Pole. Harvey originated from a low-pressure system off the coast of the Pacific Northwest, which is particularly apropos for this week’s blog because much of the recent work on Antarctic temnospondyls has been done by paleontologist Chris Sidor at the University of Washington / Burke Museum over in Seattle.
Antarctica in the Mesozoic
Overview of study
The first temnospondyl from Antarctica was collected from Graphite Peak (point 'k' on the map on the right) by Peter Barrett in 1967. The specimen, comprising a jaw fragment, comes from the lower Fremouw Formation (early or middle Triassic in age), and was subsequently described in a few papers (Barrett et al., 1968; Cosgriff, 1971) but not fully identified until 1974, when it was referred to the new brachyopoid Austrobrachyops jensensi by Colbert & Cosgriff. The holotype of this taxon is a partial pterygoid, the validity and identity of which has since been called into question. In the same paper, the authors erected a second taxon, Cryobatrachus kitchingi, for a partial skull roof that was identified as a lydekkerinid. Subsequent evaluation has also called the validity of this taxon into question. Additional material was referred to both of these taxa by Cosgriff & Hammer (1984), who additionally identified some material of an indeterminate rhytidosteid. Material of large temnospondyls from the upper Fremouw was described by Hammer (1990) but left relatively unresolved taxonomically.
A hiatus in the study of polar temnos then ensued until new work on temnospondyl material from the upper Fremouw (middle Triassic). Sidor et al. (2007) described a partial skull referred to the mastodonsaurid Parotosuchus, a genus known to extend as far north as Russia. One of the specimens described by Hammer (1990) was then reappraised by Sidor et al. (2008), who erected a new taxon, Kryostega collinsoni. Most recently, another large temnospondyl from the same area as Kryostega, Antarctosuchus polyodon, was described (Sidor et al., 2014b).
Brachyopoidea: Austrobrachyops jenseni
By date of collection of materials, Austrobrachyops was the first temnospondyl to be discovered in Antarctica. It is represented in part by the relatively small jaw fragment (AMNH [American Museum of Natural History] 9301) featured above. The original describers of the specimen did not attempt to identify it other than as a ‘labyrinthodont amphibian.’ It was revisited again by Colbert (1971) but not identified until the work of Colbert & Cosgriff (1974), when it was referred to A. jenseni. Although it was the first specimen of any temnospondyl from Antarctica, the jaw fragment is not the holotype of the species – an extremely fragmentary pterygoid (from the roof of the mouth) from a different locality is. As a result, it is not actually clear that the referred jaw and the holotype belong to the same taxon. Unsurprisingly, not a lot could be made in the way of interpretations based on such fragmentary material, although the authors placed it within the Brachyopoidea, a clade of short-faced temnospondyls that were most abundant in the southern hemisphere and that outlived all other clades (made it to the Cretaceous in Australia).
Subsequently collected material was referred to this species by Cosgriff & Hammer (1984). This consisted of three fragments from the skull region (not necessarily the same skull) and one partial clavicle (from the pectoral girdle). Again, these specimens were not collected from the same locality as the previously described material, and there is no overlap in the preserved regions of the skeleton between any specimens. Cosgriff & Hammer’s referral was based largely on the most parsimonious assignment to a superfamily (the Brachyopoidea) and then by extension of ecological exclusion (no room for many species of a similar ecology) to the only known brachyopoid (Austrobrachyops). They also suggested that it could actually be synonymous with a previously existing brachyopoid (Blinasaurus [itself now split among various other brachyopoid genera] from Australia). It may not even matter; in a systematic review of the brachyopoids by Warren & Marsicano (2000), the taxon was designated as a nomen dubium (‘doubtful name’), essentially a status of invalidity, because the holotype could not be clearly proven to be a temnospondyl of any flavor. However, that first jaw fragment is still considered to be a brachyopoid but of uncertain relationships (Warren & Marsicano, 2000).
Lydekkerinidae: Cryobatrachus kitchingi
Cryobatrachus was described based on a fairly complete skull, far more substantial than Austrobrachyops. Unfortunately, the skull is very thin and was originally exposed in ventral profile, so only the underside of the skull could be explored. This is, of course, not normally how we study skulls, and the sutural patterns on the ventral surface are not always the same as those of the dorsal surface (angled contacts, processes, etc.).
A latex mold was taken from the negative impression of another specimen (far more fragmentary) in order to examine the ornamentation. This ornamentation is fairly comparable to that of lydekkerinids, found in South Africa and Australia (amongst other regions), which was one line of evidence for associating Cryobatrachus with lydekkerinids. This specimen is not from the same locality as the holotype, and only qualitative comparisons can be made with the holotype with respect to important features of the skull. Since then, doubt has been cast on the validity and identification of Cryobatrachus; Schoch & Milner (2000) declared it to be a nomen dubium, while Sidor et al. (2007) mentioned that the holotype may actually be a juvenile mastodonsaurid. This jives with major disparities in the cranial anatomy between reconstructions of Cryobatrachus and Lydekkerina (the namesake for the lydekkerinids). As far as I know, the specimen has not been formally revisited since.
Colbert & Cosgriff also referred a bunch of specimens to Cryobatrachus, including a partial jaw, clavicles, and a few limbs. However, the only line of rationale for assigning them to this genus is that they were small temnospondyl elements ornamented in a similar fashion to the aforementioned latex peel, which in turn is of questionable relation to / with the holotype of Cryobatrachus. The ornamentation consists mostly of small, round pits, which are not particularly informative as far as temnospondyl ornamentation goes. It should go without saying (though often does not), that size is not a very useful feature for differentiating extinct species. Small individuals could not only be small-bodied adults of a separate species, but they could also be juveniles of either existing or unknown large-bodied species. I could go on about this for a while, but I'll cover / rant about size as a taxonomic feature in temnospondyls in some future post. The main point is that if there is a large temnospondyl known, small ones should not automatically be assigned to a new species because anatomical differences could just reflect differences in age. The holotype of Cryobatrachus is a particularly small skull, only about 4.5 cm in total length, so it remains a very viable possibility that it is a pretty young individual. It seems quite probable that there is a lydekkerinid in Antarctica but that it is not presently known from particularly diagnostic material.
Cosgriff & Hammer (1984) made brief mention of some fragments (partial interclavicle, two indeterminate bits) that they assigned to the Rhytidosteidae based on the pattern of ornamentation (an instance in which it is perhaps useful!). Despite what I said above about ornamentation of Cryobatrachus, the ornamentation patterns on the pectoral girdle (and the proportions of the element) are typically pretty useful for identifying / distinguishing temnospondyls - many do not even have ornamentation on the interclavicle, for example. That being said, as far as I know, nobody has really revisited this material to evaluate its taxonomic affinities because not too much was (or can be) said about it.
Hammer (1990) also noted a possible occurrence of a benthosuchid based on a jaw fragment and what is now a somewhat dated classification of mandibular morphology in temnospondyls (Jupp & Warren, 1986). As far as I know, this has not been re-examined, and remains noted as Benthosuchidae indet. by Sidor et al. (2014b).
Stereospondyls indet.: Kryostega collinsoni
Kryostega is the second temnospondyl identified from the Upper Fremouw (Sidor et al., 2008) based on a snout with anterior (front) fragments of the jaw that was originally collected and identified by Hammer (1990) as a possible mastodonsaurid (referred to as a capitosaurid in the original paper). Although it is pretty fragmentary like the above snout of Parotosuchus, it does preserve sufficiently diagnostic features to be identified as a new taxon, such as greatly enlarged teeth along the edge of the choana (Sidor et al., 2008).
Capitosauroidea: Antarctosuchus polyodon
Antarctosuchus is the most recent temnospondyl to be named from Antarctica and also comes from the Upper Fremouw (Sidor et al., 2014b). Being represented by a large (>62 cm in length) and well-preserved skull, it was readily identifiable as a distinct capitosauroid, a diverse group of large temnospondyls with relatively long and broad snouts and posteriorly positioned eye sockets. Its overall completeness makes it arguably the most informative Antarctic temnospondyl to date, considering questionable validity of some previous reports and the fragmentary nature of some other specimens. Capitosauroids have a fairly broad geographic range across the southern hemisphere in what is known as Gondwana (e.g., Australia, India, South Africa, South America, etc.), which was expanded to include Antarctica through this study. Coincidentally, an analysis of Antarctosuchus to place it within the group positioned it close to an Indian capitosauroid, Paracyclotosaurus crookshanki. As noted by the authors, a lot more work to resolve outdated taxonomy of existing taxa (largely at the genus-level) is important for more fully understanding biogeography of Triassic temnospondyls and for interpretations of endemism vs. cosmopolitanism.
What it all means
So, what are temnospondyls doing at the south pole anyway? Antarctica may not have been directly over the South Pole, nor was it covered in thick ice sheets, but it was still pretty close. Therefore, the presence of an ectotherm (what we typically call “cold-blooded,” an animal that can only regulate its body temperature via external sources) suggests that the climate was sufficiently favourable for at least some part of the year (i.e. not iced over) for temnospondyls to hang out that far south. Because of the arrangement of the continents, Antarctica was not isolated, accounting for similarities to tetrapods from other continents as noted below and offering an accessible means for tetrapods to move across latitudes to follow favourable environmental conditions that might have changed seasonally. As they often are today, amphibians of the past were probably quite sensitive to environmental conditions and can be used as indirect proxies for paleoenvironmental conditions. The fact that temnospondyls are amongst the most abundant of Fremouw fossils suggests that they were doing pretty well for themselves down south during the Triassic.
An African connection
We don't think much of relationships between Africa and Antarctica today because there aren't many as far as animals go - penguins and elephants aren't exactly hanging out by the same watering hole. But because Africa and Antarctica were directly connected during the Triassic, their fossil record exhibits a far greater degree of overlap than we see in their modern day constituents. The lower Fremouw has been correlated with the Lystrosaurus AZ (assemblage zone) in South Africa, while the upper Fremouw has been correlated with the Cynognathus AZ , based on similarities between the tetrapod assemblages (e.g., Sidor et al., 2014a and references therein). The South African region has been far more extensively explored and studied than Antarctica, but Antarctic data points provide important insights into high latitude paleoenvironments and distributions of tetrapods across Pangea.
Foreshadowing of a last stand
It may sound counterintuitive, but hanging out in cooler climes may actually have allowed temnospondyls (excluding modern amphibians) to survive all the way to about 120 million years ago in the Cretaceous. Temnospondyls are quite diverse and widely distributed from the Carboniferous through the Triassic but get hit hard thereafter. There are few post-Triassic temnospondyls (none in North America, for example), and a number of them occur in Australia, which would have been situated at a similarly high latitude to its current position. A cooler (but probably not icy) climate may have excluded more typical warm-climate tetrapods (most prominently crocodilians [though not necessarily of modern flavours]) that otherwise outcompeted / predated on temnospondyls in other geographic regions, allowing temnospondyls to persist for longer in the southern latitudes (Warren et al., 1997). This may not be the case for all post-Triassic temnospondyls, as they do occur with crocodilians of some variety in others regions, such as Russia and Mongolia, which might reflect niche partitioning between these aquatic tetrapods.
Next week: It’s supposed to be cold next weekend too, so we’ll cover temnospondyls from the other polar region.
About the blog
A blog on all things temnospondyl written by someone who spends too much time thinking about them. Covers all aspects of temnospondyl paleobiology and ongoing research (not just mine).