Friday, February 28, 2014

Sexual dimorphism in pinnipeds - comments on two new studies in the journal Evolution

Two newly published articles in the journal Evolution investigate the evolution of sexual dimorphism in pinnipeds (seals, sea lions, and walruses). One paper by Kruger et al. (2014) examines it largely from a modern biological perspective, while the other paper by Cullen et al. (2014) incorporates fossil data. We’ll start with Kruger et al., and then move on to Cullen et al.

But first, some introductory remarks. Sexual dimorphism, for the uninitiated, is the condition where males and females of a particular species are of different sizes, color pattern, or proportion (or, all of the above). Naturally this doesn’t apply to anatomical differences in sex organs, as that is something that characterizes practically all vertebrates. Humans are somewhat sexually dimorphic – generally males are taller and more robust than females, and are characterized by some subtle skeletal differences (wider pelves in females, for example). As compared to humans, gorillas are a bit more extremely dimorphic – the males are quite a bit larger, and sport large sagittal crests for jaw muscle attachment on their skulls. Extreme sexual dimorphism has often been linked with reproductive behavior – in particular, “harem” size. In pinnipeds, the males of the least sexually dimorphic species tend to mate with only one female, while males of the most extremely sexually dimorphic species tend to arrange and defend large harems (dozens to hundreds of females) and engage in male-male combat, such as the dramatic fighting commonly seen in elephant seals.

The paper by Kruger et al. (2014) analyzed data on modern pinnipeds including male and female body size, harem size, latitude of breeding, and the length of breeding and lactation, using several phylogenetic comparative methods including phylogenetic independent contrasts and phylogenetic confirmatory path analysis (neither of which I am very familiar with). Their analysis reconstructed ancestral pinnipeds as non-dimorphic and polar in distribution (e.g. Arctic). They further found that sexual dimorphism probably preceded increases in harem size (polgyny); in their words, sexual dimorphism originated first and facilitated polygyny, rather than being a consequence of it. They identify ice breeding as the ancestral reproductive behavior for pinnipeds, rather than aquatic or terrestrial breeding.

Hypothesized sequence of events in the evolution of sexual dimorphism and polygyny in pinnipeds, from Kruger et al. (2014).

There are a few problems with this hypothesis, and most of them revolve around the lack of fossils incorporated into the analysis. Of course, many of the data categories are unknown in fossils (e.g. length of breeding period). However, many fossils indicate that early pinnipeds were in fact sexually dimorphic (Berta 1994; also, see remarks upon Cullen et al., 2014 below). Secondly, simply because many modern pinnipeds breed on ice now doesn’t necessarily mean that they always have, and suggest that this trait is probably unreliable to work with. Tusked walruses, for example, were until only one or two million years ago nearly completely temperate and even subtropical in distribution – only with the evolution of Odobenus rosmarus have walruses been confined to the arctic. Furthermore, Pliocene fossils of Odobenus sp. from Japan indicate that even Odobenus was only cold temperate in distribution roughly 2-3 million years ago. The majority of the walrus fossil record not only reflects sexual dimorphism but also ~20 million years of temperate distribution – in other words, no ice. The early Miocene – the period of pinniped diversification – was substantially warmer than present with smaller icecaps; it makes me wonder if the abundance of modern ice breeding seals is due to recent (Pliocene-Holocene) diversifications into Arctic and Antarctic niches alongside rapidly expanding ice caps. Consider this: the basalmost phocids – monk and elephant seals – are all terrestrial breeding, and several members of the Phoca-Pusa-Halichoerus species complex are terrestrial breeders. It makes much more sense to me that Antarctic lobodontines and Arctic phocines evolved ice breeding habits since the Pliocene during rapid cooling and ice cap growth, rather than all pinnipeds evolving from an ice breeding ancestor and retaining that behavior for 27 million years (in fact, the complex distribution of ice and terrestrial breeding within phocines suggests that if anything, this behavior is really flexible at a macroevolutionary level). A further problem is that the most primitive known fossil pinnipeds, the enaliarctines, are all known from temperate ice-free latitudes. The moral of this story: fossils are important!

New specimen of Enaliarctos emlongi described by Cullen et al. (2014) and interpreted as an adult female.

The new study by Cullen et al. (2014), on the other hand, does incorporate fossil data. They report on a new skull of the early pinniped Enaliarctos emlongi from the Nye Mudstone of coastal Oregon, and performed a geometric morphometric analysis of sexual dimorphism in modern and fossil pinniped skulls. The skull was in fact initially tentatively referred to Enaliarctos emlongi by Annalisa Berta in her 1991 paper on Enaliarctos material from the Emlong collection. The skull is a bit squashed, but appears to be smaller and a bit more gracile than the adult male holotype. Berta (1991) originally considered this specimen to represent a juvenile, although Cullen et al. argue that the sutures are fully closed in the referred specimen, although it’s not immediately obvious from the photographs (a common problem with material in the Emlong collection is that it is consistently dark in color; generally it’s a good idea to coat specimens with ammonium chloride, as has been done by Barnes, Fordyce, Deméré, Berta, and others working on that collection). Sexually dimorphic features they highlight include a narrower rostrum in females, more strongly pronounced nuchal and sagittal crests in males, a proportionally wider palate in males (this probably goes hand in hand with a narrower/broader rostrum in general), and more widely flaring mastoid processes of the squamosal.

Sexually dimorphic features in modern and fossil pinnipeds: Enaliarctos emlongi (left) and Arctocephalus (right); first and third columns are females, second and fourth columns are males.

Morphometric analysis indicated strong sexual dimorphism both in skull size and shape for most pinnipeds, with the exception of numerous extant phocids (true seals - Pusa, Monachus, Erignathus, and Leptonychotes) – of the phocids analyzed, only the gray (Halichoerus) and hooded seals (Cystophora) were strongly dimorphic (Elephant seals, although extraordinarily dimorphic, were not part of the analyzed data set). All otariids, the walrus, and all fossil pinnipeds investigated were dimorphic. When plotted on a cladogram of modern pinnipeds, the ancestral character state is ambiguous owing to widespread lack of dimorphism in the true seals. However, when they incorporated fossil taxa – Enaliarctos and Desmatophoca only – it indicated that sexual dimorphism was primitive for all pinnipeds, and secondarily lost in phocids – and secondarily regained in gray and elephant seals.

Ancestral character state reconstruction of sexual dimorphism in pinnipeds; black=extreme dimorphism, white=little to no dimorphism. Top tree includes extant pinnipeds only, and bottom tree shows the influence of the inclusion of fossil taxa.
Overall, this publication by Cullen et al. is vastly superior in its treatment of sexual dimorphism in pinnipeds, particularly for its inclusion of fossils – which is unsurprising, since the authors are all paleontologists. I strongly suspect that Cullen et al. are correct, and this paper serves to reinforce earlier suggestions that enaliarctines were sexually dimorphic. However, there are a few minor nitpicky things that bear mentioning.

First, it is important to note that this study is not the first to propose that enaliarctines were sexually dimorphic. In fact, the first study to demonstrate sexual dimorphism was the reevaluation of Pteronarctos by Annalisa Berta (1994) – in that paper, she examined a large collection of material from the Emlong collection and concluded that Pteronarctos piersoni and Pacificotaria hadromma were prematurely named, and fall within the range of variation expected for a single species of pinniped (based on examination of variation in extant Callorhinus ursinus), and synonymized both species with Pteronarctos goedertae (regardless, later works by Barnes have been uncharitably dismissive of this hypothesis). Berta (1994) ascribed much of the variation between species to sexual dimorphism, and identified the holotypes of P. piersoni and P. hadromma as females (also dismissed by recent work by Barnes), in addition to figuring and describing additional female specimens of Pteronarctos goedertae from the Emlong collection. Curiously, this acknowledgement of sexual dimorphism in Pteronarctos was not mentioned or cited by Cullen et al. (2014), despite citing the paper. Sexual dimorphism in enaliarctines has also been suggested in various papers by Larry Barnes (1989, 1990, 1992, 2008), and demonstrated in the enaliarctine-like proto-walrus Proneotherium repenningi (Deméré and Berta 2001).

Secondly, the entire crux of this paper hinges upon the identification of the referred skull, USNM 314290 as being conspecific with Enaliarctos emlongi. I’ve never seen the specimen (it was on loan in Canada when I visited the USNM to examine their pinnipeds in 2012), but a few things struck me with the description. For starters, it was only compared with Enaliarctos emlongi, Enaliarctos barnesi, and Enaliarctos tedfordi. What about Pteronarctos? Pteronarctos is, after all, known from really low down in the Astoria Formation – around the same level as some Enaliarctos material reported by Berta (1991). No comparisons with Pinnarctidion are made, and most problematic, no comparisons with Enaliarctos mitchelli are made – Enaliarctos mitchelli is tiny, with a transversely narrow rostrum (consistent with USNM 314290) and known from the Nye Mudstone of Oregon (Berta 1991) in addition to Pyramid Hill in California. I could easily see this specimen representing another E. mitchelli specimen – but that possibility was not evaluated.

This study by Cullen et al. (2014) is certainly an excellent contribution, and a great starting point. Future analyses can (and, should) utilize other fossil pinnipeds for which males and females are known: Allodesmus gracilis/kernensis, Dusignathus seftoni, Imagotaria downsi, Neotherium mirum, Proneotherium repenningi, Pteronarctos “spp.”, Thalassoleon mexicanus, and Valenictus chulavistensis (some of these are known from male and female mandibles only (e.g. Neotherium). And of course, there’s also canines, postcrania, and that curious baculum.

A parting comment is necessary, and this should not be misconstrued as further criticism, but a word of caution for any study regarding the phylogenetic position of fossil pinnipeds: We currently do not have a robust up-to-date phylogeny for modern and fossil pinnipeds. It’s now been twenty years since Berta and Wyss (1994) published their seminal analysis of modern and fossil pinniped phylogeny, and nobody has taken up the challenge of adding more characters and taxa to that dataset (or, even putting together a new dataset of equivalent breadth). Most subsequent studies have incorporated taxa only from a single family (and this is something I am definitely guilty of). Although not stated by the authors, the phylogenetic position of Enaliarctos and Desmatophoca are probably from Berta and Wyss (1994). Morgan Churchill and I have talked about these issues at length, and we wouldn’t be surprised if 1) desmatophocids were more closely related to otarioids than phocids, 2) morphological evidence could be mustered to support an odobenid-otariid clade, and 3) enaliarctines might actually be more closely related to otarioids than to phocoids. What can be done about this? Cullen et al. (2014) rightfully point out that enaliarctines are greatly in need of a taxonomic enema (my paraphrasing): they are probably greatly oversplit, and a detailed comprehensive study of enaliarctine morphology is in order – this is especially true if Berta’s (1991) lumping of Pteronarctos goedertae, Pteronarctos piersoni, and Pacificotaria is any indication. In addition to a better treatment of enaliarctines, we need a larger analysis of pinniped phylogeny, with more taxa. Morgan and I started a such a project several years ago and presented it at SVP, and then decided it would be best to treat each family one at a time before doing anything comprehensive – but, more on that in the future.

References and further reading:

L. G. Barnes. 1989. A new enaliarctine pinniped from the Astoria Formation, Oregon, and a classification of the Otariidae (Mammalia: Carnivora). Contributions in Science 403:1-26

L. G. Barnes. 1990. A new Miocene enaliarctine pinniped of the genus Pteronarctos (Mammalia: Otariidae) from the Astoria Formation, Oregon. Contributions in Science 422:1-20

L. G. Barnes. 1992. A new genus and species of middle Miocene enaliarctine pinniped (Mammalia, Carnivora, Otariidae) from the Astoria Formation in Coastal Oregon. Contributions in Science 431:1-27

L. G. Barnes. 2008. Otarioidea. In C. M. Janis, G. F. Gunnell, M. D. Uhen (eds.), Evolution of Tertiary Mammals of North America 2:523-541

A. Berta. 1991. New Enaliarctos* (Pinnipedimorpha) from the Oligocene and Miocene of Oregon and the role of "enaliarctids" in pinniped phylogeny. Smithsonian Contributions to Paleobiology 69:1-33

A. Berta. 1994. New specimens of the pinnipediform Pteronarctos from the Miocene of Oregon. Smithsonian Contributions to Paleobiology 78:1-30

A. Berta and A. R. Wyss. 1994. Pinniped phylogeny; pp. 33–56 in A. Berta and T. A. Deméré (eds.), Contributions in Marine Mammal Paleontology Honoring Frank C. Whitmore, Jr. Proceedings of the San Diego Society of Natural History 29.

T.M. Cullen, D. Fraser, N. Rybczynski, and C. Schroder-Adams. 2014. Early evolution of sexual dimorphism and polygyny in Pinnipedia. Evolution DOI: 10.1111/evo.12360

T.A. Deméré and A. Berta. 2001. A reevaluation of Proneotherium repenningi from the
Miocene Astoria Formation of Oregon and its position as a basal odobenid
(Pinnipedia: Mammalia). Journal of Vertebrate Paleontology 21: 279–310.

N. Kohno, L. G. Barnes, and K. Hirota. 1995. Miocene fossil pinnipeds of the genera Prototaria and Neotherium (Carnivora; Otariidae; Imagotariinae) in the North Pacific Ocean: Evolution, relationships and distribution. The Island Arc 3:285-308

O. Kruger, J.B. Wolf, R.M. Jonker, J.I. Hoffman, and F. Trillmich. 2014. Disentangling the contribution of sexual selection and ecology to the evolution of size dimorphism in pinnipeds. Evolution DOI: 10.1111/evo.12370

No comments: