impact factor Coming 2019

Frontiers journals are at the top of citation and impact metrics

Opinion ARTICLE Provisionally accepted The full-text will be published soon. Notify me

Front. Earth Sci. | doi: 10.3389/feart.2019.00013

Functional categories and ontogenetic origin of temporal skull openings in amniotes

  • 1University of Tübingen, Germany
  • 2Seckenberg Center for human evolution and Palaeoenvironment, Germany

The rise of phylogenetic systematics (Hennig, 1950) uncovered many natural groups of amniotes with Synapsida – characterized by one temporal opening, and Reptilia (Modesto and Anderson, 2004) – which contains ancestral, typically anapsid groups, without temporal openings, and Diapsida, with two temporal openings and diverse secondary modifications. The ancestrally anapsid parareptiles partly show, with certain ontogenetic, inter- and intraspecific variation, one opening or marginal excavation in their temporal region (Cisneros et al., 2004;Tsuji and Müller, 2009;Macdougall and Reisz, 2014), although evidence from other characters clearly separates them from Synapsida (Müller, 2004;Tsuji et al., 2012;Brocklehurst et al., 2018). Furthermore, the phylogenetic position of some early “classical” synapsid groups (e.g.,varanopsids) is rather controversial, as they appear within Reptilia in modern phylogenetic analyses (Angielczyck and Kammerer, 2018;Laurin and Piñeiro, 2018). Compared to historical classifications (Osborn, 1903;Williston, 1917;Goodrich, 1930), there is a common consensus that temporal openings are only a weak indication for higher taxon interrelationship, although it can be informative on lower taxonomic level (Müller, 2003;Laurin and Piñeiro, 2018;MacDougall et al., 2018). Here, I present a rather morphofunctional categorization of temporal openings and provide an ontogenetic explanation on their evolutionary origins.

Morphofunctional categorization

I) Anapsid. The ancestral amniote skull likely had an anapsid shape comparable to the skull of non-amniote groups such as the lungfish with full temporal coverage. Jaw muscles broadly originated from the internal surface of the dermatocranial bones of the temporal region and inserted on the lower jaw (Fig.1I,B,B’).
II) Monapsid. When compared to early anapsid tetrapods, temporal openings in amniotes are correlated to higher domed skulls, which provide space for longer, more voluminous, and hence stronger jaw muscles. Partly via tendons, jaw musculature mainly originates from the edges of an opening, which, due to its round shape, provides a greater structural stability than a flat bone (Case, 1924;Frazzetta, 1968;Tarsitano et al., 2001;Werneburg, 2013a). In that way, a different bite performance can be reached on shore. Fossil Synapsida were mainly predators and their temporal opening might have supported the progressing development of stronger jaw musculature for biting (Fig.1-II) (Kemp, 2005;Angielczyck and Kammerer, 2018). Adaptive radiation, lateron, resulted in the secondary evolution of diverse feeding modes, correlated to tooth morphology, for example, illustrating that the morphofuntional category of a fenestration is not a general proxy for feeding ecology.
Monapsid skulls can also appear in parareptilians and the fenestra can be formed between different bones (Tsuji and Müller, 2009;Macdougall and Reisz, 2014;Laurin and Piñeiro, 2017). The “unstable” position and shape of the fenestra in different parareptilian taxa suggests diverse jaw muscle attachments and bite performances in that group. In many cases, it is not certain whether a “skull hole” is a taphonomic artifact or a real temporal opening.
III) Diapsid. Diapsid groups usually have broader skull roofs (Weishampel et al., 2004), which permits an upper temporal fenestra to develop (Fig.1-III) (Tarsitano et al., 2001). Whether a second, lower temporal opening is ancestral for Diapsida is unclear (Evans, 2008); if not, it is called katapsid (Fig.1-9).
Different to the mostly carnivorous early Synapsida, early Diapsida were certainly insectivorous and had to handle the agile food. For that, the edges of two temporal openings provided complex attachment sites for a highly differentiated jaw musculature permitting very flexible movements of the jaw apparatus (Evans, 2008;Daza et al., 2011). Later in evolution, also other feeding modes evolved within Diapsida.
IV) Excavation. To increase force with longer fibres and to develop more diverse attachment sites, musculature can expand beyond the restrictions of the ancestral temporal anatomy resulting in deep embayments in the dermal skeleton and/or the loss of temporal arcades (Fig.1-IV). Those excavations are known as (a) emarginations, marginal reductions of bones at the edge of the whole temporal region, as well as (b) internal expansions of the temporal fenestrae themselves. Different combinations and degrees of excavations exist, like all morphofunctional categories presented herein are usually fluent.
In different clades of Synapsida, the temporal opening increased in size (Fig.1-21). By supressing the ancestral posttemporal opening in the occipital region and partly the postorbital bar, the temporalis muscle expands through the temporal fenestra, travels to the skull roof and, with that, develops longer and stronger muscle fibres. The zygomatic arch serves as origin site for the newly differentiated masseter muscle, enabling chewing along the mammalian stem and leaving an emargination in the arch (Abdala and Damiani, 2004;Werneburg, 2013a;Lautenschlager et al., 2016).
Similar evolutionary patterns can be recognized in some extinct marine reptiles (Fig.1-12,14), for example, in which the upper temporal opening expands in diameter and even a dorsal parietal crest for the origin of jaw musculature can be developed (Rieppel, 2002), like in several groups of Synapsida. High predatory behavior in the diapsid marine ichthyosaurs and sauropterygians (Liu et al., 2017) might have triggered the dorsal expansion of the jaw musculature and hence the upper temporal opening resulting in the secondary modification (i.e., closure and emargination) of the lower temporal region as well (Rieppel, 2002). Associated to expanding and hence stronger jaw musculature, a stiffening of the originally kinetic skull towards an akinetic morphology can be recognized (e.g., reduction of basicranial articulation, formation of a secondary braincase wall) (Werneburg et al., 2019).
Turtles might have also derived from diapsid (Fig.1-17) reptiles (Rieppel, 2008;Wang et al., 2013;Schoch and Sues, 2015). Neck retraction in stem turtles certainly resulted in the closure of the ancestral temporal opening(s) to resist with a more compact, anapsid skull the tension of the neck musculature in the occipital region (Werneburg, 2015;Werneburg et al., 2015a;Werneburg et al., 2015b). This novel skull anatomy, consequently, differs from skull bone configurations of early anapsid Reptilia (Müller, 2003). Crown turtles evolved longer necks and highly elaborated neck retraction modes (Herrel et al., 2008). To resist the hence further increased neck tension, marginal reductions were introduced to the temporal bones, as they provide broader neck-muscle attachment sites that enable better force distribution (Fig.1-18) (Werneburg, 2012;2015). The posterodorsal emargination, an excavation of the ancestral posttemporal (occipital) fenestra, finally enabled the jaw musculature to expand, a solution to bend around the large otic region in modern turtles, which restricts the space inside the adductor chamber (Rieppel, 1990;Werneburg, 2013b;Ferreira and Werneburg, 2019). As a result, although longer, jaw musculature is more narrow and might not influence the reduction of cranial kinesis as in mammals by higher bite force (Herrel et al., 2002). Nevertheless, skull stiffening takes place in the turtle stem already (Sterli and de la Fuente, 2010) to resist embryonic neck muscle activity (Werneburg and Maier, in press).
Large marginal excavations are not only present in Diapsida or Synapsida. They can even be formed within anapsidian parareptiles (Fig.I-6,7).
The temporal openings in diapsids experienced a great amount of modifications including the reduction or closure of the lower temporal opening (Fig.1-9/10,13-14) (Rieppel and Gronowski, 1981;Reisz et al., 1984), a fusion of both openings with the reduction of the temporal arcades (Fig.1-12) (Cundall and Irish, 2008), and a fusion of both openings with the expanded orbita (Fig.1-16) (Zusi, 1993). Including diapsids, practically almost all amniotes are related to highly diverging feeding modes from insectivory and herbivory to piscivory and carnivory, from biting and chewing to filter and suction feeding, among others (Schwenk, 2000;Silva et al., 2017). This illustrates the general and extraordinary functional plasticity of the temporal skull anatomy in amniotes (Müller, 2003), which breaks phylogenetic constrains and opens avenues for the evolution of new skull shapes (Werneburg et al., 2019). To understand diversity of the temporal region, phylogeny is less guiding than functional adaptations.

Ontogenetic plasticity

The obviously non-exhaustive and simplified functional categorization of temporal fenestrations presented herein only concerns the more proximate explanations of phylogenetically young functional adaptatations. The ultimate causation in the sense of deep time evolutionary change, however, might be detected in life history changes (Sánchez-Villagra, 2012). The majority of non-amniote vertebrates has an anapsid morphotype. Exceptions are chondrichthyans, which lost their dermatocranium completely (Kardong, 2008), and batrachian lissamphibians, which, by flattening of their ancestor’s skull, largely rearranged their jaw musculature and skull arrangment (Rieppel, 1981;Schoch, 2014). The new muscle arrangements in batrachians correlate with the loss of ossification centres through ontogeny as those are no longer recruited as attachment sites through development. This eventually resulted in the complete loss of several ‘non-used’ temporal skull bones in the adults (Schoch, 2014).
Amniotes are characterized by the amniotic egg, which enables complete development of the animal outside an aquatic environment and consequently the conquest of different terrestrial habitats (Sumida and Martin, 1997;Laurin, 2010;Skawiński and Tałanda, 2014;Brocklehurst et al., 2018). With development inside the egg [or, secondarily, inside the mother’s womb (Piñeiro et al., 2012a;Werneburg et al., 2016)], the larval stage is lost and fully formed, adult-resembling hatchings leave the eggs (Fig.1C). This is not the case in non-amniotes, in which larvae hatch and have to feed (Fig.1A). For that, jaw (and branchial) musculature inserts primarily to the embryonic neurocranium (chondrocranium), as dermatocranial bones are not yet well-developed (Fig.1A’-A’’) (Edgeworth, 1935;de Beer, 1937;Ziermann et al., 2018). As such, compared to amniotes, the primordial cartilaginous skull is functional. Dermatocranial, temporal skull bones are later influenced developmentally by the functional jaw musculature near the neurocranium and are incorporated to the feeding apparatus as further attachment sites (Fig.1A’’-B’) (Ziermann et al., 2018). Furthermore, in non-tetrapods, opercle bones also contribute to the dermatocranial armor to protect gill arches, to regulate gill ventilation, and to form a natural and broad edge of the temporal region (Fig.1A’: dotted line) (Goodrich, 1930;de Beer, 1937;Kemp, 1999). Temporal bones “just” fill the narrow “gap“ between opercle bones and the skull roof.
Also water pressure on the skull in larvae and, if aquatic, in adults might result in the formation of a fully encapsulated skull in non-amniotes (Fig.1A,B) as is indicated by the extant amniotic marine turtles (Fig.1-19). Mesozoic marine turtles such as Toxochelys showed a typical turtle-emargination of the temporal region (Matzke, 2009) related to neck retraction. With the reduced ability of neck retraction and a shorter neck in crown marine turtles, a secondary closure of the temporal armor might also correspond to water pressure (Zdansky, 1923;Werneburg, 2012), a factor not present in pure terrestrial amniotes.
By skipping the larval stage (Fig.1C), the amniote’s skull architecture and growth rate differs. The chondrocranium of amniotes becomes less functional for feeding, is hence anatomically more conserved, and the pharyngeal arches are highly modified (de Beer, 1937). The bones not spatially arranged by musculature in early development result in “the embryological failure to close sutures“ (Tarsitano et al., 2001). Jaw musculature receives more freedom to diversify and to evolve in response to different feeding requirements as it is non-functional for a long period of development until hatching/birth. Depending on functional demands, just recently developed temporal bones [still losely “floating” on the head’s surface (Fig.1C’)] are recruited by the specialized jaw musculature in ontogeny and are spatially arranged to become the newly required attachement sites (Fig.1C’’) (Rieppel, 1987;Fucik, 1991). In that way, generally, apomorphic skull bone mosaics can be explained and, more specifically, the diverse shapes of temporal skull openings can be best interpreted by the functional (future) demands of the developing musculature. As correctly pointed out by the above-mentioned phylogenetic observations, temporal fenestrations between synapsids, parareptiles, and diapsids cannot be homologized (only tissue can): the “holes” in the dermatocranium are just the result of developmental plasticity driven by functional adaptations, which, again and finally, can be informative on certain taxonomic levels.
At the dawn of amniote evolution, new habitats were conquered and new ecological adaptations were necessary (Sumida and Martin, 1997;Brocklehurst et al., 2018). The great diversity and plasticity of the temporal skull region in early amniote taxa of the Permian and Triassic, such as in parareptiles (Fig.1-4,5,6,7) (Colbert, 1946;Tsuji et al., 2012;Macdougall and Reisz, 2014;Laurin and Piñeiro, 2017), illustrates the rapid adaptive radiation of vertebrates on land (Nuñez Demarco et al., 2018). The monapsid morphotype stabilized in Synapsida (Fig.1-20,21), whereas an upper temporal opening stabilized in Diapsida (Fig.1-8), although secondary modifications occurred. These are driven by internal and external forces acting on the skull (Gregory and Adams, 1915) as best observable in turtle evolution.

Keywords: fenestration, Emargination, Amniote cranial evolution, Amniote embryos, musculature

Received: 17 Dec 2018; Accepted: 28 Jan 2019.

Edited by:

Michel Laurin, UMR7207 Centre de recherche sur la paléobiodiversité et les paléoenvironnements (CR2P), France

Reviewed by:

Graciela H. Piñeiro, Universidad de la República, Uruguay
Fernando Abdala, University of the Witwatersrand, South Africa
Juan C. Cisneros, Federal University of Piauí, Brazil  

Copyright: © 2019 Werneburg. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

* Correspondence: Dr. Ingmar Werneburg, University of Tübingen, Tübingen, Germany,