Tooth Reduction: Definitions and Hypotheses

In general, our understanding of the processes affecting dental development is restricted to species which show loss of mineralized tooth tissues [1] or complete loss of the dentition [2]. In these species the process of dental development is stopped or altered in early embryonic stages, before the onset of odontogenesis [2]. In addition to tooth loss of this type, some species show extreme reduction in the size of the teeth, resulting in what is functionally equivalent to tooth loss (as has been suggested for Stenopterygius quadriscissus) [3]. Complications in the literature have led to the term “tooth reduction” becoming problematic, as it is used indiscriminately to refer to a number of different morphological and developmental changes. For example, tooth reduction has been used to describe species which show; A) reduction in the size of the teeth relative to other species in the clade, B) species showing reduction in enamel structure, C) reduction in the number of teeth, due either to early senescence of the dental lamina or size reduction of the tooth bearing elements, and D) species which show A, B, and C in conjunction or some different combination [1,3–5]. For the purposes of this article, we define tooth reduction as a reduction in the size of the individual teeth, relative to other morphologically similar species in the same clade, or relative to an earlier ontogenetic stage. Tooth reduction can be further subdivided into a number of more specific terms; tooth height reduction, tooth crown height reduction, tooth tissue reduction, and functional tooth reduction. Here, we define tooth height reduction as a reduction of the entire exposed height of the tooth, including the crown height, the height of the ring of acellular cementum (where present), and the height of the exposed root. Tooth crown height reduction, which is the type of reduction discussed in this article, refers to reduction only of the height of the enamel crown. Tooth tissue reduction, which is not discussed here, can be defined as a reduction in the thickness, extent, and structure of specific tooth tissues, including the enamel, dentine, and cementum. Functional tooth reduction refers to situations wherein a combination of tooth height or tooth crown height reduction results in teeth too small to perform their ordinary function (specifically, teeth are considered functionally reduced if the teeth did not protrude beyond the dental groove, resulting in what is functionally equivalent to true edentulism). Finally, we use the term tooth count reduction to refer to reduction in the number of teeth.

Genetic and Evolutionary Perspectives on Tooth Reduction and Loss

In addition to semantic confusion, using tooth reduction to mean different combinations of complete tooth or tissue reduction and loss has also created a problematic understanding of the mechanisms resulting in the different morphologies. Tooth loss is generally understood to be caused by changes in one or more odontogenic pathways which produce binary effects when altered (conventionally attributed to disruption of Bmp4 or Fgf8, but this has been disputed recently [6]); switching these pathways off or prohibiting their interaction with antagonizing proteins results in a complete lack of dental development (and complete edentulism). Because of the clear effects of these changes, we have a relatively good understanding of the genetic and functional precursors necessary to produce complete tooth loss [5,7] (although see [6]). Specifically, in what we are calling the replacement hypothesis (sensu [1]), development of a novel morphology (i.e. the rhamphotheca) that functionally replaces the dentition allows for the teeth to be completely lost via one of these binary pathways. In most cases, the above mentioned mutations which result in a dramatic change in the structure of the teeth (or their complete loss) are fatal, due to the critical role played by the dentition in resource acquisition and processing [1]. In some cases, selective pressures on organisms to maintain a functioning dentition can be reduced due to a behavioral or morphological change [1,5]. To differentiate these adaptive or selectively neutral situations from aberrant cases of tooth reduction and loss (i.e. pathological or traumatic tooth loss), we use the term “evolutionary tooth reduction”. Evolutionary tooth reduction and loss occurs via a number of related mechanisms, all of which appear to involve down-tuning or shutting off of various protein pathways within the odontogenic mesenchyme and epithelium [2,8].

For example, in turtles it has been suggested that tooth loss is the product of mutations resulting in the impairment of Shh signaling in the oral epithelium causing termination of Msx2 expression in the odontogenic mesenchyme, which in turn causes a failure of odontoblast development, producing complete edentulism [2]. Fossil evidence (i.e. Proganochelys) suggests that this tooth loss began rather rapidly in the marginal teeth, leaving early turtles with only palatal teeth, which were lost in a subsequent event [2]. The pathway resulting in edentulism in birds is different from that in turtles; limiting of Bmp4 signaling in the oral epithelium, with a resultant loss of contact with the underlying odontogenic mesenchyme is proposed to halt odontogenesis ([9], but see [6]). Both cases represent a complete disruption of their respective pathways, which leads to failure of tooth development and resultant total dental agenesis. Given the binary effects of these mutations (i.e. teeth or no teeth), they are unlikely to explain the gradual evolutionary tooth reduction seen in many secondarily aquatic tetrapods.

Compared to the genetics of tooth loss, the genetic factors gradually leading to reduced tooth size (as defined above) are poorly understood. Plikus et al. [8] demonstrated that modifying the Bmp signaling pathway in the oral and dental epithelia of mice can result in marked changes in the dentition, including changes in the size of the individual teeth. Different from the conclusions of Harris et al. [9] (wherein the Bmp4 pathway was completely shut off), Plikus et al. [8] demonstrated that tuning down the Bmp pathway (by increasing levels of the protein noggin) resulted in a decreased proliferation of odontoblasts (and consequently a smaller/deteriorated dentition). Where a reduced dentition is present in odontocetes, it is suggested to have evolved under similar conditions to those hypothesized above (specifically dental reduction due to alteration but not inactivation of the Bmp4 pathway) [10]. However, in the odontocetes showing the greatest degree of tooth reduction (i.e. beaked whales), the advent of suction feeding plays a clear role in reducing the functional importance of the dentition [4], and consequently most beaked whale species have completely lost their teeth via the replacement method described above [1] (or had them extremely reduced to functional edentulism—i.e. Tasmacetus shepherdi [11]). As a result, despite morphological similarities, beaked whales and odontocetes are not a good model for the gradual tooth reduction seen in S. quadriscissus.

Fossil ichthyosaurs of the genus Stenopterygius have been suggested to possess a highly reduced (in both size and number of teeth) or essentially absent dentition [12] (Fig 1). The teeth of Stenopterygius are described as reduced from two perspectives; firstly, they are reduced in relative size and structure when compared to closely related ichthyosaur species, such as Ichthyosaurus communis [9]. In I. communis, the teeth are relatively quite large and robust, possessing distinct ridges on the crown indicating a relatively thick enamel [12–13]. While the hypothesized tooth reduction in S. quadriscissus is apparent when considered interspecifically, it appears most distinct when viewed ontogenetically (Fig 1). Specifically, juvenile specimens of S. quadriscissus appear to possess more numerous and relatively larger tooth crowns; consequently, a hypothesis of ontogenetic tooth crown height (and tooth count) reduction has been suggested for this species in the literature [3]. Early descriptions of tooth reduction in Stenopterygius used tooth size as a taxonomic character, with highly reduced or completely absent teeth used to characterize S. quadriscissus, and unreduced teeth used to characterize species such as S. megalorhinus, among others [3]. However, these claims are problematic for a number of reasons, and tooth size is not considered a valid taxonomic character by many authors [14–15] (but see [12]). In a more recent taxonomic assessment [14–15], three species of Stenopterygius are recognized from the Early Jurassic. Intraspecific variation appears to exist in tooth size among adults of all species. Given the amount of intraspecific variation observed, it is unclear whether or not ontogenetic tooth reduction actually occurs in S. quadriscissus. In addition to these issues, debate has surrounded the argument regarding tooth count reduction, mainly based on issues of interpretation. Specifically, the unique tooth-attachment condition in ichthyosaurs has led to suggestions that the pattern of tooth count reduction seen may be a taphonomic artifact [16]. Stenopterygius is known to have a highly derived form of tooth attachment known as “aulocodonty” [17], wherein the teeth are set in a dental groove, but are not ankylosed to the bone [18]. As a result, it has been proposed that the “absent” dentition was in reality removed post-mortem by currents, scavenging, or some other process [16].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1.

Image showing A) juvenile Stenopterygius quadriscissus (SMNS 54026) with relatively large, unreduced teeth (specimen mandible length: 368.67 mm). B) Large Adult S. quadriscissus (SMNS 53001), showing highly reduced, non-functional teeth (too short to protrude beyond the dental groove) (specimen mandible length: 523.2 mm). C) Close-up of the teeth seen in the box inset in B. Scale bars: A = 1 cm, B = 5 cm, and C = 1 cm.

https://doi.org/10.1371/journal.pone.0141904.g001

Tooth reduction has been proposed for ichthyosaur species other than Stenopterygius in the past. Sander et al. [16] suggested that the large ichthyosaur Shastasaurus liangae was completely edentulous, implying it was a suction feeder ecologically analogous to modern beaked whales (Ziphiidae). This suggestion was disputed by Motani et al. [19], who suggested that ichthyosaurs were morphologically incapable of being suction feeders. This complicates interpretations of tooth reduction in both Stenopterygius and Shastasaurus; as was described above, suction feeding can produce the necessary reduction in the function of the teeth necessary for their reduction and loss by gradual vestigial degeneration or by one of the binary mutations [16]. If ichthyosaurs such as Shastasaurus and Stenopterygius were incapable of suction feeding, then there appears to be no clear morphological or functional novelty capable of replacing the dentition (sensu [1]). Given this, convergent dental reduction in the absence of a functional replacement for the dentition could suggest that the teeth played a reduced role in feeding for many ichthyosaur species, allowing them to be lost by gradual vestigial degeneration. This possibility is explored in more detail below.

In this article we provide the first detailed exploration of ontogenetic tooth reduction in Stenopterygius quadriscissus (both in crown height and tooth count), with controls implemented to reduce the effects of taphonomic bias. We hypothesize that the extreme reduction of the teeth seen in S. quadriscissus resulted from a loss of selective pressure to maintain a functioning dentition due to the species’ feeding strategy, which allowed fixation of mutations resulting in; A) strong negative allometry, and B) early senescence of the successional dental lamina. We assess the morphological evolution of this species using a number of techniques from growth allometry, and discuss possible selective pressures that could produce a morphology similar to that seen in S. quadriscissus. This study improves our understanding of the evolutionary and developmental processes by which teeth can become reduced in size through ontogeny, as our current understanding restricts itself to evolutionary edentulism [1] or dental reduction by age-related mesenchymal stem cell senescence [20]. Additionally, we present a hypothetical example of a mechanism potentially capable of driving a gradual pattern of tooth size reduction.

Collections Abbreviations

SMNS = Staatliches Museum für Naturkunde Stuttgart, GPIT = Geologisches und Paläontologische Institut Tübingen, MMH = Urweltmuseum Hauff.

Species Description and Geological Context

We use the taxonomy presented in [14] (see also [15]) for Stenopterygius, wherein three Toarcian species are recognized: Stenopterygius quadriscissus, S. triscissus, and S. uniter. Stenopterygius was a small to medium sized ichthyosaur, growing up to a maximum of around 3.5 meters in length [12]. Morphologically, Stenopterygius maintains the standard derived ichthyosaur body plan; thunniform and adapted for pelagic swimming [12]. The majority of specimens of Stenopterygius come from the Lower Jurassic (Toarcian) deposits of Germany, particularly the Posidonia Shale [15]. The Posidonia Shale was deposited approximately 180 million years ago in a shallow epicontinental sea known as the Southwest German Basin (SWGB), one of the many Central European epicontinental Basins [26]. The SWGB was characterized by a periodically dysoxic benthic environment, resulting in black shale deposition with high (>30%) total organic carbon (TOC) levels, and a limited benthic fauna [26]. Consequently, exceptionally preserved vertebrate fossils are known from this region, including complete, articulated skeletons (see for examples [27]). Given the exceptional nature of preservation of material within the Posidonia Shale, hypotheses suggesting that tooth reduction is an artifact created by taphonomic alteration [16] seem unlikely; small, easily dispersed bones such as distal phalanges and tail fluke vertebrae are frequently found in articulation, as are large numbers of teeth.

Functional and Evolutionary Perspectives on Dental Reduction

Despite the morphological and genetic similarities, the reduced or absent teeth of different vertebrate species serve vastly different functional purposes. As described above, in most cases, tooth loss occurs following the evolution of a new morphological novelty to replace the dentition, which predisposes a species to tooth loss [1]. Interpreted in light of this “replacement hypothesis”, the ontogenetic tooth reduction seen in Stenopterygius appears unusual, as no morphological adaptation replaced the dentition. To explain this, here we suggest that the feeding strategy used by Stenopterygius produced the predisposition to tooth loss. Specifically, Stenopterygius has been previously described as a raptorial feeder within the “Smash” guild of marine reptiles (sensu [13]). Smash feeding involves a raptorial strike (rapid closing of the jaws) upon contact to crush or smash soft-bodied prey, which are then transported and swallowed [13,30–31]. The small, essentially non-functional teeth of adult Stenopterygius (particularly in specimens where the teeth do not protrude beyond the dental groove) seem to imply the force of closing the jaws was sufficient to stun or kill prey [13]. Furthermore, preliminary analysis of stomach contents suggests that even when feeding raptorially functionally edentulous specimens were capable of capturing prey [13]. Given that specimens we consider functionally edentulous (examples with teeth that do not protrude beyond the dental groove—i.e. SMNS 50187, SMNS 80115) have large agglomerations of belemnite hooklets preserved as stomach contents (pers. obs.), we propose that mutations resulting in smaller teeth would not have affected fitness in Stenopterygius (based on the only available evidence suggesting no difference in foraging success between functionally edentulous and “toothed” specimens). If the teeth served a reduced or non-existent function in feeding, mutations allowing for reduction in the size of the teeth could become fixed, due to the absence of stabilizing selection, resulting in slow vestigial degeneration (both the observed trend of negative allometry and the high adult variability observed support this suggestion, as a lack of stabilizing selection has been correlated with increased trait variability and vestigial degeneration–[32] and references therein).

Numerous examples of belemnites with shattered phragmocones from the Posidonia Shale (Fig 8) suggest some potential details into the feeding behavior of Stenopterygius, which could explain how smash feeding of the type described could allow the teeth to become subjected to neutral selection. Adult Stenopterygius appear to have been actively splitting the guards of belemnites from the softer body elements, and consuming the latter. This is supported both by the complete absence of stomach contents other than cephalopod hooklets [13, 33–36], and by the enormous number of isolated belemnite rostra found in the Posidonia Shale (and in particular by a number of examples showing a shattered phragmocone, hypothesized to have been caused by predation [37,38]) (Fig 8). As seen in Fig 8, precision bites to the weaker posterior aspect of the rostrum (the phragmocone, the proximal-most region with respect to the arms) appear to have been necessary to separate the soft tissue from the (presumably inedible) calcitic guard. Previous analyses of stomach contents support the suggestion that Stenopterygius avoided consuming the calcified rostra of belemnites; of the hundreds of Stenopterygius specimens studied, the only cephalopod remains present as gastric contents are belemnite hooklets (with not a single example containing a rostrum) [13,33–36]. This is contrary to the hypothesis advanced by [37], in which ichthyosaurs with reduced dentitions would have eaten primarily ‘naked’ cephalopods. Given that both functionally edentulous and specimens with teeth protruding beyond the dental groove have been found with belemnite hooklets preserved as stomach contents, the force of closing the jaws (or some other mechanism independent of tooth size) appears to have been sufficient to split the belemnites. Note that this does not imply smash feeding “causes” tooth reduction, merely that it allows for accumulation of mutations affecting the size and number of the teeth (by reducing the contribution of tooth size and shape to the overall fitness of the organism). This expands the “replacement hypothesis” [1], and suggests that particular feeding behaviors can reduce the fitness contribution of the dentition in the same way as morphological novelties (such as the rhamphotheca [1]).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 8. An example of a belemnite with a shattered phragmocone, interpreted as a bite mark (SMNS 60863).

Scale bar = 1 cm.

https://doi.org/10.1371/journal.pone.0141904.g008