Ontogenetic growth in the crania of Exaeretodon argentinus (Synapsida: Cynodontia) captures a dietary shift

Abstract

Background: An ontogenetic niche shift in vertebrates is a common occurrence where ecology shifts with morphological changes throughout growth. How ecology shifts over a vertebrate's lifetime is often reconstructed in extant species-by combining observational and skeletal data from growth series of the same species-because interactions between organisms and their environment can be observed directly. However, reconstructing shifts using extinct vertebrates is difficult and requires well-sampled growth series, specimens with relatively complete preservation, and easily observable skeletal traits associated with ecologies suspected to change throughout growth, such as diet.

Methods: To reconstruct ecological changes throughout the growth of a stem-mammal, we describe changes associated with dietary ecology in a growth series of crania of the large-bodied (∼2 m in length) and herbivorous form, Exaeretodon argentinus (Cynodontia: Traversodontidae) from the Late Triassic Ischigualasto Formation, San Juan, Argentina. Nearly all specimens were deformed by taphonomic processes, so we reconstructed allometric slope using a generalized linear mixed effects model with distortion as a random effect.

Results: Under a mixed effects model, we find that throughout growth, E. argentinus reduced the relative length of the palate, postcanine series, orbits, and basicranium, and expanded the relative length of the temporal region and the height of the zygomatic arch. The allometric relationship between the zygomatic arch and temporal region with the total length of the skull approximate the rate of growth for feeding musculature. Based on a higher allometric slope, the zygoma height is growing relatively faster than the length of the temporal region. The higher rate of change in the zygoma may suggest that smaller individuals had a crushing-dominated feeding style that transitioned into a chewing-dominated feeding style in larger individuals, suggesting a dietary shift from possible faunivory to a more plant-dominated diet. Dietary differentiation throughout development is further supported by an increase in sutural complexity and a shift in the orientation of microwear anisotropy between small and large individuals of E. argentinus. A developmental transition in the feeding ecology of E. argentinus is reflective of the reconstructed dietary transition across Gomphodontia, wherein the earliest-diverging species are inferred as omnivorous and the well-nested traversodontids are inferred as herbivorous, potentially suggesting that faunivory in immature individuals of the herbivorous Traversodontidae may be plesiomorphic for the clade.

Keywords: Allometry; Crania; Cyndont; Dietary ecology; Ecological differentiation; Exaeretodon; Ischigualasto formation; Ontogeny; Traversodontidae; Triassic.

Figure 1. Reconstruction of the skull of Exaeretodon argentinus indicating the measurements taken for this study.

Reconstructions are based on MCZ VPRA-4483. The skull is shown in (A) dorsal, (B) ventral, and (C) right lateral views. Abbreviations: BB, basicranial length; BSL, basal skull length; BW, maxillary bicanine width; DL, diastema length; IO, interorbital distance; MUL, muzzle length; OD, orbit diameter; OL, orbit length; OW, occipital plate width; PAL, palate length; PD, posterior postcanine distance; SW, skull width; TEL, temporal region length; TP, transverse process width; UP, upper postcanine tooth row length; ZH, zygoma height; ZW, maximum zygoma width.

Figure 2. Skulls of Exaeretodon argentinus displaying varying degrees and forms of deformation.

(A) MCZ VPRA-4470; (B) MCZ VPRA-4468; (C) MCZ-VPRA-4486; (D) PVL 2473; (E) PVSJ 103. Skulls are shown in dorsal (A, C, D), ventral (B), and right lateral (E) views. Scale bar equals 10 cm.

Figure 3. Plot comparing the skull length and number of deformed features.

Used as a proxy to compare degree of deformation between specimens.

Figure 4. Representative plots comparing the fit of the generalized linear mixed effects model (GLMM), to the ordinary least squares regression.

95% confidence intervals are reported as dotted lines in corresponding colors. (A) Similar fit with overlapping confidence intervals. GLMM has confidence intervals that are more narrow than those for the ordinary regression. (B) A similarly positive slope for both models with a poorly fit ordinary regression. (C) Contrasting slopes between models with an overall poor fit for both models, but the GLMM is able to incorporate specimens that indicate that the slope is not negative.

Figure 5. Reconstruction of ontogenetic change in the skull of Exaeretodon argentinus.

Upper (A, B) and left (C) reconstruction is based on MCZ VPRA-4470, and lower (A, B) and right (C) reconstruction is based on MCZ VPRA-4483. Reconstructions are shown in (A) dorsal, (B) ventral, and (C) lateral views. Reconstructions are not reflective of actual size differences between individuals but are meant to indicate the major morphological differences between large and small individuals. Black silhouettes reflect the minimum skull size for E. argentinus (∼15 cm), compared to the largest size (colored reconstructions). Bones are colored to reflect homology.

Figure 6. Sutural morphology in Exaeretodon argentinus.

(A) MCZ VPRA-4470 (BSL: 166.6 mm) dorsal view of the interorbital region and posterior snout, with a simple contact between the nasals-frontals-prefrontals. (B) MCZ VPRA-338-58M (BSL: 210.6 mm) dorsal view of the interorbital region and posterior snout, showing simple sutural contacts between the nasals-frontals-prefrontals. (C) MCZ VPRA-4483 (BSL: 305 mm) dorsal view of the interorbital region and posterior snout, with most sutural contacts occluded by matrix/breakages, but the nasofrontal suture is clearly interdigitated. (D) MCZ VPRA-4483 in dorsal view (close up of C) with interdigitated suture between the nasal and lacrimal anterolateral to the right orbit. (E) MCZ VPRA-4483 lateral view of the orbit and anterior portion of the zygoma showing interdigitation between the postorbital-jugal contact. (F) MCZ VPRA-338-58M left zygoma in lateral view with simple contacts between the postorbital-jugal and jugal-squamosal. Scale bar equals two cm in all panels. Abbreviations: Fr, frontal; Ju, Jugal; La, lacrimal; Na, Nasal; Po, postorbital; PrF, prefrontal; Sq, squamosal.

Figure 7. Regression lines between skull width, temporal region length, and zygoma height with basal skull length, based on results of the GLMM.

To show relative rates of growth, black circles indicate points in which zygoma height becomes larger than skull width and temporal region length (both of which occur at skull sizes over 1 m in length, which is not known for E. argentinus). Skull width here represents isometry. The black semicircle indicates that the zygomas are smaller than the temporal region and skull width in juvenile individuals, based on the lower y-intercept (β_GLMM).

Figure 8. SEM images of a tooth cast, capturing the labial face of an isolated postcanine tooth attributed to MCZ VPRA-4470.

Central panel, line drawings of the upper left postcanine in distal and labial views (from top to bottom). Grey boxes are an approximate location where the images are taken from. (A) Groove between primary and secondary cusps (mesialmost), in the apicobasal midpoint of the cast; (B) labialmost portion of the primary cusp; (C) groove between primary and secondary cusps, just below the apical margin; (D) area surrounding wear-facet (bottom left), located at the apical tip of the primary cusp. Arrows denote the mesial (anterior) direction, with striations (anisotropy) perpendicular to the mesiodistal axis of the tooth. Dark rounded areas are holes and are an artifact of bubbles in the casting material. SEM scale bar equals 1 mm.

Figure 9. Phylogeny of cynodont relationships based on Lukic-Walther et al. (2019), with the inclusion of Siriusgnathus, and gomphodontosuchine relationships based on Hendrickx et al. (2020).

Species that have had cranial ontogenetic studies are highlighted in pink. Exaeretodon argentinus is represented in bold and with an asterisk. Steak and leaf icons represent faunivory (carnivory and/or insectivory) and herbivory, respectively. Steak on leaf icon represents omnivory. Question marks indicate uncertainty in diet. Dietary assumptions are based on discussions in various literature (Crompton, 1995; Liu & Abdala, 2014; Hendrickx et al., 2020). Uncertainty in the diet of Siriusgnathus and Exaeretodon riograndensis is represented because of their close affinity to Exaeretodon argentinus, but their absence in this current study.