Brazilian fossils reveal homoplasy in the oldest mammalian jaw joint

Rawson, James R. G.; Martinelli, Agustín G.; Gill, Pamela G.; Soares, Marina B.; Schultz, Cesar L.; Rayfield, Emily J.

doi:10.1038/s41586-024-07971-3

Download PDF

Article
Open access
Published: 25 September 2024

Brazilian fossils reveal homoplasy in the oldest mammalian jaw joint

Nature (2024)Cite this article

11k Accesses
256 Altmetric
Metrics details

Subjects

Abstract

The acquisition of the load-bearing dentary–squamosal jaw joint was a key step in mammalian evolution^1,2,3,4,5. Although this innovation has received decades of study, questions remain over when and how frequently a mammalian-like skull–jaw contact evolved, hindered by a paucity of three-dimensional data spanning the non-mammaliaform cynodont–mammaliaform transition. New discoveries of derived non-mammaliaform probainognathian cynodonts from South America have much to offer to this discussion. Here, to address this issue, we used micro-computed-tomography scanning to reconstruct the jaw joint anatomy of three key probainognathian cynodonts: Brasilodon quadrangularis, the sister taxon to Mammaliaformes^6,7,8, the tritheledontid-related Riograndia guaibensis⁹ and the tritylodontid Oligokyphus major. We find homoplastic evolution in the jaw joint in the approach to mammaliaforms, with ictidosaurs (Riograndia plus tritheledontids) independently evolving a dentary–squamosal contact approximately 17 million years before this character first appears in mammaliaforms of the Late Triassic period^10,11,12. Brasilodon, contrary to previous descriptions^6,7,8, lacks an incipient dentary condyle and squamosal glenoid and the jaws articulate solely using a plesiomorphic quadrate–articular joint. We postulate that the jaw joint underwent marked evolutionary changes in probainognathian cynodonts. Some probainognathian clades independently acquired ‘double’ craniomandibular contacts, with mammaliaforms attaining a fully independent dentary–squamosal articulation with a conspicuous dentary condyle and squamosal glenoid in the Late Triassic. The dentary–squamosal contact, which is traditionally considered to be a typical mammalian feature, therefore evolved more than once and is more evolutionary labile than previously considered.

New evidence from high-resolution computed microtomography of Triassic stem-mammal skulls from South America enhances discussions on turbinates before the origin of Mammaliaformes

Article Open access 15 June 2024

An exceptionally preserved Sphenodon-like sphenodontian reveals deep time conservation of the tuatara skeleton and ontogeny

Article Open access 03 March 2022

A symmoriiform from the Late Devonian of Morocco demonstrates a derived jaw function in ancient chondrichthyans

Article Open access 17 November 2020

Main

Mammals possess a unique secondarily evolved jaw joint between the dentary and squamosal bones^1,2. The elements forming the ancestral jaw joint, the quadrate and articular, were separated from the lower jaw during the evolution of mammals from Mesozoic era fossil relatives, the non-mammaliaform cynodonts^1,2,3,4,5. The evolution of the dentary–squamosal articulation is therefore key to understanding the origin of the unique mammalian body plan. Newly discovered non-mammaliaform cynodonts from the Late Triassic of South America are well placed, both in time and in the evolutionary tree, to answer when and how often the precursor conditions to this load-bearing jaw joint first evolved. Here we use micro-computed-tomography (μCT) scanning to analyse the jaw joint morphology of three key non-mammaliaform cynodont species and identify the oldest example of a dentary–squamosal contact in the fossil record in R. guaibensis⁹, predating that seen in Mammaliaformes by approximately 17 million years^10,11,12. Our updated phylogenetic analyses show that this contact evolved independently from that of mammal precursors, demonstrating that this classical feature is more evolutionarily labile than previously thought. B. quadrangularis, the sister taxon to Mammaliaformes, does not possess a dentary–squamosal jaw articulation as was previously interpreted^6,7,8. Evolutionary experimentation in jaw joint morphology occurred across the cynodont–mammaliaform transition, coinciding with the evolution of features such as thermoregulation, insectivory and miniaturization that would become key elements of the mammalian body plan for millions of years to come^{13,14,15,16,17,18}.

Among the most important diagnostic features of mammals are a unique middle ear containing three ossicles (malleus, incus, stapes) and a load-bearing jaw hinge comprising a dentary condyle and squamosal glenoid^3,4,19,20. The definitive mammalian middle ear (also called the detached middle ear^21,22) evolved within Mammaliaformes during the Mesozoic, with fossil and developmental evidence suggesting that this event occurred at least three times independently^20,23,24,25. However, the major morphological changes needed to facilitate this innovation took place in non-mammaliaform cynodonts (hereafter, cynodonts), the precursors of mammals^5,26. The fossil record documents the evolution of the jaw from early diverging cynodonts to mammaliaforms^1,2,4, in which the postdentary bones (articular, prearticular, surangular, angular) are reduced in size compared with the tooth-bearing dentary to become a collection of thin, rod-like elements as their role in jaw articulation is incrementally reduced (Fig. 1). Eucynodonts experimented with several modifications to the skull that improved sensorial and feeding habits, coupled with the formation of a secondary jaw articulation in between the surangular boss and the articular flange of the squamosal^2,26. Further morphological changes led to the formation of the load-bearing dentary–squamosal joint, consisting of a distinct dentary condyle that articulates to the glenoid cavity of the squamosal bordered posteriorly by a glenoid ridge. In early-diverging mammaliaforms such as Morganucodon, Sinoconodon and others, the dentary–squamosal articulation lies alongside the plesiomorphic quadrate–articular joint^27,28,29. The formation of the dentary–squamosal jaw articulation freed the postdentary bones from their ancestral role as load-bearing elements and allowed them to specialize for sound detection and finally separate from the lower jaw entirely in crown mammals (Fig. 1). Understanding the phylogenetic history and timing of evolution of the dentary–squamosal contact through the cynodont–mammaliaform transition is therefore crucial to interpretation of the origin of this unique mammalian structure.

**Fig. 1: Time-calibrated relationships of Cynodontia.**

Despite a long history of study²⁰, questions remain surrounding the evolution of the jaw joint in cynodonts. Probainognathian cynodonts show variation in jaw joint morphology suggesting evolutionary experimentation among some, but not all, groups^2,4,5. Disagreement in anatomical interpretation and scarcity of fossils have hindered investigation into jaw anatomy and function in key clades. Furthermore, the method by which cynodonts maintained a functioning jaw joint before the formation of the load-bearing dentary–squamosal articulation while simultaneously reducing the postdentary bones for improved sound detection remains unclear. Both reorganization of adductor musculature^1,3,30,31 and overall body miniaturization³² have been considered as hypotheses, although these are not mutually exclusive. Comparative studies of cynodont jaw anatomy have to date largely focused on well-known taxa^1,2,3,32 that are exemplary in terms of preservation and usefulness in anatomical study, but gaps remain in our understanding of cynodont evolution. The reliably dated Candelária Sequence of the Santa Maria Supersequence of southern Brazil¹² has in recent years delivered a suite of derived Late Triassic probainognathian cynodonts^7,9,33,34,35, many of which are represented by multiple specimens including ontogenetic sequences. Of particular note is B. quadrangularis, which has been recovered by most phylogenetic analyses^{7,8,9,26,33,34,35,36,37,38,39,40,41,42,43} as the sister taxon to mammaliaforms and is purported to possess a dentary–squamosal joint similar to that of Morganucodon^5,7,8,26,28. The robust-jawed R. guaibensis is also abundant in the Candelária Sequence, being the key taxon of the Riograndia Assemblage Zone, which includes Brasilodon. This Assemblage Zone was dated using ²⁰⁶Pb/²³⁸U, revealing a maximum deposition age of 225.42 ± 0.37 million years (early Norian)⁴⁴. The phylogenetic placement of Riograndia has varied through time, being considered either a tritheledontid⁹, a non-tritheledontid ictidosaur³⁷ or allied to other traditional ictidosaurs but not forming a monophyletic clade^35,38,43,45. Regardless, it is unambiguously a derived probainognathian placed stemward of mammaliaforms. Previous descriptions have indicated that Jurassic tritheledontids may possess a type of dentary–squamosal contact^{1,4,5,9,46,47}, although this contact lacks the robust dentary condyle as seen in mammaliaforms such as Morganucodon^5,28. The phylogenetic position and preservational quality of Brasilodon, Riograndia and other recently described probainognathian cynodonts makes these taxa key to understanding the origin of the mammalian jaw joint and middle ear, but the anatomical information of these new fossils has yet to be integrated into the wider picture of cynodont morphological evolution alongside better-known species.

To better establish the evolution of jaw joint morphology at this pivotal stage in mammalian history, we reconstructed the three-dimensional anatomy of the jaw joint in Brasilodon, Riograndia and the tritylodontid Oligokyphus using a 3D dataset of cynodont fossils obtained by μCT scanning (Supplementary Table 1; available at data.Bris (Data availability)). We incorporate morphology revealed by our μCT data into an updated phylogenetic analysis (Fig. 4 and Supplementary Information), demonstrating substantial homoplastic evolution of the jaw hinge and middle ear characters in the phylogenetic predecessors of mammals. Brasilodon does not possess a dentary–squamosal articulation as previously described, whereas Riograndia possesses a clear dentary–squamosal contact similar to that of the closely related tritheledontids. Riograndia predates the earliest-known mammaliaform condyle–glenoid jaw joint by approximately 17 million years. This demonstrates that the ictidosaur–tritheledontid clade (Fig. 4) independently acquired the dentary–squamosal contact that bears resemblance to the jaw hinge of mammaliaforms. Evolutionary experimentation with the jaw hinge occurred in derived cynodonts, and some precursor characters evolved iteratively in probainognathians, and then in mammaliaforms.

The jaw joint morphology of Brasilodon

μCT scans of nine Brasilodon specimens highlight notably different jaw joint anatomy from that described previously (Fig. 2). As in many derived probainognathians^1,2,26,48, the lateral and medial ridges of the dentary in Brasilodon create a mediolaterally thickened articular process that extends posteriorly beyond the postdentary bones. The articular process was described previously as an incipient condyle comparable to the condylar process seen in mammaliaforms^6,7,8, but the 3D data show that this thickening is much less developed than in Morganucodon and Sinoconodon, despite the pronounced posterior extension (Fig. 2d–f). The μCT data of the well-preserved specimens of Brasilodon (UFRGS-PV-929-T, UFRGS-PV-1030-T, UFRGS-PV-1043-T) also do not support the presence of an incipient squamosal glenoid, and there is no clear contact surface on the squamosal to accept the lateral ridge of the dentary (Fig. 2e). Digital manipulation of the 3D model of the well-preserved dentaries of Brasilodon UFRGS-PV-1043-T to articulate the jaws shows that the articular process does not reach the squamosal when the teeth are in occlusion, even in the largest specimens. This condition is also present in the minimally distorted left side of UFRGS-PV-1030-T (Fig. 2e), and other specimens such as UFRGS-PV-929-T and UFRGS-PV-760-T, although more distorted, show the same morphology. The lateral ridge of the dentary prohibits any interaction between the surangular and squamosal, meaning that Brasilodon also lacks the secondary surangular–squamosal articulation found in many other eucynodonts^2,26. We find that Brasilodon possesses only the ancestral quadrate-articular jaw joint. The quadrate itself has a well-developed stapedial process and a pronounced neck around the articular surface resembling that of Morganucodon^7,47. μCT scans show that the quadrate surface for contacting the squamosal is strongly concave, in a similar manner to Pachygenelus and especially Morganucodon⁴⁷. Tightly attached to the quadrate is the quadratojugal, which is housed in its own notch in the squamosal separated from the more medial quadrate notch by a ventrally projecting lamina of bone. The arrangement of the quadrate, quadratojugal and squamosal in Brasilodon therefore represents the plesiomorphic condition shown in other non-mammaliaform probainognathians^42,47, contrary to previous descriptions²⁶. Our scans show that the postdentary bones of Brasilodon (Fig. 2g) generally resemble those of cynodonts like Probainognathus and Pachygenelus in size and shape^5,48, rather than the comparatively much smaller and more slender postdentaries of mammaliaforms such as Morganucodon (Fig. 2h), again contrary to previous descriptions^6,7. The articular and prearticular are fused in all of the specimens and the surangular is fused to the dorsal surface of both in all but UFRGS-PV-1030-T, one of the smallest specimens. The angular remains unfused to the other postdentary bones in all of the specimens.

**Fig. 2: Anatomy of *B. quadrangularis* informed by μCT scans.**

The jaw joint anatomy of Brasilodon presents a contradiction. The lack of a squamosal glenoid and spatial separation between squamosal and dentary indicates that the quadrate–articular joint is the main point of articulation, but the dentary possesses numerous features associated with the evolution of a dentary–squamosal joint, such as the expanded lateral and medial ridges and long articular process that extends posteriorly beyond the postdentary bones and prevents any squamosal–surangular contact (Fig. 2). Posterior extension of the articular process is a prerequisite condition for establishing a dentary–squamosal contact, but this by itself does not necessarily denote a dentary squamosal articulation. Tritylodontids possess an articular process that extends posteriorly to the postdentary bones, as in Brasilodon, but unambiguously show no evidence of a dentary–squamosal contact (Supplementary Information). The large medial ridge and posteriorly extended articular process of Brasilodon may have braced the postdentaries from above during jaw movement in the absence of a secondary joint (although such features are less developed in tritylodontids^49,50), resulting in a precursor condition to the dentary–squamosal contact in which the articular process extends posteriorly but does not yet reach the squamosal. Alternatively, Brasilodon may have possessed some form of synovial or ligamentous connection between the dentary and squamosal that provided additional jaw joint support, producing a mismatch of plesiomorphic and derived features. Such a connection has been suggested for Thrinaxodon as a precursor to the secondary surangular-squamosal articulation². There is some osteological evidence for this too: the posterior surface of the articular process of the dentary does appear to possess the type of rugosity commonly associated with cartilaginous covering.

Dentary–squamosal contact in Riograndia

Our examination of Riograndia using μCT scanning has revealed the 3D morphology of its jaw joint region in fine detail (Fig. 3). We find a dentary–squamosal contact in Riograndia (Fig. 4), making this the oldest currently confirmed in the fossil record, although it differs in morphology to the developed articulation first seen in the Late Triassic (Norian–Rhaetian) mammaliaform Haramiyavia^10,51,52, the Late Triassic (Rhaetian) to Early Jurassic (Hettangian–Sinemurian) Morganucodon^11,16,53,54 and the Early Jurassic (Sinemurian) Sinoconodon^27,53. The medial ridge of the dentary forms the roof of the postdentary trough and, together with the lateral ridge, it forms the mediolaterally broad articular process. As in Brasilodon, the articular process extends posteriorly beyond the postdentary bones and therefore prevents any articulation between the surangular and the squamosal (Fig. 3b,c,e,f). In contrast to Brasilodon and tritylodontids, the squamosal of Riograndia bears a clear indentation on the ventromedial surface that would have contacted the lateral ridge of the articular process of the dentary (Fig. 3c,d). This contact surface is well described in the Jurassic tritheledontids Pachygenelus and Diarthrognathus^1,2,27,46,55, which belong to the more inclusive ictidosaur clade including Riograndia^34,35 (Fig. 1). The 3D morphology shown by our µCT scans, particularly that of UFRGS-PV-596-T, enables us to firmly state that a similar contact between the dentary and squamosal was also present in Riograndia. The morphology of this contact differs in several important respects from a condylar dentary–squamosal jaw joint, as seen in mammaliaforms such as Sinoconodon, Morganucodon and Haramiyavia. In Riograndia, there is no distinct glenoid cavity on the squamosal posteriorly rimmed by a ridge, or a true condyle on the articular process of the dentary (Fig. 3c–f), and most of the articulating surface is on the dorsolateral portion of the articular process, rather than on the posterior. Overall, the squamosal configuration in Riograndia bears resemblance to the surangular-squamosal articulation present in some other eucynodonts such as Probainognathus². However, the flaring lateral ridge of the dentary in Riograndia excludes the surangular from directly contacting the squamosal. Despite damage to the zygomatic arch, it is clear that the articular surface in Riograndia extended over a much larger area than the surangular–squamosal articulation of more basal eucynodonts² (Fig. 3d). This condition also differs from the morphology observed in the South African Jurassic tritheledontid Diarthrognathus, which possesses a thickened and rounded posterolateral rim on the articular process of the dentary, superficially resembling a condylar surface that is absent, or at least much less developed, in Riograndia⁴⁶. Pachygenelus, another derived tritheledontid from South Africa, probably also possesses this condition^1,5,27,46,47 (Fig. 3g). It appears then that ictidosaurs (including Riograndia and tritheledontids)³⁷ show some structural variation of the dentary–squamosal contact: the early diverging Riograndia shows a simple convex articular process forming the dentary–squamosal contact, whereas the more derived Pachygenelus and Diarthrognathus show more (albeit still incomplete) resemblance to the dentary condyle and squamosal glenoid in mammaliaforms. Tritheledontids never evolved the dentary condyle or posteriorly ridged glenoid of mammaliaforms, yet it does appear that the Jurassic forms were approaching a more derived condition with a more developed contact between the dentary and the squamosal.

Fig. 3: Anatomy of the lower jaws and jaw joint of *R. guaibensis.*

**Fig. 4: Homoplastic evolution of the dentary–squamosal jaw joint in Cynodontia.**

The dentary–squamosal contact acts as reinforcement to the quadrate–articular joint in Riograndia. The postdentary bones are large, boxy and even more robust than those found in many basal probainognathians (such as Probainognathus) and tritheledontids such as Pachygenelus and Diarthrognathus^4,5,56 (Fig. 3b,c,h,i). Our μCT scans confirm the key features of postdentaries of Riograndia as described previously⁹: the fusion of the angular, articular and prearticular, and a posteriorly projected retroarticular process of the articular, and the reflected lamina on the angular is absent (or alternatively not ossified). Riograndia has a slender quadrate, and the new scans also confirm that it is more similar to that of Probainognathus in the absence of stapedial process (a plesiomorphic trait) than to Brasilodon and Morganucodon^9,47. The μCT scans of UFRGS-PV-596-T also reveal the presence of a tiny quadratojugal in Riograndia (Fig. 3b,c), to our knowledge the first time that this bone has been observed in an ictidosaur. The quadratojugal is a narrow splint of bone of which the ventral portion is accepted by a shallow depression on the anterolateral surface of the quadrate, although the abutting contact has disarticulated in UFRGS-PV-596-T. Dorsally, the quadratojugal becomes flattened and plate-like and inserts into a notch in the squamosal separated by a thin ridge of bone from the more medial notch for the quadrate. The reported absence of a quadratojugal in tritheledontids such as Pachygenelus^5,47 may therefore be due to small size and a lack of preservation rather than a genuine loss, but this requires further investigation. The tiny quadratojugal contrasts with the large, robust quadrate, and the loose connection of the quadratojugal with the squamosal would have improved the ability of the quadrate/quadratojugal to rock forwards during jaw opening to prevent disarticulation².

The robust construction of the quadrate–articular joint alongside a secondary dentary–squamosal contact may imply adaptation to higher stresses in the jaw joint of Riograndia, and perhaps ictidosaurs in general. Characteristics such as enlarged and procumbent lower incisors, reduced canines, leaf-shaped postcanines and the general robustness of the skull and jaws suggest a herbivorous diet in Riograndia⁹, and the boxy postdentaries may have been an additional adaptation to cope with increased loading at the jaw joint. The evolution of the jaw articulation across the cynodont–mammaliaform transition may therefore have been influenced by confounding ecological factors in specific taxa, rather than being a simple stepwise reduction in postdentary size. In tritheledontids like Diarthrognathus and Pachygenelus, the more derived jaw joint may have been able to distribute more stress across its thickened posterior contact surface than the simple convex morphology of Riograndia and so allowed those taxa the freedom to further reduce the size of their postdentary bones and quadrate-articular jaw joint. Although a true dentary–squamosal joint with a developed condyle and squamosal glenoid evolved only in mammaliaforms, the morphology of Riograndia demonstrates that another clade of probainognathian cynodonts independently evolved a dentary–squamosal contact some 17 million years earlier.

Brasilodon has emerged as the sister to mammaliaforms in almost every phylogenetic study since its discovery^{6,7,8,26,33,34,35,36,37,38,39,40,41,42,43}. Both tritheledontids^{27,57,58,59,60} and tritylodontids^{13,19,61,62,63,64,65} have been considered as the sister clade to mammaliaforms in previous studies (summarized previously⁶⁶); however, neither hypothesis invalidates our interpretation that the dentary–squamosal contact evolved at least twice among probainognathian cynodonts. The derived condition in Diarthrognathus compared to Riograndia demonstrates that a more developed contact surface between the dentary and squamosal must have evolved independently within tritheledontids and mammaliaforms (Fig. 4). Tritheledontids are united by a suite of shared characters not possessed by mammaliaforms^37,67,68,69, so it is unlikely that they would represent a grade of cynodonts leading towards mammaliaforms to the exclusion of Brasilodon and tritylodontids. Considering the current spread of phylogenetic results in the literature, our conclusions on homoplasy present within the jaw joints of derived probainognathians are well supported. To address what our revised morphological descriptions mean for the phylogenetic position of taxa, we coded the new data obtained in this study (Supplementary Information) and found little change to the topology of previously published trees. Brasilodon remains the sister taxon to Mammaliaformes alongside the poorly known Botucaraitherium, to the exclusion of tritylodontids and ictidosaurs successively. The strict consensus tree of eight most parsimonious trees (MPTs) shows the ictidosaur clade (Riograndia, Irajatherium, Diarthrognathus + Pachygenelus) as monophyletic and sister to the clade composed of tritylodontids, Botucaraitherium/Brasilodon plus Mammaliaformes (Supplementary Fig. 1). In four out of the eight MPTs, Riograndia was found to be the sister of Irajatherium in a clade that is itself the sister to Diarthrognathus + Pachygenelus, whereas, in the other four MPTs, Riograndia was found to be the sister of Diarthrognathus + Pachygenelus, with Irajatherium recovered as the most basal ictidosaur. We found similar results using Bayesian analysis, with Ictidosauria forming a monophyletic group that, alongside Therioherpeton, forms the sister to a clade containing tritylodontids, Botucaraitherium/Brasilodon plus Mammaliaformes (Supplementary Fig. 2). The more basal position of a monophyletic Ictidosauria compared to tritylodontids and Brasilodon in all of the generated trees cements our conclusion that the dentary–squamosal contact evolved twice within Cynodontia. The use of this character as a defining feature of mammals is therefore problematic as it arose not only within non-mammalian mammaliaforms but also separately within probainognathian cynodonts. The evolution of a distinct condyle and glenoid, specifically, separates the dentary–squamosal joint in Mammaliaformes from the more basal dentary–squamosal lateral contact in derived cynodonts⁷⁰.

Jaw joint innovation in Cynodontia

The morphology of Brasilodon, ictidosaurs and tritylodontids demonstrates that considerable experimentation with jaw joint morphology occurred within derived probainognathians, both in the two independent acquisitions of the dentary–squamosal jaw contact and the loss of the surangular-squamosal jaw joint (Fig. 4). This is in contrast to both basal probainognathians and cynognathians, the other large radiation of eucynodonts (Fig. 1), which show little variation in their jaw joint configuration. Among cynognathians, derived forms like the traversodontid Massetognathus possess a more developed articular facet on the squamosal and a more lateral articular surface on the articular boss than basal taxa like Diademodon², but the general topology of the bones remains the same. Low disparity in the cynognathian jaw joint contrasts with the general pattern of their evolution, which has previously been characterized by higher diversity, morphological disparity and character evolutionary rates compared to probainognathians⁶¹, alongside the evolution of highly complex food processing and tooth replacement systems⁷¹. The surangular-squamosal joint of basal probainognathians generally resembles that of derived cynognathians, with a laterally placed squamosal flange similar to taxa like Massetognathus². Initial descriptions of the early Carnian cynodont Probainognathus described the secondary articulation with the squamosal as involving both the surangular and the dentary^72,73, a description later repeated by Allin³. If accurate, this would constitute a third, older independent evolution of dentary–squamosal contact in probainognathians, but subsequent works^4,5,26 state that the secondary articulation in Probainognathus involved only the surangular, as in most other eucynodonts.

This raises the question of why derived probainognathians show much more plasticity in jaw articulation than other non-mammaliaform cynodonts. It has been hypothesized that the reorganization of muscles to a more mammalian arrangement reduced the mechanical load on the primary jaw joint, allowing for a reduction in the postdentary bones and promoting the evolution of secondary articulations^3,30,46,74. It is likely that the majority of muscle reorganization had already occurred at the base of Eucynodontia³¹, with the exception being the shift of the pterygoideus from the postdentary rod to the dentary^{3,30,31,75,76}. However, uncertainty surrounding osteological correlates for the pterygoideus has obscured the exact timing of this shift^3,31,56,77. One possible correlate, the angular process of the dentary (that may or may not be homologous to the mammalian angular process⁷⁷), evolved sporadically across eucynodonts and occurs in many cynognathians; it is extremely pronounced in taxa such as Massetognathus and Exaeretodon. Cynognathians may therefore have evolved a fully mammalian masticatory muscle arrangement but never evolved the diversity in jaw articulation that derived probainognathians did. Moreover, recent biomechanical studies have shown that the musculoskeletal system of the cynodont jaw can be evolutionarily labile without affecting jaw joint stress⁷⁸, and that changing the arrangement of masticatory muscles probably had little effect on the reaction forces around the quadrate–articular joint³². Rearrangement of muscles is therefore unlikely to be solely responsible for the increased innovation in derived probainognathians.

Our findings for Brasilodon and Riograndia are more consistent with the hypothesis that body-size reduction would have facilitated, and perhaps promoted, experimentation with jaw joint morphology in derived probainognathian cynodonts including homoplastic evolution of a dentary–squamosal contact³². Experimentation with jaw joint morphology in derived probainognathians, but not in other eucynodonts, is consistent with a reduction in absolute jaw joint stress accompanying smaller body sizes in these taxa. The jaw joint of Brasilodon was apparently capable of functioning without a reinforcing secondary contact, and articulation took place solely through the small quadrate–articular joint. Miniaturization has been correlated with the evolution of other key mammalian traits including metabolic rate, thermoregulation and insectivorous ecology^{13,14,15,16,17,18,79,80,81}, and it seems likely that small body size was also involved in driving evolution of the mammalian jaw joint. However, the relatively larger size of some tritylodontids like Kayentatherium suggests further complexity, and other factors such as muscle reorganization, feeding ecology and masticatory behaviour may have contributed^30,49,74,82. The relatively larger postdentaries of Riograndia alongside a secondary dentary–squamosal contact represent a different instance of evolutionary experimentation that occurred earlier than the acquisition of a secondary jaw joint for load bearing in mammaliaforms; one that appears to be characterized by greater jaw joint robustness, possibly to cope with a more mechanically demanding diet, even at small body sizes.

Conclusions

3D μCT data of Brasilodon and Riograndia illustrate homoplastic evolution of the dentary–squamosal contact within probainognathian cynodonts and reveal the oldest occurrence of this feature in stem mammals. Although only Mammaliaformes evolved a fully developed dentary condyle and squamosal glenoid, the presence of an extensive lateral dentary–squamosal contact in Riograndia and the thickened, more condyle-like posterior rim of the articular process in more derived Jurassic tritheledontids demonstrates that at least two cynodont lineages were evolving features of the jaw articulation that are typically associated with mammals. Although traditionally considered a mammalian character, both non-mammalian mammaliaforms and non-mammaliaform probainognathian cynodonts evolved a dentary–squamosal contact, making the use of this character in defining the unique features of mammals problematic. Despite occurring in multiple lineages of derived probainognathians, innovation in jaw joint morphology occurs relatively infrequently among other groups of cynodonts, and the miniaturization of the cynodont body plan may have been vital to the development of the dentary–squamosal jaw joint, demonstrating the importance of studying the cynodont–mammaliaform transition for understanding mammalian evolution.

Methods

Sampling

Many cynodont specimens from the Triassic of Brazil are available for study, but we chose to focus on those with well-preserved lower jaw anatomy, especially those with postdentary bones. Numerous specimens used in this study were loaned for scanning from the Universidad Federal do Rio Grande do Sul (Porto Alegre, Brazil). Further information for each specimen is provided in Supplementary Table 1.

Scanning and data processing

The 26 specimens used for this study were scanned either in the Nikon XTH225 ST Micro CT scanner at the University of Bristol, the Nikon Metrology HMX ST 225 or Zeiss Xradia Versa 520 at the Natural History Museum, London, the Phoenix High-Resolution X-ray x|s Nanofocus system at the University of Finland or the SkyScan 1173 High Energy Micro-CT system at the Pontifical Catholic University of Rio Grande do Sul. Transverse TIFF slice number in the scanned specimens ranged from around 1,500 to >3,000 and the scans produced had a voxel size range of 0.0057 mm to 0.026 mm. Scans were processed using the 3D visualization software Avizo 3D 2021.1 (Thermo Fisher Scientific) or Dragonfly 2022.2 (Object Research Systems) to transform scan data into 3D reconstructions. Density thresholding was used to distinguish fossil material from air and surrounding matrix. Scans were then segmented manually slice by slice, interpolating across no more than five slices at a time, to separate the relevant bones. The dentary was segmented in each case alongside any postdentary elements that could be recovered, including disarticulated or fragmented material. Where available, the quadrate, quadratojugal and squamosal were also segmented to allow description of the jaw joint. Reconstructions of complete postdentary anatomy were estimated for Brasilodon and Riograndia by manipulating individual bone surfaces and placing them into the hypothesized life position, informed by in situ position on other specimens. Further repairs for missing elements were carried out in Blender 3.4, details of which are provided below.

B. quadrangularis

Specimens included in Brasilodon were initially considered to represent three separate species^10,11,12 but, more recently, have been suggested to represent an ontogenetic series (refs. ^38,86,87 and additional work in preparation by the authors), which is how they are treated in this work. All nine Brasilodon specimens used in this study (Supplementary Table 1) were recovered from the Riograndia Assemblage Zone, Candelária Sequence, Santa Maria Supersequence in the Rio Grande do Sul State, Brazil¹¹. To create the reconstruction of shown in Fig. 4, the dentary, postdentaries and splenials of UFRGS-PV-1043-T were first moved into the hypothesized life position and minor cracks were repaired using the segmentation editor in Avizo⁸⁸. Some postdentary elements from the right jaw were mirrored over to the left jaw, which had a complete coronoid process. The coronoid was segmented from UFRGS-PV-628-T (Fig. 2g) and scaled using equivalent measurements of the jaws and skull. Once moved into the life position, the remaining missing portions were repaired in Blender v.3.4 using previously established protocols⁸⁸. The initial segmented surfaces of UFRGS-PV-1043-T before restoration can be viewed in Extended Data Fig. 1.

R. guaibensis

Nine specimens of Riograndia were used in this study, all of which originate from the Riograndia AZ of the Candelária Sequence⁶. The reconstruction of the jaw of Riograndia shown in Fig. 4 was amalgamated using elements from UFRGS-PV-596-T (dentary), UFRGS-PV-833-T (postdentaries and teeth) and UFRGS-PV-624-T (coronoid). The broken elements of the dentary of UFRGS-PV-596-T were moved into the life position relative to each other in Avizo and minor cracks were repaired in the segmentation editor⁸⁸. The remaining elements were scaled from their respective specimens using equivalent measurements of the skull and jaws and positioned in the reconstruction based on information from additional specimens. The postdentary bones of UFRGS-PV-624-T were also reconstructed in Avizo to compare to those of UFRGS-PV-833-T. Once positioned, all elements were exported to Blender v.3.4 as .stl files and restored using information from additional specimens to produce the final reconstruction⁸⁸. The various stages of this process are shown in Extended Data Fig. 2.

O. major

Seven specimens of Oligokyphus were scanned for use in this study, all of which are held at the Natural History Museum, London. These specimens were excavated by W. Künhe from the Windsor Hilll quarry in Somerset, a terrestrial fissure deposit that has been interpreted as Hettangian to early Sinemurian in age^8,50. No complete jaw or skull exists for O. major, although between the specimens scanned in this study, the dentary at least can be reconstructed with some accuracy. Our μCT scans corroborated the findings of previous descriptions, including the absence of both a dentary–squamosal joint and a surangular–squamosal joint (Extended Data Fig. 3 and Supplementary Information for more details). Details of all specimens used are provided in Supplementary Table 1.

Phylogenetic analysis

To ensure that our conclusions about character evolution within derived probainognathians are robust, we performed a phylogenetic analysis of non-mammaliaform cynodonts with updated characters based on our anatomical interpretations. The full character list and an outline of all changes made can be found in the Supplementary Information. The final data matrix includes 158 morphological characters and 42 terminal taxa of specific level. The data matrix was analysed in TNT (v.1.5)⁸⁹ under a heuristic search including random addition sequence + tree bisection reconnection, with 1,000 replicates of Wagner trees (with random seed = 0) and using tree bisection reconnection and branch swapping (holding 10 trees save per replication). All characters were considered as unordered and with the same weight. Time calibration of the strict consensus tree (Supplementary Fig. 1) was performed in R using the ‘equal’ method implemented in the paleotree package⁹⁰. The first and last appearance dates (Supplementary Table 2) were sourced from the literature^8,42,61,91. Moreover, a Bayesian phylogenetic analysis was performed in MrBayes (v.3.2.7)⁹² with four Markov chain Monte Carlo chains, using a discrete morphological character model. The analysis was run over 1 million generations and was sampled trees every 100 generations which, using a burn-in of 25%, resulted in 7,500 retained trees (Supplementary Fig. 2).

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The data matrix used in the phylogenetic analysis is provided in separate files in Mesquite and TNT formats. Videos showing rotating views of the 3D surface models of UFRGS-PV-628-T (Brasilodon), UFRGS-PV-929-T (Brasilodon), UFRGS-PV-1030-T (Brasilodon), UFRGS-PV-1043-T (Brasilodon), MPDC-1B1 (Riograndia), UFRGS-PV-596-T (Riograndia), UFRGS-PV-833-T (Riograndia) and UFRGS-PV-1319-T (Riograndia); STL files of figured specimens; and jaw reconstructions for Brasilodon and Riograndia shown in Fig. 4 (the method for reconstruction described above) are available at the University of Bristol data repository data.bris (https://doi.org/10.5523/bris.2ie8neiry701e23iayx9gdsytv).

References

Luo, Z.-X. & Manley, G. A. in The Senses: A Comprehensive Reference Vol. 2 (eds Fritzsch, B. & Grothe, B.) 207–252 (Elsevier Academic, 2020).
Crompton, A. in Studies in Vertebrate Evolution (eds Joysey, K. A. & Kemp, T. S.) 231–251 (Oliver and Boyd, 1972).
Allin, E. F. Evolution of the mammalian middle ear. J. Morphol. 147, 403–437 (1975).
Article CAS PubMed Google Scholar
Allin, E. F. & Hopson, J. A. in The Evolutionary Biology of Hearing (eds Webster, D. B. et al.) 587–614 (Springer, 1992).
Luo, Z.-X. Developmental patterns in Mesozoic evolution of mammal ears. Ann. Rev. Ecol. Evol. Syst. 42, 355–380 (2011).
Article Google Scholar
Bonaparte, J. F., Martinelli, A. G., Schultz, C. L. & Rubert, R. The sister group of mammals: small cynodonts from the Late Triassic of southern Brazil. Rev. Bras. Paleontol. 8, 25–46 (2003).
Bonaparte, J. F., Martinelli, A. G. & Schultz, C. L. New information on Brasilodon and Brasilitherium (Cynodontia, Probainognathia) from the Late Triassic of southern Brazil. Rev. Bras. Paleontol. 8, 25–46 (2005).
Article Google Scholar
Bonaparte, J. F. Evolution of the Brasilodontidae (Cynodontia-Eucynodontia). Hist. Biol. 25, 643–653 (2013).
Article Google Scholar
Soares, M. B., Schultz, C. L. & Horn, B. L. New information on Riograndia guaibensis Bonaparte, Ferigolo & Ribeiro, 2001 (Eucynodontia, Tritheledontidae) from the Late Triassic of southern Brazil: anatomical and biostratigraphic implications. An. Acad. Bras. Ciênc. 83, 329–354 (2011).
Article PubMed Google Scholar
Clemmensen, L. B. et al. The vertebrate-bearing Late Triassic Fleming Fjord Formation of central East Greenland revisited: stratigraphy, palaeoclimate and new palaeontological data. Geol. Soc. Lond. 434, 31–47 (2016).
Whiteside, D. I., Duffin, C. J., Gill, P. G., Marshall, J. E. & Benton, M. J. The Late Triassic and Early Jurassic fissure faunas from Bristol and South Wales: stratigraphy and setting. Palaeontol. Pol. 67, 257–287 (2016).
Google Scholar
Schultz, C. L. et al. Triassic faunal successions of the Paraná Basin, southern Brazil. J. South Am. Earth Sci. 104, 102846 (2020).
Article Google Scholar
Rowe, T. in Mammal Phylogeny (eds Szalay, F. S. et al.) 129–145 (Springer, 1993).
Kemp, T. The origin of higher taxa: macroevolutionary processes, and the case of the mammals. Acta Zool. 88, 3–22 (2007).
Article Google Scholar
Kemp, T. The endocranial cavity of a nonmammalian eucynodont, Chiniquodon theotonicus, and its implications for the origin of the mammalian brain. J. Vertebr. Paleontol. 29, 1188–1198 (2009).
Article Google Scholar
Kielan-Jaworowska, Z., Cifelli, R. L. & Luo, Z.-X. Mammals From the Age of Dinosaurs: Origins, Evolution, and Structure (Columbia Univ. Press, 2005).
Newham, E., Gill, P. G. & Corfe, I. J. New tools suggest a middle Jurassic origin for mammalian endothermy: advances in state‐of‐the‐art techniques uncover new insights on the evolutionary patterns of mammalian endothermy through time. BioEssays 44, 2100060 (2022).
Article Google Scholar
Hellert, S. M., Grossnickle, D. M., Lloyd, G. T., Kammerer, C. F. & Angielczyk, K. D. Derived faunivores are the forerunners of major synapsid radiations. Nat. Ecol. Evol. 7, 1903–1913 (2023).
Article PubMed Google Scholar
Rowe, T. Definition, diagnosis, and origin of Mammalia. J. Vertebr. Paleontol. 8, 241–264 (1988).
Article Google Scholar
Luo, Z.-X., Schultz, J. A. & Ekdale, E. G. in Evolution of the Vertebrate Ear (eds Clack, J. A. et al.) 139–174 (Springer, 2016).
Harper, T. & Rougier, G. W. Petrosal morphology and cochlear function in Mesozoic stem therians. PLoS ONE 14, e0209457 (2019).
Article CAS PubMed PubMed Central Google Scholar
Wang, J. et al. A monotreme-like auditory apparatus in a Middle Jurassic haramiyidan. Nature 590, 279–283 (2021).
Article ADS CAS PubMed Google Scholar
Ji, Q., Luo, Z.-X., Zhang, X., Yuan, C.-X. & Xu, L. Evolutionary development of the middle ear in Mesozoic therian mammals. Science 326, 278–281 (2009).
Article ADS CAS PubMed Google Scholar
Anthwal, N., Urban, D. J., Luo, Z.-X., Sears, K. E. & Tucker, A. S. Meckel’s cartilage breakdown offers clues to mammalian middle ear evolution. Nat. Ecol. Evol. 1, 0093 (2017).
Article PubMed PubMed Central Google Scholar
Meng, J., Bi, S., Zheng, X. & Wang, X. Ear ossicle morphology of the Jurassic euharamiyidan Arboroharamiya and evolution of mammalian middle ear. J. Morphol. 279, 441–457 (2018).
Article PubMed Google Scholar
Luo, Z.-X. Transformation and diversification in early mammal evolution. Nature 450, 1011–1019 (2007).
Article ADS CAS PubMed Google Scholar
Crompton, A. W. & Luo, Z.-X. in Mammal Phylogeny (eds Szalay, F. S. et al.) 30–44 (Springer, 1993).
Kermack, K., Mussett, F. & Rigney, H. The lower jaw of Morganucodon. Zool. J. Linn. Soc. 53, 87–175 (1973).
Article Google Scholar
Luo, Z.-X., Bhullar, B.-A. S., Crompton, A. W., Neander, A. I. & Rowe, T. B. Reexamination of the mandibular and dental morphology of the Early Jurassic mammaliaform Hadrocodium wui. Acta Palaeontol. Pol. 67, 95–113 (2022).
Article Google Scholar
Bramble, D. M. Origin of the mammalian feeding complex: models and mechanisms. Paleobiology 4, 271–301 (1978).
Article Google Scholar
Lautenschlager, S., Gill, P., Luo, Z. X., Fagan, M. J. & Rayfield, E. J. Morphological evolution of the mammalian jaw adductor complex. Biol. Rev. 92, 1910–1940 (2016).
Article PubMed Google Scholar
Lautenschlager, S., Gill, P. G., Luo, Z.-X., Fagan, M. J. & Rayfield, E. J. The role of miniaturization in the evolution of the mammalian jaw and middle ear. Nature 561, 533–537 (2018).
Article ADS CAS PubMed Google Scholar
Soares, M. B., Martinelli, A. G. & Oliveira, T. V. A new prozostrodontian cynodont (Therapsida) from the Late Triassic Riograndia Assemblage Zone (Santa Maria Supersequence) of Southern Brazil. An. Acad. Bras. Ciênc. 86, 1673–1691 (2014).
Article PubMed Google Scholar
Martinelli, A. G., Soares, M. B. & Schwanke, C. Two new cynodonts (Therapsida) from the Middle-Early Late Triassic of Brazil and comments on South American probainognathians. PLoS ONE 11, e0162945 (2016).
Article PubMed PubMed Central Google Scholar
Martinelli, A. G., Eltink, E., Da‐Rosa, Á. A. & Langer, M. C. A new cynodont from the Santa Maria Formation, south Brazil, improves Late Triassic probainognathian diversity. Pap. Palaeontol. 3, 401–423 (2017).
Article Google Scholar
Abdala, F. Redescription of Platycraniellus elegans (Therapsida, Cynodontia) from the Lower Triassic of South Africa, and the cladistic relationships of eutheriodonts. Palaeontology 50, 591–618 (2007).
Article Google Scholar
Martinelli, A. G. & Rougier, G. W. On Chaliminia musteloides (Eucynodontia: Tritheledontidae) from the Late Triassic of Argentina, and a phylogeny of Ictidosauria. J. Vertebr. Paleontol. 27, 442–460 (2007).
Article Google Scholar
Liu, J. & Olsen, P. The phylogenetic relationships of Eucynodontia (Amniota: Synapsida). J. Mamm. Evol. 17, 151–176 (2010).
Article ADS Google Scholar
Rodrigues, P. G., Ruf, I. & Schultz, C. L. Digital reconstruction of the otic region and inner ear of the non-mammalian cynodont Brasilitherium riograndensis (Late Triassic, Brazil) and its relevance to the evolution of the mammalian ear. J. Mamm. Evol. 20, 291–307 (2013).
Article Google Scholar
Guignard, M. L., Martinelli, A. G. & Soares, M. B. Reassessment of the postcranial anatomy of Prozostrodon brasiliensis and implications for postural evolution of non-mammaliaform cynodonts. J. Vertebr. Paleontol. 38, e1511570 (2018).
Article Google Scholar
Pacheco, C. P., Martinelli, A. G., Pavanatto, A. E., Soares, M. B. & Dias-da-Silva, S. Prozostrodon brasiliensis, a probainognathian cynodont from the Late Triassic of Brazil: second record and improvements on its dental anatomy. Hist. Biol. 30, 475–485 (2018).
Article Google Scholar
Wallace, R. V., Martínez, R. & Rowe, T. First record of a basal mammaliamorph from the early Late Triassic Ischigualasto Formation of Argentina. PLoS ONE 14, e0218791 (2019).
Article CAS PubMed PubMed Central Google Scholar
Benoit, J. et al. Synchrotron scanning sheds new light on Lumkuia fuzzi (Therapsida, Cynodontia) from the Middle Triassic of South Africa and its phylogenetic placement. J. Afr. Earth. Sci. 196, 104689 (2022).
Article Google Scholar
Langer, M. C., Ramezani, J. & Da Rosa, Á. A. U-Pb age constraints on dinosaur rise from south Brazil. Gondwana Res. 57, 133–140 (2018).
Article ADS CAS Google Scholar
Kerber, L., Martinelli, A. G., Müller, R. T. & Pretto, F. A. A new specimen provides insights into the anatomy of Irajatherium hernandezi, a poorly known probainognathian cynodont from the Late Triassic of southern Brazil. Anat. Rec. 305, 3113–3132 (2022).
Article Google Scholar
Crompton, A. & Parkyn, D. On the lower jaw of Diarthrognathus and the origin of the mammalian lower jaw. Proc. Zool. Soc. Lond. 140, 697–749 (1963).
Article Google Scholar
Luo, Z.-X. & Crompton, A. W. Transformation of the quadrate (incus) through the transition from non-mammalian cynodonts to mammals. J. Vertebr. Paleontol. 14, 341–374 (1994).
Article Google Scholar
Crompton, A. in Functional Morphology in Vertebrate Paleontology (ed. Thomason, J. J.) 55–75 (Cambridge Univ. Press, 1995).
Sues, H.-D. Skull and dentition of two tritylodontid synapsids from the Lower Jurassic of western North America. Bull. Mus. Compar. Zool. Harv. Coll. 151, 217–268 (1986).
Google Scholar
Kühne, W. G. The Liassic Therapsid Oligokyphus (Trustees of the British Museum, 1956).
Jenkins, F. Jr et al. Late Triassic continental vertebrates and depositional environments of the Fleming Fjord Formation, Jameson Land, East Greenland: Meddelelser om Grønland. Geoscience 32, 1–25 (1994).
Google Scholar
Luo, Z.-X., Gatesy, S. M., Jenkins, F. A. Jr, Amaral, W. W. & Shubin, N. H. Mandibular and dental characteristics of Late Triassic mammaliaform Haramiyavia and their ramifications for basal mammal evolution. Proc. Natl Acad. Sci. USA 112, E7101–E7109 (2015).
Article CAS PubMed PubMed Central Google Scholar
Luo, Z.-X. & Wu, X. C. In In the Shadow of the Dinosaurs: Early Mesozoic Tetrapods (eds Fraser, N. & Sues, H.-D.) 251–270 (Cambridge Univ. Press, 1994).
Debuysschere, M., Gheerbrant, E. & Allain, R. Earliest known European mammals: a review of the Morganucodonta from Saint-Nicolas-de-Port (Upper Triassic, France). J. Syst. Paleontol. 13, 825–855 (2015).
Article Google Scholar
Gow, C. Pachygenelus, Diarthrognathus and the double jaw articulation. Palaeontol. Afr. 24, 15 (1981).
MathSciNet Google Scholar
Crompton, A. The evolution of the mammalian jaw. Evolution 17, 431–439 (1963).
Hopson, J. A. & Barghusen, H. R. in The Ecology and Biology of Mammal-like Reptiles (eds Hotton, N. et al.) 83–106 (Smithsonian Institution Press, 1986).
Shubin, N. H., Crompton, A., Sues, H.-D. & Olsen, P. E. New fossil evidence on the sister-group of mammals and early Mesozoic faunal distributions. Science 251, 1063–1065 (1991).
Article ADS CAS PubMed Google Scholar
Luo, Z.-X. in In the Shadow of the Dinosaurs: Early Mesozoic Tetrapods (eds Fraser, N. & Sues, H.-D.) 98–128 (Cambridge Univ. Press, 1994).
Hopson, J. A. & Kitching, J. W. A probainognathian cynodont from South Africa and the phylogeny of nonmammalian cynodonts. Bull. Mus. Compar. Zool. 156, 5–35 (2001).
Google Scholar
Ruta, M., Botha-Brink, J., Mitchell, S. A. & Benton, M. J. The radiation of cynodonts and the ground plan of mammalian morphological diversity. Proc. R. Soc. B 280, 20131865 (2013).
Article PubMed PubMed Central Google Scholar
Kemp, T. The relationships of mammals. Zool. J. Linn. Soc. 77, 353–384 (1983).
Article Google Scholar
Wible, J. R. Origin of Mammalia: the craniodental evidence reexamined. J. Vertebr. Paleontol. 11, 1–28 (1991).
Article Google Scholar
Wible, J. R. & Hopson, J. A. in Mammal Phylogeny (eds Szalay, F. S. et al.) 45–62 (Springer, 1993).
Martínez, R. N., May, C. L. & Forster, C. A. A new carnivorous cynodont from the Ischigualasto Formation (Late Triassic, Argentina), with comments on eucynodont phylogeny. J. Vertebr. Paleontol. 16, 271–284 (1996).
Article Google Scholar
Kemp, T. S. The Origin and Evolution of Mammals. (Oxford Univ. Press, 2005).
Martinelli, A. G., Bonaparte, J. F., Schultz, C. L. & Rubert, R. A new tritheledontid (Therapsida, Eucynodontia) from the Late Triassic of Rio Grande do Sul (Brazil) and its phylogenetic relationships among carnivorous non-mammalian eucynodonts. Ameghiniana 42, 191–208 (2005).
Google Scholar
Sidor, C. & Hancox, P. Elliotherium kersteni, a new tritheledontid from the lower Elliot Formation (Upper Triassic) of South Africa. J. Paleontol. 80, 333–342 (2006).
Article ADS Google Scholar
Oliveira, T. V., Martinelli, A. G. & Soares, M. B. New information about Irajatherium hernandezi Martinelli, Bonaparte, Schultz & Rubert 2005 (Eucynodontia, Tritheledontidae) from the Upper Triassic (Caturrita Formation, Paraná Basin) of Brazil. PalZ 1, 67–82 (2011).
Article Google Scholar
Luo, Z.-X., Kielan-Jaworowska, Z. & Cifelli, R. L. In quest for a phylogeny of Mesozoic mammals. Acta Palaeontol. Pol. 47, 1–78 (2002).
Google Scholar
Melo, T. P., Ribeiro, A. M., Martinelli, A. G. & Soares, M. B. Early evidence of molariform hypsodonty in a Triassic stem-mammal. Nat. Commun. 10, 2841 (2019).
Article ADS PubMed PubMed Central Google Scholar
Romer, A. S. A chiniquodontid cynodont with an incipient squamosal-dentary jaw articulation. Breviora 344, 1–18 (1970).
Google Scholar
Barghusen, H. R. & Hopson, J. A. Dentary-squamosal joint and the origin of mammals. Science 168, 573–575 (1970).
Article ADS CAS PubMed Google Scholar
Crompton, A. & Hylander, W. in The Ecology and Biology of Mammal-like Reptiles (eds. Hotton, N. et al.) 83–106 (Smithsonian Institution Press, 1986).
Watson, D. The evolution of the mammalian ear. Evolution 7, 159–177 (1953).
Article Google Scholar
Barghusen, H. R. The lower jaw of cynodonts (Reptilia, Therapsida) and the evolutionary origin of mammal-like adductor jaw musculature. Postilla 116, 1–56 (1968).
Google Scholar
Abdala, F. & Damiani, R. Early development of the mammalian superficial masseter muscle in cynodonts. Palaeontol. Afr. 40, 23–29 (2004).
Google Scholar
Reed, D., Iriarte‐Diaz, J. & Diekwisch, T. A three dimensional free body analysis describing variation in the musculoskeletal configuration of the cynodont lower jaw. Evol. Dev. 18, 41–53 (2016).
Article CAS PubMed Google Scholar
Rowe, T. B., Macrini, T. E. & Luo, Z.-X. Fossil evidence on origin of the mammalian brain. Science 332, 955–957 (2011).
Article ADS CAS PubMed Google Scholar
Grossnickle, D. M., Smith, S. M. & Wilson, G. P. Untangling the multiple ecological radiations of early mammals. Trends Ecol. Evol. 34, 936–949 (2019).
Article PubMed Google Scholar
Kaiuca, J. L. et al. Weighing in on miniaturization: new body mass estimates for Triassic eucynodonts and analyses of body size evolution during the cynodont-mammal transition. Anat. Rec. 307, 1594–1612 (2024).
Article Google Scholar
Grossnickle, D. M., Weaver, L. N., Jäger, K. R. & Schultz, J. A. The evolution of anteriorly directed molar occlusion in mammals. Zool. J. Linn. Soc. 194, 349–365 (2022).
Article Google Scholar
Brink, A. S. Two cynodonts from the Ntawere Formation in the Luangwa valley of Northern Rhodesia. Palaeontol. Afr. 8, 77–96 (1963).
Google Scholar
Hoffman, E. A. & Rowe, T. B. Jurassic stem-mammal perinates and the origin of mammalian reproduction and growth. Nature 561, 104–108 (2018).
Article ADS CAS PubMed Google Scholar
Gow, C. E. New find of Diarthrognathus (Therapsida: Cynodontia) after seventy years. Palaeontol. Afr. 31, 51–54 (1994).
Google Scholar
Botha-Brink, J., Soares, M. B. & Martinelli, A. G. Osteohistology of Late Triassic prozostrodontian cynodonts from Brazil. PeerJ 6, e5029 (2018).
Article PubMed PubMed Central Google Scholar
Martinelli, A. G. & Bonaparte, J. F. in Paleontología y Dinosaurios desde América Latina (eds Porfiri, J. D. & González Riga, B.) 179–186 (Editora de la Univ. de Cuyo, 2011).
Lautenschlager, S. Reconstructing the past: methods and techniques for the digital restoration of fossils. R. Soc. Open Sci. 3, 160342 (2016).
Article ADS MathSciNet PubMed PubMed Central Google Scholar
Goloboff, P. A. & Catalano, S. A. TNT version 1.5, including a full implementation of phylogenetic morphometrics. Cladistics 32, 221–238 (2016).
Article PubMed Google Scholar
Bapst, D. W. paleotree: an R package for paleontological and phylogenetic analyses of evolution. Methods Ecol. Evol. 3, 803–807 (2012).
Article Google Scholar
Abdala, F. et al. Non-mammaliaform cynodonts from western Gondwana and the significance of Argentinean forms in enhancing understanding of the group. J. South Am. Earth Sci. 104, 102884 (2020).
Article Google Scholar
Ronquist, F. et al. MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Syst. Biol. 61, 539–542 (2012).
Article PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We thank E. Martin-Silverstone, T. Davies, M. Day, V. Fernandez and B. Clark, A. Kallonen and I. Corfe for assistance and use of CT facilities for scanning the specimens used in this study. Exploratory segmentations and observations on some of the scans were carried out by C. Salcido, S. Holpin and C. Clark. We thank H. Francischini (UFRGS) and Silvia Tachestto (MPDC) for granting us access to Brasilodon and Riograndia fossils. Specimen photos were skilfully taken by L. Flávio Lopes (UFRGS). We thank F. Abdala (CONICET-Instituto Miguel Lillo) for providing comparative photos of Pachygenelus specimens. This work was funded by NERC (Natural Environment Research Council) grant NE/K01496X/1 to E.J.R. and P.G.G. J.R.G.R. is supported by the Biotechnology and Biological Sciences Research Council-funded South West Biosciences Doctoral Training Partnership (grant BB/T008741/1). The Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, Brazil, grants 307938/2019-0; 311251/2021-8; 406902/2022-4; 308515/2023-4) and Fundação de Amparo à Pesquisa do Estado do Rio Grande do Sul (FAPERGS 19/2551-0000719-1) supported M.B.S., C.L.S. and A.G.M. The Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ) supported M.B.S. (grant E-26/201.066/2021). A.G.M. was awarded the Benjamin Meaker Visiting Professorship of the University of Bristol and is also supported by CONICET-PIBAA 1137, PICT 2020-SERIEA-01498 and MILENIO NCN2023-025.

Author information

Authors and Affiliations

Palaeobiology Research Group, School of Earth Sciences, University of Bristol, Bristol, UK
James R. G. Rawson, Pamela G. Gill & Emily J. Rayfield
Museo Argentino Ciencias Naturales “Bernardino Rivadavia”—CONICET, Buenos Aires, Argentina
Agustín G. Martinelli
Natural History Museum, London, UK
Pamela G. Gill
Museu Nacional, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil
Marina B. Soares
Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil
Cesar L. Schultz

Authors

James R. G. Rawson
View author publications
You can also search for this author in PubMed Google Scholar
Agustín G. Martinelli
View author publications
You can also search for this author in PubMed Google Scholar
Pamela G. Gill
View author publications
You can also search for this author in PubMed Google Scholar
Marina B. Soares
View author publications
You can also search for this author in PubMed Google Scholar
Cesar L. Schultz
View author publications
You can also search for this author in PubMed Google Scholar
Emily J. Rayfield
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

The study was formulated by A.G.M., P.G.G. and E.J.R. μCT scanning was organized by A.G.M., P.G.G., M.B.S., C.L.S. and E.J.R. and segmentation and interpretation of the scans was done by J.R.G.R. Phylogenetic analyses were performed by A.G.M. and J.R.G.R. The manuscript was written by J.R.G.R., A.G.M., P.G.G and E.J.R. with input from all of the authors.

Corresponding authors

Correspondence to James R. G. Rawson, Agustín G. Martinelli or Emily J. Rayfield.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature thanks Shundong Bi and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 Reconstruction of the jaw of Brasilodon quadrangularis.

a, Medial view of the postdentary bones and left dentary of UFRGS-PV-1043-T, as preserved; b, lateral view of same; c, medial view of same showing postdentary elements moved into hypothesized life position, including mirrored elements; d, fully reconstructed posterior view of the articular process of the dentary and postdentaries; e, medial view of fully restored lower jaw including all repairs (see methods) and scaled coronoid from UFRGS-PV-628-T. Abbreviations as in main text plus: spl, splenial. Scale bars represent 10 mm or 2 mm (C).

Extended Data Fig. 2 Reconstruction of the jaw of Riograndia guaibensis.

a, Medial view of the right dentary and postdentary bones of UFRGS-PV-596-T, as preserved; b, medial view of same showing fragments of dentary and postdentary elements moved into hypothesized life position; c, medial view of the left jaw of UFRGS-PV-833-T, postdentary elements moved into hypothesized life position; d, posterior left jaw of UFRGS-PV-624-T in medial view, as preserved; e, medial view of same with coronoid fragments and postdentary bone fragments restored to hypothesized life position; f, medial view of fully restored right jaw including all repairs (see methods). All scale bars represent 10 mm.

Extended Data Fig. 3 Surface models of specimens of Oligokyphus major scanned in this study.

a, Medial view of NHMUK-PV-R7119, a right dentary; b, lateral view of same; c, medial view of NHMUK-PV-R7121, a partial right dentary; d, medial view of NHMUK-PV-R7373, a partial left dentary; e, medial view of NHMUK-PV-R7204, a right dentary; f, lateral view of same; g, medial view of NHMUK-PV-R7196, a right quadrate; h, posterior view of same; i, medial view of NHMUK-PV-R7189, left posterior postdentaries; j, lateral view of same; k, lateral view of NHMUK-PV-R1790, right posterior postdentaries. Anatomical abbreviations as in main text plus: cf, contact facet; cb, coronoid boss; csa, contact surface for the angular; csp, contact surface for the splenial; tcl, lateral trochlear condyle; tcm, medial trochlear condyle Scale bars represent 10 mm (a–f) or 1 mm (g–k).

Supplementary information

Supplementary Information

Supplementary Notes, Supplementary Figs. 1 and 2 and Supplementary Tables 1 and 2.

Reporting Summary

Peer Review File

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Rawson, J.R.G., Martinelli, A.G., Gill, P.G. et al. Brazilian fossils reveal homoplasy in the oldest mammalian jaw joint. Nature (2024). https://doi.org/10.1038/s41586-024-07971-3

Download citation

Received: 08 December 2023
Accepted: 20 August 2024
Published: 25 September 2024
DOI: https://doi.org/10.1038/s41586-024-07971-3

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.