Femoral specializations to locomotor habits in early archosauriforms

Abstract

The evolutionary history of archosaurs and their closest relatives is characterized by a wide diversity of locomotor modes, which has even been suggested as a pivotal aspect underlying the evolutionary success of dinosaurs vs. pseudosuchians across the Triassic–Jurassic transition. This locomotor diversity (e.g., more sprawling/erect; crouched/upright; quadrupedal/bipedal) led to several morphofunctional specializations of archosauriform limb bones that have been studied qualitatively as well as quantitatively through various linear morphometric studies. However, differences in locomotor habits have never been studied across the Triassic–Jurassic transition using 3D geometric morphometrics, which can relate how morphological features vary according to biological factors such as locomotor habit and body mass. Herein, we investigate morphological variation across a dataset of 72 femora from 36 different species of archosauriforms. First, we identify femoral head rotation, distal slope of the fourth trochanter, femoral curvature, and the angle between the lateral condyle and crista tibiofibularis as the main features varying between bipedal and quadrupedal taxa, all of these traits having a stronger locomotor signal than the lesser trochanter's proximal extent. We show a significant association between locomotor mode and phylogeny, but with the locomotor signal being stronger than the phylogenetic signal. This enables us to predict locomotor modes of some of the more ambiguous early archosauriforms without relying on the relationships between hindlimb and forelimb linear bone dimensions as in prior studies. Second, we highlight that the most important morphological variation is linked to the increase of body size, which impacts the width of the epiphyses and the roundness and proximodistal position of the fourth trochanter. Furthermore, we show that bipedal and quadrupedal archosauriforms have different allometric trajectories along the morphological variation in relation to body size. Finally, we demonstrate a covariation between locomotor mode and body size, with variations in femoral bowing (anteroposterior curvature) being more distinct among robust femora than gracile ones. We also identify a decoupling in fourth trochanter variation between locomotor mode (symmetrical to semi‐pendant) and body size (sharp to rounded). Our results indicate a similar level of morphological disparity linked to a clear convergence in femoral robusticity between the two clades of archosauriforms (Pseudosuchia and Avemetatarsalia), emphasizing the importance of accounting for body size when studying their evolutionary history, as well as when studying the functional morphology of appendicular features. Determining how early archosauriform skeletal features were impacted by locomotor habits and body size also enables us to discuss the potential homoplasy of some phylogenetic characters used previously in cladistic analyses as well as when bipedalism evolved in the avemetatarsalian lineage. This study illuminates how the evol...

The dinosaurian femoral head experienced a morphogenetic shift from torsion to growth along the avian stem

“…Instead, femoral head re-orientation must have been brought about by femoral torsion during ontogeny ( figure 5 c,d ). Empirically, femoral torsions can be partially observed in post-hatching stages [ 19 ] ( figure 4 i ). Given that the positional relationship was conservative in the incipient embryonic hindlimb skeleton (electronic supplementary material, SI.3.1), the extent of ontogenetic twisting reached a maximum in early coelurosaurs (peramorphism; figure 4 i ).…”

“…retroversion) during ontogeny of non-pygostylian theropods, as do the similarly acting muscles in human ontogeny [ 56 ]. This hypothesis explains homoplastic covariation from a morphogenetic viewpoint: the combination of femoral head retroversion and M.ITC -related skeletal changes appeared several times independently in major archosaurian clades [ 14 , 19 , 25 , 60 ]. We infer that the evolution of the M.ITC's configuration, consequent femoral twisting in ontogeny, and other musculo-skeletal changes accompanying them was functionally advantageous for (thus would have been ‘least-resisted’ by) their locomotory biomechanics (hip-driven erect posture) especially to theropods' plesiomorphic one (hip-driven bipedalism) (electronic supplementary material, SI.3.4).…”

“…Previous studies (e.g. [13,15,19,25,[39][40][41][42][43][44][45][46][47][48]) described the evolutionary transition of the femoral head orientation as follows. The femoral head of plesiomorphic saurians (grade 'I' in figure 4) had a small angle with the distal end reference (the intercondylar line).…”

A proposed standard for quantifying 3‐D hindlimb joint poses in living and extinct archosaurs

“…For example, geometric morphometric analyses of joint shape could be combined with JCS‐derived joint pose spaces (akin to Manafzadeh et al, 2021 ) to quantitatively relate articular geometry with higher level functional inferences in a phylogenetic context (e.g., Harcourt‐Smith et al, 2008 ). Landmark‐based analyses (e.g., Lawing & Polly, 2010 ; Mitteroecker & Gunz, 2009 ) might then pinpoint which specific features altered (or conserved/constrained) ACS/JCS parameters in archosaurs (Pintore et al, 2021 ). Such ambitious integrations could yield major advances in our understanding of fundamental form‐function relationships across development, evolution, and phenotypic plasticity (e.g., Simons et al, 2019 ).…”

“…We encountered this issue with several joints, but the ambiguity among distal femoral condyles of Marasuchus/Lagosuchus is illustrative (Figure 9g-i). As is typical of early archosaurs, the lateral condyle of the femur in Marasuchus/Lagosuchus is substantially smaller than that part which is homologous to the tibiofibular crest of birds (Nesbitt, 2011;Parrish, 1986;Pintore et al, 2021). Given that the relative sizes of the two eminences on the lateral distal femur progressively changed along the avian stem lineage, which should be selected for Marasuchus/Lagosuchus?…”