Skeletal remodelling suggests the turtle's shell is not an evolutionary straitjacket

Abstract

Recent efforts to decipher the enigma of the turtle's shell revealed that distantly related turtle species deploy diverse processes during shell development. Even so, extant species share in common a shoulder blade (scapula) that is encapsulated within the shell. Thus, evolutionary change in the correlated development of the shell and scapula probably underpins the evolution of highly derived shell morphologies. To address this expectation, we conducted one of the most phylogenetically comprehensive surveys of turtle development, focusing on scapula growth and differentiation in embryos, hatchlings and adults of 13 species. We report, to our knowledge, the first description of secondary differentiation owing to skeletal remodelling of the tetrapod scapula in turtles with the most structurally derived shell phenotypes. Remodelling and secondary differentiation late in embryogenesis of box turtles (Emys and Terrapene) yielded a novel skeletal segment (i.e. the suprascapula) of high functional value to their complex shell-closing system. Remarkably, our analyses suggest that, in soft-shelled turtles (Trionychidae) with extremely flattened shells, a similar transformation is linked to truncated scapula growth. Skeletal remodelling, as a form of developmental plasticity, might enable the seemingly constrained turtle body plan to diversify, suggesting the shell is not an evolutionary straitjacket.

Is the Pelvis Sexually Dimorphic in Turtles?

“…In theory, regionalized pelvic variation is plausible, because the highly‐integrated nature of the tetrapod skeleton permits evolutionary or plastic change in traits that directly or indirectly participate in key organismal functions (Cordero and Quinteros, ; Cordero and Berns, ). For instance, females of the extremely small (mean CL: 92.6 mm) tortoise, Homopus signatus , feature a kinetic pelvis that enables expansion of the pelvic aperture during passing of remarkably large eggs (mean volume: 11.34 cm 3 ) (Hofmeyr et al, ).…”

The postembryonic transformation of the shell in emydine box turtles

“…We also sampled hatchling specimens of Emys orbicularis ( N = 24) stored in the Centro de Conservación de Especies Dulceacuícolas de La Comunitat Valenciana (CCEDCV), Spain. Hatchlings from the ISU collection originated from eggs incubated in controlled laboratory conditions to account for potential environmental effects on hatchling morphology (Cordero & Janzen, ; Cordero & Quinteros, ; Neuman‐Lee & Janzen, ). Hatchlings from the Carnegie collection also originated from eggs incubated in similarly controlled laboratory conditions (Ewert & Nelson, ).…”

“…Active plastral kinesis is enabled by modified muscle connections and articulations of the limb girdles, as defined by Pritchard (), is the most common type of shell kinesis in living turtles and it is readily recognizable by the presence of a flexible hinge joint that enables elevation of part of the plastron in response to potential predators (Agassiz, ; Bramble, ; Pritchard, , ). Neck and limb girdle musculoskeletal specializations, which begin to develop in embryos, act as biomechanical linkages that power kinesis as turtles grows (Bramble, ; Cordero & Quinteros, ; Cordero et al, ). Consequently, the kinetic plastral hinge develops gradually in juveniles (Cordero et al, ; Legler, ; Richmond, ).…”

“…Reduced buttresses are already apparent in hatchlings (Cordero et al, ; Legler, ). Additional skeletal alterations, such as a cervico‐plastral ligament and segmented scapulae, necessary for plastral kinesis to function are prepatterned in embryos (Bramble, ; Cordero & Quinteros, ). Thus, the subsequent differentiation of a plastral hinge in juvenile emydine box turtles is a delayed ontogenetic response to musculoskeletal changes that arise earlier in embryonic development (Cordero, ).…”

Reptile embryos are not capable of behavioral thermoregulation in the egg

“…For C. serpentina , we highlight the motile stages (Decker, ) and the period when offspring sex is sensitive to temperature (Yntema, ). Data for embryo body size were derived from museum specimens described in Cordero and Quinteros () for C. serpentina and from the Paleontological Collection of the University of Tübingen (Germany) for P. sinensis . We also provide diagrams of C. serpentina embryos with their extraembryonic membranes and yolk sac at key developmental stages (adapted from Agassiz, , Figure bottom) and photographs of P. sinensis embryos at multiple stages depicting size relative to the egg (Figure b) and developmental progress at stages 14 and 17 (Figures c and 3d), which corresponds to when thermal taxis was measured in Du et al ().…”