Evolutionary assembly and disassembly of the mammalian sternum.

Evolutionary transitions are frequently associated with novel anatomical structures,1 but the origins of the structures themselves are often poorly known. We use developmental, genetic, and paleontological data to demonstrate that the therian sternum was assembled from pre-existing elements. Imaging of the perinatal mouse reveals two paired sternal elements, both composed primarily of cells with lateral plate mesoderm origin. Location, articulations, and development identify them as homologs of the interclavicle and the sternal bands of synapsid outgroups. The interclavicle, not previously recognized in therians,2 articulates with the clavicle and differs from the sternal bands in both embryonic HOX expression and pattern of skeletal maturation. The sternal bands articulate with the ribs in two styles, most clearly differentiated by their association with sternebrae. Evolutionary trait mapping indicates that the interclavicle and sternal bands were independent elements throughout most of synapsid history. The differentiation of rib articulation styles and the subdivision of the sternal bands into sternebrae were key innovations likely associated with transitions in locomotor and respiratory mechanics.3,4 Fusion of the interclavicle and the anterior sternal bands to form a presternum anterior to the first sternebra was a historically recent innovation unique to therians. Subsequent disassembly of the radically reduced sternum of mysticete cetaceans was element specific, reflecting the constraints that conserved developmental programs exert on composite structures.

Convergent Evolution of Swimming Adaptations in Modern Whales Revealed by a Large Macrophagous Dolphin from the Oligocene of South Carolina.

Modern whales and dolphins are superbly adapted for marine life, with tail flukes being a key innovation shared by all extant species. Some dolphins can exceed speeds of 50 km/h, a feat accomplished by thrusting the flukes while adjusting attack angle with their flippers [1]. These movements are driven by robust axial musculature anchored to a relatively rigid torso consisting of numerous short vertebrae, and controlled by hydrofoil-like flippers [2-7]. Eocene skeletons of whales illustrate the transition from semiaquatic to aquatic locomotion, including development of a fusiform body and reduction of hindlimbs [8-11], but the rarity of Oligocene whale skeletons [12, 13] has hampered efforts to understand the evolution of fluke-powered, but forelimb-controlled, locomotion. We report a nearly complete skeleton of the extinct large dolphin Ankylorhiza tiedemani comb. n. from the Oligocene of South Carolina, previously known only from a partial rostrum. Its forelimb is intermediate in morphology between stem cetaceans and extant taxa, whereas its axial skeleton displays incipient rigidity at the base of the tail with a flexible lumbar region. The position of Ankylorhiza near the base of the odontocete radiation implies that several postcranial specializations of extant cetaceans, including a shortened humerus, narrow peduncle, and loss of radial tuberosity, evolved convergently in odontocetes and mysticetes. Craniodental morphology, tooth wear, torso vertebral morphology, and body size all suggest that Ankylorhiza was a macrophagous predator that could swim relatively fast, indicating that it was one of the few extinct cetaceans to occupy a niche similar to that of killer whales.

Finding sacral: Developmental evolution of the axial skeleton of odontocetes (Cetacea).

Axial morphology was dramatically transformed during the transition from terrestrial to aquatic environments by archaeocete cetaceans, and again during the subsequent odontocete radiation. Here, we reconstruct the sequence of developmental events that underlie these phenotypic transitions. Archaeocete innovations include the loss of primaxial/abaxial interaction at the sacral/pelvic articulation and the modular dissociation of the fluke from the remainder of the tail. Odontocetes subsequently integrated lumbar, sacral, and anterior caudal vertebrae into a single torso module, and underwent multiple series-specific changes in vertebral count. The conservation of regional proportions despite regional fluctuations in count strongly argues that rates of somitogenesis can vary along the column and that segmentation was dissociated from regionalization during odontocete evolution. Conserved regional proportions also allow the prediction of the location and count of sacral homologs within the torso module. These predictions are tested with the analysis of comparative pudendal nerve root location and geometric morphometrics. We conclude that the proportion of the column represented by the sacral series has been conserved, and that its vertebrae have changed in count and relative centrum length in parallel with other torso vertebrae. Although the sacral series of odontocetes is de-differentiated, it is not de-regionalized.

Breaking constraint: axial patterning in Trichechus (Mammalia: Sirenia).

Meristic variation is often limited in serially homologous systems with high internal differentiation and high developmental modularity. The mammalian neck, an extreme example, has a fixed (at seven) count of diversely specialized segments. Imposition of the mammalian cervical constraint has been tentatively linked to the origin of the diaphragm, which is muscularized by cells that migrate from cervical somites during development. With six cervical vertebrae, the genus Trichechus (manatee) has apparently broken this constraint, although the mechanism of constraint escape is unknown. Hypotheses for the developmental origin of Trichechus cervical morphology include cervical rib 7 repatterning, a primaxial/abaxial patterning shift, and local homeosis at the cervical/thoracic boundary. We tested predictions of these hypotheses by documenting vertebral morphology, axial ossification patterns, regionalization of the postcranial skeleton, and the relationship of thoracic ribs to sternal subunits in a large data set of fetal and adult Trichechus and Dugong specimens. These observations forced rejection of all three hypotheses. We propose alternatively that a global slowing of the rate of somitogenesis reduced somite count and disrupted alignment of Hox-generated anatomical markers relative to somite (and vertebral) boundaries throughout the Trichechus column. This hypothesis is consistent with observations of the full range of traditional cervical morphologies in the six cervical vertebrae, conserved postcranial proportions, and column-wide reduction in count relative to its sister taxon, Dugong. It also suggests that the origin of the mammalian cervical constraint lies in patterning, not in count, and that Trichechus and the tree sloths have broken the constraint using different developmental mechanisms.

Crossing the frontier: a hypothesis for the origins of meristic constraint in mammalian axial patterning.

Serially homologous systems with high internal differentiation frequently exhibit meristic constraints, although the developmental basis for constraint is unknown. Constraints in the counts of the cervical and lumbosacral vertebral series are unique to mammals, and appeared in the Triassic, early in their history. Concurrent adaptive modifications of the mammalian respiratory and locomotor systems involved a novel source of cells for muscularization of the diaphragm from cervical somites, and the loss of ribs from lumbar vertebrae. Each of these innovations increased the modularity of the somitic mesoderm, and altered somitic and lateral plate mesodermal interactions across the lateral somitic frontier. These developmental innovations are hypothesized here to constrain the anteroposterior transposition of the limbs along the column, and thus also cervical and thoracolumbar count. Meristic constraints are therefore regarded here as the nonadaptive, secondary consequences of adaptive respiratory and locomotor traits.

Anatomy and age estimation of an early blue whale (Balaenoptera musculus) fetus.

The external anatomy of a 130-mm blue whale fetus (Balaenoptera musculus) is described, and its internal anatomy is reconstructed noninvasively from microCT scans. The specimen lies developmentally at the junction of the embryonic and fetal periods. Similarly to the embryos of many odontocetes, it lacks a caudal fluke and dorsal fin, but it also exhibits an elongated rostrum, resorbed umbilical hernia, partially exposed cornea, and spatial separation of the anus and genitalia seen in early odontocete fetuses. Dermal ossification of the cranial bones has begun, but the endochondral skeleton is completely cartilaginous. The shape and position of the maxilla suggest that the earliest stages of anterior skull telescoping have begun, but there is no indication of occipital overlap posteriorly. The nasopharynx, larynx, and heart already display the distinctive morphology characteristic of Balaenoptera. This study develops a model of body length changes during blue whale development by integrating the large International Whaling Statistics (IWS) database, historical observations of blue whale migration and reproduction, and descriptions of fetal growth trends in other mammals. The model predicts an age of 65 days postconception for the specimen. The early developmental milestones of Balaenoptera mirror those of the odontocete Stenella to a remarkable extent, but the first appearance of the caudal fluke and dorsal fin are delayed relative to other morphological transitions. The accelerated prenatal growth characteristic of Balaenoptera occurs during fetal, not embryonic, development.

Fixed cervical count and the origin of the mammalian diaphragm.

Why is mammalian cervical count fixed across the historically long and ecologically broad mammalian radiation? Multiple lines of evidence, considered together, suggest a link between fixed cervical count and the muscularization of the diaphragm, a key innovation in mammalian history. We test this hypothesis by documenting the anteroposterior (AP) movement of the diaphragm, a lateral plate derivative, relative to that of the somitic thoracolumbar transition in unusually patterned mammals, by comparing the temporal occurrence of an osteological proxy for the diaphragm and fixed cervical counts in the fossil record, and by quantifying morphological differentiation within the mammalian cervical series. We then integrate these anatomical observations with details of diaphragm function and development to propose a sequence of innovations in mammalian evolution that could have led to fixed cervical count. We argue that the novel commitment of migratory muscle precursor cells (MMPs) of mid-cervical somites to a fate in the abaxial diaphragm defined these somites as a new and uniquely mammalian modular subunit. We further argue that the coordination of primaxial abaxial patterning constrained subsequent AP migration of the forelimb, thereby secondarily fixing cervical count.

Anatomical transformation in mammals: developmental origin of aberrant cervical anatomy in tree sloths.

Mammalian cervical count has been fixed at seven for more than 200 million years. The rare exceptions to this evolutionary constraint have intrigued anatomists since the time of Cuvier, but the developmental processes that generate them are unknown. Here we evaluate competing hypotheses for the evolutionary origin of cervical variants in Bradypus and Choloepus, tree sloths that have broken the seven cervical vertebrae barrier independently and in opposite directions. Transitional and mediolaterally disjunct anatomy characterizes the cervicothoracic vertebral boundary in each genus, although polarities are reversed. The thoracolumbar, lumbosacral, and sacrocaudal boundaries are also disrupted, and are more extreme in individuals with more extreme cervical counts. Hypotheses of homologous, homeotic, meristic, or associational transformations of traditional vertebral column anatomy are not supported by these data. We identify global homeotic repatterning of abaxial relative to primaxial mesodermal derivatives as the origin of the anomalous cervical counts of tree sloths. This interpretation emphasizes the strong resistance of the "rule of seven" to evolutionary change, as morphological stasis has been maintained primaxially coincident with the generation of a functionally longer (Bradypus) or shorter (Choloepus) neck.

Vertebral anatomy in the Florida manatee, Trichechus manatus latirostris: a developmental and evolutionary analysis.

The vertebral column of the Florida manatee presents an unusual suite of morphological traits. Key among these are a small precaudal count, elongate thoracic vertebrae, extremely short neural spines, lack of a sacral series, high lumbar variability, and the presence of six instead of seven cervical vertebrae. This study documents vertebral morphology, size, and lumbar variation in 71 skeletons of Trichechus manatus latirostris (Florida manatee) and uses the skeletons of Trichechus senegalensis (west African manatee) and Dugong dugon (dugong) in comparative analysis. Vertebral traits are used to define morphological, and by inference developmental, column modules and to propose their hierarchical relationships. A sequence of evolutionary innovations in column morphology is proposed. Results suggest that the origin of the fluke and low rates of cervical growth originated before separation of trichechids (manatees) and dugongids (dugongs). Meristic reduction in count is a later, trichechid innovation and is expressed across the entire precaudal column. Elongation of thoracic vertebrae may be an innovative strategy to generate an elongate column in an animal with a small precaudal count. Elimination of the lumbus through both meristic and homeotic reduction is currently in progress.

Modular evolution of the Cetacean vertebral column.

Modular theory predicts that hierarchical developmental processes generate hierarchical phenotypic units that are capable of independent modification. The vertebral column is an overtly modular structure, and its rapid phenotypic transformation in cetacean evolution provides a case study for modularity. Terrestrial mammals have five morphologically discrete vertebral series that are now known to be coincident with Hox gene expression patterns. Here, I present the hypothesis that in living Carnivora and Artiodactyla, and by inference in the terrestrial ancestors of whales, the series are themselves components of larger precaudal and caudal modular units. Column morphology in a series of fossil and living whales is used to predict the type and sequence of developmental changes responsible for modification of that ancestral pattern. Developmental innovations inferred include independent meristic additions to the precaudal column in basal archaeocetes and basilosaurids, stepwise homeotic reduction of the sacral series in protocetids, and dissociation of the caudal series into anterior tail and fluke subunits in basilosaurids. The most dramatic change was the novel association of lumbar and anterior caudal vertebrae in a module that crosses the precaudal/caudal boundary. This large unit is defined by shared patterns of vertebral morphology, count, and size in all living whales (Neoceti).