This dissertation's primary objective was to test for heterochronic patterns and processes, specifically paedomorphism and neoteny within a putative ancestral-descendant pair the wolf, Canis lupus and the dog, Canis familiaris . In addition, patterns of developmental and morphological integration were investigated. Three dimensional landmark-based data were collected on an ontogenetic series of over 700 crania belonging to Canis familiaris (100 breeds) and 310 Canis lupus. Morphometric data analysis was accomplished by the following methods: cranial shape centroid and geometric mean differences, ontogenetic allometric regression, common principal component analysis (Flury, 1988), shared correlation matrix structure (SCMS), and Heterochronic Prediction and Diagnosis (Williams, 2001, Williams et al., 2003). Heterochronic processes, including neoteny, do not describe the evolution of most dog breeds examined, although adults of a few breeds are paedomorphic, meaning they exhibit cranial morphology similar to juvenile wolves. However, there is a wide diversity of skull shapes within the globally paedomorphic breeds. Not all dog breed crania conform to the predicted cranial shape expected from a heterochronically derived descendant of Canis lupus. Some breeds' crania are indeed morphological snapshots from wolf ontogeny, having the same shape and size as a juvenile or an adult wolf. Other breeds may have a cranial shape similar to a particular stage in the wolf's ontogeny but, because of the dissociation of size and shape, their crania are not similar in size to those of the wolf at that ontogenetic stage. Still other breeds are neomorphic, i.e. similar to no stage along the wolf ontogenetic trajectory. The evolution of Afghans, Appenseller Sennenhunds, Bernese Mountain Dogs, German Shepherds, Labradors and Newfoundlands is best modeled via the heterochronic process proportioned dwarfism. Chihuahuas are the only small breed well modeled by a heterochronic process of proportioned dwarfism. St. Bernards as well as Fox Terriers, French Bulldogs, and Pugs were not well described by any heterochronic model. During canid ontogeny, pattern and intensity of integration among cranial elements varies and adult and neonatal patterns are each unique. Ancestral patterns of developmental integration do not predict Canis familiaris integration.

To separate functional from phylogenetic pelvic differences in Aves, seventy-seven birds were measured and statistically analyzed to define the characters that unite them into function groups, and to identify the characters that separate the function groups from each other. Four Major Function Groups were identified: Birds of Prey (Chapter II) Waterbirds (Chapter III), Ground birds and Stalker-wader birds (Chapter V). Birds of Prey statistically separate into Air predators and Scavengers. Waterbirds statistically divide into Surface swimmers, Wing-propelled divers, and Foot-propelled divers. A suite of pelvic characters (Level One indicators) can separate nonpasseriformes birds into function groups. These are: length to width ratio of the pelvis, iliac angle, relative iliac surface areas and pre-acetabular ilium proportions. A second suite of characters (Level Two indicators), used by themselves, shows trends, but do not completely separate the function groups. These are proportional length of lower leg components, and femur length: pelvis length ratio. Combining the identified characters in Principal Component Analysis and Discriminant Function Analysis produces a good split among the function groups. When all of the birds are combined into a single analysis (Chapter V), using Level one and Level Two characters, the plots generated have Stalker-wader birds at the center, with three groups radiating out. One arm is Ground Birds, another is Waterbirds, and the third is Birds of Prey. Pelvic musculature dissections of a representative from each group are compared (Chapter VI). The picture drawn from the muscle data is different. The muscle data suggest that Ground Birds are the central group. Ground Birds (represented by a pigeon) have leg/pelvic muscles neither specialized for specific femoral movements, nor for specific tibiotarsal movements. No particular group of muscles is hypertrophied, and no major groups are reduced or missing. The other three function groups of birds show some degree of specialization, either by reducing or accentuating muscles, or by losing muscles completely. The result is a statistical portrait of four bird function groups that has implications for the sequence of morphological changes seen in avian radiations of the Tertiary.

The skeletal elements of cartilaginous fishes are composed of a thin layer of mineralized tissue, ‘prismatic cartilage’, overlaying a hyaline cartilage core. Cartilage, even with a surface layer of mineralization, is far less stiff and strong than bone. Nevertheless, several species of stingray, including Rhinoptera and Aetobatus, subsist by crushing hard-shelled mollusks and crustaceans in their cartilaginous jaws. The jaws of these stingrays are composed of a previously undescribed form of cartilage. This tissue, ‘trabecular cartilage’, has mineralized struts which run through the central hyaline core. The struts, or trabeculae, are hollow tubes, made of calcified blocks, arranged as in a brick chimney. They serve to prevent buckling and bending of the jaws while prey is being crushed. The struts are present in late term embryos, indicating that feeding on hard prey does not cause them to form. As the animal grows the struts lengthen and thicken though they do not appear to become more numerous. Trabecular cartilage appears to have evolved at the base of the clade containing the hard prey specialists. This clade also includes Manta, a planktivorous species, which retains trabecular cartilage.

The newly resurgent field of evolutionary developmental biology integrates the study of evolutionarily important anatomical changes and developmental biology to describe the genetic and developmental changes that have led to anatomical changes. In this dissertation I describe candidate developmental mechanisms in the context of axial elongation and pectoral fin musculature evolution in fishes. Both axial elongation and increase in pectoral fin muscle subdivisions have important ecological correlates. Elongate fishes tend to be found in highly structured environments and fishes with an increased number of fin muscles tend to use fin-based locomotion to swim. Both of these morphologies have evolved multiple times within the ray-finned fish radiation. In Chapter 1 I focus on a specific predator avoidance behavior that is only seen in elongate fishes. Deeper-bodied fish tend to perform a unilateral bend of the body and swim away whereas elongate fish, when startled, bend bilaterally and hide. Although all elongate fishes perform head retraction, their specific anatomy indicates multiple explanations for how elongation occurs. In Chapter 2, I describe changes in the vertebral column in lineages of actinopterygian fishes that have elongate members. Elongation occurs through three different mechanisms: addition of abdominal vertebrae, addition of caudal vertebrae, and lengthening of all the vertebral cents. This study suggests that the number of abdominal vertebrae, number of caudal vertebrae, and length of the vertebral cents are controlled by separate developmental modules. Fin-based locomotion has evolved multiple times independently within actinopterygian fishes and is correlated with a modification of the pectoral fin musculature. In order to increase the understanding of fin muscle development, in Chapter 3 I describe the wildtype anatomy and embryology of the pectoral fin musculature in the zebrafish, Danio rerio. Zebrafish have six muscles in the pectoral fin. Early in development, the fin musculature consists of two muscle masses, one on each side of the fin. The arrector ventralis is the first muscle to individuate from the initial abductor muscle mass, and the adult musculature is present by 3 weeks postfertilization. This study provides a basic understanding of the embryology of the fin muscles and will provide a baseline for examining mutant fin muscle morphologies in zebrafish and diverse fin muscle morphologies in other species.

Phylogenetic relationships of chimaeroid fishes are investigated in detail. The six genera studied, Callorhinchus, Rhinochimaera, Harriotta, Neoharriotta, Chimaera and Hydrolagus belong to the subclass Chimaeroidei and are the only living representatives of the class Holocephali. The comparative morphology of the lateral line canals, skeleton, tooth plates, secondary sexual characteristics and musculature of all six living genera of chimaeroid fishes is described. Development of the jaws, hyoid arch and ethmoid canal is briefly described for Callorhinchus milii of the family Callorhynchidae. Using anatomical characters a phylogeny of higher level chimaeroid relationships is hypothesized. The infraorder Neochimaeroidi is erected to include Rhinochimaera, Harriotta, Neoharriotta, Chimaera and Hydrolagus on the basis of eight synapomorphies and a new classification of chimaeroid fishes is proposed.

Most recent phylogenetic analyses of snakes have recognized two major clades within Serpentes: Alethinophidia and Scolecophidia. Alethinophidians feed predominantly on relatively large vertebrate prey, which they transport into and through the mouth via reciprocating ratcheting movements of the toothed palatopterygoid jaw arches. In contrast, scolecophidians are small-prey specialists, feeding almost exclusively on small arthropods. In addition, these diminutive, fossorial snakes lack many of the key morphological features which underlie the feeding mechanisms of alethinophidians, such as toothed palatopterygoid jaw arches and a distensible lower jaw. However, the functional significance of these morphological differences has remained poorly understood because there have been no detailed descriptions of feeding behavior in Scolecophidia. I used magnified high-speed videography, videofluoroscopy, and standard histological and gross morphological preparations to study the functional morphology of the feeding apparatus in representatives of two families of Scolecophidia, Leptotyphlopidae and Typhlopidae. In Leptotyphlops (Leptotyphlopidae), a mandibular raking mechanism is used to capture, ingest and transport prey. In this mechanism, the toothed anterior portions of the mandibular rami are rotated medially about the intramandibular joints in a bilaterally synchronous fashion. In contrast, Typhlops and Rhinotyphlops (Typhlopidae) feed via a maxillary raking mechanism, in which asynchronous rotations of the toothed maxillae are used to drag prey into and through the mouth. Both mandibular raking and maxillary raking involve exceptionally rapid (3–5 Hz) movements of the tooth-bearing elements of the jaws, thereby facilitating the ingestion of large numbers of small prey within relatively brief periods of time.

Canids have long been considered to be conservative in their postcranial anatomy, so there are few studies examining individual canid taxa for locomotor adaptations. Canids are generally considered to be the most cursorial of the carnivorans. The limbs of large canids are generally adapted for rapid terrestrial locomotion, as these animals frequently rely on speed for prey capture. The prey animal is captured and killed using the jaws and teeth. Smaller canids, such as the red fox Vulpes and gray fox Urocyon, do not use their limbs primarily for fast running. The red fox appears to have many adaptations for running, including long slender legs, but these foxes do not run in their daily activities except when chased. The red fox uses its forelimbs to help in prey capture and its hind legs for leaping. The gray fox is an unusual canid since it regularly climbs trees. The limbs of the gray fox, especially the forelimb, are utilized in climbing. This dissertation contains a detailed description of the postcranial osteology and myology Vulpes and Urocyon cinercoargenteus and includes an analysis of these anatomical features in a functional framework. An examination of both the osteology and myology of the fore and hind limbs of these two foxes reveals that their behavior is reflected in a number of anatomical characters. Adaptations for leaping in the red fox include the presence of unusually long hind legs relative to the front legs, and an increase in the length of the distal bony limb elements relative to more proximal ones. In addition, the limb bones are very slender. Muscle bellies of tarsal and digital flexors and extensors are restricted to a proximal position on the limb, and muscles in general are emphasized that act along the long axis of the limbs. Adaptations of the gray fox for climbing include the presence of relatively short legs, a greater ability to rotate the radius on the ulna relative to other canids, and a relatively greater ability to abduct the hind limb. In addition, both red and gray foxes are able to retract their claws, an ability that is not generally associated with canids.

Most larval and permanently aquatic salamanders use undulatory swimming as their primary mode of steady aquatic locomotion. These swimming movements are powered by the segmented axial musculature. The hypaxial region of each segment consists of distinct muscle layers, which have a simple planar geometry and have varying architectural features. In an aquatic salamander Siren lacertina, the morphological features of the lateral hypaxial layers allow the shortening of muscle fibers to be amplified during contraction. The angled muscle fibers in these layers function to allow fiber shortening to be accompanied by substantial rotation of muscle fibers during contraction. The connective tissue sheets separating adjacent muscle segments (myosepta), allow the segment to bulge in a way that further amplifies muscle fiber rotation. The combined effect of architectural and moment arm variation ensures that muscle fibers from different layers undergo similar shortening patterns during swimming to allow for the generation of optimal tension during locomotion. In addition to steady swimming many salamanders respond to a predatory stimulus by performing a “C-start” aquatic escape response. This unsteady maneuver involves two kinematic stages, which function to propel the salamander away from the perceived threat. During metamorphosis, the tailfin of salamanders is resorbed and is thought to result in a substantial decrease in escape performance. However, in a stream salamander Eurycea bislineata , adults spend significant time in the water and behaviorally compensate for metamorphic changes in tail morphology by increasing the amplitude of escape responses. Aquatic locomotion in salamanders is not limited to axial swimming. Some salamanders also utilize their limbs to move along the substrate at slow speeds, while submerged. Structures used during aquatic walking face dramatically different mechanical loads compared to limbs used on land. The greater hydrodynamic resistance associated with water lowers the effective weight and can act to stabilize an organism throughout its gait. Therefore structures, such as the reduced limbs of S. lacertina, which would be considered ineffective on land, can be fully functional during aquatic walking.