More than 35,000 fish species dominate diverse aquatic settings due, in part, to the evolution of functional innovations. Such traits enable novel interactions with the environment, often promoting increased diversification within the lineage that it evolved. As several of these innovative traits are morphological modifications that impact prey capture and manipulation, the feeding apparatus is a rich system to examine the macroevolutionary impacts of functional novelties. My dissertation leverages functional morphology, kinematics, and phylogenetic comparative methods to explore how key functional innovations, specifically oral jaw protrusion and a major pharyngeal jaw modification, have affected the evolution and function of the fish feeding system.

Most fishes use suction feeding to capture their food, expanding the oral cavity to draw in water and prey. Protrusion of the oral jaws, an innovation among ray-finned fishes, is essential to this feeding mechanism as it increases flow forces into the mouth. Though a common mechanism in suction feeding fishes, the magnitude and direction of jaw protrusion vary greatly phylogenetically. In my first chapter, I explored the anatomical basis of oral jaw protrusion directionality in ponyfishes (Leiognathidae), a family comprised of 52 species that exhibit striking diversity in this functional trait. Using cleared and stained specimens of 20 ponyfish species representing every major lineage within the family, I measured eight traits associated with the size and positioning of oral jaw bones. Species fell into three nonoverlapping clusters with respect to directionality including dorsal, rostral, and ventral protruders. I found that the anterior–posterior position of the articular-quadrate jaw joint is a key correlate of protrusion direction. As the joint position moves from a posterior to a more anterior location, the orientation of the relaxed mandible rotates from an almost horizontal resting position to an upright vertical posture. Thus, the abduction of the mandible from the horizontal position results in ventrally directed protrusion, while the more upright mandible rotates to a position that maintains dorsal orientation. The position of the lower jaw joint may be a widespread anatomical mechanism for adjusting the directionality of jaw protrusion in ray-finned fishes, but it remains to be seen whether other groups have evolved different traits to facilitate this underappreciated aspect of diversity in fish feeding functional morphology.

In addition to the functional capacity of the oral jaws, fishes also have pharyngeal jaws – a second jaw system located posterior to the mouth cavity that chews, tears, and/or crushes prey before swallowing. This prey capture and processing sequence is generally conserved across suction feeding fishes. However, some species have evolved a novel pharyngeal jaw configuration, termed “pharyngognathy”, where several musculoskeletal modifications enable fishes to process more robust prey. Thus, the MPJ is widely considered a key innovation as it enhanced trophic diversification within several prominent fish clades. Seen in cichlids, damselfishes, wrasses, and a few other groups, these musculoskeletal alterations are believed to have increased the evolutionary independence of the oral and pharyngeal jaw systems, enabling greater diversification of the feeding system. Though aspects of this hypothesis have received considerable attention, we lack comparative analyses contrasting the evolutionary patterns and feeding functions in species with and without the MPJ. Thus, we lack a complete understand of how this novel trait impacts the fish feeding apparatus overall.

In my second chapter, I conducted comparative phylogenetic analyses of integration, disparity, and rate of evolution on feeding-related, skeletal structures in 85 Neotropical cichlid species and 30 North American centrarchid species, which lack the specialized pharyngeal jaw. Contrasting evolutionary patterns in these two continental radiations, I tested a classic decoupling hypothesis. Specifically, I asked whether the modified pharyngeal jaws in cichlids resulted in enhanced evolutionary independence of the oral and pharyngeal jaws, leading to increased morphological diversification of feeding structures. Contrary to this prediction, I found significantly stronger evolutionary integration between the oral and pharyngeal jaws in cichlids compared to centrarchids, although the two groups did not differ in patterns of integration within each jaw system. Further, though I found no significant differences in disparity, centrarchids showed faster rates of morphological evolution. These results suggest that while the modified pharyngeal jaw enhances feeding performance, the novelty does not prompt increased morphological diversification within the feeding apparatus, as has long been thought.

In my third chapter, I conducted comparative phylogenetic analyses to assess the effects of the pharyngeal jaw novelty on diversification of feeding morphology and kinematics across a phylogenetically diverse sample of spiny-rayed fishes. I quantified movements of the oral jaws and other craniofacial structures during suction feeding strikes using 689 high speed videos collected from 228 species with and without the modified pharyngeal jaw. In line with my Chapter 2 findings, I found significantly greater disparity in oral jaw functions and craniofacial shapes in fishes without the MPJ. Further, fishes lacking the specialized trait showed faster rates of oral jaw functional evolution. In light of these findings in my 2nd and 3rd chapters, a reinterpretation of the macroevolutionary consequences of the MPJ is urgently needed. The modified pharyngeal jaw is undoubtedly a functional innovation as it enhances the strength of the prey processing system, facilitating exceptional transition rates to feeding on hard and tough prey. However, the novelty also restricts the overall diversification of the feeding system, revealing that the impact of this novel trait is more nuanced than previously thought.

Though environments change across space and time, living organisms must always find ways to succeed under the mechanical constraints posed by the world around them. The nearly 25,000 species of teleost fishes inhabit nearly every imaginable aquatic habitat and have amazing diversity of shapes, sizes, and color patterns. Yet all fishes must eat, and almost every living teleost fish is capable of using ‘suction feeding’ for prey capture. Suction is a dynamic feeding mechanism that leverages the density and viscosity of water to pull in water and a prey item via rapid expansion of the head. But there are many common prey, like algae, sponges, or coral, that are not easily suctioned in by a hungry critter: and thus, some fishes use ‘biting’ to graze, scrape, or yank prey from an attachment to the substrate. My dissertation explores how the interactions between animals and their physical world shape evolutionary processes over millions of years, seeking to understand the drivers of diversity in form and function.

In my first chapter, I explored the potential for a fundamental mechanical trade-off to constrain the evolution of biters. Fishes that use biting for prey capture can also use suction, but this may create a trade-off between cranial expansion for suction and force transmission for biting. I studied how this mechanical trade-off affects diversification of both head shape and feeding kinematics. I filmed high-speed videos of feeding strikes of fishes of both groups, then used geometric morphometrics to calculate the amount of motion of the anatomy of the head during feeding, which I refer to as “cranial mobility.” I then compared rates of evolution of cranial mobility and of head shape between biters and suction feeders. I demonstrated that the trade-off results in less diversity of kinematics, which evolve more slowly in biters, but accelerates evolution of morphology, as biters evolve head shape more rapidly than suction feeders. These results show the potential for trade-offs to provide differing evolutionary constraints on the evolution of morphology and kinematics, and to create a mismatch between the adaptive landscapes of form and function.

In my second chapter, I reconstructed the history of feeding mode among 1,530 species of reef fishes and found that prior to the end-Cretaceous mass extinction, over 96% of reef-associated lineages were suction feeders. But I found that there has been a trophic revolution among fishes since then, and benthic grazers make up fully 40% of modern reef fish species diversity. Alongside ecological shifts in the structure of reefs, innovations for improved biting by fishes at the dawn of the Cenozoic may have provided opportunities for biters to thrive and resulted in a trophic revolution among reef fishes. Using body shape data for all 1,530 species, I showed that benthic grazers are evolving body shapes 1.7x faster than suction feeders. These results suggest that benthic grazing may inherently elevate evolutionary potential by providing access to attached prey with diverse functional properties. This study demonstrates that there has been a transformation the trophic makeup of reef fishes in the last 65 million years, and that biting has been a major contributor to the ecological and phenotypic diversity of reefs.

In my third chapter, I leveraged body shape data from 5,940 species of teleost fishes across 392 families from a dataset that I helped collect at the National Museum of Natural History to understand evolution of extreme body shapes in fishes. Fishes range from pancake-like, dorsoventrally flattened goosefishes to species that have been laterally compressed into a disc-like shape, such as surgeonfishes, and even to slender, elongate eels. Yet fishes of all these shapes are subject to the same physical constraints of life underwater. How, then, is it possible for some lineages to “break the rules” and evolve extreme body shapes? One intuitive possibility is that fishes are concentrated in shape space because it is hard to evolve or function with these odd shapes, and so evolution towards extreme shapes is a long, slow battle against the physics of aquatic life. Yet this work found that species further in shape space from the average body shape are evolving more rapidly, suggesting that the evolution of an extreme body shape requires the release of the constraints on the formation of a typical body shape, such as a major developmental or ecological shift, likely paired with a major shift in adaptive zone. Such a shift in adaptive zone may result in dramatic evolution of body shapes and allow the evolution of extreme or unique body shapes.

Coral reef fishes represent one of the most morphologically and functionally diverse vertebrate groups. Reef fishes contain more than 70 families of ray-finned fishes including speciose groups that are most abundant on reefs like wrasses, butterflyfish, damselfish, surgeonfish, and angelfish. However, also present in the reef fish assemblage are species with unique morphologies and highly specialized ecologies like frogfishes, pufferfish, triggerfish, and moray eels. Reef fish occupy all levels of the trophic pyramid from detritivores and algivores, microbial specialists, planktivores, invertivores, and carnivores. The trophic diversity is matched by diversity in habitat usage from species which are highly coral associated to those that exist exclusively in the open water above and near the reef. Further, reef fish are highly variable in their locomotor styles with species that swim exclusively using their body and caudal fin undulation to those that use numerous combinations of the median and paired fins for propulsion. Broadly, my dissertation uses phylogenetic comparative methods to explore the relationships between morphological, behavioral, and functional diversity in coral reef fishes, and to determine the evolutionary pathways reef fish have taken to diversify into novel conditions.

It is widely believed that because of biomechanical trade-offs, fish body shape and the mode of propulsion are strong predictors of swimming performance, with the best cruisers, maneuverers, and accelerators having different body forms and emphasizing different propulsion mechanisms. This paradigm is regularly projected onto routine swimming behavior and dominates the ecomorphological literature, despite the paucity of field measurements. In my first chapter, I measure variation in swimming behavior among 48 species of Indian Ocean coral reef fishes using recordings from a remote stereo video system. I measured average swimming speed, average swimming bout distance, frequency of turns, and percent of time spent station-holding and looked for the predicted trade-offs between them. I find little evidence of the expected relationships between swimming behaviors across species, little evidence that body shape affects swimming, and few differences between species that swim by undulating the body and those that emphasize the use of median and paired fins. Taxa widely thought of as archetypical maneuverers (Chaetodon) and cruisers (Caranx) were not outliers in any behaviors. My results indicate that swimming behavior is not easily predicted from simple measures of body shape and that alternative swimming modes can produce comparable behavioral profiles.

In my second chapter I further explore swimming diversity, though in this chapter I emphasize the effects of locomotor mode on fin shape evolution. In complex functional systems composed of many traits, selection can induce trait evolution by acting directly on individual components within the system, or indirectly through networks of trait integration. As a result, tension exists between the functional and developmental causes of trait integration and the capacity for diversification. I explore this dynamic in the evolution of fin shapes in 106 species from 38 families of coral reef fishes, a polyphyletic assemblage that shows exceptional diversity in locomotor function. Despite strong shared developmental pathways and expectations of a strong match between form and function, I find that species that share locomotor mode show substantial disparity in fin shape, and preferred locomotor mode is a poor predictor of fin shape. The evolution of fin shape is weakly integrated across the four functionally dominant fins and is weakened during transitions to derived swimming modes. As fin evolution is not strongly integrated there is substantial off-axis independent diversification, repeating a pattern found in many vertebrate systems. My study highlights the need for additional work on the functional consequences of fin shape in fishes and the diversifying impact of functions other than locomotion.

Trait integration and modularity have both been shown to be potent mechanisms in morphological evolution. While integration can provide access to morphologically novel phenotypes at the endpoints of axes of trait covariation, extreme phenotypes can also manifest via modular evolution producing novel trait combinations. It is unclear what the relative contributions of these two pathways are in accounting for the origins of novel or extreme forms. In my third chapter, I compare the importance of integration and modular evolution in generating novel forms in the body of coral reef fishes. I estimate novelty accommodating six morphological modules across the reef fish body and describe the evolutionary tendency of species to conform to axes of integration among modules. While just under a fifth of all reef fish species achieve extreme morphologies, evolution to the far reaches of integrated axes occurs slightly more often as the evolution of highly unique trait combinations. Through simulations, I show that strong evolutionary integration produces greater morphological novelty but only along axes of trait covariation, while weak integration allows species to enter the outer regions of morphospace with novel trait combinations but produces less extreme morphologies. Reef fishes closely match diversification under moderate integration, in which species evolve into novel regions of morphospace along integrated axes and through the evolution of extreme trait combinations facilitated by independence of modules.