Major evolutionary innovations in lifestyle can lead to adaptive radiations, change the nature of selective forces, and rearrange ecological interactions. One of the many determinants of lifestyle is body mass, and so change in body mass is often accompanied by change in lifestyle. In the case of mammals, their earliest members from the Late Triassic were small. They remained relatively small throughout the Mesozoic, most likely due to interactions with the diverse large bodied diapsids that dominated Mesozoic large-body ecomorphs, especially dinosaurs in the terrestrial realm. It was not until the early Cenozoic, after the extinction of the non-avian dinosaurs and large bodied marine diapsids, that the body size of mammals expanded, which lead to new mammalian ecomorphs. The initial stages of the Cenozoic mammalian exploration of large-body ecomorphs included at least two novel semiaquatic forms – large amphibious herbivores and large coastal marine carnivores. These ecomorphs were to evolve multiple times throughout the Cenozoic and led to the freshwater Hippopotamidae, the marine Pinnipedia (seals and sea lions), as well as to the fully aquatic marine mammals, the sirenians (manatees and dugongs) and cetaceans (whales and dolphins).

This dissertation employes bone histological methods and the analysis of new fossil specimens to add to our paleontological knowledge of large bodied semiaquatic mammals. The earliest occurrences of these large-bodied semiaquatic forms are poorly understood, and gaps exist in our knowledge of their evolution and distribution. To help address these deficiencies I used bone histological methods to establish the lifestyle of the earliest large-bodied mammal, the Paleocene-Eocene Coryphodon. I follow this with the description and analysis of new fossil material that expand our knowledge of the diversity and biogeography of fossil pinnipeds and their earliest stem-relatives, the geologically longest-lived lineage of semiaquatic marine carnivores.

The Late Paleocene to Early Eocene Coryphodon was a member of the Pantodonta, the first mammalian group to achieve large body size following the end-Cretaceous mass extinction. Using histological and microanatomical techniques to examine the structure and growth of its bones, I show that Coryphodon was semiaquatic and further demonstrate that it had significant adaptations to this way of life rather than merely occupying waterways in a transient manner. A framework for behavioral inference for extinct taxa is presented, which incorporates evidence from multiple timescales. I argue that this approach facilitates the framing of the causal relationships between, and interpretation of, the different kinds of evidence that can be brought to bear on our understanding of the origins of major changes in lifestyle. In the case of Coryphodon, the concurrent achievement of large body size and semiaquatic lifestyle suggests a reciprocal rather than unidirectional causal relationship between body size and specialization at the beginning of the Cenozoic.

Turning to the origins and biogeography of the pinnipeds, the Pleistocene sea lion Proterozetes ulysses belongs to a monophyletic group of North Pacific otariids that includes the living genera Eumetopias (Stellar’s sea lion) and Zalophus (California sea lion). Fossil mandibles of Proterozetes from the Port Orford Formation of Oregon are described and found to be intermediate in morphology and size to the other North Pacific otariids, most resembling Eumetopias. The presence of this distinct but morphologically intermediate taxon in the eastern North Pacific during the Pleistocene establishes the Pleistocene as a time of increased otariid diversity, in contrast to diversity decline seen in other marine mammal clades. The timing of sea lion diversification along the west coast of North America appears to follow diversification in Japan, reinforcing the possibility of an eastward dispersal trend in the North Pacific after the end-Pliocene. Pleistocene sea lions in the Eastern North Pacific similarly show multitaxon assemblages with overlapping body size and feeding styles, indicating the high diversity possible once the coastal carnivore ecomorph was achieved.

Finally, paleontological investigation of the early history of pinnipeds similarly reveals assemblages of increased diversity and overlapping body size. The pinnipedimorph Enaliarctos was a marine-adapted carnivore with dental and locomotor features intermediate between terrestrial arctoid carnivores and living pinnipeds. Newly discovered fossils, described herein, fill a major biogeographic gap in the record of the genus, previously known nowhere between southern California and central Oregon, and show that stem pinnipeds existed for longer and survived later than previously thought, from the latest Rupelian to the late Burdigalian (28.1-16.6 mya). This discovery of an increased biogeographic and temporal range of Enaliarctos indicates that the genus may have interacted with a much wider range of other taxa than previously thought, including plotopterid birds, odontocete whales, and crown pinnipeds such as early odobenids, early otariids, and desmatophocids. The expansion of the known ranges of Enaliarctos species and the description of previously undescribed morphological attributes, particularly of the mandible and lower dentition, provides insight into the origins of pinniped diversity and their possible interactions with other early Neogene coastal marine organisms. Though early pinniped relatives were smaller than living members, increasing body size does not appear to have been necessarily superior once the coastal marine ecospace was colonized.

The genitalia of males evolves rapidly and divergently in most animal groups. Explanations for this pattern typically evoke mechanisms of post-copulatory sexual selection, such as sperm competition, cryptic female choice, and sexual conflict. While the fossil record often allows for the documentation of patterns of evolutionary change, the predominantly soft-tissue nature of the genitalia is rarely observed in the fossil record. However, the baculum (os penis or os priapi), a bone located at the distal end of the penis of most mammalian species, offers the opportunity of examining the evolution of male genitalia, and thus may contribute to our understanding of post-copulatory sexual selection and the evolution of reproductive biology within mammals. Bacula are rare in the fossil record and thus difficult to be identified and classified when in isolation. Nonetheless, sometimes they are preserved in association with other skeletal material allowing accurate species-level taxonomic assignments. Here, I focus on the bacula of the Canidae, which consists of three subfamilies, the extant Caninae (wolves, coyotes, jackals, foxes, etc.) and two extinct subfamilies, the Borophaginae and the Hesperocyoninae, both of which have rich fossil records that reveal high taxonomic and ecomorphological diversity.

This dissertation consists of four chapters. In the first chapter, I briefly contextualize the theory of post-copulatory sexual selection and then I review the evidence that it acts on the baculum from a combination of studies that used functional, experimental, and comparative data.

In Chapter 2 I describe the morphology of fossil bacula of five species of Borophaginae. These descriptions add substantial knowledge regarding the baculum of extinct canids, which was previously known for only two extinct species. The five bacula were found associated with other skeletal elements that were already described in the literature and have enabled identification to the species level. These borophagine bacula share many similarities with extant canines such as being robust, having a urethral groove, and a simple distal end. Because these features have been associated with reproductive strategies in extant carnivorans it is possible to use the baculum of Borophaginae to draw implications for their reproductive biology.

In Chapter 3 I quantify variation in morphology of the baculum of canids in 2D using elliptic Fourier analysis. First, I show that the baculum of 26 (out of ~35) extant species is morphologically conserved and that the intraspecific morphological variation is larger than the interspecific differences, and therefore it is difficult to distinguish among species using bacular morphology alone. However, four extinct canids that have completely preserved bacula show morphological differences from the extant canids. In particular, one morphological feature, the primary curvature of the baculum, differentiates the baculum of three stem group canids from those of extant species. I used phylogenetic ancestral state reconstruction to show that it would be unlikely to predict the arched curvature of the baculum observed in the stem fossils from the morphology observed in extant species.

Finally, in Chapter 4, I examine the relative size of the canid baculum. Using a linear regression corrected for phylogeny, I show a slightly negative allometric relationship between baculum length and body size in canids.Further, an isometric relationship cannot be rejected. These results are in contrast with the strong negative allometric relationships that have been described for other mammalian clades. Of particular interest are the relative size of fossil bacula. The baculum of Hesperocyon gregarius, the only known in the Hesperocyoninae, is significantly longer than expected for its body size and suggests a different allometric curve for this species in comparison with other canid groups. The bacula of two borophagine species are within the same allometric line of extant canids. Finally, the baculum of the extinct dire wolf (Aenocyon dirus) is longer than expected based on its body size, suggesting that its reproductive strategies could be different from those of extant canines.

The diversity of the biota of any place and time is often the product of a long and complex history of ecological and evolutionary assembly through the fundamental processes of immigration, speciation and extinction. This biotic assembly typically occurs amid a backdrop of constant abiotic and biotic change at multiple spatiotemporal scales. Given this potentially complex abiotic and biotic history, investigating the core processes that underlie the assembly of regional biotas, as well as the abiotic, biotic and historical factors that influence them, is a major challenge at the intersection of biogeography, macroevolution and ecology.

Oceanic islands provide ideal opportunities to investigate the processes underlying the assembly of regional biotas. Given their volcanic origins, their landscapes represent initial blank slates that are then populated by successful long-distance dispersal events by various organisms, and the subsequent diversification of some of these colonist lineages. Furthermore, their relatively predictable geologic histories and lifespans (i.e., ontogenies) make the study of how the diversification of lineages are shaped over geologic time-scales especially tractable, compared to the more complex abiotic and biotic histories that often shape continental, mainland biotas. Remote hotspot archipelagoes such as that of Hawaii are especially exciting case studies, as they are composed of islands formed in sequence through the movement of the Pacific tectonic plate over a relatively stationary mantle hotspot, and so that each island captures a record of diversification at different stages. The multiple “snapshots” of diversity provided by the differently-aged islands of the Hawaiian archipelago, as well as their well-defined and quantifiable ontogenies, allow the diversities observed on each of the main islands to be leveraged to understand how diversity on islands change through the lifespan of an island, as if each island were a different point on a time series. In Chapter 1, I investigate whether, and how, the changing geologic backdrop (in this case, island area as each island grows and later subsides) has shaped the diversification (speciation, immigration and extinction) of Hawaiian plant and animal lineages on each of the islands. I used a geologically informed diversity-dependent model of species richness change that incorporates estimates of how island area has changed through time, while also taking into account island-specific ontogenetic differences among the four current main islands / island complexes of the Hawaiian archipelago, and tested my models against alternative models that do not account for island ontogenetic changes. I find that, at the broadest temporal scales, the geologic dynamics of the Hawaiian Islands have had a profound influence on the macroevolutionary history of most of the 14 endemic lineages examined, with all lineages undergoing radiations as islands grow, and most of them now undergoing long-term evolutionary declines as the islands undergo subsidence and erosion with age.

As mentioned above, the isolation and de novo formation of the Hawaiian Islands imposes a strong filter, limiting successful colonization of the islands to a subset of mainland plant and animal groups that have the traits necessary for long-distance dispersal. As a result, it has often been assumed that the Hawaiian Islands often represent a biogeographic sink for many lineages, with colonization of the remote islands representing “dead ends” in the history of dispersal of lineages that prove successful. In some cases, this is certainly true, with some lineages evolving characteristics that reduce dispersal ability relatively quickly upon successful colonization. However, molecular phylogenetic data are slowly challenging this narrative, and it is increasingly demonstrated that the Hawaiian archipelago may instead also play a fundamental role as a biogeographic source for other parts of the Pacific and even mainland areas. In Chapter 2, testing this more nuanced view of the biogeographic history of Hawaiian lineages in the context of the greater Pacific, I constructed a molecular phylogeny of one of the most species-rich angiosperm radiations in the Pacific, Peperomia (Piperaceae), using full plastome sequences generated via a high-throughput shotgun sequencing approach. The final tree consisted of about half the total number of species described in the Pacific. Incredibly, I found that the Pacific has played host to four separate and distinct colonist lineages that originated from the Neotropics, although whether these colonization events occurred through direct dispersal from South and Central America, or via indirect routes of dispersal through the Afro and Paleotropics remain unclear. Nonetheless, in support of the newer view of the role of the Hawaiian Islands in Pacific biogeography, my molecular phylogeny paints a picture of frequent dispersal within the south Pacific, with two independent radiations of Peperomia on Hawaii, and subsequent southwards dispersal of some individuals derived from one of these radiations to the Marquesas Islands. My results emphasize the importance of expanding the biogeographic lens through which many island lineages are typically viewed, and future global-level sampling will shed more light on the role of the islands of the Pacific in generating the diversity of many globally important plant lineages.

Lastly, while my first chapter focuses on the role of landscape dynamism on the tempo of diversification on Hawaii, and my second chapter focuses on the role of the Hawaiian Islands in the biogeographic history of lineages in the greater Pacific, the third chapter of this dissertation instead takes a closer look at whether the climatic niches of lineages may influence the modes of speciation among eight large radiations of endemic plant groups. The central hypothesis is that the climatic niche of lineages may play a role in shaping the inter-island dispersal of species, and hence the geographic isolation of populations on different islands. For instance, a wet forest species on one island may be expected to more easily colonize wet forests on other islands (i.e., climatic niche conservatism). If climatic niche conservatism were a dominant force in shaping the mode of diversification, one may expect cladogenetic events in allopatry to be associated with lower rates of climatic niche evolution relative to cladogenetic events in sympatry (here defined as sister lineages co-occurring on at least one island). To test this idea, I reconstructed the phylogenetic relationships within eight species-rich plant radiations on Hawaii, and inferred the geographic context of all cladogenetic events in each plant group using probabilistic biogeographic models. I then looked at whether cladogenetic events in sympatry (i.e., sister lineages that overlap in biogeographic range) versus cladogenetic events in allopatry (diverged sister lineages that occur on different islands) were associated with higher or lower rates of climatic niche evolution. I found, in a combined analysis of all groups examined, that rates of climatic niche evolution do not appear to be explained by differences in speciation mode. However, two of the plant groups examined (Schiedea and Kadua) show significant but opposing relationships between speciation mode, with lineages of Kadua showing lower rates of niche evolution in allopatry and lineages of Schiedea diverging in allopatry associated with higher rates of niche evolution. Overall, my results suggest that lineage-specific differences likely have a strong influence on rates of climatic niche evolution and their interaction with dispersal and speciation.

Overall, oceanic islands are, and will likely remain, powerful microcosms by which strategies for understanding how mainland biotas are developed. The three chapters of this dissertation highlight the need for a multi-faceted approach in studying the ecological and evolutionary assembly of regional biotas: from its response to geologic changes over long-time scales (Chapter 1), the larger biogeographic context from which biotas are assembled (Chapter 2), to the often context-specific role of climatic niche evolution in shaping the dispersal and speciation of lineages (Chapter 3).

Cassiduloids are rare and poorly known irregular echinoids, which include the sand dollars and heart urchins, that typically live buried in the sediment, where they feed on small organic particles. Cassiduloids evolved during the Marine Mesozoic Revolution, but despite their rich fossil record, species richness (diversity) is very low. The goal of this thesis is to improve our taxonomic knowledge of the group, propose hypotheses of relationship among its representatives and analyze their patterns of geographic distribution through time, thereby contributing to our understanding of their evolutionary history.

In the first chapter , I used synchrotron radiation-based micro-computed tomography (SRµCT) images of type specimens to describe a new Cassidulus species and a new cassiduloid genus that could not have been discovered with traditional techniques. I also designate a neotype for the type species of the genus Cassidulus, Cassidulus caribaearum, provide remarks on the taxonomic history of each taxon, a diagnostic table of all living cassidulid species, and extend the known geographic and bathymetric range of two species. Besides rendering novel morphological data, the SRµCT images provided significant insights in the evolution of bourrelets of these cassiduloid echinoids. However, determining how the bourrelets have evolved, as well as the evolution of all the cassiduloid traits, requires a phylogenetic framework of the group.

Therefore, in the second chapter, I reconstructed the phylogeny of the cassidulids using morphological characters and inferred their patterns of geographical distribution through time. Because morphological and geographic histories are erased by extinctions, unraveling phylogenetic relationships and biogeographic patterns based on only the living species can be challenging, especially for groups that have experienced extensive extinction such as the cassiduloids. Studies have shown that fossil taxa generally improve phylogenetic resolution because of their unique morphological information that have often been modified in Recent species. Thus, inclusion of fossils can be critical to addressing evolutionary questions. Surprisingly for marine invertebrates, there are relatively few studies that have included fossils in their phylogenetic and biogeographic analyses. I performed a cladistic analysis of 45 cassidulids based on 98 characters, which resulted in 24 most-parsimonious trees. The monophyly of the family Cassidulidae was not supported because the genera Eurhodia and Glossaster were placed within the family Faujasiidae. Analyses to determine the sensitivity of the resulting clades to missing data did not result in significantly different topologies or resolution of the tree, but the coding of partial uncertainties changed the relationship within some subclades (particularly within genera). The taxonomic implications of these results and the evolution of some morphological features are discussed. The evolutionary history of the cassiduloids has been dominated by high levels of homoplasy and a dearth of unique novel traits. The cassiduloids (as defined in this study) most likely originated in the Early Cretaceous (oldest records from the Aptian), and no conspicuous novelties were added during their evolution. Biogeographically, a time-stratified DEC model with range constraints indicates that the cassidulids had a south Tethyan and northwest Atlantic origin probably dating back to the Late Cretaceous (Campanian–Maastrichtian) or Paleocene. Most cassiduloids are endemic to small regions and their evolution has mostly been influenced by dispersal rather than vicariant events. Speciation occurred mainly within the northwest Atlantic during the Late Paleocene to the early Eocene. Despite their high diversity during the Paleogene, cassidulids and faujasiids have only seven extant species, and three of them are relicts of lineages that date back to the Eocene. Future studies of the biology of these poorly known species, some of which brood their young, will yield further insights into the evolutionary history of this group.

While the first two chapters focused on cassiduloids, the third chapter is a broader macroevolutionary study of the echinoids. Specifically, I performed an analysis of how their genus richness (diversity) has changed since their appearance in the fossil record and estimated the turnover rates (origination and extinction) throughout their evolutionary history. The ability to document macroevolutionary trends has been accelerated by the development of free-access online databases. However, despite their undeniable benefit to research, these databases are not free from error and their data need to be checked on a regular basis. Therefore, with the goal of analyzing the echinoid’s diversity dynamics since their appearance in the fossil record, I assess the quality of the echinoid entries in the Paleobiology Database (PBDB), correct errors and improve the dataset by including missing information. Assessments of data quality included cross-referencing classifications and checking stratigraphic ranges and synonymies against the literature. The entries in the PBDB were derived from 1,100 references and include ~9,500 occurrences representing 445 genera, about 65% of all valid echinoid genera. Fifty percent of the occurrences were from Europe and the USA, and 41% percent were from the Cretaceous. Genus classifications were mostly outdated and some species were misclassified. Diversity curves were then generated using three different methods to account for preservation and sampling biases. Overall, the echinoid data from the PDBD reflected major trends known for the evolution of the class (e.g. increased diversification during the Marine Mesozoic Revolution) and recovered extinction events that have affected all marine biota (e.g. the end-Cretaceous mass extinction). However, spatial, temporal and taxonomic biases exist, so we need to be mindful of these when analyzing the data. The fossil record of the echinoids dates back to the Middle Ordovician. Although affected by mass extinctions, the group’s diversity has been steadily increasing since the origin of the irregular echinoids in the Mesozoic. In addition to the echinoid diversity curve and turnover rates, contrasting and similar diversity trajectories for closely related major echinoid clades are also presented.