Almost every living and fossil group of gymnosperms has been proposed as a possible ancestor of angiosperms. A common problem with many of these proposals is their reliance on hypothetical intermediate forms. Another common problem is finding correctly-oriented organs homologous to all the important reproductive structures of angiosperms.

These problems are least troublesome for a glossopterid origin of angiosperms. Recently discovered ovule-bearing organs of these plants may represent evolutionary intermediates, or analogous plants, between glossopterids and angiosperms. According to recent reinterpretations of glossopterid ovule-bearing organs, they have structures in an orientation which may be homologous with both the outer ovular integument and the carpel of angiosperms. Considering the reproductive and vegetative features of glossopterids, the hypothesis that they may be part of a stock ancestral to angiosperms should be seriously considered.

Limestone beds in the Middle Ordovician (Trentonian) Bobcaygeon Formation, exposed near Kirkfield, Ontario, exhibit irregular, bored and encrusted surfaces indicative of early lithification. These hardgrounds were formed in extensively burrowed carbonate sediments and their hummocky surface topographies were inherited, in part, from a pre-existing pattern of burrow tunnels. A diverse community, including bryozoans, brachiopods, crinoids, edrioasteroids, and paracrinoids, colonized these hard substrates. In addition, most surfaces are riddled with small, cylindrical boreholes (Trypanites) which represent dwellings of soft-bodied organisms. Some hardground surfaces were inhabited by multiple generations of organisms. Remains of the older generations of encrusters were strongly abraded and nearly obliterated. Superimposed upon these worn remnants are well-preserved remains of rather fragile organisms, e.g. complete hybocystitid crinoids and edrioasteroids. Evidently, certain hardground surfaces were rapidly buried by muds, resulting in the in situ preservation of the last generation of attached organisms.

Slightly differing subcommunities of organisms inhabited various microhabitats provided by the irregular hardground surfaces. Thus, the relative abundance of bryozoans and echinoderms encrusting the roofs of small crevices differs from that on the exposed upper surfaces of the hardgrounds. This is the geologically oldest known hardground community in which microhabitat subdivision can be recognized. However, polarity between the subcommunities is not as pronounced as in geologically younger hard substrate faunas.

We propose that the ancestors of bats were small, nocturnal, sylvatic gliders that used echolocation for general orientation. Their echolocation calls were short, low intensity, broadband clicks, which translated into a very short operational range. In the lineage that gave rise to bats, a switch to stronger, tonal signals permitted the use of echolocation to detect, track, and assess flying insects in subcanopy settings. We propose that these animals hunted from perches and used echolocation to detect, track, and assess flying insects, which they attacked while gliding. In this way, the perfection of echolocation for hunting preceded the appearance of flapping flight, which marked the emergence of bats. Flapping flight had appeared by the Eocene when at least eight families are known from the fossil record. Stronger signals and adaptations to minimize self-deafening were central to the perfection of echolocation for locating flying prey. Echolocation constituted a key innovation that permitted the evolution and radiation of bats. At the same time, however, its short effective range imposed a major constraint on the size of bats. This constraint is associated with flight speed and the very small time intervals from detection of, and contact with a flying target. Gleaning and high duty cycle echolocation are two derived approaches to hunting prey in cluttered situations, places where echoes from background and other objects arrive before or at the same time as echoes from prey. Both had appeared by the Eocene.

All interpretable trace fossils from the Ediacarian–Cambrian transition strata of northern Siberia, Ukraine, and elsewhere represent shelters of infaunal animals feeding from the sediment surface. There is a gradation of forms ranging from (1) makers of horizontal galleries in soft sand with bilobed lower surface and proboscis extended to the surface, through (2) linear or zigzag series of short, widely U-shaped burrows in firm clay with bilobed or three-lobed lower surface, to (3) series of cylindrical chambers dug completely inside the sediment but opening to its surface. At the same time, protective skeleton originated in animals living above the sediment surface. Apparently, the diversification of predators in the earliest Cambrian forced other animals to invest energy either in digging or in a protective armor (“the Verdun Syndrome”). True mud-eaters appeared later, as documented by the late Tommotian horizontal spreite structures from central Siberia. Most, if not all, of those infaunal traces of activity were produced probably by relatives of priapulid worms. It appears that body cavities and segmentation in the Metazoa (diverse already in the Ediacarian) evolved independently of, and prior to, hydraulic burrowing.

Previous genetic studies of extant planktonic foraminifera have provided evidence that the traditional, strictly morphological definition of species in these organisms underestimates their biodiversity. Here, we report the first case where this pattern is reversed. The modern (sub)tropical species plexus Globigerinoides sacculifer is characterized by large morphological variability, which has led to the proliferation of taxonomic names attributed to morphological end-members within the plexus. In order to clarify the taxonomic status of its morphotypes and to investigate the genetic connectivity among its currently partly disjunct (sub)tropical populations, we carried out a global survey of two ribosomal RNA regions (SSU and ITS-1) in all recent morphotypes of the plexus collected throughout (sub)tropical surface waters of the global ocean. Unexpectedly, we find an extremely reduced genetic variation within the plexus and no correlation between genetic and morphological divergence, suggesting taxonomical overinterpretation. The genetic homogeneity within the morphospecies is unexpected, considering its partly disjunct range in the (sub)tropical Atlantic and Indo-Pacific and its old age (early Miocene). A sequence variant in the rapidly evolving ITS-1 region indicates the existence of an exclusively Atlantic haplotype, which suggests an episode of relatively recent (last glacial) isolation, followed by subsequent resumption of unidirectional gene flow from the Indo-Pacific into the Atlantic. This is the first example in planktonic foraminifera where the morphological variability in a morphospecies exceeds its rDNA genetic variability. Such evidence for inconsistent scaling of morphological and genetic diversity in planktonic foraminifera could complicate the interpretation of evolutionary patterns in their fossil record.

Size and shape are analyzed for Pliocene lineages of the rodent genus Stephanomys Schaub 1938. Previous phylogenetic studies were based mainly on size variation and descriptive comparisons, without any attempt to quantify shape changes. Hence, on the basis of regular size increase, Stephanomys has been considered a prime example of phyletic gradualism. In order to quantify morphological variation within the lineage, a method for analyzing complex outlines, the elliptic Fourier transform, was applied to tooth contour (upper and lower first molars). It was then possible to compare evolution in size, estimated by tooth area, as well as evolution of shape, represented by Fourier coefficients.

While size seems to change gradually through time, morphology gives a rather discontinuous evolutionary pattern for both the upper and lower molar. Such a discrepancy between the evolution of size and shape of a single structure suggests that different genetic determinisms and mechanical constraints may act on size and shape. Hence it may be misleading to infer generalized evolutionary processes from either size or shape alone.

Short-term variations in rates of taxonomic extinction and origination in the fossil record may be the result of true changes in rates of turnover, variable rates of fossil preservation, or some combination of the two. Here, positive extinction and origination rate excursions among Phanerozoic marine animal genera are reexpressed in terms of the amount of normal, background time they represent. In addition to providing a background-adjusted calibration of rate intensities, this reexpression determines the durations of sampling gaps that would be required to explain entirely all observed rate excursions as sampling artifacts. This possibility is explored by analyzing a new compilation of the timing and duration of sedimentary hiatuses in North America.

Hiatuses spanning more than approximately one million years (Myr) in the marine sedimentary rock record have a mean duration of 73 Myr. There are two major hiatus types—those that form in response to long-duration tectonic cycles and that bound the major Sloss-scale sequences (mean duration 200 Myr), and those that form in response to shorter-duration changes in sediment accommodation space and that occur within major Sloss-scale sequences (mean duration less than 23 Myr). The latter are approximately exponentially distributed and have a mean duration that is comparable to the mean duration of intervening sedimentary rock packages.

Although sedimentary hiatuses are generally long enough in duration to accommodate the hypothesis that short-term variations in rates of genus origination and extinction are artifacts of sampling failures at major unconformities (“Unconformity Bias” hypothesis), the observed evolutionary rates are not correlated with hiatus durations. Moreover, hiatuses that follow the major mass extinctions are not long in comparison to most other non-mass extinction intervals. These results do not support the hypothesis that hiatuses at major unconformities alone have artificially clustered genus first and last occurrences, thereby causing many of the documented statistical similarities between the temporal structure of the sedimentary rock record and macroevolutionary patterns. Instead, environmental changes related to the expansion and contraction of marine environments may have been the primary forcers of both biological turnover and the spatio-temporal pattern of sediment accumulation. Fully testing this “Common Cause” hypothesis requires quantifying and overcoming lingering taxonomic, biostratigraphic, facies, and numerous other biases that are both inherent in geologic data and imposed by imperfect knowledge of the fossil record.

The feces of the Shasta ground sloth (Nothrotheriops shastense), preserved by the arid climate of the lower Grand Canyon, were collected at various levels and examined by microhistological analyses to identify and quantify plant taxa in the diet. Over 500 pieces of different Shasta sloth coprolites were examined. Sloth dung from the nearby Muav Caves was examined and compared with that from Rampart Cave.

Seventy-two genera of plants were identified in the sloth dung deposited discontinuously from over 40,000 to about 11,000 yr BP. The major plant taxa in the Rampart Cave sloth diets were desert globemallow (Sphaeralcea ambigua = 52%), Nevada mormontea (Ephedra nevadensis = 18%), saltbushes (Atriplex spp. = 7%), catclaw acacia (Acacia greggii = 6%), Cactaceae spp. (= 3%), common reed (Phragmites communis = 5%), and yucca (Yucca spp. = 2%).

Six of the most abundant plants in sloth diets were collected in the environs of Rampart Cave and were analyzed for their energy, fiber and nutrient values. The simulated diets of Rampart Cave sloths averaged 1679 cal/g in digestible gross energy and 7.9% for digestible protein. Apart from a substantial increase in digestible energy and in mormontea there was no unusual change in the sloth diet immediately prior to the time of their extinction.

The ecological role of Nothrotheriops shastense is less dramatically different from that of extant desert herbivores than was previously believed.

Numerical cladistic analysis of 73 cranial and postcranial characters has resulted in a highly corroborated hypothesis describing the phylogenetic pattern of early avian evolution. Using “non-avian theropod” dinosaurs as a comparative outgroup and root for the tree, the analysis confirmed Archaeopteryx to be the sister-group of all remaining avian taxa, or Ornithurae. This latter taxon is subdivided into two lineages, the Hesperornithiformes and the Carinatae. The carinates, in turn, were also resolved into two sister-groups, the Ichthyornithiformes and the modern birds, or Neornithes. This paper provides morphological data corroborating the divergence of the two basal clades of the Neornithes: the Palaeognathae (tinamous and ratites) and Neognathae (all other modern birds). The phylogenetic relationships of four important Cretaceous taxa were also investigated, but these fossil taxa were too fragmentary to determine their phylogenetic position unambiguously. Alexornis and Ambiortus are both carinates, but their relationships cannot be resolved in greater detail. The relationships of the Enantiornithes may lie within the Carinatae or these two taxa may be sister-groups. Gobipteryx is a neornithine and possibly the sister-group of the Palaeognathae.

This analysis indicates that major patterns of morphological change took place at the time of origin of the ancestors of the Ornithurae and the Carinatae. Ornithurine innovations included major changes throughout the skeleton, whereas those of the carinates, while substantial, were primarily restricted to the pectoral girdle and forelimb. The phylogenetic results, in conjunction with the known ages of fossil taxa, indicate that the early lineages of birds very likely arose in the Jurassic. The early cladistic events within the neornithine lineage are also more ancient than generally recognized, and may well extend back to the early Cretaceous.

We compared phylogenies derived from morphological data for two cheilostome bryozoan genera, Stylopoma and Metrarabdotos, with genetic differences between species (Stylopoma) and the stratigraphic occurrence of fossils (both genera). Correspondence between species of Stylopoma defined by protein electrophoresis and on preservable skeletal morphology is excellent, despite great morphological variability within colonies and the predominance of quantitative over discrete characters. Moreover, agreement between genetic and morphological classifications increased greatly when morphological discrimination was pushed to the limit, despite inability to consistently assign all specimens to species with high confidence. This “splitting” strategy also maximized the correlation between genetic distances and the distances between species in cladistically derived phylogenies.

Fossil and living species of both genera are sufficiently abundant and widespread to provide credible limits for inferred ancestral relationships. Inclusion of fossils in cladistic analyses of Stylopoma increased the consistency of cladistic hypotheses by up to 30% and provided a more effective means of rooting trees than comparison with living species of the most closely related genus (“outgroup”). Moreover, in the case of Metrarabdotos, failure to incorporate stratigraphic information turned the cladogram virtually upside down, so that postulated ancestors first appear in the fossil record 6–16 m.y. after their putative descendants became extinct.

Stratigraphically rooted trees suggest that most well-sampled Metrarabdotos and Stylopoma species originated fully differentiated morphologically and persisted unchanged for > 1 to > 16 m.y., typically alongside their putative ancestors. Moreover, the tight correlation between phenetic, cladistic, and genetic distances among living Stylopoma species suggests that changes in all three variables occurred together during speciation. All of these observations support the punctuated equilibrium model of speciation.

Onshore-offshore patterns of faunal change occurred at many taxonomic scales during the Paleozoic Era, ranging from replacement of the Cambrian evolutionary fauna by the Paleozoic fauna to the environmental expansion of many orders and classes. A simple mathematical model is constructed to investigate such change. The environmental gradient across the marine shelf-slope is treated as a linear array of discrete habitats, each of which holds a set number of species, as observed in the fossil record. During any interval of time, some portion of the species in each habitat becomes extinct by background processes, with rates of extinction varying among both clades and habitats, as also observed in the record. After extinction, species are replaced from within the habitat and from immediately adjacent habitats, with proportions dependent on surviving species. This model leads to the prediction that extinction-resistant clades will always diversify at the expense of extinction-prone clades. But if extinction intensity is highest in nearshore habitats, extinction-resistant clades will expand preferentially in the onshore direction, build up diversity there, and then diversify outward toward the offshore. Thus, onshore-offshore patterns of diversification may be the expectation for faunal change quite independently of whether or not clades originate onshore. When the model is parameterized for Paleozoic trilobites and brachiopods, numerical solutions exhibit both a pattern of faunal change and a time span for diversification similar to that seen in the fossil record. They also generate structure similar to that seen in global diversification, including logistic patterns of growth, declining origination but constant extinction within clades through time, and declining overall extinction across clades through time.

Climatic change has long been regarded as an important factor in evolutionary history. In particular, periods of enhanced extinction in marine taxa (especially those from warmer waters) have frequently been linked to decreases in seawater temperature. Studies of the physiology of marine invertebrates and fish alive today have revealed well-developed abilities to cope with temperature change, and there would thus appear to be a dichotomy between the rates of temperature change associated with extinction in geological history and the very much faster rates (by several orders of magnitude) with which many marine organisms can cope today. Nevertheless, evidence from ecology and biogeography indicates that temperature, or some temperature-associated factor, does play a significant role in determining the limits to performance, and hence distribution. The resolution of the dichotomy between the evidence from paleontology and physiology may come through a consideration of the role of the previous evolutionary history of the fauna, the influence of sudden temperature events, or the impact of climatic change on individual competitive ability, community structure, and ecosystem functioning. Studies of the energetics of marine invertebrates in relation to temperature and the evolutionary history of polar faunas indicate that we should beware of anthropocentric judgements in attempting to understand the role of climatic change in evolutionary history, and be critical in distinguishing the role of temperature per se from temperature-associated ecological factors. Present evidence suggests that climatic change in the sea, at least at the rates currently believed to be typical, is unlikely to cause extinction by direct physiological impact. It is more likely that extinction is caused by ecological factors; temperature change is thus only one of several factors that may promote those ecological changes that are currently the best candidates for the proximate cause of extinction in the sea.

Biometric analyses of the ontogenies of 31 species of fossil echinoids support a hypothesized relationship between regulatory gene changes and environmental stability. In 15 of 17 pairs of related species, the larger species, undergoing slower (neotenic) and/or prolonged (hypermorphic) growth, apparently inhabited the stabler environment. If true, this relationship connects biological processes at a number of levels and explains or agrees with some important macroevolutionary observations, such as onshore evolutionary innovation and Cope's Rule.

Radiocarbon-calibrated amino acid racemization ages of 428 individually dated shells representing four molluscan taxa are used to quantify time-averaging and shell half-lives with increasing burial depth in the shallow-water carbonate lagoon of Rib Reef, central Great Barrier Reef, Australia. The top 20 cm of sediment contains a distinct, essentially modern assemblage. Shells recovered at depths from 25 to 125 cm are age-homogeneous and significantly older than the surface layer. Taxon age distributions within sedimentary layers indicate that the top 125 cm of lagoonal sediment is thoroughly mixed on a sub-century scale. The age distributions and shell half-lives of four taxa (Ethalia, Natica, Tellina, and Turbo) are found to be largely distinct. Shell half-lives do not coincide with any single morphological characteristic thought to infer greater durability, but they are strongly related to a combined durability score based on shell density, thickness, and shape. These results illustrate the importance of bioturbation in tropical sedimentary environments, indicate that age estimates in this depositional setting are sensitive to taxon choice, and quantify a taxon-dependent bias in shell longevity and death assemblage formation.

The evolution of septal complexity in fossil ammonoids has been widely regarded as an adaptive response to mechanical stresses imposed on the shell by hydrostatic pressure. Thus, septal (and hence sutural) complexity has been used as a proxy for depth: for a given amount of septal material greater complexity permitted greater habitat depth. We show that the ultimate septum is the weakest part of the chambered shell. Additionally, finite element stress analyses of a variety of septal geometries exposed to pressure stresses show that any departure from a hemispherical shape actually yields higher, not lower, stresses in the septal surface. Further analyses show, however, that an increase in complexity is consistent with selective pressures of predation and buoyancy control. Regardless of the mechanisms that drove the evolution of septal complexity, our results clearly reject the assertion that complexly sutured ammonoids were able to inhabit deeper water than did ammonoids with simpler septa. We suggest that while more complexly sutured ammonoids were limited to shallower habitats, the accompanying more complex septal topograhies enhanced buoyancy regulation (chamber emptying and refilling), through increased surface tension effects.

The periosteal cortex in the shaft of limb bones is described histologically in three ichthyosaurian genera, Omphalosaurus, Stenopterygius, and Ichthyosaurus. The primary periosteal deposits are composed of typical woven-fibered tissue that was accreted as spongy bone in young individuals, and more or less compact bone in older individuals. During growth, the bone tissue was extensively remodeled with a quantitative imbalance between resorption and redeposition. As a result, the cortex was made cancellous, if previously compact, or still more spongy, if already cancellous. This pattern of remodeling explains why compact cortices are generally lacking in the long bones of ichthyosaurs. The presence of woven-fibered tissue strongly suggests that the limb bones, and probably also the body as a whole, had a rapid postnatal growth in ichthyosaurs, that might have been related to a high, “endotherm-like” metabolic rate. This hypothesis bears on the ecological interpretation of the ichthyosaurs: they could have been capable of sustained, fast swimming and long-range movements, rather than being slow-moving creatures as commonly supposed.

A model is developed to explore the potential responses of paleocommunities to disruptions of primary production during times of mass extinction and ecological crisis. Disruptions of primary production are expected to generate bottom-up cascades of secondary extinction, and these are predictable given species richnesses, functional diversity, and trophic link distributions. If, however, consumers are permitted to compensate for the loss of trophic resources by increasing the intensities of their remaining biotic interactions, top-down driven catastrophic increases of secondary extinction emerge from the model. Both bottom-up and top-down effects are themselves controlled by the geometry of the food webs. The general Phanerozoic trends of increasing taxonomic and ecological diversities, as well as the varying strengths of biotic interactions, have led to food webs of increasing complexity. The frequency of catastrophic secondary extinction increases as food web complexity increases, but increased complexity also serves to dampen the magnitude of the secondary extinctions. When intraguild competitive interactions are included in the model, competitively inferior taxa are observed to possess greater probabilities of survival if the guilds are embedded in simple subnetworks of the overall food web. The result is the emergence of post-extinction guilds dominated by those inferior taxa. These results are congruent with empirical observations of “disaster taxa” dominance after some mass extinction events, and provide a mechanism for the reorganization of ecosystems that is observed after those events. The model makes the testable prediction that dominance by disaster taxa, however, should be observed only when bottom-up disruptions have caused ecosystems to collapse catastrophically.

Stable isotope analysis of mammalian tooth enamel is a valuable method for examining resource partitioning in modern and ancient environments where there is a mixture of C3 and C4 plants. However, before 7 Ma North American ecosystems were composed predominantly of C3 plants, complicating isotopic assessment of resource partitioning. Study of modern African and North American ecosystems has shown that niche partitioning among mammals may be discerned in communities dominated by C3 plants, suggesting that a similar approach may work for ancient C3 ecosystems. Here, such analyses are applied to explore resource use and niche partitioning in two ancient C3-dominated communities, one from California and one from Florida. Each locality, Black Hawk Ranch (California) and the Love Bone Bed (Florida), occurs in Miocene deposits that accumulated prior to the rapid increase in C4 ecosystems 7 Myr ago. δ13C and δ18O values were obtained from the tooth enamel of eight species from Black Hawk Ranch, and 15 species from the Love Bone Bed. Results from the 197 bulk isotope samples showed significant differences in δ13C among taxa at the Love Bone Bed, but no significant differences were observed among taxa at Black Hawk Ranch. At both localities, equids generally have more positive δ13C values than co-occurring taxa, suggesting that equids occupied more open habitats, whereas antilocaprids, camelids, and proboscideans have more negative values, implying utilization of more closed communities. One result of note is the positive δ13C values of Pediomeryx (Yumaceras) hamiltoni from the Love Bone Bed, which suggests that P. (Y.) hamiltoni incorporated abundant fiber, possibly grass, in the diet similar to the horses from this locality. The lack of significant differences among taxa at Black Hawk Ranch may indicate a relatively homogeneous flora, or presence of abundant resources permitting niche overlap, whereas the opposite is implied by the presence of significantly different isotope values among taxa at the Love Bone Bed. The results from this study highlight the utility of isotopic techniques allowing discernment of resource partitioning in C3-dominated landscapes such as those that persisted for the millions of years before the rapid increase in C4 ecosystems that occurred during the late Miocene.

Evolutionary “dwarfism” or “nanism” is the phenomenon in which a significant decrease in mean body size of a lineage (an ancestral-descendant sequence of populations) occurs through time. A detailed analysis of several Late Quaternary dwarfed marsupial lineages from Australia is given. Based on linear tooth dimensions of the dwarfed lineages, four points are considered: 1) percent dwarfing; 2) evolutionary rates of change of tooth dimensions; 3) variation within lineages before and after dwarfing; and 4) several aspects of multivariate dental allometry. [The lineages include Macropus titan (fossil)—M. giganteus (Recent) and Osphranter cooperi (fossil)—O. robustus (Recent) in the family Macropodidae (kangaroos), and Sarcophilus laniarius (fossil)—S. harrisii (Recent) in the family Dasyuridae (Tasmaniandevil).]

Dental measurements led to these conclusions: 1) Species with the largest body size show the greatest size reduction, and the species with the smallest body size change the least. 2) Evolutionary rates for this reduction in Australian lineages are similar to comparable Post-Pleistocene dwarfed lineages in Europe. 3) Tooth width, especially posterior width, changes more rapidly than length. The first molar changes relatively slowly, especially in length. 4) Variability is higher in the dwarfed forms than in the larger ancestors. 5) Multivariate allometric rates of dwarfing are consistent with results for rates of change calculated in darwins for the relation between change in length and maximum width (and less so for the relation between M1 and M4 reduction). This pattern of dwarfing allometry is broadly similar to within-species allometry, and is quite dissimilar to synchronous interspecific allometry.

Brief consideration is also given to taxonomy of dwarfing lineages and to problems of concurrent megafaunal extinctions. It is concluded that dwarfism is an adaptive process which is probably the result of a density-dependent factor(s) (i.e. a resource limited system).

The evolutionary history of the Order Carnivora is marked by episodes of iterative evolution. Although this pattern is widely reported in different carnivoran families, the mechanisms driving the evolution of carnivoran skull morphology remain largely unexplored. In this study we use coordinate-point extended eigenshape analysis (CP-EES) to summarize aspects of skull shape in large fissiped carnivores. Results of these comparisons enable the evaluation of the role of different factors constraining the evolution of carnivoran skull design. Empirical morphospaces derived from mandible anatomy show that all hypercarnivores (i.e., those species with a diet that consists almost entirely of vertebrate flesh) share a set of traits involved in a functional compromise between bite force and gape angle, which is reflected in a strong pattern of morphological convergence. Although the paths followed by different taxa to reach this “hypercarnivore shape-space” differ because of phylogenetic constraints, the morphological signature of hypercarnivory in the mandible is remarkably narrow and well constrained. In contrast, CP-EES of cranial morphology does not reveal a similar pattern of shape convergence among hypercarnivores. This suggests a lesser degree of morphological plasticity in the cranium compared to the mandible, which probably results from a compromise between different functional demands in the cranium (e.g., feeding, vision, olfactory sense, and brain processing) whereas the mandible is only involved in food acquisition and processing. Combined analysis of theoretical and empirical morphospaces for these skull data also show the lower anatomical disparity of felids and hyaenids compared to canids and ursids. This indicates that increasing specialization within the hypercarnivorous niche may constrain subsequent morphological and ecological flexibility. During the Cenozoic, similar skull traits appeared in different carnivoran lineages, generated by similar selection pressures (e.g., toward hypercarnivory) and shared developmental pathways. These pathways were likely the proximate source of constraints on the degree of variation associated with carnivoran skull evolution and on its direction.