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.

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.

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.

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.

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.

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.

Many modern paleobiological analyses are conducted at the generic level, a practice predicated on the validity of genera as meaningful proxies for species. Uncritical application of genera in such analyses, however, has led—perhaps inadvertently—to the unjustified reification of genera in an evolutionary context. While the utility of genera as proxies for species in evolutionary studies should be evaluated as an empirical issue, in practice it is increasingly assumed (rather than demonstrated) that genera are suitable proxies for species. This is problematic on both ontological and epistemological grounds. Genera are arbitrarily circumscribed, non-equivalent, often paraphyletic, and sometimes polyphyletic collections of species. They are useful tools for communication but have no theoretical or biological reality of their own and, whether monophyletic or not, cannot themselves operate in the evolutionary process. Attributes considered important for understanding macroevolution—e.g., geographic ranges, niche breadths, and taxon durations—are frequently variable among species within genera and will be inflated at the generic level, especially in species-rich genera. Consequently, the meaning(s) of results attained at the generic level may not “trickle down” in any obvious way that elucidates our understanding of evolution at the species level. Ideally, then, evolutionary studies that are actually about species should be pursued using species-level data rather than proxy data tabulated using genera. Where genera are used, greater critical attention should be focused on the degree to which attributes tabulated at the generic level reflect biological properties and processes at the species level.

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.

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.

The Paleobiology Database (PBDB; https://paleobiodb.org) consists of geographically and temporally explicit, taxonomically identified fossil occurrence data. The taxonomy utilized by the PBDB is not static, but is instead dynamically generated using an algorithm applied to separately managed taxonomic authority and opinion data. The PBDB owes its existence to many individuals, some of whom have entered more than 1.26 million fossil occurrences and over 570,000 taxonomic opinions, and some of whom have developed and maintained supporting infrastructure and analysis tools. Here, we provide an overview of the data model currently used by the PBDB and then briefly describe how this model is exposed via an Application Programming Interface (API). Our objective is to outline how PBDB data can now be accessed within individual scientific workflows, used to develop independently managed educational and scientific applications, and accessed to forge dynamic, near real-time connections to other data resources.

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).

Echinoderms have long been characterized by the presence of ambulacra that exhibit pentaradiate symmetry and define five primary body axes. In reality, truly pentaradial ambulacral symmetry is a condition derived only once in the evolutionary history of echinoderms and is restricted to eleutherozoans, the clade that contains most living echinoderm species. In contrast, early echinoderms have a bilaterally symmetrical 2-1-2 arrangement, with three ambulacra radiating from the mouth. Branching of the two side ambulacra during ontogeny produces the five adult rays. During the Cambrian Explosion and Ordovician Radiation, some 30 clades of echinoderms evolved, many of which have aberrant ambulacral systems with one to four rays. Unfortunately, no underlying model has emerged that explains ambulacral homologies among disparate forms. Here we show that most Paleozoic echinoderms are characterized by uniquely identifiable ambulacra that develop in three distinct postlarval stages. Nearly all “aberrant” echinoderm morphologies can be explained by the paedomorphic ambulacra reduction (PAR) model through the loss of some combination of these growth stages during ontogeny. Superficially similar patterns of ambulacral reduction in distantly related clades have resulted from the parallel loss of homologous ambulacra during ontogeny. Pseudo-fivefold symmetry seen in Blastoidea and the true fivefold symmetry seen in Eleutherozoa result from great reduction and total loss, respectively, of the 2–1–2 symmetry early in ontogeny. These ambulacral variations suggest that both developmental and ecological constraints affect the evolution of novel echinoderm body plans.

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.

Despite extensive paleoecological analyses of spatial and temporal turnover in species composition, the fidelity with which time-averaged death assemblages capture variation in species composition and diversity partitioning of living communities remains unexplored. Do death assemblages vary in composition between sites to a lesser degree than do living assemblages, as would be predicted from time-averaging? And is the higher number of species observed in death relative to living assemblages reduced with increasing spatial scale? We quantify the preservation of spatial and temporal variation in species composition using 11 regional data sets based on samples of living molluscan communities and their co-occurring time-averaged death assemblages. (1) Compositional dissimilarities among living assemblages (LA) within data sets are significantly positively rank-correlated to dissimilarities among counterpart pairs of death assemblages (DA), demonstrating that pairwise dissimilarity within a study area has a good preservation potential in the fossil record. Dissimilarity indices that downplay the abundance of dominant species return the highest live-dead agreement of variation in species composition. (2) The average variation in species composition (average dissimilarity) is consistently smaller in DAs than in LAs (9 of 11 data sets). This damping of variation might arise from DAs generally having a larger sample size, but the reduction by ∼10–20% mostly persists even in size-standardized analyses (4 to 7 of 11 data sets, depending on metric). Beta diversity expressed by the number of compositionally distinct communities is also significantly reduced in death assemblages in size-standardized analyses (by ∼25%). This damping of variation and reduction in beta diversity is in accord with the loss of temporal resolution expected from time-averaging, without invoking taphonomic bias (from differential preservation or postmortem transportation) or sample-size effects. The loss of temporal resolution should directly reduce temporal variation, and assuming time-for-space substitution owing to random walk within one habitat and/or temporal habitat shifting, it also decreases spatial variation in species composition. (3) DAs are more diverse than LAs at the alpha scale, but the difference is reduced at gamma scales because partitioning of alpha and beta components differs significantly between LAs and DAs. This indicates that the effects of time-averaging are reduced with increasing spatial scale. Thus, overall, time-averaged molluscan DAs do capture variation among samples of the living assemblage, but they tend to damp the magnitude of variation, making them a conservative means of inferring change over time or variation among regions in species composition and diversity. Rates of temporal and spatial species turnover documented in the fossil record are thus expected to be depressed relative to the turnover rates that are predicted by models of community dynamics, which assume higher temporal resolution. Finally, the capture by DAs of underlying variation in the LA implies little variation in the net preservation potential of death assemblages across environments, despite the different taphonomic pathways suggested by taphofacies studies.

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 ability of paleobiologists to draw paleoecological inferences based on spatial faunal variability within a single stratigraphic interval depends ultimately on the spatial resolving power of the fossil record. This paper evaluates the potential spatial resolution of fossil assemblages by examining modern skeletal remains of molluscs on a benthic transect, along which there is a marked decrease in seagrass cover, in Smuggler's Cove, St. Croix, U.S. Virgin Islands. The sampling transect began in a Thalassia-covered area approximately three meters deep, extended into slightly deeper water with lighter seagrass cover, and ended on an open, bioturbated sandy tract at a depth of nearly six meters.

Two-way cluster analysis and polar ordination of 37 samples of molluscan remains, taken at 10-meter intervals along the 360-meter transect, reveal patterns of variation that are shown by correlation analyses and consideration of the autecologies of individual species to be related to measured changes in vegetation. There is a transition from dominance primarily by epifaunal gastropods living on seagrass blades to dominance by infaunal, burrowing bivalves as grass cover becomes lighter. Some non-systematic variability exists in faunal distributional patterns within areas where the environment does not vary systematically, but this does not mask the regular faunal transitions related to environmental changes. Correspondence between the dead and live faunas is difficult to ascertain because of the scarcity of live fauna in collected samples.

The results suggest that spatial faunal transitions in fossil remains at even the fine scale evaluated in this study are potentially preservable in the fossil record.

The 18 m.y. history of the subfamily Equinae (exclusive of Archaeohippus and “Parahippus”) in North America consisted of a 3-m.y. radiation phase, a 9-m.y. steady-state diversity phase, and a 6-m.y. reduction phase. During the steady-state phase, species richness varied between 14 and 20, with two maxima at about 13.5 and 6.5 Ma. Species richness of the tribes Hipparionini and Equini was about equal through the middle Miocene, but hipparionines consistently had more species in the late Miocene and early Pliocene. Overall mean species duration was 3.2 m.y. (n = 50), or an average extinction rate of 0.31 m.y.-1 During the radiation phase, speciation rates were very high (0.5 to 1.4 m.y.-1), while extinction rates were low (<0.10 m.y.-1). Speciation and extinction rates both averaged about 0.15 m.y.-1 during the steady-state phase, with extinction rates having more variation. Extinction rates increased fourfold during the reduction phase, while speciation rates declined slightly. Late Hemphillian extinctions affected both tribes severely, not just the three-toed hipparionines, and were correlated with global climatic change.