Open Access
How to translate text using browser tools
1 June 2004 Chapter 11: Global Climate and the Evolution of Large Mammalian Carnivores during the Later Cenozoic in North America
ROBERT M. HUNT Jr.
Author Affiliations +
Abstract

Taxon ranges of larger mammalian carnivores can be grouped into seven temporal intervals during the later Cenozoic. These intervals are of varied duration and seem to correspond to periodic faunal reorganizations that accompanied the progressive climatic deterioration occurring from the late Eocene to the Pleistocene. Recent oxygen isotope records from deep-sea cores serve as proxy for the pattern of global climate during the Cenozoic and compare reasonably well with the large carnivore intervals. Intervals A, B, and the early part of C characterize a time of cooler global climate (δ18O: 1.3 to 3.0‰) following the early Eocene climatic optimum. The later part of Interval C, following the mid-Miocene climatic optimum, and Intervals D through F record a gradual climatic deterioration (δ18O: 2.0 to 3.8‰) from the mid-Miocene to early Pliocene. Interval G (δ18O: 3.8 to 5.0‰) corresponds to the extreme global cooling of the later Pliocene and Pleistocene. Glacioeustatic decline in sea level during these intervals probably made possible the entrance of migrant Eurasian carnivores and other mammals into the New World via the Bering route. The periodic emergence of this land bridge and the effect of the climatic oscillations of the later Cenozoic on the mammalian fauna appear responsible for the faunal shifts.

INTRODUCTION

An improved record of Neogene and Quaternary fossil mammals has afforded a new perspective on faunal succession in North America during the last 40 million years. Well-constrained taxonomic ranges have resulted from (1) new discoveries of Neogene mammals from more refined lithostratigraphic contexts, and (2) continuing study of historic collections, most notably that of Childs Frick at the American Museum of Natural History. Taking advantage of these conditions, this study focuses on the evolutionary history of North American carnivorans. In this case, range diagrams for large carnivores demonstrate distinct, episodic transformations at guild level: such reorganizations seem to have occurred at much greater frequency during the late Cenozoic, seemingly in step with climatic deterioration as charted by oxygen isotope curves.

Large carnivores, at the apex of the ecological pyramid, influence the dynamic of prey populations (e.g., Schaller, 1972; Sinclair and Norton-Griffiths, 1979). The nature of social organization of a species, its body size, territorial extent, style of predation, and mode of locomotion, among other such factors, determine its interaction with various prey species. Thus, the identification of intervals characterized by stable, well-defined carnivore assemblages contributes to an improved understanding of mammalian faunal dynamics overall. Here I provide a summary of taxon ranges for North American large carnivores, calibrated against the Magnetic Polarity Time Scale (MPTS; Berggren et al., 1995). These taxa are arranged in faunal intervals from ∼40 Ma to the Holocene (fig. 11.1, table 11.1). Seven intervals are recognized, each with its characterizing carnivore taxa, and defined by first appearances (FADs) and last occurrences (LADs). The implications of these intervals and their temporal spacing are discussed below. The usage followed here employs the term “carnivoran” for members of the Order Carnivora and “creodont” for members of the Hyaenodontidae and Oxyaenidae that currently comprise the Order Creodonta. The term “carnivore” refers to any flesh-consuming mammal.

FAUNAL ASSOCIATIONS OF LARGER NORTH AMERICAN CARNIVORES

Intervals A to G (late Eocene to Pleistocene)

Large (>10–20 kg) carnivorans take the stage with the advent of the White River carnivore fauna in the Chadronian North American Land Mammal Age (NALMA) (Hunt, 1996: fig. 2). Duchesnean and earlier associations of carnivores included only archaic early Tertiary groups placed in the miacoid Carnivora and the hyaenodont and oxyaenid Creodonta. Viverravid and miacid species comprise the miacoids (Matthew, 1909; Flynn, 1998), none larger than a modern coyote (Canis latrans). The largest of the miacoids was the viverravid Didymictis vanclevei from the early Eocene of the Huerfano basin, Colorado, with a basilar length of skull of ∼15–16 cm; the very small viverravid Viverravus minutus from the Bridger Basin, Wyoming, had a basilar length of ∼4 cm, and most miacids measured <12 cm.

With the arrival of the Chadronian carnivoran-creodont fauna at ∼37 Ma, there is an evident increase in the average body size of carnivores. The largest carnivorans in North America were nimravid cats and coyote-sized daphoenine beardogs (Daphoenus vetus), accompanied by particularly large hyaenodont creodonts (Hyaenodon megaloides, H. horridus, Hemipsalodon grandis; Mellett, 1969, 1977). Hyaenodon attained its maximum species diversity in the Chadronian (Mellett, 1977). This hyaenodont-nimravid association was previously recognized as the first of four sequential mid- to late Cenozoic associations of large carnivorous mammals that characterized the last 40 million years in North America (Hunt and Tedford, 1993). Here I amplify our initial designations, extending and refining the definition of these associations, and describing their characterizing taxa in greater detail.

Hyaenodont-Nimravid Association (Interval A: ∼37–23.7 Ma)

The hyaenodont-nimravine nimravid-daphoenine amphicyonid association is typical of Chadronian, Orellan, Whitneyan, and early Arikareean faunas of North America. Of the intervals recognized here, its 13.3 million year extent is by far the longest. Hyaenodont and nimravine nimravid diversity during this interval is remarkable, with nine species of hyaenodonts and ten of nimravids currently recognized (Mellett, 1977; Bryant, 1991, 1996). If size alone is considered, the dominant carnivores of this association are the large creodonts, Hemipsalodon grandis (basilar length, ∼41 cm), Hyaenodon megaloides (skull length, ∼40 cm), and H. horridus (skull lengths, 25–35 cm). The largest nimravids occur toward the end of the interval in the Whitneyan and early Arikareean. Skull lengths of the larger nimravine genera range from ∼13 to ∼26 cm (Bryant, 1996; personal observations): Dinictis, ∼13 to 18 cm; Pogonodon, ∼20 cm; Nimravus, ∼15 to 23 cm; Hoplophoneus, ∼15 to 26 cm.

Daphoenine amphicyonids commonly accompany these groups and are represented by four genera (Daphoenus, Daphoenictis, Brachyrhynchocyon, Paradaphoenus) and at least six species (Hunt, 1996). The largest is Daphoenus vetus, with males attaining basilar skull lengths of ∼20 cm in the Orellan when the species is particularly well-represented in the White River Group of the Great Plains. Daphoenus achieves a maximum basilar length of ∼23 cm by the time of its last occurrence in the early Arikareean.

Hyaenodonts, nimravine nimravids, and Daphoenus became extinct within the early Arikareean interval. The last records of hyaenodonts and Daphoenus in late Oligocene rocks of the lower Arikaree Group in the Great Plains are dated at ∼28.0 Ma and ∼28.6 Ma, respectively; Daphoenus survived to ∼27 Ma in the Pacific Northwest in the John Day beds of Oregon. Nimravine cats persist only a short while longer to ∼25– 24 Ma. Hence these long-dominant groups appear to be extinct by ∼27–24 Ma. Nimravines are represented by at least three genera (Nimravus, Pogonodon, Eusmilus) in lower Arikaree Group sediments, contemporaneous with the last occurrences of hyaenodonts and Daphoenus. Although creodonts became extinct in North America at ∼28 Ma, they survived in southern Asia and Africa into the Miocene.

In the richly fossiliferous sediments of the central Great Plains in Nebraska, South Dakota, Wyoming, and Colorado, UNSM paleontologists have found no fossils of these groups above the lower Arikaree Group. Extensive explorations for fossil mammals in rocks of the upper Arikaree Group in Nebraska, Wyoming, and South Dakota have failed to produce a hyaenodont, nimravid, or Daphoenus. Thus the hyaenodont-nimravid association is apparently extinct by late Arikareean time (estimated to begin at ∼23 Ma) in the North American midcontinent.

Dominance of Amphicyonines-Hemicyonines (extending over Intervals B and C)

Immigrant Eurasian amphicyonine beardogs and hemicyonine ursids characterize the large carnivoran assemblage in North America from the beginning of the Miocene to ∼9 Ma at the end of the Clarendonian. During the late Hemingfordian, Barstovian, and Clarendonian, the dominant large carnivorans are large amphicyonines and hemicyonines, joined in the later part of the period by a few borophagine canids. However, this era can be subdivided into two intervals. The first interval (Interval B), from ∼23.7 to ∼17.5 Ma (late Arikareean through early Hemingfordian), is characterized by temnocyonine and daphoenine amphicyonids that seem to be primarily if not entirely North American endemics, diluted by the arrival of the earliest species of amphicyonines and hemicyonines. The second interval (Interval C), from ∼17.5 to 9 Ma (medial Hemingfordian through Clarendonian), sees the extinction of all temnocyonines and daphoenines; the larger carnivorans are amphicyonine, hemicyonine, and felid immigrants, and the endemic borophagine canids.

Temnocyonine-Daphoenine-Entelodont Association (Interval B: ∼23.7–17.5 Ma)

Beginning in the earliest Arikareean, the first temnocyonine amphicyonids are recorded in North America at Logan Butte in the John Day beds of Oregon (∼29–29.5 Ma), in the Sharps Formation of the Wounded Knee area, South Dakota (LACM Locality 1872, estimated at ∼28–29.5 Ma), and in pumice-bearing levels of the Gering Formation at Wildcat Ridge, Nebraska (∼28.3 Ma). These early temnocyonines attained the size of coyotes or small wolves (∼15–30 kg). Identified by a uniquely specialized dentition, temnocyonines are present throughout the Arikareean NALMA, and become extinct in the latest Arikareean. The last documented occurrences of temnocyonines are in Upper Harrison sediments in northwest Nebraska and southeastern Wyoming. Temnocyonines actually appear late in Interval A and extend into the early part of Interval B and are the only group of large carnivorans, other than the endemic hesperocyonine canids, spanning the Arikareean NALMA. Terminal species of large temnocyonines (∼80 kg) in the latest Arikareean attained skull lengths of ∼30 cm.

With the extinction of the hyaenodonts, nimravids, and Daphoenus, the mid-Arikareean interval (∼25–23 Ma) displays a dearth of large carnivores. Remnocyonines diversity at this time, seeming to fill this void.1 However, the failure of other carnivorans to increase in size and occupy the available meat-eating and carcass-consuming niches allows a group of artiodactyls, the Entelodontidae, to successfully enter this domain.Entelodonts increase in size from the late Oligocene into the early Miocene. By the time of their extinction at ∼17.5 Ma, they were of enormous size, large males approaching ∼800 kg. Documented evidence of scavenging by entelodonts is now known from a number of late Arikareean sites in North America (Joeckel, 1990; Hunt, 1990); it is likely that they were important carcass processors, able to crush bones of large ungulates. Both entelodonts and temnocyonines possess large, robust premolars suitable for durophagy. Their worn premolars commonly exhibit flattened tips with crimped enamel rims similar to the blunted premolars of living durophagous hyenas.

In the mid- to late Arikareean, beardogs of the genus Daphoenodon appear together with temnocyonines at many localities in the Great Plains and in Florida. At the carnivore den site at Agate National Monument a large temnocyonine was found in a burrow only a meter from dens with individuals of Daphoenodon superbus (Hunt et al., 1983; Hunt, 1990). The genus Daphoenodon includes several lineages, all probably descended from the earliest species, D. notionastes, known only from Florida (Frailey, 1979) and the Gulf Coast (Albright, 1996). Some species maintain a short-legged subdigitigrade stance, but one lineage evolves a long-legged digitigrade predator, Borocyon, that survives to the end of Interval B and attains great size (basilar skull lengths, 28– 30 cm). Borocyon last occurs in the early Hemingfordian Runningwater Formation of Nebraska, and with its demise, daphoenine amphicyonids become extinct and are replaced by immigrant amphicyonines.

Endemic New World daphoenines (Daphoenodon, Borocyon) and temnocyonines coexist with Old World amphicyonines (Ysengrinia, Amphicyon, Cynelos) within Interval B. These are the largest terrestrial carnivorans (∼50–100 kg) evolved on the continent up to this time (Hunt, 1998b). The immigrant amphicyonines Ysengrinia, Cynelos, and Amphicyon appear at ∼23.0, ∼19.2, and ∼18.8 Ma, respectively, and herald the beginning of a Eurasian amphicyonine migration into North America that continues into the mid-Miocene.

Accompanying the amphicyonines are immigrant hemicyonine ursids. The oldest North American records of hemicyonines are referable to Cephalogale, discovered in late Arikareean sediments of western Nebraska and southeastern Wyoming (Hunt, 1998a). New World species of Cephalogale range from ∼23 to ∼17.5 Ma. The youngest records of the genus are found in early Hemingfordian sites in western Nebraska and Florida where they are as large as a small Ursus americanus.

An enormous predatory mustelid, Megalictis ferox, also is confined to Interval B. Characterized by a robust, short-limbed postcranial skeleton indicative of powerful musculature, coupled with a short-faced cranium, shearing carnassials, strong canines, and well-developed crushing premolars, Megalictis apparently functioned as a giant wolverine-like predator (Hunt and Skolnick, 1996).

Canids are endemics in Intervals A–C, restricted to North America. Throughout Interval B, canids are relatively small animals, the hesperocyonines and borophagines reaching skull lengths of ∼14–18 cm in only a few species (Wang, 1994; Wang et al., 1999).

Amphicyonine-Hemicyonine-Borophagine Association (Interval C: ∼17.5–9 Ma)

With the disappearance of endemic daphoenines at the end of Interval B (∼17.5 Ma), the Old World amphicyonines (Amphicyon, Cynelos) became the largest carnivorans of the Hemingfordian NALMA. By the late Hemingfordian Sheep Creek fauna of western Nebraska, these two genera reached skull lengths of 33–34 cm (Cynelos idoneus) and 37–39 cm (Amphicyon frendens). These amphicyonine lineages eventually attained maximum size in the early Barstovian (Cynelos sinapius, 39–44 cm; Amphicyon ingens, 42–52 cm; Pliocyon, 29–30 cm), immediately prior to their apparent extinction by ∼14 Ma. With the disappearance of the large scavenging entelodonts, it is likely that these amphicyonines processed carcasses in addition to active hunting. The appearance of hyaenoid borophagines at this time also suggests a response to the availability of the durophagous niche, climaxing with the huge bone-crushing canid Epicyon haydeni (basilar skull lengths, late Clarendonian–early Hemphillian, 28–32 cm) in the Ogallala Formation of the Great Plains.

There is a major shift in the large amphicyonines at ∼14 Ma, the boundary between early Barstovian and early late Barstovian faunas in the Great Plains. Amphicyon, Cynelos, and Pliocyon are replaced by new lineages of large Pseudocyon and mid-sized Ischyrocyon, which then dominate the later Barstovian and Clarendonian of North America. The only known skull of Pseudocyon from the late Barstovian of Nebraska has a basilar length of 37 cm, and several mandibles also indicate its great size. Ischyrocyon, known from a number of skulls from 36 to 47 cm, attains enormous size in the Clarendonian. The earliest populations of Ischyrocyon from the Barstow syncline, California, are somewhat smaller (basilar skull lengths, 28–34 cm, N = 7).

Early Barstovian canids contemporaneous with the Amphicyon-Cynelos-Pliocyon group remain of modest size: the largest are the borophagines Tomarctus (basilar lengths, ∼19–19.5 cm), Paracynarctus (∼14–16 cm), Aelurodon (∼20 cm), and the hesperocyonine Osbornodon (∼20 cm). Coincident with the appearance of the Pseudocyon-Ischyrocyon group at ∼14 Ma, borophagine canids begin a steady and conspicuous size increase, documented by larger species of Aelurodon, Carpocyon, Protepicyon, and cynarctines in the later Barstovian. With the extinction of the great amphicyonids near the end of the Clarendonian, the late Clarendonian and Hemphillian are typified by large species of the borophagine canids Epicyon and Borophagus.

In the later Hemingfordian, near the beginning of Interval C, the carnivorous Eurasian hemicyonine Phoberocyon appears in eastern North America, represented by dental and postcranial remains at Thomas Farm in Florida (Tedford and Frailey, 1976) and possibly in Delaware (Emry and Eshelman, 1998:158). During the early Barstovian, Phoberocyon is succeeded by the Old World hemicyonine Plithocyon, known from a large sample from the Barstow syncline, California, and from rare remains from New Mexico and Nebraska. These hemicyonine ursids differed from the living ursine bears in their less specialized, hypercarnivorous dentitions (relative to ancestral amphicynodont ursids; Hunt, 1998a) and their long-footed, digitigrade stance, suggesting that they were active pursuit predators.

Early in Interval C, at ∼16.5–17 Ma, the first felids enter North America from Eurasia and persist as relatively small lynx- to leopard-sized cats until the appearance of the large lion-sized felid Nimravides in the Clarendonian. Also, true ursine bears (Ursavus), although never very large at this time (<50 kg), become evident in Hemingfordian and Barstovian (but not Clarendonian) faunas, probably inhabiting well-vegetated environments as isolated individuals.

Borophagine Canid-Ursid Indarctos-Felid Association (Interval D: ∼9–6 Ma)

During the early Hemphillian (∼9–7 Ma), the large canids Epicyon, Borophagus, and the smaller Carpocyon are joined by the great immigrant ursavine bear Indarctos. Apparently Indarctos oregonensis, the large borophagine dogs (Epicyon haydeni), the large long-limbed felid Nimravides, and the barbourofeline nimravid Barbourofelis have replaced the great amphicyonines and hemicyonines, marking a radical alteration in the large carnivore assemblage. The “cats” and borophagines are species that have continued to increase in size from the Clarendonian faunas into the Hemphillian. At ∼7 Ma, later in the Hemphillian interval, appear additional Eurasian immigrant species: the machairodont felid Machairodus, the arctoid “dog” Simocyon, and the mustelid Eomellivora (Tedford et al., 1987). The earliest tremarctine ursid Plionarctos also appears at this time (∼7 Ma; Tedford and Martin, 2001; Hunt, 1998a). Intervals D and E have been commented on by others (Tedford et al., 1987; Webb and Opdyke, 1995) as a time of renewed immigration from the Eurasian mainland across the Bering route into North America, and a number of species of smaller carnivores and mammals also participate in this migration.

Borophagine Canid-Ursid Agriotherium-Felid Association (Interval E: ∼6–4.5 Ma)

The late Hemphillian is populated by new faunal elements, recording a continuing dilution of the North American carnivoran fauna by Old World species. At ∼6 Ma the migrant ursid Agriotherium and mustelid Plesiogulo enter, soon followed by the machairodont felid Megantereon. The nimravid cats and simocyonine arctoids are absent and presumably extinct. A diverse group of smaller felids (Lynx, Adelphailurus, Pratifelis) and the large Nimravides are joined for the first time in North America by several machairodont felids (Machairodus, Megantereon), and this coexistence of machairodonts and the less specialized felines continues into the late Pleistocene. Modern procyonids (Procyon) also make their first appearance. The canids Epicyon, Borophagus, and Carpocyon continue through this interval, suggesting that the endemic canid assemblage is largely unperturbed, simply receiving occasional pulses of immigrant stocks from Eurasia.

However, a number of extinctions of characteristic taxa occur at the end of the interval at ∼4.5–4.7 Ma (Hemphillian-Blancan boundary). The felids Machairodus and Nimravides, the ursid Agriotherium, and the canids Epicyon and Carpocyon are missing from Interval F, representing a significant shift in the carnivoran guild in the early Pliocene.

Early Canid-Felid-Ursid Association with Machairodont and Hyaenoid Canid (Interval F: ∼4.5–3 Ma)

At ∼4.5 Ma the earlier Blancan faunas include species of felids, canids, and ursids that reflect at least to some degree the modern representatives of these families in North America. Ursus appears near the beginning of the interval (∼4.3–4.5 Ma), as does Puma (Martin, 1998). Ursus is accompanied by the tremarctine bear Plionarctos, marking the initiation of a shared occupation of North America by Ursus and tremarctines that will extend to the Holocene. The canine canids that arise in the Hemphillian begin to radiate into the lines leading toward modern foxes, coyotes, and wolves, but most Blancan canines (the ancestral coyote Canis lepophagus, the small wolflike Canis edwardii, and foxes) are small.

Archaic holdovers that remain in the interval include the tremarctine ancestor Plionarctos, the machairodont Megantereon, and the hyaenoid canid Borophagus.

Modern Canid-Felid-Ursid Association with Machairodont and Hyaenid (Interval G: ∼3 Ma–Holocene)

The presence of modern groups ancestral or closely related to the living North American canids (Canis, Vulpes, Urocyon), felids (Felis, Lynx, Puma, Panthera), and ursids (Ursus sensu lato) characterize the interval. The tremarctine ursid Tremarctos first occurs at ∼2.5 Ma in California and Idaho, and its sister taxon the arctodont bear Arctodus, the largest land carnivoran to have ever existed, appears late in the interval at ∼1 Ma, accompanying the much smaller species of New World Ursus.

The late Cenozoic radiation of the canine canids (subfamily Caninae), stemming from the small early and mid-Miocene Leptocyon, was initiated in the Hemphillian. Only small foxes (Urocyon, Vulpes) coexisted with coyote-sized “Canisdavisi (Berta, 1987) in the Hemphillian, and so canines did not yet rival in size the large carnivorans of the latest Miocene in North America. Canine diversification took place in North and possibly Central America from ∼6 to 3 Ma (Berta, 1987). By ∼3.5 Ma in the Blancan a coyote-sized canid, Canis lepophagus, was widely distributed and is a possible ancestor of Canis latrans (Kurtén and Anderson, 1980). No wolf-sized canids were present in the New World at this time. Early wolves, Canis etruscus, evolved in the Old World in the early Pleistocene from the C. arnensisC. lepophagus group of Holarctic coyotelike canines (Kurtén and Anderson, 1980). A small wolf-like canid does not occur in North America until the appearance of Canis edwardii in the late Blancan and Protocyon texanus (known only from the American Southwest) in the early Irvingtonian. These small wolves are followed by the larger C. armbrusteri in the late Irvingtonian, and the arrival of Canis lupus, the gray wolf, in the latest Irvingtonian and Rancholabrean. The large dire wolf (Aenocyon dirus) of North America is contemporary with the gray wolf in the Rancholabrean. Thus, large wolf-sized canids are a relatively recent phenomenon in North America, appearing only in the early Pleistocene (Irvingtonian) and continuing to the present. These wolves do not become important in the large carnivoran assemblage of North America until the extinction of the hunting hyena Chasmaporthetes at ∼1.5 Ma in the early Irvingtonian and the disappearance of the canid Borophagus at or shortly after ∼2 Ma.

An archaic and rather exotic component remains early in the interval: in the later Blancan faunas the ailurid Parailurus in the Pacific Northwest and a number of aeluroid carnivores appear from ∼3.5 to 2.8 Ma, including the Old World hyaenid Chasmaporthetes (the only hyaenid to ever reach the New World), the machairodont Homotherium, and the felines Dinofelis and Miracinonyx. The machairodont Megantereon persists from the early Blancan and is believed to evolve into the late Blancan and Irvingtonian Smilodon at ∼2.4 Ma.

The advent of the mass extinctions at the Pleistocene-Holocene boundary sees the demise of many of the great carnivorans of Intervals F and G: gone are Homotherium in the Wisconsinan; Miracinonyx at ∼19,000– 20,000 years; Arctodus at ∼12,000–13,000 years; Smilodon at ∼8,000 years. Earlier last occurrences in the interval included the hyaenids at ∼1.5 Ma and the felid Dinofelis at ∼2–2.3 Ma.

Oxygen Isotopes, Cenozoic Global Climate, and Intervals A–G

Although the fossil history of larger carnivorans and creodonts is punctuated by numerous temporal gaps in the Cenozoic nonmarine sedimentary record of North America, the richer sampling of lineages made possible by renewed field efforts and the opening of the Frick Collection in New York, together with a progressively more refined lithostratigraphy and biochronology of North America mammals, make possible an improved assessment of faunal turnover in these groups. Most striking in the compilation of table 11.1 is the relative duration of these intervals through time. Intervals A through C are significantly longer than those to follow (13.3, 6.2, 8.5 m.y. versus 3.0, 1.5, 1.5, and 3.0 m.y.). Although the boundary between intervals is not always sharply defined, with some taxon ranges overlapping and others beginning or ending diachronously, the composition of the guilds of larger carnivores reflects an evident measure of faunal stability over these temporal intervals.

Recent advances in our understanding of global climate and tectonism during the Cenozoic provide a template against which we can evaluate Intervals A through G. Earth's climate has been profoundly affected by the orbital behavior of the planet, and by crustal movements and deformation generated by plate tectonics. Insight into Cenozoic climate change has been considerably enhanced by the development of deep-ocean oxygen and carbon isotopic records that supply detailed information on long-term as well as abrupt, or “transient”, shifts in global climate (Miller et al., 1987, 1991; Prentice and Matthews, 1988; Denton, 1999; Zachos et al., 2001). Earlier discussions of Cenozoic climate tended to emphasize a nearly continuous climatic cooling from the early Paleogene into the Pleistocene. Newer high-resolution isotopic data have revealed a more complex and varying climatic signal characterized by intervals of broad climatic stability of variable duration that record cooling and warming events. These intervals are themselves made up of shorter oscillating periods of cooling and warming, some exceedingly brief (104–105 years).

Here I compare the North American Intervals A–G (fig. 11.1) with the recently published deep-sea stable oxygen isotope record for benthic foraminifera compiled from over 40 ocean drilling sites that sampled the Cenozoic record of sedimentation (fig. 11.2, from Zachos et al., 2001). These intervals compare reasonably well with the δ18O curve; several large carnivore faunal turnover events occur at evident shifts in oxygen isotope values of the deep-sea curve. Moreover, subevents (e1–e3) within Intervals C and D appear to correspond to abrupt, or “transient”, oxygen isotope enrichment events suggestive of marked climatic cooling. Next I discuss the more detailed correspondence of Intervals A–G with the Zachos et al. (2001) curve.

Oxygen isotope curves for the Cenozoic (Miller et al., 1987; Prentice and Matthews, 1988; Zachos et al., 2001) record an evident shift toward greater enrichment in δ18O ratios (e.g., 0.0 to 1.8‰, Zachos et al., 2001) from the end of the early Eocene (∼49 Ma) to the beginning of the late Eocene (∼37 Ma). The shift from mid- to late Eocene corresponds to the advent of the Chadronian carnivore fauna in North America characterized by the large carnivore guild of Interval A (Hunt, 1996: fig. 2). Prior to Interval A, the Paleocene-Eocene carnivores of North America, the miacoids and creodonts, were not large mammals, most much less than 50 kg and many much smaller (∼1–20 kg).2 The shift from a miacoid-creodont dominated fauna in the Paleocene and earlier Eocene to the fauna of Interval A corresponds to the global climatic changes now documented at the end of the middle Eocene (∼37 Ma, Berggren et al., 1995; Prothero and Emry, 1996: 679).

A more moderate shift occurred in North America at the Eocene-Oligocene boundary at ∼33.7 Ma (extinction of brontotheres, the creodont Hemipsalodon, several species of hyaenodonts, the archaic amphicyonids Daphoenictis and Brachyrhynchocyon, oromerycid artiodactyls, and cylindrodont rodents), here designated as the Chadronian-Orellan boundary between subintervals A1 and A2 (fig. 11.2). In the Old World the Eocene-Oligocene mammalian faunal shift was a more pronounced event (Grand Coupure, ∼33.7 Ma, Montanari et al., 1988; Lévêque, 1993).

From the EECO (early Eocene climatic optimum, 50–52 Ma) a 17-million-year trend toward cooler climate is indicated by a 3.0‰ enrichment in δ18O from 50 to 34 Ma (early Eocene to early Oligocene). Prior to the late Eocene this enrichment is attributed to a 7°C. decline in deep ocean temperature (Zachos et al., 2001). All later δ18O change is regarded as the combined effects of ice-volume and temperature, “particularly for the rapid enrichment event at 34 Ma” (Zachos et al., 2001).

The Oligocene oxygen isotope curve from ∼34 to 26 Ma shows δ18O values of 2.4 to 3.0‰, indicating a marked enrichment in δ18O over the Eocene interval where values were between 0.0 and 2.0‰ (Zachos et al., 2001). Permanent ice sheets with mass as great as 50% of the present-day ice sheet are inferred during this early to mid-Oligocene interval (Zachos et al., 1994, 2001). From ∼33.7 Ma to ∼26 Ma, the mammals of the Orellan, Whitneyan, and early Arikareean NALMAs represent a uniform faunal aggregate of species evolving without major interruption in faunal content; this is likewise an interval of uniform δ18O values until the late Oligocene warming event—the Zachos et al. (2001) curve places this event at ∼26–24 Ma, immediately prior to the Miocene Mi-1 glaciation.

Within Interval A2, a warming spike at ∼28.3 Ma coincides with an episode of major erosional incision in the central Great Plains in which much of the upper White River Group is removed at a number of localities in the late Oligocene, followed by deposition of the lower Arikaree Group. This event is coincident with a prominent fall in sea level recorded during the initiation of supercycle TB1 in the stratigraphic record (Haq et al., 1988: fig. 14; Woodburne and Swisher, 1995: fig. 3).

From ∼26 to 24 Ma, a marked warming trend is recorded in the δ18O values, which fall to between 1.4‰ and 2.0‰. From ∼26 Ma until the mid-Miocene (∼14–15 Ma), oxygen isotope values are believed to indicate reduced global ice volume and cooler bottom-water temperatures, punctuated by several brief glaciations (Mi events). These glaciations and concomitant lowering of sea level may have allowed brief migrations at these times via the Bering land bridge.

At approximately the Oligocene-Miocene boundary (23.7–24.0 Ma), a marked cooling event occurs (Mi-1 glaciation, fig. 11.2), evidenced by an enrichment spike in the δ18O ratio to 2.6‰. The turnover of large carnivorans between Intervals A and B is closely tied to the Oligocene-Miocene boundary, and to the transformation of the early to late Arikareean mammal fauna within North America, an event that coincides with the strong global temperature decline recorded in the oxygen-isotope curves. This also coincides with a major erosional interval in the Great Plains between lower Arikaree Group sediments and the upper Arikaree Group. 40Ar/ 39Ar dating of a tuff in the Harrison Formation in northwestern Nebraska indicates an age of ∼23 Ma (sanidine, 22.9 ± 0.08; Izett and Obradovich, 2001) above a major unconformity with lower Arikaree sediments. The composition of the Harrison mammal fauna demonstrates a profound alteration of taxa from the early Arikareean assemblage that preceded it. Thus, the agreement of the faunal shift, the regional erosional event within the Arikaree Group, and the δ18O enrichment spike at ∼24 Ma are particularly compelling in suggesting a major climatic shift at the Oligocene-Miocene boundary that may be coupled with a marked global lowering of sea level that takes place from ∼23 to 20 Ma (Prentice and Matthews, 1988). Sea level fall associated with the Mi-1 glaciation probably made possible a migration event via the Bering route in which late Arikareean carnivores (Ysengrinia americana, Cephalogale, Zodiolestes, Cynelos) and other mammals (chalicothere Moropus, entelodont Dinohyus, rhinoceros Menoceras, small dromomerycids and moschids) entered the New World.

Thereafter, a warming climate persists from the early Miocene until the mid-Miocene climatic optimum (fig. 11.2, MMCO; Zachos et al., 2001), which ends at ∼14 Ma. Within this interval additional warming is indicated from ∼17.5 to ∼16.2–16.0 Ma (Box Butte–Sheep Creek interval of the Hemingfordian NALMA). An abrupt δ18O enrichment event occurs at ∼16 Ma, followed again by climatic warming from ∼16 to ∼15 Ma. From ∼15 Ma a steady climatic deterioration begins that continues to the present, during which δ18O values increase from 1.6‰ in the mid-Miocene to 3.0–5.0‰ in the Pleistocene, accompanied by the development of the Antarctic and Arctic ice sheets.

At ∼15–14 Ma the steady and pervasive enrichment in δ18O values marks the end of the mid-Miocene climatic optimum (Zachos et al., 2001). A dramatic shift in the large amphicyonids (e1 event, fig. 11.2) and the progressive rise of the borophagine canids at this time suggests that the great beardogs of the late Hemingfordian and early Barstovian were somehow tied to the stable climatic interval of the early and mid-Miocene. Within Interval C, the e1 event marks a pronounced transition within the large carnivoran community. At ∼14 Ma the Amphicyon-Cynelos-Pliocyon group was replaced by equally large Pseudocyon and Ischyrocyon species, preceded somewhat earlier by extinction of the last hesperocyonine canids. The ∼14 Ma event marks the beginning of an δ18O enrichment phase that persists until ∼9 Ma, with δ18O values steadily increasing from +2.0 to +3.0 ‰. The extinction of amphicyonids and hemicyonine ursids coincides with the 9 Ma date.

From ∼9 to ∼6 Ma (Interval D), there is a period of more uniform δ18O values in the Zachos curve corresponding to the Hemphillian NALMA. The preceding period of global cooling may have contributed to aridification of the North American continental interior at this time, and development of the interior grassland faunas of the late Miocene. Two subevents during Interval D appear to coincide with abrupt oxygen isotope enrichment spikes. They occur within a brief warming trend from ∼9 to ∼7 Ma. At ∼8 Ma the appearance of the migrant Old World ursid Indarctos corresponds to subevent e2. At ∼7 Ma, an abrupt enrichment event accompanies a significant sea level fall that coincides with the immigration of several Eurasian carnivorans (Machairodus, Simocyon, Plesiogulo, ?Plionarctos) via the Bering route into North America. Similar abrupt enrichment events also occur in the Zachos δ18O curve at ∼6 Ma and ∼4.5 Ma, which mark the boundaries between Intervals D–E and E–F. Zachos et al. (2001) recognized a “subtle warming trend” from 6 to ∼3.2 Ma, “when oxygen isotope values increased reflecting the onset of Northern Hemisphere Glaciation”. Sea level fall in Beringia may have contributed to the migration events recorded at these times.

It becomes evident that as δ18O values continue to increase on average through the Pliocene and Pleistocene, many instances of sea level lowering provide avenues across Beringia for mammal migration. Episodic appearances in the New World of large carnivores with Eurasian affinities are likely tied to such climatic events.

From 4.5 to ∼3.2 Ma the climate in the early Pliocene remains somewhat stable and warm, but after ∼3 Ma a cooling trend begins. From ∼3.2 to ∼2 Ma the effect of global cooling and the full development of both Antarctic and northern hemisphere ice masses undoubtedly influenced the mammalian fauna of the northern continents, and resulted in the shift to a modern assemblage of large carnivores in which large canids, ursids, and felids became the dominant species. The marked shift to oxygen isotope values >+3.0 occurs from ∼3.2 to 2 Ma and corresponds to the transition from Interval F to G.

At the beginning of the Pleistocene at ∼1.8 Ma the oscillations in the δ18O values increase noticeably, and a succession of glacial-interglacial events continues from ∼2.4 Ma to the present. Such climatic oscillations are likely related to (and perhaps largely responsible for) shifts in carnivoran ranges, and also for certain speciation events and extinctions, on the northern continents.

Comparison of Intervals A–G with Cenozoic Mammal Migration Events

Intervals A to G are broadly comparable with land mammal migration episodes in the Cenozoic of North America identified by Webb and Opdyke (1995). Timing of these migration events was compared to the Prentice-Matthews (1988) δ18O curve for the Cenozoic, based on equatorial planktonic foraminifera. Correspondences were also noted with the sequence stratigraphic onlap-offlap chronology of Haq et al. (1988). Boundaries of Intervals D through G generally correspond to the migration episodes of Webb and Opdyke for the Hemphillian and Blancan land mammal ages, including the Clarendonian-Hemphillian boundary. More recent work on faunas of the late Oligocene–early Miocene Arikaree and Hemingford Groups of the central Great Plains has improved the resolution of taxon ranges of large carnivores and other mammals for Intervals A–C.

Webb and Opdyke (1995) noted important mammalian immigration events in the early Miocene, two first-order and one second-order, placed at 21, 20, and 18 Ma. These occurred in the latest Arikreean (six genera), early Hemingfordian (14 genera), and medial Hemingfordian (10 genera). They remarked on the close temporal spacing of the two events at ∼21 and ∼20 Ma. The placement of the three events can be revised here so that they occur at 23.0, 19.2, and 18.8–18.5 Ma. The ∼23 Ma date is based on 40Ar/39Ar dating of the Agate Ash (Izett and Obradovich, 2001) in the Harrison Formation, marking the initiation of upper Arikaree Group deposition in the Great Plains. The ∼19.2 Ma date (FT, zircon) establishes the earliest Upper Harrison deposition in the same region (Hunt et al., 1983); and ∼18.8–18.5 Ma defines the earliest Runningwater Formation sediments based on paleomagnetic calibration (MacFadden and Hunt, 1998). The mammalian taxa associated with these intervals are listed in table 11.2.

Webb and Opdyke's (1995) “latest Arikareean” event is better placed at or near the Oligocene-Miocene boundary at ∼23–23.7 Ma (Interval A–B boundary). This event is coincident with a major sea level fall (Haq et al., 1988) that could have made possible the introduction of Eurasian large carnivores and other mammals (table 11.2) into North America via the Bering land bridge. In the central Great Plains, Interval B includes 3 principal lithostratigraphic units: the two formations of the upper Arikaree Group (Harrison Formation, “Upper Harrison” beds) and the oldest unit of the Hemingford Group (Runningwater Formation). Mammals from these rock units have collectively been named the “Runningwater Chronofauna” (Webb and Opdyke, 1995) because of an evident faunal continuity through the interval, particularly among the ungulate lineages (e.g., camels, oreodonts, rhinos, horses, protoceratids, dromomerycids).

Within this chronofauna, Webb and Opdyke (1995) identified one of the most significant (first-order) migration events in the later Cenozoic, involving the appearance of 14 genera of mammals in the Runningwater Formation of western Nebraska. This event occurs within Interval B, herein dated at ∼18.8–18.5 Ma based on the paleomagnetic calibration of the initiation of Runningwater deposition in western Nebraska by MacFadden and Hunt (1998). A faunal shift among large amphicyonine carnivores (from subgroup I to II, fig. 11.1), the appearance of new lineages of daphoenines (Borocyon), the first appearance of the Eurasian ursine bear Ursavus, and the extinction of temnocyonines and the giant mustelid Megalictis (Hunt and Skolnick, 1996) mark this event. Thus, Interval B can be subdivided into B1 and B2 segments (fig. 11.1: B1, ∼23 to ∼18.8–18.5 Ma; B2, ∼18.8–18.5 to 17.5 Ma).

A profound change in the style of sedimentation is observed in the Great Plains at the boundary between Intervals B1 and B2, influencing faunal composition of the upper Arikaree units relative to the fauna of the Runningwater Formation. Arikaree Group sedimentation is primarily fine-grained eolian volcaniclastic sands and silts deposited over a level aggrading landscape east of the Rocky Mountain uplifts (Hunt, 1990). Fluvial deposits are largely limited to broad shallow streams that reworked the fine-grained eolian materials but remained confined to localized linear valley tracts. In contrast, Runningwater sediments include a much greater proportion of granitic, epiclastic coarse sediment derived from the uplifts to the west (Cook, 1965; Yatkola, 1978), deposited within an enormous and deep paleovalley complex that can be traced from southeastern Wyoming through northwest Nebraska to southwestern South Dakota. This “sudden” appearance of riparian fluvial environments at ∼18.8 Ma is the signature of Runningwater sedimentation, and these beds are replete with fish, alligator, salamander, aquatic turtles, frogs, and mammals in numerous local channel deposits that remain richly fossiliferous. Such depositional environments are very rare in Arikaree rocks. It is uncertain how many of the “migrant” mammals of the Runningwater are the result of the new wetter environments or whether they truly represent relatively synchronous introduction of Old World taxa into North America at this time. Many of these first appearances might be predicted in such riparian settings (e.g., the aquatic arctoid Potamotherium, otter Mionictis, mustelid Leptarctus, the ursine bear Ursavus, weasels and other small arctoid carnivorans, the procyonid Edaphocyon, small merycodont antilocaprids, and several insectivores).

Nonetheless, there is an integrity to the faunas of the upper Arikaree Group and Runningwater beds that is terminated by the hiatus between this latter formation and the Box Butte Formation of the medial Hemingfordian, which rests disconformably on the terminal paleosol (“Platy Bench”) of the Runningwater at many localities in western Nebraska (Galusha, 1975). Box Butte mammals show a greater affinity with the superjacent fauna of the Sheep Creek Formation (Tedford et al., 1987; Galusha, 1975). The break between the Runningwater and Box Butte faunas was recognized by Webb and Opdyke (1995) as a first-order migration event (with 10 migrant genera), corresponding to the Interval B–C boundary of this study. In the Great Plains this boundary is marked by the appearance of hypsodont horses with cement-covered cheek teeth, the arrival of the first true felids, a new large aceratherine rhinoceros and new mustelids, and the extinction of many taxa belonging to Interval B (table 11.2).

Interval C represents a period of relative stability in North American Miocene faunas, culminating in the mid-Miocene climatic optimum (MMCO), after which the steady increase in δ18O values heralds the initiation of global cooling and the development of large-volume Antarctic ice. Near the onset of climatic deterioration, the e1 subevent records the replacement of the large amphicyonine fauna (Amphicyon, Cynelos, Pliocyon) of the early Barstovian by the large amphicyonines (Pseudocyon, Ischyrocyon) of the medial-late Barstovian (Hunt, 1998b). Webb and Opdyke (1995) remarked on the notable decline in diversity of browsing and mixed feeding ungulates following the mid-Miocene cooling, and their replacement by grazing forms. The cooling event seems to favor the radiation of borophagine canids in North America as appropriate predators of the grazing ungulates.

The Interval C–D boundary marks the end of the Clarendonian NALMA, where a shift to the Hemphillian faunas has long been recognized (Tedford et al., 1987). Among the most striking extinctions at this time are the loss of the great amphicyonines and hemicyonines that characterize much of the Miocene of North America. Similarly, the aelurodontine and cynarctine canids also fail to survive the Clarendonian (Wang et al., 1999).

Within Interval D, Webb and Opdyke recognized as a second-order event the migration of the great bear Indarctos at ∼8 Ma, and also events at ∼7 and ∼6 Ma. The ∼8 Ma event (e2 event, fig. 11.2) corresponds to Mi-7 (Miller et al., 1991) and the sea level fall TB3.2 (Haq et al., 1988). The events at ∼7 Ma (e3 event, fig. 11.2) and ∼6 Ma (Interval D–E boundary) also now appear to correspond to oxygen isotope excursions of the Zachos et al. (2001) δ18O curve.

Also identified by Webb and Opdyke is a first-order migration episode at ∼5 Ma that appears to correspond to Interval E (from ∼6 to ∼4.5 Ma). Moreover, their event at ∼2.5 Ma, during the development of major Northern Hemisphere ice sheets, correlates approximately with the start of Interval G at ∼3– 3.2 Ma. Despite slight differences in the timing of these events, there is an evident correspondence between the migration events for Cenozoic mammals identified by Webb and Opdyke (1995) and the large carnivore intervals proposed here.

The correlation of mammal migration events to the oxygen isotope record from marine deposits of the global ocean was discussed by Webb and Opdyke (1995) as the probable result of development of land bridges (in the Neogene via the Bering Route) coincident with lowered sea level, accompanied by climate shifts that in and of themselves transformed the North American ecosystem. The continuing refinement of mammalian biochrons in North America, based on new collections and intensive study of older materials, and accompanied by more detailed field geological efforts, will provide a steadily improving database for such investigations.

Acknowledgments

For making available the Frick Collection of carnivores at the American Museum over several decades, I am especially grateful to M. C. McKenna and R. H. Tedford. Field and laboratory research was funded in part by the H. B. Meek Fund of the University of Nebraska and by grants from the National Geographic Society. The two figures were prepared by UNSM illustrator Angie Fox, whose skills and helpful suggestions were essential to the project. My colleague M. R. Voorhies shared his expertise on Barstovian and Clarendonian faunas in several discussions that improved my understanding of these intervals.

Malcolm McKenna served as mentor, advisor, and friend during my graduate study at the American Museum from 1965–70; his scholarship so ably evidenced in his studies in mammalian paleontology has served as an example many of us have attempted to follow, yet not equaled in originality and insight. I appreciate the invitation of Gina Gould and Susan Bell to contribute this paper.

REFERENCES

1.

L. B. Albright 1996. Insectivores, rodents, and carnivores of the Toledo Bend local fauna: an Arikareean (earliest Miocene) assemblage from the Texas Coastal Plain. Journal of Vertebrate Paleontology 16:3458–473. Google Scholar

2.

W. A. Berggren, D. V. Kent, C. C. Swisher, and M. P. Aubry . 1995. A revised Cenozoic geochronology and chronostratigraphy. In W.A. Berggren, D.V. Kent, M.P. Aubry, and J. Hardenbol (editors), Geochronology, time scales and global stratigraphic correlation: 129–212. Tulsa: SEPM (Society for Sedimentary Geology) Special Publication 54. Google Scholar

3.

A. Berta 1987. Origin, diversification, and zoogeography of the South American Canidae. In B.D. Patterson and R.M. Timm (editors), Studies in neotropical mammalogy: essays in honor of Philip Hershkovitz. Fieldiana: Zoology (n.s.) 39:455–471. Google Scholar

4.

H. N. Bryant 1991. Phylogenetic relationships and systematics of the Nimravidae (Carnivora). Journal of Mammalogy 72:56–78. Google Scholar

5.

H. N. Bryant 1996. Nimravidae. In D.R. Prothero and R.J. Emry (editors), The terrestrial Eocene-Oligocene transition in North America: 453– 475. New York: Cambridge University Press. Google Scholar

6.

H. J. Cook 1965. Runningwater Formation, middle Miocene of Nebraska. American Museum Novitates 2227:1–8. Google Scholar

7.

G. H. Denton 1999. Cenozoic climate change. In T.G. Bromage and F. Schrenk (editors), African biogeography, climate change, and human evolution: 94–114. New York: Oxford University Press. Google Scholar

8.

R. J. Emry and R. E. Eshelman . 1998. The early Hemingfordian (early Miocene) Pollack Farm local fauna: first Tertiary land mammals described from Delaware. In R.N. Benson (editor), Geology and paleontology of the lower Miocene Pollack Farm fossil site, Delaware. Delaware Geological Survey Special Publication 21:153–173. Google Scholar

9.

J. J. Flynn 1998. Early Cenozoic Carnivora (“Miacoidea”). In C. Janis, K. Scott, and L. Jacobs (editors), Evolution of Tertiary mammals of North America: 110–123. New York: Cambridge University Press. Google Scholar

10.

D. Frailey 1979. The large mammals of the Buda local fauna (Arikareean: Alachua County, Florida). Bulletin of the Florida State Museum, Biological Sciences 24:2123–173. Google Scholar

11.

T. Fremd, E. A. Bestland, and G. J. Retallack . 1994. John Day basin paleontology field trip guide and road log. 1994 Society of Vertebrate Paleontology Annual Meeting, Seattle, WA. Kimberly, OR: John Day Fossil Beds National Monument Publication 94-1. Google Scholar

12.

T. Galusha 1975. Stratigraphy of the Box Butte Formation, Nebraska. Bulletin of the American Museum of Natural History 156:1–68. Google Scholar

13.

B. U. Haq, J. Hardenbol, and P. R. Vail . 1988. Mesozoic and Cenozoic chronostratigraphy and eustatic cycles. Society of Economic Paleontologists and Mineralogists Special Publication 42:71–108. Google Scholar

14.

R. M. Hunt Jr. 1990. Taphonomy and sedimentology of Arikaree (lower Miocene) fluvial, eolian, and lacustrine paleoenvironments, Nebraska and Wyoming. In M.J. Lockley and A. Rice (editors), Volcanism and fossil biotas. Geological Society of America Special Paper 244:69–111. Google Scholar

15.

R. M. Hunt Jr. 1996. Amphicyonidae. In D.R. Prothero and R.J. Emry (editors), The terrestrial Eocene-Oligocene transition in North America: 476–485. New York: Cambridge University Press. Google Scholar

16.

R. M. Hunt Jr. 1998a. Ursidae. In C. Janis, K. Scott, and L. Jacobs (editors), Evolution of Tertiary mammals of North America: 174–195. New York: Cambridge University Press. Google Scholar

17.

R. M. Hunt Jr. 1998b. Amphicyonidae. In C. Janis, K. Scott, and L. Jacobs (editors), Evolution of Tertiary mammals of North America: 196– 227. New York: Cambridge University Press. Google Scholar

18.

R. M. Hunt Jr. and R. I. Skolnick . 1996. The giant mustelid Megalictis from the early Miocene carnivore dens at Agate National Monument, Nebraska: earliest evidence of dimorphism in New World Mustelidae (Carnivora, Mammalia). Contributions to Geology, University of Wyoming 31:135–48. Google Scholar

19.

R. M. Hunt Jr. and R. H. Tedford . 1993. Phylogenetic relationships within the aeluroid Carnivora and implications of their temporal and geographic distribution. In F. Szalay, M.C. McKenna, and M.J. Novacek (editors), Mammal phylogeny: placentals: 53–73. New York: Springer-Verlag. Google Scholar

20.

R. M. Hunt Jr., X-X. Xue, and J. Kaufman . 1983. Miocene burrows of extinct beardogs: indication of early denning behavior of large mammalian carnivores. Science 221:364–366. Google Scholar

21.

G. A. Izett and J. D. Obradovich . 2001. 40Ar/39Ar ages of Miocene tuffs in basin-fill deposits (Santa Fe Group, New Mexico, and Troublesome Formation, Colorado) of the Rio Grande system. The Mountain Geologist 38:277–86. Google Scholar

22.

R. M. Joeckel 1990. A functional interpretation of the masticatory system and paleoecology of entelodonts. Paleobiology 16:459–482. Google Scholar

23.

B. Kurtén and E. Anderson . 1980. Pleistocene Mammals of North America. New York: Columbia University Press. Google Scholar

24.

F. Lévêque 1993. Correlating the Eocene-Oligocene mammalian biochronological scale from SW Europe with the marine magnetic anomaly sequence. Journal of the Geological Society of London 150:661–664. Google Scholar

25.

B. J. MacFadden and R. M. Hunt Jr. . 1998. Magnetic polarity stratigraphy and correlation of the Arikaree Group, Arikareean (late Oligocene-early Miocene) of northwestern Nebraska. Geological Society of America Special Paper 325:143–165. Google Scholar

26.

L. D. Martin 1998. Felidae. In C. Janis, K. Scott, and L. Jacobs (editors), Evolution of Tertiary mammals of North America: 236–242. New York: Cambridge University Press. Google Scholar

27.

W. D. Matthew 1909. The Carnivora and Insectivora of the Bridger Basin, middle Eocene. Memoirs of the American Museum of Natural History 9:291–567. Google Scholar

28.

M. C. McKenna and S. K. Bell . 1997. Classification of mammals above the species level. New York: Columbia University Press. Google Scholar

29.

J. S. Mellett 1969. A skull of Hemipsalodon (Mammalia, Deltatheridia) from the Clarno Formation of Oregon. American Museum Novitates 2387:1–19. Google Scholar

30.

J. S. Mellett 1977. Paleobiology of North American Hyaenodon (Mammalia, Creodonta). Contributions to Vertebrate Evolution 1:1–134. Google Scholar

31.

K. G. Miller, R. G. Fairbanks, and G. S. Mountain . 1987. Tertiary oxygen isotope synthesis, sea level history, and continental margin erosion. Paleooceanography 2:11–19. Google Scholar

32.

K. G. Miller, J. D. Wright, and R. G. Fairbanks . 1991. Unlocking the icehouse: Oligocene-Miocene oxygen isotopes, eustasy, and margin erosion. Journal of Geophysical Research 96:B46829–6848. Google Scholar

33.

A. Montanari, A. Deino, R. Drake, B. Turrin, D. DePaolo, G. Odin, G. Curtis, W. Alvarez, and D. Bice . 1988. Radioisotopic dating of the Eocene-Oligocene boundary in the pelagic sequence of the northeastern Apennines. In I. Premoli-Silva, R. Coccioni, and A. Montanari (editors), The Eocene-Oligocene boundary in the Marche-Umbria basin (Italy), Ancona, Italy: 195–208. International Union of Geological Sciences, Commission on Stratigraphy. Google Scholar

34.

M. L. Prentice and R. K. Matthews . 1988. Cenozoic ice-volume history: development of a composite oxygen isotope record. Geology 16:963–966. Google Scholar

35.

D. R. Prothero and R. J. Emry . 1996. Summary. In D.R. Prothero and R.J. Emry (editors), The terrestrial Eocene-Oligocene transition in North America: 664–683. New York: Cambridge University Press. Google Scholar

36.

G. B. Schaller 1972. The Serengeti lion: a study of predator-prey relations. Chicago: University of Chicago Press. Google Scholar

37.

A. R. E. Sinclair and M. Norton-Griffiths . (editors). 1979. Serengeti: dynamics of an ecosystem. Chicago: University of Chicago Press. Google Scholar

38.

R. H. Tedford and D. Frailey . 1976. Review of some Carnivora (Mammalia) from the Thomas Farm local fauna (Hemingfordian: Gilchrist County, Florida). American Museum Novitates 2610:1–9. Google Scholar

39.

R. H. Tedford and J. Martin . 2001. Plionarctos, a tremarctine bear (Ursidae: Carnivora) from western North America. Journal of Vertebrate Paleontology 21:2311–321. Google Scholar

40.

R. H. Tedford, M. F. Skinner, R. W. Fields, J. M. Rensberger, D. P. Whistler, T. Galusha, B. E. Taylor, J. R. Macdonald, and S. D. Webb . 1987. Faunal succession and biochronology of the Arikareean through Hemphillian interval in North America. In M.O. Woodburne (editor), Cenozoic mammals of North America: 153– 210. Berkeley: University of California Press. Google Scholar

41.

X-M. Wang 1994. Phylogenetic systematics of the Hesperocyoninae (Carnivora: Canidae). Bulletin of the American Museum of Natural History 221:1–207. Google Scholar

42.

X-M. Wang, R. H. Tedford, and B. E. Taylor . 1999. Phylogenetic systematics of the Borophaginae (Carnivora: Canidae). Bulletin of the American Museum of Natural History 243:1–391. Google Scholar

43.

S. D. Webb and N. D. Opdyke . 1995. Global climatic influence on Cenozoic land mammal faunas. In S.D. Webb, and N.D. Opdyke (editors), Effects of past global change on life: 184–208. Washington, DC: National Academy Press. Google Scholar

44.

M. O. Woodburne and C. C. Swisher III. . 1995. Land mammal high-resolution geochronology, intercontinental overland dispersals, sea level, climate, and vicariance. In W.A. Berggren, D.V. Kent, M.P. Aubry, and J. Hardenbol (editors), Geochronology, time scales and global stratigraphic correlation: 335–364. Tulsa, OK: SEPM (Society for Sedimentary Geology) Special Publication 54. Google Scholar

45.

D. A. Yatkola 1978. Tertiary stratigraphy of the Niobrara River valley, Marsland Quadrangle, western Nebraska. Nebraska Geological Survey Special Paper 19:1–66. Google Scholar

46.

J. Zachos, L. D. Stott, and K. C. Lohmann . 1994. Evolution of early Cenozoic marine temperatures. Paleooceanography 9:2353–387. Google Scholar

47.

J. Zachos, M. Pagani, L. Sloan, E. Thomas, and K. Billups . 2001. Trends, rhythms, and aberrations in global climate 65 Ma to Present. Science 292:686–693. Google Scholar

 Fig. 11.1. Taxon range diagram of larger late Eocene to Pleistocene carnivores (carnivorans and creodonts) in relation to the North American Land Mammal “Ages” (NALMAs) and Lyellian epochs, calibrated against the Magnetic Polarity Time Scale of Berggren et al. (1995). North American Larger Carnivore Turnover Events (NALCTE) are indicated at ∼37–36, 33.7, 23.8, 18.8, 17.5, 9.0, 6.0, 4.5, and 3.0 Ma (see text for discussion). Temporal intervals between the turnover events are designated Intervals A through G. Amphicyonine Group I, Ysengrinia, Cynelos; II, Amphicyon, Cynelos, Pliocyon; III, Pseudocyon, Ischyrocyon. Hemicyonine I, Cephalogale; II, Phoberocyon; III, Plithocyon; IV, new Clarendonian genus

i0003-0090-285-1-139-f101.gif

 Fig. 11.1. Continued.

i0003-0090-285-1-139-f102.gif

 Fig. 11.2. The Cenozoic oxygen isotope curve of Zachos et al. (2001) and global climatic events relative to the guilds of larger carnivores designated by Intervals A through G. Subevents e1–e3 within Intervals C and D are discussed in text (modified from Zachos et al., 2001)

i0003-0090-285-1-139-f02.gif

TABLE 11.1 Late Eocene to Recent Faunal Associations of Large Carnivores in North America

i0003-0090-285-1-139-t01.gif

TABLE 11.2 First and Last Occurrences of Early Miocene Mammals in the Harrison, Upper Harrison, and Runningwater Lithostratigraphic Units of Western Nebraska and Southeastern Wyoming

i0003-0090-285-1-139-t02.gif

[1] Nonmarine rocks in North America dating from ∼25 to 23 Ma have been difficult to confidently identify. In the John Day Formation of north-central Oregon, recent dating of tuffs (Fremd et al., 1994) has established that the interval between the Tin Roof Tuff (∼25.1–25.3 Ma) and the ATR Tuff (22.6 ± 0.13 Ma) falls within this time interval. However, no large carnivores have been recently collected from this interval nor can fossils from the older 19th-century collections be placed with certainty in these beds. The Arikaree Group of the Great Plains and Rocky Mountain basins may include sediments of this age, but reliable radiometric dates have not been obtained because of the scarcity of datable tuffs. In intervals within the Arikaree Group that may fall within the ∼25 to 23 Ma window, temnocyonine beardogs are the only large carnivorans that have been found.

[2] Mesonychians, long considered creodonts but now removed from that order (McKenna and Bell, 1997), existed in North America from the early Paleocene to mid-Eocene and undoubtedly filled a scavenging, omnivorous, if not strictly carnivorous role for most species. Placed in three families (Mesonychidae, Hapalodectidae, Triisodontidae), they were nearly all larger than contemporary miacoids and many creodonts.

[3] Issued June 30, 2004

ROBERT M. HUNT Jr. "Chapter 11: Global Climate and the Evolution of Large Mammalian Carnivores during the Later Cenozoic in North America," Bulletin of the American Museum of Natural History 2004(285), 139-156, (1 June 2004). https://doi.org/10.1206/0003-0090(2004)285<0139:C>2.0.CO;2
Published: 1 June 2004
Back to Top