A series of 33 Late Cretaceous (earliest Campanian through Maastrichtian) paleoshoreline maps was developed to document the migrational evolution of the western edge of the North American Western Interior Seaway. The maps represent a geologic span of roughly 18 million years, and portray the estimated positions of the strandline for each standard Western Interior ammonite zone, beginning with the Clioscaphites choteauensis zone and continuing to the end of the Mesozoic. We attempted correlation of all significant mammal-bearing localities known from the Western Interior with the ammonite-based marine zonation. First approximations of correspondence between ammonite zones and North American Land-Mammal “Ages” (NALMAs) include: Lancian ( Sphenodiscus through “Triceratops” zones); “Edmontonian” (a name not yet faunally defined; Didymoceras cheyennense through Baculites clinolobatus zones); Judithian (the smooth, late form of Baculites sp. through Exiteloceras jenneyi zones); and Aquilan ( Scaphites hippocrepis through Baculites asperiformis zones). Correlations emphasize use of provincial biostratigraphic terminology designed specifically for use in the Western Interior. On the basis of temporal constraints suggested herein, known mammalian fossils from the upper Fruitland and/or lower Kirtland Formations of New Mexico probably are of “Edmontonian,” not Judithian age. Although considerable latitudinally based taxonomic diversification of Judithian mammals is now recognizable across the Western Interior, comparative data are inadequate to defend a similar statement for the remaining Late Cretaceous NALMAs. Quantitative evaluation of geographic patterns of shoreline change suggests occurrence of a general, regional regression of the sea during the entire geologic interval represented in the study. We favor explanation by a slow sea-level depression resulting from topographic evolution of the world’s mid-oceanic system of ridges and volcanic plateaus. Local and subregional asynchronous episodes of shoreline transgressions, stillstands, and regressions are superimposed upon the general regressive trend, and probably represent influences of local tectonism, not eustatic changes in sea level. Strandline evolution of the epeiric sea during the last 20 million years of the Cretaceous in the North American Western Interior is inconsistent with: (1) existence of geologically brief (1 to 10 m.y.) global fluctuations in sea level; and (2) the concept that the late Campanian was represented by an unusually high global sea level.

With the discovery and description of the Bug Creek faunas in 1965, it became necessary to reexamine the sequence of mammal ages (Lancian-Puercan) spanning the Cretaceous-Tertiary boundary. Bug Creek faunal assemblages have been viewed as being in part coeval with the Lancian assemblages or slightly younger. Because of the lack of Lancian sites above the Bug Creek–type sites and the discontinuous nature of the sediments preserving the latter type faunas, it appears that the Bug Creek faunas postdate Lancian faunas. Although the Bug Creek faunas cannot be well constrained biostratigraphically, the appearance of a number of new taxa and the continuation of these or closely related taxa into the Puercan is strong evidence for a biochronologic sequence of Lancian–Bug Creek–Puercan. When the faunal contents of Lancian, Bug Creek, and earliest Puercan (Pu 1 ) sites are scrutinized, it is clear that important faunal introductions occur with the commencement of Bug Creek assemblages. In contrast, between Bug Creek and earliest Puercan (Pu 1 ) assemblages there are very few major faunal introductions and thus more faunal continuity. Therefore, the definition and concept of the Puercan Land Mammal Age is modified, and the Bug Creek faunas are formally defined and characterized as the Protungulatum/Peradectes Interval-Zone (Pu 0 ) of the Puercan Land Mammal Age. The Protungulatum/Peradectes Interval-Zone (Pu 0 ) postdates the (latest Cretaceous) Lancian Land Mammal Age and commences the (latest Cretaceous?–early Paleocene) Puercan Land Mammal Age. This interval-zone is defined as including faunas that occur during the time between the first appearance of the arctocyonid ungulate Protungulatum and the first appearance of the didelphid marsupial Peradectes. Certain biochronological criteria (first appearances and “cladochronology”) within this interval-zone strengthen the view that the original sequence (from oldest to youngest) of Bug Creek Anthills, Bug Creek West, and Harbicht Hill is correct. Therefore, three informal biochrons are recognized, the Protungulatum/Mimatuta (bk 1 ), Mimatuta/Oxyprimus (bk 2 ), and Oxyprimus/Peradectes (bk 3 ) biochrons. Dinosaurs and Lancian mammals are found at all Pu 0 localities. Although the possibility of reworking cannot be completely dismissed, the abundance of Lancian mammals and the nature of dinosaur material at certain Pu 0 sites suggest some temporal overlap. Similarly, the stratigraphic placement of palynological change and an iridium anomaly relative to Pu 0 faunas remains equivocal.

Although Paleocene mammals have been known from western Canada for nearly 70 years, only during the last 15 years have concerted efforts been made to discover, develop, and describe collections documenting their evolution. Whereas much remains to be done, a faunal succession has been reconstructed, based on collections from 41 mammalian local faunas in Alberta and Saskatchewan, ranging from early Puercan to late Tiffanian age (the first 8 to 10 m.y. of the Tertiary). Latest Cretaceous mammalian local faunas in Saskatchewan show that evolution of progressive “Paleocene aspect” mammals, including condylarths, had begun during the North American Cretaceous as evidenced by the occurrence of fossils in stratigraphie settings free from the complexities that make uncertain the age of faunally similar assemblages in Montana. The Ravenscrag Formation, southwestern Saskatchewan, yields the oldest (early Puercan) Tertiary mammals known from Canada (Rav W-l: MHBT Quarry, Pine Cree Park and Croc Pot localities), including the first discovery of the ancestral primate Purgatorius outside of eastern Montana, and the oldest unarguable carnivoran. Torrejonian mammals are known from the Coalspur Formation (Rocky Mountain Foothills; Diss locality) south of Edson, Alberta. Cochrane 2 (Porcupine Hills Formation, west of Calgary) has yielded unexpectedly diverse earliest Tiffanian mammals, and early Tiffanian mammals have recently been discovered in the Paskapoo Formation near Drumheller (Hand Hills West, lower level), and east of Innisfail, Alberta (Aaron’s Locality). Middle Tiffanian localities occur in the Paskapoo Formation near Red Deer (DW 1 to 3, Mel’s Place, Joffre Bridge Road Cut, Mammal Site No. 1, Erickson’s Landing), and in the Hand Hills (Hand Hills West, upper level); the Police Point local fauna (Ravenscrag Formation), Alberta, appears to be late middle or early late Tiffanian in age. Late Tiffanian mammals at Roche Percée, Saskatchewan (Ravenscrag Formation), and Canyon Ski Lodge, Crestomere School, and Swan Hills, Alberta (Paskapoo Formation) conclude the Paleocene mammalian record known from Canada. Species lists for each locality are presented. The early and middle Tiffanian mammalian record from Alberta fails to show a decline in species numbers seen at several American localities representing this interval: instead of the global cooling sometimes hypothesized to account for this decline, it now seems to be a result of biological or sedimentological events acting on a local scale, sampling error, or some combination of these factors.

South America, Central America, the southeastern United States, Arctic Canada, Europe, Asia, and Africa all have been suggested as possible or probable biogeographic sources for taxa that appeared in the Western Interior of North America during the late Paleocene and early Eocene. Recent compilations of the geographic and temporal distributions of Paleocene and Eocene mammals and new data, derived primarily from recent collections from early Tiffanian (late Paleocene) quarries in the Crazy Mountains Basin of south-central Montana, permit tests of these hypotheses, particularly those involving a southern New World origin. Significant first appearances of mammalian higher taxa in the Western Interior occur in the earliest Tiffanian, late Tiffanian, earliest Oarkforkian, and earliest Wasatchian. Those that appear in the earliest Tiffanian probably were derived from late Torrejonian forms in the same region. It appears, therefore, that there was not a pronounced geographic shift in North American mammalian faunas across the Torrejonian-Tiffanian boundary as suggested in some southern New World origin hypotheses. It has been suggested that Palaeanodonta, Dinocerata, and Notoungulata (represented by Arctostylopidae), which appear in the late Tiffanian in the Western Interior, originated in South America, but the evidence is inconclusive and highly controversial. New higher taxa that appear in the Western Interior at the beginning of the Clarkforkian, particularly Rodentia and Tillodontia, probably originated in Asia and dispersed across Beringia. Most of the suprageneric taxa that first appear at the beginning of the Wasatchian in the Western Interior (Perissodactyla, Artiodactyla, Adapidae, Omomyidae, and Hyaenodontidae) also probably appeared in Asia and Europe at essentially the same time; there is no evidence for heterochrony. Recent paleontological discoveries and paleogeographic evidence suggest that the ultimate origins of some or all of these taxa lay in either Africa or the Indian subcontinent. The latter biogeographic source has not been seriously considered previously.

The Bighorn Basin has produced the richest and most diverse early Eocene mammalian faunas in the world and is the principal source of our knowledge of skeletal anatomy in these mammals. Until recently, most of our information on postcranial anatomy in early Eocene mammals came from the works of Matthew and his contemporaries. Considerable new evidence has been unearthed in the last 25 years, but very little of it has yet been described or even reported in the literature. Since 1979, a USGS–Johns Hopkins project working in the Wasatchian part of the Willwood Formation has collected more than 150 skeletal associations (representing more than 25 genera in 20 families), varying from several bones to virtually complete, articulated skeletons. Among these are important new specimens—some of them the first or the most nearly complete skeletons known—of Palaeanodon, Alocodontulum, Microsyops, Phenacolemur, Cantius, Chriacus, Anacodon, Oxyaena, Prototomus, Didymictis, Vulpavus, Miacis, Phenacodus, Hyracotherium, Homogalax, Wasatchia, and Diacodexis. Comparison of characters such as limb proportions, long bone and joint structure, and ungual shape with those in extant forms whose behavior is documented enables inferences of locomotor capabilities in extinct mammals. A wide range of terrestrial adaptations is apparent in Willwood mammals, which include fossorial palaeanodonts, a large digger/rooter ( Ectoganus ), ambulatory ( Oxyaena, Didymictis) or graviportal forms ( Coryphodon ), incipient cursors ( Phenacodus, Pachyaena ), more specialized cursors ( Hyracotherium ), small cursorial/saltatorial types ( Diacodexis, Wasatchia), and small saltatorial mammals (leptictid insectivores). Arboreal locomotion was of at least two types: quadrupedal climbing and leaping (adapid primates), and scansorial claw-climbing (small arctocyonids and miacid carnivorans) that involved extreme tarsal mobility. Some postcranial modifications are strikingly similar to those in extant relatives of these Eocene mammals, suggesting that modification of skeletal form occurred well in advance of dental evolution.

Willwood Formation paleosols are ranked on a scale of 0 to 5 on the basis of their relative maturity (= relative time required to form). In the lateral dimension, the least mature soils were developed more proximal to ancient channel belts, whereas the more mature paleosols formed in areas more distant to channel belts. Quantitative study shows that both mammalian taxonomic composition and taphonomic completeness vary systematically with the maturity of these paleosols. Species-level differences in taxonomic composition are identified for pedofacies sequences located at the 442-m and 546-m levels of the Willwood Formation. At 442 m, Cantius frugivorus and Hyopsodus sp., cf. H. minor account for practically all of the adapiform primate and hyopsodontid condylarth faunas in stage 3 to 4 paleosols (which are distally located with respect to the ancient channel belt). Laterally adjacent and stratigraphically equivalent stage 1 to 2 paleosols (proximally located with respect to the ancient channel belt), are instead dominated by Cantius sp. nov. and Hyopsodus sp., cf. H. miticulus. Intermediate proportions of these taxa occur at localities in paleosols of intermediate maturity (stage 2 to 3 paleosols) at the 442-m level. At 546 m, the otherwise relatively rare species Hyopsodus powellianus makes up nearly 50 percent of the hyopsodontid fauna at some localities developed in stage 1 paleosols; elsewhere in this pedofacies the species Hyopsodus minor and H. lysitensis make up the overwhelming majority of the Hyopsodus. Also at 546 m, the adapiform primates Cantius abditus and “Copelemur”feretutus exhibit reversals in relative abundance from proximal to distal localities across the pedofacies; Cantius is more abundant in proximal localities and “Copelemur” is dominant in distal localities. Ordinal-level differences in taxonomic composition were detected at localities in two distinct pedofacies lying at or slightly above Biohorizon C (= “Graybullian-Lysitean” boundary). There, Condylarthra and Artiodactyla are more common in immature (stages 1 to 2) than mature (stage 4) paleosols, whereas the reverse is true for Primates, Carnivora, Rodentia, and Perissodactyla. Lateral controls on completeness of skeletal elements, as related to lateral variation in sedimentation rate, are also evident. Proportions of less complete skeletal elements are considerably higher at localities developed in mature paleosols, where sedimentation rates were low. These findings underscore the inherent relatedness of geographic distribution of taxa, taphonomy, and sedimentology and suggest that intrabasinal differences in microhabitat had a significant effect on the local taxonomic composition of the Willwood mammalian fauna.

Changes in sample size may confound interpretation of faunal change in the fossil record. The record of early Eocene mammals from the Wasatchian Land-Mammal Age in the Clark’s Fork Basin, Wyoming, shows a high correlation between the square root of sample size and species richness. This correlation suggests that Schankler’s (1980) Biohorizon A, a faunal turnover composed mainly of disappearances, is largely an artifact of sampling fluctuation. Within the interval of Biohorizon A, sample size drops from record high to record low values. Monte Carlo simulation of the drop in sample size across Biohorizon A demonstrates the role of sampling variation in producing artifacts of faunal change. The distribution of missing species resulting from the simulations provides a reliable estimate of the number of species that are likely to be missing at a specified sample size, even if they were present in the original population. Results of the simulations indicate that most of the 16 disappearances observed in the 200-m interval above Biohorizon A can be explained by low sample size alone. For each species that disappeared in the actual record, the frequency of absence in the simulations is a basis for comparing the likelihoods of two hypotheses: (1) that the species was present but not represented by fossils, and (2) that the species was absent Evaluation of the likelihood ratio for these hypotheses indicates that Arctodontomys wilsoni-A. nuptus, Phenacodus vortmani, and Homogalax n. sp.-H. semihians are the lineages most likely to have disappeared over the interval of low sample size. Reconsideration of the biostratigraphic correlation between the central Bighorn Basin and the Clark’s Fork Basin based on the first appearances of Homogalax protapirinus and Tetonius matthewi/steini supports a higher stratigraphic position for the interval corresponding to Biohorizon A in the Gark’s Fork Basin than its original placement.

The Buck Spring Quarries, located in the southern part of the type area of the Lost Cabin Member of the Wind River Formation, Wind River Basin, Wyoming, provide one of the richest assemblages of fossil vertebrates known from the latest Wasatchian Land-Mammal Age (ca. 50.5 Ma, Lostcabinian, early Eocene) of North America. More than 100 species of mammals, reptiles, birds, amphibians, and fishes are known. The quarries uniquely preserve associated skeletal remains, and complete skulls and dentitions of a large percentage of the vertebrates. The fossils come from a 2-m-thick sequence, which is composed primarily of mudstones, bioturbated limestone lenses, and laminated limestone/mudstone couplets. These sediments were deposited in a well-drained swamp or ponded area between 250 and 600 m away from a low-sinuosity stream. The fossils are especially common in limestones, where they accumulated as a result of (1) natural death (articulated specimens), (2) predator activity (coprolites and kill sites), and (3) very limited hydraulic transport of smaller bones. Rarefaction estimates from surface and quarry collections suggest similar patterns of species richness, which are among the highest known for the Paleogene and compare favorably with penecontemporaneous Lostcabinian assemblages and slightly younger Gardnerbuttean (early Bridgerian) ones. The mammalian assemblage is dominated by small species of mammals, and has a body-size distribution and species diversity similar to modern tropical communities. The abundance of arboreal mammals indicates that a multistoried woodland habitat was in close proximity to the quarry area.

The first Eocene vertebrate assemblage known from the Great Basin, the Elderberry Canyon Local Fauna, occurs in rocks referred to the Sheep Pass Formation near Ely, Nevada. Approximately 40 taxa are now known, including small anuran amphibians, small reptiles, birds, and mammals. The mammalian component consists of: the insectivorans Apatemys bellus, Pantolestes longicaudus, a tiny apternodont, at least one nyctitheriid, and at least four other taxa representing dormaalid and/or erinaceid erinaceomorphs; an epoicotheriid palaeanodont, cf. Tetrapassalus mckennai; the primates Notharctus tenebrosus, Trogolemur myodes, and two species of uintasoricines; the rodents Reithroparamys delicatissimus, R. cf. R. huerfanensis, Sciuravus sp., Microparamys sp., Pauromys sp., Mattimys sp., and two new genera; the hyaenodont Sinopa minor; two viverravid carnivores including Viverravus; the condylarth Hyopsodus paulus; the perissodactyls Hyrachyus modestus, Hyrachyus affinis, Helaletes nanus, Isectolophus latidens, and a new genus and species; and the artiodactyl Antiacodon pygmaeus. Greatest faunal similarity is with the Black’s Fork Member (or lower), Bridger Formation, and other early Bridgerian localities such as Powder Wash in the Douglas Creek Member of the Green River Formation in northeastern Utah. The age of the Elderberry Canyon Local Fauna can confidently be called early Bridgerian. The Elderberry Canyon Fauna is preserved in carbonate rocks believed to have been deposited in a shallow, warm, heavily vegetated, permanent, hardwater lake. The mammals lived on marshy wetland terrain adjacent to the lake, although some faunal elements may have been transported in from more distant habitats.

Two new genera and species of plagiomenids (Mammalia, ?Dermoptera, Plagiomenidae) are described from the North American Uintan (middle Eocene) and Chadronian (early Oligocene). A third genus and species, Ekgmowechashala philotau, from the early and middle Arikareean (late Oligocene) of the northern United States, is removed from the primate family Omomyidae and placed in the Plagiomenidae. All three newly recognized plagiomenids are placed in the Ekgmowechashalinae, sister subfamily to the subfamily Plagiomeninae (new rank). Ekgmowechashaline plagiomenids are somewhat primate-like, as is the plagiomenine genus Worlandia, but the Plagiomenidae are usually considered to be allied to the living colugos of southeast Asia, order Dermoptera. Analysis of that relationship is placed outside the scope of this paper. Tarka stylifera, the earliest known ekgmowechashaline, occurs in the type section of the Tepee Trail Formation, early Uintan (Shoshonian: late medial Eocene) of northwestern Wyoming. This locality falls in paleomagnetic Chron C20R, interpreted to be close to 47.5 Ma in age. A second, more primitive but later-occurring ekgmowechashaline genus and species, Tarkadectes montanensis, is from a nominally early Oligocene level (Chadronian) in the Kishenehn Formation of northern Montana. Ekgmowechashala is known from lower dentitions from the early Arikareean Sharps Formation of South Dakota and probably from an upper dentition reported from middle Arikareean rocks in the John Day Formation of Oregon. Ekgmowechashala is placed with the other two genera because of lower cheek-tooth morphology, but it lacks the enlarged incisor of Tarka. Ekgmowechashalines are hypothesized here to be primarily frugivores, folivores, and nectar- and exudate-feeders. Until now, known undoubted plagiomenids were restricted to the Paleocene and early Eocene (Wasatchian). The newly recognized post-Wasatchian occurrences are all in the northern part of the United States and are in keeping with previously known plagiomenid geographic distribution, which ranged from northern Wyoming to the Canadian arctic and possibly beyond.

Abstract Although species are the basic units of many paleontological and evolutionary studies, the term “species” applied to the fossil record does not convey the same concept to all workers. Simpson’s “evolutionary species” incorporated time into the species concept, but considered each non-branching lineage as a separate species; longer lineages with more continuous fossil records may require subdivision into successional species. One’s perception of paleontological species affects, and is affected by, evolutionary philosophy and models of how new species form and evolve. For example, if species actually arise abruptly and persist for much longer periods essentially unchanged (punctuated equilibria), discrimination of paleontological species should be a relatively simple matter. Alternatively, if there is continuous change within and between successive species (gradualism), species boundaries would be nebulous, and would have to be imposed arbitrarily. We summarize our study of omomyid primates and cite other supportive evidence which suggests that, where the record is sufficiently dense, gradual evolution (requiring arbitrary boundaries) is common between species and even genera.