The Upper Cretaceous (Campanian) Belly River Group (BRG) of southern Alberta has a complex internal stratigraphic architecture derived from differential geometries of its component formations that resulted from regionalized tectonic influences and shifting source areas. A full understanding of BRG architecture has been compromised heretofore by a limited understanding of subsurface data in southwestern- and southeastern-most Alberta. In this study outcrop exposures throughout southern Alberta are tied to reference well logs and subsurface cross-sections allowing a more precise understanding of BRG architecture and how it relates to well-known vertebrate fossil producing areas. Modifications to an existing stratigraphic model of the BRG show that the Oldman and the Dinosaur Park formations have reciprocal north-to-south wedge-shaped geometries and a diachronous contact that become prominently expressed south of Twp 12. The updated model also demonstrates that the Oldman Formation thickens stratigraphically up-section to the south, and that the Foremost-Oldman contact is, essentially, a datum across much of southern Alberta. Identification of the Oldman Formation in the subsurface remains based on its relatively high gamma-ray response in mudstone successions, but it is also recognized that many of its sandstones exhibit relatively low gamma-ray responses like those in underlying and overlying formations. Nomenclature and subdivisions of the Oldman Formation are revised to accommodate this updated understanding, and modifications are also made to the definition of the Judith River-Belly River discontinuity, a newly recognized surface that marks the onset of accommodation and eustatic rise in sea-level in the northern Western Interior Basin at ~76.3 Ma.

Terrestrial strata of the Judith River−Belly River wedge, widely exposed in the plains of north-central Montana, southern Alberta, and southwestern Saskatchewan, were pivotal in early stratigraphic investigations of the Western Interior of North America and are renowned to this day for their spectacular preservation of Late Cretaceous fossils, most notably dinosaurs. Correlation of the Judith River Formation in Montana with the Foremost, Oldman, and Dinosaur Park Formations (= Belly River Group) in Canada has been challenging for a variety of reasons, including lithostratigraphic complexities, legacy bentonite ages of limited comparability, and distinctly different stratigraphic models on opposite sides of the international border. An updated model calibrated with U-Pb zircon ages provides an improved framework for stratigraphic analysis. New geochronology indicates that the Oldman−Dinosaur Park discontinuity in Dinosaur Provincial Park correlates in age with the mid-Judith discontinuity in the Judith River Formation in Montana, which is interpreted as an expansion surface linked to a major pulse of accommodation and onset of the Bearpaw transgression at ca. 76.3 Ma. The regionally expressed shift in alluvial facies marking the mid-Judith discontinuity can be traced in well logs from Montana to southern Canada, where it loses distinction and transitions to a subsurface signature typical of the Oldman−Dinosaur Park discontinuity, which in turn can be traced north to Dinosaur Provincial Park and beyond. Across this expanse, both discontinuities parallel the Eagle/Milk River shoulder at approximately the same stratigraphic height, confirming their chronostratigraphic significance. These findings have clear implications for regional correlation and the evolution of alluvial depositional systems in a foreland basin setting, and they afford an opportunity to evaluate existing interpretations and advance understanding of the stratigraphy and paleontology of the Judith River−Belly River wedge. The term “Judith River−Belly River discontinuity” should be used henceforth to refer to the chronostratigraphically significant stratal discontinuity that subdivides the Judith River−Belly River wedge throughout the plains of north-central Montana, southern Alberta, and southwestern Saskatchewan.

The 100 m thick stratigraphic section exposed at Dinosaur Provincial Park (DPP; southern Alberta) contains bentonites that have been used for more than 30 years to date DPP’s rocks and fossils using the K–Ar decay scheme. Limited reproducibility among different vintages of K–Ar and 40 Ar/ 39 Ar ages inhibited the development of a high-resolution chronostratigraphy. Here, we employ and further test a recently completed U–Pb geochronology and associated age-stratigraphy model to update temporal constraints on the Park’s bentonites, formational contacts, and other markers. In turn, we document rock accumulation rates and calibrate ages and durations of informal megaherbivore dinosaur assemblage zones and other biozones. Weighted mean 206 Pb/ 238 U ages from five bentonites range from 76.718 ± 0.020 to 74.289 ± 0.014 Ma (2σ internal uncertainties) through an interval of 88.75 m, indicating a duration of ∼2.43 Myr and an overall rock accumulation rate of 3.65 ± 0.04 cm/ka. An increase in rate above the Oldman–Dinosaur Park formational contact conforms to a regionally expressed pattern of increased accommodation at ∼76.3 Ma across Alberta and Montana. Palynological biozone data suggest a condensed section/hiatus in the uppermost portion of the Oldman Formation. Dinosaur assemblage zones exhibit durations of ∼700–600 kyr and are significantly shorter than those in the overlying Horseshoe Canyon Formation. A decreased rate in dinosaur assemblage turnovers in the last eight million years of the Mesozoic in western Canada may be explained by withdrawal of the Western Interior Seaway and the expansion of ecologically homogenous lowlands in its wake.

The spectacular fossil fauna and flora preserved in the Upper Cretaceous terrestrial strata of North America’s Western Interior Basin record an exceptional peak in the diversification of fossil vertebrates in the Campanian, which has been termed the ‘zenith of dinosaur diversity’. The wide latitudinal distribution of rocks and fossils that represent this episode, spanning from northern Mexico to the northern slopes of Alaska, provides a unique opportunity to gain insights into dinosaur paleoecology and to address outstanding questions regarding faunal provinciality in connection to paleogeography and climate. Whereas reliable basin-wide correlations are fundamental to investigations of this sort, three decades of radioisotope geochronology of various vintages and limited compatibility has complicated correlation of distant fossil-bearing successions and given rise to contradictory paleobiogeographic and evolutionary hypotheses. Here we present new U–Pb geochronology by the CA-ID-TIMS method for 16 stratigraphically well constrained bentonite beds, ranging in age from 82.419 ± 0.074 Ma to 73.496 ± 0.039 Ma (2σ internal uncertainties), and the resulting Bayesian age models for six key fossil-bearing formations over a 1600 km latitudinal distance from northwest New Mexico, USA to southern Alberta, Canada. Our high-resolution chronostratigraphic framework for the upper Campanian of the Western Interior Basin reveals that despite their contrasting depositional settings and basin evolution histories, significant age overlap exists between the main fossil-bearing intervals of the Kaiparowits Formation (southern Utah), Judith River Formation (central Montana), Two Medicine Formation (western Montana) and Dinosaur Park Formation (southern Alberta). Pending more extensive paleontologic collecting that would allow more rigorous faunal analyses, our results support a first-order connection between paleoecologic and fossil diversities and help overcome the chronostratigraphic ambiguities that have impeded the testing of proposed models of latitudinal provinciality of dinosaur taxa during the Campanian.

The Middle-Late Jurassic transition period is a critical period for the evolution of terrestrial vertebrates, but the global fossil record from this time is relatively poor. The Shishugou Fauna of this period has recently produced significant fossil remains of dinosaurs and other vertebrate groups, some representing the earliest known members of several dinosaurian groups and other vertebrate groups and some representing the best-known specimens of their group. These discoveries are significant for our understanding of the origin and evolution of several vertebrate lineages. Radiometric dating indicates that the fauna is aged approximately 159-164 Ma. Comparisons with other similarly-aged terrestrial faunas such as Shaximiao and Yanliao show both taxonomic similarities and differences between these faunas and indicate that the Junggar deposits might have preserved the most complete vertebrate fossil record for a Middle-Late Jurassic Laurasian terrestrial fauna.

Mapping of fossil sites represents an important aspect of palaeontology, because the data collected are required for interpreting the taphonomic and depositional history of the site, as well as the palaeoecology and behavior of the organisms. Methods for mapping and documenting certain vertebrate fossil sites, such as trackways, have drastically changed in recent years, with the integrated technologies of photogrammetry, laser scanning, and geographic information systems becoming standard practice, and providing digital, three-dimensional, and georeferenced data for analyses. Contrasting this technological revolution, the methods for mapping vertebrate bone accumulations, such as bonebeds, have changed little in recent decades, and are largely limited to two dimensions, are non-georeferenced, and produce static maps. Here, we present a novel test case in the mapping of two ceratopsid (Dinosauria: Ornithischia) monodominant bonebeds (mass death assemblages) that are documented digitally, fully georeferenced, and in three dimensions, using a combination of high-resolution (at centimetre-scale) global positioning system, photogrammetry, and geographic information systems. Importantly, accompanying spatial data (i.e., size and orientation) are collected in the field in the traditional manner and directly compared with values calculated from the digital map. Parameters describing bone length and orientation exported from the digital map are largely reflective of measured field data, with both size and orientation distributions being statistically indistinguishable, but with disproportionate error for elements smaller than 10 cm. Protocols and methods tested here will hopefully add to the discussion about the future of fossil bonebed mapping, specifically incorporating digital, three-dimensional, and fully georeferenced data into a powerful analytical tool.

Lower Eocene (Wasatchian-aged) sediments of the Margaret Formation on Ellesmere Island in Canada's High Arctic preserve evidence of a rainforest inhabited by alligators, turtles, and a diverse mammalian fauna. The mammalian fossils are fragmentary and often poorly preserved. Here, we offer an alternative method for their identification. Among the best preserved and extensive of the Eocene Arctic forests is the Strathcona Fiord Fossil Forest, which contains permineralized in situ tree stumps protruding from a prominent coal seam, but a paucity of vertebrate fossils. In 2010 and 2018, we recovered mammalian tooth fragments at the fossil forest, but they are so incomplete as to be undiagnostic by using their external morphology. We used a combination of light microscopy and SEM analysis to study the enamel microstructure of two tooth fragments from the fossil forest-NUFV2092B and 2092E. The results of our analysis indicate that NUFV2092B and 2092E have Coryphodon-enamel, which is characterized by vertical bodies that manifest as bands of nested chevrons or treelike structures visible in the tangential section under light microscopy. This enamel type is not found in other mammals known from the Arctic. Additionally, when studied under SEM, the enamel of NUFV2092B and 2092E has rounded prisms that open to one side and are surrounded by interprismatic matrix that is nearly parallel to the prisms, which also occurs in Coryphodon enamel, based on prior studies. The tooth fragments reported here, along with some poorly preserved bone fragments, thus far are the only documented vertebrate fossils from the Strathcona Fiord Fossil Forest. However, fossils of Coryphodon occur elsewhere in the Margaret Formation, so its presence at the fossil forest is not surprising. What is novel in our study is the way in which we identified the fossils using their enamel microstructure.

The non-marine Horseshoe Canyon Formation (HCFm, southern Alberta) yields taxonomically diverse, late Campanian to middle Maastrichtian dinosaur assemblages that play a central role in documenting dinosaur evolution, paleoecology, and paleobiogeography leading up to the end-Cretaceous extinction. Here, we present high-precision U–Pb CA–ID–TIMS ages and the first calibrated chronostratigraphy for the HCFm using zircon grains from (1) four HCFm bentonites distributed through 129 m of section, (2) one bentonite from the underlying Bearpaw Formation, and (3) a bentonite from the overlying Battle Formation that we dated previously. In its type area, the HCFm ranges in age from 73.1–68.0 Ma. Significant paleoenvironmental and climatic changes are recorded in the formation, including (1) a transition from a warm-and-wet deltaic setting to a cooler, seasonally wet-dry coastal plain at 71.5 Ma, (2) maximum transgression of the Drumheller Marine Tongue at 70.896 ± 0.048 Ma, and (3) transition to a warm-wet alluvial plain at 69.6 Ma. The HCFm’s three mega-herbivore dinosaur assemblage zones track these changes and are calibrated as follows: Edmontosaurus regalis – Pachyrhinosaurus canadensis zone, 73.1–71.5 Ma; Hypacrosaurus altispinus – Saurolophus osborni zone, 71.5–69.6 Ma; and Eotriceratops xerinsularis zone, 69.6–68.2 Ma. The Albertosaurus Bonebed — a monodominant assemblage of tyrannosaurids in the Tolman Member — is assessed an age of 70.1 Ma. The unusual triceratopsin, Eotriceratops xerinsularis, from the Carbon Member, is assessed an age of 68.8 Ma. This chronostratigraphy is useful for refining correlations with dinosaur-bearing upper Campanian–middle Maastrichtian units in Alberta and elsewhere in North America.

The Battle Formation (BFm) is a widespread Upper Cretaceous marker horizon in western Canada that records a time of low sediment input and marks the boundary between the Edmontonian and Lancian land-vertebrate ages. Here, we present the first high-precision U–Pb CA–ID–TIMS age of 66.936 ± 0.047/0.060/0.140 Ma for the Battle bentonite, an altered vitric ash in the upper portion of the BFm at Knudsen’s Farm in the Red Deer River valley of Alberta. This age supersedes those previously reported, confirms that rates of sediment accumulation for the formation were very low (∼1.40 cm/ka), and allows us to interpolate an age range of ∼66.88–67.20 Ma for the BFm. Our data also provide a maximum age of ∼66.88 Ma for the base of the overlying Scollard Formation, a dinosaur-rich unit. We combine our age data with calibrated magneto- and palynostratigraphic data to assess chronostratigraphic correlations among the Scollard Formation of Alberta, the Frenchman Formation of Saskatchewan, and the Hell Creek Formation in eastern Montana. Whereas the combined data support previous interpretations that equate the age ranges of the Scollard Formation, Frenchman Formation, and the upper one third of the Hell Creek Formation in eastern Montana, they also indicate that all of the lower one third (L3) and part of the middle one third (M3) of the Hell Creek Formation in Montana are chronostratigraphically equivalent to all or part of the sub-BFm unconformity and the BFm in Alberta. Accordingly, a minimum age of ∼67.20 Ma is assessed for the base of the Hell Creek Formation in its type area.

The famous "Dragon's Tomb" Saurolophus bonebed at Altan Uul II preserves multiple complete skeletons with skin impressions, and is exceptional among Nemegt Formation dinosaur localities as the only mass accumulation of articulated individuals. Here, we present results from an on-going paleontological project focused on the Baruungoyot-Nemegt succession in order to document this unique locality. Sedimentological data allow us to confirm that the fossils and host sediments were deposited in the thalweg of a meandering paleochannel that experienced a variety of high and lower energy flow conditions. A variety of taphonomic data were collected from exposed fossils, including those displaced by poaching. Three-dimensional maps of the host sediments and fossils were generated, and 13 m2 of exposed, in-situ bones were mapped. Previously undocumented portions of at least 13 additional Saurolophus individuals and a partial subadult Tarbosaurus skeleton were identified. The Dragon's Tomb is a monodominant bonebed with at least three size classes (juvenile, subadult, and large adult) of Saurolophus preserved. The bonebed has a minimum size of ~ 2000 m2, suggesting that over 100 Saurolophus carcasses may have contributed to the thanatocoenose. Although the data provide evidence for a catastrophic mass death of a social group of Saurolophus and provide the first evidence of gregariousness in this taxon, the cause of death cannot be determined.

Portions of the dinosaur-rich Upper Cretaceous Baruungoyot-Nemegt succession are exposed in four geographic areas of the Nemegt Basin, Mongolia: Nemegt, Altan Uul, Bügiin Tsav, and Hermiin Tsav. Lithostratigraphic correlations of these areas employ marker beds and non-random stratigraphic patterns of lithology, grain-size variation, and inferred paleoenvironments. Correlations suggest a minimum thickness of 350m for the succession. Exposures at Nemegt, Altan Uul, and Bügiin Tsav form a 100km east-west transect along the northern margin of the basin, exposing 265m of section. From east to west the transect exposes higher portions of the succession, extending from the uppermost Baruungoyot Formation and encompassing an almost complete section of the Nemegt Formation, which has a minimum thickness of 235m and can be consistently divided into three informal members: lower, middle, and upper. The lower Nemegt is well exposed at Nemegt and Altan Uul and is dominated by fluvial deposits. The middle and upper Nemegt are well exposed at Altan Uul 2 and Bügiin Tsav and consist of alluvial plain, paludal, lacustrine, and fluvial deposits. As confirmed by earlier studies, the Baruungoyot and Nemegt formations exhibit an interfingering transition at Nemegt. The Hermiin Tsav section is a geographic outlier, tens of kilometers west-southwest of the northern transect. It exposes about 150m of section that overlaps with, but is mostly lower, stratigraphically, than the northern transect. The uppermost Hermiin Tsav section correlates with the lowermost Nemegt section, and also exhibits an interfingering contact between the Baruungoyot and Nemegt formations. The remaining section at Hermiin Tsav comprises sediments of the Baruungoyot Formation (135m) that were deposited in alluvial fan, fluvial, paludal, lacustrine, and eolian environments, and can be divided into two informal members designated as lower and upper. The distribution of facies through the composite section records changes in paleoenvironments during the Late Cretaceous that may provide insights into the origins of the dinosaurian biostratigraphy.

Dinosaur Provincial Park (DPP; southern Alberta) yields a well-preserved and uniquely-diverse assemblage of late Campanian dinosaur fossils. Because the assemblage plays a central role in documenting Late Cretaceous peak-dinosaur-diversity, a chronostratigraphic framework for DPP’s fossils has evolutionary and paleoecological significance that extends beyond DPP’s boundaries. Here we employ high-precision CA-TIMS U-Pb zircon geochronology to constrain the ages of five bentonites that are semi-evenly spaced through the exposed bedrock section. Weighted mean 206Pb/238U dates for the bentonites range from 76.7 Ma to 74.3 Ma (with 2σ internal errors of ≤30 kyr), indicating that the section is younger than previously understood and that it encompasses a longer time span (2.4 Myr). Combined with biostratigraphic data, these results confirm a previous hypothesis that the Oldman–Dinosaur Park formation contact becomes younger toward southeastern Alberta. A significantly lower sediment accumulation rate near the Oldman-Dinosaur Park formation contact suggests the presence of a ~230 kyr hiatus (or stacked hiatuses) near the contact. Coincident first occurrences of five palynomorph taxa—that otherwise show stratigraphically staggered first occurrences in southeastern Alberta—support an interpretation that the hiatus is limited to the uppermost few meters of the Oldman Formation. Previously amassed sedimentological and stratigraphic data suggest that the hiatus may be a consequence of overall lower rates of sediment accumulation and interrupted sedimentation along the northern edge of the Oldman sedimentary lobe. Rates of sediment accumulation stay largely constant through the Dinosaur Park Formation, but may decline upwards into the Bearpaw Formation. Applying a currently accepted ammonite biochronology to DPP’s age-calibrated strata correlates the bottom of the section with the middle of the Baculites scotti Zone, and the top of the section with the approximate base of the Baculites compressus Zone. Four previously recognized dinosaur-assemblage zones at DPP are also calibrated and show durations of approximately 600–700 kyr, similar to those of ammonite biozones.

A monodominant bonebed of Avimimus from the Nemegt Formation of Mongolia is the first oviraptorosaur bonebed described and the only recorded maniraptoran bonebed from the Late Cretaceous. Cranial elements recovered from the bonebed provide insights on the anatomy of the facial region, which was formerly unknown in Avimimus. Both adult and subadult material was recovered from the bonebed, but small juveniles are underrepresented. The taphonomic and sedimentological evidence suggests that the Avimimus bonebed represents a perimortem gregarious assemblage. The near absence of juveniles in the bonebed may be evidence of a transient age-segregated herd or 'flock', but the behaviour responsible for this assemblage is unclear. Regardless, the Avimimus bonebed is the first evidence of gregarious behaviour in oviraptorosaurs, and highlights a potential trend of increasing gregariousness in dinosaurs towards the end of the Mesozoic.

Sphenacodon Marsh, known from the Lower Permian of New Mexico, Utah, and Arizona, is one of the best represented pelycosaurs in fossil collections, yet its skull has never been completely described. Virtually complete, disarticulated cranial remains of Sphenarodon ferocior from the Lower Permian Cutler and Abo Formations of north-central New Mexico are described in detail for the first time. These descriptions provide the basis for the most detailed comparisons to date within the genus Sphenacodon and the subfamily Sphenacodontinae. No morphological differences exist between the skulls of Sphenacodon and Ctenospondylus, and only minor morphological differences exist between Sphenacodon and Ctenospondylus on the one hand and Dimetrodon on the other. On the basis of cranial morphology alone, Sphenacodon and Ctenospondylus appear to be more closely related than the current taxonomy reflects. Macromerion is undiagnostic below the family level and materials referred to Neosaurus, Bathygnathus, and "Oxyodon" are undiagnostic below the subfamily level. The retention of all of these names except "Oxyodon" is recommended as a paleogeographic convenience. "Oxyodon" has been shown by Paton (1974) to be preoccupied, and pelycosaur material assigned to that taxon should be therefore reclassified as Sphenacodontinae incertae sedis.

We describe early Eocene (Wasatchian) occurrences of the isectolophid Homogalax, tapiroids Heptodon posticus, Heptodon cf. H. posticus, and Heptodon sp., as well as early middle Eocene (Bridgerian) fossils of the brontothere Palaeosyops from localities in the Margaret Formation of the Eureka Sound Group on Ellesmere Island, Nunavut, Arctic Canada. Their occurrence on Ellesmere Island considerably extends the geographic range of these taxa, previously known from mid-latitude localities in British Columbia (only Heptodon), the Western Interior of the United States, and Asia (Homogalax, Heptodon, and Palaeosyops). We also place the fossil localities near Bay Fiord on central Ellesmere Island into a refined lithostratigraphic framework based upon data from three measured stratigraphic sections. Our stratigraphic data confirm the presence of two, stratigraphically distinct fossil assemblages — a late Wasatchian-aged lower assemblage and a Bridgerian-aged upper assemblage that were previously hypothesized by others based on faunal differences — that are separated by a 478 m thick stratigraphic gap that appears to lack fossil vertebrates. From a paleoenvironmental perspective, occurrence of the tapiroid Heptodon in the Eocene Arctic corroborates an hypothesis put forward by others that tapiroids are proxies for densely forested habitats, although they were adapted to a range of temperatures including near (or at) freezing temperatures of Eocene Arctic winters. Further, Arctic occurrences of tapiroids and brontotheres imply that these typical mid-latitude ungulate mammals were adapted to Arctic environments, thereby increasing the probability of Trans-Beringian dispersal during early and middle Eocene time.

Upper Cretaceous dinosaur bonebeds are common in Alberta, Canada, and have attracted continuous scientific attention since the 1960s. Since its inception, the Royal Tyrrell Museum of Palaeontology has documented the presence of hundreds of these sites and has been involved directly in the scientific study of many tens. Because many of these bonebeds have been used to address questions about the paleobiology and paleoecology of dinosaurs, questions have arisen about bonebed origins and preservation in the Cretaceous of Alberta. This study of 260 bonebeds delineates broad paleoenvironmental settings and associations, and taphonomic signatures of assemblages as a first step in assessing patterns of dinosaur bonebed origins in the Upper Cretaceous of Alberta. Bonebeds are known predominantly from the Belly River Group and the Horseshoe Canyon, lower St. Mary River, Wapiti, and Scollard formations. In these units, bonebeds are mostly associated with river channel and alluvial wetland settings that were influenced by a subtropical to warm-temperate, monsoonal climate. Most bonebeds formed in response to flooding events capable of killing dinosaurs, reworking and modifying skeletal remains, and burying taphocoenoses. The "coastal-plain-flooding hypothesis," proposed in 2005, suggested that many bonebeds in the Dinosaur Park Formation formed in response to the effects of recurring coastal-plain floods that submerged vast areas of ancient southern Alberta on a seasonal basis. It remains the best mechanism to explain how many of the bonebeds were formed and preserved at Dinosaur Provincial Park, and here, is proposed as the mechanism that best explains bonebed origins in other Upper Cretaceous formations across central and southern Alberta.