Historical Background

The two scaphite species redescribed in this paper are based on material that was originally collected in the mid-19th century during field expeditions to South Dakota and Montana, in what was then called the Nebraska Territory (fig. 3). These expeditions resulted in impressive monographs that covered all aspects of the natural history of the area, including paleontology and stratigraphy. The history of these expeditions, spanning more than 50 years, has been described in detail by Goetzmann (1966), Waage (1968), Hartman (1984, 1999), Foster (1994), and Martin et al. (2007). We are interested in determining the localities where the original ammonite specimens were collected, inasmuch as these sites represent the type localities for the two species.

Fig. 3.

Drainage map of southwest South Dakota showing the rivers and creeks mentioned in early accounts of western exploration of part of what was then known as the Nebraska Territory. The numbers 1 and 2 indicate Sage Creek and Bear Creek, respectively, two of the collecting localities of Meek and Hayden in 1853.

The description of Scaphites nodosus ( =  Hoploscaphites nodosus) was published by David Dale Owen in 1852 in the “Report of a geological survey of Wisconsin, Iowa, and Minnesota; and incidentally of a portion of Nebraska Territory made under instructions from the United States Treasury Deparment” (fig. 4). In 1849, Owen employed one of his “subagents,” John Evans, to explore the Upper Missouri River. Evans accompanied a party from the American Fur Company, who facilitated his travel, and eventually reached the White River Badlands east of the Black Hills in Nebraska Territory, now western South Dakota.

Fig. 4.

Hoploscaphites nodosus (Owen, 1852), macroconch with part of the hook missing, holotype, UC 6381, Pierre Shale, probably from the Sage Creek area, Pennington County, South Dakota. A. Right lateral; B. apertural; C. ventral. Specimen is illustrated natural size.

Owen (1852: 195) published the “substance” of Evan's report. According to his report, he noted the presence of Hoploscaphites nodosus “in the Fox Hills, which form the dividing ridge between the Cheyenne and Moreau River.” However, in the description of this species and the figure caption accompanying the illustration of Scaphites nodosus, Owen (1852: 581, pl. 8, fig. 4) listed the locality of this species as “Sage Creek, a southern tributary of the Cheyenne,” not the dividing ridge between the Cheyenne and Moreau rivers (fig. 3). According to Owen (1852: 195), Evans did, indeed, visit Sage Creek, noting the presence of Inoceramus and Baculites at the site, and added that

Of the two localities mentioned by Owen, it is much more likely that the holotype of Hoploscaphites nodosus came from the vicinity of Sage Creek rather than the dividing ridge between the Cheyenne and Moreau rivers. The Pierre Shale crops out extensively in the Sage Creek area and contains many specimens of this species. It is less likely that the specimen was collected to the north (the dividing ridge between the Cheyenne and Moreau rivers) because most of the exposures in this area are stratigraphically higher.

The holotype of Hoploscaphites nodosus is covered with patches of orange-red material, especially on the body chamber. Similarly colored fossils are sometimes found in medicine bags of Indians from the Northern Great Plains (Landman, 1983). Such fossils were rubbed with a red substance and adorned with pieces of hide and strands of beads. These “sacred stones” or “rock medicine” were used in ceremonies by tribal medicine men to cure the sick. Many of these objects were brought back by explorers during expeditions to the American West in the 19th century. If the holotype of H. nodosus were such an amulet, then Evans would have acquired it by trade rather than collection, and its provenance would be even more difficult to establish. (As a curiosity, a specimen of H. brevis has been documented in the Hopewell Mounds, Ross County, Ohio, in deposits dating from 100 BCE to 300 CE; J Murphy, personal commun.)

The mineralogy of the red material on the holotype of Hoploscaphites nodosus was determined using X-ray diffraction analysis. For comparison, we analyzed similar red material on a specimen (YPM 11169) of Clioscaphites vermiformis (Meek and Hayden, 1862) purported to be from a medicine bag, according to the label (fide Landman, 1983: 23, top figure). The results indicate that the red material on the specimen of C. vermiformis is a mixture of hematite and quartz grains whereas the red material on the holotype of H. nodosus is cinnabar. The cinnabar granules are suspended in a clear resin, possibly lacquer or glue. Unlike hematite, cinnabar does not occur naturally in the Northern Great Plains. It is possible that cinnabar was a pigment in commercial paint in the 19th century. Such paint may have been applied to the specimen after it was brought back east to protect it or prepare it for museum display. In any event, the fact that the red material on the holotype of H. nodosus is cinnabar indicates that the specimen probably did not originate in a medicine bag, but was collected by Evans in the field.

Owen did not cite the repository of the holotype of Hoploscaphites nodosus nor any of the other specimens he described. The holotype is presently in the Field Museum and is listed as coming from the William F.E. Gurley Collection. Gurley was State Geologist of Illinois from 1893–1897 and his collection was acquired by the University of Chicago (Winchell, 1900). In 1900, Gurley became Associate Curator at the Walker Museum of Paleontology at the University of Chicago and continued there until his death in 1943. The fossil collection of the Walker Museum was transferred to the Field Museum in 1965.

Additional specimens of Hoploscaphites nodosus and what was later called Scaphites nodosus var. brevis ( =  Hoploscaphites brevis) (fig. 5) and Scaphites nodosus var. quadrangularis ( =  H. brevis and H. plenus, microconchs) (fig. 6) were collected by Fielding B. Meek and Ferdinand V. Hayden. This dynamic duo first visited what is now western South Dakota in the summer of 1853, while still assistants of James Hall, then State Geologist of New York. They explored Sage Creek and Bear Creek, both of which are tributaries of the Cheyenne River (fig. 3). Their experience in Sage Creek was recorded by Meek in his journal entry for Monday, June 27, 1853 (housed in the Meek archives of the Smithsonian Institution):

Fig. 5.

Hoploscaphites brevis (Meek, 1876), macroconch, holotype, USNM 367, Pierre Shale, probably from the Sage Creek area, Pennington County, South Dakota. A. Right lateral; B. apertural; C. ventral; D. left lateral. Specimen is illustrated natural size.

Fig. 6.

A–C. Hoploscaphites plenus (Meek, 1876), microconch, USNM 365 ( =  Scaphites nodosus var. quadrangularis, paratype, cast), illustrated in Meek (1876: pl. 25, fig. 4), said to be from the Pierre Shale on the “Yellowstone River, Montana, 150 miles above its mouth,” and probably, more specifically, from the Cedar Creek Anticline, Dawson County, Montana. A. Right lateral; B. ventral; C. left lateral. D–G. Hoploscaphites brevis (Meek, 1876), microconch, USNM 386690 ( =  Scaphites nodosus var. quadrangularis, paratype) illustrated in Meek (1876: pl. 25, fig. 2a–c), said to be from the Pierre Shale on the “Yellowstone River, Montana, 150 miles above its mouth,” but probably from the Sage Creek area, Pennington County, South Dakota. The aperture is restored in Meek's illustration. D. Right lateral; E. apertural; F. ventral; G. left lateral. H–K. Hoploscaphites plenus (Meek, 1876), microconch, USNM 366 ( =  Scaphites nodosus var. quadrangularis, holotype), illustrated in Meek (1876: pl. 25, fig. 3a–c), said to be from the Pierre Shale on the “Yellowstone River, Montana, 150 miles above its mouth,” and probably, more specifically, from the Cedar Creek Anticline, Dawson County, Montana. It exhibits a repaired injury just adapical of the point of recurvature. H. Right lateral; I. apertural; J. ventral; K. left lateral. Specimens are illustrated natural size. The position of the base of the body chamber in D, G, H, and K is estimated due to poor preservation.

According to a biography of Hayden, the two men amassed 13 boxes of mostly Cretaceous fossils (Foster, 1994).

The material was sent home to James Hall who later sold a portion of the collection to compensate for his expenses (Foster, 1994). A part of this collection was purchased by the American Museum of Natural History in 1875 for 65,000 (Landman and Winston, 1999). The reason for this purchase is stated in The Seventh Annual Report of the American Museum of Natural History (1875: 6): “As this work was done from the basis of the New York geological formations, and large collections were made from the Western States for the fuller and more complete illustration and the fixing of the New York geological nomenclature, it became a matter of just pride with us to secure… the interesting and authentic examples of a work so extensive and important.” The AMNH collection contains several specimens of Hoploscaphites nodosus and H. brevis from the Hall collection. For example, AMNH 9520/1, a macroconch of H. nodosus, is recorded as having come from the Hall collection and as having been collected at Sage Creek (unpublished Catalogue of the Palaeontological Collection, vol. 3; not listed in Whitfield and Hovey, 1898). It is possible that this specimen was part of Meek and Hayden's original haul.

After the joint expedition of Meek and Hayden in 1853, Hayden made many more excursions into the field. In the summer of 1854, he explored the region around Fort Union, collecting fossils along the Yellowstone River and at the mouths of the Powder, Tongue, and Bighorn rivers. He revisited Sage Creek in February 1855. His experience was described by Foster (1994: 64): “Despite storms that dumped two feet of snow and dropped the thermometer twenty below zero, they reached Sage Creek where Hayden spent nearly a week. He also collected around Pineas Spring, on the north fork of what is now called the Bad River.” [Pineas Spring, variously spelled Poeno Spring or Pino's Spring, was a popular camp site on the Fort Pierre–Fort Laramie trail located at the head of present day Poeno Creek, a small tributary of the North Fork of the Bad River (fide Waage, 1968)]. Hayden spent another five days at Sage Creek and Bear Creek in May 1855, and filled two carts with fossils. Hayden was also invited by Alexander Culbertson, the agent in charge of all the forts on the upper Missouri and Yellowstone rivers to spend the summer of 1855 around Fort Benton, where he collected fossils along the Missouri River and its tributaries.

Hayden returned to St. Louis in January 1856 with the collections from 1854–1855. According to Foster (1994: 66), “he brought an extraordinary harvest: six tons of specimens accumulated over two years, including more than a thousand pounds of fossils he wanted to work up with Meek, which he expected would occupy them for five years.” Meek saw some of the collection in early February 1856 and wrote “They are grand—magnificent. All of them as perfect as recent shells. There are not less than 50 species most of which are new” (quoted in Foster, 1994: 67). In the Meek archives of the Smithsonian Institution, there is a newspaper clipping attached to a letter from Hayden to Meek dated February 6, 1856. The clipping, from a St. Louis newspaper, describes Hayden's collections and is entitled “Remarkable Cabinet of Natural Curiosities”:

The description of the scaphites collected by Meek and Hayden was published in installments starting in 1856. Some of the material was described by Hall and Meek (communicated 1854; imprinted 1855; published 1856) who listed Scaphites nodosus and analyzed the ontogenetic development of its suture, but did not illustrate any specimens. In 1856, Meek and Hayden published five articles in Proceedings of the Academy of Natural Sciences of Philadelphia (designated as Meek and Hayden, 1856a–232233234e, in the References). In one of the papers (1856e), they listed Scaphites nodosus as coming from Formation 4 ( =  Pierre Shale). Meek and Hayden (1860) included S. nodosus, S. nodosus var. brevis, and S. nodosus var. quadrangularis in their systematic catalog of fossils collected in the Nebraska Territory by the exploring expeditions under the direction of Lieutenant G.K. Warren. Meek (1864) included these same species in his checklist of the invertebrate fossils of North America.

Meek's full description of Scaphites nodosus var. brevis and S. nodosus var. quadrangularis did not appear until 1876. Although Hayden made additional field trips to western South Dakota in 1856, 1857, and 1866, Meek's descriptions were based on material collected in 1854–1855. Hayden (in Meek, 1876: iii, iv) commented that “The accumulation of the materials which compose this volume was commenced in the spring of 1854, and the greater number of the new species of fossils were discovered by the writer of this letter during that and the succeeding year.”

Meek (1876) provided the following information about the locality of the holotype of Scaphites nodosus var. brevis (USNM 367) (fig. 5): “Dr. Owen's type-specimen came from near Cheyenne River, Dakota, where it was found in the upper part of the Fort Pierre group. Ours is from the same horizon on the Yellowstone River, Montana, 150 miles above its mouth.” The same locality appears on the label for this specimen. In the Catalogue of the Type Specimens of Fossil Invertebrates in the Department of Geology, U.S. National Museum (Schuchert, 1905: 588), the locality is further restricted to the Yellowstone River, near Miles City, Montana, although this was probably added later on (in fact, a distance of 150 miles above the river mouth, as measured along the river, is midway between Miles City and Glendive, Montana). In contrast, the entry for this specimen in the handwritten catalog of the Department of Paleobiology, U.S. National Museum, states that the specimen was collected from the south fork of the Cheyenne River, that is, in western South Dakota. Of these two localities, the latter is much more in agreement with the source of most of our specimens and that of the holotype of Hoploscaphites nodosus. In contrast, the exposures of the Pierre Shale near Glendive, Montana, in the Cedar Creek Anticline, while richly fossiliferous, are higher up in the section and yield specimens of H. plenus rather than H. brevis.

Meek (1876: 429) cited the same Yellowstone River locality for Scaphites nodosus var. quadrangularis. In the Catalogue of the Type Specimens of Fossil Invertebrates in the Department of Geology, U.S. National Museum (Schuchert, 1905: 588), the locality is given as both “Fort Pierre (Cret.). Yellowstone River, near Miles City, Montana, and Cheyenne River, South Dakota.” The illustrated specimens are a mixture of microconchs of two species: Hoploscaphites brevis (Meek, 1876: pl. 25, fig. 2a–c  =  USNM 386690) (fig. 6D–G) and H. plenus (Meek, 1876: pl. 25, fig. 3a–c  =  USNM 366; fig. 4  =  USNM 365) (fig. 6H–K and 6A–C, respectively). In the handwritten catalog of the Department of Paleobiology, U. S. National Museum, a note states that these three specimens were renumbered. We suspect that the two microconchs of H. plenus are from the vicinity of the Yellowstone River, Montana, and the microconch of H. brevis is from the vicinity of the Cheyenne River, South Dakota.

Other specimens labeled as Scaphites nodosus var. quadrangularis in the USNM collections also represent a mixture of species. One tray contains six specimens with the number 366 (the number “370” is crossed out) and the locality “Upper Yellowstone 150 mi from mouth.” The specimens are all Hoploscaphites plenus and, as such, the stated locality is probably correct. However, another specimen (USNM 315704) labeled as a paratype of Scaphites nodosus var. quadrangularis is actually a macroconch of H. nodosus, although it is also labeled as coming from the Yellowstone River, Montana. The specimen (USNM 12291) illustrated by Whitfield (1880: pl. 13, figs. 10, 11) (fig. 7A–D) as Scaphites nodosus var. quadrangularis is a microconch of H. brevis. It is said to be from the Pierre Shale “on the Cheyenne River, near Rapid and French Creeks, and also from beds which contain a mingling of the fossils of this and the succeeding group on Old Woman Fork, Black Hills.”

Fig. 7.

Hoploscaphites brevis (Meek, 1876), microconchs. A–D. USNM 12291 ( =  Scaphites nodosus var. quadrangularis), illustrated in Whitfield (1880: pl. 13, figs. 10, 11), said to be from the Pierre Shale “on the Cheyenne River, near Rapid and French Creeks, and also from beds which contain a mingling of the fossils of this and the succeeding group on Old Woman Fork, Black Hills.” The aperture is restored in Whitfield's illustration. A. Right lateral; B. apertural; C. ventral; D. left lateral. E–H. GSC 5342a ( =  holotype, Hoploscaphites landesi Riccardi, 1983, pl. 1, figs. 13, 14), “? Demaine Sandstone, Bearpaw Formation, and South Saskatchewan River opposite mouth of Swift Current River, Saskatchewan.” E. Right lateral; F. apertural; G. ventral; H. left lateral. I–L. GSC 5342b ( =  paratype, Hoploscaphites landesi Riccardi, 1983, pl. 1, figs. 15–17), same locality as E–H. I. Right lateral; J. apertural; K. ventral; L. left lateral. M–Q. GSC 5342c ( =  paratype, Hoploscaphites landesi Riccardi, 1983, pl. 1, figs. 20–22), same locality as E–H. M. Right lateral; N. apertural; O. close-up of the aperture with part of the lower jaw exposed; P. ventral; Q. left lateral. Specimens are illustrated natural size.

The confusion regarding localities can be traced to several sources. Hayden's explorations in 1854–1855 included both the Pierre Shale along the Cheyenne River, South Dakota, and the Pierre Shale along the Yellowstone River, Montana. Therefore, the specimens may have become mixed up in the field, on the trip back, or at the museum. Indeed, as noted, the suite of specimens was renumbered at some point, presumably at the museum. In addition, Meek considered the two sites as representing the same stratigraphic horizon. He also considered all of the loosely uncoiled specimens ( =  microconchs) as representing the same species (which he named Scaphites nodosus var. quadrangularis). As a consequence, he may not have realized that the specimens were mixed up at all, and, in effect, perpetuated the confusion. To further complicate matters, Meek (1876: xxi) wrote most of his massive monograph while wintering in Florida, “away from many desired facilities.”

The scaphite collections at the Academy of Natural Sciences of Philadelphia (ANSP) contain the same errors with respect to these species and localities. The academy purchased these collections in 1857 from the Smithsonian Institution. According to Foster (1994: 75), the academy paid Hayden $1,200 for an assortment of vertebrate and invertebrate fossils collected in 1854–1855. One tray in the ANSP collections contains an unnumbered specimen of Hoploscaphites brevis and a fragmentary, unidentifiable specimen with the number 366 (one of the same numbers as in the Smithsonian collections). There are three labels in the tray, all of which cite the Smithsonian and list the locality as the Yellowstone River, Montana. In addition, there are two other trays, one of which contains fragmentary specimens from the Pierre Shale of Montana, with the number 366, and the other one of which contains a partial specimen of H. brevis with the label “Ft. Pierre Gp., Montana.”

Zonation

Hoploscaphites nodosus and H. brevis are present in the Pierre Shale in Montana, North Dakota, South Dakota, Wyoming, Colorado, and Kansas, and in the Bearpaw Shale (referred to as the Bearpaw Formation by the Geological Survey of Canada) in Saskatchewan, Alberta, and Montana. Both of these formations are subdivided into finer lithostratigraphic units, which vary from region to region. Recently, Martin et al. (2007) elevated the Pierre Shale in South Dakota and eastern Wyoming from formational to group status and, concomitantly, elevated the members of the Pierre Shale to formational status. To maintain consistency with previously published reports, we retain the original nomenclature of the Pierre Shale as used, for example, by Crandell (1958) and Gill and Cobban (1966).

The ammonite zonation of the Upper Cretaceous of the Western Interior of North America was first summarized by Cobban and Reeside (1952). Since then, it has undergone many revisions (e.g., Robinson et al., 1959; Scott and Cobban, 1965, 1975, 1986a, 1986b; Gill and Cobban, 1966; Cobban 1951a, 1958a, 1958b, 1969; Cobban and Hook, 1979; Kennedy and Cobban, 1991; Kennedy et al., 2001; Landman and Waage, 1993), culminating in one of the most refined zonal schemes of any system in the geological record (Cobban et al., 2006). It consists of 66 zones, which are stacked one on top of another. However, it is important to note that the strata themselves are seldom continuously fossiliferous. In particular, the ammonites in the Pierre Shale and Bearpaw Shale tend to occur in concretionary horizons that are separated by intervals that are either nonfossiliferous or contain fossils that are too fragmentary to identify with any certainty. Thus, a zone ususally includes an interval in which the zonal fossils are missing. This is a common and practical approach in establishing a zonation. For example, in defining the ammonite zones of the Pierre Shale at Red Bird, Wyoming, Gill and Cobban (1966: A28) stated “an attempt will be made to estimate the vertical range of certain ammonites at Red Bird including strata that do not contain fossils. . . .” As an illustration, in establishing the top of the Baculites eliasi Range Zone and the base of the overlying Baculites baculus Range Zone, Gill and Cobban (1966: A33) noted that “Unit 87, 15 feet above the highest collection of B. eliasi (D1969), contains abundant Inoceramus typicus, a pelecypod characteristic of the Range Zone of Baculites baculus. This 15-foot interval is arbitrarily divided between the two zones… .” Thus, the top of the B. eliasi Range Zone and the base of the overlying B. baculus Range Zone are defined at a horizon that does not contain either fossil. Although this approach is unavoidable, it nevertheless introduces a certain amount of error into the zonal scheme, which affects the limits of resolution in correlating strata from one geographic region to another.

Recognition of a biostratigraphic zone also depends upon the proper identification of the fossil taxon on which the zone is based. As a result of taxonomic revisions over time, the upper and lower boundaries of a zone can change and new zones and subzones can be added, leading to a more refined zonation. For example, in the early version of the ammonite zonation of the Upper Cretaceous of the Western Interior (Cobban and Reeside, 1952), the Baculites compressus Zone occupies a thick stratigraphic interval. Baculites compressus (Say, 1820) is characterized by an elliptical cross section, smooth to weakly ribbed venter, and lack of ornament on the flanks, except in large specimens. Subsequently, Baculites cuneatus Cobban, 1962a, was described. This form was previously included in B. compressus, but differs from it in having a trigonal cross section with a very narrow venter, and moderately strong ribbing on the venter and flanks in both small and large specimens. In addition, B. compressus var. robinsoni Cobban, 1962a, was described from central Montana and southern Canada. It differs from B. cuneatus in having a broader venter, stronger ornament, and less deeply incised suture. According to Cobban (1962a), this variety is either a geographic subspecies of B. compressus or transitional to B. cuneatus. As a consequence of these taxonomic revisions, later versions of the ammonite zonation of the Upper Cretaceous (e.g., Gill and Cobban, 1966; Scott and Cobban, 1986a) include a thinner B. compressus Zone overlain by the B. cuneatus Zone.

Hoploscaphites nodosus and H. brevis are widespread in the Baculites compressus–B. cuneatus zones of the Western Interior of North America (fig. 2). They also extend into the upper part of the underlying Didymoceras cheyennense Zone and into the lower part of the overlying Baculites reesidei Zone, but their exact distribution in these zones is not yet known. These four zones occupy the middle of the upper Campanian, according to the informal tripartite subdivision of the Campanian in North America (Cobban et al., 2006), and represent what is known as the Bearpaw cycle (Kauffman and Caldwell, 1993). Cobban et al. (2006) reported the radiometrically determined ages of bentonites from these zones as follows: 74.67 ± 0.15 Ma for a bentonite from the Didymoceras cheyennense Zone, 73.52 ± 0.39 Ma for a bentonite in the lower part of the Baculites compressus Zone, and 72.94 ± 0.45 Ma for a bentonite from an unspecified level in the B. reesidei Zone. Thus, the duration of the B. compressus–B. cuneatus zones ( =  the difference in age between the bentonite in the lower part of the B. compressus Zone and the bentonite from an unspecified level in the B. reesidei Zone) is approximately 580,000 years. The range of H. nodosus and H. brevis ( =  the difference in age between the bentonite in the upper part of the D. cheyennense Zone and the bentonite at an unspecified level in the B. reesidei Zone) is approximately 1.73 Ma. Both species show a slight increase in size throughout this interval, expecially in the B. cuneatus Zone, but are otherwise unchanged.

The Baculites compressus–B. cuneatus zones are present in western Canada. They have been reported from the Bearpaw Shale in southern Alberta (Williams, 1930; Williams and Dyer, 1930; Russell and Landes, 1940; Russell, 1950; Gill and Cobban, 1966; Riccardi, 1983; Larson et al., 1997; Tsujita and Westermann, 1998) and southern Saskatchewan (Fraser et al., 1935; Robinson, 1945; Caldwell, 1968; Riccardi, 1983). In addition, Wickenden (1945: 49) noted the presence of the Baculites compressus Zone in the Riding Mountain Formation in southwestern Manitoba, which is equivalent to the Bearpaw Shale in Alberta and Saskatchewan.

The Baculites compressus–B. cuneatus zones are also present in central and eastern Montana (figs. 7, 8) (Dobbin and Erdmann, 1955; Cobban, 1962a, 1962b; Jensen and Varnes, 1964; Izett et al., 1971; Gill et al., 1972; Larson et al., 1997). They occur in the Bearpaw Shale in Glacier, Chouteau, Meaghar, Blaine, Fergus, Wheatland, Golden Valley, Musselshell, Stillwater, Big Horn, Rosebud, Garfield, Teton, and Valley counties; in the Livingston Formation in Park County; and in the Pierre Shale in Carter County. The combined thickness of the two zones near Mosby, Garfield County, and Porcupine Dome, Rosebud County, is 30 m and 45 m, respectively (figs. 8, 9).

Fig. 8.

Stratigraphic section of the Bearpaw Shale near Mosby, east-central Montana, spanning the Scaphites leei II to Baculites eliasi zones. The Baculites compressus–B. cuneatus zones are approximately 30 m thick.

Fig. 9.

Stratigraphic section of the Bearpaw Shale on Porcupine Dome, southeast Montana, spanning the Baculites sp. (smooth) to B. grandis zones. The Baculites compressus–B. cuneatus zones are approximately 45 m thick.

The Didymoceras cheyennense–Baculites cuneatus zones are absent in east-central Wyoming. In their study of the Red Bird section of the Pierre Shale in Niobrara County, Wyoming, Gill and Cobban (1966: 32) noted that “only 6 feet of shale in the upper part of the lower unnamed shale member (unit 68) seems to be assignable to these zones.” No phosphatic pebbles or other features indicating slow deposition are present. They hypothesized that the absence of these ammonite zones may be due to submarine erosion. Gill et al. (1970) documented a widespread unconformity farther west in the south-central part of Wyoming at the base of the Teapot Sandstone Member of the Mesaverde Formation, which may correspond to this interval. In contrast, the Baculites compressus–B. cuneatus zones are present in the Pierre Shale in Weston and Laramie counties, Wyoming.

In South Dakota, the Baculites compressus–B. cuneatus zones are well exposed in the western two-thirds of the state. They have been reported from the Pierre Shale in Butte, Pennington, Custer, Shannon, Meade, Haakon, Ziebach, Mellette, Jones, Stanley, Dewy, Hughes, Lyman, Buffalo, and Brule counties (Searight, 1937; Crandell, 1958; Robinson et al., 1959; Cobban, 1962a, b; Cobban and Larson, 1997; Black, 1964; Stoffer, 1998, 2003; Fox, 2007; Hanczaryk and Gallagher, 2007; Martin et al., 2007). The combined thickness of the two zones is approximately 30 m at AMNH locs. 3408 and 3409 along Elk Creek, Meade County. At AMNH loc. 3212, along Cedar Creek, Shannon County, the combined thickness of the two zones is approximately 14 m (Stoffer, 1998). In some exposures in Meade County, e.g., AMNH loc. 3274, we have noted that both B. compressus and B. cuneatus are present in the same horizon, and even in the same concretion. It is possible that such strata represent a transitional interval between the two zones. The B. compressus–B. cuneatus zones are also present in the Pierre Shale in Barnes County, North Dakota, and in Dawes County, Nebraska (Tharalson, 1993).

In Colorado, Scott and Cobban (1965, 1975, 1986a, 1986b) mapped the Pierre Shale along the Front Range of the Rocky Mountains. In the Colorado Springs–Pueblo area, the Baculites compressus and B. cuneatus zones are present in the the cone-in-cone zone of Lavington (1933), and are 24 m and 12 m thick, respectively. Between Loveland and Round Butte, the two zones occur in the unnamed shale member of the Pierre Shale and are 16 m and 8 m thick, respectively. The two zones are also present near Kremmling, Grand County, Colorado, in the unnamed shale member of the Pierre Shale between the Carter Sandstone Member below and the Gunsight Pass Member above (Izett et al., 1971; Izett and Barclay, 1973; Cobban et al., 1992; Kennedy et al., 2000a). The lowest occurrences of B. compressus, B. cuneatus, and B. reesidei in this area are approximately 116 m, 128 m, and 149 m above the top of the Carter Sandstone Member, respectively (Izett et al., 1971), implying that the B. compressus and B. cuneatus zones are 12 m and 21 m thick, respectively. In Wallace County, Kansas, the B. compressus–B. cuneatus zones have been reported from the Weskan Shale Member and lower part of the overlying Lake Creek Shale Member of the Pierre Shale (Elias, 1933).

Correlation

The biostratigraphic interval containing Hoploscaphites nodosus and H. brevis in the Western Interior corresponds to the middle upper, but not uppermost Campanian (fig. 10B). It is useful to correlate this interval with sections elsewhere in North America and Western Europe to determine the geographic and biostratigraphic distribution of these ammonites, as well as related forms. We discuss the biostratigraphy of the Western Interior, the Gulf Coastal Plain, the Atlantic Coastal Plain, Tercis, France, and the Vistula River Valley, Poland, in that order. Correlation between these regions is possible based on the biostratigraphic distribution of nostoceratid ammonites and inoceramids (fig. 10B). In our discussion, certain taxa appear in quotation marks, e.g., “Acanthoscaphites,” to indicate that the generic name is used informally. The Campanian/Maastrichtian boundary is placed in the upper part of the Baculites eliasi Zone, following the interpretation of Walaszczyk et al. (2002) and Walaszczyk (2004), based on the definition of the boundary at the Global Standard stratotype Section and Point (GSSP) for the Campanian/Maastrictian boundary at Tercis, France. On the other hand, strontium isotope analyses have suggested that the boundary actually occurs in the B. jenseni Zone (McArthur et al., 1992).

Fig. 10.

A. Acanthoscaphites praequadrispinosus Błaszkiewicz, 1980 (38: pl. 19, figs. 6–8), holotype, IG 1.310.II.17, macroconch, left lateral, Nostoceras (N.) hyatti Zone, Piotrawin, Vistula River Valley, Poland. Forms similar or identical to this species occur in the Baculites reesidei-B. jenseni zones of the Western Interior of North America. The specimen is illustrated natural size.

Fig. 10.

B. Comparison of the biostratigraphy of the upper part of the Campanian in the Western Interior of North America, the Gulf and Atlantic Coastal plains, Tercis, France, and the Vistula River Valley, Poland. The shaded area indicates the biostratigraphic interval corresponding to the ranges of Hoploscaphites nodosus and H. brevis. The Campanian/Maastrichtian boundary is placed in the upper part of the B. eliasi Zone, following the interpretation of Walaszczyk et al. (2002) and Walaszczyk (2004), based on the definition of the boundary at Tercis, which is the site of the global standard stratotype section and point (GSSP) for the Campanian/Maastrictian boundary. The thicker range bars for N. (N.) helicinum and N. (N.) hyatti II indicate a higher abundance. ¹Cobban et al. (2006); ²Walaszczyk et al. (2001); ³Riccardi (1983); ⁴Kennedy et al. (1992); ⁵Odin and Cobban (2001); ⁶Küchler and Odin (2001); ⁷Walaszczyk et al. (2002); ⁸Blaszkiewicz (1980); ⁹Walaszczyk (2004).

In the Western Interior of North America, the ammonite zonation of the upper Campanian consists of nine zones (Cobban et al., 2006). Hoploscaphites nodosus and H. brevis range from the upper part of the Didymoceras cheyennense Zone to the lower part of the Baculites reesidei Zone. Several species that resemble H. nodosus and H. brevis are present in the B. reesidei–B. jenseni zones. They include “Jeletzkytes” furnivali Riccardi (1983: 19, pl. 4, figs. 3, 4, 7–9; text figs. 7c, 8), and undescribed forms similar to or identical with “Acanthoscaphites” praequadrispinosus Błaszkiewicz (1980: 38, pl. 19, figs. 2, 3, 6–8; pl. 20, figs. 1–3, 6–8; pl. 21, figs. 1–6) (fig. 10A), H. vistulensis Błaszkiewicz (1980: 34, pl. 17, figs. 4–6, 8, 9), and H. minimus Błaszkiewicz (1980: 35: pl. 23, fig. 4; pl. 24, fig. 3; pl. 25, figs. 3, 4). In addition, Nostoceras (N.) hyatti Stephenson, 1941, a widespread late Campanian ammonite, has been reported from the B. jenseni Zone in Huerfano County, Colorado (Kennedy et al., 1992; Odin and Cobban, 2001). It has since been interpreted as N. (N.) hyatti subspecies II by Küchler and Odin (2001).

Walaszczyk et al. (2001) established five inoceramid zones in the upper Campanian of the Western Interior. The top three zones are, in ascending order, the “Inoceramus” altus Zone, the “I.” oblongus Zone, and the “I.” redbirdensis Zone. The “I.” altus Zone correlates with the Didymoceras cheyennense–B. cuneatus zones (Walaszczyk et al., 2001: fig. 6). Walaszczyk et al. (2002: 277) added that

The “I.” oblongus Zone correlates with the B. reesidei and B. jenseni zones (Walaszczyk et al., 2001, fig. 6), although the range of “I.” oblongus appears to be restricted to the B. reesidei Zone (Walaszczyk et al., 2001: fig. 5). Indeed, Walaszczyk et al. (2002: 277) noted that “the ‘I.’ oblongus Zone characterizes roughly the ammonite zone of Baculites reesidei. The precise correlation of its lower and upper boundaries is, however, unknown.” The “I.” redbirdensis Zone approximately correlates with the B. eliasi Zone (Walaszczyk et al., 2001: fig. 6).

The biostratigraphic interval corresponding to the ranges of Hoploscaphites nodosus and H. brevis on the Gulf Coastal Plain is devoid of ammonites and is unzoned. Kennedy et al. (1992) identified the upper Campanian Nostoceras (N.) hyatti Zone in the Saratoga Chalk in southwestern Arkansas (Kennedy and Cobban, 1993), the Coon Creek Tongue of the Ripley Formation in Tennessee (Cobban and Kennedy, 1994), and the Nacatoch Sand in Texas (Kennedy and Cobban, 1993). All three sites contain N. (N.) hyatti subspecies II and N. (N.) helicinum (Shumard, 1861). Species similar to or identical with “Acanthoscaphites” praequadrispinosus are also present: in Arkansas—called Jeletzkytes nodosus by Kennedy and Cobban (1993: fig. 17.22–25); in Tennessee—called Scaphites reesidei by Wade (1926: 183, pl. 61, figs. 3–7); J. nodosus by Cobban and Kennedy (1994: pl. 9, figs. 7–11); and J. reesidei by Larson (in press: pl. 10, figs. 3a–8b); and in Texas—called S. rugosus by Stephenson (1941: 425, pl. 89, figs. 15–18).

The biostratigraphic interval corresponding to the ranges of Hoploscaphites nodosus and H. brevis on the Atlantic Coastal Plain is also devoid of ammonites and is marked by an unconformity at the base of the Navesink Formation. However, the fauna of the Mount Laurel Sand at Biggs Farm, Delaware (Kennedy and Cobban, 1994), includes Didymoceras cheyennense Meek and Hayden, 1856b, Spiroxybeloceras meekanum (Whitfield, 1877), Anaklinoceras reflexum Stephenson, 1941, and Baculites undatus Stephenson, 1941. All of these forms suggest a correlation with the D. cheyennense Zone and, possibly, the lower part of the B. compressus Zone in the U.S. Western Interior. The Nostoceras (N.) hyatti Zone is present in the phosphatized fauna at the base of the Navesink Formation at Atlantic Highlands, New Jersey (Cobban, 1974; Kennedy and Cobban, 1993; Kennedy et al., 1992, 1995, 2000b). The ammonite fauna at this site consists of 11 species including N. (N.) hyatti subspecies II, N. (N.) helicinum, and forms similar to or identical with “Acanthoscaphites” praequadrispinosus (called Scaphites nodosus by Whitfield, 1892: 261: pl. 44, figs. 13, 14; S. nodosus? by Weller, 1907: pl. 107, figs. 1, 2; S. aff. S. leei by Reeside, 1962: 126, 127, pl. 71, figs. 8–11; Hoploscaphites sp. by Cobban, 1974: 18, pl. 11, figs. 13–19; text fig. 14; and Jeletzkytes cf. nodosus by Kennedy et al., 2000b: 20, figs. 9J–P, 10, 11, 12C–F).

At Tercis, France, the Nostoceras (N.) hyatti Zone marks the top of the Campanian, although Walaszczyk (2004: fig. 5) suggested that it also extends into the lower Maastrichtian. Küchler and Odin (2001: fig. 2) analyzed the biostratigraphic distribution of the index species in this section, as well as that of other nostoceratids and diplomoceratids, indicating the position of fossils by level numbers, spaced at intervals of approximately 1 m. The base of the N. (N.) hyatti Zone, marked by the lowest occurrence of N. (N.) cf. hyatti, is at level 66.4. The lowest occurrences of N. (N.) hyatti subspecies I and II are at levels 66.5 and 94.2, respectively. The lowest occurrence of N. (N.) helicinum is at level 67, but this species is more abundant higher up in the section in association with N. (N.) hyatti subspecies II (thicker range bars, fig 10B).

Walaszczyk et al. (2002: 276, chart) recognized five inoceramid zones in the upper Campanian–lower Maastrichtian at Tercis, the uppermost four of which are, in ascending order, the “Inoceramus” altus Zone, the “I.” oblongus Zone, the Trochoceramus costaecus Zone, and the “I.” redbirdensis Zone. The bases of these four zones, defined by the lowest occurrences of their respective index fossils, are at level 66.5, between levels 80 and 88, at level 96.7, and at level 108.

Walaszczyk et al. (2002) correlated these inoceramid zones to the ammonite zones in the U.S. Western Interior. According to them, the “Inoceramus” altus Zone corresponds to the Didymoceras cheyennense–Baculites cuneatus zones. However, there is some confusion regarding the correlation of the “I.” oblongus, Trochoceramus costaecus, and “I.” redbirdensis zones. In Walaszczyk et al. (2002: 276, chart), the “I.” oblongus Zone correlates with the B. reesidei Zone, the T. costaecus Zone correlates with the Baculites jenseni Zone, and the “I.” redbirdensis Zone correlates with the B. eliasi Zone. In contrast, in Walaszczyk (2004: fig. 5), the“I.” oblongus Zone correlates with the B. reesidei Zone and the lower two-thirds of the B. jenseni Zone, the T. costaecus Zone correlates with the upper one-third of the B. jenseni Zone and the lower one-third of the B. eliasi Zone, and the “I.” redbirdensis Zone correlates with the upper two-thirds of the B. eliasi Zone. We have followed the correlation scheme in Walaszczyk (2004: fig. 5), since this publication presumably reflects the latest thinking on the subject (fig. 10B).

In the Vistula River Valley, Poland, Błaszkiewicz (1980) established the Didymoceras donezianum and Nostoceras pozaryski zones at the top of the Campanian. Kennedy et al. (1992) synonymized N. pozaryski with N. (N.) hyatti, and Küchler and Odin (2001) further assigned this species to N. (N.) hyatti subspecies II. Błaszkiewicz (1980: 14) noted that “In the assumed concept of the zone, its lower boundary is determined by the appearance of Acanthoscaphites praequadripinosus sp. nov., while the first occurrence of the index species is considered as pronouncedly later.” Walaszczyk (2004: 97) added that N. (N). hyatti “first appears distinctly higher and ranges up to approximately the top of the zone” (although elsewhere Walaszczyk [2004: 100], based on a personal communication from M. Machalski, stated that this species may actually appear at the base of the zone). According to Walaszczyk (2004: fig. 5), the N. (N.) hyatti Zone is upper but not uppermost Campanian, and is immediately overlain by the Belemnella lanceolata lanceolata Zone, which was previously assigned to the lower Maastrichtian by Błaszkiewicz (1980).

In Poland, the Nostoceras (N.) hyatti Zone contains “Acanthoscaphites” praequadrispinosus (fig. 10A), Hoploscaphites vistulensis, and H. minimus. In addition, Machalski (personal commun., 2009) reported the presence of scaphites that are “transitional between nodosus and praequadrispinosus” from either the top of the Didymoceras donezianum Zone or the bottom of the N. (N.) hyatti Zone. The ribbing in these specimens is similar to that in H. nodosus, but the size of the adult shell more closely matches that of “A.” praequadrispinosus.

Walaszczyk (2004: fig. 5) recognized five inoceramid zones in the upper Campanian in the Vistula River Valley, Poland, the uppermost four of which are, in ascending order, the “Inoceramus” altus Zone, the “I.” inkermanensis Zone, the Trochoceramus costaecus Zone, and the “I.” redbirdensis Zone, each of which is defined at the base by the lowest occurrence of the respective index fossil. The “I.” altus and “I.” inkermanensis zones correspond to the Nostoceras (N.) hyatti Zone, the T. costaecus Zone correlates with the lower part of the Belemnella lanceolata lanceolata Zone, and the “I.” redbirdensis Zone correlates with the upper part of the B. lanceolata lanceolata Zone and the lower part of the B. occidentalis Zone. In terms of the ammonite zonation of the Western Interior, the “I.” altus Zone correlates with the Didymoceras cheyennense–Baculites cuneatus zones, the “I.” inkermanensis Zone correlates with the B. reesidei Zone and the lower two-thirds of the B. jenseni Zone, the T. costaecus Zone correlates with the upper one-third of the B. jenseni Zone and the lower one-third of the B. eliasi Zone, and the “I.” redbirdensis Zone correlates with the upper two-thirds of the B. eliasi Zone (Walaszczyk, 2004: fig. 5).

In summary, based on the proposed correlation using inoceramids and nostoceratid ammonites (Błaszkiewicz, 1980; Kennedy et al., 1992; Küchler and Odin, 2001; Walaszczyk, 2004; Walaszczyk et al., 2001, 2002; Cobban et al., 2006), there is a slight discrepancy in the position of the interval containing Hoploscaphites nodosus and H. brevis between North America and Western Europe (fig. 10B). In the Western Interior of North America, this interval correlates with the “Inoceramus” altus Zone, which, by indirect correlation, corresponds to unzoned strata below the Nostoceras (N.) hyatti Zone on the Gulf and Atlantic Coastal plains. In contrast, in Western Europe, the interval containing H. nodosus and H. brevis ( =  “I.” altus Zone) correlates with the lower part of the N. (N.) hyatti Zone. In Poland, this zone contains “Acanthoscaphites” praequardripinosus, H. vistulensis, and H. minimus. Similar forms are present in the N. (N.) hyatti Zone in North America above the interval containing H. nodosus and H. brevis.

The biostratigraphic discrepancy between North America and Western Europe raises several questions. Why are H. nodosus and H. brevis present in North America but absent in Poland? Does this difference reflect biogeographic factors or is it the result of differential preservation? With respect to the inoceramid-nostoceratid biostratigraphy, why is the lowest occurrence of “A.” praequadrispinosus and related species lower in Poland than in North America? Does this indicate a real difference in the timing of appearance of these species, with implications for ammonite migration, or is it simply the result of imprecision in the biostratigraphic correlation? How well integrated are the ammonite and inoceramid biostratigraphies? How similar are the ranges of the index species between North America and Western Europe? How do unfossiliferous intervals contribute to the uncertainty of the correlation?

Answering these questions involves further study on several fronts: (1) The taxonomy of the scaphites on both sides of the Atlantic (e.g., “Acanthoscaphites” praequadrispinosus, Hoploscaphites vistulensis, H. minimus, “Jeletzkytes” furnivali, Scaphites reesidei, and S. rugosus) must be reexamined to determine whether these species are the same or different. (2) The biostratigraphy of the inoceramids and nostoceratids must be reevaluated with the goal of specifying areas of uncertainty. For example, in Poland, the base of the Nostoceras (N). hyatti Zone is unclear. In addition, the ranges of scaphites in this zone are not well documented due, undoubtedly, to the rarity of material. (3) Future biostratigraphic studies must take into account unfossiliferous portions of the section, and the common practice, at least in the U.S. Western Interior, of drawing zonal boundaries at horizons lacking index fossils, which invariably leads to some uncertainty. For example, in the Western Interior, the correlation of the “Inoceramus” oblongus Zone with the Baculites reesidei and B. jenseni zones requires further analysis, including a more precise determination of the lowest and highest occurrences of these species, as well as gaps in their ranges. (4) The relative abundance of a species in the section can provide additional data, perhaps permitting a more refined zonation (as per Küchler and Odin, 2001). For example, if the relative abundance of N. (N). hyatti is taken into account (see thicker range bars, fig. 10B), rather than simply its range, the biostratigraphic distribution of this species would line up better between North America and Western Europe. These points will be further addressed in future studies.

Mode of Occurrence

Hoploscaphites nodosus and H. brevis generally occur in carbonate-cemented concretions that are distributed in discontinuous layers (fig. 11). They are usually associated with a rich molluscan fauna, dominated by inoceramids and baculites. The scaphites are generally preserved in three dimensions, with the phragmocone filled with sparry calcite, and the body chamber filled with sediment. In the concretions from the Baculites compressus–B. cuneatus zones of the Pierre Shale in western South Dakota and east-central Montana, most specimens retain their original aragonitic shell structure, although the quality of preservation varies (Cochran et al., 2010). In contrast, in the sandstone concretions from the B. compressus–B. cuneatus zones of the Pierre Shale along the Front Range of the Rocky Mountains, Colorado, the original shell is usually missing.

Fig. 11.

Fossiliferous concretion BHI 4788 from the Baculites cuneatus Zone of the Pierre Shale, Meade County, South Dakota. The concretion contains an adult specimen of B. cuneatus Cobban, 1962, an adult macroconch of Hoploscaphites brevis (Meek, 1876), and “Inoceramus” sagensis Owen, 1852.

To better understand concretion formation, we examined concretions from the Baculites compressus–B. cuneatus zones of the Pierre Shale in Meade and Pennington counties, South Dakota. These concretions cover an area of at least 1,000 km², based on our observations of outcrops along Elk, Sage, and Cedar creeks, all of which are tributaries of the Cheyenne River. This area represents a relatively offshore environment approximately 250 km from the western shoreline of the Western Interior Seaway (fig. 1). However, the position of the shoreline in this area is not well constrained and it may have been farther east than supposed. Although it is difficult to establish the exact depth of the water in this area, Gill and Cobban (1966) suggested that the seaway was probably less than 70 m deep at this time. As a comparison, Feldman (1972) estimated that the Timber Lake Member of the Fox Hills Formation, which also contains abundant scaphitiferous concretions, was deposited at a depth of 18–55 m.

Landman and Klofak (2006) reported the results of a detailed analysis of a concretion (Kathy's concretion) from AMNH loc. 3274 from the Baculites compressus –B. cuneatus zones in the Pierre Shale, Meade County, South Dakota (figs. 12, 13). The concretion is ovoid, and approximately 1 m long by 0.5 m wide. It consists of light grey, finely bioturbated mud, without any indication of bedding structures. In addition to Hoploscaphites nodosus and H. brevis, the concretion contains approximately 40 macroinvertebrate species including Baculites compressus and B. cuneatus, which represent the most abundant group in terms of numbers of specimens, one species of placenticeratid, 16 species of bivalves (both infauna and epifauna), 10 species of gastropods, 2 species of scaphopods, and evidence of crustaceans, bryozoans, encrusting algae, foraminifera, small tubes, and fish scales, and bits of wood. In most of the molluscs, the original mineralogy of the shells is preserved.

Fig. 12.

Scaphites from a single concretion (Kathy's concretion), Baculites compressus–B. cuneatus zones, Pierre Shale, AMNH loc. 3274, Meade County, South Dakota. A, B. Hoploscaphites brevis (Meek, 1876), large macroconch, AMNH 50429. A. Left lateral, uncoated; B. ventral hook. A chunk of shell is missing from the adapical part of the body chamber, probably due to an injury, and the rest of the shell is broken and deformed. C, D. Hoploscaphites brevis (Meek, 1876), large macroconch, AMNH 51068. C. Right lateral; D. ventral. E, F. Hoploscaphites brevis (Meek, 1876), small macroconch, AMNH 51067. E. Apertural; F. left lateral. G, H. Hoploscaphites brevis (Meek, 1876), small microconch, AMNH 55876. G. Right lateral; H. ventral. I, J. Hoploscaphites brevis (Meek, 1876), large microconch, AMNH 51065. I. Right lateral; J. apertural. K, L. Hoploscaphites nodosus (Owen, 1852), microconch, AMNH 55873. Note that the base of the body chamber is 90° adapical of the line of maximum length. K. Right lateral; L. ventral. Specimens are illustrated natural size.

Fig. 13.

Preservation of fossils from the Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. A–J. Single concretion (Kathy's concretion), AMNH loc. 3274. A. Left lateral of a phragmocone of Hoploscaphites brevis (Meek, 1876), AMNH 56778. The specimen is broken at the adapical end of the body chamber, probably due to an injury. B, C. Hoploscaphites brevis (Meek, 1876), small macroconch, AMNH 51067. B, Left lateral; C. apertural. A chunk of shell is missing from the adapical end of the body chamber, but the rest of the specimen is still intact. D. Lower jaw attributed to Hoploscaphites, with a tear in the right wing, AMNH 56808. E. Lower jaw attributed to Hoploscaphites showing damage (arrow) along the lateral margin, AMNH 56803. F. Juvenile of Baculites, with the ammonitella (arrow) still attached, AMNH 56824. G. Isolated piece of Inoceramus shell (arrow) with prismatic microstructure, AMNH 56798. H. Agerostrea falcata (Morton, 1830) with the impression of a baculite (arrow), AMNH 56797. I. Juvenile of Hoploscaphites with most of the body chamber preserved, AMNH 56842. J. Fragment of Baculites with scalloped edges (arrows) along the margin, AMNH 56833. K, L. Hoploscaphites nodosus (Owen, 1852), large macroconch, with the impression of a worm tube (arrow) on the body chamber, AMNH 56763, AMNH loc. 3274. M, N. Baculites compressus (Say, 1820) with impressions of worm tubes (arrow), on the body chamber, AMNH 56870, AMNH loc. 3408. Specimens are illustrated natural size unless indicated otherwise by a scale bar.

The distribution of fossils in the concretion is patchy, with concentrations of shelly material surrounded by less fossiliferous matrix. In these concentrations, the shells are loosely packed but in contact, so that, for example, the broken phragmocone of a baculite abuts the broken body chamber of a scaphite. Adult scaphites of both dimorphs are present. The body chambers of these specimens are always broken, although enough of the body chamber is still preserved in these specimens to identify them as adults. The concretion also contains large scaphite phragmocones in which only a small piece of body chamber is still attached. These specimens cannot be unequivocably identified as adults, and could represent instead large juvenile or submature individuals. In addition, there are juvenile scaphites as small as 4 mm in diameter, with most of the body chamber intact. In large scaphites, the chambers of the phragmocone are either hollow or filled with sediment or calcite, whereas in small scaphites, the chambers of the phragmocone are usually filled with calcite. The shell wall of the scaphites consists of a thin, outer prismatic layer and a thicker nacreous layer (e.g., 15 µm versus 200 µm in AMNH 66261). The body chambers of large baculites (adults?) are always broken. Smaller specimens are more complete and include juveniles 5–10 mm in length with the ammonitella still attached. Some of the smaller specimens (10–20 mm in length) are oriented approximately parallel to each other. As in the scaphites, the chambers of large baculite phragmocones are either hollow or filled with sediment or calcite, whereas the chambers of small baculite phragmocones are always filled with calcite.

The concretion contains dozens of isolated ammonite jaws, almost all of which are deformed. The outer calcitic layer is usually missing or recrystallized. In those specimens in which it is missing, the chitinous part of the jaws lies in direct contact with the rest of the concretion, without any gap. After ammonites, the most abundant fossils are bivalves including a mixture of epifaua (e.g., Limopsis parvula [Meek and Hayden, 1856c], “Inoceramus” altus Meek, 1871) and infauna (e.g., Protocardia subquadrata [Evans and Shumard, 1857], Solemya subplicata [Meek and Hayden, 1856c]). The epifaunal bivalves are always disarticulated, whereas the infaunal bivalves are nearly always articulated, although not in life position. In large inoceramid shells, the beak is usually preserved, but the commissure is broken. In general, the inner nacreous layer (composed of aragonite) is still present, but the outer prismatic layer (composed of calcite) is broken up and scattered as pieces throughout the concretion (fig. 13G).

Most of the breakage in the adult scaphites is probably due to predation. In general, a lethal injury in these forms is manifested by the loss of the adapical end of the body chamber, precluding repair of the shell (Radwański, 1996; Larson, 2003; Keupp, 2006; Klompmaker et al., 2009). Usually, what remains is the phragmocone and a small part of the adapical end of the body chamber terminating in a jagged edge. For example, both AMNH 66238 and 66252 are subadult to adult specimens of Hoploscaphites brevis that consist of only the phragmocone and a small piece of the body chamber. In other specimens, the adapical part of the body chamber is missing, but the rest of the shell is still intact, although deformed due to subsequent compaction. For example, in AMNH 50429 (fig. 12A, B), a large macroconch of H. brevis, a chunk of shell is missing from the adapical end of the body chamber, but the rest of the shell is intact. The broken pieces were probably held together by the periostracum or the mantle tissue lining the inside of the shell. Similarly, in AMNH 55876 (fig. 12G, H), a small microconch of H. brevis, a chunk is missing from the venter and left side of the adapical end of the body chamber, but the right side of the body chamber is intact, although fractured in multiple places. Many of the adult baculites in the concretion exhibit the same pattern of breakage near the base of the body chamber, suggesting that these animals also suffered from a similar form of predation. The shells of the scaphites and baculites may have floated for a short time following an injury. Specimens in which most of the shell is still intact probably settled to the bottom soon after death.

Thus, these shells may have been added to the sea floor through attrition by predation from fish or mosasaurs. Other shells may have been added as a consequence of an environmental disturbance, such as a turbidity current caused by a storm or river discharge. Such turbidity currents may have entrained and killed small baculites and scaphites living just a few meters above the bottom, resulting in minor breakage at the aperture and the current alignment of some of the baculites. Shells may have accumulated in one spot due to irregularities on the sea floor. Such irregularities may have been produced by current induced scours, burrows, or scouring around preexisting shell debris (see Tsujita, 1995, for a similar explanation for the formation of fossiliferous concretions in the Campanian-Maastrichtian Bearpaw Shale in southern Alberta). In contrast to the cephalopods that lived in the water, the gastropods and bivalves lived on the bottom, probably at or near the site. However, they were slightly reworked after death due to currents and bioturbation, as indicated by the fact that epifaunal bivalves are disarticulated and infaunal bivalves are not in life position.

Shells must have rested on the sea floor for some time before burial, or else were buried and then exhumed. For example, in AMNH 56763, a macroconch of Hoploscaphites nodosus from another concretion at the same site, a calcareous worm tube is present on the inside surface of the body chamber, indicating that the empty shell must have rested on the sea floor before burial (fig. 13K, L). In AMNH 56797, a specimen of Agerostrea falcata (Morton, 1830), the left valve shows the impression of a small baculite, indicating that the oyster must have settled on the dead ammonite (fig. 13H). In AMNH 56833, a fragment of the adapical part of the body chamber of a small baculite, the aperture shows a series of irregular notches, suggesting that it was peeled back by a crab to consume the soft tissue (fig. 13J). In many ammonite jaws, the calcitic layer is missing with no gap between the chitinous part of the jaw and the rest of the concretion, implying that the calcitic layer was destroyed prior to concretion formation. Most of these jaws are deformed with ragged margins, and some of them are even pierced with holes, suggesting that they may have been eaten by a predator, and subsequently secreted as faecal matter.

The sequence of concretion formation has been studied at many sites, e.g., Reeside and Cobban (1960)—the Albian Mowry Shale in Wyoming; Waage (1964)—the upper Maastrichtian Fox Hills Formation in north-central South Dakota; Raiswell (1976)—the lower Lias in northeast England; Gautier (1982)—the Campanian Gammon Ferruginous Member of the Pierre Shale in Montana and the Black Hills region; Maeda (1987)—the Upper Cretaceous Yezo Group in northwestern Hokkaido; Carpenter et al. (1988) – the Fox Hills Formation in south-central North Dakota; and Ludvigson et al. (1994)—the Upper Cretaceous Greenhorn marine cycle in Iowa and South Dakota (see also Berner, 1968; Allison, 1990; and Maeda and Seilacher, 1996). These studies suggest that the cementation of concretions occurred at shallow burial depths below the sediment-water interface. This conclusion probably applies to the concretions in our study as well.

In summary, Kathy's concretion represents a time-averaged deposit of organisms derived from the local community. The shells probably accumulated in irregularities on the sea floor and were reworked by bioturbation and weak currents. Some ammonites, such as the adult scaphites, may have died as a result of predation and settled to the bottom. Other ammonites, such as the juvenile baculites, may have died following short-term environmental disturbances. Therefore, this accumulation reflects successive additions derived from multiple sources. Some material may have been buried rapidly, while other material, such as the ammonite jaws, may have rested on the sea floor for some time. However, taphonomic experiments with modern coleoid jaws (Kear et al., 1995) suggest that, by analogy, the ammonite jaws could not have survived exposure on the sea floor for more than several tens of years, and possibly less.

Other concretions from the Baculites compressus–B. cuneatus zones in the Pierre Shale in Meade and Pennington Counties, South Dakota, vary in shape, size, and fossil content, but contain some or all of the same faunal elements. These concretions represent the same biofacies extending over an area of at least 1,000 km². These accumulations can be interpreted as mixed death assemblages, similar to the “within habitat time averaged assemblages” of Kidwell and Bosence (1991) and the “Type C assemblages” of Tsujita (1995). They formed during relatively short periods of time (months to years) and reflect a composite mixture of material derived from both background attrition (predation) and mass mortalities (environmental disturbances).

Given the rate of accumulation of the shells, these concretions can provide valuable information about the structure and composition of scaphite populations. For example, examining the scaphites in single concretions provides a more accurate method of determining the range of variation within a single species, and the size difference between conspecific dimorphs, than examining a sample of specimens from different horizons in the same zone. Nevertheless, it is important to bear in mind that these concretions do not represent census assemblages. Such concretions are present elsewhere, for example, in the Lower Nicolletti Assemblage Zone in the Fox Hills Formation of north-central South Dakota (Waage, 1964; Landman et al., 2003). These concretions are characterized by monospecific clusters of macroconchs (presumably females) of a single species of scaphite, occasionally with jaws preserved inside the body chamber. The interpretation is that the females probably died immediately after spawning, and were rapidly buried.

Repositories

Specimens are identified by species, museum repository number, study number, locality number, geographic area, and stratigraphic unit. Study numbers range from 1 to approximately 2000. The repository is indicated by a prefix, as follows: Department of Geology, University of Alberta (UA), Edmonton, Alberta, Canada; Division of Paleontology (Invertebrates), American Museum of Natural History (AMNH), New York, New York; Academy of Natural Sciences of Philadelphia (ANSP), Philadelphia, Pennsylvania; Black Hills Museum of Natural History (BHI), Hill City, South Dakota; Geological Survey of Canada (GSC), Ottawa, Ontario, Canada; Department of Geology, the Field Museum (variously given as FMNH, FMUC, UC, and UCP), Chicago, Illinois; Geological Museum of the Polish Geological Institute (IG), Warsaw, Poland; Memphis Pink Palace Museum (MPPM), Memphis, Tennessee; Museum of Paleontology, University of Kansas (KUMIP), Lawrence, Kansas; Museum of Paleontology, University of Michigan (UMMP), Ann Arbor, Michigan; Timber Lake and Area Museum (TLAM), Timber Lake, South Dakota; Tyrrell Museum of Paleontology (TMP), Drumheller, Alberta; the U.S. Geological Survey (USGS), Denver, Colorado; U.S. National Museum (USNM), Washington, D.C.; and Yale Peabody Museum (YPM), New Haven, Connecticut. Locality numbers are listed in appendix 1.

Terms and Methods

Most of the terms used to describe scaphites are reviewed by Riccardi (1983), Landman and Waage (1993), Landman and Cobban (2007), and Machalski (2005b). The adult shell consists of a closely coiled phragmocone and a slightly to strongly uncoiled body chamber (fig. 14A, B). The adult phragmocone is the part of the phragmocone that is exposed in the adult shell, as compared to the part that is concealed inside. The point of exposure is the most adapical point of the adult phragmocone. The body chamber consists of the shaft, beginning near the last septum, and a hook terminating at the aperture. The point of recurvature is the point at which the hook curves backward.

Fig. 14.

Scaphite terminology. A. Macroconch, right lateral view. The shell is oriented in the probable floating position when the body was withdrawn into the body chamber. The umbilical seam of the shaft in macroconchs is straight with a slight umbilical bulge. Abbreviations: H₁  =  whorl height ¼ whorl adapical of the long axis; H₂  =  whorl height along the long axis; H₃  =  whorl height at the ultimate septum; H₄  =  whorl height at midshaft; H₅  =  whorl height right before the point of recurvature; H₆  =  whorl height at the point of recurvature; H₇  =  whorl height at the aperture; LMAX =  maximum length along the long axis; apt. <  =  apertural angle; septal. <  =  septal angle. B. Microconch, right lateral view. The microconch is approximately 85% of the size of the macroconch or, inversely, the macroconch is approximately 120% the size of the microconch. The umbilical seam of the shaft in microconchs is curved and follows the curvature of the venter. Specimens are photographed from lateral, ventral, and apertural views, as shown. Asterisks indicate the up position in ventral and apertural views. C. Close-up of the umbilicus of the macroconch showing the umbilical diameter measured parallel to the long axis (UD_L) and the umbilical diameter measured at the base of the body chamber (UD_P). D. View of the venter of the body chamber at midshaft, with the adoral direction toward the top, showing the width of the venter (V₄). E. The angle of orientation of the aperture (orientation <) is defined as the angle of the aperture with respect to the vertical, as determined using the method in Trueman (1941). x  =  center of buoyancy; ·  =  center of mass. F. Median cross section through the ammonitella showing measurements of the ammonitella diameter (AD) and ammonitella angle (AA). G. Dorsoventral cross section of a juvenile specimen. Abbreviations: D  =  diameter, WH  =  whorl height, WW  =  whorl width, UD  =  umbilical diameter.

Several measurements were made on the adult shell (fig. 14, tables 1–10). The maximum length of the adult shell (LMAX) was measured from the venter of the adult phragmocone to the venter of the hook. The umbilical diameter of the adult shell (UD) was measured in two ways: (1) through the center of the umbilicus, parallel to the line of maximum length (UD_L), and (2) through the center of the umbilicus, passing through the position of the last septum (UD_P), as indicated on the umbilical shoulder. Whorl width (W) and whorl height (H) were measured at seven points on the adult shell: (1) on the phragmocone 90° adapical of the line of maximum length, (2) on the phragmocone, along the line of maximum length, (3) at the position of the last septum, (4) at midshaft, (5) just adapical of the point of recurvature, (6) at the point of recurvature, and (7) at the aperture. The width of the venter at midshaft (V₄) was measured between the ventrolateral margins on opposite sides of the venter. All measurements were made using electronic calipers on actual specimens, rather than on photos. Some measurements were estimated (as indicated by asterisks in tables 1–10) in cases of poor preservation. These estimates were not used in computing averages and standard deviations or in plotting graphs.

TABLE 1

Septal approximation at maturity in Hoploscaphites nodosus (Owen, 1852) and Hoploscaphites brevis (Meek, 1876)

TABLE 2

Angle of orientation of the aperture in Hoploscaphites nodosus (Owen, 1852) and Hoploscaphites brevis (Meek, 1876)

TABLE 3

Measurements of the adult shells of Hoploscaphites nodosus (Owen, 1852), macroconchs

TABLE 4

Ornament of the adult shells of Hoploscaphites nodosus (Owen, 1852), macroconchs

TABLE 5

Measurements of the adult shells of Hoploscaphites nodosus (Owen, 1852), microconchs

TABLE 6

Ornament of the adult shells of Hoploscaphites nodosus (Owen, 1852), microconchs

TABLE 7

Measurements of the adult shells of Hoploscaphites brevis (Meek, 1876), macroconchs

TABLE 8

Ornament of the adult shells of Hoploscaphites brevis (Meek, 1876), macroconchs

TABLE 9

Measurements of the adult shells of Hoploscaphites brevis (Meek, 1876), microconchs

TABLE 10

Ornament of the adult shells of Hoploscaphites brevis (Meek, 1876), microconchs

We calculated several ratios to describe the shape of the adult shell and facilitate comparisons among specimens. The ratios of whorl width to whorl height were calculated at seven points on the shell (W₁/H₁–W₇/H₇), as described above, and provide a measure of the degree of whorl compression. The ratio of ventral width to whorl height at midshaft (V₄/H₄) and the ratio of ventral width to whorl width at midshaft (V₄/W₄) were also calculated and provide additional measures of the degree of whorl compression. The ratio of whorl height at midshaft to whorl height on the phragmocone along the line of maximum length (H₄/H₂), and the ratio of whorl height at midshaft to whorl height at the point of recurvature (H₄/H₆) both describe the ontogenetic change in whorl height. The ratio of maximum length to whorl height on the phragmocone along the line of maximum length (LMAX/H₂) is a measure of the degree of uncoiling. For example, Landman (1987: 211) used this ratio to characterize the degree of uncoiling in Turonian-Santonian scaphites in the U.S. Western Interior. The mean value ranged from 3.31 in the most primitive species (Scaphites larvaeformis Meek and Hayden, 1859) to 2.55 in the most advanced species (C. vermiformis [Meek and Hayden, 1862]), reflecting an evolutionary trend toward tighter coiling. The ratio of maximum length to whorl height at midshaft (LMAX/H₄) is a measure of the degree of curvature of the body chamber in lateral view. For example, if the outline of the body chamber in lateral view is a semicircle, the ratio of LMAX/H₄ equals 2. This ratio only applies to macroconchs because the umbilical seam of the body chamber usually coincides with the line of maximum length in these forms, and thus the whorl height is the distance from the line of maximum length to the venter of the body chamber (equivalent to the radius in the case of a semicircle).

The apertural angle is defined as in Landman and Waage (1993) and Machalski (2005b). This angle is measured on photographs of specimens in lateral view. A line is drawn along the umbilical shoulder and another line is drawn along the apertural margin. The apertural angle is the angle of intersection between these two lines, extending from approximately the point of recurvature to the end of the shell. We restricted this measurement to macroconchs, but even in these forms, there is room for error. The umbilical shoulder sometimes shows a bulge or sag, making it difficult to know where to draw the line.

The septal angle describes the position of the last septum relative to the line of maximum length. As in the apertural angle, the septal angle is measured on photographs of specimens in lateral view. The position of the last septum is defined as the position of the median saddle in the ventral (external) lobe. A line is drawn through the position of the last septum (marked on the flanks for convenience) and the center of the umbilicus. Another line is drawn through the center of the umbilicus parallel to the line of maximum length. The septal angle is the angle of intersection between these two lines, extending from the position of the last septum to the line of maximum length. Negative angles are defined as above (adapical of) the line of maximum length; positive angles are defined as below (adoral of) the line of maximum length.

A number of terms are used to describe ornamentation. Primary ribs originate near the umbilicus, whereas secondary ribs originate on the flanks or venter, either by splitting or intercalation (strictly speaking, the latter are called intercalatories). The density of ribs is measured in three ways (Landman and Cobban, 2007). The first method is to count the number of ribs/cm on the venter at a particular spot on the shell, for example, on the midshaft or hook. When the ribs are widely spaced, the rib number may be a fraction (usually ending in 0.25, 0.50, or 0.75), but when ribs are closely spaced, the number is usually an integer. The second method is to count the number of secondary ribs between a given number of primary ribs, and divide the former by the latter. In practice, the number of secondary ribs is usually regarded as the number of ventral ribs, although this actually includes primary ribs. The third method is to count the number of secondary ribs into which a primary rib splits and the number of secondary ribs intercalated between consecutive primary ribs. These three methods yield slightly different information.

In addition to ribs, the ornamentation includes tubercles, which are small conical swellings. They may develop into bullae (elongated in a radial direction), nodes (swollen and blunt), clavi (elongated in a spiral direction), or spines (elongated and pointed) (for an illustration of these ornamental features, see Arkell et al., 1957: L90, fig. 133). Bullae are sometimes difficult to distinguish from raised ribs. Umbilicolateral tubercles occur near the umbilicolateral margin and ventrolateral tubercles occur near the ventrolateral margin (fig. 14). We counted the number of both types of tubercles on the phragmocone and body chamber of the adult shell. We recorded the distance between tubercles, as measured between tubercle crests, following the curvature of the shell on the umbilicolateral margin for umbilicolateral tubercles, and on the ventrolateral margin for ventrolateral tubercles. The height of a tubercle is measured from its base to its tip, although this measurement is usually an underestimate due to imperfect preservation. We also recorded the number of ribs that join a tubercle dorsally, the number of ribs that branch from it ventrally, and the number of ribs that intercalate between consecutive tubercles.

Septal approximation was measured as described in Landman and Waage (1993). We introduced three refinements to their method. (1) In each specimen studied, we measured the interseptal distances two or three times to determine the measurement error. This error was usually less than 0.3 mm, representing less than 5% of the value of the interseptal distance. (2) We measured the interseptal distances in several specimens representing juvenile and submature stages in order to determine the pattern of change during “normal” growth, and thus, by comparison, to give a better idea of septal approximation at maturity. In general, in the early ontogeny of these scaphites, the septa are evenly spaced (in contrast, septal crowding is common in the early ontogeny of Carboniferous ammonoids, see Kraft et al., 2008). Our results indicate that during “normal” growth, the change in interseptal distance between successive chambers is ±1 mm. Therefore, septal approximation at maturity is defined at the point at which the reduction in interseptal distance is >1 mm. (3) The variation in the magnitude of septal approximation is the ratio, expressed as a percentage, of the interseptal distance of the last chamber to that of the last “normal” chamber. Values between 30% and 40% indicate strong approximation, values around 50% indicate moderate approximation, and values between 60% and 80% indicate weak approximation.

Sutures were drawn using a camera lucida. In general, the last suture or next to last suture is illustrated. Most of our discussion about sutures refers to the shape of the first lateral saddle and lobe. The suture terminology is that of Wedekind (1916), as reviewed by Kullmann and Wiedmann (1970). However, as Korn et al. (2003) have pointed out, this terminology requires revision, based on their study of the evolution of the suture line in prolecanitids.

In addition to describing adult morphology, we also described the ontogenetic development of these scaphites. The methods are reviewed by Landman and Waage (1993) and Landman and Cobban (2007). In general, adult shells were incrementally broken down to expose earlier and earlier whorls. Specimens were photographed at each stage of this process to document the ontogenetic change in whorl shape and ornamentation. In addition, we prepared dorsoventral cross sections perpendicular to the plane of symmetry and passing through the axis of coiling. We measured the ontogenetic change in the degree of whorl compression, W/H; the relative umbilical diameter, UD/LMAX; and the whorl expansion rate, (LMAX_θ/LMAX_θ-π)², where θ is the angle at any given diameter.

Photographs of adult shells are natural size unless otherwise indicated. Small tick marks on the photos mark the base of the body chamber, where visible. The base of the body chamber is defined as the position of the median saddle in the ventral lobe. Specimens are photographed in lateral, ventral, and apertural views (fig. 14B). In lateral view, specimens are photographed with the aperture near the bottom, approximating their position in life.

Shell Morphology

The ammonitella consists of the initial chamber and approximately two-thirds of a whorl (figs. 15, 16). In two specimens of Hoploscaphites brevis (AMNH 58522 and 58523), the ammonitella diameter is 678 µm and 742 µm and the ammonitella angle is 281° and 302°, respectively. These values are similar to those reported for scaphites from the Fox Hills Formation by Landman and Waage (1993). The outer whorls of the ammonitella are covered with a tuberculate microornamentation (fig. 15). Such tubercles were first observed in scaphites by Smith (1905: 642, fig. 1.13) in what he called S. nodosus var. brevis, and have since been documented in nearly all Mesozoic ammonites (Landman et al., 1996). The ammonitella is prismatic in microstructure except for the primary varix, which consists of nacre. Internally, the first septum (proseptum) in scaphites bears a necklike attachment (fig. 16), as previously described in Landman and Bandel (1985) and Landman and Waage (1993).

Fig. 15.

Embryonic whorls of Hoploscaphites nodosus (Owen, 1852) and H. brevis (Meek, 1876), AMNH loc. 3274, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. A–C. Hoploscaphites brevis (Meek, 1876), AMNH 58539. A. The embryonic shell is covered with tubercles. B. Most of the tubercles are irregularly distributed but some are aligned in longitudinal rows. C. The tubercles vary in diameter and range from 5–10 µm. D–F. Hoploscaphites nodosus (Owen, 1852), AMNH 66259. D. Ventral surface of the embryonic shell just adapical of the primary constriction. The surface is covered with a thin dorsal layer and retains the outline of the dorsal sutures of the succeeding whorl. E. Muscle scar in the dorsal lobe. F. The muscle scar is demarcated by a fine-grained texture.

Fig. 16.

Embryonic whorls of Hoploscaphites brevis (Meek, 1876), AMNH loc. 3274, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. A–C. Median cross section, AMNH 58522. A. The embryonic shell ends in the primary varix (PV). The ammonitella diameter is 742 µm and the ammonitella angle is 281°. B. Close-up of the primary varix (PV). The arrow here and in C and D indicates the adoral direction. C. The postembryonic shell emerges from beneath the primary varix (PV) and consists of an outer prismatic (OP) and inner nacreous layer (NA). D–E. Median cross section, AMNH 58531. D. The proseptum (1), its necklike attachment (N), and the flange (F) are visible. E. Close-up of the dorsal portion of the proseptum (1), its necklike attachment (N), and the flange (F). F. Close-up of the ventral portion of the proseptum (1) and its necklike attachment (N).

The juvenile shell is planispirally coiled with a small umbilicus. The whorl section is initially depressed and becomes more compressed through early ontogeny. This change is similar to that in many other ammonites and may correlate with changes in habitat and/or swimming ability (Jacobs, 1992). The ratio of umbilical diameter to shell diameter decreases steeply in early ontogeny until a shell diameter of approximately 10–20 mm, after which the ratio remains nearly the same until the beginning of the mature body chamber. The rate of whorl expansion in juveniles is approximately 2.60.

It is difficult to determine the angle of the body chamber in juveniles because the apertural margin is very thin and almost never preserved intact. We measured an angle of 217.5° in a juvenile of Hoploscaphites brevis (D  =  25.1 mm), and an angle of 202° in a juvenile of H. nodosus (D  =  37.5 mm), although the apertural margin is broken in both of these specimens (AMNH 58559 and 58560, respectively). In a study of 37 well-preserved juveniles of Hoploscaphites comprimus (Owen, 1852) ranging from 1.8 to 9.6 mm in phragmocone diameter, Landman and Waage (1993) reported an average body chamber angle of 237° with a range from 220° to 264°. Based on all these measurements, we estimate that the body chamber angle in juveniles of H. nodosus and H. brevis is nearly two-thirds of a whorl (230°), which is referred to as mesodomic in the terminology of Westermann (1996). In contrast, Landman (1987) documented a higher body chamber angle in juveniles of Scaphites preventricosus and Clisoscaphites vermiformis. In a sample of 31 juveniles of C. vermiformis and 9 juveniles of S. preventricosus, the body chamber angle averages 266° and ranges from 235° to 302°.

The ornament consists of ribs and umbilicolateral and ventrolateral tubercles (fig. 17). The ribs represent flexures in the shell wall and are, therefore, expressed on the steinkern. However, they are sharper on the actual shell. Ribs beome slightly more closely spaced in passing from the adapical to the adoral part of the exposed phragmocone (fig. 18). Thereafter, they become slightly more widely spaced on the midshaft and slightly more closely spaced again on the hook. Ribs are rectiradiate to slightly rursiradiate on the adapical portion of the shaft, becoming more prorsiradiate on the adoral portion of the shaft. They bend backward on the inner flanks, forward on the midflanks, back again on the outer flanks, and weakly forward at the ventrolateral margin. Ribs become straight or slightly concave near the point of recurvature. Branching and intercalation occur at the umbilical shoulder, on the inner one-third of the flanks at the site of the umbilicolateral tubercles, if they are present, and on the outer one-third of the flanks at the site of the ventrolateral tubercles. Ribs cross the venter of the shaft with a slight forward projection, which weakens on the hook.

Fig. 17.

Details of the ornamentation in Hoploscaphites nodosus (Owen, 1852). The venter is toward the top in all views except for G in which it is toward the bottom. A–C. BHI 4216, macroconch, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. A. Umbilicolateral tubercle on the right side of the exposed phragmocone. One rib joins it dorsally and three ribs branch from it ventrally. B. Ventrolateral tubercle on the right side of the exposed phragmocone. Two ribs join it dorsally and three ribs branch from it ventrally. C. Ventrolateral tubercle on the right side of the exposed phragmocone. One rib joins it dorsally and three ribs branch from it ventrally. D, E. BHI 4178, macroconch, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. D. Umbilicolateral tubercle on the right side of the exposed phragmocone. The tubercle occurs in the interspace between ribs, with one rib joining it dorsally and three ribs branching from it ventrally. E. Ventrolateral tubercle on the left side of the exposed phragmocone. The tubercle occurs in the interspace between ribs, with one rib joining it dorsally and three ribs branching from it ventrally. F. BHI 4121, macroconch, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. Umbilicolateral and ventrolateral tubercles on the left side of the exposed phragmocone. G. AMNH 9520/3, microconch, Pierre Shale, Belle Fourche River, South Dakota. Umbilicolateral tubercle on the left side of the body chamber, with a steeply dipping adapical side on the right and a more gently sloping adoral side on the left. H–I. AMNH 45343, macroconch, AMNH loc. 3212, Baculites compressus–B. cuneatus zones, Pierre Shale, Shannon County, South Dakota. H. Ventrolateral tubercles on the right side of the exposed phragmocone. The tips of the tubercles are broken revealing that the tubercles are hollow. I. Ventrolateral tubercle on the left side of the exposed phragmocone broken in half and filled with matrix.

Fig. 18.

Spacing of ribs on the venter starting from near the adoral end of the phragmocone (bottom) to the aperture (top). Ribs become more widely spaced on the midshaft and more closely spaced on the hook. A. Hoploscaphites nodosus (Owen, 1852), macroconch, BHI 4257, Baculites cuneatus Zone, Pierre Shale, Meade County, South Dakota. B. Hoploscaphites nodosus (Owen, 1852), macroconch, AMNH 9520/2, Pierre Shale, Sage Creek, Pennington County, South Dakota. C. Hoploscaphites brevis (Meek, 1876), macroconch, holotype, USNM 367, Pierre Shale, probably from the Sage Creek area, Pennington County, South Dakota. D. Hoploscaphites brevis (Meek, 1876), macroconch, YPM 35576, YPM loc. A6520, Pierre Shale, Sage Creek, Pennington County, South Dakota. E. Hoploscaphites brevis (Meek, 1876), macroconch, BHI 4235, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. F. Hoploscaphites brevis (Meek, 1876), macroconch, YPM 35584, YPM loc. A6520, Pierre Shale, Sage Creek, Pennington County, South Dakota.

The tubercles are hollow (nonfloored) and are, therefore, expressed on the steinkern. The umbilicolateral tubercles usually appear midway on the exposed phragmocone whereas the ventrolateral tubercles usually appear in early ontogeny. The umbilicolateral and ventrolateral tubercles increase in size in passing from the phragmocone to the body chamber. One of the distinctive features of the scaphites from the Baculites compressus–B. cuneatus zones is the distribution of the ventrolateral tubercles on the exposed phragmocone. The tubercles in this area are widely and irregularly spaced and commonly occur in pairs or clusters (fig. 19). This contrasts, for example, with Hoploscaphites crassus (Coryell and Salmon, 1934) from the B. baculus Zone in which the tubercles are closely and regularly spaced.

Fig. 19.

Spacing of ventrolateral tubercles on the adult shells of Hoploscaphites nodosus (Owen, 1852) and H. brevis (Meek, 1876). A. Hoploscaphites brevis (Meek, 1876), large macroconch, BHI 4790, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. B. Hoploscaphites nodosus (Owen, 1852), large microconch, BHI 4222, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. C. Hoploscaphites nodosus (Owen, 1852), large macroconch, BHI 4216, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. D. Hoploscaphites brevis (Meek, 1876), small macroconch, BHI 4292, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. The photo in D has been flipped so that it faces the same way as the others. Symbols: •  =  tubercle on the phragmocone; o  =  tubercle on the body chamber.

The ventrolateral tubercles attain their maximum size at midshaft. In Hoploscaphites nodosus, these tubercles develop into subspinose clavi, with a flattened, steep adapical face and a more gently sloping adoral face. This implies that during tubercle formation, the apertural lip flared backward and outward along the ventrolateral margin. Although the pattern of growth lines is not visible on any of these tubercles, the mode of tubercle formation may have been similar to that in Aspidoceras, as described by Checa and Martin-Ramos (1989: fig. 4A–C). Starting on the adoral end of the shaft, the ventrolateral tubercles abruptly diminish in size, much more so than the umbilicolateral tubercles. In general, there is an increase in rib density at the sites of the umbilicolateral and ventrolateral tubercles. Several ribs join an umbilicolateral or ventrolateral tubercle dorsally, and an additional number of ribs branch from it ventrally. Occasionally, a tubercle occurs in the interspace between ribs.

Muscle attachment areas are indicated by thin layers of myostracum on steinkerns of the body chambers (fig. 20). The ventral muscle attachment area is ovoid and appears several millimeters adoral of the midventral saddle. It is 1–5 mm in length, depending on the size of the adult shell. The dorsal muscle attachment area is wedge shaped and drapes over the umbilical shoulder on each side of the shell just adoral of the last suture.

Fig. 20.

Muscle scars and other features on the adult shells of Hoploscaphites nodosus (Owen, 1852) and H. brevis (Meek, 1876). A, B. Unpaired ventral muscle scar (arrows) just adoral of the last suture. A. Hoploscaphites nodosus (Owen, 1852), macroconch, BHI 4948, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. B. Hoploscaphites nodosus (Owen, 1852), microconch, BHI 4949, Baculites compressus Zone, Bearpaw Shale, Rosebud County, Montana. C. Dorsal muscle scar (arrows) on the umbilical shoulder on the right side of the body chamber just adoral of the last suture, Hoploscaphites brevis (Meek, 1876), microconch, BHI 4283, Baculites cuneatus Zone, Pierre Shale, Meade County, South Dakota. D. Septal approximation (arrow), Hoploscaphites nodosus (Owen, 1852), macroconch, BHI 4948, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. E. Absence of septal approximation (arrow), Hoploscaphites nodosus (Owen, 1852), macroconch, BHI 4215, Baculites cuneatus Zone, Pierre Shale, Meade County, South Dakota. F. Trace of ventral muscle scar (arrow) on the phragmocone, Hoploscaphites brevis (Meek, 1876), BHI 4393, microconch, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. G, H. Constriction (arrows) at the apertural margin, Hoploscaphites brevis, macroconch, BHI 5130, Baculites compressus–B. cuneatus zones, Pierre Shale, Pennington County, South Dakota. G. Ventral view; H. left lateral view.

The sutures in Hoploscaphites nodosus and H. brevis do not differ significantly from those of other Campanian and Maastrichtian species of Hoploscaphites (fig. 21). The ventral lobe (E) is broad and contains a median saddle. The first lateral sadde (E/L) is broad and asymmetrically bifid. The first lateral lobe (L) is symmetrically to asymmetrically bifid, owing to the relative enlargement of its inner branch.

Fig. 21.

Sutures in Hoploscaphites nodosus (Owen, 1852) and H. brevis (Meek, 1876). A. Hoploscaphites brevis (Meek, 1876), macroconch, third from last suture, BHI 7034, Pierre Shale, Baculites cuneatus Zone, Pennington County, South Dakota. B. Hoploscaphites brevis (Meek, 1876), macroconch, next to last suture, BHI 4205, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. C. Hoploscaphites nodosus (Owen, 1852), microconch, fourth from last suture, USNM 536232, USGS Mesozoic loc. 21224, Bearpaw Shale, Big Horn County, Montana. D. Hoploscaphites nodosus (Owen, 1852), microconch, second from last suture, BHI 4122, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. E. Hoploscaphites brevis (Meek, 1876), microconch, third from last suture, USNM 536262, Pierre Shale, Pennington County, South Dakota. F. Hoploscaphites brevis (Meek, 1876), microconch, third from last suture, BHI 4255, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. Abbreviations: x  =  tubercle; E  =  ventral lobe; E/L  =  first lateral saddle between ventral and lateral lobes; L  =  lateral lobe.

Scaphites, like many other ammonites, exhibit determinate growth (Bucher et al., 1996). Seilacher and Gunji (1993: 241) have described the phenomenon of “mophogenetic countdown,” which they defined as “an irrevocable sequence, or program, of morphogenetic events that proceeds step by step and terminates with growth cessation as the final command.” This sequence reflects the growth of the reproductive organs and the attainment of maturity. It is also expressed in the morphology of the shell. For example, in modern nautilus, the disappeance of the color bands on the venter indicates the beginning of the countdown. It also coincides with a reduction in septal spacing and an increase in septal thickness (Collins and Ward, 1987). Landman (1989) discussed this phenomenon in Pteroscaphites from the Turonian-Santonian of the U.S. Western Interior. The countdown began in these scaphites at an unusually small size, initiating a set of mature modifications. However, because of the small size of these scaphites, their mature modifications closely reflect the morphology of the juvenile shell and, therefore, differ from the mature modifications of closely related, but larger species.

The morphogenetic countdown begins in scaphites at the point at which the shell departs from the spiral coil and develops into the shaft. This marks the beginning of an essentially irreversible process, unless a fatal injury intervenes, culminating in the formation of the mature shell. The shaft passes into the hook, which recurves backward forming an apertural angle of 35°–85°, depending on the species. The apertural opening is reduced relative to the whorl section of the hook, due to a constriction in the shell (fig. 22). The shell wall bends sharply inward, at nearly a right angle, forming a thick varix, and then reflects outward, terminating in a flared lip. The formation of a varix is generally interpreted as reflecting a pause in growth (Bucher et al., 1996). Sediments commonly accumulate in the narrow gap between the flared lip and the rest of the shell wall. The constriction and varix disappear on the dorsum where the apertural margin forms an elongate, broadly rounded projection (fig. 22). In steinkerns, the constriction and varix in the shell wall appear as a constriction at the apertural margin.

Fig. 22.

Morphology of the apertural margin. A, B. Cross section of the shell wall at the apertural margin of a mature macroconch of Hoploscaphites plenus (Meek, 1876), BHI 4807, Bearpaw Shale, Garfield County, Montana. A. Sketch showing the location of the cross section (rectangle). B. Close up of the apertural margin. The exterior direction is toward the right (note the outline of the ribs) and the adoral direction is toward the top. As the shell approaches the apertural margin, it bends sharply inward, at nearly a right angle, forming a thick varix, and then reflects outward, terminating in a flared lip. Sediments commonly accumulate in the narrow gap behind the flared lip. C–E. Hoploscaphites nodosus (Owen, 1852), large macroconch, AMNH 58564, AMNH loc. 3274, Pierre Shale, Meade County, South Dakota. C. Apertural view showing the kidney-shaped aperture filled with dark matrix. D. Apertural view, slightly rotated, showing the broadly rounded projection on the dorsal side. E. Left lateral view.

In the 19th century, it was not universally accepted that the formation of an uncoiled body chamber implied a cessation of growth. Hyatt (1894: 613) argued that the uncoiled body chamber in scaphites and other heteromorph ammonites was resorbed during ontogeny. He based this on the fact that he observed small shells with uncoiled body chambers, which he interpreted as belonging to the same species as large shells with similarly uncoiled body chambers, thus concluding that the uncoiled body chamber must have been resorbed and reformed several times during ontogeny. However, it is much more likely that these small and large shells simply represented adults of different species or dimorphs of the same species. In contrast to Hyatt, Pompeckj (1894) stated that he did not find any evidence for resorption in his study of ammonites with “abnormal” body chambers. Smith (1905), based on his ontogenetic study of scaphites, agreed with Pompeckj's conclusion, and reported that uncoiled body chambers never appeared at early ontogenetic stages. However, like Hyatt, Smith argued that the presence of an uncoiled body chamber indicated that scaphites showed “degeneration or senility.”

The initiation of the morphogenetic countdown in scaphites, as indicated on the shell by the beginning of the shaft, corresponds internally to the secretion of the septum approximately two-thirds of a whorl back, based on our observations of the angular length of the body chamber in juveniles (see above). Therefore, the last seven or so septa must have been secreted during the formation of the adult body chamber. Of these, the last two or three septa correspond to the formation of the hook, culminating in the development of the constriction and varix at the apertural margin. Commonly, these last few septa are more closely spaced, which is generally interpreted as an indication of a decrease in the rate of growth, beginning at the onset of maturation (Bucher et al., 1996).

Scaphites occur as dimorphs, which are referred to as macroconchs (M) and microconchs (m). They are interpreted as sexual in nature, the macroconch being the female, and the microconch being the male, following the traditional view as outlined by Lehmann (1981) and Davis et al. (1996). The dimorphs in scaphites are distinguished by differences in size and morphology (Makowski, 1962; Cobban, 1969; Landman and Waage, 1993; Machalski, 2005a, 2005b; but see Kin, 2009, for a counterargument, and Machalski, in press, for a rebuttal). Dimorphism is more marked in Campanian-Maastrichtian scaphites such as Hoploscaphites than in older scaphites such as Clioscaphites. In general, dimorphism appears only in adult shells. Juvenile shells of conspecific dimorphs are nearly identical, except in rare instances. For example, Landman and Waage (1993: fig. 67) noted that in Hoploscaphites nicolletii, the ventral ribs starting at approximately 10 mm shell diameter, are coarser and stubbier in microconchs than in macroconchs. In scaphites from the Baculites compressus–B. cuneatus zones, the dimorphs are distinguished at maturity by (1) the presence or absence of an umbilical bulge, (2) the outline of the umbilical shoulder relative to that of the venter in side view, (3) the umbilical diameter, (4) the change in whorl height from the mature phragmocone to the midshaft of the body chamber (H₄/H₂), (5) the change in whorl height from the midshaft of the body chamber to the hook (H₄/H₆), (6) the change in the degree of whorl compression from the point of recurvature to the aperture, and (7) the adult size. Some of these criteria are more reliable than others in distinguishing between dimorphs.

The presence of a bulge on the umbilical shoulder indicates that the specimen is a macroconch. However, the absence of a bulge is equivocal. Although this feature is invariably absent in microconchs, it is also absent in many specimens that are identified as macroconchs on the basis of other criteria. The umbilical bulge appears on the adapical end of the shaft and extends adorally for several millimeters. In some specimens, the umbilical bulge is so large that it occludes part of the umbilicus. The significance of the bulge is unknown, but it may be related to the position of the reproductive organs, in analogy with the nidamental gland in present-day nautilus.

The outline of the umbilical shoulder of the body chamber relative to that of the venter in side view is usually a reliable means of distinguishing between dimorphs. In general, the umbilical shoulder of the body chamber is straight in macroconchs whereas it is curved in microconchs, matching the curvature of the venter. However, there is some variation in this feature, especially in small, compressed specimens of Hoploscaphites brevis. The umbilical shoulder is straight in these specimens, suggesting that they are macroconchs, but other features of the shell, including a gradual rather than abrupt increase in whorl height in passing from the phragmocone to the midshaft, suggest that they are microconchs instead (listed as “microconchs, transitional to macroconchs,” in table 9). In a few instances, these specimens appear as macroconchs on the right side and microconchs on the left side, although the reason for this disparity is unknown.

Landman and Waage (1993: 90, fig. 48L–P) documented a similar phenomenon (straight umbilical shoulder and a gradual increase in whorl height) in small specimens of Hoploscaphites nicolletii. However, they interpreted these specimens as macroconchs rather than microconchs because the umbilical shoulder of the body chamber is sharply rather than broadly rounded, which is more characteristic of macroconchs. Examples of other ammonites showing a mixture of features of both dimorphs have been reported from the Jurassic of Europe. Parent et al. (2008: 186) noted this phenomenon in aspidoceratids and speculated that it may be due to “a weak sexual determination (hermaphroditism?) or a sex change in the subadult stage.” These authors also cited an example of a perisphinctid that was identified as a macroconch on the basis of its ornament in the subadult stage, but which developed an apertural lappet at maturity, suggesting to them that the specimen had changed into a microconch during ontogeny. However, in modern cephalopods, hermaphroditism is unknown (Hanlon and Messenger, 1996).

In general, the umbilical diameter is larger in microconchs than in macroconchs. For example, a plot of UD_P versus LMAX in Hoploscaphites nodosus reveals a cluster of microconchs on the upper left characterized by high values of UD_P and low values of LMAX, and a cluster of macroconchs on the lower right characterized by lower values of UD_P and higher values of LMAX (fig. 23A). The umbilical diameter measured through the center of the umbilicus passing through the position of the last septum (UD_P) is a more reliable means of distinguishing between dimorphs than the umbilical diameter measured through the center of the umbilicus parallel to the line of maximum length (UD_L). The reason is that UD_P takes into account the difference in the curvature of the umbilical shoulder, provided that the base of the body chamber occurs below the line of maximum length. For example, in H. nodosus, the average value of UD_P is significantly higher in microconchs (7.2 mm) than in macroconchs (4.1 mm), whereas the average value of UD_L is nearly the same in both dimorphs (tables 3 and 5). Similary, a plot of UD_P versus UD_L in this species mostly differentiates the two dimorphs based on differences in UD_P (fig. 23B).

Fig. 23.

A. Plot of LMAX (mm) versus UD_P (mm) in Hoploscaphites nodosus (Owen, 1852) reveals a cluster of microconchs on the upper left and a cluster of macroconchs on the lower right. B. Plot of UD_P (mm) versus UD_L (mm) in Hoploscaphites nodosus (Owen, 1852) distinguishes the two dimorphs mainly on the basis of UD_P.

In addition to the curvature of the umbilical shoulder and the size of the umbilical diameter, dimorphs differ in the shape of their shells. The whorl height of the body chamber increases more markedly in macroconchs than in microconchs. This difference is expressed in the ratio of whorl height at midshaft to whorl height on the phragmocone along the line of maximum length (H₄/H₂) and by the ratio of whorl height at midshaft to whorl height at the point of recurvature (H₄/H₆). In our sample of Hoploscaphites nodosus, the average values of H₄/H₂ and H₄/H₆ are significantly higher in macroconchs (1.32 and 1.06, respectively) than in microconchs (1.21 and 0.99, respectively) (tables 3 and 5). In a plot of H₄/H₆ versus H₄/H₂, the two dimorphs appear as distinct clusters with some overlap between them (fig. 24A). Similarly, in our sample of H. brevis, the average values of H₄/H₂ and H₄/H₆ are significantly higher in macroconchs (1.37 and 1.18, respectively) than in microconchs (1.30 and 1.07, respectively) (tables 7 and 9). In a plot of H₄/H₆ versus H₄/H₂ in this species, the two dimorphs also appear as distinct clusters (fig. 24B). However, there is more overlap between them because of the existence of “transitional” specimens.

Fig. 24.

A. Plot of H₄/H₂ versus H₄/H₆ in Hoploscaphites nodosus (Owen, 1852). B. Plot of H₄/H₂ versus H₄/H₆ in H. brevis (Meek, 1876). In each species, the two dimorphs appear as distinct clusters with some overlap between them.

In both dimorphs, the shell becomes more compressed in passing from the end of the phragmocone to the body chamber (fig. 25). The ratio of whorl width to whorl height reaches a minimum value at midshaft. The whorl section becomes more depressed in passing into the hook, but the two dimorphs diverge thereafter. In macroconchs, the whorl section maintains the same degree of depression or becomes slightly more depressed toward the aperture, whereas in microconchs, the whorl section becomes less depressed. The fact that the whorl section at the aperture is more depressed in macroconchs than in microconchs may be related to differences in the pattern of sexual maturation. It is possible that the more depressed aperture in macroconchs reflects the development of the ovaries and the formation of eggs.

Fig. 25.

Plot of the ratio of whorl width to height (W/H) at seven points on the adult shell of Hoploscaphites nodosus (Owen, 1852) and H. brevis (Meek, 1876) from ¼ whorl adapical of the long axis (1) to the aperture (7), as shown in figure 14. The shell attains the maximum degree of compression near midshaft (points 3 and 4). A–D. Macroconchs. A. H. nodosus, AMNH 9520/2, Pierre Shale, Pennington County, South Dakota. B. H. nodosus, BHI 4215, Baculites cuneatus Zone, Pierre Shale, Meade County, South Dakota. C. H. brevis, BHI 4124, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. D. H. brevis, BHI 4230, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. E–H. Microconchs. E. H. nodosus, AMNH 58521, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. F. H. nodosus, YPM 35631, YPM loc. A1461, Bearpaw Shale, Rosebud County, Montana. G. H. brevis, AMNH 55884, AMNH loc. 3415, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. H. H. brevis, YPM 1861,YPM loc. A6520, Pierre Shale, Sage Creek, Pennington County, South Dakota.

As in other scaphites, dimorphs differ in the size of the adult shell. Macroconchs attain a larger size than microconchs but the size ranges of the two dimorphs overlap. The extent of size overlap depends upon the number of specimens in the sample, with larger samples showing a greater degree of overlap. The extent of size overlap also depends on the geographic and temporal range of the sample. For example, in our sample of Hoploscaphites nodosus from the Baculites compressus–B. cuneatus zones, the extent of size overlap is 44% of the total combined size range of both dimorphs. However, this sample contains an unusually small macroconch. If this specimen is excluded, the extent of size overlap decreases to 22%.

It is perhaps of greater biological interest to determine the size difference between dimorphs within the same species that lived in the same area at the same time and would have potentially mated with each other. The best approximation of these conditions is obtained in a concretion in which both dimorphs are preserved together. As a reminder, however, the cooccurrence of dimorphs in the same concretion does not guarantee that they lived together, only that they were preserved together. A concretion from Pennington County, South Dakota, contains both dimorphs of Hoploscaphites brevis preserved side by side (YPM 35581 and 35582). The macroconch is 15% larger than the microconch. Concretion S from AMNH loc. 3274 from Meade County, South Dakota, also contains both dimorphs of H. brevis (figs. 26, 27). The average size of two macroconchs is 23% larger than that of two microconchs. We therefore conclude that, on average, the macroconch is 20% larger than the conspecific microconch (“the 20% size rule”).

Fig. 26.

Hoploscaphites brevis (Meek, 1876) from a single concretion (Concretion S), AMNH loc. 3274, Pierre Shale, Baculites compressus–B. cuneatus zones, Meade County, South Dakota. A–C. Small microconch, AMNH 58518. A. Right lateral; B. apertural; C. left lateral. D, E. Small macroconch, AMNH 58517. D. Apertural; E. left lateral. F–I. Small macroconch, AMNH 58515. F. Right lateral; G. apertural; H. ventral; I. left lateral. Specimens are illustrated natural size.

Fig. 27.

Hoploscaphites brevis (Meek, 1876) from a single concretion (Concretion S), AMNH loc. 3274, Pierre Shale, Baculites compressus–B. cuneatus zones, Meade County, South Dakota (continuation of fig. 26), containing a mixture of macroconchs and microconchs. A–D. Small microconch, AMNH 58514. A. Right lateral; B. apertural; C. ventral; D. left lateral. E–F. Large, coarsely ornamented macroconch, AMNH 58516. Note the small scaphite in the aperture. E. Right lateral; F. apertural. Specimens are illustrated natural size.

Closer spacing or approximation of the last few septa is common in both dimorphs (fig. 28). We examined the pattern of septal approximation in 13 specimens each of Hoploscaphites nodosus and H. brevis (4 macroconchs and 9 microconchs in H. nodosus, and five macroconchs and eight microconchs in H. brevis) (table 1). In H. nodosus, the pattern of septal approximation is the same in both dimorphs. Septal approximation either does not occur or occurs over only one or two chambers. The magnitude of approximation is more or less the same in both dimorphs and ranges broadly from 23% to 73%. In contrast, septal approximation occurs in nearly all specimens of H. brevis and the pattern of septal approximation differs between dimorphs. In macroconchs, septal approximation occurs over two or three chambers and the magnitude of approximation is strong, ranging from 33% to 38%. In microconchs, approximation usually occurs over only one or two chambers and the magnitude of approximation is weak, ranging from 57% to 86%.

Fig. 28.

Septal approximation at maturity in Hoploscaphites nodosus (Owen, 1852) and H. brevis (Meek, 1876). A, B. H. nodosus, macroconchs. A. YPM 35598, YPM loc. C3347, Pierre Shale, Belle Fourche River, South Dakota. Septal approximation occurs over one chamber and is relatively weak (the interseptal distance of the last chamber is 65% that of the last “normal” chamber). This specimen also exhibits a marked pathology due to an injury, which probably explains the large variation in septal spacing. B. BHI 4215, Baculites cuneatus Zone, Pierre Shale, Meade County, South Dakota. No septal approximation is apparent, but the base of the body chamber is adapical of the line of maximum length. C, D. H. nodosus, microconchs. C. BHI 4122, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. No septal approximation is apparent. Note that the values of the interseptal distances in successive chambers are within ± 1 mm of each other. This specimen also exhibits an injury. D. BHI 4895, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. Septal approximation occurs over one chamber and is weak (the interseptal distance of the last chamber is 63% that of the last “normal” chamber). E. H. brevis, macroconch, BHI 4202, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. Septal approximation occurs over two chambers and is strong (the interseptal distance of the last chamber is 33% that of the last “normal” chamber). F. H. brevis, macroconch, USNM 536262, Pierre Shale, Pennington County, South Dakota. Septal approximation occurs over one chamber and is weak (the interseptal distance of the last chamber is 83% that of the last “normal” chamber). Measurements are listed in table 1.

Landman and Waage (1993) observed the pattern of septal approximation in species of Hoploscaphites from the Fox Hills Formation, including H. nicolletii (Morton, 1842) and H. spedeni (Landman and Waage, 1993). They discovered that in H. nicolletii, septal approximation tends to occur over more chambers in macroconchs than in microconchs. In macroconchs, septal approximation generally occurs over the last two chambers, whereas in microconchs, it occurs over only the last chamber. In contrast, there is no difference in the pattern of septal approximation between dimorphs in H. spedeni. Like H. nodosus, H. spedeni is a robust species with an inflated whorl section. It is possible that septal approximation is more strongly manifested in specimens with compressed whorl sections, like H. brevis and H. nicolletii. In such specimens, changes in cameral volume more directly translate into differences in interseptal distance.

Dimorphs also differ in the ratio of the volume of the phragmocone to the volume of the body chamber at maturity. Neige (1996) investigated this relationship in dimorphs of Hoploscaphites nicolletii. He dissected adult shells to expose earlier stages in order to calculate this ratio through ontogeny. He assumed a body chamber angle of 230°, based on the average value in 37 juveniles of the closely related species H. comprimus as documented by Landman and Waage (1993: fig. 25). At each stage, he measured the volume of the phragmocone and the volume of the body chamber and calculated the ratio between them. In juveniles of both dimorphs, the ratio is constant throughout ontogeny (0.21). However, at maturity, the two dimorphs diverge. In microconchs, the ratio remains the same whereas in macroconchs it increases (0.39). Neige assumed that both dimorphs maintained neutral buoyancy throughout their lifetime and, therefore, the higher ratio in macroconchs did not indicate an increase in buoyancy at maturity. He argued instead that the higher ratio in macroconchs may have served to compensate for (1) the retention of liquid in some of the last few chambers of the phragmocone, (2) an increase in the density of tissues at maturity, for example, related to the formation of the enigmatic “hooks” in macroconchs as described by Landman and Waage (1993: figs. 43–46), or (3) an increase in the volume of tissues at maturity, which protruded outside the aperture.

Jaws

Scaphite jaws occur in the Baculites compressus–B. cuneatus zones, but generally are not as well preserved as those in the B. baculus Zone in the Pierre Shale near Glendive, Montana (Grier and Grier, personal commun., 2003), nor in the Hoploscaphites nicolletii and H. nebrascensis zones in the Fox Hills Formation in north-central South Dakota (Landman and Waage, 1993). These differences may reflect differences in the rate of burial of the ammonites after death. In all of these deposits, the lower jaw is more common than the upper jaw, probably due to the larger size and bulkier shape of the lower jaw.

The morphology of the scaphite jaws in the Baculites compressus–B. cuneatus zones is similar to that of scaphite jaws described elsewhere by Landman and Waage (1993) and Kennedy et al. (2002, and references therein). The terminology of the jaws is reviewed by 200Landman et al. (2006, 2007b). The upper jaw is U-shaped and consists of two narrow wings that converge anteriorly to a beaklike apex (fig. 29E). The lower jaw is convex on the ventral side and consists of two mirror-image triangular wings that meet along the plane of bilateral symmetry, called the commissure (also symphysis or central groove). The commissure is bordered on each side by a flange, which increases in height in a posterior direction, reaching its maximum height just before the posterior margin (fig. 29F). The apex of the lower jaw is weakly projected forward, and the anterior margin is slightly concave on either side. The lateral margins are broadly rounded, and the posterior margin is bilobate, with a slight indentation at the commissure (fig. 29H).

Fig. 29.

Jaws of Hoploscaphites. A, B. Lower jaw near the aperture of Hoploscaphites brevis (Meek, 1876), small microconch, BHI 4264, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. A. Ventral view of the lower jaw, splayed out in the aperture, anterior direction of the jaw toward the top; B. left lateral view. C, D. Lower jaw near the aperture of Hoploscaphites brevis (Meek, 1876), small microconch, AMNH 56874, AMNH loc. 3415, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. C. Right lateral; D. lower jaw folded together, folio style, showing a gap at the commissure, anterior direction of the jaw toward the top. E. Isolated upper jaw attributed to Hoploscaphites, anterior view, AMNH 56805, Kathy's concretion, AMNH loc. 3274, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. F. Left lateral view of a lower jaw associated with Hoploscaphites brevis (Meek, 1876), anterior direction of the jaw toward the top, BHI 5795, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. G. Left lateral view of an isolated lower jaw attributed to Hoploscaphites, with postmortem damage including a puncture (arrow), anterior direction of the jaw toward the top, AMNH 56866, AMNH loc. 3408, Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. H, I. Hoploscaphites brevis (Meek, 1876), small microconch, GSC 5342c, “?Demaine Member, South Saskatchewan River opposite mouth Swift Current Creek, Sask.” ( =  paratype, Hoploscaphites landesi Riccardi, 1983, pl. 1, figs. 20–22). H. Right lateral; I. close up of the left side of the lower jaw in the aperture, anterior direction of the jaw toward the bottom. J. Dorsal (inner) surface of the aptychus of an isolated lower jaw attributed to Hoploscaphites, anterior direction of the jaw toward the top, AMNH 56883, AMNH loc. 3274, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. K. Isolated lower jaw attributed to Hoploscaphites, anterior direction of the jaw toward the top, with postmortem damage including punctures (arrows), AMNH 56846, Kathy's concretion, AMNH loc. 3274, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota.

Each wing of the lower jaw is composed of two layers. The inner layer consists of coarsely crystalline black material, which presumably represents diagenetically altered chitin (Gupta et al., 2008). The ventral surface of this layer is ornamented with radial striations, and lirae and growth lines that parallel the posterior margin (fig. 29F, G). The lirae and growth lines slant diagonally forward on the flange. The outer layer consists of calcite and represents the aptychus. This layer is 110 µm thick in AMNH 56883 (fig 29H) and 140 µm thick in AMNH 58563. The ventral surface of the aptychus is not exposed on any of our specimens, but the dorsal surface reflects the ornamentation on the ventral surface of the black layer (fig. 29H). Based on investigations of the lower jaw in other scaphites, the microstructure of the aptychus consists of thin lamellae that parallel the dorsal surface (Kruta et al., 2009).

The chitinous layer of the lower jaw is much more commonly preserved than the calcitic layer. Landman and Waage (1993) noted this same phenomenon in scaphite jaws from the Fox Hills Formation and stated that “The calcitic layer is usually absent on even slightly weathered surfaces, and in freshly exposed specimens it may be partly or entirely absent as a result of removal prior to formation of the enclosing concretion.” Similarly, 200Landman et al. (2006) noted that the calcitic layer is invariably absent on the lower jaws of Placenticeras from North America. However, because such a layer is present on the lower jaws of the closely allied genus Metaplacenticeras from Hokkaido, Japan, these authors hypothesized that the calcitic layer in Placenticeras was diagenetically lost. The rarity of calcitic aptychi in concretions that otherwise contain aragonitic fossils suggests that the calcitic plates were susceptible to mechanical breakage and chemical dissolution during fossilization.

The jaws in the Baculites compressus–B. cuneatus zones usually occur as isolated elements in association with whole or broken scaphites. Landman and Klofak (2006) recorded dozens of jaw fragments in association with broken scaphites and other molluscs in a concretion from the Pierre Shale of Meade County, South Dakota. These jaws show evidence of compaction, warping, folding, tearing, puncturing, and exfoliation (fig. 28 I), indicating that they must have remained on the sea floor for an extended period of time, perhaps months to years. In taphonomic experiments with modern coleoids, which were designed to simulate the conditions following death and before burial, Kear et al. (1995) reported that the jaws of these animals were still more or less intact after a year ( =  the duration of the experiments), although the posterior edges of the jaws had begun to fracture and disintegrate after 10 weeks. The fragmentary nature of the ammonite jaws may also indicate that they were redepoited as faecal matter, having first been consumed and digested by a predator.

Although isolated jaw fragments are abundant, occurrences of jaws inside or closely associated with the body chambers are relatively rare. We have observed only a few specimens, all of which are microconchs of Hoploscaphites brevis, in which the jaws occur just inside or just outside the aperture. In AMNH 56874, the lower jaw is folded together folio style, and is oriented diagonally across the broken aperture (fig. 29C, D). The two wings are separated at the commissure, and it appears that the upper jaw is nestled within the gape of the lower jaw. In BHI 4264, the lower jaw is splayed out butterfly style, exposing the ventral surface and the commisure (fig. 29A, B). The edge of the left wing is folded inward and the edge of the right wing is torn. The jaw is oriented perpendicularly to the broken aperture. In GSC 5342c, the left wing of the lower jaw, including the flange, is almost perfectly preserved, except for minor abrasion along the lateral margin (fig. 29H, I). The jaw covers more than half of the aperture, with the apex of the jaw touching the dorsal projection of the shell.

The function of the calcitic layer in the lower jaw is unknown. It may have served as reinforcement to strengthen the jaw for crushing or biting, but the layer is relatively thin (100–150 µm thick). In addition, the presence of a groove, rather than a thickening, along the middle of the lower jaw would have compromised any structural benefit from a mineralized layer (Landman et al., 2007b). Alternatively, the calcitic layer may have served as an operculum for protection against predators. However, it is unclear how the lower jaw would have rotated into a position to fulfill this function (for a review of the various arguments, see Lehmann and Kulicki, 1990; Lewy, 2000; Engeser and Keupp, 2002; Landman et al., 2007b; and Kruta and Landman, 2008).

Pathologic Specimens

Pathologic specimens are common among the scaphites in the Baculites compressus–B. cuneatus zones (figs. 34, 35). Pathologies have been widely documented in both Paleozoic and Mesozoic ammonites (Holder, 1956; Guex, 1967, 1968; Landman and Waage, 1986; Bond and Saunders, 1989; Kröger, 2002a, 2002b; Larson, 2003, 2007; Keupp, 2006). Many descriptive terms, known as “forma-types,” have been introduced to categorize the type of pathology, as recently reviewed by Hengsbach (1996).

Fig. 34.

Sublethal injuries in Hoploscaphites nodosus (Owen, 1852) and H. brevis (Meek, 1876), Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. A, B. Hoploscaphites brevis, microconch, BHI 4898. A. Right lateral; B. ventral, showing temporary loss of symmetry on the shaft (arrow), with the ventrolateral tubercles migrating to the left side and then to the right side, before reestablishing themselves on the hook. C, D. Hoploscaphites brevis, microconch, transitional in morphology to macroconch, BHI 4316. C. Right lateral; D. close-up of swollen scar (arrow in C) on the right side of the adoral part of the shaft, with widely spaced ribs. E. Hoploscaphites brevis, microconch, BHI 4256, showing a V-shaped scar, followed by a large inflated area with distorted ribs, terminating at the aperture (area enclosed in brackets). A large piece of shell is also missing from the adapical part of the body chamber. F, G. Hoploscaphites brevis, microconch, BHI 4896, with a single row of ventrolateral tubercles (area enclosed in brackets). F. Right lateral; G. ventral. H, I. Hoploscaphites nodosus, microconch, BHI 4895, with a smooth healed area on the venter of the body chamber (area enclosed in brackets). H. Right lateral; I. ventral. J, K. Hoploscaphites brevis, AMNH 56871, broken phragmocone with an inflated scar (arrow) at a whorl height of 6.5 mm. J. Right lateral; K. apertural. The scar is covered with distorted ribs that point backward, forming an acute angle with the midventer. Specimens are illustrated natural size unless indicated otherwise by a scale bar.

Fig. 35.

Sublethal injuries in Hoploscaphites nodosus (Owen, 1852). A. Hoploscaphites nodosus (?), microconch, left lateral view, YPM 35631, with a single row of midventral tubercles instead of two rows of ventrolateral tubercles, YPM loc. A1461, Bearpaw Shale, Rosebud County, Montana. B. Hoploscaphites nodosus, macroconch, right lateral view, YPM 35598, with a spiral furrow on the right side of the shell beginning just adoral of the point of exposure, YPM loc. C3347, Pierre Shale, Belle Fourche River, South Dakota. C, D. Hoploscaphites nodosus, microconch, BHI 4912, with a round “blister” (arrow) on the venter just adapical of the point of recurvature, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. C. Ventral; D. left lateral. Specimens are illustrated natural size.

If the pathology cannot be attributed to an injury, it may be due to disease, attachment of epizoans to the shell, or infestation by parasites. Examples of such deformities include specimens that depart from bilateral symmetry, without any evidence of a scar in the shell (Landman and Waage, 1986: “Morton's syndrome”; Hengsbach, 1996: “symmetropathy”; Keupp and Ilg, 1992: “forma undaticarinata”). In BHI 4896, a microconch of Hoploscaphites brevis, the right row of ventrolateral tubercles on the body chamber disappears, although the left row persists (fig. 34F, G). There is no indication of a scar in the shell, suggesting that the disappearance of the tubercles may be due to a parasitic infestation or disease. A similar explanation may account for the deformity in YPM 35631 (fig. 35A), a microconch of H. nodosus (?). In this specimen, both rows of ventrolateral tubercles migrate toward the center, beginning just adoral of the point of exposure. The two rows subsequently merge into a single row, which oscillates back and forth before stabilizing on the midventer on the adoral part of the phragmocone. At the point of recurvature, the tubercles abruptly diminish in size and reappear as two closely spaced rows. This deformity also produced a bulge on the midventer on the adoral end of the phragmocone, which caused a displacement of the ventral lobe of the suture to the left side. In addition, there is a scar 3.5 cm long ×1.5 cm wide, part of which is inflated, on the dorsal half of the flanks on the left side of the adoral part of the phragmocone and adapical part of the body chamber. It is possible that this scar represents a repaired break at the aperture. However, this scar is adoral of the initial displacement of tubercles, suggesting that the original cause of this displacement was disease or parasitic infestation.

A pathology due to an injury is easier to recognize because of the presence of a scar, which formed during the lifetime of the animal. Some of these scars are the results of injuries at the apertural margin. If the injury was minor, the ornament on the repaired shell is “normal,” but is ususally offset with respect to the ornament on the adapical side of the break. If the injury was severe, the ornament on the repaired shell may be distorted or absent altogether, depending upon the ability of the mantle to withdraw to the position of the break, as documented in other ammonites (Kröger, 2002a). If the ornament is absent, it implies that the epithelial mantle could not withdraw, probably because the shell break was too extensive, and that the shell was repaired instead by a part of the mantle that could not secrete ornament. The repaired shell is sometimes inflated, indicating that the mantle expanded to fill the broken part of the shell. If the scar persists to the aperture, it suggests that the mantle edge was permanently damaged. If the scar eventually disappears and merges with the rest of the shell, it suggests that the mantle was able to recover following the attack.

Examples of repairs at the aperture involving permanent damage to the mantle are grooves that follow the spiral growth of the shell (Hölder, 1956: forma verticata; Landman and Waage, 1986: spiral furrow). In general, the injury also affects the surrounding ornament, sometimes resulting in the loss of ribs or tubercles (Guex, 1967: “compensation ornamentale”). YPM 35598, a macroconch of Hoploscaphites nodosus, shows a spiral furrow on the right side of the shell (fig. 35B). The furrow originates just adoral of the point of exposure and extends to the point of recurvature. The ribs on the flanks bend backward at the furrow, which runs along the lateral side of the ventrolateral tubercles. The right side of the venter and most of the right flanks of the hook are missing, suggesting a fatal injury of the already weakened specimen. USNM 12284, a macroconch of H. nodosus, which was originally illustrated by Whitfield (1880, 441, pl. 13, fig. 12), shows a spiral furrow on the left side (the side not figured by Whitfield). The furrow originates on the middle of the exposed part of the phragmocone at approximately the line of maximum length and continues to the point of recurvature. The ventrolateral tubercles and associated ribs are absent along the furrow, but the rest of the ornamentation on the left side is unaffected. Most of the apertural part of the shell is missing, again perhaps due to a fatal injury.

Examples of repairs at the aperture involving nonpermanent damage to the mantle include BHI 4898 and BHI 4316, both of which are microconchs of Hoploscaphites brevis. In BHI 4898, there is a quadrangular patch of repaired shell 1 cm long on the left side of the adapical end of the body chamber (fig. 34A, B). The scar is slightly inflated and stands out from the rest of the shell. The ornament is distorted on the adapical end of the scar, but becomes more “normal” toward the adoral end of the scar. In addition, there is a temporary loss of symmetry on the shaft, with the ventrolateral tubercles migrating first to the left side and then to the right side, before reestablishing themselves on the hook. In BHI 4316, there is a swollen scar on the right side of the adoral part of the shaft (fig. 34C, D). The scar is approximately 1.5 cm long and spans an angle of approximately 40°. The scar is smooth on the dorsal side but is ribbed on the lateral and ventral sides. The ribs on the scar are coarser and more widely spaced than those on the adjacent part of the shell, but become more “normal” toward the adoral end of the scar. The presence of ornament on the scar indicates that the mantle epithelium must have been able to withdraw an angular distance of 40° back from the aperture in order to repair the break.

Scars in the shell also result from damage to the body chamber adapical of the aperture. Such injuries have been referred to as “back breaking” if they occur on the adapical part of the body chamber and “body piercing” if they occur on the flanks of the body chamber (Larson, 2003, 2007). Keupp (2006) introduced the term “forma aegra fenestra” for these windowlike scars. As with injuries at the aperture, the presence of ornament on the scar indicates that the break was repaired by the mantle edge. If, on the other hand, the scar is smooth, it was repaired by the posterior part of the mantle, which could not secrete ornament. The repaired area is usually inflated and appears as a “blister” surrounded by a sharp fracture (Keupp, 2006: “forma inflata”). However, the demarcation between the blister and the rest of the shell is not as apparent if the outer shell has spalled off, leaving an internal mold.

There are several specimens that show “blisters” on the body chamber. In BHI 4912, a large microconch of Hoploscaphites nodosus, there is a round “blister” approximately 2.5 cm in diameter at an angular distance of approximately 80° from the mature aperture (fig. 35C, D). The blister extends from the middle of the venter to the middle of the left flanks. It is smooth and surrounded by “normally” ornamented shell. This scar represents the repair of a hole in the body chamber by the posterior part of the mantle. In BHI 4895, a microconch of H. nodosus, there is a smooth area along the venter approximately 4 cm long (fig. 34H, I). The ventrolateral tubercles are present on the right side but are missing on the left side. Ribs appear on the flanks, as well as on the venter on the adapical and adoral ends of the blister. Although much of the outer shell has spalled off, the most likely explanation is that this scar represents the repair of a hole in the body chamber by the posterior part of the mantle. In addition, there is a jagged hole on the venter and part of the left side of the body chamber at the point of recurvature, which probably represents the final, fatal injury.

Similar scars are also present in juvenile shells. In AMNH 56871, a broken phragmocone presumably of Hoploscaphites brevis, there is an inflated scar at a whorl height of 6.5 mm (fig. 34J, K). The scar is approximately 6 mm long and extends from the middle of the venter to the outer flanks on the left side. The scar is covered with distorted ribs that point backward, forming an acute angle with the midventer. The scar is sharply demarcated on the adapical, adoral, and ventral sides and is surrounded by more “normal” ornament. The presence of distorted ribs suggests that the break occurred close enough to the aperture that the peristomal epithelium was able to withdraw far enough back to repair it.

In BHI 4256, a microconch of Hoploscaphites brevis, the coiled portion of the shell is perfectly intact, but the body chamber is severely deformed (fig. 34E). The pathology is probably due to two injuries, implying that the animal was attacked, survived, repaired its shell, and was attacked again. However, the exact sequence of events is difficult to piece together. The adoral part of the body chamber is ornamented on the flanks, but the venter is bumpy and devoid of ornament. In addition, on the right side on the adoral end of the shaft, there is a sharp V-shaped chevron, followed by a large inflated area with distorted ribs, terminating at the aperture. The first attack may have occurred adapical of the aperture and was repaired by the posterior part of the mantle. The second attack may have occurred at the aperture and resulted in the formation of the V-shaped scar and blister with distorted ribs, culminating at the aperture. This specimen also suffered a third, fatal injury, as indicated by a break in the adapical part of the body chamber.

Lethal injuries are indicated by missing pieces of shell, although it is not always easy to distinguish these breaks from postmortem damage. In addition, incomplete shells are generally overlooked because of the tendency to collect whole, perfect specimens. However, the consistent position of the injuries is a clue that they occurred during life, rather than after death (Radwański, 1996; Larson, 2003, 2007; Keupp, 2006; Klompmaker et al., 2009). In general, lethal injuries in scaphites occur on the venter and sides on the adapical part of the body chamber, precluding repair of the shell. The resulting accumulation of shell fragments on the sea bottom is sometimes difficult to distinguish from a shell hash produced by hydrodynamic processes. However, the presence of jagged shell fragments in association with scaphites showing lethal injuries suggests that such an accumulation is a result of predation rather than mechanical breakage (fig. 36F; for further discussion about “kitchen middens,” see Radwański, 1996). Lethal injuries have been documented in Late Cretaceous scaphites from the Western Interior (Larson, 2003), the Atlantic Coastal Plain (Landman et al., 2007a), and the Vistula River Valley, central Poland (Radwański, 1996).

Lethal injuries are common in Hoploscaphites nodosus. In USNM 536235, a large macroconch, there is a triangular gap approximately 4 cm long on the venter and ventrolateral part of the right side of the body chamber at midshaft (fig. 36E). In USNM 536234, a large macroconch, a chunk of shell 3 cm × 4 cm is missing from the venter and the ventrolateral parts of both flanks on the adapical end of the body chamber (fig. 37C, D). In BHI 4287a, a large macroconch, the venter and nearly all of the right side of the shaft of the body chamber is missing (fig. 37A, B). However, despite the removal of such a large part of the body chamber, the rest of the specimen is perfectly intact. In BHI 7033, a large microconch, there is a triangular gap spanning the venter and most of the flanks on the left side on the adapical end of the body chamber (fig. 36D). This specimen also shows a chunk of shell missing on the left side of the body chamber near the point of recurvature.

Fig. 36.

A–E. Lethal injuries in Hoploscaphites nodosus (Owen, 1852) and H. brevis (Meek, 1876). A–C. Hoploscaphites brevis. A. BHI 4267, macroconch, right lateral view, with a puncture at one-third whorl height (arrow) on the right flanks on the adapical end of the body chamber, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. B. BHI 4900, macroconch, left lateral view, with a gap approximately 1 cm long on the venter and the ventrolateral flanks on the adapical end of the body chamber (area enclosed in brackets), Baculites compressus Zone, Pierre Shale, Meade County, South Dakota. C. BHI 4901, microconch, transitional in morphology to macroconch, right lateral view, with a gap almost 2.5 cm long on the adapical end of the body chamber, including both the venter and ventrolateral part of the shell (area enclosed in brackets), Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. D. Hoploscaphites nodosus, large microconch, BHI 7033, left lateral view, with a triangular shaped gap spanning the venter and most of the flanks on the left side on the adapical end of the body chamber (area enclosed in brackets), Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. E. Hoploscaphites nodosus, macroconch, USNM 536235, right lateral view, with a triangular shaped gap approximately 4 cm long on the venter and ventrolateral part of the right side of the body chamber at midshaft (area enclosed in brackets), USGS Mesozoic loc. 9154, Pierre Shale, near Kremmling, Grand County, Colorado. F–H. Postmortem shell damage. F. AMNH 63449, fragments of the body chamber of Hoploscaphites nodosus, probably resulting from predation, AMNH loc. 3274, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. G, H. Minute holes on the left flanks of the phragmocone of a juvenile of Hoploscaphites brevis, AMNH 56819, AMNH loc. 3274, Baculites compressus–B. cuneatus zones, Pierre Shale, Meade County, South Dakota. G. Close-up of holes; H. left lateral view. Specimens are illustrated natural size unless indicated otherwise by a scale bar.

Fig. 37.

Hoploscaphites nodosus (Owen, 1852), macroconchs, lethal injuries with chunks of shell missing from the adapical part of the body chamber (areas enclosed in brackets in lateral views, arrows in ventral views). Despite the removal of large pieces of shell, the specimens are otherwise intact. A, B. BHI 4287a, Baculites cuneatus Zone, Pierre Shale, Meade County, South Dakota. A. Right lateral; B. ventral. C, D. USNM 536234, USGS Mesozoic loc. D5034, Pierre Shale, Baculites compressus Zone, Larimer County, Colorado. C. Right lateral; D. ventral. Specimens are illustrated natural size.

Lethal injuries are equally common in Hoploscaphites brevis. In BHI 4900, a small macroconch, there is a gap approximately 1 cm long on the venter and the ventrolateral flanks on the adapical end of the body chamber (fig. 36B). The edges of the gap are sharp and irregular. In BHI 4901, a small microconch, there is a similar gap, almost 2.5 cm long, on the adapical end of the body chamber, affecting both the venter and ventrolateral parts of the flanks (fig. 36C). In AMNH 50429, a large macroconch, most of the shaft of the body chamber was removed by a predator (fig. 12A, B). Subsequently, during fossilization, this specimen was twisted and crushed, producing a jigsaw of shell fragments, which were probably held together by the remaining pieces of periostracum. In BHI 4267, a macroconch, there is a rectangular hole 1 cm long at one-third whorl height on the right flanks on the adapical end of the body chamber (fig. 36A). In AMNH 56819 (fig. 36G, H), a small piece of a phragmocone, two tiny holes, 0.75 mm and 1 mm in diameter, appear on the left flanks at a shell diameter of 12 mm. These holes may represent lethal injuries but, surprisingly, they occur on the phragmocone and not on the body chamber.

The position of injuries on the scaphite shell provides clues about the mode of life of these animals including the kind of predator and direction of attack. Injuries at the aperture indicate that scaphites were sometimes attacked from the front, based on the orientation of the living animal (see section on mode of life). Such injuries in ammonites have been attributed to fish, reptiles, crustaceans, and cephalopods (Landman and Waage, 1986; Larson, 2003; Keupp, 2006; Klompmaker et al., 2009). Injuries on the venter and ventrolateral flanks on the adapical end of the body chamber indicate that scaphites were also attacked from behind (Larson, 2003). This is a particularly vulnerable area because it is out of eyesight, although the exact position of the eyes is unknown. In addition, the whorl section of the body chamber is subquadrate and compressed, making it relatively easy for a predator to grasp and bite the shell. As a result, these injuries are usually fatal. It is, therefore, no coincidence that the adapical end of the body chamber is the most heavily armored part of the scaphite shell, featuring subspinose ventrolateral tubercles (Riccardi, 1983; Landman and Waage, 1993; Monks, 2000). Such tubercles effectively increased the size of the scaphite, discouraging small predators (for a discussion of the antipredatory function of ammonite ornamentation, see Kröger, 2002b). Injuries on the venter and ventrolateral flanks on the adapical end of the body chamber in ammonites have been attributed to fish, reptiles, and cephalopods (Larson, 2003; Klompmaker et al., 2009).

Holes in the flanks of the body chamber indicate that scaphites were also attacked from the side. Radwański (1996) noted such holes in late Maastrichtian scaphites from Poland, and attributed them to crabs, lobsters, and stomatopods. Larson (2003) described similar holes in Campanian and Maastrichtian scaphites from the U.S. Western Interior and attributed them to fish and reptiles. Keupp (2006) reviewed the occurrence of such punctures in a wide variety of ammonites and pointed out that in many specimens, the holes are restricted to only one side. He argued that such holes could not have been produced by grasping predators, such as vertebrates and claw-bearing crustaceans, but rather by stomatopods that would have inflicted ballistic damage on one or both sides of the body chamber. Because all living stomatopods are benthic, he reasoned that ammonites exhibiting such punctures must have lived close to the bottom. This conclusion is consistent with other evidence about the habitat of scaphites (see below).

Mode of Life

Inferences about the mode of life of scaphites are based on observations of their morphology (shell, septa, muscle scars, jaws, hooks, radula), in analogy with modern cephalopods whose mode of life is known. Assuming that scaphites maintained near neutral buoyancy during life, we calculated the angle of orientation of the aperture in embryonic, juvenile, submature, and mature shells (fig. 14). We estimated the centers of buoyancy and mass using the same approach as Trueman (1941). Because scaphites are bilaterally symmetrical, we used photos of specimens in lateral view and cut out the whole specimen and body chamber separately (requiring two photos). Using a computer program (Image J), we determined the centers of the areas of the whole specimen and of the body chamber, and treated these as proxies for the centers of buoyancy and mass, respectively. We then drew a line through the centers of mass and buoyancy ( =  the vertical) and another line through the apertural margin. The angle of orientation of the aperture is defined as the angle of the aperture with respect to the vertical.

In an embryonic shell of Hoploscaphites brevis, the angle of orientation of the aperture is 106.0° (table 2). In a juvenile shell of the same species, the angle of orientation of the aperture is 87° (table 2). In a specimen of H. nodosus in which the adult body chamber was removed to expose the phragmocone, the angle of orientation of the aperture of the “juvenile,” assuming a body chamber angle of 230°, is 88° (table 2). Determination of the angle of orientation of the aperture in submature shells with incomplete shafts is difficult because of the lack of specimens preserved at this stage. In AMNH 55873, a broken microconch of H. nodosus, the base of the body chamber is approximately 90° adapical of the line of maximum length (fig. 12K, L), presumably corresponding to the aperture at the point of recurvature. In principle, the apertural margin at this point would have followed the outline of the ribs (for a similar assumption in nostoceratid ammonites, see Okamoto, 1988: 275). Using this specimen as a model, we reconstructed the “submature” shell in two microconchs of H. brevis. The angle of orientation of the aperture in these two “submature” specimens averages 121° and ranges from 117.5° to 125° (table 2). Determination of the angle of orientation of the aperture in mature shells is much easier because the base of the body chamber is usually visible. We examined 10 mature specimens each of H. nodosus and H. brevis (five macroconchs and five microconchs in each sample for a total of 20 specimens). The angle of orientation of the aperture averages 102.2° and ranges from 90.5° to 109° (table 2).

In summary, our results indicate that the angle of orientation of the aperture is approximately 90° throughout ontogeny. However, the method employed involves at least three assumptions: (1) The phragmocone is completely filled with air. Liquid in the last few chambers effectively decreases the size of the phragmocone and increases the size of the body chamber (Saunders and Shapiro, 1986). We recalculated the centers of buoyancy and mass in a few of the specimens in our sample assuming a smaller phragmocone (minus the last two chambers) and a larger body chamber. The angle of orientation of the aperture increases by approximately 10°. (2) There is a uniform distribution of soft tissues in the body chamber. The presence of an aptychus-type jaw near the aperture effectively lengthens the body chamber, reducing the angle of orientation. Similarly, an extension of the soft body outside of the aperture decreases the angle of orientation (Kakabadzė and Sharikadzė, 1993; Monks and Young, 1998). (3) There is no variation in shell thickness. The presence of a varix at the apertural margin decreases the angle of orientation.

The angle of orientation of the aperture probably varies within an acceptable range of values throughout ontogeny. This assumption lies at the foundation of most studies about the theoretical morphology of ammonoids (e.g., Okamoto, 1988). In the scaphites from the Baculites compressus–B. cuneatus zones, we have noted a correlation between the septal angle and the apertural angle in adults. This correlation is stronger in macroconchs of Hoploscaphites brevis than in macroconchs of H. nodosus (fig. 38A, B). It is particularly apparent in the combined sample of both species (fig. 38C). The apertural angle is higher in specimens in which the last septum occurs below the line of maximum length. This suggests that the two angles covary such that the aperture achieved a preferred orientation at maturity irrespective of the position of the last septum.

Fig. 38.

Plot of septal angle versus apertural angle in macroconchs of Hoploscaphites nodosus (Owen, 1852) and H. brevis (Meek, 1876). The apertural angle is higher in specimens in which the last septum falls below the line of maximum length. This suggests that the two angles covary such that the aperture achieved a preferred orientation at maturity irrespective of the position of the last septum. Estimates are excluded. A. Hoploscaphites nodosus (Owen, 1852). (The smallest macroconch of H. nodosus [USNM 536270] is excluded). B. Hoploscaphites brevis (Meek, 1876). C. Total sample of Hoploscaphites nodosus and H. brevis.

The high angle of orientation of the aperture in adult scaphites is incompatible with a nektobenthic mode of life involving scavenging or searching for food on the bottom, as exemplified by modern nautilus (Klinger, 1981; cf. Monks, 2000). This interpretation is supported by several morphological features of the adult shell. The apertural margin exhibits a constriction ending in a very thin lip. If the soft body extended downward to feed off the bottom, the lip would have easily broken off. In fact, the apertural margin is commonly intact in the thousands of specimens we have examined. In addition, the apertural opening is reduced in size by the constriction and accompanying varix as well as the dorsal extension of the shell. This reduction in the size of the opening would have impeded unrestricted movement of the soft body, further reducing access to the bottom. Trueman (1941: 374) reached this same conclusiton in his study of the mode of life of scaphites: “it is extremely difficult to understand how these forms could have been benthonic crawlers, as has commonly been suggested; the buoyant effect of the air-chambers would make such a mode of life almost impossible.”

Lewy (1996) proposed an alternative explanation for the shape of the scaphite body chamber at maturity. According to him, the upwardly facing aperture and terminal constriction would have so impeded movement of the soft body that it would have resulted in death. Subsequently, the body chamber would have served as a floating egg case, wth the newly hatched young feeding on the carcasse of the dead female. He based this interpretation on the similarity of the scaphite body chamber to the egg case of present-day argonauts. Although the resemblance between these two shells is striking, as noted by previous authors, it is due to evolutionary convergence. The ammonite shell is secreted by the mantle, whereas the argonaut egg case is secreted by the glands of the expanded dorsal arms (Saul and Stadum, 2005).

Using nautilus as a model for swimming in ammonites suggests that scaphites were probably poor swimmers (but nautilus may not be the best model to use; see Jacobs and Landman, 1993). Scaphites lack large retractor muscles, which would have provided the propulsive power necessary for strong swimming. Such muscles may have existed in other ammonites based on the presence of scars on the internal surfaces of the shells (Doguzhaeva and Mutvei, 1996). In scaphites, there are only two types of muscle scars, the paired dorsal muscle scars and the unpaired midventral muscle scar (fig. 20). In addition, scaphites lack a hyponomic sinus on the midventer, which would have prevented the animal from extending its hyponome below the shell in order to swim forward. Therfore, scaphites may have been limited in their movement to swimming backward or downward (Westermann, 1996). Juvenile scaphites, with their compressed, closely coiled shells, may have been relatively efficient at slow, continuous swimming (Jacobs, 1992; Jacobs and Chamberlain, 1996). The uncoiling of the body chamber at maturity undoubtedly increased the stability and coefficient of drag, and decreased the hydrodynamic efficiency.

The poor swimming ability of scaphites, especially as adults, may have prevented them from migrating over long distances. Instead, they may have exploited a low-energy lifestyle, remaining at a single site for an extended period of time, subject to current activity. This interpretation is supported by two studies of scaphite species in the Western Interior Seaway. Jacobs et al. (1994) documented subtle but distinct differences in the shape of the adult shells of Scaphites whitfieldi from two regions approximately 225 km apart in South Dakota and Wyoming. If the sites are time equivalent, it suggests that the populations from the two regions did not intermingle. In addition, Cochran et al. (2003) investigated strontium isotope ratios in Hoploscaphites nebrascensis (Meek, 1876) from various biofacies (e.g., brackish water, nearshore, offshore) in the Pierre Shale and Fox Hills Formation in South Dakota. The strontium isotope ratios in specimens from a single biofacies are similar to each other, and different from those in specimens from other biofacies. If the sites are time equivalent, this suggests that the scaphites did not migrate among biofacies.

The rate of growth, up to the onset of maturity, was probably the same in both dimorphs within a species, as indicated by the fact that the early whorls are indistinguishable between them. The uncoiling of the body chamber indicates the initiation of the mature growth program. One remarkable observation is that, after many years of study, we have never found a single scaphite preserved at the submature stage, that is, with the shaft of the body chamber in the process of formation. It is possible that the apertural margin at this stage was very thin and easily fragmented during fossilization. Another explanation is that the animal was very vulnerable to predation at this stage, and did not commonly survive. An example of a specimen that was killed at this stage is AMNH 55873 (fig. 12K, L). It is a broken microconch of Hoploscaphites nodosus with a lethal break on the adapical part of the shaft. What is unusual about this specimen is that the base of the body chamber is approximately 90° adapical of the line of maximum length, implying that the aperture was at the point of recurvature when the injury was inflicted. An additional explanation for the lack of shells at the submature stage is that the shaft may have formed rapidly, so that the chances of preserving it in the process of formation were low. Landman and Waage (1993) documented a reduction in the thickness of the shell wall and a weakening of the ribbing at midshaft in macroconchs of H. nicolletii, possibly reflecting an increase in the rate of growth. Of course, the growth rate decreased and eventually stopped at the attainment of maturity, as indicated by a reduction in the spacing of ribs and tubercles on the hook and the formation of a constriction and varix at the aperture. This stage also coincides with a decrease in septal spacing. The differences in the patterns of septal spacing between dimorphs, as described above, further imply that macroconchs experienced a longer period of slower growth. The larger size of macroconchs generally implies that they reached maturity later than microconchs. Depending on the age difference between them, macroconchs may have mated with microconchs from another year's cohort.

One of the most notable features of the scaphites from the Baculites compressus–B. cuneatus zones is that the largest ventrolateral tubercles appear on the midshaft (Riccardi, 1983; Landman and Waage, 1993; Monks, 2000) (fig. 17). In Hoploscaphites nodosus, these tubercles usually develop into subspinose clavi. The presence of such large tubercles on the midshaft would have protected the animal against predation from behind, presumably from its blind side. (The eyes may have been located in the reentrant on each side of the dorsal projection.) The ventrolateral tubercles diminish in size near the adult aperture. This may imply that attacks from below and from the front were less common. It is possible that the calcified aptychi of the lower jaws may have provided additional protection in the apertural region.

The presence of epizoans on ammonite shells also provides clues to the mode of life of these animals. Surprisingly, we have never observed a scaphite encrusted with epizoans that grew on it during its lifetime. The absence of such epizoans may be due to antifouling adaptations such as a thick periostracum or mucuslike covering (Landman et al., 1987). In contrast, specimens of scaphites with epizoans that settled on the ammonites after they died are not uncommon. For example, AMNH 56763 is a macroconch of Hoploscaphites nodosus with worm tubes on the inside surface of the body chamber, indicating that the specimen must have rested on the sea floor for an extended period of time (fig. 13K, L). There are reports of epizoans on the shells of other cephalopods in the Western Interior Seaway, but most of these specimens were encrusted by epizoans after death. Gill and Cobban (1966) documented calcareous worm tubes and traces of other organisms on the steinkerns of body chambers of baculites from the Pierre Shale at Red Bird, Wyoming. Kase et al. (1998) reported limpets on Placenticeras meeki from the Pierre Shale of South Daktoa and Montana. One specimen is covered with 102 limpets on the left side and 21 on the right side. In most of these specimens, the limpets settled on the shells after the animals died and floated to the surface. Specimens of Placenticeras are also commonly encrusted with bryozoans (L. Larson and T. Linn, personal commun., 2008), but it is unclear whether the bryozoans settled on the ammonites before or after death. Dunbar (1928) documented bryozoans on a specimen of what he called Sphenodiscus lenticularis mississippiensis Hyatt from the Ripley Formation of Mississippi. The bryozoans settled on the ammonite while it was still alive based on the fact that the bryozoans are partly covered by the secretion of the next whorl. Landman et al. (1987) illustrated a specimen of Eutrephoceras dekayi (Morton, 1834) from the Fox Hills Formation encrusted by bryozoans and serpulids. These epizoans settled on the animal while it was alive and were later overgrown by subsequent whorls.

There is no direct evidence about the feeding habits of scaphites based on finds of stomach contents. Like many other members of the Aptychophora Engeser and Keupp, 2002 ( =  the aptychus-bearing ammonites), scaphites may have preyed upon small organisms in the water column, including decapod crustaceans, copepods, and newly hatched ammonites (Kennedy et al., 2002; Landman et al., 2007b). They may have also eaten larger prey, including juvenile ammonites and fish (Larson, 2003: 10, 11, fig. 13). Recent finds of radulae in Baculites and the scaphite Rhaeboceras have shed new light on this subject (Kruta et al., in prep). Although the morphology of the radula is basically the same in both genera, the teeth are shorter and more robust in Rhaeboceras than in Baculites (the large hooklike elements in Rhaeboceras previously interpreted as radulae are something else; see below). The difference in the shape of these teeth suggest that Rhaeboceras, and perhaps scaphites, in general, fed on slightly larger prey than did Baculites.

Reproduction in scaphites has received little attention. The high angle of orientation of the aperture in scaphites, as in many other heteromorphs (e.g., didymoceratids) presents challenges to copulation. In modern nautilus, the aperture is oriented at an angle of 30 ± 15°, allowing the two sexes to meet head on, with the male grasping the female using its jaws and cirri (Arnold, 1987). However, this method of copulation would not have been possible in scaphites without considerable gymnastics. It is much more likely that in scaphites, as in many modern coleoids, the male possessed a specially modified arm (the hectocotylus) that delivered the spermatophores to the female. The deployment of such an arm would have allowed the two sexes to mate head on or side by side. The best evidence of the existence of such a specially modified arm in scaphites is the presence of hooklike structures in association with Hoploscaphites nicolletii and H. nebrascensis from the upper Maastrichtian Fox Hills Formation (Landman and Waage, 1993: figs. 43–46) and the much larger, but otherwise similar structures in association with Rhaeboceras from the Campanian Bearpaw Shale (Kennedy et al., 2002). Hooklike structures are also present in association with species of Hoploscaphites from methane seeps in the Pierre shale in the Baculites scotti–B. compressus zones.

Habitat

Hoploscaphites nodosus and H. brevis inhabited the central and northern part of the Western Interior Seaway. During the late Campanian, the sea flooded southern Alberta and central Montana, but the western two-thirds of Wyoming were emergent. The migration of the western shoreline at the time is illustrated in Gill and Cobban (1973). The western shoreline does not show a consistent pattern of displacement along strike due to local variation in tectonic activity, sedimentation rates, sediment supply, and accommodation rates (Krystinik and DeJarnett, 1995). Gill and Cobban (1973: 34) calculated that the transgression of the strandline across Montana occurred at a rate of approximately 115 km per million years (Gill and Cobban, 1973: 34).

Reconstructing the habitat of scaphites is difficult because these animals lived in the water column but are preserved in the sediments. We are interested in determining the environment in which the scaphites lived and, in particular, the depth below the sea surface and the height above the bottom. This information is derived from five sources: (1) the biofacies distribution of the scaphites; (2) the geographic distribution of the scaphites in the basin; (3) the presence of fauna associated with the scaphites and the inferred habitat and mode of life of such fauna; (4) the isotopic composition of the scaphites, especially in comparison with that of other faunal elements whose habitat and mode of life are known; and (5) the mechanical properties of the scaphite shells, as indicators of implosion depths and, by implication, habitat depths.

In evaluating most of this information, it is important to determine the origin of the fossil deposit. Is it a condensed sequence spanning hundreds of thousands of years? Do the fossils represent animals that lived at approximately the same time and in the same place or do they represent a time average of successive generations? In brief, does the fossil deposit reflect the living community? These issues have been widely discussed with respect to benthic organisms (e.g., Kidwell, 1991; Kidwell and Bosence, 1991). Ammonites pose special problems because they lived in the water column and, by definition, represent transported or “exotic” elements, necessarily creating a mixed fossil assemblage on the sea floor (Maeda and Seilacher, 1996). In addition, ammonites are prone to postmortem drift, in analogy with modern nautilus (House, 1987).

The most important piece of evidence to help distinguish between ammonites that sank soon after death and those that floated for an extended period of time is the presence of jaws preserved nearby or inside the ammonite shells. This association has generally been interpreted as indicating that the animals lived, died, and were preserved at approximately the same site (Tanabe, 1979; Landman et al., 2007a; Wani et al., 2005; Wani, 2007). If an ammonite floated after death, its soft body would have fallen out, and the jaw would have either been lost or preserved at a different site than the shell. Field studies of ammonite assemblages have suggested that postmortem drift is not as important a phenomenon as previously thought. In his study of the distribution of ammonite shell morphotypes in the Western Interior Seaway, Batt (1989) noted that certain mophotypes were restricted to particular parts of the basin and, therefore, concluded that postmortem drift was a relatively insignificant factor for most ammonites in shallow epicontinental seas.

Hoploscaphites nodosus and H. brevis occur in Alberta, Saskatchewan, Montana, North Dakota, South Dakota, Wyoming, Colorado, and Kansas. They are unknown outside of the Western Interior Basin of North America, including the Gulf Coastal Plain, although their absence in this area may be due to a lack of preservation. The abundance of scaphites diminishes toward the north. Such a decrease in molluscan diversity has previously been observed by Kauffman and Caldwell (1993), and presumably reflects the global decrease in diversity from the tropics to the poles. For example, at Porcupine Dome in east central Montana, scaphites are abundant and are associated with a rich molluscan fauna (Cobban, 1962a). This fauna becomes sparser toward the north, consisting mostly of only baculites and placenticeratids. Tsujita and Westermann (1998) noted that in the Baculites compressus Zone, scaphites are absent in western Alberta, and rare in eastern Alberta (Cypress Hills area). According to them, scaphites are absent altogether in Alberta in the transitional beds between the B. compressus and B. cuneatus zones and in the B. cuneatus Zone. They postulated that this absence was due to the existence of stagnant bottom waters trapped on the west side of the Sweetgrass Arch, which may have formed a slight submarine high during this time (Stelck, 1975).

The scaphites occur in a variety of biofacies ranging from nearshore to offshore environments. They are present in sandy, nearshore environments, as represented by the unnamed shale member of the Pierre Shale along the Front Range of the Rocky Mountains (Scott and Cobban, 1965, 1975, 1986a, 1986b), and the unnamed shale member of the Pierre Shale near Kremmling, Colorado (Izett et al., 1971; Izett and Barclay, 1973; Cobban et al., 1992; Kennedy et al., 2000a). They are also present in more offshore deposits, as represented by the Pierre Shale and Bearpaw Shale in South Dakota and Montana, respectively. For example, they occur in the DeGrey Member of the Pierre Shale in Buffalo County, South Dakota, which was located several hundred kilometers to the east of the shoreline at the time (fig. 1). This site may have represented a shallow area east of the forebulge region (Cross and Pilger, 1978; Metz, 2008). The scaphites also occur in methane seep facies in the Pierre Shale in Custer County, South Dakota (see below).

In all probability, Hoploscaphites nodosus and H. brevis occupied all biofacies simultaneously. However, the level of biostratigraphic resolution available is inadequate to determine whether there is any diachroneity, that is, whether species appeared first in nearshore facies and subsequently migrated to offshore facies, or vice versa. The upper Maastrichtian Fox Hills Formation in north-central South Dakota is subdivided into assemblage zones, with several assemblage zones per ammonite zone (Landman and Waage, 1993), permitting a higher biostratigraphic resolution. The bio- and lithostratigraphic distributions of the scaphites in these strata were compiled by C. MacClintock (personal commun., 2008). According to his data, the lowest occurrence of both H. nicolletii and H. spedeni is in the offshore muds of the Mobridge Member of the Pierre Shale. Both species range into the nearshore silts of the Little Eagle lithofacies of the Trail City Member of the Fox Hills Formation and then disappear. These species are replaced by H. comprimus and H. nebrascensis. However, the lowest occurrence of these species differs depending on the lithofacies. In the upper part of the Little Eagle lithofacies, the lowest occurrence of these species is in the Abyssinus concretions, whereas in the muddy Irish Creek lithofacies, the lowest occurrence of these species is higher (just below the Cucullaea Assemblage Zone). This discrepancy may imply that these two species originated in the environment represented by the Little Eagle lithofacies, and then migrated into the environment represented by the Irish Creek lithofacies, perhaps reflecting a change in living conditions.

The scaphites in the Baculites compressus–B. cuneatus zones are almost always associated with an abundant, diverse benthic fauna. A rich association of scaphites and benthic fauna has been documented in sandstones and sandy shales in the Pierre Shale in Grand County, Colorado (Izett et al., 1971). These deposits represent a nearshore environment and contain 72 species, most of which are bivalves, in addition to 13 species of cephalopods (Sava, 2007), so that the total number of species equals 85 (this figure might be slightly inflated due to the assignment of incomplete specimens to new species [S. Klofak, personal commun., 2009]). The excellent preservation of the fossils suggests that they did not experience much postmortem transport. We have documented a more depauperate fauna consisting of 26 benthic species in addition to scaphites and baculites in a single concretion from the Baculites compressus–B. cuneatus zones in the Pierre Shale, Meade County, South Dakota, which represents a more offshore environment (Landman and Klofak, 2006). The concretion contains dozens of ammonite jaw fragments, indicating that the fauna was buried rapidly (within months to years). An association of scaphites and benthic fauna has also been observed by Tsujita and Westermann (1998) from the same biostratigraphic interval in the Bearpaw Shale in Montana. In other horizons in the Pierre Shale, for example, in the B. grandis Zone, scaphites are also usually associated with a diverse benthic fauna (Gill and Cobban, 1966).

The common association of scaphites and benthic species has generally been interpreted as indicating hospitable, that is, oxygen-rich episodes in the history of the Western Interior Seaway (Tsujita and Westermann, 1998: fig. 12). These faunal associations suggest that the same environmental factors that promoted the development of the benthic community also favored the proliferation of scaphites. Both the scaphites and the benthic species were probably part of a larger community including worms, crustaceans, fish, and mosasaurs. The association between scaphites and benthic species further suggests that scaphites lived close to the sea floor (Landman and Waage, 1993: 25; Tsujita and Westermann, 1998: 144), but this interpretation is difficult to prove (see below). During periods of dysoxia, the scaphites, as well as the rest of the community, were excluded from the habitat.

The baculites provide an interesting comparison with the scaphites. The baculites are the most common ammonites in the Pierre Shale and Bearpaw Shale. They occur in association with benthic species but are not restricted to such associations. A notable example is the Sharon Springs Member of the Pierre Shale at Red Bird, Wyoming (Gill and Cobban, 1966). This unit contains fishes, mosasaurs, and baculites, with little or no benthic fauna aside from large flat-shelled inoceramids. The environment is interpreted as representing an anaerobic bottom with oxygenated water above (Byers, 1979). The presence of baculites in this unit suggests that they were relatively independent of the bottom. The baculites are also more geographically widespread than the scaphites in the Western Interior Seaway, making them more useful as index fossils. For example, in the Baculites compressus–B. reesidei zones in the Bearpaw Shale in Alberta, the baculites are much more widespread than the scaphites (Tsujita and Westermann, 1998: fig. 10). All this evidence suggests that the baculites lived higher up in the water column than the scaphites, and possibly made occasional forays into the dysoxic bottom waters below (Westermann, 1996).

The habitat of scaphites has also been determined on the basis of isotopic analyses of the shells (Forester et al., 1977; Tourtelot and Rye, 1969; Rye and Sommer, 1980; Wright, 1987). Whittaker et al. (1987) analyzed the oxygen and carbon isotopic composition of molluscs, including scaphites, from the lower Campanian Lea Park Formation of south-central Saskatchewan. They noted that the isotopic composition of scaphites was more similar to that of benthic inoceramids than to that of baculites, which suggested to them that scaphites lived closer to the sea floor than did baculites. He et al. (2005) examined the isotopic composition of molluscs from the Baculites scotti–B. clinolobatus zones in the Pierre Shale and Bearpaw Shale of the Western Interior. They noted that the isotopic composition of scaphites was very variable and overlapped that of didymoceratids and baculites. They assumed that baculites lived in the upper part of the water column and that didymoceratids, whose isotopic values were similar to those of benthic inoceramids, lived near the bottom. They, therefore, concluded that scaphites lived throughout the water column. Moriya et al. (2003) investigated the isotopic composition of ammonites as well as planktic and benthic foraminifera and benthic molluscs in the Campanian Yezo Group of Hokkaido, Japan. Although they did not explicitly analyze scaphites, they sampled a wide variety of ammonites including uncoiled heteromorphs. They concluded that the oxygen isotopic values of the ammonites more closely matched those of the benthic organisms than those of the planktic foraminifera, suggesting that the ammonites lived near the sea floor.

In such isotopic studies, it is important to consider the state of preservation of the samples. For example, in Whittaker et al. (1987), the samples consist of aragonitic shell material with nacreous microstructure. The authors used X-ray diffraction analysis to determine the suitability of these samples for isotopic study. However, Cochran et al. (2010) have argued that even material that is still mostly aragonite may have neverthess undergone alteration, yielding erroneous isotopic values. They proposed a screening protocol emphasizing scanning electron microscopy to assess the state of preservation of the nacre. In addition, in studies that compare the isotopic compositions of different organisms to document their respective habitats, it is important that the samples represent the same time interval. This can be approached through the examination of material from the same biostratigraphic zone or, even better, from the same concretion. Otherwise, the results can be misleading, reflecting variation in both environment and time. For example, He et al. (2005) compared the isotopic values of didymoceratids and scaphites in order to document their respective habitats, but the specimens they analyzed were collected from different biostratigraphic zones.

Habitat depths of scaphites have also been calculated based on the mechanical properties of the scaphite shell, especially its septa and siphuncle. Hewitt (1996: table 2) calculated the maximum habitat depth of several hundred ammonites including nine specimens of what he called Hoploscaphites spp. and Jeletzkytes spp. He stated that the most precise estimates were based on the mechanical strength of the septa. For Hoploscaphites spp., he reported a range of values from 53 to 185 m (both of these values are preceded by a question mark in his table). For Jeletzkytes spp., he reported a range of values from 20 (preceded by a question mark in his table) to 133 m. Tsujita and Westermann (1998) also calculated the maximum habitat depth of 11 scaphite specimens from the Baculites compressus–B. reesidei zones of the Bearpaw Shale using the same equations for septal strength as employed by Hewitt (1996). They calculated an average of 118 m for seven specimens of Hoploscaphites nodosus, 74.4 m for three specimens of H. furnivali, and 71.5 m for one specimen of H. sp. α. All these values represent maximum habitat depths and the scaphites may have, in fact, lived in shallower water. Landman et al. (2007a) estimated a habitat depth of 20–30 m for Discoscaphites iris (Conrad, 1858) from the upper Maastrichtian Tinton Formation of New Jersey, based on stratigraphic evidence and an estimate of sea level at the time.

Methane Seeps

Among the most interesting occurrences of scaphites in the Western Interior of North America are the Tepee Buttes. These unusual features in the Pierre Shale have been recognized since the late 19th century (Gilbert and Gulliver, 1895). Tepee Buttes are conical hills usually composed of limestone, up to 60 m in diameter, and 10 m in height (Fenneman, 1931). They have been documented from Montana to the Colorado–New Mexico border, and from the Front Range of the Rocky Mountains to western Kansas. They range in age from the late middle Campanian to the early Maastrichtian (Howe, 1987; Kauffman et al., 1996; Metz, 2008).

The presently accepted interpretation of the Tepee Buttes is that they were the sites of cold methane seeps. The source of the methane was nutrient rich brines or connate waters from the Pierre Shale and, possibly, the underlying Niobrara Formation. The methane was oxidized by chemosynthetic bacteria thereby increasing the concentration of CO₂ in the water, thus promoting authigenic precipitation of carbonate minerals. The presence of methane has been confirmed by isotopic analyses of the carbonate cements. Kauffman et al. (1996) reported that the carbonate cements in vent deposits from Colorado are extremely depleted in δ¹³ C (−40‰ to −45‰). Similarly negative values have been reported from Cretaceous vents in the Canadian Arctic (Beauchamp and Savard, 1992). Such negative values indicate that the carbon was derived from methane oxidation.

The seeps support an abundant and diverse community, including bivalves (notably, inoceramids and aggregations of chemosymbiotic-harboring lucinids), gastropods, ammonites, echinoderms, crabs (Bishop, 2000), sponges, tube worms, chemosynthetic bacteria, foraminifera, and radiolarians. Howe (1987) reported a diversity of approximately 30 molluscan species at single seeps from the upper Campanian of Colorado. In their studies of these sites, Kauffman et al. (1996: fig. 3) documented a zonation of macrofaunal and microfaunal assemblages distributed from the center of the vents to the adjacent sea floor, reflecting, according to them, an environmental stress gradient with decreasing H₂S. The most diverse assemblage, consisting mostly of molluscs, occurred on the upper part of the flanks, downcurrent from the immediate vicinity of the methane emissions. In contrast, the sea floor adjacent to the vents was relatively depauperate, and may have been characterized by dysoxic water conditions.

It is unknown if these seeps represented topographical highs on the sea floor, or formed below or at the sediment-water interface (Shapiro and Fricke, 2002). However, the presence of fallen slump blocks on the sides dipping away from the central conduits suggests some relief (Kauffman et al., 1996). A single seep may have perisisted over a time span of 1.25 million years (Kauffman et al., 1996). During this interval, the seep may have episodically stopped, started, collapsed, and restarted again nearby, depending on the source of the methane and the intricacies of the plumbing network.

The geographic distribution of cold seep deposits in the Western Interior of North America may be related to underlying structural features, as well as the overall tectonics of the foreland basin. In Colorado, the seep deposits are aligned along early Laramide basement faults parallel to the Front Range of the Rocky Mountains (Howe, 1987). In South Dakota, they form a ring around the Black Hills, suggesting that a series of faults existed in this area during the Late Cretaceous. Metz (2008) analyzed the geographic and temporal distribution of cold seeps in the Western Interor Basin, and suggested that they are associated with the development of the forebulge depozone. She argued that the formation of cold seeps coincided with transgressive episodes, which involved an increase in sediment loading near the orogenic belt, and an increase in the degree of flexure of the forebulge, which, in turn, promoted the migration of cold seep fluids from the underlying sediment. In contrast, the cessation of seep activity coincided with regressive episodes, which involved a decrease in sediment loading near the orogenic belt, and a decrease in the degree of flexure of the forebulge.

We examined and collected from methane seeps in the Didymoceras cheyennense and Baculites compressus zones in Custer County, South Dakota, as well as from older seeps in Butte County, South Dakota. The seep from the Didymoceras cheyennense Zone is exposed in cross section at AMNH loc. 3418. The outcrop is nearly vertical and is approximately 13 m high and 20 m wide. The central area consists of multiple, anastomosing pipelike conduits composed of limestone, surrounded by grey shale with orange-weathering partings. Fossils are abundant for a distance of approximately 6 m on either side of the central area. Some of the fossils are preserved in the sediment and others are preserved in limestone concretions. In both instances, they retain their original shell. The sediments on the outer margins of the vent, up to 20 m away, are much darker black, with fewer fossils.

The fossils immediately surrounding the central area include Hoploscaphites nodosus, H. brevis, Baculites corrugatus Elias, 1933, Didymoceras cheyennense (Meek and Hayden, 1856b), Spiroxybeloceras meekanum, Nympholucina occidentalis (Morton, 1842), “Inoceramus” altus, “Inoceramus” sagensis Owen, 1852, crinoids, and sponges (fig. 39). Many of the shells are broken, probably as a result of predatory activity. For example, a microconch of H. brevis shows a missing chunk from the adapical end of the body chamber, indicating a lethal injury (fig. 39I, J). In addition, a macroconch of Baculites corrugatus shows a series of small scalloped edges along the entire length of the body chamber on both sides, indicating that an animal, possibly a crab, systematically gnawed off pieces of the baculite shell after death (fig. 39A, B). Several pieces of evidence suggest that the shell debris in the seep was buried rapidly: (1) the phragmocones of many baculites are hollow; (2) all of the lucinids are articulated; and (3) isolated jaws of scaphites are present.

Fig. 39.

Fossil invertebrates from a cold seep in the Pierre Shale, Didymoceras cheyennense Zone, AMNH loc. 3418, Custer County, South Dakota. A, B. Baculites corrugatus Elias, 1933, mature macroconch, AMNH 58552, uncoated. Note the series of small scalloped edges along the entire length of the body chamber on both sides (arrows), probably reflecting predation. A. Ventral, with apertural projection on top; B. right lateral, with apertural projection on top. C. Didymoceras cheyennense (Meek and Hayden, 1856b), fragment, AMNH 63440. D. Lower jaw attributed to Hoploscaphites, AMNH 63423. E. Nympholucina occidentalis (Morton, 1842), AMNH 66246. F. Whorl cross section at the adoral end of a fragment of Baculites corrugatus Elias, 1933, AMNH 66245. G. Crassatella evansi Hall and Meek, 1856, AMNH 66241. H. “Inoceramus” sagensis Owen, 1852, AMNH 66248. I, J. Finely ornamented fragment of a microconch of Hoploscaphites brevis (Meek, 1876), AMNH 66242. I. Right lateral; J. ventral. Specimens are illustrated natural size unless indicated otherwise by a scale bar.

The seeps from the Baculites compressus Zone appear as carbonate mounds. We recorded five such mounds within a radius of 100 m, representing a vent field. The two largest mounds (AMNH locs. 3419 and 3420) are each approximately 5–10 m across and 3 m high. The three smaller mounds (AMNH locs. 3457, 3457a, 3457b) are grouped around AMNH loc. 3420, and are each 2–4 m across and 1–2 m high. The mounds consist of heavily cemented limestone, superficially resembling Paleozoic reefs. This high degree of cementation suggests a late stage in vent development, as described by Beauchamp and Savard (1992) for Late Cretaceous vents from the Canadian Arctic.

Because of the heavy cementation of the mounds, it is difficult to establish a zonation from the center of the vents to the margins. The mounds are riddled with conduits filled with limestone and deposits of sparry calcite. The fauna in the mounds is extremely diverse (fig. 40) and includes Hoploscaphites nodosus, H. brevis, Baculites compressus, Baculites corrugatus, Placenticeras sp., Nympholucina occidentalis, “Inoceramus” altus, “Inoceramus” sagensis, gastropods, boring teredos, sea urchins, articulate crinoids, bryozoans, sponges, serpulid worm tubes, and stromatolitelike sheets (microbialites). In contrast, the surrounding grey shale contains few fossils.

Fig. 40.

Invertebrate fossils from cold seeps in the Pierre Shale, Baculites compressus Zone, AMNH locs. 3419, 3420, Custer County, South Dakota. A. Articulate crinoid columnals, AMNH 66260. B. Small gastropods, AMNH 61130A. C. Nympholucina occidentalis (Morton, 1842), AMNH 66251. D. Sponge, AMNH 66249. E–G. Hoploscaphites brevis (Meek, 1876), part of the body chamber of a microconch, AMNH 61122. E. Right lateral; F. ventral; G. left lateral. H–J. Hoploscaphites brevis (Meek, 1876), phragmocone of a mature macroconch, AMNH 61135. H. Right lateral; I. apertural; J. ventral. K–O. Baculites compressus (Say, 1820), AMNH 58558. K. Right lateral; L. dorsal; M. ventral; N. left lateral; O. whorl cross section at the adoral end. P–T. Baculites compressus (Say, 1820), AMNH 58544. P. Right lateral; Q. dorsal; R. ventral; S. left lateral; T. whorl cross section at the adoral end. Specimens are illustrated natural size.

The species of scaphites at the methane seeps are the same as those elsewhere in the Western Interior Seaway. However, the presence of scaphites at these seeps poses some interesting questions: How did the scaphites arrive at the seeps? Did they live their entire lives at the seeps or did they wander back and forth between the seeps and the surrounding sea floor? How did the scaphites die at the seeps—by H₂S poisoning or by predators? Although we are planning further studies in the future, some data are already available to answer these questions. We have recovered juvenile and adult scaphites at the seeps from both the Didymocerass cheyennense and Baculites compressus zones. Thus, both juvenile and adult scaphites lived, died, and were buried at the seeps, although it is unclear whether a single individual lived its entire life there. In addition, we have observed specimens with both lethal and sublethal injuries, indicating that the scaphites were subject to predators, which must also have lived at the site.

To further answer these questions, we are analyzing the isotopic composition of the scaphites from the seep in the Didymoceras cheyennense Zone. Previous studies have shown that carbonate cements precipitated at methane seeps are strongly depleted in carbon (e.g., Kauffman et al., 1996). Presumably, this depletion is also reflected in the shells of the animals living at the vents, provided that the shell material is well enough preserved. These results may reveal whether the scaphites lived at the vents all the time or only visited occasionally.