RESEARCH ARTICLE
Intraspecific variation throughout ontogeny in the Cenomanian (Late Cretaceous) ammonoid taxon Neogastroplites muelleri Reeside and Cobban
2025 Volume 29 Pages 332-350
Details
2025 Volume 29 Pages 332-350
Intraspecific variation of Neogastroplites muelleri Reeside and Cobban throughout ontogeny was studied based on 100 well-preserved specimens ranging in diameter from 10 to 75 mm that were extracted from a single calcareous concretion in the Mowry Shale Formation of the Colorado Group (lower Cenomanian) in central Montana, USA. The embryonic shell is exceptionally large among Ammonitina, and the width of the initial chamber exhibits a wide range of intraspecific variation in relation to its diameter, and thus a large variation exists in the thickness of the initial chamber. There is a positive correlation during the post-hatching stage between shell growth ratios (whorl width to whorl height and umbilical diameter to shell diameter) and relative initial chamber thickness, such that those specimens with more depressed initial chambers have greater growth ratios than those that are more compressed. This may imply that the initial chamber morphology could constrain subsequent shell growth. Due to the differences in these growth ratios, the degree of intraspecific variation of relative whorl thickness and relative umbilical size increases as the shell grows. Our analysis suggests that this high level of intraspecific variation in N. muelleri is directly related to the large morphological differences observed in the initial chambers, and these corresponding differences are maintained in the shell growth ratios of the subsequent post-hatching stages. The large embryonic shell suggests that the inhospitable environment of the Mowry Sea may have reduced the survival of newly hatched ammonoids, thus allowing Neogastroplites to survive by evolving a larger hatching size. Although the cause of the large variation observed in the relative initial chamber thickness is unclear, the wide intraspecific variation observed in post-hatching shell morphology implies that the physical and biological environment did not reduce the survival of individuals with different shell morphologies. This observation suggests that the Mowry Sea may have been an environment lacking external constraints for post-hatching Neogastroplites, when compared to other post-hatching environments.
The hoplitid Neogastroplites McLearn, 1931 is an endemic ammonoid genus of the Mowry Sea, which existed as an extreme southward extension of the Arctic Ocean into North America during latest Albian to early Cenomanian time (Williams and Stelck, 1975; Slattery et al., 2015; Figure 1). Because several concretions have yielded thousands of individual Neogastroplites shells ranging from smooth and compressed to spinous and depressed, Reeside and Cobban (1960) recognized that these specimens co-occurring in the same concretion represented a continuum of morphologies of a single species rather than several distinct taxa. In the Western interior region, the genus includes four species, i.e., from oldest to youngest, Neogastroplites haasi Reeside and Cobban, 1960, N. cornutus (Whiteaves, 1885), N. muelleri Reeside and Cobban, 1960, and N. americanus (Reeside and Weymouth, 1931), each of which forms a stratigraphically continuous fossil zone in the Mowry Shale Formation as well as other geographical areas covered by the Mowry Sea. Neogastroplites muelleri is the third species in its evolutionary lineage (Reeside and Cobban, 1960; Kennedy and Cobban, 1976). Each species is widely known for its extremely high level of intraspecific variation (Reeside and Cobban, 1960; Kennedy and Cobban, 1976; Reyment and Kennedy, 1998; Reyment and Minaka, 2000; Yacobucci, 2004b).
Reeside and Cobban (1960) suggested that the absence of competitors permitted the high levels of variability seen in Neogastroplites, because it was likely the dominant ammonoid in the Mowry Sea and occupied almost all niches that other ammonoid species would ordinarily have filled. However, this hypothesis is unsubstantiated, because the introduction of a potential competitor, e.g. the engonoceratid Metengonoceras, apparently did not cause a decrease in variability in Neogastroplites (Yacobucci, 2004b).
High levels of intraspecific variation in ammonoids may be related to differences in growth rates and growth patterns throughout ontogeny, and therefore, intraspecific variability should be studied throughout all growth stages (De Baets et al., 2015a). Ammonoids and other shelled mollusks are an ideal group to study intraspecific variability during growth because they preserve a record of their ontogeny due to the accretionary mode of growth of their shell (e.g. Bucher et al., 1996; Tajika et al., 2018). However, the intraspecific variation of Neogastroplites throughout ontogeny has never been thoroughly investigated.
In this paper, we describe the ontogenetic shell development to understand the characteristics of intraspecific variation in Neogastroplites muelleri based on the Jenks collections housed at the New Mexico Museum of Natural History and Science, Albuquerque in Albuquerque, New Mexico, USA (NMMNH) and the National Museum of Nature and Science in Tsukuba, Ibaraki, Japan (NMNS).
A total of 100 specimens of Neogastroplites muelleri were examined for this study. They were extracted by one of the authors (L.C.E.) from a single calcareous concretion, ~1.2 m in diameter, from the lower Cenomanian N. muelleri Zone of the Mowry Shale Formation of the Colorado Group at USGS Locality 24065 (= map locality 20 of Reeside and Cobban, 1960) near Teigen in sec. 4, T14N, R25E, Petroleum County, Montana, USA (Figure 1). The concretion also contained two Metengonoceras specimens as well as abundant fish remains.
MethodsAnalysis of early shell morphology.—A total of 31 specimens was selected for observation of early internal shell morphology (Figure 2). These specimens were cut along a median plane that halved the siphuncular tube or its holes of the septa, which are approximately 0.1 mm in diameter and distributed along the median plane. The surfaces of these ‘median-sectioned’ specimens were etched with 5% acetic acid for a few minutes, washed with distilled water and allowed to air dry.
The etched surfaces of three of these specimens (NMNS PM 45650–45652) were coated with gold using an ion coater, and the early internal features were observed using a JEOL model JSM-5310 scanning electron microscope (SEM). The initial chamber diameter (Di) and shell size (= ammonitella diameter: AD) at the primary constriction were measured from the SEM images.
For the remaining 28 specimens (NMMNH P-103518–103545), acetate peels of the etched surfaces were made by pressing a sheet of triacetylcellulose film (0.034 mm in thickness) onto the etched surface while flooded with acetone. Subsequently, these 28 ‘median-sectioned’ specimens were cut perpendicular along a line passing through the position of the proseptum on the ventral side and the center of the initial chamber to prepare the ‘half cross-sectioned’ specimens. These half cross-sectioned surfaces were also etched with 5% acetic acid for a few minutes, and acetate peels of the etched surfaces were prepared. The early internal features were observed under a profile projector (Nikon Model V-20B). The initial chamber diameter (Di), one half of initial chamber width (Wi/2) and ammonitella diameter (AD) were measured using a digital micrometer (accuracy, ± 0.001 mm) attached to the profile projector on images magnified 100 times, and the ratio (Wi/Di: relative initial chamber thickness) was calculated.
Biometric analysis of shell morphology.—A total of 97 specimens was examined for biometric analysis of shell morphology. For the 28 half cross-sectioned specimens (NMMNH P-103518–103545, see above), four standard geometric parameters of the shell (Figure 2), i.e., shell diameter (D), umbilical diameter (U), whorl height (H) and one half of whorl width (W/2), were measured on images magnified 50 times at intervals of every half whorl using a digital micrometer (accuracy, ± 0.001 mm) attached to the profile projector. From this data, three ratios, relative umbilical size (U/D), relative whorl thickness (W/H) and whorl expansion rate (WER = [D/D’] 2, where D’ equals the diameter of the preceding whorl; see Klug et al., 2015), were calculated. The remaining 69 specimens (NMNS PM 35140–35173, NMMNH P-94572–94575, 94576a, 94576b, 94577–94579, 103486, 103493–103517) were scanned using X-ray computed tomography (inspeXio SMX-225CT FPD HR, Shimadzu) at NMNS with settings of 0.022–0.091 mm Boxel size, 225 kV and 70 μA. Four parameters (D, U, H, W) were measured at intervals of every half whorl using an X-ray CT image of the cross section (File S1) and the three ratios (U/D, W/H, WER) were calculated.
Analysis of individual relative growth.—The relationship between X and Y in two parts of a growing system is empirically expressed as Y=bXα, where α is the growth ratio (relative growth coefficient) and b is the initial growth constant (e.g., Thompson, 1917; Yamagishi, 1977). We calculated the allometric equations between each shell geometric parameter (D, D’, U, W, H) for each specimen to determine α.
Analysis of intraspecific variation.—We used the coefficient of variation (CV), which is the standard deviation divided by the mean, as a measure of the range of intraspecific variation of shell characters (e.g. De Baets et al., 2015b). Because U, H, and W were measured at various shell diameters (D) during individual growth, we first calculated the allometric equations between these geometric parameters and D to obtain the U, H and W at a particular diameter and then calculated U/D and W/H. For WER, we calculated the allometric equations between the D and WER and calculated WER at a particular diameter. Using these values, we calculated the CV of the three parameters at a particular diameter. We also calculated the CV of the Di/Wi to show variation in initial chamber morphology.
Measurement error.—In order to know the range of human error in the measurement of the early initial shell features, the initial chamber diameter (Di) of a specimen (NMMNH P-103518) was measured 20 times by one of the authors (Y.S.). The measurements were taken using a digital micrometer (accuracy, ±0.001 mm) and the profile projector to magnify the replica film of the median section 100 times. The measurement data ranged from 0.754 to 0.764 mm (mean 0.760 mm, standard deviation 0.003 mm), thus indicating small variation and high measurement precision. Therefore, considering that there was little human error, measurements of not only the initial shell features but also the standard geometric parameters of the shell were performed only once.
Uncertainty related to missing the median plane.—The ‘median-sectioned’ specimens were prepared by cutting in half the siphuncular tube or its holes of the septa in half, which are approximately 0.1 mm in diameter and distributed along the median plane of the whorls succeeding the initial chamber. The initial chamber is spindle-shaped, circular in median section and elliptical in cross section. Even if the median section of the specimen deviated from the true median section by approximately 0.05 mm, which is the radius of the siphuncular tube, the initial chamber diameter (Di) slightly decreases because the curvature is small. For example, in specimen NMMNH P-103518, the Di at a position 0.05 mm off the median plane decreased by approximately 0.005 mm when measured on image magnified 100 times using a digital micrometer (accuracy, ± 0.001 mm) attached to the profile projector. Therefore, in the ‘median-sectioned’ specimen, a deviation from the median plane of about half the diameter of the siphuncular tube, i.e., about 0.05 mm, does not significantly affect the value of the Di. In the ‘cross-sectioned’ specimens, the shape of the siphuncular tubules can be observed, allowing the location of the median plane to be estimated and one half of initial chamber width (Wi) and whorl width (W) to be accurately measured, even in specimens that are off the median plane.
The examined specimens range from smooth, compressed forms to spiny, depressed forms (Figures 3, 4). The cross-sections of the whorls are symmetrical but may appear slightly asymmetrical due to ribs, nodes or spines. Flanks vary from slightly convex in compressed forms to sub-angular in depressed forms. In some of the depressed forms, the venter is well defined, slightly convex, and merges with the outer part of the flank to form a broad ventral zone. The umbilicus in compressed forms is narrow and fairly deep with a gently convex, oblique wall, while in depressed forms, it is fairly narrow and deep with a gently convex, sub-vertical wall. The umbilical shoulders are always rounded. Ornamentation consists of strong, rounded, prorsiradiate, sigmoidal ribs. Intercalation occurs at mid-flank. Ribs bend gently forward on the ventrolateral shoulders, before crossing venter in a convex arch. Although not examined statistically, the number of ribs per whorl tends to be greater in the compressed forms than in depressed forms. In depressed forms, the primary ribs form a node or spine elongated radially at mid-flank, a transverse node (clavus) on the ventral shoulder, and a siphonal node at mid-venter. In compressed forms, the nodes are weakened.
All ammonoids possess a microscopic early shell, termed the ammonitella, which consists of an initial chamber and approximately one planispiral whorl that terminates at the primary constriction; the prosiphon, caecum and siphuncular tube are also observed within the early shell (e.g. Ohtsuka, 1986; Landman et al., 1996; De Baets et al., 2015b). These early shell features are also observed in the Neogastroplites muelleri specimens described herein (see Figure 2).
The caecum is elliptical in lateral view, and its adapical end is connected to the initial chamber wall by a long, straight prosiphon (Figures 2, 5). The siphuncle is located sub-centrally in each camera in the first half whorl from the initial chamber, but then gradually shifts toward the venter within the second whorl and maintains a ventral position thereafter (Figure 5).
The measurements of the specimens of Neogastroplites muelleri studied herein are shown in File S2.
Initial chamber and ammonitella.—The initial chamber of Neogastroplites muelleri is spindle-shaped (Figure 6) with a diameter (Di) ranging from 0.703 to 0.827 mm, width (Wi) ranging from 1.034 to 1.284 mm, and the ratio Wi/Di ranging from 1.31 to 1.67. Di has no clear relationship with Wi [Figure 7A; r (correlation coefficient) = 0.084, P (p-value) = 0.673], but Di has a weak negative linear relationship with Wi/Di (Figure 7C; r = −0.440, P < 0.05) and Wi has a positive strong linear relationship with Wi/Di (Figure 7D; r = 0.858, P < 0.01), i.e., the more compressed specimens have a larger Di and smaller Wi, and conversely, the more depressed specimens have a smaller Di and larger Wi. The diameter (AD) ranges from 1.428 to 1.746 mm and has a positive linear relationship with Di (Figure 7B, r = 0.781, P < 0.01).
Relative umbilical size.—At a shell diameter of about 2 mm, each specimen exhibits a narrow to fairly narrow umbilicus (U/D = 0.15–0.21, Figure 8A). As the shell grows, the U/D of specimens with a wider umbilicus gradually decreases, but the U/D of specimens with a narrower umbilicus at first increases up to a diameter of about 5 mm then gradually decreases (Figure 9A). The U/D is 0.10–0.20 at a diameter of 40 mm (Figure 8A), but specimens with a wider umbilicus at a shell diameter of about 2 mm tend to have a smaller U/D than specimens with a narrower umbilicus (Figure 9A).
Relative whorl thickness.—At the ammonitella stage before the primary constriction, each specimen exhibits a very depressed whorl (W/H = 1.52–2.11; Figure 8B). As the shell grows, the whorl section becomes progressively more compressed (Figure 9B), and at a diameter of about 40 mm, each specimen exhibits a fairly compressed to fairly depressed whorl (W/H = 0.62–1.43; Figure 8B).
Whorl expansion rate.—At the ammonitella stage, each specimen exhibits a whorl expansion rate (WER) of 1.86–2.49 (Figure 10). After the primary constriction, the WER slightly decreases, gradually increases up to a diameter of 10 mm as the shell grows, then slightly decreases, reaching a WER of 1.85–2.68 at a diameter of 40 mm (Figures 10, 11).
The values of the growth ratio (α) for each Neogastroplites muelleri specimen’s allometric equation are shown in File S3.
U to D.—The α values of the umbilical diameter (U) to shell diameter (D) for those specimens examined from the post-ammonitella stage after the primary constriction up to 5 mm in D has a wide range of variation, i.e., from 0.74 to 1.51 (N = 28) with a mean of 0.95, indicating negative to positive allometry (Figure 12A). For specimens larger than 5 mm in D, the α value ranges from 0.46 to 0.98 (N = 90) with a mean of 0.77, indicating nearly isometry to negative allometry (Figure 12B), and thus, the relative umbilical size (U/D) either remains unchanged or decreases with growth.
H to D.—The α value of whorl height (H) to shell diameter (D) for the post-ammonitella stage for the examined specimens ranges from 0.91 to 1.28 (N = 92) with a mean of 1.04, indicating nearly isometry to positive allometry (Figure 13A), and thus, the relative whorl height (H/D) remains either unchanged or increases with growth.
W to D.—The α value of whorl width (W) to shell diameter (D) for the post-ammonitella stage for the examined specimens ranges from 0.68 to 0.98 (N = 92) with a mean of 0.79, indicating nearly isometry to negative allometry (Figure 13B), and thus, the relative whorl width (W/D) remains either unchanged or decreases with growth.
W to H.—The α value of whorl width (W) to whorl height (H) for the post-ammonitella stage for the examined specimens ranges from 0.62 to 0.97 (N = 92) with a mean of 0.77, indicating nearly isometry to negative allometry, and thus, the relative whorl thickness (W/H) decreases with growth. The α value tends to be greater for depressed shells than for the compressed ones (Figure 13C).
D to D’.—The α value of shell diameter (D) to shell diameter before the half whorl (D’) of the post-ammonitella stage for the examined specimens ranges from 0.92 to 1.12 (N = 66) with a mean of 1.02, indicating nearly isometric growth (Figure 13D).
Intraspecific variation through ontogenyRelative umbilical size.—The coefficient of variation (CV) of the relative umbilical size (U/D) is 0.08 at a diameter of 3 mm, but gradually increases with growth, becoming 0.17 at a diameter of 20 mm, and then remains almost constant (Figure 14; Files S4, S5).
Relative whorl thickness.—The CV of the relative initial chamber thickness (Wi/Di) is 0.08. After the primary constriction, the CV of the relative whorl thickness (W/H) is 0.08 at a diameter of 2 mm but then gradually increases with growth to 0.23 at a diameter of 30 mm (Figure 14; Files S4, S5).
Whorl expansion rate.—The CV of the whorl expansion rate (WER) is 0.03 at a diameter of 2 mm and gradually increases with growth to 0.08 at a diameter of 30 mm (Figure 14; Files S4, S5).
Most recent workers consider that the ammonitella, consisting of the initial chamber and approximately one planispiral whorl terminating at the primary constriction, was formed within the egg-capsule as an embryonic shell and that the ammonoid hatched directly without going through a post-hatching larval stage, similar to extant cephalopods (e.g. Landman et al., 1996; De Baets et al., 2015b).
The initial chamber size (Di) and ammonitella size (AD) of Neogastroplites muelleri (Di = 0.703–0.827 mm, AD = 1.428–1.746 mm; Table 1) are exceptionally large compared to Jurassic and Cretaceous Ammonitina (Di = 0.24–0.75 mm, AD = 0.48–1.52 mm); relatively large compared to Lytoceratina (Di = 0.32–1.05 mm, AD = 0.88–1.90 mm); and much larger than those of Phylloceratina (Di = 0.38–0.67 mm, AD = 0.63–1.30 mm) except for some Boreal taxa (Di = 1.14–1.54 mm, AD = 2.37–2.90 mm) and Ancyloceratina (Di = 0.25–0.58 mm, AD = 0.52–0.95 mm) (Landman et al., 1996; De Baets et al., 2015b). According to those authors, a statistically significant positive linear relationship exists between the Di and AD (e.g. Landman et al., 1996).
| taxon | N | CV | register number | locality | horizon | ||
|---|---|---|---|---|---|---|---|
| Di | Wi | Wi/Di | |||||
| Neogastroplites muelleri | 28 | 0.036 | 0.063 | 0.068 | NMMNH P-103518–103545 | Montana | Cenomanian |
| Damesites damesi | 13 | 0.048 | 0.037 | 0.039 | NMNS PM 45653–45665 | Hokkaido | Santonian |
| Tetragonites glabrus | 18 | 0.041 | 0.046 | 0.041 | NMNS PM 45666–45683 | Hokkaido | Santonian |
| Yezoites pseudoequalis | 10 | 0.032 | 0.052 | 0.039 | NMNS PM 45684–45693 | Hokkaido | Coniacian |
It has been reported that the size of the initial chamber and ammonitella changed with latitude and global climate changes. Drushchits and Doguzhaeva (1981) reported that the sizes of these same parts of late Mesozoic boreal ammonoids are larger than those of warm-water species. Landman et al. (1996) and Laptikhovsky et al. (2013) showed that these sizes decreased during the Turonian global warming but began to increase again during the gradual Turonian–Maastrichtian cooling. A factor affecting ecophenotypic plasticity of embryogenesis as well as egg and hatchling sizes in Recent cephalopods is temperature (e.g. Boyle and Rodhouse, 2005). Paleotemperature may be an important factor influencing historical changes in the evolution of embryonic shell and egg sizes in ammonoids (Laptikhovsky et al., 2013). The Mowry Sea existed as a southward extension of the Arctic Ocean into North America during latest Albian to early Cenomanian time (Williams and Stelck, 1975; Slattery et al., 2015; Figure 1), and its water temperature may have influenced the size of the Neogastroplites embryonic shell.
Some models have suggested that the increased egg size may be an evolutionary response to the reduced chances of juvenile survival because of the physical and biological environment faced by hatchlings (Smith and Fretwell, 1974; Kolding and Fenchel, 1981; Sibly and Calow, 1986; Clutton-Brock, 1991). Compared to contemporary sites elsewhere, the Mowry fauna was of markedly low diversity (Reeside and Cobban, 1960), an observation thought to be due to its restricted connection to the Arctic Ocean and its relatively inhospitable environment (Yacobucci, 2004b). This environment probably reduced the survival of newly hatched ammonoids, but the newly evolved larger hatching size of Neogastroplites likely at least partly compensated for this harmful environment.
Intraspecific variation in initial chamber thicknessThe coefficient of variation (CV) of the relative initial chamber thickness (Wi/Di) of Neogastroplites muelleri is 0.068. Previously, the intraspecific variation of this particular variable has rarely been reported for ammonoids (e.g. De Baets et al., 2015b, table 5.2). We investigated the CV of Di, Wi, and Wi/Di for the relatively large number of specimens of the following three Cretaceous ammonoid taxa stored at NMNS (Table 1): 1) Damesites damesi (Jimbo, 1894) belonging to Ammonitina, 2) Tetragonites glabrus (Jimbo, 1894) belonging to Lytoceratina, and 3) Yezoites pseudoequalis (Yabe, 1910) belonging to Ancyloceratina. The procedures for preparation and measurement of the ‘median-sectioned’ and ‘cross-sectioned’ specimens of these taxa were the same as those used for Neogastroplites specimens in this study.
The CV of Di for Neogastroplites muelleri (0.036) is similar to those of the three comparative species (0.032–0.048), but the CVs of Wi and of Wi/Di (0.063 and 0.068, respectively) are much larger than those of the comparative species (0.037–0.046 and 0.039–0.041, respectively). This indicates that N. muelleri has a larger variation in Wi than in Di, which has resulted in a larger variation in Wi/Di, although it is unclear why the variation in Wi is so large.
The CV of Di and Wi for several Late Cretaceous species of Scaphites from the Western Interior of North America (0.032–0.121 and 0.052–0.099, respectively; Landman (1987, tables 2, 5) is similar or much larger than those of Neogastroplites muelleri, but the Wi/Di and its CV cannot be calculated because these measurements were not taken on the same specimens. In cases in which both Di and Wi were measured on the same Jurassic oppeliid specimens, both CVs (0.026–0.086 and 0.019–0064, respectively; Palframan, 1966, 1996) were similar or greater than those of N. muelleri, but the CV of Wi/Di was smaller (0.031–0.054). However, since the measurement methods and accuracy differ significantly from those in this study, caution is required when comparing CV values.
Intraspecific variation through ontogenyFor Neogastroplites muelleri, the Wi/Di has a positive linear relationship with the growth ratio of W to H (Figure 15A, r = 0.653, P < 0.01) and U to D (Figure 15B, r = 0.564, P < 0.01), i.e., those with more compressed initial chambers have smaller growth ratios of W to H and U to D than those with more depressed initial chambers. Thus, those ammonitella with a larger initial chamber diameter and smaller initial chamber width develop more compressed shells with a narrower umbilicus after hatching than those with a smaller initial chamber diameter and wider initial chamber width. This observation may imply that the initial chamber morphology could constrain subsequent shell growth.
The intraspecific variation of relative whorl thickness (W/H) gradually increases with growth, at least up to a shell diameter of 30–40 mm (Figure 14). Because the growth ratio of W to H in a depressed shell tends to be greater than that in a compressed shell (Figure 13C), intraspecific variation necessarily increases throughout ontogeny (Figures 8b, 9b, 14). The growth ratio of U to D also varies among specimens, tending to be larger for depressed shells than for compressed shells (Figure 12B), such that the intraspecific variation in relative umbilical size (U/D) increases with growth (Figures 8a, 9a, 14). Because the growth ratio of D to D’ differs slightly among specimens (Figure 13D), the intraspecific variation in whorl expansion ratio (WER) increases slightly with growth (Figures 10, 11, 14).
In summary, our analysis suggests that the high level of intraspecific variation in Neogastroplites muelleri is directly related to the large morphological differences in the initial chambers and that the corresponding differences are maintained in the subsequent shell growth ratios of the post-hatching stages.
Intraspecific variation in the shell of Neogastroplites muelleri increases with growth and reaches the highest level at shell diameters of 30–40 mm (Figure 14). Adult individuals larger than 75 mm diameter are relatively few and not available for this study, but our specimens all fall within the median growth stage. Several previous studies have reported that intraspecific variation is the greatest during the middle growth stage (Dagys and Weitschat, 1993; Korn and Klug, 2007; De Baets et al., 2013; Inose and Watanabe, 2025), and the results of our study are consistent with these observations.
As suggested earlier, this high level of variability likely resulted from the absence of competitors, which allowed this ammonoid to become dominant (Reeside and Cobban (1960)). However, according to the fossil record, the entry of the latecomer Metengonoceras into the Mowry Sea had no adverse effect on the variability of Neogastroplites (Yacobucci, 2004b).
The high level of intraspecific variation was maintained throughout the evolutionary lineage of Neogastroplites for at least 1 to 2 million years (Kennedy and Cobban, 1976, p. 40, text-fig. 9). The Mowry Shale Formation contains a variety of depositional environments, with widely varying fossil occurrences and preservation depending on the location (Reeside and Cobban, 1960). The fact that similar morphological variation has been maintained over a wide range of time and space suggests that morphological variation was not directly linked to the differing local environments of the Mowry Sea. Yacobucci (2004b) pointed to environmental stresses and instabilities, such as variable oxygen levels, salinity and current energy levels, as well as intrinsically ‘plastic’ developmental growth programs, as explanations for the increased variability. The wide intraspecific variation observed in post-hatching shell morphology implies that the physical and biological environment did not reduce the survival of individuals with different shell morphologies. It might suggest that the Mowry Sea was an environment lacking external constraints for post-hatching Neogastroplites.
Buckman’s first law of covariationIt is well known that in ammonoids, the more evolute a shell, the more depressed and ornamented it will be. This is known as Buckman’s first law of covariation (Westermann, 1966; Monnet et al., 2015). This law has been observed in many previous studies of intraspecific variation of Neogastroplites (Reeside and Cobban, 1960; Reyment and Kennedy, 1998; Reyment and Minaka, 2000; Yacobucci, 2004b). In our study, we also observed that specimens with more evolute and depressed shells tended to be more ornamented (Figures 3, 4), but a statistical analysis of this phenomenon was not performed.
According to Hammer and Bucher (2005), since the size of lateral and ventral ornamentation correlates with the size of the aperture (its width and height, respectively), the diameter of the aperture in a very compressed shell is small in the lateral direction, whereas in a depressed shell, its diameter is small in the dorsoventral direction. They suggested that Buckman’s first law of covariation is simply a statement of proportionality that needs no special explanation. That is, the law highlights mechanical constraints on the construction of ammonoid ornamentation (Hammer and Bucher, 2005).
Hammer and Bucher (2005) also predicted that more compressed specimens should have stronger ventral ornamentation based on proportionality. They tested this hypothesis for the Middle Triassic ammonoid Pseudodanubites halli (Mojisisovics) from the Fossil Hill Formation, northwestern Nevada, USA., and demonstrated that the ventral rib height correlates with whorl height, but the correlation between the lateral rib height and whorl width is weak or not significant. In the case of Neogastroplites, we observed that compressed forms have weak siphonal nodes (Figure 3B), while depressed forms have strong siphonal nodes (Figure 3J), which is the opposite of the prediction of Hammer and Bucher (2005).
Yacobucci (2004a) demonstrated a surprising lack of correlation between shell shape and variability of rib character for Cenomanian acanthoceratid ammonoid genera of the Western Interior and speculated that the morphogenesis of ornamentation is more constrained by genetic and/or developmental processes than by shell shape. The variability in the strength of the siphonal nodes seen in Neogastroplites tends to support Yacobucci (2004a) rather than Hammer and Bucher (2005).
1. Ontogenetic shell development and intraspecific variation of Neogastroplites muelleri throughout ontogeny were studied based on 100 well-preserved specimens (diameter ranges from 10 to 75 mm) extracted from a single calcareous concretion in the lower Cenomanian of the Mowry Shale Formation of the Colorado Group at USGS Locality 24065, Petroleum County, Montana, USA.
2. The embryonic shell of Neogastroplites muelleri is exceptionally large among Ammonitina, which suggests that the inhospitable environment of the Mowry Sea likely reduced the survival of newly hatched ammonoids, thus allowing Neogastroplites to survive by evolving a larger hatching size. The width of the initial chamber exhibits a wide variation in relation to its diameter, and thus, a large variation exists in the thickness of the initial chamber.
3. In the post-hatching stages, the relative whorl thickness becomes progressively more compressed as the shell grows and the relative umbilical size decreases. However, in some cases, the relative umbilical size increases at first and whorl expansion rate decreases slightly.
4. There is a positive correlation during the post-hatching stage between shell growth ratios (whorl width to whorl height, umbilical diameter to shell diameter) and relative initial chamber thickness such that those specimens with more depressed initial chambers have greater growth ratios than those that are more compressed. This may imply that the initial chamber morphology constrained subsequent shell growth. Due to the differences in these growth ratios, the degree of intraspecific variation of relative whorl thickness and relative umbilical size increases as the shell grows.
5. Our analysis suggests that this high level of intraspecific variation in Neogastroplites muelleri may be related to the large morphological differences observed in the initial chambers and the corresponding differences in shell growth ratios of the post-hatching stages.
6. The wide intraspecific variation observed in post-hatching shell morphology implies that the physical and biological environment did not reduce the survival of individuals with different shell morphologies. This observation suggests that the Mowry Sea may have been an environment lacking external constraints compared to other environments for post-hatching stages of Neogastroplites.
7. With the exception of siphonal node strength, Buckman’s first law of covariation was observed. That is, specimens with more evolute and more depressed shells tend to be more ornamented, but a statistical analysis of this trend was not performed in this study.
We are deeply indebted to the New Mexico Museum of Natural History and Science (Albuquerque, New Mexico, USA) for kindly providing the opportunity to examine the specimens. We thank K. De Baets (University of Warsaw, Warsaw, Poland) and H. Maeda (Kyushu University, Fukuoka, Japan) for valuable comments on the first draft. Thanks are extended to G. Shinohara, S. Nomura and T. Kutsuna for their efforts in the proper maintenance of the micro-CT scanner and software in the National Museum of Nature and Science Research Wing (NMNS, Tsukuba).
File S1, X-ray CT images of whorl cross sections of 69 specimens (NMNS PM 35140–35173, NMMNH P-94572–94575, 94576a, 94576b, 94577–94579, 103486, 103493–103517) used in the biometric analysis of shell morphology of Neogastroplites muelleri are illustrated. They were scanned using X-ray computed tomography (inspeXio SMX-225CT FPD HR, Shimadzu) at the National Museum of Nature and Science, Tsukuba with settings of 0.022–0.091 mm Boxel size, 225 kV and 70 μA. File S2, measurements (in mm) of herein studied specimens of Neogastroplites muelleri. D, shell diameter; Di, initial chamber diameter; AD, ammonitella diameter; U, umbilical diameter; H, whorl height; W, whorl width; Wi, initial chamber width; WER, whorl expansion rate. File S3, growth ratios between each geometric parameter of the shell for Neogastroplites muelleri specimens studied herein; D, shell diameter; D’, shell diameter before the half whorl; U, umbilical diameter; H, whorl height; W, whorl width. The coefficient of determination for most of the regression lines is greater than 0.98. File S4, calculated values of geometric parameters at a particular diameter of herein studied specimens of Neogastroplites muelleri, which were calculated from allometric equations between measured geometric parameters and shell diameter. D, shell diameter; U, umbilical diameter; H, whorl height; W, whorl width; WER, whorl expansion rate. File S5, coefficient of variation (CV) of herein studied specimens of Neogastroplites muelleri. D, shell diameter; Di, initial chamber diameter; U, umbilical diameter; H, whorl height; N, sample size; W, whorl width; Wi, initial chamber width; WER, whorl expansion rate.
L. C. E. collected the fossils and contributed to the geological aspect of the study. J. F. J. prepared the specimens. Y. S. conducted the study of intraspecific variation. All authors contributed to the writing of the paper.
