Ammonoid extinction versus nautiloid survival: Is metabolism responsible? Open Access

Understanding the relationship between environmental perturbations, physiological traits of organisms, and their evolutionary consequences is important because it allows us to predict the impact of anthropogenic climate change on current ecosystems. Mass extinction events can provide important insights, and a large number of studies have attempted to determine the biotic and/or abiotic factors, kill mechanisms, and selective nature of such events (e.g., Schulte et al., 2010). One of the “Big Five” mass extinction events that has attracted the attention of both the general public and researchers is the Cretaceous-Paleogene (K-Pg) mass extinction.

In the marine realm, ammonoids and nautiloids, both of which possessed an external shell, are iconic examples of selective extinction. Although ammonoids were more diverse and abundant than nautiloids during the Late Cretaceous, ammonoids became extinct at the K-Pg boundary (or shortly thereafter; see Landman et al., 2014), whereas nautiloids have survived to the present day. One possible explanation for this selective extinction is the difference in life cycles between the two groups. Ammonoids hatched at a smaller size (~1 mm) and probably spent some time in the plankton after hatching (Tajika et al., 2018), whereas nautiloids hatched at a larger size (~10 mm) and were nektobenthic after hatching (Gallagher, 1991). Therefore, ammonoids would have been more vulnerable to surface-water acidification following the Chicxulub asteroid impact, as proposed by Alegret et al. (2012) and Henehan et al. (2019).

Metabolic rate is another trait that can affect differential survival (Tajika et al., 2020). It dictates the energy requirements of an organism to sustain minimum activity and, thus, is linked to the energy allocation within an environment. Evidence for metabolic rate is not directly preserved in the fossil record, and a proxy must be used to reconstruct the metabolic rates of extinct organisms, especially those with no living descendants. Chung et al. (2021) proposed a new proxy to determine the metabolic rate in modern cephalopods using the stable carbon isotope composition of the shell. The principles of this method are: (1) the isotope ratio ¹³C/¹²C (expressed as δ¹³C) is a function of the isotope composition of carbon in the dissolved inorganic carbon (DIC) reservoir in which the shell formed as well as carbon incorporated via metabolism (McConnaughey et al., 1997), and (2) the fraction of metabolic carbon in the δ¹³C of the shell is correlated with metabolism-related factors such as temperature and developmental stage. We apply this method to determine the metabolic rate in a variety of modern shell-bearing cephalopods and compare our results with published estimates of metabolic rate based on oxygen consumption in order to validate the approach. We then apply this method to Late Cretaceous shell-bearing cephalopods, namely ammonoids and nautiloids, to reveal their enigmatic ecology.

For modern shell-bearing cephalopods, we selected 10 species (Nautilus macromphalus, Nautilus pompilius, Argonauta argo, Illex illecebrosus, Dosidicus gigas, Architeuthis dux, Sepia elegans, Sepia orbignyana, Sepia officinalis, and Spirula spirula), drawing on data from Chung et al. (2021). The sources of these specimens are listed in the Supplemental Material¹.

For fossil shell-bearing cephalopods, we chose three time-rock intervals: the upper Maastrichtian Discoscaphites iris Zone of the Owl Creek Formation in Mississippi, USA; the lower Maastrichtian Baculites baculus–Baculites grandis Zones of the Pierre Shale in Montana, USA; and the lower Albian Douvilleiceras mammillatum Zone of the Ambarimaniga Formation in Madagascar. For the upper Maastrichtian and lower Albian localities, we compiled data for δ¹³C_shell for ammonoids, nautiloids, and bivalves based on previously published papers (Hoffmann et al., 2019; Sessa et al., 2015). For the lower Maastrichtian locality, in addition to the data published by Landman et al. (2018), we collected shell samples and analyzed them for δ¹³C (Fig. 1; see the Supplemental Material for more details). All samples were composed of aragonite and were examined under scanning electron microscope (SEM) to determine the degree of alteration. We employed the preservation index (PI) of Cochran et al. (2010) and selected only samples with PI ≥3 (Supplemental Material).

The Maastrichtian study sites represent relatively shallow water. The upper Maastrichtian Owl Creek Formation in Mississippi was deposited in a nearshore environment between 70 and 150 m deep (Sessa et al., 2015). The lower Maastrichtian Pierre Shale in Montana was deposited in the Western Interior Seaway at a depth of <70 m (Landman et al., 2018). This locality was probably also influenced by freshwater influx associated with the progradation of the Sheridan Delta into the area during the late Campanian and early Maastrichtian (Landman et al., 2020). The fossils from the lower Albian Ambarimaniga Formation occur in a condensed section reflecting multiple transgressive systems tracts (Zakharov et al., 2016). The exact depth of deposition is unclear, but Hoffmann et al. (2019) suggested a depth of ~525 m, much deeper than the two Maastrichtian localities.

where ε is the δ¹³C fractionation between the shell and DIC (+2.7‰ ± 0.6‰ [aragonite], 1.0‰ ± 0.2‰ [calcite], dominated by HCO³−) (Romanek et al., 1992), C_meta is the fraction of metabolic carbon incorporated into the shell, and δ¹³C_shell, δ¹³C_meta, and δ¹³C_DIC are the δ¹³C values corresponding to the shell, metabolic carbon, and seawater DIC, respectively.

We calculated C_meta in modern shell-bearing cephalopods using the same values of δ¹³C_meta and δ¹³C_DIC as Chung et al. (2021) (see the Supplemental Material for more details). Because no data on the δ¹³C_meta of extinct cephalopods are available, we used −17‰, based on the value of siphuncular material in modern Nautilus (~−21‰ to −12‰; Crocker et al., 1985; Pape, 2016; see Tajika et al., 2022, for discussion). In addition, because direct measurements of δ¹³C_DIC in ancient oceans are not possible, we calculated the average value of δ¹³C_DIC using the δ¹³C_shell of co-occurring bivalves. Following Tobin and Ward (2015), we assumed that the δ¹³C_meta and C_meta of bivalves equal −19‰ and 10%, respectively. To test for a significant difference in C_meta among cephalopods, both for modern taxa and for fossil taxa from a single time interval, we performed an analysis of variance (Welch’s ANOVA) for all taxa with sample sizes >10. This test was followed by a multiple comparisons test to determine which pair of taxa exhibited significant differences. The Welch’s t-test was performed for the lower Maastrichtian taxa. All statistical tests were performed using the SciPy library within the Python 3.9.7 programming environment.

The calculated fraction of metabolic carbon (C_meta) in the δ¹³C_shell of modern cephalopods is shown in Figure 2 and the Supplemental Material. The Welch’s ANOVA for six species (N. macromphalus, N. pompilius, A. argo, D. gigas, Sepia officinalis, and Spirula spirula) reveals that there are significant differences among them (p < 0.000001). The two species of Nautilus exhibit values of C_meta (6%–35%) that are significantly lower than those of the other four species. The values of C_meta in Nautilus are followed by those of two internally shelled coleoids, Spirula spirula and Sepia officinalis (7%–35%); an octopod with an egg case, A. argo (25%–36%); and a gladius-bearing squid, D. gigas (34%–58%). The values within each group are significantly different from each other. Within any single taxon, the variation in C_meta is high. Additional details are reported in the Supplemental Material.

The calculated fractions of metabolic carbon in ammonoids and nautiloids from the three time-rock intervals are shown in Figure 3 and in the Supplemental Material. In the upper Maastrichtian Discoscaphites iris Zone of the Owl Creek Formation in Mississippi, the single data point for the nautiloid Eutrephoceras cf. dekayi falls at the lower end of the range for all ammonoid species (Fig. 3A). In the lower Maastrichtian Baculites baculus–Baculites grandis Zones of the Pierre Shale in Montana, there is a significant difference between the nautiloid Eutrephoceras dekayi (10%–23%) and all ammonoid species including Hoploscaphites macer and Hoploscaphites criptonodosus (7%–55%) (Fig. 3B). If we assume that the δ¹³C_DIC we used to calculate the C_meta values for each time interval is correct, we can compare the fraction of metabolic carbon between taxa in each of these two time intervals. Our results reveal that the values for E. dekayi are significantly lower than those for all late Maastrichtian ammonoid species (Supplemental Material). In contrast, the values for the nautiloid Cymatoceras sp. from the lower Albian D. mammillatum Zone of the Ambarimaniga Formation in Madagascar are nearly the same as those for the co-occurring ammonoids (Fig. 3C; Supplemental Material).

As noted above, calculation of the fraction of metabolic carbon (C_meta) is dependent on the value of δ¹³C_DIC. We calculated the δ¹³C_DIC using co-occurring bivalves, assuming 10% contribution of metabolic carbon and −19‰ for δ¹³C_meta (Tobin and Ward, 2015). However, ammonoids and nautiloids were nektobenthic, whereas bivalves lived on or in the substrate. Hence, for the upper Maastrichtian Owl Creek Formation, we used δ¹³C data for planktic and benthic foraminifera (Sessa et al., 2015) to assess the effect of water-column variation in δ¹³C_DIC on our results. We hypothesized that the δ¹³C of planktic and benthic foraminifera is comparable to the δ¹³C_DIC of surface and bottom water, respectively, and used the average and median values of δ¹³C to calculate the maximum possible range of δ¹³C_DIC of the water column (0.76‰–1.61‰).

This approach cannot be applied to the early Maastrichtian because data on foraminifera are unavailable. Indeed, the bivalve-based δ¹³C_DIC for the early Maastrichtian (~−1.4‰) is lower than that for the late Maastrichtian. We attribute this difference to the proximity of the study area to the Sheridan Delta and the likelihood of freshwater input into the seaway, thus lowering the δ¹³C_DIC (for further discussion, see Cai et al., 2020). However, assuming that the range of δ¹³C_DIC in the early Maastrichtian is similar to that of the late Maastrichtian, the range of δ¹³C_DIC is −1.05‰ to −1.94‰, respectively. Figure 4 shows the comparison between early and late Maastrichtian cephalopods, incorporating the potential variation in δ¹³C_DIC. These results suggest that the variation in δ¹³C_DIC does not significantly affect the comparative results.

The early Albian site from Madagascar represents a deeper-water setting than the two Maastrichtian sites. For example, Hoffmann et al. (2019) estimated that the nautiloid Cymatoceras sp. inhabited a maximum water depth of 250 m and that the ammonoids Cleoniceras and Desmoceras inhabited a maximum water depth of 450–500 m. Thus, all these cephalopods lived in a deeper-water habitat and may have universally experienced a lower metabolic rate compared to those at the Maastrichtian sites. In addition, the spatial variation of δ¹³C_DIC over ~200 m (the difference in habitat depth between the nautiloids and ammonoids) may further mask the actual difference in metabolic rates between the two groups.

The metabolic rate of modern Nautilus is significantly lower than that of all the coleoids studied, based on the stable carbon isotope proxy. This result agrees well with the estimates of metabolic rates based on oxygen consumption experiments, summarized by O’Dor and Webber (1991). There is no difference in the fraction of metabolic carbon among different species of Nautilus. The metabolic rate of the Maastrichtian nautiloid Eutrephoceras is also lower than that of all the co-occurring ammonoids examined. Intraspecific variation of the metabolic rate in ammonoids appears as high as (or higher than) that in modern coleoids. It is noteworthy that the metabolic rate is significantly different even among species within a single ammonoid genus such as Eubaculites (Fig. 3).

A higher metabolic rate entails a higher energy requirement per time unit to sustain minimal life activities. Therefore, a higher rate can be a disadvantage during a drastic food shortage. At the end of the Cretaceous, the Chicxulub asteroid impact as well as the Deccan Traps eruptions may have produced a transient episode of surface-water acidification (Henehan et al., 2019). This may have resulted in the decline of primary producers and planktic organisms, a likely food source for many ammonoids (see Kruta et al., 2011). Nautiloids may have been able to survive such a food-impoverished environment due to their lower metabolic rate and broader diet (for a discussion about the scavenging diet of modern nautilus, see Ward and Wicksten, 1980). The difference in egg size between the two groups is also linked to the availability of food resources. The larger embryonic size (~10 mm) of Maastrichtian nautiloids versus those of ammonoids (~1 mm) guaranteed an abundant supply of nourishment during the embryonic stage, preparing the postembryonic organism for the reality of a harsh, new world.

We thank M. Slovacek (formerly American Museum of Natural History) for preparing specimens, T. Linn (Glendive, Montana) and K. Ikuno (Museum of Nature and Human Activities, Hyogo, Japan) for help in fieldwork, and N. Izumoto (Atmosphere and Ocean Research Institute, University of Tokyo) for isotope analysis. This study was supported by KAKENHI (Japan Society for the Promotion of Science) grants (20J00376 and 21K14028 [both to Tajika], 17J11417 and 17K14413 [both to Nishida], 16H02944 [to Nishida and Ishimura], and 19KK0088 [to Ishimura, Nishida, and K. Sato), and a Tokyo Geographical Society Research Grant (to Tajika and Nishida). Landman was supported by National Science Foundation grant 1924807.