Late Palaeozoic mollusc reproduction: cephalopod egg-laying behavior and gastropod larval palaeobiology




Royal H. Mapes [], Department of Geological Sciences, Ohio University, Athens 45701, U.S.A; Alexander Nützel [], Bayerische Staatssammlung für Paläontologie und Geologie, Richard Wagner Str. 10, 80333 München, Germany;


Faunal analysis of an oxygen-depleted marine Lower Carboniferous succession (Late Mississippian Ruddle Shale) suggests how some cephalopod taxa laid their eggs during the Late Palaeozoic. At the Ruddle Shale collecting site in Arkansas, USA, the facies and overall fauna suggest severe oxygen depletion at the sediment/water interface. The Ammonoidea, with their small egg size, were probably laid in suspended gelatinous egg-filled masses in the water column above the bottom or by attachment of the egg masses to floating debris. The ammonitella embryos developed within the suspended or attached egg mass; hatched individuals became part of the free-swimming plankton biota. Based on shell morphology the Bactritoidea probably followed the same reproductive pattern. Coiled nautiloids (the Nautilida) and most orthoconic nautiloids (mostly the Pseudorthocerida) probably did not lay their eggs in the mid water column or as floatant attachments. This conclusion is based on the fact that, with one exception, all shells recovered of these two nautiloid orders are well past hatching. Gastropods in the Ruddle Shale are very small and cannot be visually detected in the field. However, microgastropods are abundant in washed residues. Most specimens are much smaller than 1 mm. The largest caenogastropod specimen is 1.3 mm high. These caenogastropods represent isolated larval shells and a successful metamorphosis was impossible because of oxygen depletion on the bottom. Allegations that a size of more than 1 mm is too large for pelagic larvae are refuted by examples of planktotrophic larval shells of modern gastropods (more than 1 mm high) and Triassic caenogastropods (up to 2 mm high) from the Cassian Formation (Northern Italy, South Alps). Repository information is given for the type-material of the gastropod species Nuetzelina striata Bandel, 2002 and Anozyga arkansasensis Bandel, 2002 which were both erected based on specimens from the Ruddle Shale that were illustrated by Nützel & Mapes in 2001.

A variety of egg-laying behaviors has been reported for modern cephalopods (e.g. Boyle 1983, 1987; Boletzky 1987a,b; Westermann 1996; Norman 2000). Most modern cephalopods lay their eggs on the sea bottom, either as single individual egg (Nautilus) or in clusters (e.g. squids, cuttlefish, and some Octopus). However, some octopuses brood their eggs in the mid- and upper water column, and some modern coleoids have relatively small eggs in floating egg masses. These different traits as well as location of egg laying have wide implications for dispersal and selection mechanisms (Hewitt 1988).

Despite intense research about cephalopod ontogeny, there is no direct fossil evidence of the reproductive behavior associated with egg laying by fossil cephalopods. Numerous studies about early ontogenetic shells of cephalopods, especially about those of Ammonoidea (ammonitellas) are concerned with size, shell ultrastructure, morphology, and phylogenetic implications (e.g. Kulicki 1996; Landman et al. 1996). However, little has been written on the reproductive behavior and place of egg laying or embryo development (Faulkner et al. 1992; Faulkner 1993; Tanabe et al. 1993; Westermann 1996). We are aware of only one attempt to resolve details of the egg laying of fossil nautiloids (Chirat & Rioult 1998). The early ontogenetic development, taxonomy, shell ultrastructure, and morphology of the Bactritoidea were discussed in several studies (Erben 1964a, b; Mapes 1979; Doguzhaeva 1996, 1999, 2002; Hecht et al. 1996). Initially, bactritoids were thought to have a post-hatching veliger larval stage of development (Erben 1964a); however, this idea has been completely abandoned based mostly on the ultrastructural details at the time of hatching of the bactritella (Doguzhaeva 1996, 1999, 2002).

In fossil cephalopods, the study of well-preserved early ontogenetic stages as well as lithofacies is needed to infer egg-laying behaviour and life histories. The Lower Carboniferous Ruddle Shale Member (Arkansas, USA) has yielded such very well-preserved early ontogenetic mollusc shells. The well-preserved pelagic mollusc assemblage from this shale comprises mainly cephalopods and planktonic gastropod larval shells, which allow the reconstruction of life histories and reproductive behavior of Orthocerida nautiloids, ammonoids, and especially bactritoids as well as of gastropods.

Locality, biostratigraphy, and methods

The fossil-bearing site is a shale exposure belonging to the Ruddle Shale Member of the Moorefield Formation. Informally, the site is known as the Buffalo Wallow locality. The site is a set of open glades and ravines (Fig. 1A) located about 2 km north and 7.2 km east southeast of Batesville, Independence County, Arkansas (Mapes 1979, locality M-11; Nützel & Mapes 2001). The shale, including a particularly fossiliferous horizon that yielded the specimens described herein, belongs to the Lower Carboniferous (Mississippian). The horizon falls within the Goniatites granosusDombarites choctawensis Zone of late Visean (early Chesterian) age (Malinky & Mapes 1982). Conodonts also indicate an early Chesterian age for the fossil horizon (T.L. Thompson, personal communication, 2000; see Nützel & Mapes 2001).

Figure 1.

inline imageA–C. Outcrop and lithofacies of the Ruddle Shale Member at the Buffalo Wallow location (study area). inline imageA. Outcrop overview. inline imageB. Flaky weathered shale with faint lamination and concretion horizon; rucksack as scale. inline imageC. Flaky weathered shale, detail; scale: pocket knife in upper left corner. inline imageD–F. Micrographs of thin sections of concretions; black matter is pyrite; slight lamination and peloidal texture (probably faecal pellets) is visible. inline imageF. Concretion with bivalve shell accumulation at surface; shells are extremely thin and covered with thick cements. inline imageG. Monaxon spicule cluster that is probably a hold fast of an unidentified sponge (OUZC 5165).

The shale exposed at the Buffalo Wallow locality is about 50 m thick (Drahovzal 1966). In general the Ruddle Shale member is as thick as 75 m. Both the base and the upper contact with the overlying Batesville Sandstone are covered at this locality. Macrofossils have only been recovered from the lower 10 m exposed at the Buffalo Wallow locality (the rest of the exposed section at that site is unfossiliferous). The macrofauna of this fossil-bearing unit has been collected for more than 40 years by students and faculty of the Universities of Iowa and Arkansas and especially by R. H. M. and students of Ohio University; extensive collections of the fauna are available for study at the Ohio University palaeontology collections repository. Within this fossil-bearing unit, a horizon of about 1 m thickness was repeatedly sampled for microfossils. It yielded the minute cephalopods and gastropods that form the subject of this report. This horizon is exposed about 4 m above the base of the exposed section and about 1 m below a thin, discontinuous limestone unit on the north end of the exposure. It also has produced numerous larger well-preserved pyritized and limonitized cephalopods by surface collecting and was thus chosen for bulk sampling. The shale above and below this horizon was not sampled for microfauna.

Over a span of approximately 20 years, bulk shale samples, totalling more than 400 kg from the fossil-bearing interval, have been disaggregated by the methods described by Mapes & Mapes (1982). After disaggregation of the shale, usually a 60-mesh sieve was used to concentrate the fossil-bearing residue; however, for conodonts, a 230-mesh sieve was usually used on some samples. The fraction 0.2–1 mm of the dried residue from the 400 kg of shale (about 1 kg of residue) was completely picked for all fossils.

Geological setting and lithology

Geological setting and lithology of the Ruddle Shale Member were described and discussed by Nützel & Mapes (2001). The Ruddle Shale is a fine-grained, dark gray clay stone that weathers flaky so that primary sediment structures are obscured in the field (Fig. 1A–C). However, lamination can be recognized in the shale as well in weathered carbonate concretions. Horizons of flat calcareous and iron-rich concretions are common in the lower portion of the beds exposed at the Buffalo Wallow locality. The concretions are reddish or dark with yellow weathering colour. Thin sections of the concretions show that they consist of laminated mudstones with peloidal texture. The peloids probably represent faecal pellets. Disseminated pyrite is moderately abundant. Some concretions are fossiliferous with the dominant fossils being cephalopods, arthropods, and pteriomorph bivalves. The bivalves in the concretions are very thin-shelled and can be enriched at the surface of the concretion (Fig. 1F). Some of the ammonoid cephalopods retain mandibles within the body chambers indicating that concretion formation was very early in the depositional history of the unit. Weathering commonly oxidizes the pyritized fossils and concretions within the shale. The cephalopods, gastropods and other Mollusca are preserved as shell-covered, solid iron-oxide/pyrite casts that appear light to dark brown. Sometimes even the shell has been replaced by iron oxide but is preserved in great detail. For no reason we can determine, the cephalopod shell debris from the same samples are mostly darker than the gastropod shells; however, they are preserved in a similar fashion. Overall, the Ruddle Shale Member at this site is not very fossiliferous.

Deposition of the Ruddle Shale took place in an offshore marine environment, below the storm wave base. A relatively deep-water, off-shore environment (outer shelf to basinal edge) is suggested by the ammonoid and bactritoid cephalopod biofacies (Boardman et al. 1984; Hecht et al. 1996). This interpretation is supported by the low energy soft mud bottom without signs of storm deposition or winnowing, and the presence of parallel laminations in early diagenetic concretions. The presence of pyrite and the lack of bioturbation support the conclusion that the sediment at the water–sediment interface on the bottom was a reducing environment. The lack of a benthic fauna suggests that this reducing environment extended into the water column for an unknown distance above the water–sediment interface (see below for additional discussion of these interpretations).


The illustrated material, except for the gastropods, is housed in the Ohio University Zoological Repository (OUZC) at Ohio University, Athens, Ohio, USA; accession numbers of reposited specimens are OUZC 5163–5178. Moreover, repository numbers OUZC 5179–5182 are assigned to the non-gastropod material illustrated by Nützel & Mapes (2001, fig. 8J–Q) including two bivalves, one ostracode and one foraminiferid specimen.

Figure 8.

 Embryonic and post-embryonic shells of three different and unidentified bactritoid species. These specimens are considered different species because they have different ornament, septal spacing (position of septa marked with arrows), protoconch shapes and shell dimensions; all are about ×50. Embryonic shells of bactritoids will have up to four septa (= three chambers and the rounded protoconch; see Fig. 7) and a long body chamber (indicated with ‘b–c’) that gradually narrows to the hatching constriction; post-embryonic shells will have more than four septa and a body chamber that begins expanding past the hatching constriction. Many of the bactritoids from the Ruddle Shale are crushed and somewhat distorted, making it difficult to see the hatching constriction. inline imageA–C. Bactrites sp. A is characterized by the size and shape of the protoconch, smooth initial shell, rate of expansion, and similarities in septal spacing: the conch is an embryonic shell with three well-developed septa (at arrows) and has a slightly flattened, relatively long body chamber (OUZC 5171). B is a post-hatchling individual with five septa (at arrows) (OUZC 5172). C is a four-septa (at arrows) bactritella that appears to have died almost immediately after hatching (OUZC 5173). inline imageD–E. Bactrites sp. B is identified by the conch diameter, expansion rate, and the box work-like ornament on the initial chambers. D is an embryonic bactritella with three septa and a short body chamber. The position of the most oral septum is indicated by a shallow constriction; this septum was probably not secreted at the time of the animal's death, and this would make the body chamber longer than indicated (OUZC 5174). E is a post-hatchling bactritella without a protoconch with five septa and a long body chamber, which is not shown to its full extent (OUZC 5175). inline imageF–H. Bactrites sp. C is identified on the basis of the relatively large round protoconch, smooth shell on the initial chambers, and septal spacing. F is a post-hatchling bactritella with five septa; the complete and relatively long body chamber is not completely shown (OUZC 5176). G is an embryonic bactritella with two septa (OUZC 5177), and H is a somewhat crushed embryonic bactritella with only one septum (OUZC 5178).

All illustrated and non-illustrated gastropod material including that illustrated by Nützel & Mapes (2001, figs 5–8) is housed in the Bayerische Staatssammlung für Paläontologie in Munich, Germany, under the numbers BSPG 2008 XIV. This includes the type material of the gastropod taxa Nuetzelina striata and Anozyga arkansasensis which were both erected by Bandel (2002) based on specimens illustrated by Nützel & Mapes (2001).

Figure 5.

 Early whorls of Ammonoidea. The ammonitella terminates at the primary constriction (arrows). inline imageA. Specimen from the Lower Carboniferous Ruddle Shale (OUZC 5168) with smooth ammonitella. inline imageB. Schematic drawing based on an illustration of a Cretaceous specimen (Landman et al. 1996, fig. 1A); the ammonitella of this Late Cretaceous specimen (originally with pustules, omitted here) consists of the shell up to the primary constriction and is about 0.7 mm in diameter; the morphology of this lytoceratid ammonitella is not significantly different from the Ruddle Shale goniatitid specimens.

General characteristics and palaeoecology of the Ruddle Shale fauna

We hypothesize that mollusc egg laying and hatching on the sea bottom was impossible during deposition of the Ruddle Shale due to oxygen depletion. Support for this hypothesis is dependant on known modern cephalopod reproductive behavior and our interpretation of the palaeoenvironmental conditions in the water column and at the water–sediment interface in the Ruddle Sea. The following facies and faunistic characteristics of the Ruddle Shale suggest a major oxygen deficiency in the base of the water column at the water–sediment interface (see also Nützel & Mapes 2001).

  • 1 Presence of pyrite and common pyritization of fossils indicating a reducing environment (pyrite now present as iron oxides due to weathering; pyrite is still present in some concretions).
  • 2 Presence of phosphate concretions.
  • 3 Fine parallel lamination in early diagenetic concretions indicating the lack of bioturbation of the sediment.
  • 4 Lack of a typical Late Palaeozoic benthic community, e.g. of brachiopods, echinoderms (crinoids), trilobites, corals, etc.
  • 5 Lack of normal sized adult gastropods contrasted by an abundance of minute (less than 1–2 mm) gastropods (mostly planktonic larval shells).
  • 6 Abundance of a relatively rich nektonic assemblage of cephalopods (contrasted by poor to almost absent benthic assemblage).
  • 7 The lack of scavenging that allowed soft parts (mandibles) to be preserved within some ammonoid shells.

About 40 identified invertebrate taxa have been recovered from the Ruddle Shale: 11 gastropod taxa (all very small – most are less than 1.0 mm, seven of them as isolated larval shells) (Figs 2A–C, 3A–D); 26 cephalopod species (10 orthoconic nautiloids (Fig. 4), four coiled nautiloids, eight ammonoids (Figs 5–6), and three bactritoids (Fig. 8)), two bivalve species, rare remains of an unidentified sponge (Fig. 1G) (three sponge spicule holdfast tufts from the 60 kg shale sample), a single unidentified species agglutinated tubular foraminifer, a single unidentified ostracode species, and a spathocarid arthropod. In addition, three limonite after pyrite tubes that filled burrows in the bottom sediment were recovered from a 60-kg shale sample taken from the main fossil-bearing interval. These burrows imply there was rare infaunal activity.

Figure 2.

 Relatively large multi-whorled caenogastropod protoconchs from the Mississippian Ruddle Shale (A–C), Recent (D), and the Late Triassic (E, F); A–F at same magnification (see scale bar). The larval shells from the Ruddle Shale Member (A–C) are isolated with no teleoconch attached; the dimensions of planktotrophic larval shells are in a comparable size range. inline imageA–C. Isolated larval shells from the Ruddle Shale Member. inline imageA. Caenogastropod larval shell ‘Zygopleura’ from Nützel & Mapes (2001, fig. 6H) (BSPG 2008 XIV 1). inline imageB. ‘Ianthimorpha’ from Nützel & Mapes (2001, fig. 7C) (BSPG 2008 XIV 2), with a height of about 1.3 mm, this is the largest known gastropod (larval) shell from the Ruddle Shale Member. inline imageC. Imoglobid species 2 from Nützel & Mapes (2001, fig. 8G) (BSPG 2008 XIV 3); this shell was interpreted as isolated larva of an unknown taxon; based on this specimen, Bandel (2002) erected the new genus Nuetzelina and interpreted the ornamented part of the shell as teleoconch. inline imageD. Recent juvenile shell of the caenogastropod Cypraea with a rather large planktotrophic larval shell with reticulate ornament. inline imageE–F. Relatively large protoconchs of Late Triassic caenogastropods from the Cassian Formation. In both specimens, the larval shell is attached to the teleoconch. inline imageE. Zygopleura hybridissima (from Nützel 1998 pl. 20, fig. P). inline imageF. Protorcula marshalli (from Nützel 1998 pl. 27, fig. I); with nearly 2 mm height this is the largest larval shell known for all Mesozoic gastropods. inline imageG. Several hundred randomly chosen microgastropods from the Ruddle Shale Member with match as a scale to show minute size of the gastropod fauna.

Figure 3.

inline imageA–C. Cluster of minute gastropod shells, mostly protoconchs of a single naticopsid species (BSPG 2008 XIV 4); it contains also a specimen representing a caenogastropod larval shell ‘Zygopleura sp.’, which shows that this cluster does not represent fossil spawn. inline imageD–F. Two Carboniferous representatives of the caenogastropod family Pseudozygopleuridae; this family has an extremely characteristic larval shell with strongly curving axial ribs. In the Ruddle Shale, these larval shells are always isolated (D, from Nützel & Mapes 2001 fig. 6D; BSPG 2008 XIV 5) i.e. they are not attached to a teleoconch. Pseudozygopleurids with preserved protoconchs attached to an adult shell are known from various Late Palaeozoic sites; inline imageE–F shows an example from the Upper Carboniferous (Pennsylvanian) Ames Shale, West Virginia (Nützel 1998, pl. 29, Q–R).

Figure 4.

 Early post-hatched orthoconic nautiloid probably belonging to either Reticycloceras sp. or Mitorthoceras sp. The initial portion of the shell is cicatrix bearing and smooth with later embryonic growth having coarse growth lines. The nepionic constriction (= time of hatching) is indicated by the arrow, note the sublethal repaired damage sustained after hatching. Only a small amount of growth occurred before the demise of the animal; this is indicated by the crushing of the shell, which indicates there is a lack of massive cameral and/or siphuncular deposits that are typically found in the phragmocones of more mature specimens of these genera (OUZC 5167).

Figure 6.

 Embryonic and post-embryonic shells of ammonoid cephalopods from the Ruddle Shale (see also Fig. 5); arrows mark primary constrictions that mark the time of hatching. inline imageA–B. Early juvenile specimen, not determinable to genus with confidence. The post-hatchling specimen shows the non-pustulate ammonitella; the hatching constriction and a sublethal repair at the constriction are indicated at the arrow. Sublethal repairs of the embryonic shell of ammonoids (and most all other embryonic shells of other cephalopod classes) are very rare. Thus, the presence of repaired shell damage is considered to be an indication of a post-hatched shell. Note the initiation of ribbing after the primary constriction; such ornament changes after hatching are common in Carboniferous ammonoids. inline imageA. Lateral view. inline imageB. Ventral views (OUZC 5169). inline imageC–E. Juvenile specimen (OUZC 5170) of Girtyoceras with slightly more than 2.5 whorls past the primary or hatching constriction (arrows). The ribbing is lost in more mature specimens (pieces of body chambers of this genus have been surface collected from the locality that indicate mature diameters of over 200 mm were attained). inline imageC. Early whorls in oblique lateral view, including hatchling ammonitella with virtually no growth past the hatching constriction (at arrow). inline imageD. Lateral view. E. oblique ventral view; hatching constriction marked with arrow.

The coarser sieve fractions of the bulk samples yielded cephalopods at different stages of growth and other faunal elements such as fragments of an arthropod probably belonging to Spathocarus and planktonic pelecypods. However, and of great importance, no gastropods were observed in these larger size fractions and no large gastropods have been recovered in extensive surface collecting of the exposure. To date more than 1000 gastropod specimens have been recovered only from sieve sizes of less than 1 mm. These gastropods occur with the embryonic cephalopods, conodonts, sponge spicule clusters, and other faunal elements indicated above.

Nektonic and planktonic molluscs, especially the Cephalopoda, dominate the fauna of the Ruddle Shale. Cephalopods are by far the most abundant fossils that can be found by surface collecting in the field. Even cephalopod mandibles are present and some are in situ in the body chambers of shells encased by early diagenetic carbonate concretions (Doguzhaeva et al. 1998).

At least two pteriomorph bivalve species are present at various stages of growth including isolated planktonic larval shells (prodissoconchs) (see Nützel & Mapes 2001). They were identified by Gordon (1964) as Posidonia nasuta (Girty) and Caneyella richardsoni Girty, which either lived as part of the plankton or represent an opportunistic benthos taxon adapted to minor fluctuations of the available oxygen. Nützel & Mapes (2001) reported an abundant and fairly diverse microgastropod assemblage, which commonly represents larval stages (Figs 2A–C, 3A–D) (see below). Other pelagic biota present in the Ruddle Shale include spathocarid arthropods that can attain carapace lengths of over 100 mm, shark teeth and other fish debris, and conodonts. These faunal elements document a rich pelagic assemblage that lived in the oxygenated water column above the bottom.

Faunal elements in the Ruddle Shale Member that are usually considered to be benthic are limited to rare sponge remains, ostracodes, foraminifera, and small burrow infillings. The sponge remains occur as small (2.0 mm in length) isolated clusters of monaxon sponge spicules that probably served as holdfasts in the soft mud or to a floating substrate such as terrestrial plant debris or algae (Fig. 1G). If this organism was not attached to a floatant at the surface, but was instead a sessile animal at the sediment/water interface, then its presence suggests that a few organisms could withstand the extremely low-oxygen bottom conditions for a limited period of time or that there were fluctuations of the oxygen concentration that allowed temporary colonization of the bottom. No articulated sponges have been recovered. The recovered benthic microfauna consists of a single species of agglutinated foraminifera as well as an ostracode species (Nützel & Mapes 2001). Intense and repeated collecting failed to produce fragments of any other sessile or mobile benthic fauna such as brachiopods, corals, bryozoans, crinoids, or trilobites.

Thus, the facies and fauna support the conclusion that the Ruddle Shale ocean bottom waters were oxygen depleted (Nützel & Mapes 2001). However, in the fossil-bearing interval, a low diversity assemblage of foraminifera and ostracodes as well as some of the post-larval diminutive gastropods (bellerophontoids and vetigastropods) and sponges suggest episodic, short-term presence of low oxygen concentrations that sometimes reached the sediment/water interface. The pioneer juvenile gastropod fauna (bellerophontoids and vetigastropods) were killed relatively quickly, well before they could reach sexual maturity. This suggests episodic changes from exaerobic to quasi-anaerobic conditions with a rapid return back to exaerobic conditions in a generally anaerobic regime throughout the region (see Nützel & Mapes 2001).

Palaeoecological interpretation of the Ruddle Shale microgastropods

The washed residues from the studied fossil bearing horizon of the Ruddle Shale yielded more than 1000 microgastropod specimens representing at least 11 species-level taxa (Figs 2A–C, G, 3A–D) (see also Nützel & Mapes 2001). The gastropods are generally very well-preserved. Most (more than 90%) of the studied gastropod specimens are smaller than 1 mm with only a few attaining a larger size. The largest specimen is 1.3 mm high (Fig. 2B) and gastropods are absent at a mesh size of > 1 mm (Nützel & Mapes 2001). Not a single gastropod was found during repeated field collecting expeditions for more than 20 years because the gastropods are so small that they are overlooked.

Virtually all gastropods from the Ruddle Shale have an intact initial whorl. Most have a few additional whorls and the largest have a total of up to five to six whorls. This and the small size qualify the gastropods from the Ruddle Shale as protoconchs (larval shells and early juveniles). Gastropod protoconchs reflect ontogeny and life histories to a high degree (e.g. Shuto 1974; Jablonski & Lutz 1980, 1983). About the first whorl of gastropods represents the embryonic shell (protoconch I). Basal gastropod clades, such as the Vetigastropoda, hatch with this protoconch I and have a non-feeding planktonic veliger stage, but they do not add shell during this larval stage (i.e. they do not produce a larval shell). Thus, after metamorphosis, shells of basal gastropod clades (see Frýda et al. 2008) consist of a protoconch I (embryonic shell) and an adult shell (teleoconch) only. The more derived gastropod clades (Caenogastropoda, Neritimorpha, Heterobranchia) hatch with a protoconch I and are able to form plankton feeding veliger larvae.

There are several caenogastropods and one neritimorph larval shell present in this fauna (Nützel & Mapes 2001). Within each of these more derived clades, different life mode strategies can be present including: planktotrophic larval development, lecitotrophic development (non-feeding larvae), and direct development (hatching of a crawling juvenile). Basically, these strategies can be inferred from the protoconch morphology, i.e. the size of the initial whorl (which serves as a proxy for the amount of yolk), the number of larval whorls (high number in planktotrophic species) and other features such as changes in shape and ornamentation.

Dimensions and morphology of the minute gastropods from the Ruddle Shale indicate that this assemblage represents early juveniles and isolated planktonic larval shells (see also Nützel & Mapes 2001) and this is also important for the reconstruction of the cephalopod life histories. The neritimorph (naticopsid) and caenogastropod larval shells are always isolated (Figs 2A–C, 3A–D), i.e. these gastropods did not complete their life cycle (no metamorphosis), and they died when they attempted to assume a benthic life mode. The larvae descended to the bottom from the upper part of the water column and died because conditions on the bottom were unfavourable for metamorphosis. The minute but seemingly post-larval vetigastropods and bellerophontoids represent juveniles. They were either floating at the air/water interface or attached to sea weed or came in as non-feeding larvae and were able to live for a short while but failed to grow to a mature size. It is very likely that the gastropod assemblage as a whole was unable to attain maturity and reproduce on the sea bottom at or near the studied site probably due to oxygen deficiency. Oxygen deficiency is also suggested by the absence of benthic organisms and facies characteristics as previously indicated.

A cluster containing about 50 microgastropods represents an interesting new discovery from the Ruddle Shale (Fig. 3A–C). It is almost monospecific, containing almost exclusively Naticopsis sp. protoconchs. However, it also contains at least one caenogastropod larval shell of the ‘Zygopleura’-type sensu Nützel & Mapes (2001) (Fig. 3B). Therefore, this cluster cannot represent spawn. This is important because if it was spawn, this cluster would represent an argument against our point that reproduction of gastropods (and other molluscs) was impossible on the Ruddle Shale sea bottom. The dimensions of the clustered gastropod shells suggest that they represent planktonic larval shells. It is unclear which process concentrated these shells. This cluster probably represents a coprolite from a pelagic organism, which fed on planktonic gastropod larvae. A less likely scenario is that it is an agglomeration produced by an unknown benthic concentration process.

Ruddle Shale microgastropods as planktonic larvae and maximum size of larval shells

Bandel (2002) doubted that some of the small gastropod shells from the Ruddle Shale represented only the larval stage of development because some specimens are as large as 1.0–1.3 mm. Bandel argued that this size is too large for planktonic gastropod larvae. Specifically, Bandel did not accept that some Ruddle Shale gastropod shells, which were left in open nomenclature by Nützel & Mapes (2001) as ‘Ianthimorpha’, ‘Zygopleura’, and Imogloba, represent larval shells of caenogastropod taxa (Bandel 2002, pp. 160, 173, 174). Instead, Bandel interpreted these shells as adults or post-metamorphic juveniles and even named some of the open nomenclature specimens illustrated by Nützel & Mapes (2001) from the Ruddle Shale (specifically Nuetzelina striata Bandel, 2002, and Anozyga arkansasensis Bandel, 2002) (see also Nützel & Pan 2005). The shell illustrated in Fig. 2C was interpreted as an isolated larval shell of an unknown taxon of the caenogastropod family Imoglobidae by Nützel et al. (2000) and Nützel & Mapes (2001). However, based on our published images, Bandel (2002) erected the new genus Nuetzelina and interpreted the ornamented part of the shell as teleoconch.

Bandel's claim that these particular gastropods grew to adulthood and were benthic is in conflict with our interpretation of the Ruddle Shale as being deposited under strongly oxygen-depleted conditions. It also conflicts with our interpretation of cephalopod reproductive behaviour provided herein. Thus, the question arises how large a planktonic gastropod larval shell can be. Below, we now argue fourfold: (1) that all gastropods from the Ruddle Shale are small and that larger (normal sized) gastropods are lacking, (2) that there are modern as well as Early Mesozoic planktonic gastropod larvae that are as large or even larger than even the largest Ruddle Shale microgastropods, (3) that some of the more characteristic shells from the Ruddle Shale are undoubtedly isolated pseudozygopleurid larval shells that are well-known from other Palaeozoic sites where they are attached to adult shells, and (4) that there is strong evidence for oxygen deficiency on the bottom.

1. Very small size of the Ruddle Shale gastropod fauna (absence of large gastropods)

By far most of the gastropods of the Ruddle Shale are smaller than 1 mm (Fig. 2G), which would be normal for gastropod larvae. Only a few specimens are as large as 1.3 mm, but even this relatively large size lies within the range of shells built by planktonic veliger larvae (Figs 2, 3). Larger gastropods are completely absent. The small bellerophontoids and vetigastropods from the Ruddle Shale fauna were probably not planktonic veliger larvae but early juveniles (Nützel & Mapes 2001). Their presence can be explained by either being attached to floating algae or other debris or adhering to the water/air interface. However, many of the cephalopods from the same samples, which yielded the gastropods, are much larger (up to 30 mm in diameter) and essentially mature ammonoid specimens that are more than 140 mm in diameter were recovered during surface collecting. This shows that the small size of the gastropods is not a preservational artefact due to selective pyritization of small shells. None of the Ruddle Shale gastropods were able to grow to a normal adult size; they did not reach maturity and could not reproduce at that place.

2. Size comparison of the Ruddle Shale microgastropods with modern and Mesozoic larval shells

Bandel (2002) claimed that the maximum shell size for a planktonic life mode was exceeded for some selected Ruddle Shale gastropods (e.g. Anozyga arkansasensis at 1.2 mm maximum size). However, modern higher caenogastropods commonly have larval shells that exceed larval shell sizes of 2.0 mm (e.g. Bandel et al. 1997; Riedel 2000). Here, we illustrate the larval shell of a modern caenogastropod (Cypraea sp.) (Fig. 2D) in comparison with the largest Ruddle Shale gastropods ever found (Fig. 2A–C) and some Triassic protoconchs at the same scale. It is obvious that that they all fall into the same size range. Thus, veliger larvae can remain planktonic up to 2.0 mm shell height not only in modern higher gastropods but also in Palaeozoic and Early Mesozoic gastropods. Bandel himself (1991) reported larval shells of Early Mesozoic caenogastropods from the Late Triassic Protorcula subpunctata which are up to 1.3 mm. Nützel (1998) reported larval shells of the same species and of Protorcula marshalli that are up to 1.9 mm (Fig. 2F). Each of the Recent (Fig. 2D) and Triassic (Carnian, Cassian Formation) (Fig. 2E–F) gastropod specimens illustrated here comprises both the protoconch and the early teleoconch. The protoconchs terminate in a sinusigera with an abrupt ontogenetic change of the shell ornament. This suggests metamorphosis and shows that planktonic veliger larvae formed these multi-whorled protoconchs. These examples show that Bandel's contention that some of the Ruddle Shale gastropods are too large to represent planktonic larvae must be rejected.

3. Presence of clearly isolated pseudozygopleurid larval shells

Additional evidence that the caenogastropod larval shells are in the planktonic larval development stage is seen in the pseudozygopleurid larval shells (Fig. 3D–F). Pseudozygopleuridae form a highly diverse Late Palaeozoic caenogastropod family, which is characterized by extremely diagnostic larval shells (e.g. Knight 1930; Hoare & Sturgeon 1978; Nützel 1998). Nützel & Mapes (2001) reported about 30 pseudozygopleurid larval shells in the initial collection, and since that time additional shells have been recovered (see Fig. 3D). These larval shells are highly characteristic and can be assigned to the caenogastropod family Pseudozygopleuridae with certainty (compare Fig. 3D with Fig. 3E–F). Not a single one of the protoconchs of these shells is attached to a juvenile teleoconch. Hence, the conclusion that pseudozygopleurids from the Ruddle Shale were unable to perform metamorphosis is inescapable, and it is a reasonable assumption that all the other caenogastropods present in the Ruddle Shale are also in a premetamorphic stage of development.

4. Oxygen-deficient facies

The fourth point in dealing with our and Bandel's (2002) size/planktonic veliger larval interpretation is that Bandel did not consider the palaeoenvironment that was present on the bottom of the Ruddle sea at that time. Facies and fauna other than gastropods suggest oxygen depletion due to pyrite, phosphate, parallel bedding laminations, lack of typical Late Palaeozoic benthos, dominance of nekton, and lack of bottom scavenging as discussed above (see also Nützel & Mapes 2001). This contradicts Bandel's interpretation that some of the caenogastropods were adult bottom dwellers.


The cephalopod fauna of the Ruddle Shale is relatively diverse. Goniatite ammonoids (Figs 5–6) in the Ohio University collections include Dombarites choctawensis, ‘Lusitanocerasgranosum, Girtyoceras limatum, Lusitanites subcirculare, Metadimorphoceras varians, and rare specimens of Ruddellites drahovzali (identifications based on Gordon 1964; Drahovzal 1966; Malinky & Mapes 1982; Manger 1988; D. Korn, personal communication, 2008; and others). With the exception of the latter genus, which is only known from sub-adults and specimens that are probably mature, these genera have been recovered through all stages of growth from the ammonitella to what are probably mature individuals. The report by Drahovzal (1966) of Cravenoceras from this exposure is now considered to be a case of contamination as thirty years of macro collecting and observation of thousands of micro and macro goniatites have not yielded additional Cravenoceras specimens to support the claim by Drahovzal that Cravenoceras extends this low in the world wide biostratigraphic zonation. Reports of Sudeticeras sp. and Neoglyphioceras newsomi from this collecting site as reported by Gordon (1964, table 2) have not been substantiated.

Three different bactritoid taxa have been recovered by surface collecting: Bactrites carbonarius, Bactrites cf. B. alhosoensis, and Bactrites Morphotype 13 (Mapes 1979). In addition, microsamples have yielded three different kinds of bactritoid protoconchs called bactritellas (Fig. 8); these bactritellas are not systematically treated here, as that is beyond the scope of this investigation. These microbactritoids include embryological specimens, hatchlings, and very early juveniles. Larger bactritoid body chamber fragments and more rarely segments of the phragmocone are recovered as part of the macrofauna collections. Gordon (1964, table 2) reported the presence of Bactrites smithianus from this stratigraphic unit, but this taxon was not observed by Mapes (1979) and in subsequent collections.

Coiled nautiloids in the Ruddle Shale are very rare. To date no hatchlings of this group of cephalopods have been recovered and there are only a few relatively mature specimens in the collection. At least three genera are present in the Ohio University collections including Solenochilus sp., Epistroboceras sp., and ?Peripetoceras sp. Additionally, Gordon (1964, table 2) reported the presence of Endolobus sp. from this stratigraphic unit.

Segments of several different taxa of orthoconic nautiloid (Pseudorthocerida) are common in the macrofaunal collections; however, these segments never retain the initial growth stages. These more mature orthoconic nautiloid phragmocone fragments recovered by surface collecting in the Ohio University collections are identified as Mitorthoceras perfilosum, M. crebriliratum, ?Reticycloceras sp., Rayonnoceras sp. and Brachycycloceras sp. Additionally, Gordon (1964, table 2) reported the occurrence of Euloxoceras greenei praecursor, Adnatecoceras alaskense, Dinocycloceras cf. ballianum, Mitorthoceras sp., an unidentified annulate orthocone, and a number of indeterminate orthocones from this stratigraphic unit.

Only three protoconch-bearing segments of an orthoconic nautiloid (Pseudorthocerida) have been recovered by processing the shale for the microfauna. Two of the specimens are missing the oral portions of the shell. The third specimen (Fig. 4) has changes in the ornamentation and repaired sublethal damage that suggests this individual is a very early juvenile. Kröger (personal communication, 2005) suggested that these protoconchs were similar to those observed in Reticycloceras; however, it is possible that they belong to Mitorthoceras, which Gordon (1964) reported from this stratigraphic interval. The protoconch of Mitorthoceras is as yet unknown.

Ontogeny and life histories of the cephalopods

Any analysis of reproduction of fossil cephalopods must be primarily based on studies of modern cephalopod reproduction because the normal fossilization processes do not typically preserve the eggs. However, it is possible that extinct cephalopod groups followed non-actualistic ontogenetic pathways. Moreover, even the egg-laying behavior of many modern cephalopods is unknown. Of those that are known, a variety of egg-laying behaviors has been reported (see the compilations edited by Boyle in 1983, 1987). Most modern cephalopods lay their eggs on the sea bottom, either as single individual egg (e.g. Nautilus) or in clusters (e.g. squids, cuttlefish, and some Octopus). However, some octopuses are known to brood their eggs in the mid- and upper water column (for more extended discussions and examples see Boyle 1983, 1987; Boletzky 1987a, b; Westermann 1996; Norman 2000).

The reproductive strategy of relatively small eggs in floating egg masses is found in some modern coleoids (see Westermann 1996 for an extended discussion). Some ommastrephid squids extrude their eggs in a gelatinous mass in the water column (Seibel et al. 2000). Free-floating gelatinous egg masses of the modern ommastrephid squid Nototodarus gouldi were reported from New Zealand waters at a water depth of 10 to 30 m (O'Shea et al. 2004). They have a diameter of 1–2 m and contain an estimated several thousand randomly distributed eggs. The young squid hatch directly from this gelatinous mass into the water column. Other pelagic species, such as those that belong to the Enoploteuthidae, release eggs singly into the water column (Young 1985); the hatchling squids become part of the pelagic fauna. Some cephalopods nurture and protect the eggs, whereas, others follow a ‘lay it and forget it’ strategy. These different traits as well as the preferred locations where egg laying takes place have wide implications for dispersal and selection mechanism (r- vs. K-type selection) (Hewitt 1988).

The ontogeny of the Ammonoidea has received much attention for many years (for example see the 1957 and the 1964 American and Russian treatises on cephalopods, respectively), and many details of the ontogeny of the shells, including the presumed embryological growth of the shell have been well documented (e.g. Kulicki 1996; Landman et al. 1996). The ontogeny of coiled nautiloids is also moderately well known because of their close relationship to modern Nautilus and the laboratory experiments that have been successfully done to determine details of the reproduction and early ontogenetic growth of this modern model (Carlson 1991; Uchiyama & Tanabe 1998). Although a modern model for coiled nautiloids is available, many palaeobiological details of fossil nautiloids remain unknown, and some modern information derived from Nautilus cannot be directly applied to the fossil forms. The details of the growth that took place in the orthoconic nautiloids are moderately well documented because of the taxonomic importance of the deposits that most of these animals developed inside their phragmocones. Details regarding their reproduction are unknown, and only inferences based on what is known about modern Nautilus have been suggested (Chirat & Rioult 1998). The Bactritoidea have received some attention (Erben 1964a,b; Mapes 1979; Doguzhaeva 1996, 1999, 2002; and others). These animals are thought to be evolutionarily and phylogenetically intermediate between the Ammonoidea and the Orthocerida (Palcephalopoda) (e.g. Engeser 1996; Kröger et al. 2005; Kröger & Mapes 2007), and for this reason, a mixture of early ontogenetic growth and reproductive strategies used by either subclass is possible. Additionally, the Bactritoidea are believed to have given rise to all the modern Coleoidea (octopus, squids, etc.) (e.g. Jeletzky 1966; Engeser 1996). By examining the reproductive strategies of these modern cephalopods, some insight can be gained about the reproductive strategies of the fossil forms, particularly the cephalopod fauna from the Ruddle Shale Member at the ‘Buffalo Wallow’ locality. Thus, a brief overview of the embryologic, hatchling and early juvenile characteristics of the different cephalopod groups that have been recovered at the Ruddle Shale site is warranted.

Ammonoids are now believed to develop in small (about 1–2 mm) eggs with a small yolk. The shell essentially forms all at one time in the egg and a constriction is present at the aperture that marks the time of hatching (Figs 5, 6). After hatching, the shell ornament changes dramatically on some taxa, and many times there is a dramatic increase in repaired shell damage (for more details see Tanabe et al. 1993; Landman et al. 1996 and their citations). Eggs are thought to have been laid in large numbers, and the site of egg laying is thought to be in the water column in a gelatinous mass or attached in groups to floating material (Westermann 1996). Another possibility is that some ammonoids laid their egg masses on the bottom (Faulkner et al. 1992; Faulkner 1993). Hatching time is thought to have been short because of the small size of the hatchling shell called the ammonitella and the presumably small size of the yolk.

Modern Nautilus, with its large (about 25 mm in diameter) single egg with a large yolk serves as a model for fossil coiled nautiloids. In the embryo of Nautilus, a calcareous plate forms as a single unit with the cicatrix scar in the center of the plate. At the edges of the plate incremental growth occurs with the animal facing the large yolk. Continued growth occurs incrementally including the placement of about seven septa and the body chamber when hatching occurs. The time of hatching is marked by a faint constriction at the aperture, the conch diameter is about 25 mm, and after hatching there is a gradual change of ornament and an increase in sublethal shell repairs (see Arnold et al. 1987 for additional details). Eggs are laid singly in aquariums, and are usually attached to hard substrates. Hatching time is long (6 months to almost a year) (Carlson 1991; Uchiyama & Tanabe 1998). The presumption that most coiled nautiloids in the Carboniferous closely followed this reproductive model is not correct as many newly hatched fossil nautiloids have been recovered in sediments that are mud bottoms, not the hard substrates that modern Nautilus prefers. This suggests that some fossil nautiloids were able to lay eggs on the soft sediment or on hard substrates, such as empty shells, on soft muddy substrates. We find no support for the possibility that they laid their eggs in the water column, at least in the case of the Ruddle Shale locality.

Pseudorthocerid reproduction is thought to follow that of Nautilus. Some taxa have a constriction or change their ornament at a position that could be considered the time of hatching (e.g. Kröger & Mapes 2004). The eggs of some coleoids are known to stretch as the embryo develops (Boletzky 1987a). The eggs of the Pseudorthocerida were also probably relatively long and could have stretched in order to accommodate the long embryonic shell that developed sequentially like that of Nautilus. Because the presumed embryonic shells in different fossil taxa in these kinds of nautiloids are variable in length, egg size must have been variable. It is unknown if the eggs were laid singly or in clusters, and the place where the eggs were reposited or attached is also unknown.

Bactritoids have small hatchlings, similar to those documented in the Orthocerida (Kröger & Mapes 2007). The completely developed bactritoid embryonic shell at the time of hatching includes a spherical to subspherical protoconch, three chambers, and a relatively long body chamber that gradually decreases in diameter towards the aperture (Fig. 7) (Mapes 1979; Hecht et al. 1996; Doguzhaeva 1996, 2002). To accommodate the mature embryo, the eggs were probably relatively long and cylindrical with a small diameter (1–3 mm in length and 0.5 mm in diameter). The stretching of the egg capsule is also a possibility in this group. Initially, the embryo probably secreted the entire protoconch without a septum as is postulated for ammonitellas, and then the bactritoid embryo proceeded to create the complete embryonic shell by incremental growth. This includes the sequential formation of the first septum that separates the protoconch from the body chamber, with the eventual creation of three additional chambers and a long body chamber for a fully formed embryonic shell. Hatching time is indicated by the wide constriction (i.e. the place where the diameter of the shell attains the maximum amount of reduction in diameter in the body chamber) (Figs 7, 8). This constriction is equivalent to the nepionic constriction seen in Nautilus and the primary constriction seen in ammonitellas. Subsequent growth in the post-hatching phase includes body chamber lengthening, increased diameter of the conch, and additional chamber formation until maturity is reached.

Figure 7.

 Schematic diagram showing the generalized early ontogenetic sequence of shell development of a typical Carboniferous bactritoid. inline imageA. Embryonic shell, before hatching; it consists of an initial protoconch, a primary septum, one or more additional septa, and a body chamber. inline imageB. Bactritoid hatchling; it consists of an initial protoconch, a primary septum, three additional septa, and a body chamber with a terminal constriction. inline imageC. Post-hatchling; additional septa built within the constricted shell of the hatchling and additional chambers in the expanding shell that is deposited after the hatching constriction; additional body chamber.


The facies and fauna of the Ruddle Shale Member at the Buffalo Wallow location suggest oxygen depletion during deposition. There were probably episodes with slightly increased oxygen concentrations that allowed a low diversity benthic assemblage of a few minute species. Even with episodic low oxygen increases, spawning and hatching on the sea bottom were impossible for molluscs (cephalopods and gastropods) due to oxygen depletion.

The rich microgastropod assemblage consists of isolated planktonic caenogastropod and neritimorph larval shells as well as early juveniles of other clades. Planktonic caenogastropod larval shells are generally smaller than 1 mm but some are as large as 1.3 mm. No larger, sexually mature gastropod individuals lived at the locality.

Many of the cephalopods (all Ammonoidea, Bactritoidea, and one pseudorthocerid taxon) occur as hatchlings and/or embryonic shells despite oxygen depletion in the palaeoenvironment. Nautilida and most Pseudorthocerida have not been recovered as hatchlings or embryos but only as adults or juveniles, suggesting that these Nautiloidea did not lay their eggs in the water column but probably attached them to the substratum in an area distant from the collecting site where bottom conditions were probably less stressed. This mode of reproductive palaeobiology for the Nautilida and certain Pseudorthocerida is closer to that of modern Nautilus than to that of ammonoids, bactritoids, and fossil coleoids.

As is the case of some modern coleoids, the ammonoids and bactritoids probably laid their relatively small eggs in gelatinous masses suspended in the water column above the bottom or by attachment of the egg masses to floating objects over the collecting site (Fig. 9). The natural mortality of the hatchlings, early juveniles and embryos in the eggs is reflected by the sinking of their carcasses in various stages of ontogenetic growth into the oxygen depleted water column to the bottom where they were preserved. More mature specimens of Ammonoidea, Bactritoidea, and the pseudorthocerid taxon are also present, suggesting that some individuals hatched, grew to maturity, reproduced, and died in the water column above the collection site.

Figure 9.

 Schematic reconstruction of the pelagic ecosystem producing the Ruddle Shale minute mollusc taphocoenosis with special reference to bactritoid and ammonoid egg-laying behavior (larger, adult cephalopods are omitted for clarity). The upper water column is well oxygenated with abundant living cephalopods and their egg masses as well as gastropod and pelecypod veliger larvae. However, the bottom water is anoxic or has very low oxygen concentrations. The fact that bactritoids and ammonoids occur at all ontogenetic stages including pre-hatchlings and hatchlings suggests that reproduction occurred nearby, probably in floating or suspended egg masses. The taphocoenosis is largely devoid of benthic organisms that would be typical for an oxygenated Late Palaeozoic invertebrate community, e.g. brachiopods, echinoderms or normal sized, adult gastropod and pelecypods.
A1. Living bactritoid hatchling and (A2) dead shell.
B1. Living bactritoid embryo (pre-hatchling) and (B2) dead shell.
C1. Living ammonoid hatchling and (C2) dead shell.
D1. Living planktotrophic gastropod veliger larva and (D2) dead shell.
E1. Living planktotrophic pelecypod veliger larva and (E2) dead shell.

We suggest that caution be used in employing these conclusions for late Palaeozoic cephalopods in general as these interpretations are based on a single locality and only a few cephalopod taxa are involved. It is likely that other reproductive strategies have been employed at different times by other cephalopods. However, the presence of abundant hatchlings and pre-hatchlings in an offshore setting with oxygen depletion on the bottom is convincing evidence that floating egg masses of ammonoids and bactritoids were present (Fig. 9).


Acknowledgements. –  We would like to thank Ohio University for supporting part of this research. Additionally, R.H.M. is grateful for the numerous students who collected the exposure over the past 32 years. We thank Björn Kröger (Berlin) and Peter Alsen (Copenhagen) for their helpful reviews. Additionally we wish to thank Dieter Korn (Berlin) for his assistance in providing updated ammonoid identifications for the ammonoids from the Buffalo Wallow site. Partial support of this research to R.H.M. was provided by the National Science Foundation Grant EAR 012549. A.N. acknowledges technical assistance by Barbara Seuß (Erlangen) and research grants by the Deutsche Forschungsgemeinschaft (projects NU 96/3-1, 3-2; NU 96/10-1).