It is generally accepted that heteromorph ammonites were poor swimmers, on the basis of overall shape and position of the aperture, although the dispersive juvenile stages and some adult forms may have been quite mobile planktonic drifters (Ward 1986, 1979, Shigeta 1993, Westermann 1996). Klinger (1980) demonstrated that helically coiled and orthocone ammonites must have been oriented in life with the head downwards, and suggested that these animals may have foraged on the sea floor, perhaps jetting vertically upwards when disturbed. Ebel (1992) reconstructed a number of heteromorph ammonites (such as Scaphites and Nipponites) with large bodies unable to retract into the shell, and that overall they were negatively buoyant (the float being too small in proportion to the body). Instead, Ebel (1992) proposed that these animals crawled along the sea floor on muscular arms like modern Octopus.
Our model contrasts with Ebel (1992) in proposing that instead of the body being much larger than analogy with Nautilus would suggest, it was much smaller, and more like a gastropod. However, like Ebel (1992) and Klinger (1980) we believe many heteromorph ammonites were nektobenthonic. In support of this interpretation there is good evidence that at least some ammonites lived on the sea floor. Gut contents reveal that some ammonites fed on benthic foramanifera, ostracods, crinoids, crustaceans and other ammonites, perhaps also scavenging (Lehmann 1973, Lebrun 1996). The absence of a hyponomic sinus in most ammonites (and all heteromorphs) suggests the hyponome was weakly developed, if present at all, and the sort of manoeuvrable jetting seen in Nautilus is unlikely. Indeed Géczy (1960) has argued that ammonites probably lacked the ability to swim. That Jurassic and Cretaceous heteromorph ammonites are usually found in clay or marl rather than sandy facies also suggests that the heteromorph ammonites were substrate dependent (Marcinowski & Wiedmann 1990). There are examples of ammonite specimens that seem to have been preserved as part of the benthos (e.g., a school of Polyptychoceras, a hamiticonic form) "nesting" within the shell of an Eupachydiscus (Matsumoto & Nihongo 1979). Finally, periodic disappearance of some heteromorphs during times of bottom water anoxia has been noted (e.g., Scaphites and Hamites from the Western Interior sea, Batt 1989). In addition it may be noted that the Hamitidae are essentially confined to shelf sediments (Westermann 1996).
With a small, mobile body within the long body chamber, a heteromorph ammonite can be visualised as having been rather like a small octopus with a mobile burrow or cave. When feeding, the aperture would have been angled toward the sediment allowing the animal to forage the sediment, perhaps pulling itself along with its arms, digging up small molluscs, worms and crustaceans. When alarmed, the ammonite would have withdrawn deep into the shell, and waited for the danger to pass. Because of the change in position of the centre of mass, the shell would have rotated away from the substrate. Rather than swimming away, the ammonite may have relied on the mechanical defences of spines (making the shell difficult to crush) and ribs (to hinder peeling). The long body chamber of the ammonite would also have served to make the animal difficult to pull out. Similar adaptations are well documented in gastropods (Vermeij 1993). This mode of life is similar to that suggested by Kakabadzé and Sharikadzé (1993), but since the orientation of the shell relative to the substrate becomes a function of the position of the animal within the shell and the geometry of the shell itself; the requirement for movement of cameral fluids around the shell is eliminated.
Other interpretations of heteromorph ammonite morphology include partial emergence of the mantle cavity from the shell (Jacobs and Landmann 1993); investment of the shell within mantle tissues (Doguzhaeva and Mutvei 1993); and even a free living ammonite animal using the shell as an egg case (Lewy 1996). However, for all these interpretations the evidence is circumstantial at best. Our model has the advantage of requiring less divergence from modern cephalopods in terms of behaviour or anatomy, being in accord with the more conservative views on ammonite ecology, and suggesting the possibility of lifestyles hitherto unconsidered.