Exceptional soft-tissue preservation of Jurassic Vampyronassa rhodanica provides new insights on the evolution and palaeoecology of vampyroteuthids

In their original description of V. rhodanica, Fischer & Riou¹⁶ determined that the previously undescribed genus was a Jurassic relative of V. infernalis. This assignment was based on the configuration of the arm crown and armature, fin type, presence of luminous organs, lateral eyes, and the absence of an ink sac. Assuming this assignment is correct, then V. rhodanica is a member of the suborder Vampyromorphina, which includes the family Vampyroteuthidae^22,29.

Reappraisal of the anatomy shows that V. rhodanica and V. infernalis both have 8 arms and uniserial suckers flanked by cirri. They both possess V. infernalis-like sucker attachments^34,36, which are broader at the base and taper up to a radially symmetrical sucker.

Both species have distinctly modified arms though the morphology differs in each. V. infernalis, has retractable filaments in the position of arm pair II^27,33,34, though there is no evidence of these appendages in V. rhodanica. Instead, the species has elongate dorsal arms (arm pair I) with a unique configuration of suckers and cirri on the distal section.

The suckers and cirri of V. rhodanica are more numerous than those of V. infernalis^27,37. They are also more closely positioned. Proportionally, the suckers of both species have a consistent ratio to mantle length³⁷, though the diameter of the cirri and infundibulum are greater in V. rhodanica. The V. infernalis-like attachment^1,3,34 is present in both species, though in V. rhodanica, the distal part of the neck protrudes into the acetabular cavity. Of note, the sucker stalks on the dorsal arms of V. rhodanica are more elongate than those on the other arms (Figs. 2b,c, and 3a,b). This variation in suckers and their attachments suggests a specialized function between the dorsal and sessile appendages. On the longer dorsal arms, the larger sucker diameter, and more elongate stalks (Figs. 2b and 4) indicate the potential for increased mobility over their extant relatives, and possibly facilitated additional manipulation and prey capture capability.

In addition to the arm crown specialization, V. rhodanica has a more streamlined shape than V. infernalis, which is caused by a proportionally narrower head. Their muscular body is narrower and more elongate than the gelatinous V. infernalis^16,27,37 suggesting a higher energy locomotory style. This is consistent with increased predation relative to the modern form. Observations in this study support many assertions of Fischer & Riou¹⁶ about the characters in V. rhodanica, though the presence of luminous organs cannot be confirmed. Rather than luminous organs much larger than those present in the deep-sea, extant V. infernalis, it is possible that these structures represent displaced cartilage prior to fossilization (Supplementary Fig. 6).

Two other genera from the La Voulte-sur-Rhône locality, Gramadella and Proteroctopus are, like V. rhodanica, considered to be Incertae sedis Vampyromorpha²². All three share morphological similarities that include an elongated mantle fused with the head, and a longer dorsal arm pair with armature on the distal ends^1,16,22,38. Neither the second nor fourth arm pair have been modified. Each has one pair of fins. In Gramadella, the fins are lateral and skirt-like^16,38. In V. rhodanica and Proteroctopus these fins are located posteriorly^1,16.

V. rhodanica shows the greatest length variation between the dorsal and sessile arms (Fig. 4), though proportionally, Gramadella, and Proteroctopus have longer dorsal arms^1,31. Fischer & Riou³¹ and Kruta et al.¹ described biserial suckers in their descriptions of Gramadella, and Proteroctopus, respectively. In Proteroctopus, these suckers have a proportionally smaller diameter than the uniserial row in V. rhodanica, and do not exhibit the same tapered pattern.

None of these specimens shows evidence of an ink sac, though it is present in contemporaneous genera from the same assemblage (Mastigophora, Rhomboteuthis and Romaniteuthis)^8,16. That this character occurs only in some taxa from the same assemblage suggests variation in ecology, possibly associated with the steep, bathymetric relief in the La Voulte-sur-Rhône paleoenvironment¹¹. The mosaic of characters found within the coleoid taxa at La Voulte-sur-Rhône suggests that Mesozoic vampyromorphs co-occurred in different ecological niches during the mid-Jurassic.

Today, extant V. infernalis is uniquely adapted to a low-energy, deep-sea mode of life^27,28,29,39, though the timing of character acquisition and progression of this ecology is unclear²⁴. It is hypothesised that the vampyromorph Necroteuthis Kretzoi 1942 was already exploiting this niche by the Oligocene²⁹, and that the initial shift to offshore environments was possibly driven by onshore competition^24,29. The data obtained here suggests that V. rhodanica, the purportedly oldest-known genus of the Vampyromorphina group, was an active predator following a pelagic mode of life.

Indeed, several anatomical details, mainly found in the brachial crown, seem to support this hypothesis. Though we cannot directly compare functionality of the arm crown elements with other Jurassic taxa, we can infer function based on observation in modern forms. In Octopoda, the sister group to Vampyromorpha, suckers are attached to the arm by a cylindrical layer of muscle, encircling oblique musculature^40,41, that connects the arm musculature and the lateral margin of the acetabulum^34,40,41,42. This facilitates a variety of functions including locomotion, manipulation, and prey retention⁴³. The sucker attaches by flattening the infundibulum against the surface and then the encircling epithelium creates a watertight seal^{36,40,41,42,43,44,45}. Contraction of the radial acetabular muscles provides the pressure differential required to create the suction force^43,44,46.

The stalked sucker attachments^2,34 of decabrachians (Fig. 3d, and Supplementary Fig. 4) are muscular³⁵ and connect the musculature of the arm with the base of the sucker, forming part of the acetabulum^33,34. Tension on the sucker stretches this muscular attachment, which pulls locally on the acetabular base. This facilitates a greater pressure differential inside the sucker, allowing the teeth on the sucker ring to maintain the hold⁴⁷.

Extant V. infernalis lack decabrachian-like stalks^2,18 and the neck of the attachment joins to the base of the acetabulum (Fig. 3c, and Supplementary Fig. 4), rather than being inserted into it¹⁸. The infundibulum is not distinct, and the suckers do not provide strong suction²⁷. Instead, suckers function by secreting mucus to coat detritus—marine snow captured by retractable filaments—which is then moved to the mouth by cirri^7,27.

A mosaic of these characters is present in V. rhodanica (Fig. 3a,b), therefore, suggesting their potential for increased attachment and hold on prey over extant V. infernalis. These include a larger infundibular diameter, a neck attachment integrated with the acetabular muscles, and the elongated stalks of the dorsal suckers.

Additionally, the paired, filamentous cirri observed in extant cirrates⁴⁸ are present in V. rhodanica (Fig. 4, and Supplementary Fig. 2). In extant forms they are understood to have a sensory function and are used in the detection and capture of prey⁴⁸. In V. infernalis, they serve to transport the food proximally along the arms to the mouth²⁷. The greater diameters of cirri, and placement along the entire arm in V. rhodanica (Fig. 4), suggests an increased sensory function in these fossil forms.

The shape of the arms also contributes to the suction potential⁴⁹ in coleoids. Functional analysis in Octopoda highlights a positive correlation between distal tapering of the arms and their flexibility. A tapered, flexible arm facilitates more precise adhesion than a cylindrical-shaped one and requires a greater force for sucker detachment⁴⁹. Suckers detach sequentially, rather than the more simultaneous release observed in models of arms with less taper variation. The tapered diameter of the suckers, like those seen on the sessile arms of V. rhodanica, potentially facilitated this kind of sequential detachment⁴⁹ allowing them more adherence force and flexibility. Though V. rhodanica has just two suckers on the distal tips of their dorsal arms, the most distal is marginally smaller in diameter than the proximal one. On the dorsal arms, this tapering is observed in conjunction with a well-developed axial nerve cord (Fig. 2b). In extant forms, the nerve cord facilitates complex motor functions⁴². The combination of these characters in V. rhodanica suggests their arms had increased potential to be actively used in prey capture⁵⁰ over extant V. infernalis.

Though arm crown characters offer insight on the ecology of V. rhodanica, in fossil coleoid phylogenies only a few characters are based on the suckers^{1, 3}. Two studies that have attempted to create a phylogeny using morphological characters that include both fossil and extant taxa return V. rhodanica and V. infernalis as sister taxa^1,3. These matrices are, by necessity, heavily influenced by the gladius⁵¹ and more than 50% of the characters are based on this feature^1,3. Indeed, the authors¹ note that the lack of gladius data for some fossil forms, including V. rhodanica, creates an inherent bias in the phylogenetic matrix. Fischer & Riou¹⁶ suggested that V. rhodanica and V. infernalis are related on the basis of the observable morphological characters in the family Vampyroteuthidae, though without morphological information on the gladius, a recent systematic synthesis of fossil Octobrachia²² positioned V. rhodanica as Vampyromorpha Incertae sedis.

X-ray CT analysis in this study did not allow a reconstruction of the gladius. Nevertheless, it does provide new data on soft tissues, and permits comparisons between extant and fossil taxa. Specifically, we can add distinct states to 4 of the 132 characters in the existing phylogenetic matrix from Sutton et al.³ that was modified and used in Kruta et al.¹. These four characters (#89–#92) represent the suckers, and sucker attachments. Detailed examination revealed that the sessile and dorsal arms have the Vampyroteuthis-like attachment. In the dorsal arms, this is more elongated, though it cannot be considered pedunculate like those seen in modern decabrachians. Indeed, the attachment type (plug and base³⁴) is the same, only the length varies. As previously discussed, this variation may have functional implications.

When updated with these new data, the matrix from this study returns the same topology seen in Kruta et al.¹ that supports the positioning of V. rhodanica and V. infernalis as sister taxa. Further, it strengthens their relationship as they both share a sucker attachment that is not clearly attached to the arm muscles, a state that was previously considered autapomorphic in V. infernalis. However, it is important to note that no additional characters were added for the gladius, which is the cornerstone of coleoid systematics⁵². Indeed, just 29 of the 132 matrix characters can so far be coded for V. rhodanica, with only 9 of these relating to the 74 states of the gladius.

Assuming the phylogenetic work so far is correct, then both species belong to the family Vampyromorphina, and are joined by the Oligocene fossil Necroteuthis hungarica²⁹. While the lack of gladius characters precludes a full phylogenetic understanding of this group, preservation and observation of the soft tissues allow us to infer information regarding palaeobiology.

The data obtained in this study demonstrates that the characters observed in V. infernalis, including the sucker attachments and lack of ink sac, were present in Jurassic Vampyromorpha. Comparative anatomy of V. rhodanica and extant V. infernalis revealed that the fossil taxon displayed more morphological variation and were more diversified than previously understood. The assemblage of characters observed in V. rhodanica are consistent with a pelagic predatory lifestyle and corroborate the likelihood of a distinctly different ecological niche. These findings support the hypothesis that a shift towards a deep-sea environment occurred prior to the Oligocene^5,29.

Source: Ecology - nature.com