Main

The holotype of O. finneyi shows the cranial end of the left stylohyal resting on the tympanic bone (Fig. 1c–e). However, the stylohyal on the right side is in a different position, the tip of the stylohyal extends beyond the tympanic on both sides of the skull, and both tympanics are crushed. In our opinion, the skull is too deformed to provide evidence of the spatial relationships of these bones in life. Micro-computed tomography (MCT) images of the skull make clear the extent of crushing and fragmentation (Fig. 1b, d). Veselka et al.8 noted that the stylohyal–tympanic contact on the left side might be a taphonomic artefact, but nevertheless favoured the interpretation that O. finneyi was an echolocating bat. Available evidence indicates otherwise.

Figure 1: Location and orientation of the stylohyal in an extant echolocating bat and in O. finneyi , as revealed by MCT scans.
figure 1

a, Lateral view of the skull of Myzopoda aurita (United States National Museum (USNM) 449282), an extant echolocating bat; note the depth of the braincase and ring-like tympanic (indicated in orange) where the stylohyal (red) articulates with the base of the skull. The cranial end of the stylohyal is expanded to form a bifurcated tip, and the stylohyal is fused to the tympanic along the length of its course across that bone. b, Lateral view of the skull of the holotype of O. finneyi (Royal Ontario Museum (ROM) 55351A). The cranium is crushed flat and lies directly under a thin, dense sediment layer (shown in blue). This layer is free of any sculptured bone fragments; the layers below it include multiple bone fragments that are all that remain of the braincase and rostrum roof. c, Ventral view of the same specimen (ROM 55351A). Both ear regions are preserved but crushed flat. The right and left stylohyals lie at different angles relative to the tympanic ring, indicating that neither was fused to the tympanic. d, e, Individual MCT slices through the basicranial region of the same specimen, with d through a plane slightly dorsal to e. The stylohyals are marked with numerals indicating thirds from the cranial end (1) to the distal end (4). The right stylohyal runs perpendicular to the tympanic and is fractured between points 2 and 3. The left stylohyal runs parallel to the edge of the tympanic ring (90° offset from the right stylohyal) and shows evidence of fractures between points 1 and 2, 2 and 3, and 3 and 4. Neither element shows unambiguous articulation with the tympanic or any other bone. Scale bars, 10 mm.

PowerPoint slide

Full size image

Four osteological traits have been postulated as indicators of laryngeal echolocation in bats: (1) an enlarged orbicular apophysis on the malleus3,4; (2) an enlarged cochlea3,4,5,6,7; (3) an enlarged paddle-like or bifurcated cranial tip on the stylohyal3,4; and (4) an articulation between the stylohyal and the tympanic8. Studies in other groups (for example, talpid moles9) indicate that large orbicular apophyses may occur in non-echolocating lineages, hence this trait cannot be considered a definitive indicator of echolocation8. However, the hypothesis that relative cochlear size is a good indicator of the echolocation abilities of bats3,4,5,6,7,8 has not been refuted. The cochlea of O. finneyi falls outside the size range seen in living echolocating bats and is similar to the proportionally smaller cochleae of bats that lack laryngeal echolocation4,8, suggesting that it did not echolocate.

Data presented by Veselka et al.8 indicate that cranial expansion of the stylohyal and an articulation between this structure and the tympanic are 100% correlated in extant bats. Previous reports that two families of echolocating bats (Nycteridae and Megadermatidae) lack stylohyal modifications3,4,10,11 overlooked expansions of the stylohyal where it articulates with the tympanic. We found uniform presence of expansion and flattening of the stylohyal in both families. Observed correlations across all extant bat families indicate that this is a definitive marker of laryngeal echolocation, and that expansion and flattening of the cranial stylohyal should be considered a fundamental part of the stylohyal–tympanic articulation rather than an independent feature. In O. finneyi, the stylohyal is rod-like and has no cranial expansion or flattening other than a tiny knob at the proximal end. We hypothesize that this knob might be an ossified, fused typanohyal, which in some non-echolocating bats (for example, Rousettus, Eonycteris12) and insectivores (Echinosorex, Erinaceus13) is connected to the stylohyal by a thin ligament or cartilage; regardless, it is not comparable to the condition seen in any extant echolocating bat. In contrast with Veselka et al.8, we conclude that O. finneyi did not have a stylohyal–tympanic articulation as it clearly lacks one of the definitive components of this feature: a modified stylohyal with an expanded and flattened cranial end.

Reconstructions of behaviours of extinct animals require careful consideration of preservation artefacts in fossils as well as patterns of form and function among extant animals. Our analyses show that the only two unambiguous pieces of evidence available at this time (cochlear size and stylohyal morphology) support the hypothesis that O. finneyi was not an echolocating bat. Because postcranial morphology indicates that O. finneyi could fly and phylogenetic analyses place it on the most basal branch within Chiroptera4, the ‘flight first’ hypothesis for the origin of flight and echolocation in bats3,4 remains the best-supported hypothesis for the origins of these key features.

Methods

Micro-computed tomography (MCT) images of O. finneyi (Fig. 1b–e) were obtained with an MCT apparatus using a special ‘region of interest’ algorithm (RayScan 200 XE, RayScan Technologies). CT data for Myzopoda aurita (Fig. 1a) were provided by the University of Texas CT laboratory. Image processing was done with VGStudio MAX 2.0.1 (Volume Graphics).