The discovery of a fossil tortoise on Bermuda (Meylan and Sterrer 2000) provided another example of the extraordinary ability of tortoises to colonize remote oceanic islands over long distances of ocean. The Pleistocene Bermuda tortoise, Hesperotestudo bermudae (Meylan and Sterrer 2000), was hitherto known from a single but well-preserved specimen consisting of the anterior portion of carapace and plastron plus skull, vertebrae, and pectoral elements (Fig. 1). This was discovered in a dune deposit at a relatively high elevation (for Bermuda) in the middle of the main island of Bermuda (Olson et al. 2005, fig. 4) that was correlated with the Middle Pleistocene Upper Town Hill Formation, about 310,000 years old (Meylan and Sterrer 2000, Olson et al. 2005). This dune formed during the Marine Isotope Stage (MIS) 9 interglacial high stand of the sea when sea levels were from +3 to +5 m higher than at present (Hearty and Kaufman 2000).

View Figure

• Download as Powerpoint
• Download Figure

Olson et al. (2006) postulated that the specimen of H. bermudae may have been deposited near the point of extinction of the species, as no tortoises are known from any of the more abundant vertebrate fossil localities of younger age (Hearty et al. 2004; Olson et al. 2005). It was further suggested that the only time available for the colonization and evolution of H. bermudae would have been during the preceding glacial period MIS 10, for which no vertebrate fossil sites have yet been discovered on Bermuda (Olson et al. 2006). Any evidence that corroborates the existence of a viable population of tortoises on Bermuda would thus be significant.

Here we document a second specimen of H. bermudae, although it is one for which critical data have regrettably been lost. The specimen (Fig. 2) consists of the proximal end of a right radius (BAMZ 2004 228 011) that was included with a lot of fossilized bird bones that was found in the Bermuda Aquarium, Museum and Zoo, bearing only the number 0434 from a registry that can no longer be located. All but 1 of the bird bones are from the gadfly petrel known as the Cahow (Pterodroma cahow), once the most abundant breeding seabird on Bermuda. The other bone is a paratype of the endemic fossil hawk Bermuteo avivorus (Olson 2008). Most of these bones were encrusted with calcareous sand that included tubes of serpulid marine worms, and it was assumed that the bones had been collected in an underwater cavern that was connected with the sea. Amino acid analyses of the encrustations indicate a minimum age early in the last interglacial (MIS 5) about 120,000 years ago (Hearty and Kaufman 2008). The bones themselves could be much older, however, dating to some glacial period when sea levels were lower. The tortoise bone had no such incrustations, however, and so may have been reworked from another level than the bird bones.

View Figure

• Download as Powerpoint
• Download Figure

Despite the ambiguities surrounding the bone here referred to H. bermudae, it provides several points of information concerning the species. It is from an individual about twice the size of the holotype (c. 500 mm straight carapace length vs. 270 mm for the holotype), thus establishing something about the degree of size variation in the species. Its likely deposition during a glacial period and at a necessarily lower elevation than the holotype indicates a more expanded duration and distribution of the species than could be inferred from the original specimen. The reality of a formerly viable, evolving population of Pleistocene tortoises on Bermuda has thus received additional supporting evidence.