Pteropus allenorum, Helgen & Helgen & Wilson, 2009

Helgen, Kristofer M., Helgen, Lauren E. & Wilson, Don E., 2009, Pacific Flying Foxes (Mammalia: Chiroptera): Two New Species of Pteropus from Samoa, Probably Extinct, American Museum Novitates 3646, pp. 1-40 : 6-14

Pteropus allenorum , new species

HOLOTYPE: The holotype of Pteropus allenorum is ANSP 1234, a skin in alcohol ( fig. 3 ) with the cranium and mandible removed and cleaned ( fig. 7 ), collected at Apia (13 ° 499S, 171 ° 449W), a harbor and settlement on the island of Upolu (and today the capital of Independent, or Western, Samoa; fig. 2 ), apparently by H.C. Caldwell in April 1856 (see below), and donated to the museum by W.S.W. Ruschenberger. Judging from its craniodental development ( fig. 7 ; see Helgen, 2004a; Giannini et al., 2006), this unsexed specimen is a nearly mature subadult.

A faded tag bearing scripted ink writing accompanies the skin in alcohol, which we take to be the original or at least the oldest tag associated with it. Both this tag and another— affixed to the specimen, less faded, and bearing writing in pencil—give the provenance of the specimen as ‘‘Apia, Upolu’’. Written in ink on the mandible is the faded annotation ‘‘Apia’’. The accession catalog, skull box, and a penciled label inside the skull box give the locality as ‘‘Navigator Islands, Apia, Apola’’. (The ‘‘Navigator Islands’’ is a 19th-century appellation for the Pacific archipelago today known as Samoa, incorporating the modernday political boundaries of both Independent Samoa and American Samoa; see Wilkes, 1844; Keesing, 1934.) The relatively large and high island of Upolu (area 1100 km 2, maximum elevation ca. 1100 m), home to the harbor of Apia, is the second largest island in the Samoan archipelago (after the adjacent island of Savai9i, with area 1820 km 2 and maximum elevation ca. 1850 m).

The faded tag in the alcohol jar with the holotype also bears the date ‘‘ April 1856 ’’. We consider this most likely to be the specimen’s date of collection, rather than the date of accession at ANSP. (Other specimens listed on the same page of the ANSP mammal accession catalog list a ‘‘ Date of Presentation’ ’ to the museum a decade later, in 1865 or 1866.) Tags associated with the jar of alcohol and the skull bear the name ‘‘ Dr. W.S.W. Ruschenberger’ ’ or ‘‘ W.S. W. R.’’, who is listed in the ANSP accession catalog under the column of ‘‘Donor’’, rather than ‘‘Collector’ ’.

Amongst ‘‘Donations to the Museum’’ received in 1857, the Proceedings of the Academy of Natural Sciences of Philadelphia gives the following entries, listed consecutively ( Anonymous, 1858: i):

A collection of Echinodermata, Acelephae, and Mollusca in alcohol, from the Navigator Islands. Presented by Drs. W. S. W. Ruschenberger, and Henry Clay Caldwell, U.S. N[avy].

A specimen of Pteropus from the same locality.

Based on its unique and concordant label data (noting its collection in Samoa in 1856 and presentation to the museum by Ruschenberger) we strongly suspect that this latter specimen mentioned amongst the Proceedings donations in 1857 is the holotype of P. allenorum , and not any other specimen of Pteropus currently in the collections (or listed in the catalogs) at ANSP.

Fowler (1901) discussed and described ichthyological collections from Samoa likewise donated to ANSP by Ruschenberger and Caldwell, of which he noted: ‘‘The following specimens were collected many years ago by Dr. H.C. Caldwell, by whom they were presented to the Academy.’’ Fowler also described the new taxon Mugil caldwelli , noting that he ‘‘named this species for Dr. Caldwell, who collected the type.’’ We have been unable to discover if Ruschenberger actually took part in the collecting efforts in Samoa in 1856, or if he was simply a financial sponsor of these exploratory efforts. It is clear, however, that Caldwell was directly responsible for the collection of some zoological specimens during this voyage to Samoa. We suspect based on the evidence at hand that he (or his assistants and colleagues during his visit to Apia) was the collector of the holotype of Pteropus allenorum .

The holotype is the only specimen of Pteropus allenorum known to us.

DIAGNOSIS: Pteropus allenorum is a relatively small to medium-sized ( figs. 4–8 ) member of the genus Pteropus (forearm 116 mm in the young holotype), probably with a brown head, tinged with russet; a golden-brown mantle, dusky brown back, and warm brown limbs and wing membranes; very small cheekteeth (with an upper cheektooth size gradient such that P3. P4, M1, according to overall bulk), but proportionally large canines and incisors; a moderately elongate rostrum; and a relatively gracile mandible.

DISTRIBUTION: Pteropus allenorum is recorded historically only from the Samoan island of Upolu. We speculate that, like most native elements in the Samoan avifauna ( Steadman, 2006b), the actual historic or prehistoric distribution of this species was not limited solely to this single island, but probably extended to Savai9i and to other islands of Samoa, if not to adjacent archipelagos (even if Upolu truly was its last place of occurrence). Further excavations of subfossil material in Samoa and further study of subfossil material from the adjacent archipelagos of Tonga and Fiji may help to clarify the past distribution of this species. It is not yet reported from the subfossil record of Tonga, the only Polynesian archipelago where the chiropteran subfossil record has been studied in some detail ( Koopman and Steadman, 1995), although in light of the elucidation of this species in the historical fauna of Samoa, closer study of Pteropus osteological material from Tongan excavations is certainly warrant- ed. Although likely extinct, we suggest that P. allenorum should be sought after during future biotic inventory efforts in Samoa on the chance that an overlooked extant population survives somewhere in the archipelago (see Discussion, below).

ETYMOLOGY: We have chosen the specific epithet allenorum to honor the name of Allen, in the plural. The epithet simultaneously acknowledges Harrison Allen (1841–1897), a zoologist, anthropologist, and physician ( Hrdlicka, 1914), who assembled much of the ANSP chiropteran collection in the late 19th century, and Allen Drew, who kindly hosted the Helgens during a visit to Philadelphia in 2006, during which the holotype of allenorum was first examined. We suggest ‘‘Small Samoan Flying Fox’’ as an appropriate common name.

DESCRIPTION: As noted above, the only available specimen of Pteropus allenorum is represented by a skin stored in alcohol with an accompanying skull that, although broken, preserves most cranial features. The skin is fragmentary and fragile but includes the head skin (separated from the rest of the body), most of the dorsal skin of the body, the limbs and wing membranes (somewhat decayed and partly discolored), and some other small clumps of fur. Because it has been preserved in alcohol for more than 150 years and its overall state of preservation is poor, it is difficult to characterize the original external appearance of this specimen. Based on the single specimen available, we think that the best that can be stated is that the holotype probably had a brown head tinged with russet, a golden-brown mantle, a dusky brown back, and warm brown limbs and wing membranes. We suggest that the general appearance of P. allenorum was probably that of a rather ‘‘furry’’ flying fox, similar in pelage quality to P. samoensis (see photograph in Flannery, 1995), in which the fur is longish and not strongly adpressed, as opposed to many flying foxes (such as P. tonganus ), in which the fur tends to be shorter, sleeker, and often clearly adpressed dorsally. The lengths of the hairs in the fur on the mid-back reach to about 25 mm. The fur is sparser and paler on the front of the face, and there is no darkened eye-ring encircling the eye. The forearm in the holotype measures 116 mm, and we expect that fully grown adults would have a forearm length of ca. 116–125 mm, smaller than either extant Samoan congener, Pteropus samoensis and P. tonganus (see table 1).

The holotype cranium is broken behind the orbits, but most of it is preserved in two intact pieces, which we carefully reconstructed to prepare an image of the overall skull ( fig. 7 ). Compared to other Pteropus , the rostrum is of ‘‘moderate length’’ sensu Andersen (1912). Despite its youth, the postorbital processes are rather well developed. The back of the palate forms a broad ‘‘U’’ shape. We estimate the condylobasal length of the (nearly mature but broken) holotype skull of allenorum to be 50 mm and the zygomatic width to be ca. 26 mm; measurements of full-grown adults would thus somewhat exceed these values, particularly in zygomatic width.

A striking feature of P. allenorum is the small size of the teeth relative to the size of the skull, even when perfectly unworn, as in the holotype. As indicated above, the very small skull and teeth of P. allenorum allow for its instant discrimination against the sympatric large-toothed forms P. tonganus and P. samoensis ( figs. 4–8 ; table 1). The upper dentition is largely complete in the holotype (right C1 is loose from the jaw but preserved in the box; right P1 is missing, represented only by an empty alveolus). The upper incisors are proportionally very broad ( fig. 7 ). The canines are long and narrow, with a moderately developed posterior cingulum. P1 is present. The cheekteeth posterior to P1 are relatively very small and narrow. The length of the maxillary toothrow (C1–M2) measures 18.0 mm, markedly smaller than in Samoan congeners (table 1; fig. 5 ). The soft palate is not preserved.

The holotype mandible is complete and gracile in overall appearance. It has comparatively weak posterior processes ( fig. 7 ), resembling P. fundatus of Vanuatu ( Felten and Kock, 1972), yet it is more gracile in the reduced size of the coronoid and angular processes despite its slightly larger overall size compared to that species. The incisors and canines and most of the right premolars and molars have been dislodged from the mandible but are preserved separately in the skull box (we reinserted these to provide the dorsal view of the mandible and its dentition, fig. 7 ). The lower canines and incisors are much larger (both in absolute and relative terms) than the corresponding teeth in P. fundatus . The posterior premolars and molars are small and rather narrow, but p1 is especially well developed.

Overall, the skull of P. allenorum requires closest comparison with P. fundatus , a flying fox of similar cranial and dental size, also endemic to a remote Pacific archipelago ( Vanuatu). The premolars and molars of P. allenorum match those of P. fundatus closely in both shape and absolute dimensions (table 2). More striking dental contrasts between P. allenorum and P. fundatus lie in the comparative size of the anterior dentition. In P. allenorum the incisors are much broadened and the canines vertically and anteroposteriorly relatively more elongate than in P. fundatus ( fig. 7 ). We suggest that this juxtaposition of relatively less massive cheekteeth but more massive incisors and canines in P. allenorum relative to P. fundatus probably reflects salient differences in feeding mode and ecology between these two ecomorphologically distinctive taxa. P. allenorum also can be distinguished immediately from P. fundatus by its darker coloration ( P. fundatus is a pale flying fox; fig. 9 ) and its larger body and skull size (the forearm measures 95–102 mm [n 5 21] and condylobasal length measures 44–49 mm [n 5 17] in adult P. fundatus ; Felten and Kock, 1972; Flannery, 1995; compare to P. allenorum in table 1). Given their morphological similarity and geographic proximity in the remote Pacific, P. allenorum and P. fundatus may be close relatives within the genus Pteropus . However, we stress that similarities between them may not necessarily reflect a close phylogenetic relationship, but perhaps instead a shared pattern of ecomorphological convergence on isolated Pacific archipelagos. A study drawing on molecular sequence data from a wide taxonomic and geographic sample of Pteropus is needed to distinguish between these possible scenarios (see Giannini et al., 2008).

The dentition of Pteropus allenorum is considerably less reduced than that of Pteropus scapulatus (of Australia and southern New Guinea) and the presumed phylogenetic allies of that species ( P. woodfordi and P. mahaganus of the Solomon Islands and P. gilliardorum of the Bismarck Archipelago), which have relatively small and nearly featureless cheekteeth and together are thought to constitute a distinctive Australian and Pacific group of specialist nectar-feeding flying foxes ( Thomas, 1888; Andersen, 1912; Sanborn, 1931; Van Deusen, 1969; Flannery, 1995; Bonaccorso, 1998; Helgen, 2004a; although see Giannini et al., 2008). However, Pteropus allenorum shares with these species a similarly gracile mandible, relatively large canines, small cheekteeth, and similar body size (e.g., fig. 7 ), attributes suggestive of similarities in lifestyle, although likely convergently derived.

Like P. allenorum , Pteropus vetulus Jouan, 1863 (a New Caledonian endemic), is a small Pacific Pteropus with very small cheekteeth, broad upper and lower incisors, and a conspicuously large p1. Like P. allenorum , it is also a rather darkly colored Pteropus with thick and furry rather than adpressed dorsal pelage (see photograph published by Flannery, 1995: 299). Pteropus allenorum is distinguished from P. vetulus by its larger skull and body size (the forearm measures 92–112 mm in adult P. vetulus ); dorsoventrally longer canines (short and deep in P. vetulus ); simpler upper first molar (labial edge subdivided into two distinct cusps in P. vetulus ); proportionally more elongate and evenly sloping rostrum (blunter, with more precipitous slope downward from the braincase in lateral profile in P. vetulus ); upper cheektooth size (bulk) gradient such that P3. P4, M1 (P3, P4. M 1 in P. vetulus ); less pronounced occipital cresting (strongly marked, with a subrectangular posterior braincase conformation in P. vetulus ); and more gracile mandible with a less developed angular process and lower coronoid process (slightly more robust with broader and higher respective processes in P. vetulus ). Selected wing and leg measurements in the holotype of P. allenorum are as follows (in mm, measured from wet skin): pollex (with claw) 48.5, (without claw) 42.6; D2M 58.7; D2P1 17.5; D2P2 12.5; D3M 78.4; D3P1 57.3; D3P2 67.1; D4M 77.9; D4P1 50.6; D4P2 45.5; D5M 85.0; D5P1 37.7; D5P2 38; tibia ca. 52; hindfoot (with claws) 30, (without claws) 28.

NATURAL HISTORY: That Pteropus allenorum has such a markedly reduced dentition relative to its extant and extinct Samoan congeners undoubtedly reflects differences in its overall diet relative to those species. It may

TABLE 2 Measurements of Premolars and Molars in Pteropus allenorum and P. fundatus, Two Relatively Small Pacific Pteropus L indicates length; W, width

TABLE 3 Principal Components Analysis Comparing 36 Pteropus Skulls from Samoa Factor loadings, eigenvalues, and percentage of variance for the first three principal components in a principal components analysis are included. Principal components are extracted from a covariance matrix of 10 log-transformed cranial and dental variables; see figure 8 , where the first two components are plotted.

have specialized as a nectar-feeder or by eating inflorescences or smaller fruits and nuts than those typically utilized by these larger bats, as the smaller-toothed species P. mahaganus and P. fundatus ( fig. 7 ) do in the Solomon archipelago and Vanuatu, respectively ( Bonaccorso, 1998; Flannery, 1995). However, no notes about ecological attributes or context of collection are associated with the holotype of Pteropus allenorum , such that nothing is firmly recorded of its basic biology.

The morphological resemblance between P. allenorum and P. fundatus of northern Vanuatu ( fig. 7 ) may indicate that similar small-bodied flying foxes were formerly more widespread throughout the Pacific theatre. If so, their discovery in the modern faunas or subfossil records of Fiji, Tonga, and on other islands of the Vanuatu and Samoan archipelagos might be expected in the future. For example, the current geographic restriction of P. fundatus to the small outlying Banks and Torres island groups of Vanuatu ( Flannery, 1995) seems likely to be a relictual distribution indicative of ‘‘pseudoendemism’’ in that archipelago ( Steadman, 1997, 2006a). In this vein, we note with interest that Hickey (2007) discussed a small, dark-colored, inflorescenceeating Pteropus observed on the large island of Malekula in Vanuatu. Hickey (2007) speculated that if not a juvenile P. tonganus , this might be a small flying fox related to P. fundatus . The chiropteran faunas of Vanuatu and Fiji undoubtedly remain incompletely inventoried, so future discoveries are certainly to be expected ( Helgen and Flannery, 2002; Helgen, 2004c, 2005).

Figures

Fig. 3. The fragmentary holotype skin of Pteropus allenorum (ANSP 1234, preserved in alcohol).

Fig. 7. Skulls of smaller-toothed, medium-sized Pteropus species of the southwest Pacific region. A, Partially cleaned skull of Pteropus fundatus, endemic to Vanuatu (AM M26897, adult male, Mota); B, reconstructed skull of Pteropus allenorum, endemic Samoa (ANSP 1234, unsexed subadult, Upolu); C, skull of Pteropus mahaganus, endemic to the Solomon Archipelago (AM M6280, adult male, Bougainville). Scale bar 5 10 mm.

Fig. 2. A map of Samoa, showing political boundaries and principal islands, with inset showing the location of Apia on the island of Upolu, the type locality of Pteropus allenorum. Adapted from Steadman (2006b).

Fig. 4. Bivariate ecomorphological contrasts in Samoan Pteropus. A plot of rostral (orbitonasal) length versus zygomatic width discriminates the four Pteropus species recorded from Samoa. Sample represents all adult specimens (or nearly adult in the case of the unique holotype of P. allenorum) from the Samoan archipelago. Open triangles indicate P. samoensis; closed diamonds, P. tonganus; closed square, P. allenorum; open circles, P. coxi.

Fig. 5. Bivariate ecomorphological contrasts in Samoan Pteropus (continued). A plot of maxillary toothrow length versus the distance across the upper canines also discriminates the four Pteropus species recorded from Samoa. Sample and symbols as for figure 4.

Fig. 6. Maxillary toothrows of the four Pteropus species of Samoa, to scale. A, Pteropus allenorum (ANSP 1234, unsexed subadult); B, Pteropus samoensis (ANSP 1867, unsexed adult); C, Pteropus coxi (USNM 3791, adult, probably male); D, Pteropus tonganus (AMNH 68738, adult male). Scale bar 5 5 mm.

Fig. 8. Multivariate morphometric comparisons in Samoan Pteropus (principal components analysis), drawing from 10 log-transformed craniodental measurements, divide the four Pteropus species recorded from Samoa into four discrete quadrant clusters. Sample and symbols as for figure 4. In this case, overall size can be visualized on the first component (increasing from right to left), while the loadings on the second component serve as an indication of general ‘‘robustness’’ (increasing from top to bottom). See text and table 3. Parenthetical numbers on the axes indicate the proportion of variance for each principal component (see table 3).

Fig. 9. Skins of the endemic Pteropus of Vanuatu. A, P. fundatus, AM M26896, adult female, Mota. B, P. anetianus, USNM 278062, adult female, Espiritu Santo.