Pteropus coxi, 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 : 16-21

Pteropus coxi , n. sp.

USNM 3790 USNM 3791

‘‘Samoan Isle’’, skull, accompanying skin (3952) missing ‘‘Samoan Isle’’, skull, accompanying skin (3953) missing

Pteropus tonganus Quoy and Gaimard, 1830

USNM 3954 ‘‘Samoan Island’’, adult male, skin with skull in situ

USNM 3955 ‘‘Samoan Isle’’, adult, skin without skull

USNM 38681/3956 ‘‘Samoan Isle’’, subadult, skin (38681) and skull (3956)

remains unknown. Future subfossil excavations in Samoa may clarify its prehistoric or historical distribution.

ETYMOLOGY: The specific epithet honors Dr. Paul A. Cox of the Institute for Ethnomedicine in Jackson Hole, Wyoming, in recognition of his research on flying foxes, tropical ecology, and biodiversity conservation in Pacific archipelagos in general and in Samoa in particular. We suggest ‘‘Large Samoan Flying Fox’’ as an appropriate common name for this species.

DESCRIPTION: Nothing is known with certainty about the external appearance of P. coxi , because the skins originally associated with both skulls have been lost and cannot be traced (L.K. Gordon, in litt.). However, a few clues are available that may help to envision its appearance. First, the skull of P. coxi is larger and more robust than its sympatric congeners on Samoa, an indication that it was probably a heavier bat than P. samoensis and P. tonganus , with a somewhat longer forearm. Another clue to its appearance may be found in the original description of Pteropus samoensis by Peale (1848). The U.S. Exploring Expedition collected P. samoensis , P. tonganus , and P. coxi in Samoa in 1839-1841, but Peale (1848) implicitly referred all of these to P. samoensis in his original description of P. samoensis , noting:

The species was first discovered on the island of Tutuila, and subsequently in all the islands of the Samoan Group; we obtained numerous specimens, and found the young animals somewhat-lighter coloured than the old ones, but in other respects there is but little variation in colour or size.

In preparing his description of P. samoensis , we assume from his comments that Peale had all or most Samoan Pteropus specimens from the U.S. Exploring Expedition at hand (of these, those that are specifically localized indicate their collection on Tutuila, Upolu, and Olusinga [5 Olosega] Islands; table 4). The skins of both specimens of P. coxi (in addition to the still-available skulls) were apparently still extant in 1848 and were apparently referred to P. samoensis . As such, we suggest that these specimens can be assumed to have formed part of the original hypodigm of Pteropus samoensis (as can several specimens of P. tonganus collected in Samoa during the expedition; see below, where we designate a lectotype for P. samoensis to preserve its traditional taxonomic association). Peale’s (1848: 20–21) statement about variation (reiterated in the observations of Cassin [1858] a decade later) might allow us to infer that the skins of P. coxi were not especially different in overall pelage patterning and form from his series of true P. samoensis , which he characterized as ‘‘head … tawny, the front gray; ears small, rounded, black; neck of the old animals rufus [sic], in the younger animals tawny, body and throat reddishbrown; hair erect, and somewhat woolly, most smooth on the back; wing-membranes black; irides brown.’’ Although P. coxi is undoubtedly a larger bat than P. samoensis , Peale’s (1848) statement about the lack of noticeable size variation in his series also suggests that the skins of P. coxi were probably not markedly larger in apparent size as compared to true P. samoensis (and P. tonganus ).

The skull of P. coxi is larger and more robustly constructed than that of P. samoensis , with a broader rostrum and palate; a longer maxillary toothrow featuring more massive individual teeth; broader and thicker zygomata and postorbital processes; and a dentary with a deeper ramus, more massive coronoid, condylar, and angular processes, and more heavily sculpted surfaces for muscular attachments. The canines are more massive in general (vertically taller, broader, more anteroposteriorly elongate), and the upper canines bear more prominent internal ridging. In the holotype, the right upper canine bears a groove separating a small secondary cusp from the main body of the canine, and a similar, incipient secondary cusp can be seen on the left upper canine. The paratype, a somewhat younger animal judging from toothwear, also with massive canines, does not have these same secondary cusps. This trait may be individual in its variability, but it could also be a sexual difference. Males and females differ in canine size and robusticity in many species of pteropodids, including many if not most species of Pteropus ( Andersen, 1912: 75: ‘‘canines in males of nearly all species longer and heavier than in females’’). Individual premolars and molars are both wider and longer in P. coxi than in P. samoensis (e.g., table 1). P1 (upper) is very small but persists in the adult dentition. The first lower premolar (p1) is large. In general cranial robustness, particularly in the conformation of the mandible, P. coxi resembles P. anetianus of Vanuatu, with which it is most closely allied in our morphometric comparisons (e.g., fig. 14 ), but the two species differ both in cranial and dental size (greater in P. coxi ) as well as in the proportional length of the rostrum (longer in P. coxi ). In its extreme cranial robustness, large teeth, and incipient (if variable) secondary cusps in the upper canines, P. coxi might be regarded as ecomorphologically convergent to some extent on the cranially robust and large-toothed Pacific ‘‘monkey-faced bats’’ of the Solomon archipelago (genus Pteralopex ) and Fiji (genus Mirimiri ), which occur in the broader region but not in Samoa, and to which flying foxes are not immediately related ( Hill and Beckon, 1978; Flannery, 1995; Parnaby, 2002b; Helgen, 2005). (These large bats are thought to feed on nuts, hard, thick-skinned fruits, and perhaps tree exudates; Flannery, 1995; Fisher and Tasker, 1997; Helgen, 2005.)

NATURAL HISTORY: Nothing is known of the habits or biology of P. coxi , the largest of Polynesian flying foxes, and we assume that the species is now extinct, perhaps for longer than a century, as the only known specimens were collected in 1839–1841.

However, published observations from the early 1980s by the botanist Paul Cox, discussing a flying fox of unusual size in Samoa, could indicate that P. coxi survived until recently. In his paper on the bats of Samoa, Cox (1983) noted an encounter with ‘‘ Pteropus samoensis ’’:

I will never forget the first time I saw one of these giant bats in the rainforest. One day, while climbing a tree, I saw what appeared to be an eagle flying away from a liana flower. The bat I saw in my field glasses appeared to have a wingspan of five feet or more and lacked the white fur on the back of the neck that characterizes the locally common flying fox, P. tonganus . This large bat was black and its behavior was completely unusual. I later thoroughly enjoyed watching them soar, eagle-like, high above the forest in midday sun.

In this paper, Cox (1983) concluded that P. samoensis is perhaps the largest species of Pteropus anywhere—a strange claim, considering that P. samoensis is a medium-sized member of the genus, and even sympatric Pteropus tonganus is slightly heavier (despite equivalent forearm lengths in Samoa [table 1], body mass averages 383 g in three adult Samoan P. samoensis and 410 g in nine adult Samoan P. tonganus at USNM). After an attempt to collect verifying voucher specimens, and further reflection, Cox (1984a) noted:

Two specimens [later] shot by hunters were confirmed as P [teropus] samoensis by Dr. Karl Koopman at the American Museum of Natural History. These had wingspans of only 3.5 feet, the size considered typical for the species. The larger bats that I observed appeared to behave like normal P. samoensis , however their size raises the possibility that my observations include a second, but undescribed endangered species. These findings emphasize the paucity of available information and the urgency of a thorough investigation.

Understandably, most subsequent authors have attributed Cox’s observations of giant black bats to somewhat exaggerated or fantastical descriptions of Pteropus samoensis (e.g., Wilson and Engbring, 1992), a conclusion that Cox apparently later accepted ( Cox, 1999). However, we raise the possibility (however slight) that Cox instead encountered the species that we describe here as P. coxi , a very large species of Samoan flying fox that is probably closely related to P. samoensis , with which it must occur (or have occurred) sympatrically. If Cox truly observed P. coxi in the forests of Upolu in 1981, then he has provided the only known description of its external features—i.e., that it is ‘‘black’’, with a wingspan of ca. ‘‘5 feet’’ (5 1.5 m) ( Cox, 1983; later revised to ‘‘4 feet’’ [5 1.2 m] by Cox, 1999). Certainly neither P. samoensis nor P. tonganus achieves such a large body size in Samoa ( Wilson and Engbring, 1992). That Cox’s clearest observations of this large bat took place in upland forests during the day ( Cox, 1983, 1984a), when P. samoensis also flies (e.g., Andersen, 1912; Wilson and Engbring, 1992; Thomson et al., 1998, 2002), provides the only ecological information that might conceivably refer to P. coxi .

Cox’s (1984a) prescient call regarding the ‘‘urgency of a thorough investigation’’ into the possibility that more than two extant species of Pteropus persist in Samoan forests has gone unheralded. This urgency has increased with the passage of time and with the discoveries reported herein. To us, Cox’s account raises the slight possibility that populations of P. coxi could persist in Samoa, especially in remote montane habitats on the high forested islands of Upolu and Savai9i. This possibility should be pursued in further biological explorations in the archipelago and in interviews with local communities throughout Samoa. Should P. coxi persist in Samoa as a rare species, some field studies that have attempted to document the biology of ‘‘ Pteropus samoensis ’’ in recent years (e.g., Cox, 1983; Wilson and Engbring, 1992; Banack, 1996, 1998; Brooke, 2001) could conceivably be based on studies of more than one biological species (i.e., P. samoensis and P. coxi ), even after distinctions between P. samoensis and P. tonganus , sometimes confused in the past, were made clear to Samoan fieldworkers ( Wilson and Engbring, 1992). Although it is tempting to associate this newly elucidated taxon with Cox’s observations in the early 1980s, we suggest that these are best viewed with skepticism in the absence of any additional evidence for the recent survival of P. coxi . A dark brown color phase is the most common coloration pattern in Samoan populations of Pteropus samoensis ( fig. 11 ), and the wingspan of flying foxes is difficult to estimate from any distance. As such, Cox’s (1983, 1984) observation of very large, blackish bats in flight during the day could easily be based on sightings of P. samoensis . Furthermore, while 20th-century specimens of P. samoensis and P. tonganus from Samoa are not uncommon in museums (e.g., AM, AMNH, ANSP, USNM, ZMUC), we know of no museum specimens of P. coxi collected since the type series was taken more than 160 years ago. Regardless of whether more than two species of Pteropus survive today in Samoa, it is clear to us from our own experiences and from our close reading of relevant literature that it is indeed only P. samoensis and P. tonganus that have been regularly encountered by field biologists working in Samoa during the past 15 years (e.g., Wilson and Engbring, 1992; Craig and Syron, 1992; Elmqvist et al., 1992, 1994; Craig et al., 1994a, 1994b; Morrell and Craig, 1995; Grant and Banack, 1995, 1999; Banack, 1996, 1998; Pierson et al., 1996; Grant et al., 1997; Miller and Wilson, 1997; Richmond et al., 1998; Thomson et al., 1998, 2002; Brooke et al., 2000; Nelson et al., 2000; Webb et al., 2000; Brooke, 2001; Banack and Grant, 2002, 2003).

A LECTOTYPE FOR PTEROPUS

Figures

Fig. 14. Bivariate plot of the two canonical variates in a discriminant function analysis contrasting samples identified as Pteropus samoensis (Fiji and Samoa, triangles), Pteropus anetianus (Vanuatu, closed circles), and Pteropus coxi (Samoa, open circles). Parenthetical numbers on the axes indicate the proportion of variance for each variate (see table 5).

Fig. 11. Color variation in the extant Pteropus spp. of Samoa. A, P. tonganus (USNM 566603, adult female, ‘‘Samoa’’), dorsal coloration pattern; B, P. samoensis (USNM 338624, adult male, Tutuila), dark phase (most common); C, P. samoensis (USNM 565827, adult female, Tutuila), white phase (very rare); D, P. samoensis (USNM 3947, unsexed young adult, ‘‘Samoan Archipelago’’), pale, straw-colored phase (less common in Samoa, more common in Fiji).