Leptotyphlops carlae, Hedges, 2008

Hedges, S. Blair, 2008, At the lower size limit in snakes: two new species of threadsnakes (Squamata: Leptotyphlopidae: Leptotyphlops) from the Lesser Antilles, Zootaxa 1841 (1), pp. 1-30 : 5-9

Leptotyphlops carlae sp. nov.

Fig. 1A

Barbados Threadsnake

Holotype. USNM 564819 , adult female, collected on 9 June 2006 under rock on ground near Bonwell , St. Joseph Parish, Barbados ( Fig. 2A ), 280 m (13° 11.196’ N, 59° 32.445’ W), by S. Blair Hedges and Carla Ann Hass. Field tag number 267708.

Paratypes. USNM 564818 , from same locality ; BM 89.7 .5.27, from Barbados (no specific locality) ; BM 1969.792 , from Codrington College , St. John Parish , Barbados, 100 m (13° 10.543’ N, 59° 28.481’ W); Cali- fornia Academy of Sciences ( CAS) 49279, “St. John, Antigua,” collected by W. K. Fisher in July, 1918, on the Barbados-Antigua Expedition. This locality is interpreted to be in error; it is likely from St. John Parish, Barbados (see discussion below) .

Diagnosis. A small species of the genus (104 mm maximum total length, TOL) and member of the West Indian bilineatus group ( Thomas et al., 1985) in having the ocular and lip separated by labial scales. It is most closely related to Leptotyphlops bilineatus ( Martinique) and an undescribed species from St. Lucia (see below) in having fewer than 290 middorsal scales, large eyes, and a dark body color with two pale dorsolateral stripes. In color pattern it differs from both in having continuous stripes from eyes to tip of tail with an additional pair of narrow middorsal lines ( Fig. 2A–B ). In scale characters ( Figs. 3–4 , Table 1) it differs from both species in having a narrower PA1 (width/length 1.56–1.74 vs. 1.84–2.28), a wider contact (suture) of PF and RO (PF-RO/PF-PN 0.78–1.2 vs. 0.31–0.61), and a shorter PA1-PA2 suture (0.63–0.68 vs. 0.69–0.85 % snoutvent length, SVL). In addition, it differs from L. bilineatus in having a narrower head (2.38–2.49 vs. 2.53– 2.68 % SVL), a shorter OC (0.65–0.73 vs. 0.78–0.84 % SVL), and shorter anterior middorsals, PF+FR+IP+IO (2.06–2.16 vs. 2.28–2.55). From the new St. Lucia species, it differs in having more middorsal scales (185– 192 vs. 176–183), a shorter OC suture (SO-OC/PN-OC 1.42–2.08 vs. 2.23–3.31), and a narrower SO (0.87– 0.96 vs. 0.99–1.16). Measurements and scale counts, including those of the holotype, are in Table 1.

Description. Body with 14 scale rows, reducing in region of vent; middle of tail with ten rows; four supralabials and four infralabials. Body stout; head narrow, tapering from behind eyes to snout and covered with numerous tubular scale organs; rostral narrow, tapering posteriorly to rounded margin, widest (and slightly protuberant) at tip of snout, narrower and concave on ventral portion; spine at tip of tail. In average width, PF> IO> IP> F. Eye large, 43–51% distance from naris to posterior edge of eye. Coloration in life dark brown to black above and below, with pale grayish-yellow dorsolateral lines extending from PA1 to tip of tail, defining a reddish-brown and slightly iridescent middorsal zone three scales in width, with narrower (inner) pair of lines along outer border of middorsal scale row; lower sides and venter pale grayish-brown, beginning 1.8 scales below dorsolateral lines; two pale spots on each PN scale, with occasional white markings on SO and PF scales; 5–11 white scales around vent. Live weight, 0.60 g (holotype).

Etymology. The species name is dedicated to my wife, Carla Ann Hass.

Comments. Until now, the two BM specimens of this species, one from 1889 ( Fielden, 1889; Boulenger, 1893) and the other from 1963 ( Underwood, 1963), have been the only documented evidence of the occurrence of Leptotyphlops on Barbados. Both have been confused with L. bilineatus of Martinique. The second BM specimen requires discussion because it has had a confused history. It was first reported by Underwood ( Underwood, 1963) as coming from “the vicinity of Codrington College,” St. John Parish, having been collected by Father E. J. Pearce. Several years later, Emsley ( Emsley, 1966) reported on what appears to be the same specimen, although he makes no mention of the earlier report by Underwood. However, Underwood is mentioned in the note as having examined the specimen for him and compared it with the 1889 BM specimen and one from Martinique. Although Emsley ( Emsley, 1966) mentioned that the specimen was in the collection of the Department of Zoology, University of the West Indies, Trinidad, it must be the same specimen as BM1969.762, now in the Natural History Museum (NHM), London. The collection dates (January, 1963), collector, and ecological notes correspond. Also, a check of the collection in Trinidad (A. Hailey, pers. comm.) confirmed that the specimen is no longer there. The possibility that this specimen is different from the one reported initially by Underwood ( Underwood, 1963) was considered because the date of publication of his book is reported in all literature as “1962,” which would predate the collection of the specimen. The 1962 date derives from a date of May, 1962 in the foreword of the book, written by Hector Wynter. However, Underwood’s supplement to the book, dated June, 1964, states that 12 months have passed since publication of the original book, giving a likely date of June, 1963 (not 1962) as the publication date of the original book. This would be consistent with the mention of the Barbados specimen collected in January, 1963. Therefore, I consider BM1969.762 to be the same snake reported by Underwood ( Underwood, 1963) and Emsley ( Emsley, 1966).

The CAS paratype (CAS 49279) has an equally confusing history. It has gone unnoticed by herpetologists studying West Indian snakes for nearly a century since it was collected, despite being in a major U.S. collection. The specimen locality is “St. John, Antigua,” collected by W. K. Fisher in July, 1918. However, it is 400– 500 km north of the ranges of L. bilineata , L. breuili , and L. carlae , and it is consistent in all aspects with L. carlae of Barbados. After some investigation, it was revealed that the collector and date of the specimen implied that it was collected on the Barbados-Antigua Expedition of the University of Iowa ( Nutting, 1919). Also, the major city of Antigua is St. John’s, whereas the Parish in Barbados where L. carlae is found is St. John, again consistent with the locality of the specimen. Reptiles were not a major group studied on the expedition, and the collector (Fisher) studied echinoderms. The few other reptiles collected on Antigua came from only two sites, English Harbor and Monk’s Hill (not St. John’s). The account of the expedition mentions field work in St. John (Parish), Barbados ( Nutting, 1919). Considering all of this information, I conclude that this specimen is from St. John Parish, Barbados, not “St. John, Antigua.”

The two new specimens were collected adjacent to a small patch of secondary forest near Bonwell in the Parish of St. Joseph. This locality and the only other known locality, Codrington College ( Fig. 2A ), lie within a small, geologically unique area ( Speed, 1994) in east-central Barbados lacking a Pleistocene reef cap. This upland area comprises ~ 40 km 2 or about one-tenth of the total land area of Barbados and corresponds to the parishes of St. Andrew, St. Joseph, and a small portion of St. John close to the coast. Numerous other localities in these parishes were searched intensely for several days in June 2006, including coastal and upland habitats, without success. The imprecise locality for CAS 49279, “St. John,” is consistent with the species being restricted to the same region of Barbados as the other specimens. This was the first part of Barbados to emerge above sea level about one million years ago ( Speed, 1994) and establishes the maximum age for the colonization of Barbados by Leptotyphlops , although molecular clock time estimates in anoline lizards inhabiting Barbados have conflicted with this geologic date ( Thorpe et al., 2005).

There is little known about the ecology of these new Antillean snakes. Threadsnakes are burrowers and usually feed on the adults and larvae of ants and termites ( Greene, 1997). It is almost certain that this native species, which evolved in the presence of forests, requires such forest habitat for survival. However, Barbados is one of the ten most densely populated countries in the World ( World Resources Institute, 2006) and has essentially no original forest remaining ( FAO, 2005). Assuming that secondary forest is sufficient to provide habitat for the survival of this species, and if such habitat comprises 5–10% of the east-central sub-reef region, then the suitable habitat for L. carlae is probably is no more than a few square kilometers. An additional concern is the recent introduction to Barbados of the Flowerpot Blindsnake, Ramphotyphlops braminus Daudin , a parthenogenetic species native to the Australasian region. On our visit in 2006 we examined specimens of this species collected in Bridgetown (St. Michael) and near Lowland (Christ Church) indicating that it is probably now widespread in Barbados, at least in urban areas. As a potential competitor, it could pose an additional threat to the survival of L. carlae .

Figures

FIGURE 1. Two species of threadsnakes (Leptotyphlops) from the Lesser Antilles. (A) Leptotyphlops carlae (101 mm TL), adult female paratype (USNM 564818). (B) Leptotyphlops breuili (108 mm TL), adult male paratype (USNM 564811).

FIGURE 2. Distribution and pattern variation in threadsnakes (Leptotyphlops) from three islands in the Lesser Antilles. (A) Distribution and generalized color pattern differences (head, midbody, and tail). The location of samples used in sequence analyses is indicated. (B) Head pattern in three individuals from each island, illustrating variation (MNHN 2006.0516, USNM 564808, 564809; USNM 564810, 564815, 564816; BM1969.792, USNM 564819, 564818). Dashed line in Barbados indicates geologic region of sub-reef outcrop.

FIGURE 3. Variation in scalation of threadsnakes (Leptotyphlops) from three islands in the Lesser Antilles. (A) Head scalation (top and side of head) in representatives from each island (USNM 564809, 564810, 564819). Scale bar = 1 mm. Numbered head scales are rostral, RO (1); prefrontal, PF (2); frontal, FR (3); interparietal, IP (4); interoccipital, IO (5), postnasal, PN (6); preocular, PO (7); parietal-I, PA1 (8); parietal-II, PA2 (9); and ocular, OC (10). (B) Plot of parietal-1 scale width versus length. (C) Plot of parietal-2 scale width versus length.

FIGURE 4. Variation in the number of middorsal scales in Leptotyphlops from the Lesser Antilles. The two groups identified for L. bilineatus are suspected to be different species.