Murina harrisoni Csorba and Bates 2005

Francis, Charles M. & Eger, Judith L., 2012, A review of tube-nosed bats (Murina) from Laos with a description of two new species, Acta Chiropterologica 14 (1), pp. 15-38 : 29-32

Murina harrisoni Csorba and Bates 2005

( Figs. 3a , 4a , 5a , 9 , 10 , 11 ; Tables 1 , 2; Map Fig. 2a )

Murina tiensa Csorba et al. 2007: 3 ; type locality Kim Hy Nature Reserve, Bac Kan Province, Vietnam.

Specimens examined from Laos

EBD 24974 ♀, 29 March 1998, Nam Khan, Louangphabang, Laos (20°09’N, 103°24’E).

Additional specimens examined

SMF 53218, 17 June 1977, Ban Tham Tap Tao, SSW of Fang, Chiang Mai, NW Thailand (19°59’N, 99°59’E); CM 88162 Huai Kha Khang , Uthai Thani, Thailand (15°29’N, 99°18’E) ; ROM 107739 ♂, 107749 ♂, 107750 ♀ 5–6 June 1997, Yok Don National Park , Dak Lak, Vietnam (12°52’N, 107°42’E) ; ROM 116463 ♀, 116468 ♂ 1–2 May 2005, Shiwandashan National Reserve , Guangxi, China (21°50’N, 107°53’E) ; MSNG (field number 309), Biapo , Karen State, ≈ 70 km NE Toungoo, Myanmar (estimated location 19°20’N, 96°40’E) .

Description

Externally, this is a moderately large species with orange-brown fur that is paler at the base, unlike M. huttoni or M. cyclotis which have darker bases to the fur. The underside is somewhat paler. The interfemoral membrane is extensively covered with orange-brown hairs. At the foot, the membrane inserts on the side of the toe, about half way between the base of the toe and the base of the claw.

In cranial profile, the skull of EBD 24974 has a moderate rostral depression with a thick rostrum, closely resembling the holotype of M. harrisoni , though some of the other specimens from Vietnam and China have a thinner rostrum ( Fig. 9 ), more similar to specimens that were referred to M. tiensa by Csorba et al. (2007). In EBD 24974, the upper incisors are in line with each other; the outer incisor is about half the height of the inner one ( Fig. 4a ). In some of the other specimens the inner incisor is similar in height to the outer, but this appears to be due to wear as the tip of the inner incisor is clearly very worn. The upper premolars are similar in height to each other, about half the height of the canine; the anterior upper molars (M 1 and M 2) have moderately developed mesostyles which are visible in profile, at least on unworn specimens ( Fig. 4a ), although relatively smaller than in M. huttoni , and the labial side of the teeth is distinctly indented in the middle ( Fig. 4a left). The talonid of the anterior lower molars (M 1 and M 2; Fig. 5a ) is relatively larger than in M. cyclotis ( Fig. 5d ), but not as large as in M. huttoni ( Fig. 5c ).

Discussion

These specimens represent the first published records for this species from Laos, Thailand, Myanmar and China. Our examination of these specimens suggests that M. tiensa is a junior synonym of M. harrisoni .

Genetic analyses suggest that all of the specimens we examined are the same species, despite morphological similarities of some to the holotype of M. harrisoni and others to the holotype of M. tiensa . DNA barcode sequences for the Lao specimen are nearly identical to those from the two specimens from China, while the three specimens from Vietnam are the same as each other and differ from the China / Lao specimens by about 4%. These were labelled as Murina JLE sp. F and Murina JLE sp. E respectively in figure 4 of Francis et al. (2010).

We also obtained cytochrome b sequences for these same specimens as well as the holotypes of both M. tiensa and M. harrisoni (J. L. Eger, unpublished data) and generated a neighbour-joining tree incorporating two additional specimens reported on Genbank as M. tiensa from Vietnam ( Fig. 10 ). These results indicate that the Lao specimen clusters with the holotype of M. tiensa as well as the Chinese specimens. Although the Vietnamese ‘sp. E’ differ by ≈ 3%, they morphologically resemble specimens referred to M. tiensa ( Fig. 9 ).

The sequence of the holotype of M. harrisoni clearly clusters with the other specimens, but appears to diverge by about 5–6%. However, this divergence must be interpreted cautiously due to degradation of the DNA. DNA for this specimen was extracted from a small tissue sample taken from the specimen after it had been stored in 70% ethanol for several years. It was only possible to retrieve short sequences which were subsequently assembled. Separate runs did not always agree in the sequences. We analysed separately all available base pairs (1008 base pairs out of the potential total 1140 bp, excluding two areas from 725–751 and 954–1058 which were particularly unreliable), as well as a shorter subset truncated at 724 base pairs. The shorter subset showed less divergence from other specimens in the tree, suggesting some of the divergence was due to sequencing errors. Any sequencing errors would make the sequence appear to diverge from the other specimens, such that the true sequence may be more similar to the other specimens than indicated. In any case, even a difference of ≈ 5% is within the range of observed intraspecific variation in other Murina for mitochondrial genes and is smaller than the differences among any pairs of other currently recognized species of Murina ( Francis et al., 2010) .

Csorba et al. (2007) listed only two morphological differences between M. tiensa and M. harrisoni . In the skull profile, they suggested that the anterior part of the rostrum is almost straight in M. tiensa , while it is more bulbous in M. harrisoni with a less thickened rostrum (compare the top and bottom skulls in Fig. 9 ). They also suggested that the insertion point of the wing membrane is at the base of the toe in harrisoni and at the base of the claw (tip of the toe) in M. tiensa . Our specimens suggest that neither of these characters is reliable for differentiating species. We found variation in the rostrum shape among the specimens that we have examined ( Fig. 9 ), which does not match up with any genetic differences ( Fig. 10 ). The wing membrane of all the individuals we examined, irrespective of skull shape, is inserted on the side of the toe, roughly half way between the base (where the toes separate) and the base of the claw ( Fig. 11 ). We note that the exact position can sometimes be difficult to discern, depending upon how the specimen was preserved and how it is stretched, which can lead to confusion.

We thus conclude that the names M. harrisoni and M. tiensa both refer to the same species. The appropriate, prior name for the species is M. harrisoni as it was described first.

Specimen 88162 from the Carnegie Museum was originally reported by McBee et al. (1986) as the first record of Murina leucogaster from Thailand along with a 2nd specimen which we did not examine but which was presumably the same species. Given that there are no other records for M. leucogaster from Thailand ( Bumrungsri et al., 2006), that species should be removed from the list of bats known from Thailand. The specimen SMF 53218 from Thailand was originally reported by Yenbutra and Felten (1986) as Murina huttoni , but agrees well with the other specimens we have examined of M. harrisoni .

The specimen from Myanmar in the MSNG (field number 309) was reported by Thomas (1893: 926) as M. leucogaster but our examination also indicates it should be referred to M. harrisoni . Although the skull has not been extracted, the mouth was preserved in an open position, and the distinctive dental characters are clearly visible. The pale bases to the fur and the insertion point of the membrane also help to differentiate the specimen from M. huttoni . As such, M. leucogaster should also be removed from the list of species from Myanmar, although we note that most previous authors (e.g., Hill, 1992) had overlooked the record in Thomas (1893).

Distribution and ecology

Although reported from only a few specimens, the known locality records are widely distributed throughout the region ( Fig. 2a ). Specimens from Thailand and Myanmar were collected in hill areas, but precise altitude and habitat information is not available. Vietnamese specimens were collected in dry open dipterocarp forest; those from China in Shiwandashan National Reserve specimens at an altitude of 550 m, in subtropical, montane secondary forest.

Figures

FIG. 3. Skull profiles of specimens of eight species of Murina from Laos: (a) M. harrisoni, EBD 24974; (b) M. fionae sp. nov. ROM 106383; (c) M. huttoni ROM 106419; (d) M. cyclotis, ROM 106431; (e) M. annamitica sp. nov. ROM 106467; (f) M. eleryi, ROM 106411; (g) M. walstoni ROM 110450; (h) M. feae, ROM 106477. Scale bar = 5 mm

FIG. 4. Upper left toothrows of specimens of eight species of Murina from Laos: (a) M. harrisoni, EBD 24974; (b) M. fionae sp. nov. ROM 106383; (c) M. huttoni ROM 106419; (d) M. cyclotis, ROM 106431; (e) M. annamitica sp. nov., ROM 106467; (f) M. eleryi, ROM 106411; (g) M. walstoni ROM 110450; (h) M. feae, ROM 106477. Scale bar = 5 mm

FIG. 5. Lower left toothrows of specimens of eight species of Murina from Laos: (a) M. harrisoni, EBD 24974; (b) M. fionae sp. nov. ROM 106383; (c) M. huttoni ROM 106419; (d) M. cyclotis, ROM 106431; (e) M. annamitica sp. nov., ROM 106467; (f) M. eleryi, ROM 106411; (g) M. walstoni ROM 110450; (h) M. feae, ROM 106477. Scale bar = 5 mm

FIG. 9. Skull profiles for five specimens of M. harrisoni showing variation in the shape of the skull profile. From top to bottom, these are HZM 1.31525 ♀ from Vietnam, which was referred to M. tiensa by Csorba et al. (2007); ROM 107749 ♂ from southern Vietnam; ROM 116468 ♂ from southern China; EBD 24974 ♀ from Laos; and HZM 1.36316 ♀, the holotype of M. harrisoni from Cambodia. Scale bar = 5 mm

FIG. 10. Neighbour joining tree based on cytochrome b sequences for several specimens of M. harrisoni, including the holotype of M. tiensa, along with three specimens of M. huttoni as an outgroup. The branch for the sequence of M. harrisoni is shown with a dotted line because its reliability is uncertain due to degraded DNA. The sequence HZM 1.36136a includes all available base pairs (1008bp), while the sequence HZM 1.36136b is a subset representing the first 724 base pairs of the sequence. The sequences GQ168922 and HNHM 2007.28.1 were taken from Genbank, where they are identified as M. tiensa. The remainder were sequenced at the Royal Ontario Museum (J. L. Eger, unpublished data)

FIG. 11. Ventral view of left foot of M. harrisoni (EBD 24974) from Laos, showing insertion point of wing membrane on the side of the toe, slightly less than half way between the base of the toe and the base of the claw. Note that the two first toes are joined near the base by a membrane, but the actual base of the toe is closer to the body

FIG. 2. Distribution of known specimen records of Murina from Laos and surrounding countries for (a) M. harrisoni, (b) M. walstoni, (c) M. fionae sp. nov., and (d) M. annamitica sp. nov. Solid circles (●) represent locations for specimens examined by the authors from Laos and surrounding countries (see Appendix and text for specimen details). Solid stars represent type localities for each species. Hollow circles represent records from the published literature. On map (a), the open star marks the type locality for M. tiensa, here considered a synonym of M. harrisoni, while the open circle marks the location of additional material referred to M. tiensa by Csorba et al. (2007)