210 



G. CSORBA AND P. JENKINS 



I ' ' ' ■ I 



cc 



o 



LL 



R. s. stheno 



I ' ' ' ' 



R. malayanus 



R. s. microglobosus # 



■ i ■ ■ ■ ■ ' 



-1.5 -1 -0.5 0.5 



FACTOR1 



1 1.5 



Fig. 3 Principal components analysis of R. s. microglobosus. R. s. stheno and R. malayanus specimens based on 10 external and craniodental characters. 



swellings component' in which the two measurements of the ros- 

 trum (rostral swelling width and median anterior rostral swelling 

 width) were the most important, and Fl pertained to the other 

 characters (Table 2). 



Table 2 Factor loadings of external and craniodental characters obtained 

 by Principal Component Analysis 



Character 





Factor 1 



Factor 2 



Forearm length 





-0.873 



-0.344 



Greatest skull length 





-0.985 



-0.040 



Maxillary toothrow length 





-0.954 



0.075 



Zygomatic width 





-0.887 



0.283 



Mastoid width 





-0.972 



-0.003 



Mandible length 





-0.943 



0.126 



Lower toothrow length 





-0.955 



0.096 



Interorbital width 





0.740 



0.569 



Rostral swelling width 





-0.050 



0.929 



Median anterior rostral swelling 



width 



-0.206 



0.894 



Variance explained 





67.67% 



22.17% 



DISCUSSION 



The new records of R. stheno extend the known distribution of the 

 species to North Vietnam, and represent a new subspecies which is 

 characterised by its generally smaller, narrower skull and above all, 

 by the small, globular anterior median rostral swellings. 



Specimens of R. stheno from Thailand in the collection of The 

 Natural History Museum also proved to belong to the new subspe- 

 cies. It seems possible also, that specimens recorded by Osgood 

 (1932: 219) refer to the same subspecies as described here. His 

 specimens, listed as 'Rhinolophus sp.\ derived from Tonkin (North 

 Vietnam) and Osgood stated that '. . . it is possible that the present 

 [form] is a northern representative of the larger Malayan form 

 stheno\ An alternative suggestion, that Osgood's specimens might 

 be referable to R. borneensis was, however, made by Hill & 

 Thonglongya (1972). This supposition is equally probable, as af- 



firmed by specimens of R. borneensis which were also collected 

 during the recent expeditions to Vietnam. It appears likely that the 

 section on R. stheno in Lekagul & McNeely ( 1 977) also refers to the 

 new subspecies; unfortunately, however the accompanying photo- 

 graph is of a specimen in which the diagnostic characters are not 

 visible on the damaged rostrum. 



According to the literature, R. stheno and the closely related R. 

 malayanus may be distinguished by the shape of the rostral swell- 

 ings. On the basis of our data set, the width of the interorbital 

 constriction also distinguishes the two species (Table 1). 



As regards the external characters, according to Koopman (1994) 

 there is a definite gap between the two species in forearm length (45- 

 48 mm against 4CM3 mm) but McFarlane & Blood ( 1 986) concluded 

 'that there is a probability of overlap between specimens of the two 

 species'. Indeed, during the examination of larger series derived 

 from different geographical regions only very slight differences may 

 be observed between extreme values of forearm length of small R. 

 stheno and large R. malayanus. Furthermore the ratio of first to 

 second phalange of the third digit in R. s. microglobossus shows 

 overlap in size between the smaller/?, malayanus and the larger/?, s. 

 stheno. as figured by McFarlane & Blood (1986). 



Acknowledgements. We are very grateful the Society for Environ- 

 mental Exploration - Vietnam, especially the co-ordinators Mike Hill and 

 Neville Kemp assisted by Dung Ngoc Can of the Institute of Ecology and 

 Biological Resources, Hanoi, but also the other members involved in the bat 

 surveys, including Maysie Harrison, Daniel Fitton, Vicky Jenkins and Joe 

 Walston, and to Pham Due Tien for his generous help in the field. We are 

 indebted to Prof. Cao Van Sung, Institute of Ecology and Biological Re- 

 sources, Hanoi for organising and co-ordinating the research programs in 

 Vietnam, to Dr. Gyorgy Topal for his helpful advice during the preparation of 

 the manuscript, to Dr. Jacques Cuisin, Museum National d'Histoire Naturelle, 

 Paris for access to the collection and to Peter Ujhelyi for the drawings. In 

 particular we thank Dr. Wieslaw Bogdanowicz, Polish Academy of Sciences 

 for his very helpful review comments and Clive Moncrieff, NHM for his 

 constructive criticism of the statistical methodology. The work of GCS was 

 supported by the Royal Society of London and the Hungarian National 

 Scientific Fund (OTKA) grant no. F 17700. 



