OUP user menu

Phylogenetics of Small Horseshoe Bats from East Asia Based on Mitochondrial DNA Sequence Variation

Gang Li, Gareth Jones, Stephen J. Rossiter, Shiang-Fan Chen, Stuart Parsons, Shuyi Zhang
DOI: http://dx.doi.org/10.1644/05-MAMM-A-395R2.1 1234-1240 First published online: 29 December 2006


We undertook analyses of mitochondrial DNA gene sequences and echolocation calls to resolve phylogenetic relationships among the related bat taxa Rhinolophus pusillus (sampled across China), R. monoceros (Taiwan), R. cornutus (main islands of Japan), and R. c. pumilus (Okinawa, Japan). Phylogenetic trees and genetic divergence analyses were constructed by combining new complete mitochondrial cytochrome-b gene sequences and partial mitochondrial control region sequences with published sequences. Our work showed that these 4 taxa formed monophyletic groups in the phylogenetic tree. However, low levels of sequence divergence among the taxa, together with similarities in body size and overlapping echolocation call frequencies, point to a lack of taxonomic distinctiveness. We therefore suggest that these taxa are better considered as geographical subspecies rather than distinct species, although this should not diminish the conservation importance of these island populations, which are important evolutionarily significant units. Based on our findings, we suggest that the similarities in body size and echolocation call frequency in these rhinolophids result from their recent common ancestry, whereas similarities in body size and call frequency with R. hipposideros of Europe are the result of convergent evolution.

Key words
  • Chiroptera
  • mitochondrial DNA
  • phylogenetics
  • Rhinolophidae
  • Rhinolophus

Phylogenetic relationships among several species of horseshoe bat of the pusillus subgroup (Rhinolophus pusillus, R. monoceros, R. cornutus, and R. c. pumilus) have been discussed by numerous researchers for many years (Corbet and Hill 1992; Csorba 1997; Csorba et al. 2003; Koopman 1994; Simmons 2005; Yoshiyuki 1989, 1990).

Rhinolophus pusillus (the least horseshoe bat) was described initially from Java by Temminck (1834). Temminck (1834) also 1 st described R. cornutus (the little Japanese horseshoe bat) from Japan, whereras R. c. pumilus (the Okinawa least horseshoe bat) is endemic to the Ryukyu Islands of Japan, situated between Japan and Taiwan and was 1st described by Andersen (1905). R. monoceros (the Formosan lesser horseshoe bat) was described from Taiwan (where it is endemic) by Andersen (1905).

The separate species status of R. pusillus and R. cornutus was acknowledged by Yoshiyuki (1989, 1990), but questioned by Corbet and Hill (1992), who considered that these taxa might be conspecific, with R. pusillus including the types cornutus and pumilus. R. monoceros was speculated to be conspecific with R. cornutus, R. pusillus, or both for more than a decade by different authors (Corbet and Hill 1992; Csorba et al. 2003; Koopman 1994). More recently, Guillén-Servent et al. (2003) used complete mitochondrial cytochrome-b (Cytb) gene sequences to reconstruct the phylogenetic tree of the family Rhinolophidae, but did not include R. cornutus, R. monoceros, and R. c. pumilus, and, therefore, did not solve the systematic puzzle of these small horseshoe bats of East Asia. Csorba et al. (2003) reviewed the genus Rhinolophus and considered R. monoceros, R. pusillus, and R. cornutus as independent species, with R. c. pumilus a subspecies of the latter. In the most recent assessment, Simmons (2005) agrees with this treatment, but acknowledges that all 4 species might be conspecific.

Mitochondrial DNA (mtDNA) data can provide valuable information on the evolutionary histories of closely related taxa and have been widely used to evaluate phylogenetic relationships of bats (e.g., Baker et al. 1994; Hoofer and Van Den Bussche 2001; Kawai et al. 2002). In this study, we combined new complete Cytb gene sequences and partial sequences of the mtDNA control region of R. pusillus, R. monoceros, and R. cornutus with those of other species of Rhinolophus and Hipposideros to analyze the phylogenetic relationships among these species. Additionally, we included a published sequence of R. c. pumilus, obtained from GenBank. We also obtained recordings of echolocation calls from R. pusillus and R. monoceros, and compared these with data obtained by other authors on R. cornutus. By analyzing divergence in mtDNA gene sequences and echolocation call frequencies, we aimed to resolve the taxonomic positions of these taxa.

Materials and Methods

Taxonomic sampling.—Rhinolophus pusillus was sampled over a large area of its range in China, which extends from Beijing to Hainan across more than 21 degrees of latitude. Samples of R. cornutus and R. monoceros were obtained or collected from Japan and Taiwan, respectively (see Appendix I for locality details). Samples were either 3-mm wing membrane biopsies from live animals (which were released after being punched) or liver biopsies taken from voucher specimens. For all bats, forearm lengths were measured with dial calipers to the nearest 0.1 mm for the morphological identification. All animals were handled in accordance with guidelines for animal care and use established by the American Society of Mammalogists (Animal Care and Use Committee 1998), and all voucher specimens of bats captured in China were deposited at the Institute of Zoology, Beijing.

DNA extraction and amplification.—All tissue was preserved in 75% ethanol until genomic DNA was extracted using a standard phenol-chloroform protocol (Maniatis et al. 1982) and stored at 4°C. Complete mitochondrial Cytb sequences and partial control region sequences were obtained from 8 individuals of R. pusillus, 4 of R. cornutus, and 5 of R. monoceros, as well as 1 individual from each of the following 5 additional Rhinolophus species: R. hipposideros, R. affinis, R. pearsonii, R. ferrumequinum, and R. luctus. We also amplified the corresponding sequences in 2 hipposiderid species (Hipposideros armiger and H. pratti), which were used as outgroups. Published mitochondrial sequences of R. monoceros and R. c. pumilus also were obtained from the GenBank database. Sampling localities and Genbank accession numbers are given in Appendix I.

Complete Cytb gene sequences were amplified via polymerase chain reaction from each individual DNA sample. The amplification process was conducted as follows: 94°C (5 min), 35 cycles at 94°C (50 s), 50°C (40 s), 72°C (50 s), 72°C (5 min). Polymerase chain reaction mixtures were prepared in 30-µl volumes with a final concentration of 0.4 µM of each primer, 0.2 µM of each deoxynucleoside triphosphate, 1,5 µM MgCl2, and 1 U of Taq DNA polymerase. Primers L14724 (5′-GGT CTT AGG CAA AAA ATT GGT GCA ACT C-3′—Kocher et al. 1989) and H15915R (5′-TCA GCT TTG GGT GTT GAT GG-3′—Irwin et al. 1991) were used for amplification. To improve amplification performance for some species, we also designed the primer Bat_Cytb_l (5′-TAG AAT ATC AGC TTT GGG TG-3′), which also was used with L14724 (Kocher et al. 1989).

For the control region, the primers DLH 16750 (5′-CCT GAA GTA GGA ACC AGA TG-3′—Wilkinson and Chapman 1991) and THRL 16272 (5′-CCC GGT CTT GTA AAC C-3′—Stanley et al. 1996) were used with the following thermal profile: 94°C (2 min), 34 cycles at 94°C (30 s), 55°C (30 s), 72°C (30 s), 72°C (10 min). We amplified and sequenced approximately 500 control region base pairs.

Genetic analyses and phylogenetic reconstructions.—Sequence data were aligned using CLUSTALX 1.81 (Thompson et al. 1997) with the default parameters, and sequence variation and divergence were calculated using MEG A3 (Kumar et al. 2004) under the Kimura 2-parameter model.

Before phylogenetic analysis, the most appropriate substitution model was determined for sequences using the program MODELTEST 3.06 (Posada and Crandall 1998). We combined the Cytb and control region sequences to reconstruct a maximum-likelihood phylogenetic tree using PAUP* 4.0b (Swofford 2002). These analyses were performed using heuristic searches with tree bisection and reconnection branch swapping and incorporating all codon positions and substitutions. To assess the robustness of the tree topologies, 2,000 bootstrap replicates were carried out for the neighbor-joining and maximum-parsimony analyses as well as 100 replicates for the maximum-likelihood analysis. In addition, a maximum posterior probability tree was constructed using the program MrBayes 3.1 (Huelsenbeck and Ronquist 2001). Five million generations were used for 6 simultaneous Markov chains, and the trees were sampled after one million generations, when the chains approached equilibrium. In the control block of the input file, we set the code parameter to vertebrate mitochondrial sequences.

Echolocation calls analyses.—Echolocation calls were recorded using a Pettersson D980 bat detector (Pettersson Elektronik AB, Uppsala, Sweden), and 10× time-expanded calls were downloaded to either a Sony WMD6C cassette recorder (Sony, Tokyo, Japan) or a Sony TCD-D8 DAT recorder. No measurable differences in call frequency occurred depending on which recording method was used. Echolocation calls were digitized using the sound analysis software BatSound Pro, v3.0 (Pettersson Elektronik AB). The maximum intensity of the constant-frequency component of the 2nd harmonic in the power spectrum was used for measuring the resting frequency (in kHz) of a call. A 4,096-point fast Fourier transformation and a Hanning window were used within a 5-kHz frequency range, giving a frequency resolution of 64 Hz. Call frequencies were stable within individual bats, so 1 call was chosen at random from each bat for analysis. R. pusillus was recorded from Yunnan, Guangdong, and Huibei provinces in China, and R. monoceros was recorded throughout its range in Taiwan. All bats emitted calls typical of horseshoe bats, with a long constant-frequency portion initiated and terminated by brief frequency-modulated sweeps.


Genetic analyses.—The complete mitochondrial Cytb gene and partial control region sequences of the bats were deposited in GenBank (Appendix I). The mitochondrial Cytb sequences contained 1,140 base pairs, beginning with the codon ATG and ending with AGA. No gaps were found in the Cytb sequences among the species of Rhinolophus and Hipposideros studied. The number of conserved sites of Cytb gene sequences was 1,045 (92%). Correspondingly, 8% variable sites and 4.4% parsimony-informative sites were found. To carry out the phylogenetic analyses, 494 base pairs of control region sequences were combined, which included 411 conserved sites (83%) and 17% variable sites. Sixty-two (12.55%) of the sites yielded information for the parsimony analysis. As expected, the control region was more variable than the coding gene.

Pairwise genetic distances based on complete Cytb and partial control region sequences show the high homogeneity of these mitochondrial sequences (Appendix II). Sequence divergence between R. cornutus and R. c. pumilus was found to be only 1.6% at Cytb and 5.2% at control region sequences. Within R. pusillus samples from different areas of China, sequence divergence ranged from 0.2% to 1.7% at Cytb sequences (X̄ = 1.2%) and from 0.4% to 5.8% (X̄ = 4.2%) based on the control region. Sequence divergence between R. monoceros and R. cornutus ranged from 2.1% to 2.9% (X̄ = 2.38%) at Cytb and from 5.3% to 7.1% (X̄ = 6.1%) at the control region. Between R. monoceros and R. pusillus, genetic distance based on Cytb sequences ranged from 1.2% to 2.2% (X̄ = 1.66%) and based on control region sequences ranged from 3.1% to 7.8% (X̄ = 5.7%). The average Cytb and control region sequence divergences were 3.6% and 7.7%, respectively, between R. c. pumilus and R. pusillus.

Our results show that the genetic distances among R. pusillus, R. monoceros, R. cornutus, and R. c. pumilus are very low, ranging from 1.6% to 4% (X̄ = 2.31%) for the complete mitochondrial Cytb gene and 5.7% to 7.8% (X̄ = 6.2%) for the partial control region. Moreover, sequence divergence within R. pusillus from China is often as high as that between currently recognized species. On the other hand, much higher sequence divergence values were recorded between these and other rhinolophid species. Approximate divergence values of these taxa versus the other Rhinolophus species studied were 11% for Cytb and 14% for the control region.

Phylogenetic reconstructions.—For the combined sequences, the TVM+G+I model was selected and the relative base frequencies were: A = 0.32, T = 0.24, G = 0,12, and C = 0.32. Rate matrices were: A-C = 0.57, A-G = 5.01, A-T = 0.79, C-G = 0.07, C-T = 5.02, and G-T = 1.0. The proportion of invariable sites equaled 0.54 and the gamma distribution shape parameter equaled 1.44. We used these selected models and parameters for maximum-likelihood analyses and Markov chain simulation.

Based on the combined sequences, both maximum-likelihood (Fig. 1A) and maximum posterior probability (Fig. 1B) trees showed that R. pusillus, R. monoceros, R. cornutus, and R. c. pumilus together formed a monophyletic group that was supported strongly by high maximum-likelihood bootstrap and posterior probability values, and that all 4 taxa formed separate monophyletic groups. The sister relationship between R. monoceros and R. pusillus from mainland China also was strongly supported by each tree with high bootstrap values (82 for maximum likelihood, 98 for neighbor joining [distance measure used GTR model with gamma distribution], 86 for maximum parsimony, and a posterior probability value of 1.0). These 2 species are more closely related to each other than to the taxa of the Japanese archipelago (R. cornutus) and Ryukyu archipelago (R. c. pumilus). Both maximum-likelihood and Bayesian analyses indicate with high bootstrap and statistical support an earlier divergence between R. c. pumilus and the other small horseshoe bats studied. Conversely, the neighbor-joining and maximum-parsimony bootstrap values support a sister relationship between R. cornutus and R. c. pumilus, supporting the current taxonomic status of these taxa. Each tree rejected a close relationship between these species and the similarly sized R. hipposideros from Europe, indicating that similar echolocation frequency and morphological features between these bats are convergent characters.

Fig. 1

A) Maximum-likelihood tree (InL = −7,992.58, with branch lengths) was constructed based on the combined mitochondrial complete cytochrome-b gene and partial control region of species of Rhinolophus under TVM+G+I model using program PAUP*4.0b (Swofford 2002). B) Maximum posterior probability tree (without branch lengths). Hipposideros species were designated as outgroups. The evolution model used in the Bayesian analysis also was selected by MODELTEST 3.06. Numbers near the node indicate posterior probabilities and bootstrap values of major nodes. The 1st number is the posterior probability value, and the 2nd, 3rd, and 4th numbers are bootstrap values following maximum-likelihood (100 replicates), neighbor-joining (2,000 replicates), and maximum-parsimony (2,000 replicates) methods, respectively. An asterisk (*) indicates that the method did not support that particular clade.

Echolocation calls analyses.—Echolocation calls from R. pusillus in China contained a frequency with most energy at 106.71 kHz ± 2.47 SD (n = 24 bats, range 102–111 kHz). Averages obtained for males and females were 105.95 ± 2.72 kHz (n = 8) and 108.48 ± 2.57 kHz (n = 5), respectively. The average forearm length of males was 36.9 ±1.4 mm (n = 17, range 35.1–40.2 mm), and of females was 37.2 ±1.5 mm (n = 8, range 34.9–39.7 mm). R. monoceros from Taiwan emitted calls with a frequency of most energy at 112.72 ± 2.55 kHz (n = 30, range 107–118 kHz). Males called with most energy at 111.71 ± 2.22 kHz (n = 15) and had an average forearm length of 37.3 ± 1.2 mm (n = 15, range 35.6–39.2 mm). Female calls contained most energy at 113.73 ± 2.51 kHz (n = 15), and their average forearm length was 37.9 ±1.1 mm (n = 15, range 35.0–39.9 mm). R. cornutus from Yakushima in Japan had a mean frequency of most energy at 108 kHz (D. A. Hill, pers. comm.). Wallin (1969) gives the forearm length of R. cornutus as 38.0–40.7 mm and that of R. C. pumilus as 37.0–40.0 mm. All taxa therefore show considerable overlap in body size and, as far as is known, in echolocation call frequency.


We examined the phylogenetic relationships of horseshoe bats within the pusillus subgroup across much of their range in East Asia. R. pusillus was sampled from across China. The identification of 2 individuals captured in a cave near Beijing, previously considered northeast of the known range of this species, was confirmed by genetic analysis. The individual from Hainan Island (currently recognized as subspecies R. pusillus parcus) was not found to be divergent from the mainland populations of this species, and was nested within bats captured on other Chinese localities, despite Hainan's location 29.5 kilometers off the coast of southeastern China (19°3′N, 109°49′E).

The Ryukyu population of R. c. pumilus was allocated to a subspecific level by Andersen (1905); however, based on wing color and skull and dental characters, Yoshiyuki (1989) ranked this taxon as a distinct species. Other specimens from southern China (Guangdong, Guangxi, and Fujian provinces) were identified as R. c. pumilus, but were referred to R. c. pusillus later (Corbet and Hill 1992). Hill and Yoshiyuki (1980) analyzed the morphology of R. c. pusillus and R. cornutus and suggested that differences occurred in the structure of the connecting process between R. c. pusillus and R. cornutus, but speculated that there was a high likelihood that these bats were conspecific. Corbet and Hill (1992) and Koopman (1994) later regarded R. cornutus and R. c. pusillus as 2 distinct species based on their allopatric distributions.

Rhinolophus monoceros is currently recognized as an insular endemic species and its distribution is restricted to Taiwan. Andersen (1905) described the type specimen of R. monoceros and suggested that it was a new species that could be differentiated from R. cornutus by the shape of the lancet in the nose leaf. However, other researchers suggested that R. monoceros might be conspecific with R. cornutus, R. pusillus, or both (Corbet and Hill 1992; Csorba 1997; Koopman 1994).

Bradley and Baker (2001) surveyed Cytb divergence levels across 4 genera of rodents and 7 genera of bats and showed that genetic distances can be used to broadly evaluate the systematic status of taxa. Specifically, they suggested that genetic distance values <2% usually indicate conspecific populations; 2–11% encompass conspecific populations and separate species, and so require further study; and >11% usually correspond to distinct species. Levels of divergence in bats also were generally higher than equivalent taxonomic levels in rodents (Bradley and Baker 2001). Ditchfield (2000) analyzed mitochondrial Cytb sequences from 275 individual bats of 17 species and showed low levels of sequence divergence at Cytb with <4% divergence (usually 1–2.5%) within bat species, whereas Ruedi and Mayer (2001) analyzed complete Cytb sequences of the genus Myotis, and showed >10% sequence divergence (average 15%) among congeners. In this current study, sequence divergence based on complete mitochondrial Cytb was consistently low among R. pusillus, R. monoceros, R. cornutus, and R. c. pumilus, not exceeding 4%, and averaging just 2.3%. This high level of sequence homogeneity indicates that these bats have diverged relatively recently in their evolutionary history. In our data set, the pairwise genetic distances of Cytb gene between other Rhinolophus species (excluding the divergence within the pusillus group of East Asia) ranged from 11% to 15% and the divergence values of Rhinolophus and outgroup (Hipposideros) averaged approximately 20%, with a minimum of 17%.

It would be unwise to determine species status solely on the basis of sequence divergence at one gene locus. Indeed, some bat species that look very different in terms of their morphology differ only slightly in mtDNA sequences (e.g., Eptesicus serotinus and E. nilssoniiMayer and von Helversen 2001). Neither genetic nor morphological data alone are sufficient to distinguish species. The body sizes (as measured by forearm lengths) and echolocation call frequencies of R. pusillus, R. monoceros, and R. cornutus all overlap considerably, supporting their classification as 1 species. Echolocation call frequencies can be informative in resolving taxonomic differences among bat species. For example, many cryptic species of bats are difficult to distinguish by morphological criteria, but differ considerably in the frequency of their echolocation calls (Jones and Barlow 2004). Although R. monoceros calls at a slightly higher frequency and is slightly larger than R. pusillus, the differences are small, and overlap is extensive. Call frequency of the similar-sized R. hipposideros in Britain varies between 109 and 117 kHz and is partly explained by variation in sex and age (Jones et al. 1992), so intraspecific variation in call frequency can be substantial.

The major obstacle in determining whether allopatric populations should be given species rank is that reproductive isolation cannot be proven, because the populations are spatially separated. Whether reproductive isolation occurs among these taxa could only be determined if populations established secondary contact in the future. Helbig et al. (2002) provide useful criteria for deciding whether allopatric taxa of birds should be given species status. They proposed that species status should be assigned if taxa are fully diagnosable in each of several discrete or continuously varying characters related to different functional contexts (e.g., structural features and vocalizations) or DNA sequences, and the sum of the character differences corresponds to or exceeds the level of divergence seen in related species that coexist in sympatry. These criteria do not appear to be met in our study species. Allopatric taxa can be deemed worthy of consideration as allospecies if at least 1 character is fully diagnostic or if taxa are fully diagnosable by a combination of 2 or 3 characters (Helbig et al. 2002). Allospecies also are usually unambiguously phenotypically divergent, which does not seem to be the case here. Our results show that the current classification of these taxa is inconsistent: R. c. pumilus is considered as a subspecies of R. cornutus, whereas R. monoceros is considered as a species distinct from R. pusillus even though these pairs of taxa show the same average level of genetic divergence. The most-parsimonious and consistent explanation is that R. monoceros, R. cornutus, and R. c. pumilus are all island subspecies of the more widely distributed R. pusillus. However, this view should not devalue the importance of the island populations from a conservation perspective. As seen by the small frequency differences between calls of R. pusillus and R. monoceros, the island populations of these bats appear to be diverging from the mainland stock of R. pusillus, so their importance as evolving populations in the process of speciation cannot be overlooked. Indeed, the monophyly of all 4 taxa, together with the geographical distances between these islands and China, together indicate that genetic similarity stems from recent common ancestry rather than recurrent gene flow, and thus each population represents a separate evolutionarily significant unit (Moritz 1994).

A recent phylogenetic analysis of the Rhinolophidae placed R. pusillus, R. cornutus, and R. monoceros in the pusillus subgroup, and R. hipposideros in the hipposideros subgroup (Guillén-Servent et al. 2003). This study also reported deep branching between these 2 subgroups, suggesting that divergence occurred approximately 15 million years ago. Thus, the similarity of echolocation calls and body size (forearm length of 34.4-39.3 mm, after Jones et al. [1992]) between R. hipposideros and the bats in the R. pusillus subgroup studied here is best explained by convergent evolution.

Estimates of the accumulated mutation rate for vertebrate mtDNA range from 2% to 5% per million years (Arbogast and Slowinski 1998; Brown et al. 1979; Shields and Wilson 1987), with a mutation rate of 2% per million years for Cytb (Arbogast and Slowinski 1998). Based on this approximation, divergence among R. pusillus, R. monoceros, R. cornutus, and R. c. pumilus probably occurred approximatley 1 million years ago. The current wide geographic distribution of R. pusillus includes India, Nepal, Myanmar, south China, Vietnam, Laos, Thailand, Malaysia, and Indonesia, whereas R. cornutus is endemic to the Japanese archipelago, where it is widely distributed (Csorba et al. 2003). The Japanese archipelago separated from the Eurasian continent in the Neocene, but the Japanese island of Hondo became connected to the Asian continent via a land bridge during 3 periods between 300,000 and 1 million years ago. Land bridges also formed between the Ryukyuan islands and the Japanese archipelago in the late Pleistocene (Ujiie 1998), providing opportunities for faunal exchange. We therefore suggest that the common ancestor of R. pusillus, R. cornutus, and R. c. pumilus reached Japan via a land bridge in the middle of the Pleistocene Epoch and subsequently began to diverge as the Japanese archipelago became isolated from the rest of the Asian mainland and Ryukyu Islands. Similarly, although the Taiwan Strait formed approximately 4 million years ago (Hsu 1990), a land bridge has connected Taiwan to mainland Asia approximately 5 times in the past 4 million years (Creer et al. 2001; Fairbanks 1989; Gascoyne et al. 1979; Yu 1995), allowing colonization of R. pusillus into Taiwan. An apparent lack of exchange during the most recent land bridge (see Chen et al. 2006) has allowed these populations to diverge in allopatry, forming monophyletic groups.

The weight of evidence based on sequence divergence values, morphological measurements, echolocation call data, phylogenetic analyses, and information about geological time of separation all indicate that R. pusillus, R. monoceros, R. cornutus, and R. c. pumilus are best considered as populations of the same species. However, these populations are almost certainly reproductively isolated by sea barriers, and represent allopatric, evolving populations that merit substantial conservation effort.


We thank K. Kawai and F. Dai for collecting bat tissue samples from Japan. We also are grateful to Renting National Park and Yangmingshan National Park in Taiwan for granting permission to sample R. monoceros. This study was financed by the National Natural Science Foundation of China (grant 30270169) and the National Geographic Society (grant 7806-05) to SZ, a Joint Project Grant between the Royal Society (London) and the Chinese Academy of Sciences, and a Darwin Initiative grant (grant 14-008) to GJ.

Appendix I

View this table:

Specimens examined, with collection localities and the corresponding GenBank accession numbers for species of Rhinolophus (R.) with species of Hipposideros (H.) as outgroups A single asterisk (*) indicates that the sequences obtained from GenBank were from Lin et al. (2002), and double asterisks (**) that they were from Nikaido et al. (2001). The numbers in the 1st column correspond to the columns and rows in Appendix II. Cytb = cytochrome b.

GenBank accession no.
SpecimensCollection localityCytbControl region
1 R. monocerosSheding, Renting, TaiwanDQ297578DQ297602
2 R. monocerosGueishan, Kenting, TaiwanDQ297579DQ297603
3 R. monoceros*TaiwanNC_005433NC_005433
4 R. monocerosYangmingshan, TaiwanDQ297580DQ297604
5R. monocerosYangmingshan, TaiwanDQ297581DQ297605
6 R. monocerosTaiwanDQ297576DQ297600
7 R. pusillusYunnan, ChinaDQ297574DQ297598
8 R. pusillusSichuan, ChinaDQ297595DQ297618
9 R. pusillusSichuan, ChinaDQ297589DQ297613
10 R. pusillusHainan, ChinaDQ297590DQ297614
11 R. pusillusGuizhou, ChinaDQ297577DQ297601
12 R. pusillusGuangdong, ChinaDQ297597DQ297620
13 R. pusillusHubei, ChinaDQ297583DQ297607
14 R. pusillusBeijing, ChinaDQ297588DQ297612
15 R. cornutusKashi, Nagasaki, JapanDQ297594DQ297617
16 R. cornutusKashi, Nagasaki, JapanDQ297593DQ297621
17 R. cornutusOkupirika, Hokkaido, JapanDQ297591DQ297615
18 R. cornutusOkupirika, Hokkaido, JapanDQ297592DQ297616
19 R. c. pumilus**RyukyuNC_005434NC_005434
20 R. pearsoniiSichuan, ChinaDQ297587DQ297611
21 R. affinisGuizhou, ChinaDQ297582DQ297606
22 R. hipposiderosUpper Langford, United KingdomDQ297586DQ297610
23 R. ferrumequinumYunnan, ChinaDQ297575DQ297599
24 R. luctusHuibei, ChinaDQ297596DQ297619
25 H. armigerGuizhou, ChinaDQ297585DQ297609
26 H. prattiGuangxi, ChinaDQ297584DQ297608

Appendix II

View this table:

Sequence divergence matrix based on complete mitochondrial cytochrome-b (Cytb) sequences (1,140 base pairs, above the diagonal) and partial control region sequences (494 base pairs, below the diagonal) for species of Rhinolophus. Numbers designating rows and columns correspond to numbers for specimens in Appendix I (1-6 represent R. monoceros, 7-14 R. pusillus, 15-18 R. cornutus, 19 represents R. c. pumilus, and 20, 21, 22, 23, and 24 correspond, respectively, to R. pearsonii, R. affinis, R. hipposideros, R. ferrumequinum, and R. luctus). The boldface numbers are values of genetic distance among the little or lesser horseshoe bats of East Asia.

2 0.0080.0070.0060.0070.0070.0170.0180.0190.0220.0210.0210.0170.0190.0290.0250.0260.0260.0310.110.1170.1410.1180.133
3 0.0140.0230.0030.0050.0040.0130.0140.0150.0190.020.0170.0130.0150.0250.0230.0220.0220.0250.110.1170.140.1170.131
4 0.0270.0350.0210.0030.0010.0120.0130.0140.0180.0190.0160.0120.0140.0240.0220.0210.0210.0260.1090.1160.1390.1170.132
5 0.0310.040.0250.0120.0020.0130.0140.0150.0190.020.0170.0130.0150.0250.0210.0220.0220.0290.1120.
6 0.0350.0440.0330.0330.0250.0130.0140.0150.0190.020.0170.0130.0150.0250.0230.0220.0220.0270.110.1170.1370.1180.133
7 0.0310.040.0380.0550.060.060.0030.0110.0140.0130.0110.0090.0110.030.0280.0270.0270.0340.1110.1190.1410.1190.136
8 0.0310.040.0380.0550.060.060.0040.0120.0150.0160.0130.010.0120.0310.0290.0280.0280.0340.1120.120.1420.1230.137
9 0.060.0690.0620.0640.0670.0750.0510.0510.0140.0150.0120.0090.0110.0320.030.0290.0290.0350.1140.1230.1410.1170.134
10 0.0490.0570.0460.0530.0510.0550.0380.0420.0330.0170.0140.0120.0120.0350.0330.0330.0330.0390.1140.1170.1360.120.132
11 0.0460.0550.0440.0550.0550.060.0440.0440.040.0310.010.0130.0150.0360.0340.0330.0330.040.1130.1230.1420.1190.137
12 0.0550.0640.0580.0690.0680.0640.040.0440.0490.0360.0250.0110.0120.0330.0320.0310.0310.0370.1120.120.140.1160.134
13 0.0550.0640.0530.0620.0730.0730.0490.0490.0530.040.0510.0510.0020.0280.0260.0250.0250.0320.1110.1210.1360.1170.132
14 0.060.0690.0580.0660.0780.0780.0490.0490.0580.0440.0510.0560.0120.030.0280.0270.0270.0340.1110.1210.1340.1150.13
15 0.0550.0640.0620.060.0680.0710.0570.0570.0680.0640.0660.0710.0690.0640.0070.0060.0060.0180.1170.1220.1430.1170.13
16 0.060.0680.0620.060.0680.0660.0620.0620.0730.0640.0710.080.0690.0690.0120.0040.0040.0160.1160.120.1440.1150.131
17 0.0530.0570.0550.0530.0620.0640.0550.060.0710.0570.0690.0730.0660.0660.0210.01600.0150.1140.1150.1390.1120.128
18 0.0550.0590.0570.0570.0660.0690.0570.0620.0730.060.0710.0760.0690.0640.0250.0210.0040.0150.1140.1150.1390.1120.128
19 0.0730.0820.0660.0590.0730.080.0750.080.0770.0680.080.0870.0730.0780.0510.0550.0490.0530.1140.1220.1360.1120.127
20 0.1620.1690.1490.1540.1640.170.1540.1590.1640.1560.1540.1480.1540.1610.1480.1510.1410.1430.1460.1210.1310.1230.13
21 0.1670.1740.1590.1590.1690.1720.1770.1830.1820.1670.1770.1640.1580.1690.1660.1720.1660.1720.1640.1470.140.1210.135
22 0.1570.1630.1520.1550.160.1630.1470.1520.140.140.1550.1450.1550.1550.1570.1620.1550.160.150.1840.2020.1250.129
23 0.1680.1750.170.1680.1730.1730.1680.1730.1520.1420.1580.160.1650.1710.1570.160.150.1550.1550.1660.1870.1380.126
24 0.1780.1880.1780.1860.1850.1880.170.1750.1670.160.170.1730.1830.1830.1620.1750.1670.1730.1750.1860.210.170.155


  • Associate Editor was Jesús Maldonado.

Literature Cited

View Abstract