An integrative approach to characterize Malagasy bats of the subfamily Vespertilioninae Gray, 1821, with the description of a new species of Hypsugo Author Goodman, Steven M. Author Rakotondramanana, Claude Fabienne Author Ramasindrazana, Beza Author Kearney, Teresa Author Monadjem, Ara Author Schoeman, M. Corrie Author Taylor, Peter J. Author Naughton, Kate Author Appleton, Belinda text Zoological Journal of the Linnean Society 2015 2015-04-30 173 4 988 1018 https://academic.oup.com/zoolinnean/article-lookup/doi/10.1111/zoj.12223 journal article 10.1111/zoj.12223 0024-4082 5335115 PIPISTRELLUS RACEYI BATES ET AL ., 2006 Molecular genetics This endemic species is known from two distinct portions of Madagascar . The samples used in the intraspecific genetic analysis include ten animals from dry deciduous western forests and nine specimens from humid eastern area. The K2P distance within P . raceyi when these lineages were pooled was 0.004 ( N = 7, Table 1 ). No haplotypes were shared between the eastern and western populations ( N = 17, Table 9 ). These populations exhibited a divergence of 0.74% (K2P), reflecting five fixed mutations in the trimmed cytochrome b sequences (680 bp). When the full sequence set of P . raceyi was examined ( N = 17, Table 9 ), diversity was reasonably high in the western population (Hd 0.722, Pi 0.00287), but was noticeably low in the eastern population (Hd 0.429, Pi 0.00057), which was only represented by two different haplotypes and distinguished by a single base change. Table 8. Descriptive statistics of baculum total length (TL) measurements in Malagasy vesper bats of the genera Hypsugo , Neoromicia , and Pipistrellus , with most individuals represented in the molecular genetic portion of this study. For species represented by ≥ 3 individuals, data are expressed as mean ± SD (standard deviation), minimum– maximum, N = number of individuals. This species is notably divergent from the other Vespertilioninae samples included in the analysis, specifically members of the genus Pipistrellus , and without more extensive taxonomic and geographic sampling, little can be advanced concerning possible sister taxa or the geographical region the ancestral form that colonized Madagascar was from. Bates et al . (2006) suggested, based on baculum morphology, that P . raceyi might be closely related to Asiatic P . endoi Imaizumi, 1959 . This interesting hypothesis warrants further molecular investigation.

Species	TL
N . malagasyensis	2.10, 2.25, N = 2
N . matroka	2.26 ± 0.128
2.12–2.49, N = 6
N . robertsi	3.12, N = 1
P . hesperidus	1.73 ± 0.153
1.60–2.10, N = 4
P . raceyi (east)	6.10, 10.00, N = 2
P . raceyi (west)	8.90, N = 1
H . bemainty	1.70 ± 0.097
1.50–1.79, N = 7

Morphometrics Measurements presented in Table 2 . Animals from eastern populations are in general larger than those from western populations. Figure 8. Dorsal (above) and lateral (below) views of bacula from two species of Pipistrellus . A, Pipistrellus raceyi (FMNH 217886, total length 8.90 mm) from Kirindy Forest (CNFEREF); B, Pipistrellus raceyi (FMNH 222722, total length 6.10 mm) from near Andasibe; C, Pipistrellus hesperidus (FMNH 176094, average total length of four specimens 1.9 mm) from Parc National de Kirindy Mitea (redrawn from Bates et al ., 2006 , Fig. 10A); D, Pipistrellus hesperidus (FMNH 209270, total length 1.90 mm) from Grotte de Sarodrano; E, Pipistrellus hesperidus (FMNH 217905, total length 1.70 mm) from Grotte de Sarodrano; F: Pipistrellus hesperidus (TM 48624, UP-840, total length 2.18 mm) from South Africa. Table 9. Intraspecific diversity and divergence for the three most intensively sampled taxa in the genera Hypsugo and Pipistrellus . Numbers in parentheses indicate standard deviation (SD) for estimates of haplotype diversity (Hd) and nucleotide diversity (Pi). There were no shared haplotypes between the east and west populations of Pipistrellus raceyi (five fixed mutations between lineages) or between the African and Malagasy populations of Pipistrellus hesperidus (22 fixed mutations between lineages)

Species	Population	Hd (SD)	Pi (SD)	% divergence (K2P)
H . bemainty P . raceyi	n/a* ( N = 13) East ( N = 8) West ( N = 9)	0.962 (0.050) 0.429 (0.028) 0.722 (0.159)	0.00519 (0.00072) 0.00057 (0.00022) 0.00287 (0.00439)	n/a 0.74%
P . hesperidus	Madagascar ( N = 28) Africa†	0.744 (0.078) n/a	0.00181 (0.00035) n/a	1.89%

* Hypsugo bemainty was only sampled as a single population. †Only two samples were available from the African population of Pipistrellus hesperidus and intraspecific diversity could not be calculated. Craniodental morphology As mentioned under this header for P . hesperidus , members of this genus occurring on Madagascar are easily differentiated from Neoromicia based on upper toothrow counts. Further, Bates et al . (2006) have provided a few characters to differentiate P . hesperidus from P . raceyi , the other member of the genus occurring on Madagascar : (1) the second upper incisor (I 3 ) ≤½ the length of the first upper incisor (I 2 ) (in hesperidus ) vs. I 3 nearly the length of I 2 (in raceyi ); (2) I 2 is unicuspid (in hesperidus ) vs. bicuspid (in raceyi ); and (3) that the first upper premolar (P 3 ) is notably small (in hesperidus ) as compared to large and prominent (in raceyi ). A reevaluation of these characters based on the specimens of P . hesperidus sequenced in this study ( N = 13) found: characters 1 – consistent in all cases, character 2 – almost uniformly consistent, with the exception of MNHN 1986.1074 with a seemingly bicuspid I 2 , character 3 – consistent in all cases, with some variation in size of the P 3 , which was always smaller than in P . raceyi . Bates et al . (2006) also noted that the relative length of I 3 to I 2 between P . hesperidus and Hypsugo anchietae [= H . bemainty sp. nov. , see below] were similar and the best means to differentiate these two taxa based on craniodental characters was the former has a distinctly thin zygomata and without a jugal eminence (fig. 11 in Bates et al ., 2006 ). Bioacoustics Bioacoustics (average values, sexes combined, N = 4, which include from the west 2 ♂♂ and 1 ♀ and from the east 1 ♂ ): PF: 56.6 (range 53.4–58.2) kHz, Fmax: 114.8 (range 91.7–128.3) kHz, Fmin: 46.0 (range 41.7– 51.4) kHz, Dur: 2.6 (range 2.3–3.1) ms, IPI: 67.3 (range 14.8–125.0) ms ( Table 6 ). The single sequenced female with associated bioacoustic recordings had several parameters that fell outside the range for three males ( Fig. 6 ). Further work is needed to verify this pattern. Bacular morphology In their description of P . raceyi , Bates et al . (2006) provided details on its bacular morphology. The length of this structure in P . raceyi is distinctly longer than any other Malagasy vesper and measures 8.90 and 10.00 mm ( Table 8 ). It has a long and narrow shaft, with a slight curvature, and variable deflection at the distal end that seems to indicate some intraspecific variation ( Fig. 8A, B ). A specimen obtained outside of Andasibe (FMNH 222722) had a distinctly shorter baculum ( 6.10 mm ), which may be related to age and lack of ossification of the structure ( Table 8 , Fig. 8B ). Known geographical range Localities from which sequenced specimens of this species are known are presented in Figure 1 . One specimen from another site referred to this species includes Forêt de Mikea , 22°16.00′S , 43°28.70′E ( FMNH 176165 ) ( Bates et al ., 2006 ) . Taxonomic comments Notable genetic differences were found between eastern and western populations of P . raceyi , with the type specimen being from the eastern locality of Kianjavato ( Fig. 1 ). Further, statistically significant differences were identified between specimens from these two different portions of the island in several external measurements ( Table 2 ) and in all cranial and dental measurements ( Tables 3 and 4 ). Further work is needed to assess the level of genetic divergence and possible reproductive isolation between eastern and western populations; herein we maintain them as a single taxon, with divergent populations.