Male Ambulyx kuangtungensis. Photo: © NHMUK. Male Ambulyx kuangtungensis f. adhemariusa , Luoyang, Henan, China. Photo: © Kenichiro Nakao.


Oxyambulyx kuangtungensis Mell, 1922, Dt. ent. Z. 1922: 114. Type locality: China, northwest Kwangtung [Guangdong].

Synonym. Oxyambulyx kuangtungensis Mell, 1922.

Synonym. Oxyambulyx kuangtungensis formosana Clark, 1936.

Synonym. Oxyambulyx kuangtungensis hoenei Mell, 1937.

Synonym. Oxyambulyx kuangtungensis melli Gehlen, 1942.

Synonym. Oxyambulyx takasago Okano, 1964.

Synonym. Ambulyx adhemariusa Eitschberger, Bergmann & Hauenstein, 2006.

Note. The holotype and a paratype of Ambulyx adhemariusa fall within part of a tight DNA barcode cluster of 'typical' Ambulyx kuangtungensis from Fujian, Hubei, Shaanxi, Sichuan, Yunnan and Xizang/Tibet and should be considered, at best, a form. Furthermore, Ambulyx kuangtungensis 'form' adhemariusa is not restricted to the type locality but is quite widespread, though rare, throughout the range of the species, as several posted images can testify.

Kitching (6.i.2017) expands on this in the Sphingidae Taxonomic Inventory --- STI, saying "Synonymized with Ambulyx kuangtungensis by Brechlin & Haxaire, 2014, Entomo-Satsphingia 7(2): 47. Reinstated as a species by Eitschberger, 2016, Neue ent. Nachr. 70: 149. Eitschberger et al. (2006) differentiated Ambulyx adhemariusa from Ambulyx kuangtungensis primarily on the basis of a complete, or nearly complete, basal band on the forewing upperside. In comparison, the basal band in Ambulyx kuangtungensis, as in all other species of the genus except Ambulyx japonica, comprises at most two separate elements, a dark, semicircular mark on the costa and a circular or ovoid spot near the posterior margin (although either or both may be absent as part of normal intraspecific individual variation). They also noted that the male genitalia were very similar to those of Ambulyx kuangtungensis, although there were a 'great number of differences'. However, these differences were not described, but only noted to exist in sacculus, harpe, uncus, valve and aedeagus, and the reader was simply referred to the illustrations to determine these differences for themselves. The only difference explicitly described by Eitschberger et al. (2006) was that during preparation of the genitalia, the scales on the valve (most likely the stridulatory scale patch, though this is not clear) detached as a unit in Ambulyx adhemariusa, whereas in Ambulyx kuangtungensis and other species studied in that paper, they detached individually.

Brechlin & Haxaire (2014) later synonymized Ambulyx adhemariusa with Ambulyx kuangtungensis. They argued that the form of the basal band on the forewing upperside was not diagnostic as, contrary to the statement by Eitschberger et al. (2006), who only illustrated the holotype, it was not present in all 11 paratypes. However, they did not state how they had arrived at this conclusion. Other than that feature, Brechlin & Haxaire (2014) noted that no other diagnostic differences in the habitus had been presented or were evident. They also criticized the lack of explicitly described diagnostic features of the male genitalia by Eitschberger et al. (2006) and questioned the validity of the valve scale character. In the same paper, Brechlin & Haxaire (2014) then described a closely related new species, Ambulyx latifascia, which although it had a similar continuous basal band differed in other respects and detailed descriptions of these, for both the habitus and male genital structure, were given. Brechlin & Haxaire (2014) also justified their synonymy of Ambulyx adhemariusa and their description of Ambulyx latifascia as a new species on the basis of results from analysis of DNA barcode data. Unfortunately, they did not provide any details in support of this conclusion.

Eitschberger (2015) took great exception to both the synonymy and new species, referring indirectly to the paper by Brechlin & Haxaire (2014) as a 'published fairytale' in his English abstract. His first criticism was levelled at the claim by Brechlin & Haxaire (2014) that not all of the paratypes had the complete basal band, which he refuted by publishing images of all 12 specimens. In this regard, Eitschberger was correct, but rather than trying to determine the source of the error, he simply asked the snide question 'Wer kann diesen dans ins Ohr geflüstert haben' ['Who can have been whispering this in their ears?']. As it turns out, Brechlin & Haxaire (2014) had simply been misled by a record in the Barcoding of Life Database (BOLD) that had been mistakenly labelled as a paratype (BC-EMEM0865 from 'Ankang, Suburn' ['Ankang suburb'?]. In fact, this specimen was actually the Ambulyx kuangtungensis whose genitalia (GP4344) Eitschberger et al. (2006) had compared with those of a paratype of Ambulyx adhemariusa (GP4342). This error in BOLD was corrected on 6.x.2016. Eitschberger (2015) did not address the lack of explicit descriptions of diagnostic features, effectively saying that if the reader could not determine those differences by comparing the illustrations, then there was nothing more to be said. He then proceeded to reinstate Ambulyx adhemariusa to species status by assertion, and synonymized Ambulyx latifascia with it 'since the agreement is just too large, not only externally but in the genital morphology, there is no escaping their synonymy'. Just how the two taxa agreed in morphology was not explained and the DNA barcode data was just dismissed as irrelevant.

Overall, the arguments and conclusions put forward by Brechlin & Haxaire (2014) are considered here to be unrefuted and Ambulyx latifascia is treated as a good species. Furthermore, Ambulyx adhemariusa is placed back into synonymy with Ambulyx kuangtungensis as a form in which the costal and inner edge elements of the basal band of the forewing upperside have expanded to form an almost complete band. DNA barcode analysis supports these treatments. Not only do the two species fall into two different BINs (Ambulyx latifascia AAB6407; Ambulyx kuangtungensis AAB6406) but the holotype and a paratype of Ambulyx adhemariusa form part of a tight cluster with 'typical' Ambulyx kuangtungensis from Fujian, Hubei, Shaanxi, Sichuan, Yunnan and Xizang/Tibet (China). Furthermore, Ambulyx kuangtungensis f. adhemariusa is not restricted to the type locality but is quite widespread, though rare, throughout the range of the species, as several published images testify."


Wingspan: 65--80mm. Some forms may resemble Ambulyx latifascia (Brechlin & Haxaire, 2014).

Resting Ambulyx kuangtungensis, Taiwan. Photo: © Felix Lin. Resting Ambulyx kuangtungensis, Taiwan. Photo: © Felix Lin. Resting Ambulyx kuangtungensis, Jiucai Ling, Guizhou, China. Photo: © Viktor Sinjaev. Resting Ambulyx kuangtungensis, Kuocang Mountain Nature Reserve, Zhejiang, China, 21.v.2015. Photo: © Chen Chuntang.



China: vi (Laojun Shan & Gaoligong Shan, Yunnan); vii (Luotian, Hubei; Henan); viii (Guangze, Fujian); 20-23.ix (Baoshan, Yunnan); ix (Xunyang, Shaanxi); 15.ix (Xishuangbanna, Yunnan); 27.ix (Lijiang, Yunnan). Taiwan: iii-vi (Hualien Hsien); 28.vii (Hualien Hsien).





Pupa of Ambulyx kuangtungensis. Image: Mell, 1922b Pupa of Ambulyx kuangtungensis. Image: Mell, 1922b

Larval hostplants. Recorded in Guangdong on Choerospondias axillaris (as Poupartia fordii) (Anacardiaceae) (Mell, 1922b). A record on Aceraceae (Chang, 1989) may be an erroneous transferred record from another Ambulyx species and requires confirmation.

In Taiwan, recorded only from Pistacia chinensis (Anacardiaceae).




China: Shaanxi (Xunyang, 1380m); Gansu; Henan (Luoyang, 1763m); Jiangsu (Baohuashan National Forest Park, Zhenjiang; Kongshan, Nanjing, 341m); Zhejiang (Kuocang Mountain Nature Reserve); Hubei (Luotian, 1729m); Sichuan; Yunnan (Lijiang, 2700m; Mengla County; Laojun Shan, 2579m; Baoshan, 1000-1500m; Gaoligong Shan); Xizang/Tibet; Guizhou (Jiucai Ling; Xinzhaidashan, Zhijin County, 1000-2000m); Fujian (Guangze, 1200m); Guangdong (Meixian; Nanling National Forest Park); Guangxi; Hainan (Duowen Ling, nr Lingao).

Taiwan: Hualien Hsien (Taroko National Park; Tianxiang); Nantou Hsien (Puli); Taichung Hsien (Devlan Trail (Guguan), 645m).


Besides southern China, it is also found in northern Burma/Myanmar (Banmauk, Sagaing Province), northern Thailand, northern Laos, northern Vietnam and Taiwan.

Global distribution of Ambulyx kuangtungensis. Map: © NHMUK.


Holarctic; eastern Palaearctic region. Pleistocene refuge: Monocentric -- Yunnan refugium.

Return to Sphingidae of the Eastern Palaearctic species list
© A. R. Pittaway & I. J. Kitching (Natural History Museum, London)