A database based on distributional records of Eurasian Zephyrus hairstreaks (Lepidoptera: Lycaenidae: Theclini) was compiled to analyse their areas of endemism (AoEs), species richness and distribution patterns, to explore their locations of past glacial refugia and dispersal routes.
Over 2000 Zephyrus hairstreaks occurrences are analysed using the NDM/VNDM algorithm, for the recognition of AoEs. Species richness was calculated by using the option ‘Number of different classes’ to count the different classes of a variable presented in each 3.0°×3.0° grid cell, and GIS software was used to visualize distribution patterns of endemic species.
Centres of species richness of Zephyrus hairstreaks are situated in the eastern Qinghai-Tibet Plateau (EQTP), Hengduan Mountain Region (HDMR) and the Qinling Mountain Region (QLMR). Latitudinal gradients in species richness show normal distribution with the peak between 25° N and 35° N in the temperate zone, gradually decreasing towards the poles. Moreover, most parts of central and southern China, especially the area of QLMR-EQTP-HDMR, were identified as AoEs that may have played a significant role as refugia during Quaternary global cooling. There are four major distributional patterns of Zephyrus hairstreaks in Eurasia: Sino-Japanese, Sino-Himalayan, high-mountain and a combined distribution covering all three patterns.
Zephyrus hairstreaks probably originated at least 23–24 Myr ago in E. Asia between 25° N to 35° N in the temperate zone. Cenozoic orogenies caused rapid speciation of this tribe and extrusion of the Indochina block resulted in vicariance between the Sino-Japanese and the Sino-Himalayan patterns. The four distribution patterns provided two possible dispersal directions: Sino-Japanese dispersal and Sino-Himalayan dispersal.
Citation: Zhuang H, Yago M, Settele J, Li X, Ueshima R, Grishin NV, et al. (2018) Species richness of Eurasian Zephyrus hairstreaks (Lepidoptera: Lycaenidae: Theclini) with implications on historical biogeography: An NDM/VNDM approach. PLoS ONE 13(1): e0191049. https://doi.org/10.1371/journal.pone.0191049
Editor: Ulrike Gertrud Munderloh, University of Minnesota, UNITED STATES
Received: June 4, 2017; Accepted: December 27, 2017; Published: January 19, 2018
Copyright: © 2018 Zhuang et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This work was supported Grant-in-Aid for Scientific Research (C) (No. 26440207 to MY), (http://www.um.u-tokyo.ac.jp/people/faculty_yago.html). This funder provided the idea, fund and resources and participated writing original draft and revision. This work was also supported by the Chinese National Natural Science Foundation (No. 40971037 to MW). This funder provided the idea, fund, resource and participated filed survey, confirmed test. Lastly, it was funded by the CSC (China Scholarship Council, Liujinfa  3022 to HZ. This funder conducted almost all aspects of this research expect, resources and validation.
Competing interests: The authors have declared that no competing interests exist.
All living things on earth are not distributed randomly, but are restricted to a defined geographical area . Two very important issues of biogeography are the identification of areas of endemism (AoEs) and centres of species richness. AoEs defined by at least two endemic taxa are generally treated as keys to link the distribution of organisms and historical geographical events [2, 3], and these areas are characterized with high species survival and speciation rates. Furthermore, these areas are widely regarded as refugia during the Last Glacial Maximum . Consequently, they are usually treated as historical units used to infer the relationship between the distribution of organisms and historical factors [5–10]. Biogeographical boundaries, which are often affected by multiple factors (e.g. biological, physical, ecological, historical), have been of particular interest due to their impact on ecology, biogeography, evolution and conservation biology . Biogeographical lines, the most famous of which is Wallace’s Line , are the boundaries defining the distribution of organisms. These lines reflect some consistency in the influence of ecological, physical, and geological factors, as well as historical events. For instance, the biogeographical line called the “Tanaka-Kaiyong line” which extends from northwest Yunnan through southeast Yunnan to Guangxi, China, and to northern Vietnam (about from 28° N, 98° E to 19° N, 108° E) was suggested by several researchers based on their floristic surveys [13–15].
Lepidoptera (moths and butterflies) form the second largest order in the class Insecta and are an ideal group for biogeographical studies due to their abundance and ubiquity [16–20]. However, the study on biogeography of Lepidoptera, especially at broad geographical scales, is not comprehensive enough due to lack of sufficiently detailed and available data in Eurasia. The Zephyrus hairstreaks (Lepidoptera: Lycaenidae: Theclini) are mainly distributed in Eurasia and are characterized by their brilliantly shining wings in a dazzling variety of metallic colours, such as metallic green, blue, purple, orange, silvery white and others. Therefore, they are often referred to as “living jewels”, enchanting naturalists, lepidopterists and amateurs alike . This tribe was selected as our study objective for several reasons: Firstly, the outstanding taxonomic work, including 197 known species belonging to 53 genera of Theclini and their detailed distributional data reported prior to 2016, has been well synthesized [22–51]. Secondly, the description and revision of genera are beginning to stabilize, although there have been two new genera and some new species proposed since a relatively detailed survey of China and Indo-China in 2007, though there were several new species reported from northwest China and Indo-China [22, 28, 29, 31, 39, 41, 46–48, 51]. Thirdly, monophyly of this tribe was been discussed based on morphological features and ecological information [21, 22] and tested using molecular analyses (unpublished research by Zhuang et al., and Hsu et al.). However, their biogeographical relations are still poorly understood. By detecting the AoEs from the data, we can obtain new insights into the origins, diffusion and distribution of this group.
East Asia, located on the west side of the Pacific Ocean, is characterized by complex geological structures due to geographical events throughout history, such as the Indo-Asian tectonic collision which caused the formation of the Himalaya, uplifted the Qinghai-Tibetan plateau, and caused the rotation of the Shan-Malay Plate, and the glacial and interglacial periods during the Quaternary. The areas, affected by the collision and global cooling, are also biodiversity hotspots and are the focus of current research .
Our study examines the AoEs, centres of species richness of Zephyrus hairstreaks and their distribution patterns in Eurasia. By integrating the information from analyses based on our database, we aim to explain where the candidates of refugia of this tribe are and how the dispersal to the current distribution patterns can be reconstructed. This garnering information is useful for conservation applications.
Materials and methods
Species distribution data
We compiled a database including 2126 records of 197 known species belonging to 53 genera of Theclini. We included 1564 records of 189 Eurasian endemic species assigned to 47 genera to analyse species richness. Moreover, 1517 records of 142 endemic species from 42 genera (excluding species from only a single location) were used to identify areas of endemism (AoEs). Geographical distribution data and taxonomic information for each species were compiled and reviewed from data in literature published prior to 2016 [24–51]. Classification and nomenclature follow the literature of “The Zephyrus Hairstreaks of the World” by Koiwaya . A list of species records and complete geographical information regarding the specimen collection sites and references can be found in the supporting information (Table in S1 Table).
Description of the study area
An area covering latitudes -10° S to 80° N and longitudes from 40° E to 180° E, including most parts of the countries of Eurasia, was subdivided into quadrats measuring 2° × 2° and 3° × 3° Operative Geographical Units (OGU) without consideration of physiographical features and the uniformity of quadrats in the margin of the survey area that are only partly land . This region covers most parts of the known distributional area of Zephyrus Hairstreaks in Eurasia.
Species richness is the most direct method to evaluate species diversity. A total of 1564 distributional records were coded as a shapefile that was then run in DIVA-GIS 7.5, using 3° × 3° grid cells. DIVA-GIS 7.5 is a free computer program for mapping and analyzing geographical distributions and biodiversity data [53–55]. Species richness was analyzed using the option “Number of different classes” to count the different classes of a variable present in each grid cell. A list of conditioned equal interval values was taken from a species richness map of the study area. Before analysis, a species accumulation curve at genus level has been tested, using our database, and the curve tends to go up and then it tends to slow down. It reflects a relatively adequate sampling.
NDM/VNDM 3, two sister programs written by Goloboff , implement the methods described in Szumik et al., and Szumik & Goloboff [57, 58] to demarcate AoEs. These programs, based on an optimality criterion, take into account the spatial component of endemism for identifying AoEs. Parameters were selected in accordance with Prado et al. . Scores above 2.0 were saved for genus-level analysis and above 3.0 for species-level analysis, and set the upper minimum score above 0.3 (except for Hayashikeia florianii (0.289)). The consensus AoEs were computed using a cut-off of 50% similarity at different taxonomic levels and the flexible consensus was selected [56, 57]. Endemicity analysis (EA) was selected to identify areas of endemism and executed by NDM, because EA has been recognized as a best method that supported areas of endemism . Analytical procedures are filed in http://dx.doi.org/10.17504/protocols.io.[dx.doi.org/10.17504/protocols.io.jr2cm8e].
AoEs were constructed from 2.0°×2.0° and 3.0°×3.0° cell sizes at each taxonomic level (genus and species). The two sizes of cells were optimized in consideration of the advantages and disadvantages of each particular size . Different taxonomic levels used in the analysis of endemicity can increase the reliability [61, 62].
Species richness of the tribe Theclini
Three centres of species richness of Zephyrus hairstreaks are situated in the eastern Qinghai-Tibet Plateau (EQTP), the Hengduan Mountain Region (HDMR), and the Qinling Mountain Region (QLMR). The highest values of the centres of species richness reached 74, 60 and 48 in one grid cell of 2°×2° in these three areas, respectively. The remaining portions of the mountains and islands also contain a relatively high abundance, such as the tri-border region of China, Laos and Vietnam (TCLV, value 29), Taiwan island, China (value 25), the Wuling Mountain Region (WLMR) located in central China and the Tianmu Shan region (TMS) (value both 24), Japan (value 23), the Russian Far East (value 20), the Korean peninsula (value 19), the Changbai Mountain region and the Nanling Mountain region (NLM) of China (value both 17), the eastern Himalaya (EHMLY, value 16), Hainan Island, China, and Sakhalin Island, Russia (values of both 6). Some species were also unevenly distributed in the chain of the Himalaya (Fig 1).
Latitudinal gradients in species richness
The value of species richness of Zephyrus hairstreaks in each grid cell is variable in Eurasia. Although for certain grid cells, where the species have not been observed, this may either represent real absences or sampling errors, the database meets the analysis criterion and reflects reasonable real world results. Latitudinal gradients in species richness show a normal distribution (Kolmogorov-Smirnov Z = 1.281, P = 0.075 > 0.05) with one clear peak between 25° N and 35° N in the temperate zone (Fig 2), and a gradual decrease towards the poles. The prevalent consensus regarding species richness is that diversity generally peaks in the tropics near the equator and declines towards the poles.
Areas of endemism
The NDM/VNDM 3 analytic approach identified fourteen different AoEs (Figs 3 and 4) at two taxonomic levels (genus and species) and two different Operative Geographical Units (2.0°×2.0° and 3.0°×3.0°). At the genus level, there were three consensus endemic areas (Fig 3a–3c) searched, using the heuristic algorithm of NDM/VNDM 3 and supported by the values of 2.13–2.38 (Fig 3a in 2.0° grid size), 2.26–3.26 (Fig 3b in 3.0° grid size), and 2.26–2.51 (Fig 3c in the 3.0° grid size), respectively. The consensus endemic areas of Fig 3a and 3c were formed of the same three genera (Neogonerilia, Saigusaozephyrus and Shaanxiana) but in different proportions (Table in S2 Table), with both of these two regions being associated with the eastern Qinghai-Tibet Plateau and the Qinling Mountain Region (EQTP-QLMR). Areas of endemism from Fig 3b (QLMR-EQTP-WLMR-TMS-HDMR-NLM) mainly covered central and southern China. At the specific level, there were eleven overlapping or non-consensus endemic areas identified, based on two different Operative Geographical Units (2.0° and 3.0°). They were assembled over different regions; the Qinling Mountain Region (QLMR), the eastern Qinghai-Tibet Plateau (EQTP), the Wuling Mountain Region (WLMR), the Tianmu Shan region (TMS), the Hengduan Mountain Region (HDMR), the Nanling Mountain Region (NLM), the eastern Himalaya (EHMLY) and the tri-border region of China, Laos and Vietnam (TCLV). In the 2.0° grid size, there were four consensus areas (Fig 3d–3g): QLMR-EQTP-HDMR (score range 3.04–4.27), QLMR-EQTP (score range 5.71–6.46), EQTP (score range 7.64–8.14) and HDMR (score range 3.67–3.92), all of which were derived from an optimality criterion to demarcate AoEs. There were seven consensus areas (Figs 3h and 4i–4m) in 3.0° grid size; QLMR-EQTP-WLMR (score range 3.07–3.32), EQTP-HDMR-EHMLY-TCLV (score range 3.38–3.63), HDMR-TCLV (score range 3.84–4.09), QLMR-EQTP-TMS (score range 4.34–4.59), EQTP-HDMR (score range 3.97–4.22), QLMR-EQTP-HDMR (score range 17.33–17.58) and EHMLY (score range 3.30–3.55).
Areas of endemism at generic level: 2.0° (a) and 3.0° (b, c); Areas of endemism at specific level 2.0° (d-g) and 3.0° (h). Lower cases indicate those areas assembled over different regions: (a, c, e): QLMR-EQTP. (b): QLMR-EQTP-WLMR-TMSM-HDMR-NLM. (d, m): QLMR-EQTP-HDMR. (f): EQTP. (g): HDMR. (h): QLMR-EQTP-WLMR.
The core areas of endemism
Three core consensus AoEs (EQTP, HDMR and EHMLY) (Fig 5) had high consensus scores of endemicity and could be identified at the specific level. EQTP is supported as an independent AoE in the 2.0° grid size by 15 species with differing consensus scores; Chrysozephyrus linae (0.303–0.324), Chrysozephyrus marginatus (0.688), Chrysozephyrus sakula (0.625), Cordelia koizumii (0.545), Gonerilia pesthis (0.393), Hayashikeia courvoisieri (0.688), Hayashikeia florianii (0.289), Howarthia caelestis (0.364–0.391), Howarthia nigricans (0.719), Howarthia sakakibarai (0.639), Neozephyrus coruscans (0.000–0.344), Neozephyrus helenae (0.719), Teratozephyrus chibahideyukii (0.719), Teratozephyrus hecale (0.463) and Uedaozephyrus kuromon (0.438). Compared with that, there were five species in the consensus AoE (HDMR) in the 2.0° grid size with different contribution scores; Chrysozephyrus yunnanensis (0.750), Euaspa mikamii (0.750), Neozephyrus dubernardi (0.833), Shirozuozephyrus nansarae (0.667) and Shirozuozephyrus nyishwini (0.667). Among the three core AoEs, EQTP and HDMR at the same time have high species richness. In the 3.0° grid size, SHEMY was also identified as a core consensus AoE, supported by five species; Chrysozephyrus letha (0.714), Chrysozephyrus uedai (0.633), Neozephyrus suroia (0.455), Shirozuozephyrus jakamensis (0.800) and Shirozuozephyrus khasia (0.700). Additionally, the consensus area of QLMR-EQTP-HDMR (Fig 5) was obtained from the analysis results using two grid sizes (2.0° and 3.0°) with its highest support scores (up to 17.32973–17.57973) made up of 28 species (in the 3.0° grid size) (Table in S2 Table) based on NDM/VNDM 3. QLMR-EQTP (Fig 5) was strongly supported as an AoE for the reason that, at the generic level, it was supported by three genera with different scores in the 2.0° and 3.0° grid sizes; Neogonerilia (0.600/0.800), Saigusaozephyrus (0.680/0.605) and Shaanxiana (0.850/0.857). Furthermore, at the species level, there were twelve species supporting an endemic area in the 2.0° grid size (Table in S1 Table); Chrysozephyrus gaoi (0.469–0.591), Chrysozephyrus souleanus (0.000–0.306), Chrysozephyrus tatsienlurnsis (0.444–0.455), C. koizumii (0.750–0.763), G. pesthis (0.750–0.833), Gonerilia thespis (0.000–0.621), N. coruscans (0.643–0.682), Saigusaozephyrus atabyrius (0.646–0.854), Shaanxiana takashimai (0.630–0.792), Shirozua melpomene (0.403–0.419), T. hecale (0.000–0.537) and Wagimo sulgeri (0.000–0.556). It is worth noting that while both EQTP and HDMR are centres of species richness and AoEs which are located adjacent to each other, they appear independent, differing with each other. For instance, both EQTP and HDMR are supported as independent AoEs in the 2.0 grid size, and moreover, EQTP-HDMR (Fig 5) has a lower support rate (only one time) and values (3.974–4.224) compared with those of QLMR-EQTP (three times and their values as follows: 2.130–2.380, 2.262–2.512 and 5.714–6.464). QLMR-EQTP-HDMR is treated as a large and important region for species richness and AoEs of Zephyrus hairstreaks and among these areas QLMR has a particularly closer connection with EQTP (because QLMR lacks independence and was three times recognised together with EQTP) more than other AoEs.
QLMR: Qinling Mountain Region. EQTP: Qinghai-Tibet Plateau. HDMR: Hengduan Mountain Region. EHMLY: Eastern Himalaya. TCLV: Tri-border region of China, Laos and Vietnam. WLMR: Wuling Mountain Region. TMS: Tianmu Shan. NLM: Nanling Mountain.
Areas of endemism and species richness
Prado et al.  stated that using smaller grid cells appears to be more restrictive and rigorous in identifying AoEs but it may lead to ignoring some important areas, so we identified these regions using different grid sizes and different taxonomic levels. The results revealed the most regions in the mountains of central and southern China, especially the area of QLMR-EQTP-HDMR (Fig 5) as AoEs. These regions played a significant role as centres of survival and speciation during the Quaternary global cooling , and these AoEs (Fig 5) appear consistent with those obtained from other biological assemblages in plants, birds and aphids [63, 64]. The highest species richness value within the area of EQTP reached 74. The Hengduan Mountain Region (HDMR) also recorded 60 species of Zephyrus hairstreaks. The two areas are not only core AoEs but also centres of species richness. In addition, the species compositions within the two regions appear markedly different (Table in S2 Table). EHMLY (Fig 5) is also a core AoE, independent from HDMR and EQTP, although it lacks high species richness. Eliot  deduced the greatest abundance and variety of Lycaenidae in the SE Asian Subregion of the Oriental Region, and he suggested that the SE Asian Subregion extends from SE China to Ceylon and as far east as Weber’s Line. Its characteristic species are lowland or submontane in habit and appear to be centred in Sundaland. However, centres of species richness of Zephyrus hairstreaks and core AoEs are located in the Sino-Himalayan and Sino-Japanese subregions.
General distribution and diffusion of Zephyrus hairstreaks
Almost all of the Zephyrus hairstreaks are distributed in the Northern Hemisphere and a major part of them are concentrated in Asia, except for three species in North America and two species in Europe. Our data used in the analyses showed that the northern boundary of distribution of Zephyrus hairstreaks is located along the Sea of Okhotsk in the Russian Far East in Asia, with the southern boundary located in Java in Indonesia, the western boundary is in the eastern region of Afghanistan, and the eastern boundary is in northern Japan and its surroundings, In addition, parts of northern and northwestern China are covered by large areas of desert and inaccessible areas, so there are no distributional records of Zephyrus hairstreaks in our database, along with areas of over-urbanization and areas incompletely investigated, all of which are represented as gaps (Fig 1). Nevertheless, our database maximizes the use of published available information and investigations of our team and fulfills the criteria of quality of biodiversity databases discussed by Hortal et al., . At last, there are four major distribution patterns of Zephyrus hairstreaks in Eurasia (Fig 6): 1) Sino-Japanese distribution: the species belonging to this distribution pattern are characteristic with extensions from EQTP and HDMR northward to Manchuria-Pacific (Northeast China, the Korean Peninsula, the Russian Far East and Japan), southward to South China (WLMR, TMS and NLM), and even to the west of Europe (e.g. Neozephyrus, Shirozua, Iozephyrus, Thecla and Howarthia). 2) Sino-Himalayan distribution: the species originated on the EQTP and HDMR, from where it dispersed to northwestern India along the Himalayan chain and eastward to east China, Indochina, Sundaland, Taiwan island, Japan and the west of Wallacea (e.g. Euaspa, Shirozuozephyrus, Chaetoprocta and Austrozephyrus). 3) High mountain distribution: the species are distributed only in the areas of HDMR, EQTP and QLMR (e.g. Kameiozephyrus, Saigusaozephyrus and Uedaozephyrus). 4) Combined distribution: the species exhibit a combined distribution pattern covering all three other patterns (Chrysozephyrus).
(a): Sino-Japanese distribution. (b): Sino-Himalayan distribution. (c): High-mountain distribution. (d): Combined distribution covering all three above patterns (Chrysozephyrus).
Evidence suggests that the Indo-Asian collision occurred in the early Cenozoic Era, 50–55 Myr ago or even earlier , and it almost simultaneously caused the uplift of the Himalaya-Tibetan plateau region and the Eurasian plate, which also underwent glacial/interglacial cycles during the Quaternary . Lepidoptera, as the second largest order of Insecta, diverged prior to the Cretaceous/Paleogene (K/Pg) event (65 Mya) and the extant families of Lycaenidae quite likely evolved in the Cenozoic Era . In addition, the extrusion of the Indochina block [68–70] which happened during the period of 23 to 24 Mya  was proposed to explain the formation of the boundary line called the “Tanaka-Kaiyong line” which extends from 28° N, 98° E southward to approximately 18° 45′or 19° N, 108° E [13, 14, 72, 73]. Of particular note, this biogeographical line, to some extent, exists between the core AoEs of EQTP and HDMR. This fact indicates that Zephyrus hairstreaks may have originated at least 23–24 Mya, and the Cenozoic orogeny as a result of the Indo-Asian collision likely caused rapid speciation in two AoEs, EQTP and HDMR. The significant differences between EQTP and HDMR (identified as independent AoEs respectively with lower correlation values) probably can be attributed to vicariance events which happened in these areas, for instance, the aforementioned extrusion of the Indochina block.
Areas of endemism and centres of species richness of Zephyrus hairstreaks are located in central and southern China, the Himalaya as well as northern Indo-China which is in good agreement with Quaternary vegetation reconstructions [4, 74] that had been supposed as the existence of refugia in China. Furthermore, four distribution patterns inferred two dispersal routes of endemism and diversity, which strongly suggests that Zephyrus hairstreaks probably originated in E. Asia between 25°N to 35°N in the temperate zone (based on our database). Dispersal may have played a leading role in the closer relationship of species between QLMR and EQTP, while vicariance events are the most likely cause of the significant differences between EQTP and HDMR. In addition, the Cenozoic orogeny (the uplift of the Himalaya-Tibetan plateau region and the extrusion of the Indochina block during the Indo-Asian collision) [66, 71], probably accelerated speciation, especially in the core areas endemism.
S1 Table. Distributional data of Zephyrus hairstreaks.
We would like to express our special thanks to Dr. M. Owada (National Museum of Nature and Science, Tsukuba, Japan), Mr. LP. Zhou (Shaanxi, China) and Ms. XL. Fan for providing relevant references and/or materials. We also give our appreciation to Dr. HM. Xu (Guangdong Academy of Agricultural Sciences, Guangdong, China) and Mr. WT. Wang (Agricultural Bureau of Yizhang County, Hunan, China), who contributed to this study. Special thanks go to Mr. Neil Moffat for looking over the manuscript.
- 1. Ebach MC, Goujet DF. The first biogeographical map. Journal of Biogeography. 2006; 33: 761–769.
- 2. Cracraft J. Historical biogeography and patterns of differentiation within the South America avifauna: Areas of endemism. Ornithological Monographs. 1985; 36: 49–84.
- 3. Platnick NI. On areas of endemism. Australian Systematic Botany. 1991; 4: 11–12.
- 4. López-Pujol J, Zhang FM, Sun HQ, Ying TS, Ge S. Centres of plant endemism in China: places for survival or for speciation? Journal of Biogeography. 2011; 38: 1267–1280.
- 5. Crisp MD, Laffan S, Linder HP, Monro A. Endemism in the Australian flora. Journal of Biogeography. 2001; 28: 183–198.
- 6. Linder HP. Plant diversity and endemism in sub-Saharan tropical Africa. Journal of Biogeography. 2001; 28: 169–182.
- 7. Tribsch A, Schönswetter P. Patterns of endemism and comparative phylogeography confirm palaeoenvironmental evidence for Pleistocene refugia in the Eastern Alps. Taxon. 2003; 52: 477–497.
- 8. Tribsch A. Areas of endemism of vascular plants in the Eastern Alps in relation to Pleistocene glaciation. Journal of Biogeography. 2004; 31: 747–760.
- 9. Carnaval AC, Moritz C. Historical climate modelling predicts patterns of current biodiversity in the Brazilian Atlantic Forest. Journal of Biogeography. 2008; 35: 1187–1210.
- 10. Médail F, Diadema K. Glacial refugia influence plant diversity patterns in the Mediterranean Basin. Journal of Biogeography. 2009; 36: 1333–1345.
- 11. Mackey BG, Berry SL, Brown T. Reconciling approaches to biogeographical regionalization: a systematic and generic framework examined with a case study of the Australian continent. Journal of Biogeography. 2008; 35: 213–229.
- 12. Wallace AR. “On the physical geography of the Malay Archipelago (1863)”. Alfred Russel Wallace Classic Writings. Paper 9. 2010; http://digitalcommons.wku.edu/dlps_fac_arw/9.
- 13. Tanaka T. Species problem in citrus: A critical study of wild and cultivated units of citrus based upon field studies in their native homes (Revisio Aurantiacearum IX). Japan Society for the Promotion of Science. Tokyo, Japan. 1954; 3: 141–152.
- 14. Li XW, Li J. The Tanaka-Kaiyong Line: An important floristic line for the study of the flora of East Asia. Annals Missouri Botanic Garden. 1997; 84: 888–892.
- 15. Zhu H. The floras of southern and tropical southeastern Yunnan have been shaped by divergent geological histories. PLoS ONE. 2013; 8: 1–8.
- 16. Eliot JN. The higher classification of the Lycaenidae (Lepidoptera): a tentative arrangement. Bulletin of the British Museum (Natural History), Entomology. 1973; 28: 457–465.
- 17. Saigusa T, Nakanishi A, Shima H, Yata O. Phylogeny and geographical distribution of the swallow-tail subgenus Graphium (Lepidoptera: Papilionidae). Entomologia Generalis. 1982; 8: 59–69.
- 18. Holloway JD. Macrolepidoptera diversity in the Indo-Australian tropics: geographic, biotopic and taxonomic variations. Biological Journal of the Linnean Society. 1987; 30: 325–341.
- 19. Holloway JD, Nielsen ES. Biogeography of the Lepidoptera. In: Handbook of zoology, a natural history of the phyla of the animal kingdom, Lepidoptera, Moths and Butterflies, Vol. 1: evolution, systematics, and biogeography (ed. by Kristensen N.P.). Walter de Gruyter, Berlin; 1998. pp. 423–462.
- 20. Morrone JJ. Biogeographic areas and transition zones of Latin America and the Caribbean islands based on panbiogeographic and cladistic analysis of the entomofauna. Annual Review of Entomology. 2006; 51: 467–494. pmid:16332220
- 21. Shirôzu T, Saigusa T. The “Zephyrus” hairstreaks of Japan. Lepidopterological Society of Japan; Osaka. 1980. 9 pp.
- 22. Koiwaya S. Mushi-Sha’s Iconographic Series of Insect 5: The Zephyrus hairstreaks of the world (I, II). Kyoichiro Ueda, supervisor. Fujita Hiroshi, editor. published by Mushi-Sha; 2007.
- 23. Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J. Biodiversity hotspots for conservation priorities. Nature. 2000; 403: 853–858. pmid:10706275
- 24. Shirôzu T, Yamamoto H. A generic revision and the phylogeny of the tribe Theclini (Lepidoptera: Lycaenidae). Sieboldia. 1956; 1: 329–421.
- 25. Kim S. A new species of the genus Favonius from Korea. Journal of the Lepidopterists′ Society of Korea. 2006; 16: 33–35.
- 26. Sinev SY. Catalogue of the Lepidoptera of Russia. KMK Press, St. Petersburg-Moscow; 2008.
- 27. Hasegawa T. Favonius koreanus, the twelfth species of the genus. Gekkan-Mushi, 2009; 461: 9–14 (in Japanese).
- 28. Hsu YF. A new species of Antigius (Lepidoptera: Lycaenidae: Theclini) from Taiwan. Zootaxa. 2009; 1983: 45–53.
- 29. Hsu YF. A new species of Shaanxiana (Lepidoptera: Lycaenidae) from southern China. Zootaxa. 2015; 4027: 130–134. pmid:26624170
- 30. Kim S, John S, Sohn J, Lee Y. Observation notes on Favonius koreanus from Korea. Journal of the Lepidopterists′ Society of Korea. 2009; 18: 1–6.
- 31. Wang M, Owada M. Description of a new species of the genus Ussuriana Tutt (Lepidoptera, Lycaenidae) from China. Butterflies and moths. 2009; 60: 125–127.
- 32. Singh AP. Butterflies of Kedarnath Musk Deer Reserve, Garhwal Himalaya, India. Journal of Threatened Taxa. 2009; 1: 37–48.
- 33. Shizuya H, Harada M. Early stages of Chrysozephyrus vittatus, Shirozuozephyrus khasia and Favonius letha (?) from Myanmar. Gekkan-Mushi. 2010; 473: 13–16 (in Japanese).
- 34. Katayama T, Saito K. Description of a new subspecies of Ravenna nivea (Nire, 1920) (Lepidoptera: Lycaenidae, Lycaeninae) from Central Vietnam. Butterflies (Teinopalpus). 2011; 59: 8–10.
- 35. Koiwaya S. Descriptions of nine new subspecies of the tribe Theclini (Lycaenidae) from China, Myanmar and eastern India. Gekkan-Mushi. 2011; 485: 4–8 (in Japanese).
- 36. Nakamura N, Wakahara H. Notes on the butterflies of Laos (VII): Description of a new subspecies of Ussuriana michaelis (Oberthür, 1880) (Lepidoptera: Lycaenidae, Lycaeninae) from central Laos. Butterflies (Teinopalpus). 2011; 57: 22–27.
- 37. Saito M. A new record of Shirozuozephyrus triloka (Hannyngton, 1910) from Nepal. Butterflies (Teinopalpus). 2011; 57: 57. (in Japanese).
- 38. Harada M, Yoshida Y, Ohshima Y, Wang M. Zephyrus hairstreaks observed at the Nanling area of Guangdong, China. Gekkan-Mushi. 2012; 497: 9–14 (in Japanese).
- 39. Koiwaya S, Shizuya H. A new genus and a new species of the tribe Theclini (Lycaenidae) from Myanmar. Gekkan-Mushi. 2012; 502: 28–31 (in Japanese).
- 40. Wang M, Tang DM. Butterflies of Guangxi Maoershan National Nature Reserve. Guangxi Nationalities publishing house; 2012.
- 41. Hasegawa T, Saito K. Description of new genus and new species of the subtribe Theclina (Lepidoptera: Lycaenidae) from Myanmar. Butterflies (Teinopalpus). 2013; 63: 4–8.
- 42. Koiwaya S. In search of the immature stages of Zephyrus hairstreaks from northern Yunnan. Gekkan-Mushi. 2013a; 512: 10–27 (in Japanese).
- 43. Koiwaya S. In search of the immature stages of Araragi panda. Gekkan-Mushi. 2013b; 509: 4–16 (in Japanese).
- 44. Asano S. Four Zephyrus species and rainy season butterflies from Chin State, Myanmar. Gekkan-Mushi. 2014; 521: 6–13 (in Japanese).
- 45. Xu HM, Wang M. Preliminary report on Lycaenidae species from Nanling (Lepidoptera: Lycaenidae). Science paper Online, China. 2013; 1–10 (in Chinese).
- 46. Hasegawa T, Saito K. Description of two new species of the subtribe Theclina (Lepidoptera: Lycaenidae) from Central Vietnam. Butterflies (Teinopalpus). 2014; 67: 4–11.
- 47. Koiwaya S. Descriptions of two new species and three new subspecies of Theclini (Lycaenidae) from Western and Northern China. Gekkan-Mushi. 2014; 521: 21–30 (in Japanese).
- 48. Koiwaya S, Monastyrskii AL. Description of a new species of the genus Shirozuozephyrus Koiwaya, 2007 (Lepidoptera, Lycaenidae) from Da Lat plateau C. Vietnam. Butterflies (Teinopalpus). 2010; 56: 4–8.
- 49. Yokochi T, Seki Y, Hasegawa T, Miyagawa T, Saito K, Katayama T, et al. Report of butterfly fauna in mount Hon Ba, Khang Hoa (I), Vietnam. Butterflies (Teinopalpus). 2015a; 69: 9–25.
- 50. Yokochi T, Seki Y, Hasegawa T, Miyagawa T, Saito K, Katayama T, et al. Report of butterfly fauna in mount Hon Ba, Khang Hoa (II), Vietnam. Butterflies (Teinopalpus). 2015b; 70: 17–30.
- 51. Zhuang HL, Yago M, Wang M. Theclini butterflies from Weixi, China, with description of two new species (Lepidoptera: Lycaenidae). Zootaxa. 2015; 3985: 142–150. pmid:26250028
- 52. Morrone JJ. On the identification of areas of endemism. Systematic Biology. 1994; 43: 438–441.
- 53. Hijmans RJ, Guarino L, Cruz M, Rojas E. Computer tools for spatial analysis of plant genetic resources data: 1. DIVA-GIS. Plant Genetic Resources Newsletter. 2001; 127: 15–19.
- 54. Hijmans RJ, Guarino L, Mathur P. DIVA-GIS. Version 7.5 [Software]. Manual. University of California Davis, California; 2012. http://www.diva-gis.org/
- 55. Scheldeman X, van Zonneveld M. Training manual on spatial analysis of plant diversity and distribution. Bioversity International, Rome; 2010.
- 56. Goloboff P. NDM/VNDM ver. 2.5 [Software]. Programs for identification of areas of endemism. 2005. http://www.lillo.org.ar/phylogeny/endemism/
- 57. Szumik CA, Cuezzo F, Goloboff P, Chalup AE. An optimality criterion to determine areas of endemism. Systematic Biology. 2002; 5: 806–816.
- 58. Szumik C, Goloboff PA. Areas of endemism: an improved optimality criterion. Systematic Biology. 2004; 53: 968–977. pmid:15764564
- 59. Prado JRD, Pamella GGB, Leandro PG, Gustavo SL, Edson FDA, Paulo ROR, et al. Species richness and areas of endemism of oryzomyine rodents (Cricetidae, Sigmodontinae) in South America: an NDM/VNDM approach. Journal of Biogeography. 2015; 42: 540–551.
- 60. Escalante T, Morrone JJ, Rodríguez-tapia G. Biogeographic regions of North American mammals based on endemism. Biological Journal of the Linnean Society. 2013; 110, 485–499.
- 61. Cracraft J. Patterns of diversification within continental biotas: hierarchical congruence among the areas of endemism of Australian vertebrates. Australian Systematic Botany. 1991; 4: 211–227.
- 62. Glasby CJ, Alvarez B. Distribution patterns and biogeographic analysis of Austral Polychaeta (Annelida). Journal of Biogeography. 1999; 26: 507–533.
- 63. Lei FM, Qu YH, Lu JL, Liu Y, Yin ZH. Conservation on diversity and distribution patterns of endemic birds in China. Biodiversity and Conservation. 2003; 12: 239–254.
- 64. Huang XL, Lei FM, Qiao GX. Areas of endemism and patterns of diversity for aphids of the Tibetanan Plateau and the Himalayas. Journal of Biogeography. 2008; 35: 230–240.
- 65. Hortal J, Lobo J, Jiménez-Valverde A. Limitations of biodiversity databases: case study on seed-plant diversity in Tenerife, Canary Islands. Conservation Biology. 2007; 21: 853–863. pmid:17531062
- 66. Yin A, Harrison TM. Geologic evolution of the Himalayan-Tibetan orogeny. Annual Review of Earth and Planetary Sciences. 2000; 28. 211–280.
- 67. Wahlberg N, Wheat CW, Peña C. Timing and Patterns in the Taxonomic Diversification of Lepidoptera (Butterflies and Moths). PLoS ONE. 2013; 8: 1–8.
- 68. Audley-Charles MG. Dispersal of Gondwanaland: relevance to evolution of the angiosperms. In “Biogeographical Evolution of the Malay Archipelago” (ed. by T.C. ). Clarendon Press, Oxford; 1987. pp. 5–25.
- 69. Fortey RA, Cocks LRM. Biogeography and palaeogeography of the Sibumasu terrane in the Ordovician: a review. Biogeography and Geological Evolution of SE Asia (ed. by Hall R. and Holloway J.D.). Backhuys Publishers, Leiden; 1998. pp. 43–56.
- 70. Metcalfe I. Palaeozoic and Mesozoic geological evolution of the SE Asian region: multidisciplinary constraints and implications for biogeography. Biogeography and Geological Evolution of SE Asia (ed. by Hall R. and Holloway J.D.). Backhuys Publishers, Leiden; 1998. pp. 25–41.
- 71. Che J, Zhou WW, Hu JS, Yan F, Papenfuss TJ, Wake DB, et al. Spiny frogs (Paini) illuminate the history of the Himalayan region and Southeast Asia. Proceedings of the National Academy of Sciences of the United States of America. 2010; 107: 13765–13770. pmid:20643945
- 72. Zhu H, Yan LC. Notes on the realities and significances of the “Tanaka line” and the “Ecogeographical diagonal line” in Yunnan. Advance in Earth Sciences. 2003; 18: 870–876.
- 73. Li H, He DM, Bartholomew B, Long CL. Re-examination of the biological effect of plate movement-Impact of Shan-Malay Plate displacement (the movement of Burma-Malay Geoblock) on the biota of the Gaoligong Mountains. Acta Botanica Yunnanica. 1999; 21: 407–425.
- 74. Harrison SP, Yu G, Takahara H, Prentice IC. Diversity of temperate plants in east Asia. Nature, 2001; 413: 129–130. pmid:11557970