Cryptic diversity in the subgenus Oxyphortica (Diptera, Drosophilidae, Stegana)

Phylogenetic relationships of the subgenus Oxyphortica were reconstructed based on two mitochondrial genes (COI and ND2). The results revealed the paraphyly of Oxyphortica and supported high levels of cryptic diversity within this subgenus. By integrating morphological characteristics and molecular evidence, we identified 17 new species as members of Oxyphortica: S. (O.) amphigya sp. nov., S. (O.) armillata sp. nov., S. (O.) ashima sp. nov., S. (O.) bawo sp. nov., S. (O.) crypta sp. nov., S. (O.) gelea sp. nov., S. (O.) hengduanmontana sp. nov., S. (O.) jinmingi sp. nov., S. (O.) mengbalanaxi sp. nov., S. (O.) mouig sp. nov., S. (O.) setipes sp. nov., S. (O.) shangrila sp. nov., S. (O.) tsauri sp. nov., S. (O.) valleculata sp. nov., S. (O.) wanhei sp. nov., S. (O.) yangjin sp. nov. and S. (O.) hypophaia sp. nov. To test the early morphological identifications and confirm the species boundaries, different species delimitation methods, including Automatic Barcode Gap Discovery (ABGD) and Bayesian Phylogenetics and Phylogeography (BP&P), were used, together with traditional distance. All species boundaries were clearly defined. As Oxyphortica species are mainly distributed across Southwest China (e.g., 20 spp. from the Hengduan Mountains), the complex climate and topographic landforms of the area may be responsible for the high levels of species diversity and endemism.

The subgenus Oxyphortica encompasses 38 species (Bächli, 2021), which mainly live near streams and feed on tree trunk sap, moss, or fungi. These species mostly inhabit the tropical to subtropical broad-leaved forests of the Oriental Region. Nevertheless, the habitats of some species extend beyond these areas-the habitats of S. (O.) nigripennis (Hendel, 1914) and S. (O.) dendrobium Chen & Aotsuka, 2004 extend northward into the southern part of the Palaearctic Region (Kyushu, Japan) and the habitat of S. (O.) convergens (de Meijere, 1911) extends southward into the northern part of the Australian Region (New Guinea) (Brake & Bächli, 2008;Wang et al., 2017;Huang et al., 2018). In this context, Southwest China encompasses a centre of diversity of Oxyphortica (55.3% of the known species of this subgenus occur in this area, and 39.5% of them are endemic).
Accurate species identification is essential for a comprehensive phylogenetic assessment. However, complex evolutionary history often limits the inference of reliable phylogenies and the delimitation of species boundaries, especially in rapidly diversified taxa. Complicating factors include morphologically cryptic species (Low et al., 2016;Machado et al., 2017), horizontal gene transfer (Hurst & Jiggins, 2005), incomplete lineage sorting (Yeates et al., 2011) and introgression caused by multiple hybridisation events (Fougère-Danezan et al., 2015). Nevertheless, compared with traditional taxonomy, the application of new molecular analytical approaches has made unveiling cryptic species diversity easier and faster. In fact, molecular data recently revealed high levels of cryptic diversity of Oxyphortica along the Hengduan Mountains and its adjacent regions (Wang et al., 2017;Huang et al., 2018), where the complex montane topography may have promoted and maintained species differentiation.
During recent field trips, a large number of Oxyphortica species with similar morphological characteristics were collected from Southwest China. Several of these species vary among different mountain gullies, even if the gullies are close. Considering the limited molecular data analysis and unclear taxonomic status of numerous species, in this study, we aim to assess the boundaries and taxonomic status of 17 new species identified by morphological characteristics and molecular data and briefly discuss the cryptic diversity of Oxyphortica in Southwest China.

Sampling and morphological evaluations
Specimens were collected by net sweeping from tussocks or tree trunks along streams in forests. Following collection, the specimens were immediately preserved in 75% ethanol. In the laboratory, all samples collected from the field were initially identified based on morphological characteristics, and then targeted specimens were selected for thorough morphologilical examination. Detailed information is listed in Table 1. Genitalia of the specimens were removed for further identification and photographed using a microscope Mshot Camera (Mshot, China). A small piece of tissue was excised for DNA extraction using a TIANGEN TM DNA Extraction Kit (Tiangen Biotech, Beijing, China). Finally, using a combination of morphological characteristics and molecular data, new species were recognised. Type specimens were air-dried and deposited in the Kunming Institute of Zoology, Chinese Academy of Sciences, Kunming, China (KIZ); the Department of Zoology, Tibet Museum of Natural Science, Lasa, China (TMNS); and the Department of Entomology, South China Agricultural University, Guangzhou, China (SCAU). To unify Drosophilidae terminology and facilitate exchanges between research fields, the male terminalia terminology followed that of Rice et al. (2019).
The electronic version of this article in portable document format will represent a published work according to the International Commission on Zoological Nomenclature (ICZN), and hence the new names contained in the electronic version are effectively published under that Code from the electronic edition alone. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. The ZooBank LSIDs (Life Science Identifiers) can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix http://zoobank.org/. The LSID for this publication is: urn:lsid:zoobank.org: pub:1D19FB6F-DBF2-4982-9C9A-D8FBFC1F3F1E. The online version of this work is archived and available from the following digital repositories: PeerJ, PubMed Central and CLOCKSS.

PCR amplification and sequencing
Two mitochondrial (cytochrome c oxidase subunit I (COI) and NADH dehydrogenase subunit 2 (ND2)) markers were employed for PCR amplification, which involved the following steps: initial pre-denaturation for 3 min at 94 C, followed by 35 cycles of 30 s at 94 C, 1 min at 49-56 C and 1 min at 72 C, and subsequent post-extension for 5 min at 72 C. The amplification and sequencing primers are listed in Table 2. Sequencing of the PCR products was performed using an ABI 3700 sequencer; a standard sequencing reaction using a BigDye TM Terminator Kit (Perkin-Elmer, Waltham, MA, USA) was performed with 25 cycles of denaturation for 10 s at 96 C, annealing for 5 s at 50 C and extension for 4 min at 60 C, subsequently followed by purification with 75% isopropyl alcohol. The sequence accession numbers are listed in Table 1.

Phylogenetic analyses
The mitochondrial sequences were aligned separately for each gene using the programme MAFFT v. 7 (http://mafft.cbrc.jp/alignment/server/) and then concatenated (COI + ND2). A total of six data blocks of the concatenated data (i.e., first-, second-and third-codon positions of two mitochondrial genes) were subjected to an evaluation of the best partitioning scheme and substitution models (GTR+I+G: COI 1st + COI 2nd + ND2 1st + ND2 2nd ; TIM+I+G: COI 3rd + ND2 3rd ) for phylogenetic analysis in PartitionFinder 2 (Lanfear et al., 2017) using the ′greedy′ algorithm and the Bayesian information criterion.      Miller, Pfeiffer & Schwartz, 2010). Two independent runs were implemented in parallel, each with 20,000,000 generations. A sampling frequency of every 1,000 generations was employed, and 5,000 early-phase samples were discarded as burn-in. When the average standard deviation of split frequencies was <0.01, the analysis was considered convergent. The ML analysis was performed using IQ-Tree (Nguyen et al., 2015) under the aforementioned substitution models. Node support was assessed via ultrafast bootstrap (UFBP) analysis with 1,000 replicates. MEGA 5 was used to estimate pairwise genetic distances (p-distance) (Tamura et al., 2011).

RESULTS
Both phylogenetic analyses and the species delimitations were performed using 138 COI and 137 ND2 sequences. Within this dataset, 620 nucleotide sites of COI sequences contained 366 conserved, 254 variable and 234 parsimony-informative sites, whereas 1,034 nucleotide sites of ND2 sequences contained 411 conserved, 620 variable and 577 parsimony-informative sites. All new sequences characterised in this study were uploaded to GenBank. The accession numbers are listed in Table 1.
The topology of the phylogenetic tree varied among the different datasets and methods (Figs 1, S1-S5). The concatenated data (COI + ND2) analyses supported the reciprocal monophyly of all species, whereas three subgenera (Stegana, Steganina and Orthostegana) was found to be nested in the subgenus Oxyphortica, rendering the latter paraphyletic. The Oxyphortica is mainly composed of four clades (Fig. 1). Clade I comprises 12 known species located at the base of the phylogenetic trees (PP = 1.00; UFBP = 100).  (Table S1), consistent with the species definition based on morphological characteristics.
Description. Arista with four to six dorsal and three to five ventral branches. Clypeus yellowish brown. Mesoscutum and scutellum yellow to yellowish brown. Katepisternum brown above, yellow below. Abdominal first tergite yellow; second to sixth tergites mostly dark brown, yellow medially, narrowly yellow bands on anterior margins on fourth and fifth tergites (Fig. 5D). Sternites mostly brown, yellow on second. Male terminalia (Fig. 6): Epandrium with ca. 20 setae on each posterior portion. Surstylus with a row of ca. 11 setae. Pregonites small and inconspicuous. Aedeagus with pubescence distally. Measurements.   Diagnosis. This species is very similar to S. (O.) amphigya sp. nov. in male terminalia, but can be distinguished by the foretarsi second to fourth subapically lacking small, suberect, fringe-like seta on anterior surface (Fig. 3E); surstylus round ventrally, with seven small prensisetae on dorstal 3/4 (Fig. 7B). Description. Arista with four to six dorsal and three to four ventral branches. Clypeus brown. Mesoscutum and scutellum yellow to brownish yellow. Katepisternum brownish above, yellow below. Abdominal first tergite yellow; second and third tergites yellow medially, brown laterally; fourth to sixth tergites mostly brown, yellow on anterior margins (Fig. 5E). Sternites yellow, with narrowly brown band on lateral margins of fourth to sixth. Male terminalia (Fig. 7): Epandrium with ca. 21 setae on each posterior portion. Surstylus with a row of ca. 12 setae. Aedeagal sheaths submedially with two pairs of prensisetae and five pairs of setae, elongated to semiring-shaped, distally with one small prensisetae and a few setae on each side. Pregonites small, inconspicuous. Aedeagus with pubescence distally.  annularly fused and deeply concave posteromedially (Fig. 8C, phap); aedeagus with a row of setae on apical margin (Fig. 8C).
Description. Arista with seven to eight dorsal and four ventral branches. Clypeus dark brown. Mesoscutum mostly yellow, sometimes light brown laterally. Katepisternum brown above, yellow below. Scutellum brownish yellow. Foreleg first tarsus with ca. seven fringe-like setae; first to fourth tarsi with ca. 11 setae and subapically each with one valgus fringe-like setae on anterior surface (Fig. 3G). Abdominal first and second tergites mostly yellow, dark brown on second tergite posterior margin; third yellow medially, dark brown laterally; fourth to sixth dark brown to black (Fig. 5G). Sternites yellow on first to third, dark brown to black on fourth to sixth. Male terminalia (Fig. 9) Etymology. The name denotes "dauntless" in Tibetan.  Diagnosis. This species differs from other species in having the surstylus very narrow, with two small prensisetae dorsally (Fig. 10C); aedeagus with numerous serrated processes on lateral margins (Figs. 10C, 10D).
Description. Arista with six dorsal and four ventral branches. Clypeus brown. Mesoscutum yellow, slightly brownish laterally and posteriorly. Scutellum brownish yellow. Katepisternum entirely yellow. Foretarsi second to fourth each with one small fringe-like seta on anterior surface (Fig. 3H). Abdominal tergites all yellow medially, brown laterally (Fig. 5H) Diagnosis. This species is very similar to S. (O.) wanhei sp. nov. in the male terminalia structure, but can be distinguished from the latter by the apical processes of aedeagus and aedeagal sheaths are nearly equilong (Figs. 11C, 11D), foreleg first tarsus subapically to apically with three recurvate fringe-like setae on anterior surface (Fig. 3I).
Description. Arista with six dorsal and four ventral branches. Clypeus brown. Mesoscutum and scutellum brownish yellow. Katepisternum brown anteriorly, yellow posteriorly. Abdominal first tergite yellow; second and third tergites yellow medially, brown laterally; fourth to sixth mostly brown, with yellow anteromedially (Fig. 5J). Sternites yellow. Male terminalia (Fig. 12): Epandrium with two irregular rows of ca. 20 setae along each posterior margin. Cercus with pubescence. Surstylus with two stout prensisetae dorsally, pubescence medially, and ca. five setae ventrally. Pregonites present. Aedeagus apically bifurcated and curved. Aedeagal sheaths anterolaterally elongated, curved and fused with phallapodeme, posteriorly protruded. Diagnosis. This species is very similar to S. (O.) wanhei sp. nov. in the male terminalia, but can be distinguished from the latter by the apical processes of aedeagal sheaths slightly curved in lateral view (Fig. 13D); foretarsi first and second with two rows of ca. 10 long, recurvate fringe-like setae; first to fourth subapically each with one valgus fringe-like seta on anterior surface (Fig. 4A).
Etymology. Patronym after the collector Mr. Jinming Lu (SCAU). Diagnosis. This species somewhat resembles S. (O.) yangjin sp. nov. in the male terminalia (Figs. 14, 21), but can be distinguished from the latter by the foretarsi first and second with two to three rows of ca. 22 long, recurvate fringe-like setae; first to fourth apically each with one short, valgus fringe-like seta on anterior surface (Fig. 4B).
Etymology. The name means "dreamlike homeland", from the language of the Dai ethnic group in Xishuangbanna, Yunnan, China.
Description. Arista with four to six dorsal and three to four ventral branches. Clypeus brown. Mesoscutum yellow, with brown longitudinal stripes. Katepisternum yellow. Scutellum brown, yellow along margin. Foretarsi first to fourth with a row of ca. 13 long, recurvate fringe-like setae, and subapically each with one short, valgus fringe-like seta on anterior surface (Fig. 4C). Abdominal first tergite yellow; second tergite yellow medially, dark brown laterally; third tergite mostly dark brown, with one pair of yellow patches medially; fourth to sixth dark brown (Fig. 5M). Sternites yellow. Male terminalia (Fig. 15): Epandrium with ca. 19 setae on posterior portion per side. Surstylus with numerous setae submedially and ventrally. Aedeagal sheath slender.
Etymology. The name means "the Almighty" in the language of the Va ethnic group in Yunnan.
Stegana ( Diagnosis. This species somewhat resembles S. (O.) tsauri sp. nov. in the male terminalia, but can be distinguished from the latter by the foretarsi first and second with two rows of ca. 30 long, recurvate fringe-like setae on anterior surface (Fig. 4D); abdominal tergites yellow medially, light brown laterally (Fig. 5N).
Description. Arista with seven to eight dorsal and three to five ventral branches. Clypeus brown. Mesoscutum, katepisternum and scutellum yellow. Abdominal first to fourth tergites yellow medially, brown to dark brown laterally; fifth and sixth dark brown, with yellow anteromedially (Fig. 5O). Sternites entirely yellow. Male terminalia (Fig. 17 Figs. 19C,19D), but can be distinguished from the latter by the foretarsi first to third with two rows of ca. 27 long, slightly curved fringe-like setae on anterior surface (Fig. 4G), and surstylus with two stout prensisetae dorsally (Fig. 19B).
Description. Arista with five to six dorsal and three to four ventral branches. Clypeus brown. Mesoscutum and scutellum yellow. Katepisternum yellowish. Abdominal first tergite yellow; second and third tergites yellow medially, brown laterally; fourth and fifth yellow medially and brown along posterior margins and laterally in male, nearly brown to dark in female; sixth brown to dark brown (Fig. 5Q) (Fig. 20D), abdominal fifth and sixth tergites yellow medially, narrowly black along lateral margin (Fig. 5R), and foretarsi first subapically with two rows of ca. 10 recurvate fringe-like setae, second with two short fringe-like setae on anterior surface (Fig. 4H).
Description. Arista with seven to nine dorsal and four to five ventral branches. Clypeus brown. Mesoscutum and scutellum brownish yellow. Katepisternum brown anteriorly, yellow posteriorly. Abdominal first tergite yellow; second to fourth tergites yellow medially, black laterally (Fig. 5R) Diagnosis. This species differs from S. (O.) mengbalanaxi sp. nov. in having the foretarsi first and second with two rows of ca. 12 long, slightly recurvate fringe-like setae on anterior surface (Fig. 4I), and abdominal tergites brown on first to third, dark brown on fourth to sixth (Fig. 5S).
Description. Arista with five dorsal and four ventral branches. Clypeus brownish yellow. Mesoscutum and scutellum yellow. Katepisternum yellow, slightly brown anteriorly.

DISCUSSION
In contrast to the claim that all supraspecific units should be monophyletic (Komarek & Beutel, 2006), the phylogenetic analyses conducted herein recovered the paraphyly of the subgenus Oxyphortica (Fig. 1), as described by Li et al. (2013). Within this context, the nigripennis and convergens species groups were defined as monophyletic in both concatenated data and each gene analyses (Figs 1, S1-S5), whereas Clades I and IV were not in the COI analysis (Figs S2, S3). Therefore, a combined dataset of mitochondrial and nuclear markers may help to obtain a reliable phylogenetic framework and plausibly reflect the subdivision of groups within this subgenus. Moreover, this study highlights the existence of high levels of cryptic diversity in the subgenus Oxyphortica, which is concordant with the ever-growing number of species. The diversity of this subgenus is highly concentrated around Southwest China, highlighting the importance of climatic and topographical temporal and spatial differences for species diversification.
Our results indicate that genetic divergence has not been accompanied by appreciable morphological changes. This occurs, for example, in the allopatric species pairs in Clade I, S. (O.) chuanjiangi Zhang & Chen, 2017 (southwestern Yunnan) and S. (O.) dawa Zhang & Chen, 2017 (Beibeng, Motuo in Xizang), which are distinguished by the aedeagus that is wider than it is long and not slightly bifid apically (Figs. 8D, 10D in Wang et al., 2017). This situation contrasts with the striking diversification in genitalia that occurs in other allopatric drosophilids (e.g., the sibling species Amiota cuii Chen & Toda, 2001and A. nozawai Chen & Watabe, 2005, Pseudostegana bifasciata Chen & Wang, 2005 and Ps. acutifoliolata Li, Gao & Chen, 2010), suggesting recent divergence, with insufficient time to accumulate substantial morphological differentiation (Struck et al., 2018;Fišer, Robinson & Malard, 2018). In fact, it is generally assumed that, in the early stages of speciation, natural selection primarily acts on traits that are closely associated with survival, such as physiological, immunological, or behavioural traits, rather than on morphology (Struck et al., 2018), particularly in extreme environments (Xu & Che, 2019).
Among the newly recognised species, significantly similar male genitalia structures are shared by the species of Clades IV-A, -B and -C (Fig. 1) ) zhulinae spans about 300 km in the north-south direction (Xincheng, Yingjiang to Nanling, Lancang in Yunnan). These results further illustrate the rich species diversity of Oxyphortica that may occur even in narrow geographical areas.
Accurate species identification is a primary prerequisite for biological research. However, incongruence among gene trees and substitution patterns often occurs. The ABGD analyses suggested two MOTUs for S. are endemic. This high species diversity and rich endemism are largely attributed to the complex climate and the north-south longitudinal mountain and canyon landforms in the Hengduan Mountains, known as the 'folded zone'. Positioned in the northern part of the Indo-Chinese Peninsula and in the eastern border of the Qinghai-Tibet Plateau, Yunnan is greatly influenced by the tropical and East Asian monsoons and by atmospheric circulation formed by the Himalayan and Qinghai-Tibetan climate. This creates great climatic diversity, including a tropical zone (river valley), a subtropical zone (mid-lying mountains), a temperate zone (sub-high mountains), a frigid zone and an alpine desert zone (everlasting snow zone at the highest point of the mountains) (Ji et al., 1999). The north-south longitudinal mountains facilitate the southward movement of northern species, whereas the longitudinal river valleys at low elevations stimulate the northward diffusion of southern species. Together, these conditions indicate that the Hengduan Mountains are a centre of species aggregation and differentiation.
Previous dating studies have shown that the origin of the subgenus Oxyphortica dates back to the late Oligocene to early Miocene (Li et al., 2013). The Miocene and subsequent Pliocene are known for rapid speciation and adaptive radiation of different taxa in Southwest China (Wen et al., 2014;Favre et al., 2015;Päckert et al., 2015;Lu et al., 2018;Ding et al., 2020), which resulted from the extensive uplift of both the Qinghai-Tibet Plateau and Hengduan Mountains (Wang et al., 2012). This rapid orogenic movement may have provided numerous ecological opportunities for the increase of parapatric speciation via niche differentiation without competitors (Moritz et al., 2000;Gavrilets & Vose, 2005;Favé et al., 2015;Hazzi et al., 2018). The high endemism and the short internode divergence suggest that Oxyphortica may have experienced rapid adaptive radiation in Southwest China, where 16 new species were recorded. This study underlines the huge biodiversity of this region and how poorly it is known.