﻿Revisiting the type species of the genus Homidia (Collembola, Entomobryidae)

﻿Abstract Homidiacingula Börner, 1906, the type species of the genus Homidia Börner, 1906, is widespread from India to Southeast Asia, but its detailed morphological characteristics have not yet been described. We examined the morphology of specimens of H.cingula from Indonesia and southwestern China and confirmed their conspecific status by comparing their DNA barcoding sequences. We also compared the morphology of H.cingula with other two closely related species, confirming the valid species status of H.subcingula Denis, 1948. Our study provides new taxonomic and molecular data for the genus Homidia.


Introduction
The genus Homidia Börner, 1906 (Collembola, Entomobryidae, Entomobryinae) was first described as a subgenus of Entomobrya Rondani, 1861 (Börner 1906).It was later recognized as a distinct genus by Denis (1929).Homidia, with 76 reported species, is widely distributed in the Northern Hemisphere, particularly in East and Southeast Asia (Bellinger et al. 1996(Bellinger et al. -2023)).Homidia can be easily distinguished from other genera of Entomobryinae by several key characteristics, including 8+8 eyes, dental spines in adults, a subapical mucronal tooth larger than the apical one, and eyebrow-like macrochaetae on the anterior part of the fourth abdominal segment (Börner 1906;Szeptycki 1973).
The type species of the genus, Homidia cingula Börner, 1906, was originally described from Java (Buiterizor) and has since been recorded in India, Bangladesh, Thailand, Malaya, Singapore, Sumatra, and Vietnam (Yoshii 1989).Homidia cingula is characterized by its distinctive colour pattern, which includes pigmented abdominal segments II and III.This colour pattern is also seen in three other related species: Entomobrya kali Imms, 1912 from India, Homidia subcingula Denis, 1948 from Vietnam, and Homidia glassa Nguyen, 2001 also from Vietnam.Handschin (1925) initially questioned whether H. kali was a synonym of H. cingula, and this was later confirmed by Mitra (1976) who re-examined the syntypes of E. kali.However, these species have notable differences in colour pattern based on their original descriptions (Table 1): thoracic patches and broad band on Abd.IV posteriorly in H. cingula; narrow band on Abd.IV posteriorly and transverse band on Abd.V in H. subcingula; and a pair of small metathoracic patches in H. glassa.Although Denis (1929) noted differences between H. cingula and H. subcingula (such as a broad vs narrow stripe on the posterior part of the fourth abdominal segment), Mitra (1976) suggested that H. cingula may represent juveniles and H. subcingula the darker form of adults.Nguyen (2001) only identified one difference between H. subcingula and H. glassa: two small patches on the metathorax in the latter.Unfortunately, the type material of H. cingula was destroyed (Weidner 1962).Further examination of the type species is essential to resolve the taxonomic uncertainties surrounding these species and to improve our understanding of the genus Homidia.
This study focuses on specimens of H. cingula collected from Indonesia (Java, Sulawesi) and China, as well as the types of H. subcingula.We also employ molecular barcoding techniques to obtain genetic sequences for H. cingula specimens from Java and China and compare their genetic distances.A detailed description of H. cingula is provided.

Morphological examination
Juvenile and adult specimens were cleared in lactic acid, mounted in Marc André II solution, and studied using Leica DMLB and Nikon 80i microscopes.Illustrations were enhanced in Adobe Photoshop CS5.Dorsal body chaetae nomenclature follows Szeptycki (1979), Zhang and Deharveng (2015), and Zhang et al. (2019), labial palp nomenclature follows Fjellberg (1999), and labial chaetae nomenclature follows Gisin (1967).The dorsal chaetotaxy is given per half-tergite in the descriptions; the solid and hollow circles represent the primary and secondary chaetae, respectively.

DNA barcoding
DNA was extracted using an Ezup Column Animal Genomic DNA Purification Kit (Sangon Biotech, Shanghai, China) following the manufacturer's standard protocols.Primers used were LCO1490/HCO2198, which are commonly used for metazoans (Folmer et al. 1994) Antenna 2.5-3.0 times as long as cephalic diagonal.Antennal segments ratio as I: II: III: IV = 1: 1.2-1.3:1.2-1.3:1.6-1.9.Smooth straight mic at antennal base three dorsal and three ventral on Ant.I, one external, one internal and one ventral on Ant.II and absent on Ant.III and IV.Ant.III organ with two rod-like sensilla (Fig. 3).Ant IV with apical bulb bilobed (Fig. 4).

22
(1976) considered them as synonyms and doubtfully thought that Börner's H. cingula represented juveniles of H. subcingula.However, the re-examination of the type specimen (2.4 mm) of H. subcingula in MNHN by the senior author (F.Zhang) shows that both taxa are valid species, although the type of H. cingula is in poor condition and many details are impossible to visualize.Homidia subcingula differs from the H. cingula in the narrow strip on posterior Abd.IV, Abd.V pigmented, 11+11 mac on Abd.I, and 6+6 (m3ea present) central mac on Abd.II.In addition, the validation of H. glassa, which also possesses pigmented Abd.II-III, is questionable.Chaetotaxy of Abd.I-III of H. glassa could be closer to H. subcingula, but this differs from that of H. cingula (Table 1).

Molecular results
Our results show that pairwise genetic distances range from 0.03 to 0.212 among 10 Homidia species (Table 2).The genetic distance between specimens of H. cingula from Yunnan (China) and Java is 0.03 (Table 2).The small genetic divergence (Hebert et al. 2003) indicates that these specimens belong to the same species (Fig. 23).

Discussion
Both morphological and molecular analyses confirm that the Homidia specimens from Indonesia and southwestern China, collected ca 3,000 km apart, are the same species.Species possessing distinct colour pattern (Abd.II-III and posterior half of Abd.IV pigmented) are widely distributed in Southeast and South Asia.Genetic divergence of the individuals from the most southern and the most northern regions is very low (ca 3%).Their colouration and wide distribution perfectly match the original descriptions and subsequent records of H. cingula.Therefore, we consider the species examined in this study to be H. cingula, although the type material described by Börner has been destroyed (Weidner 1962).Re-examination of types of H. subcingula verifies its validity based on colour pattern and chaetotaxy of Abd.IV (
Homidia cingula is characterized by dark transverse bands on Abd.II-III, 5+5 central mac on Abd.II, and 1+1 central mac on Abd.III.It has similar colour pattern to H. subcingula from Dalat, Vietnam (dark Abd.II and III).Mitra

Figure 23 .
Figure 23.Neighbour-joining tree based on mtDNA COI sequences.Node numbers are bootstrap values.

Table 1 .
Morphological comparison of three Homidia species.
(Zhang et al. 2019) body lengths of H. cingula and H. subcingula were approximately equal, thus disproving the hypothesis that H. cingula represents juveniles of H. subcingula.We doubt the validity of H. glassa, whose characteristics is very similar to H. subcingula except for its rough description of colouration.Applying colouration to distinguish Homidia species remains a powerful tool in modern taxonomy.Following this rule, the photographs of "H.cingula" from Taiwan (photographer: H.-J. Cheng) seem different from our understanding of H. cingula: both anterior and lateral margins are pigmented (anterior part pale in H. cingula), and the posterior half patch of Abd.IV is divided into two parts (connected in H. cingula).A molecular comparison could easily resolve this problem.Combining with the first instar chaetotaxy(Zhang et al. 2019), our revisiting of H. cingula provides valuable information for the diagnoses of the genus Homidia.Further collections of H. cingula from type locality (Buitenzorg, Indonesia) could be conducted to assign the neotype material.