Nesticus kosodensis Yaginuma, 1972 bona species. Molecular and morphological separation from N. latiscapus Yaginuma, 1972 with notes on cave scaffold-web spiders subspecies in Japan (Araneae, Nesticidae)

The troglophilic spider subspecies Nesticus latiscapus kosodensis Yaginuma, 1972 is compared with its nominal species N. latiscapus Yaginuma, 1972 using an integrated taxonomic approach. Both morphological and molecular analysis suggest the two taxa are distinct, rejecting the status of subspecies. In the light of these results, N. kosodensis stat. nov. is elevated to species level and both species are redescribed. We expect that future revisions of Japanese Nesticidae will find that many currently designated subspecies are in fact distinct species. Taxonomic revision of long-established and unrevised subspecies may help to improve the conservation efforts aiming to protect the Japanese endemic cave fauna.


Introduction
Defining the boundaries among species is a key factor in modern systematics (Wiens 2007). Due to its arbitrariness across different taxonomic groups, the correct separation between geographic forms in polytypic species (= subspecies) from closely related but distinct congeners may be tricky. Although the use of subspecies is accepted by the International Code of Zoological Nomenclature (ICZN 2012), their use in spider taxonomy has decreased substantially during the last century; they are often limited to old descriptions from the first half of the 20 th century or earlier (Kraus 2000;Nentwig et al. 2019). Revisiting established spider subspecies using a more modern approach is necessary to confirm their taxonomic status as local variants of the same species or, alternatively, to include them in a different taxonomic rank (e.g. a separate species).
Nesticidae Simon,1894 is a family of spiders (16 genera and 278 species, World Spider Catalog 2020) with a worldwide distribution. Most nesticids inhabit caves or subterranean environments, but some are found in other places characterized by relatively uniform conditions of temperature and humidity along the year (e.g. deep cracks in rocks, humid and shadowed valleys, tropical forest litter, etc.). In Japan the nesticid fauna appears to be particularly diverse, with five genera and 59 recorded species and subspecies, the wide majority of which (54 taxa) are included in the genera Nesticus Thorell, 1869and Cyclocarcina Komatsu, 1942(Tanikawa 2020. Most of the Japanese nesticid species show high levels of endemism being often restricted to one of few caves and consequently can be considered good candidates for species conservation. Nevertheless, among the Japanese Nesticus and Cyclocarcina species 10 of them (~20% of the total) are currently considered subspecies, all of which were established in the 1970s by Prof. Takeo Yaginuma: Cyclocarcina floronoides komatsui Yaginuma 1979; C. floronoides notoi Yaginuma 1979;C. floronoides tatoro Yaginuma 1979; Nesticus akiyoshiensis ofuku Yaginuma 1977;N. latiscapus kosodensis Yaginuma 1972;N. longiscapus awa Yaginuma 1978;N. longiscapus draco Yaginuma 1978;N. longiscapus kiuchii Yaginuma 1978;N. tosa iwaya Yaginuma 1976;and N. tosa niyodo Yaginuma 1976. The validity of these interspecific names has not yet been tested using more recent taxonomic techniques.
During field trips in central Japan (Yamanashi and Tokyo Prefectures), we collected several fresh specimens of the subspecies Nesticus latiscapus kosodensis Yaginuma, 1972 andits nominal species N. latiscapus Yaginuma, 1972. These spiders are found in caves and similar environments within a relatively restricted area in central Japan. Due to their presumed morphological similarities and close geographic distributions, N. latiscapus kosodensis has long been considered a variant of N. latiscapus. The aim of this study is to compare the morphology and genetic distances of these two taxonomic units, and test the concept of subspecies among Japanese cave spiders.

Materials and methods
Collecting sites were centred in Yamanashi and Tokyo Prefectures, central Japan. Spider samples were predominantly collected from caves, but some were found under rocks.
Specimens were preserved in 95% ethanol for molecular and morphological analysis. Specimen morphology was examined using a Nikon SMZ1270 stereomicroscope (for N. latiscapus) or a Nikon AZ100M stereomicroscope (for N. kosodensis stat. nov.), and photographed with a Canon EOS Kiss x8i digital camera attachment. Photos were combined using the image stacking software Helicon Focus 6.7.1. and further processed with Adobe Photoshop CC 2015. Epigynes were dissected using a sharp scalpel, and cleared by boiling them in a 20% KOH solution for several minutes until clear. Leg measurements are reported in the text as total length (femur, patella, tibia, metatarsus, tarsus). All measurements are in millimeters. The following abbreviations are used in text or figures: Be = bulge of embolus; Co = copulatory opening; Dp1-3 = dorsal processes of paracymbium; Di = distal process of paracymbium; E = embolus; Id + Fd = insemination and fertilization ducts; Ma = median apophysis; P = paracymbium; Pc1-3 = processes of conductor complex; S = spermatheca; Sc = scapus; St = subtegulum; T = tegulum; Ta = tegular apophysis; Vp = vulval pocket; Vpp = ventral process of paracymbium. Specimens used in this study are stored in the following collections: FBPC = F. Ballarin personal collection; YSPC = Y. Suzuki personal collection; NSMT = National Museum of Nature and Science, Tokyo, Japan; TSPC: T. Suguro personal collection.
In order to evaluate the genetic distance between the taxa discussed in this study and other close related congeners from the same geographical area, a total of 13 Nesticus specimens belonging to four species were included in the analysis: N. kosodensis stat. nov. (4 specimens), N. latiscapus (7 specimens), N. shinkaii Yaginuma, 1979 andN. gondai Yaginuma, 1979 (1 specimen each). The latter two species were preferentially selected on the base of their geographic distribution and results of our preliminary studies on the phylogeny of the genus Nesticus. The phylogenetic tree was rooted using the species Cyclocarcina floronoides tatoro (Yaginuma 1979), since our preliminary results also suggest a close phylogenetic relationship between Cyclocarcina and the Japanese components of the genus Nesticus. The final alignment was 663 nucleotides long.

Phylogenetic analysis
Maximum likelihood analysis (Fig. 9) confirms a close phylogenetic relationship between N. latiscapus and N. kosodensis stat. nov., with these taxa belonging to two sister clades. However, the two clades have relatively long branches, indicating a rather high difference in the number of substitutions per site. Such genetic distance is even higher than the distance among other valid species of the same genus included in the analysis, thus rejecting the hypothesis of N. latiscapus and N. kosodensis stat. nov. being conspecifics. The pairwise distance analysis (Table 1) further supports this clear separation, reporting a genetic distance of ~10% from each other, and between them and other congeners.    Fig. 3a, b vs. 6a, b); Ta weakly developed but present (lacking in N. latiscapus, see Fig. 3a, e vs. 6a, e). Females N. kosodensis stat. nov. are distinguished from females N. latiscapus by the different shape of scapus, narrower towards the tip and with convergent edges in N. kosodensis stat. nov., almost of the same width or slightly wider towards the tip, with slightly divergent edges and lobate corners in N. latiscapus (Fig. 4 vs. Fig. 7). The shape of copulatory openings is also different between the two species: the edge of the opening is curved and slanted compared to the edge of the scapus in N. kosodensis stat. nov., but straight and parallel to the edge of the scapus in N. latiscapus (Fig. 4 vs. Fig. 7).
Habitus as in Fig. 1a, b. Carapace light yellowish brown, midline and lateral edges darker. Chelicera, labium and sternum light yellowish-brown without any markings. Legs yellowish-brown with dark bands. Leg measurements as follows (based on one male): I 15.60 (4.34, 0.83, 4.62, 4.21, 1.61), II 12.10 (3.48, 0.77, 3.32, 3.18, 1.34), Palp as in Figs 2, 3, 8a. Paracymbium with five processes: ventral process recurved, dorsal process 1 sharp and weakly curved when the palp is observed ventrally and laterally, dorsal process 2 short and straight, dorsal process 3 very small and close to the basal part of dorsal process 2, hidden behind ventral process when the palp is observed ventrally. All dorsal processes strongly sclerotized. Distal process of paracymbium sharp with a smooth edge. Embolus with a well-developed bulge, median apophysis weakly protruded, tegular apophysis small and triangular.
Epigyne and vulva as in Figs 4, 8b. Scapus prolonged posteriorly, narrower towards the tip, with slightly convergent edges and bearing a U-shaped groove. Edge of copulatory openings rounded and slanted compared to the edge of the scapus. Insemination and fertilization ducts reaching spermathecae with a short and convoluted trait. Spermathecae oval. Vulval pockets strongly wrinkled, located above spermathecae.

Nesticus latiscapus
Palp as in Figs 5, 6, 8c. Paracymbium with five processes: ventral process recurved, narrower at the basal part, dorsal process 1 sharp, weakly curved when observed lat-** For the complete list of references see the World Spider Catalog 2020. erally and ventrally, dorsal process 2 short and straight, dorsal process 3 small, distal process wing-shaped, with a serrated edge. Bulge of embolus absent, median apophysis lobate, strongly protruded prolaterally, tegular apophysis absent.
Epigyne and vulva as in Figs 7, 8d. Scapus prolonged posteriorly, duck beakshaped, almost the same width or slightly wider towards the tip, ending with slightly divergent edges and lobate corners. Edge of copulatory openings straight, parallel towards the edges of the scapus. Insemination and fertilization ducts reaching spermathecae with a short and convoluted trait. Spermathecae oval. Vulval pockets wrinkled, located above spermathecae.
Distribution. Endemic to the Fuji Mountain area in central Honshu, Japan (Yamanashi Prefecture). See Fig. 10.

Remarks on autoecology of Nesticus kosodensis stat. nov. and N. latiscapus
Although lacking deep morphological adaptations to a strict hypogean lifestyle, both N. kosodensis stat. nov. and N. latiscapus share a clear preference for troglophilic habitats, being mainly found in the twilight zone of caves or other similar environments. N. latiscapus inhabits the lava tubes located in the Aokigahara area and other surroundings northwest of Mount Fuji, thus apparently preferring this typology of basaltic superficial caves rather than the more common limestone caves. N. kosondensis stat. nov. seems to dwell in a wider range of subterranean or semi-subterranean habitats along the southern area of Okuchichibu Mountains Region, from large limestone caves to deep rocks cracks and large, empty spaces under stones and screes. Despite neighbor- ing each other, these species seem to have an allopatric distribution, their ranges being approximately 30 km far from each other (Yaginuma 1972(Yaginuma , 1979. Known records of N. kosodensis stat. nov. and N. latiscapus are reported in Fig. 10.

Impact of species/subspecies delimitation on conservation of Japanese cave fauna
All Japanese nesticids subspecies are currently know from sporadic records only, being limited to one cave (C. floronoides notoi, C. floronoides tatoro, N. longiscapus awa, N. longiscapus draco, N. tosa iwaya, N. tosa niyodo) or just few localities from a restrict- ed area (C. floronoides komatsui, Nesticus akiyoshiensis ofuku) (Yaginuma 1976(Yaginuma , 1978(Yaginuma , 1979. They apparently share small differences in the morphology of genitalia and a parapatric or allopatric distribution with widely distributed conspecifics from the same area. Furthermore, their attribution to an intraspecific rank rather than a separate, well-defined species was mainly based on traits which are not clearly distinguishable (Yaginuma 1979). Thus, it is unclear if such spiders should be considered as real endemic or just local variations of widely distributed species.
Endemic subterranean organisms, including cave spiders, are considered a priority target for species conservation since they are prone to extinction due to local or global environmental changes (Mammola et al. 2018). Nevertheless, our ability to correctly separate species and subspecies may have an impact on the establishment of protective actions for cave fauna.
Incorrectly defining a valid species as a subspecies may underestimate the biodiversity of a cave, and thus fail to identify biodiversity hotspots and potentially endangered species. In turn, efforts spent for their protection may be insufficient, potentially causing the loss of endemic organisms even before they are discovered. On the other hand, correctly evaluating the existence of geographic variants of widely-distributed species may enable more efficient and effective prioritization in conservation projects. Widelydistributed organisms are often less vulnerable than endemic taxa, and have lower con- servation priority (Primack 2006). Using limited economic resources to protect least concern species may distribute funding away from more threatened species.
Revising the taxonomic position of long-established and uncertain subspecies, like in the case of the Japanese nesticid spiders, can therefore be worthwhile in terms of conservation, especially in countries hosting several unique endemic species like Japan.

Conclusions
Both morphological and molecular analysis suggest that N. latiscapus is distinct from N. kosodensis stat. nov., rejecting the status of subspecies for N. kosodensis stat. nov. In the light of these results, the elevation of N. kosodensis stat. nov. to the rank of species is herein proposed. Our results further highlight that future work should investigate the remaining Japanese nesticid subspecies, as well as other spider subspecies, including from a conservation perspective. It is likely that new, detailed studies involving both morphological and molecular comparison may lead to the rise of more of these interspecific ranks to valid species, thus further increasing the already high diversity of the family Nesticidae in Japan, and Japanese cave fauna in general. Such result may also help to promote new and more correct conservation plans to preserve the endemic cave spider fauna in Japan.