Cave-dwelling pseudoscorpions of China with descriptions of four new hypogean species of Parobisium (Pseudoscorpiones, Neobisiidae) from Guizhou Province

We summarize and discuss the 29 known cave-dwelling pseudoscorpion species from China. Four new troglomorphic pseudoscorpion species, Parobisium motianense sp. nov., P. qiangzhuang sp. nov., P. sanlouense sp. nov., and P. tiani sp. nov., belonging to the family Neobisiidae, are described based on specimens collected in karst caves in Guizhou, China. Detailed diagnosis, descriptions, and illustrations are presented. We also provide recommendations for management of caves where they occur, as well as the cave arthropod communities and the habitats that support them.


Introduction
Biospeleological studies in the South China Karst (SCK) has rapidly accelerated in recent years. Since 2017, 39 new subterranean-adapted species across several taxonomic arthropod groups have been described (Gao et al. 2017;Huang et al. 2017;Li and Wang 2017;Song et al. 2017;Tian et al. 2017Tian et al. , 2018Deuve and Tian 2018;Li et al. 2019a). Overall, at least 382 cave-dwelling arthropod species are now known from this region (Ran and Yang 2015;Tian et al. 2016;Li and Wang 2017;Gao et al. 2018;Feng et al. 2019;Li et al. 2019b;Liu and Wynne 2019). Incidentally, this work has also resulted in the identification of at least 21 troglomorphic pseudoscorpion species (refer to Feng et al. 2019;Li et al. 2019b).
The pseudoscorpion genus Parobisium was first established by Chamberlin (1930) as a subgenus of Neobisium Chamberlin, 1930, and later elevated by Chamberlin and Malcolm (1960) to generic rank. Parobisium is characterized by the absence of a galea on the movable cheliceral finger, fixed chelal finger with a compact subterminal cluster of only three tactile setae (et, it, est), and a more diffuse subbasal to basal cluster of five tactile setae (isb, ist, ib, esb, eb) (Chamberlin 1962). However, for some North American and Asian Parobisium species, the trichobothrium (est) is isolated in the distal half of the fixed finger and has a trichobothrial pattern quite similar to Bisetocreagris.
The key character used to distinguish between these two genera is that Bisetocreagris usually has elongate galeae. Mahnert and Li (2016) intimated the galea is extremely fragile in Bisetocreagris species. This implies galea may be easily broken or damaged during collection or transport (Y. Li, pers. com., 18 December 2019). Subsequently, using galeae as a diagnostic character for describing and identifying species may hinder accurate classification of this group. In general, Parobisium differs from Bisetocreagris as there is a distinct and rounded sclerotic knob, rather than an absence of galea (Morikawa 1960;Hong 1996;Mahnert and Li 2016).
During the examination of Guizhou specimens collected by Mingyi Tian (of the South China Agricultural University, Guangzhou Province) between 2013 and 2017, we identified several species of Parobisium, which may be undescribed. In most cases, we had too few specimens to formally describe the species, and in some cases we had only one specimen. Unfortunately, this can be limiting in describing new species, especially given the aforementioned considerations with the galea. To address this problem, the lead author and colleagues collected additional specimens at the caves originally sampled by M. Tian. With additional specimens, we were able to both describe these species and confirm that these Parobisium species have a distinct and rounded sclerotic knob rather than the absence of galae.
Based upon specimens collected by both M. Tian and the lead author, we describe four new species of Parobisium from caves in Guizhou Province, China. All species are subterranean-adapted, and include P. motianense sp. nov., P. qiangzhuang sp. nov., P. sanlouense sp. nov., and P. tiani sp. nov. We also provide recommendations for management of these caves and the cave arthropod communities and habitats they support.

Study area
Guizhou, located in the Yunnan-Guizhou Plateau, is the centrally located province within the SCK. The karst escarpment within this area is approximately 130,000 km 2 encompassing 73% of Guizhou Province (Rong and Yang 2004;He and Li 2016). Karst formation in Guizhou emerged from a plate group from the Proterozoic to the Quaternary Period, and consists mainly of shallow marine carbonate deposits (Zhou et al. 2017). The extensive distribution of carbonate geology and the subtropical monsoon climate provide suitable conditions for the development of karst caves. According to Zhou et al. (2017), Guizhou supports at least 4,960 caves.
The lead author and colleagues searched for and collected pseudoscorpions within three of the four Guizhou caves (Fig. 1) where M. Tian initially collected specimens. As none of these caves were subject to previous studies or exploration efforts, cave maps were not available. We have provided estimations of entrance configuration and cave length, as well as information on surface vegetation and adjacent human activities.
Motian Cave (Figs 1C,18) is located ~2 km southwest of Tangbian Town, Pingtang County. This limestone cave has one downward sloping oval entrance (~8 meters high by ~4 meters wide), approximately 2100 meters in length, and extends horizontally. The cave is surrounded by agriculture with the nearest rural residential area less than 100 m from the cave entrance.
Zharou Cave (Figs 1B,19) is located ~1 km north of Daying Town, Ziyun County. This limestone cave has one triangular entrance (~2 meters high and ~3.5 meters wide), approximately 80 meters in length, and extends horizontally. The surrounding area is largely disturbed and characterized by low shrubs and weeds (Gramineae); agricultural fields and a rural residential area are approximately 100 m away.
Sanlou Cave (Figs 1D, 20) is located ~2.5 km northwest of Daoping Town, Fuquan City. This horizontally-trending cave is approximately 200 meters in length and has an irregularly round entrance (~4 meters in diameter). Situated near a sand mining operation, this cave is the primary water source for the village of Daoping. Subsequently, a reservoir and water delivery system was built in the deepest part of the cave. The cave has also been designated as a water source protection area. Biyun Cave (Figs 1A, 21) is located in Biyun Park, Chengguan Town, Pan County and is less than 50 m from a rural residential area. This cave has two entrances. One entrance is dome-shaped (~30 meters wide at the base and 10 m high); during the rainy season, a river flows into this cave entrance. The second entrance is located about 80 m uphill from the lower entrance and is irregularly round in shape (~30 meters in diameter). This cave was developed as a tourist cave, and has an unmaintained footpath paved with concrete, which connects the two entrances.

Field sampling
From 29 July to 5 August 2019, researchers conducted direct intuitive searches (sensu Wynne et al. 2019) in the estimated deep zone of each cave by examining bat guano, dead insects, edges of pools and streams, flood detritus, and mud floors (Figs 19B,20C). Two observers spent approximately two hours searching in Zharou Cave, four hours searching in Biyun cave, and four observers spent about three hours searching Sanlou Cave.

Preparation and analysis
Specimens were preserved in 75% ethanol and deposited in the Museum of Hebei University (MHBU), Baoding, China. Photographs were taken using a Leica M205A stereomicroscope equipped with a Leica DFC550 camera and LAS software (Ver. 4.6). We used a Leica M205A stereomicroscope (with a drawing tube) for drawings and measurements. Chela and chelal hand were measured in ventral view. All measurements are in millimeters (mm) unless noted otherwise. Detailed examination of characters was done using an Olympus BX53 general optical microscope. Temporary slide mounts were prepared in glycerol.

Terminology
Cave ecosystems typically consist of four environmental zones (Howarth 1980(Howarth , 1983: (1) entrance zone-or light zone, which represents a combination of surface and cave environmental conditions; (2) twilight zone-occurring slightly deeper within the cave and has both diminished light conditions and direct influence of surface environment; (3) transition zone-aphotic, yet barometric and diurnal shifts may still occur at a significantly diminished rate, but the climate is approaching near stable conditions; and, (4) deep zone-complete darkness, high environmental stability, near stable temperature, near water-saturated atmosphere, and low to no airflow (usually in the deepest part of the cave). The deep zone represents the region most conducive to supporting subterranean-adapted animals. Although there are four primary cave specific functional groups recognized, the species discussed here have been identified as either troglobionts or troglophiles. Troglobionts are obligate cave dwellers that require the stable environmental conditions of the deep zone to complete their life cycle and exhibit morphological characteristics (i.e., troglomorphisms) indicative of subterranean adaptation (Sket 2008). Troglophiles (or troglophilous organisms) lack troglomorphic characters yet occur facultatively within caves and complete their life cycles there, but also occur in similar cave-like surface habitats (Barr 1967, Howarth 1983. Pseudoscorpion terminology and measurements mostly follow Chamberlin (1931) with some minor modifications to the terminology of the trichobothria (Harvey 1992) and chelicera (Judson 2007). The following abbreviations used for the trichobothria: b = basal; sb = sub-basal; st = sub-terminal; t = terminal; ib = interior basal; isb = interior sub-basal; ist = interior sub-terminal; it = interior terminal; eb = exterior basal; esb = exterior sub-basal; est = exterior sub-terminal; and, et = exterior terminal. Table 1. The 29 known cave-dwelling pseudoscorpion species from China. 'Category' indicates the functional group in which each species belongs -either troglobiont or troglophile. The number of caves (# Caves) may be used to infer the level of endemism. Names of administrative provinces where each species is presently known is also provided.

Key to Parobisium species of China
Venom apparatus present only in fixed chelal finger, venom duct short, not extending beyond half the distance to et. Fixed chelal finger with 96-98 teeth, movable finger with 95-97 teeth.
Measurements: (length/breadth or depth in mm; ratios for most characters in parentheses Etymology. The species name, qiangzhuang, was derived from the Latinized Mandarin phrase for "strong and hardy" qiáng zhuàng (强壮), which refers to the shape of chela.
Pedipalps (Figs 12D-E, 13H, J-K): Apex of coxa rounded, with 5 setae on each side, pedipalpal coxa with 8-10 setae. Pedipalp smooth and slender except for hand and inside lateral of femur, which is finely granular. Trochanter 2.86-2.92 times longer than wide, femur 6.50-6.59, patella 5.07-5.11 times longer than wide, pedicel about half the entire length of patella, chela (with pedicel) 6.08-6.34, chela (without pedicel) 5.44-5.62 times longer than wide, movable finger 1.51-1.57 times longer than hand (without pedicel). Fixed chelal finger with 8 trichobothria, movable finger with 4, eb and esb on lateral margin of hand; ib, ist and isb closely grouped at the base of the fixed finger; est slightly distal of finger at middle; it slightly closer to fingertip than et; on movable finger st nearer to t than to sb, the latter slightly nearer b than to st (Figs 12D, 13H, J). Venom apparatus present only in fixed chelal finger, venom duct short, not extending past half of the distance to et. Fixed chelal finger with 126-130 teeth, movable finger with 119-126 teeth.
Distribution. This species is known only from the type locality.
Pedipalps (Figs 16E-F, 17H-I, K): Apex of coxa rounded, with 5 setae on each side, pedipalpal coxa with 7-8 setae. Pedipalp smooth and slender except for hand, which is finely granular. Trochanter 2.67-2.70 times longer than wide, femur 5. 63-5.73, patella 4.52-4.58 times longer than wide, pedicel about half the entire length of patella, chela (with pedicel) 4.97-5.03, chela (without pedicel) 4.39-4.40 times longer than wide, movable finger 1.44-1.45 times longer than hand (without pedicel). Fixed chelal finger with 8 trichobothria, movable finger with 4, eb and esb on lateral margin of hand; ib, ist and isb closely grouped at the base of the fixed finger; est slightly distal of finger middle; it closer to fingertip than et; on movable finger, st nearer to t than to sb, the latter slightly nearer st than to b (Figs 16F, 17H-I). Venom apparatus present only in fixed chelal finger, venom duct short, not extending past half of the distance to et. Fixed chelal finger with 104-109 teeth, movable finger with 105-112 teeth.

Discussion
Our work has increased the number of Chinese cave-dwelling pseudoscorpions from 25 to 29 species. In addition, we found that female individuals may have a distinctly rounded sclerotic knob, while the sclerotic knob in males was not often obvious. Uncertainty of the characteristics of the sclerotic knob makes versus the damaged of galeae it difficult to identify and describe species -especially when only a single-sex specimen is available. We recommend that future researchers: (1) collect multiple specimens to help ensure both adult males and females are collected (enabling further examination and study of the sclerotic knob across additional specimens and species); (2) carefully collect pseudoscorpions to avoid damaging the galeae; (3) cautiously examine the galeae particularly when specimens are few; and, (4) use scanning electron microscopy (e.g., Cokendolpher and Krejca 2010) to more accurately examine the structure of the galeae. The latter will enable researchers to determine whether galea are reduced or altered during collection or transportation.
As with most Chinese hypogean pseudoscorpions (Table 1; Schawaller 1995;Mahnert 2003Mahnert , 2009Mahnert and Li 2016;Gao et al. 2017;; Gao et al. , these four new species are currently considered single cave endemics. However, this may be due to limited investigations in the region, rather than actual short-range distributions of these species. There is a high density of caves in Guizhou -in particular, numerous caves occur within a 5 km radius of the caves containing these species. Specifically, there are at least three caves with a 5 km radius of Zhaorou Cave, no fewer than seven caves with a 5 km radius of the Sanlou Cave, and at least 10 caves with a 5 km radius in Biyun Cave. None of these caves have been inventoried for cave-dwelling arthropods. As a result, it is possible our newly described species are actually restricted to a geological formation rather than a single cave (Schawaller 1995;Mahnert 2009;Mahnert and Li 2016; Table 1). Additional investigations will be required to make this distinction.
While it is becoming increasingly well-established that caves in China are rich in cave biological resources and support subterranean-adapted species with highly restricted distributional ranges, the county presently lacks policies or a governmental agency to protect and manage subterranean natural resources. This presents challenges for conservation and management of sensitive subterranean animal populations. Unfortunately, human activities (i.e., urbanization, mining, and other related activities) in karst areas, and the development of tourist caves, continue without prior evaluations of the potentially sensitive natural resources and/or the biodiversity they may support Figure 19. Zharou Cave, type locality of Parobisium qiangzhuang sp. nov. A entrance B Area where P. qiangzhuang specimens were collected. (Whitten 2009). Therefore, we recommend the distributions of the new species described here be more thoroughly established through sampling of the abovementioned caves surrounding the type localities. Importantly, once we determine whether these species are single cave or regional endemics, this information may be used to guide management policies to protect these animals and their habitats.
Local human activities (Figs 20A, D, 21D) are more likely to have significant impacts on these cave-dwelling species, which could result in their imperilment or potentially their extinction. In fact, the four new Guizhou pseudoscorpions and their habitats are directly threatened by human activities. All four caves were close to human settlements and/or agricultural areas (within 0 m to 100 m) and were affected to varying degrees by other human activities. Motian Cave is surrounded by agricultural activities. For Zharou Cave, agricultural activities are less than 100 m away. In both cases, pesticide and fertilizer residues may contaminate the caves via runoff (Castaño-Sánchez et al. 2019). Although the entrance of Zharou Cave is somewhat obscured by vegetation, local residents are aware of this cave, and we observed recent evidence of human activities. The entrance of the Sanlou Cave is about 50 m from sand mining operations, and the deepest part of the cave has been modified and converted to a reservoir. Additionally, the sand mining facility and the water extraction activities may ultimately affect the survival of P. sanlouense sp. nov.
Biyun Cave, located in Biyun county park, is a tourist cave. Evidence of human activity was observed throughout the cave, which included refuse, remnants of bonfires, and graffiti on the cave walls. Subsequently, cave habitats have been damaged to varying degrees. During our work, we observed at least 10 tourists visiting the cave. Fortunately, because of the muddy and steep path at the back of the cave, we suspect fewer visitors will be willing to access the area where we found P. tiani sp. nov. As a result, this habitat may be somewhat protected.
Extinction is often characterized by time lags, and at-risk populations may persist for long periods of time near extinction thresholds prior to becoming extinct (e.g., Brooks et al. 1999, Hanski and Ovaskainen 2002, Vellend et al. 2006. These "ex- Figure 20. Sanlou Cave, type locality of Parobisium sanlouense sp. nov. A Surrounding surface vegetation with cave entrance (white arrow) and sand mining operations adjacent to the cave B Entrance C Area where P. sanlouense specimens were collected D Terminus of Sanlou Cave with the potable water reservoir; specimens were collected near the reservoir. tinction debts" (see Tilman et al. 1994) may occur when populations become isolated following significant environmental perturbations or intensive human activities. Both P. sanlouense sp. nov. and P. tiani sp. nov. occur in areas of intensive human activities. This further emphasizes the need for additional surveys to determine whether these species are single cave or regional endemics. This distinction will be of critical importance in determining the sensitivity of these animals to current human activities within and at proximity to the caves where they occur. Moreover, understanding their distributions will be required to develop effective monitoring protocols -if additional evidence supports that either (or both) species are single cave endemics.
There are other measures that should be examined to protect sensitive cave-dwelling species and their habitats. An outreach campaign to help educate villagers, school children, and tourists concerning the vulnerability of cave biological resources should be considered (refer to Mammola et al. 2019). At Biyun Cave, it may be worth posting educational signs (within the village and perhaps near the entrance) to explain the sensitivity of cave natural resources and that endemic species occur within, as well as guidelines for reducing human impacts to sensitive cave resources. Figure 21. Biyun Cave, type locality of Parobisium tiani sp. nov. A Surrounding surface habitat with two entrances, entrance 1 (white arrow) and entrance 2 (red arrow) B Entrance 1 (white arrow) C Entrance 2 (red arrow) D Passage where specimens were collected (blue arrow).
As research on cave biological resources in southern China continues, numerous additional new species with restricted ranges will be described. Understanding their distributions and the functional roles they play in these often highly sensitive ecological communities will be of paramount importance for developing management plans to protect both sensitive species and their habitat. Through these and other efforts, we hope our findings and data collected in the future will be employed to help shape effective cave resource management in China.