A Case for Below-Ground Dispersal? Insights into the Biology, Ecology and Conservation of Blind Cave Spiders in the Genus Troglodiplura (Mygalomorphae: Anamidae)

Simple Summary Blind cave spiders of the genus Troglodiplura are large, enigmatic spiders, which show a number of adaptations to an underground existence, including elongated limbs and a complete lack of eyes. They are known only from the Nullarbor Plain, Australia and prior to this study the species were only known from juveniles or fragments of dead spiders. We investigated the distribution of Troglodiplura in South Australia, providing detailed observations of the behaviour of living adult and juvenile spiders in their natural habitat and in captivity. Given the adaptations of the spiders to an underground existence, plus the barriers to above-ground dispersal posed by the arid climate of the Nullarbor Plain, we expected species to be confined to single caves. However, our molecular evidence showed that the species T. beirutpakbarai was distributed across multiple caves in South Australia, each separated by between 10–27 km, and that there had been connection between the caves in recent history. These findings provide intriguing evidence for subterranean between-cave dispersal mechanisms for these spiders. A number of threats to cave fauna and the fragile cave ecosystem are posed, and we recommend further research to better define the distribution of vulnerable cave taxa and their conservation profiles. Abstract Previously described from only fragments of exoskeleton and juvenile specimens, the cave spider genus Troglodiplura (Araneae: Anamidae), endemic to the Nullarbor Plain, is the only troglomorphic member of the infraorder Mygalomorphae recorded from Australia. We investigated the distribution of Troglodiplura in South Australia, collecting and observing the first (intact) mature specimens, widening the number of caves it has been recorded in, and documenting threats to conservation. Phylogenetic analyses support the placement of Troglodiplura as an independent lineage within the subfamily Anaminae (the ‘Troglodiplura group’) and provide unequivocal evidence that populations from apparently isolated cave systems are conspecifics of T. beirutpakbarai Harvey & Rix, 2020, with extremely low or negligible inter-population mitochondrial divergences. This is intriguing evidence for recent or contemporary subterranean dispersal of these large, troglomorphic spiders. Observations of adults and juvenile spiders taken in the natural cave environment, and supported by observations in captivity, revealed the use of crevices within caves as shelters, but no evidence of silk use for burrow construction, contrasting with the typical burrowing behaviours seen in other Anamidae. We identify a range of threats posed to the species and to the fragile cave ecosystem, and provide recommendations for further research to better define the distribution of vulnerable taxa within caves and identify actions needed to protect them.


Introduction
Caves, with near uniform annual temperature and humidity [1], low light levels, low energy inputs, and low diversity and abundance of organisms, represent extreme natural ecosystems. Many organisms living within caves are able to also exist in epigean habitats, and are termed eutroglophiles; however, others are only able to exist strictly within caves and these are termed troglobionts or troglobites [2]. Many troglobionts have specifically evolved in response to these extreme environs and show a range of morphological features, including loss or reduction of eyes, thinning or changes in the structure of the cuticle, loss of pigmentation, and elongation of appendages [3][4][5]. The repeated evolution of such phenotypes, collectively known as troglomorphisms, across diverse lineages provides models that offer insights into the developmental and genetic basis of evolutionary change.
In Australia, troglobiont spiders have been recorded from a number of families, including Anapidae [6]; Mysmenidae [7]; Pholcidae [8]; Ctenidae [9,10]; Desidae [9]; Gradungulidae [11]; Zoropsidae [10]; Linyphiidae [12]; Stiphidiidae [9]; Symphytognathidae [13]; Trachycosmidae [14]; and, the subject of this paper, the Anamidae [15]. While all of these examples are regarded as troglobionts, for many this is inferred from their location in cave systems, with only some actually showing evident troglomorphic features. Of these, the genus Troglodiplura Main, 1969 (Anamidae), endemic to the Nullarbor Plain or Nullarbor Divide, is clearly troglomorphic and is the only lineage of troglomorphic Mygalomorphae [15] known from Australia. The Nullarbor Plain (part of the biogeographic Nullarbor Divide) is an extensive arid landscape in southern Australia [16], covering an area of around 200,000 km 2 , and containing hundreds of caves [17]. The region is remarkable for its harsh environment, with a mean annual rainfall of 250 mm near the coast, temperatures spanning from a minimum of around −2 • C in winter to above 40 • C in summer [18]; inland there is a conspicuous lack of large woody vegetation and therefore shade. The cave environs provide a stark contrast to the extreme climes of the surface, with many caves recording near uniform annual temperatures and relative humidity deep in the cave systems, with temperatures stable around 16 • C to 18 • C depending on their locality, and relative humidity ranging between 50% and 90%, depending on the cave structure [1,19].
Spiders of the infraorder Mygalomorphae have a range of traits that make them especially vulnerable to threatening processes, such as habitat specificity, reduced physiological tolerance to changes in abiotic factors (for example sensitivity to desiccation), longevity, low vagility and a tendency towards short-range endemism, and as such are recognised for their conservation significance [20,21]. Troglodiplura are large spiders, showing a number of adaptations to a hypogean existence, including complete loss of eyes or eye spots, and elongate appendages [12]. Species of Troglodiplura are enigmatic both biologically and ecologically; prior to this study, all mature specimens in collections were known from only fragments of exoskeleton, with fresh specimens having only been collected for one species, T. beirutpakbarai Harvey & Rix, 2020, all of which were juveniles.
Troglomorphic adaptations can render troglobiont organisms unable to survive in the epigaeic environment [22] and so restrict above-ground dispersal from cave systems. Thus, in the absence of subterranean dispersal, cave systems can act as habitat islands which present allopatric barriers to gene flow [23,24]. Organisms restricted to such isolated systems are typically characterised by extreme genetic divergence between neighbouring sub-populations, and high levels of species endemism to particular cave systems [25][26][27][28][29]. Several examples of extremely limited dispersal have been found for spiders of the family Nesticidae [30,31], for example the troglomorphic Nesticus barri Gertsch, 1984 [27], which had no shared COI haplotypes between individuals from caves at distances of greater than 12 km [28]. Similar patterns of isolation were found for the troglomorphic spider Troglohyphantes vignai Brignoli, 1971 (Linyphiidae) of the Italian Alps [32], and in Telema cucurbitina (Wang & Lee, 2010) (Telemidae) in South China karst [33], suggesting highly restricted or no dispersal. Even for spiders with a broad range, such as the troglomorphic spider Parastalita stygia (Joseph, 1882) (Dysderidae), which occurs across an area of 240 km 2 in the Dinarides mountain range in the Balkans, molecular analyses revealed deep population structuring and pronounced patterns of isolation, suggesting any dispersal is highly limited [34,35]. The foregoing case studies correspond to a "caves as islands" model [36,37]. In contrast to this model, some cave systems have variable amounts of physical linkage between caverns including via lava tubes, meso-or micro caverns, and subterranean pathways in cavernicolous limestone and/or groundwater, which may act as pathways through which troglobiont organisms can disperse [22,23]. Dispersal potential through these pathways is influenced by the presence of stratigraphic or fluvial barriers to dispersal, by the organism's vagility, or relative dispersal ability [22,23,38,39], and by its level of troglomorphism, with more highly troglomorphic species having a lower dispersal ability [32,34]. Subterranean dispersal is more commonly recorded in stygofauna and aquatic troglobionts than in terrestrial troglobionts, where water dwelling organisms may face fewer barriers to dispersal [27,[38][39][40][41]. Despite being less common, evidence of subterranean dispersal of terrestrial troglobionts has been recorded for a range of invertebrate taxa, including cave beetles of the tribe Trechini (Carabidae) in North America [38], and of the tribe Leptodirini (Leiodidae) in the Pyrenees [41][42][43]. In arachnids, evidence of subterranean dispersal was used to explain the distribution of the troglomorphic pseudoscorpion Protochelifer naracoortensis Beier, 1968 across multiple caves, albeit separated by only small distances [26]. Arnedo et al. (2007) collected troglobiont spiders of the genus Dysdera Latreille, 1804 (Dysderidae) from cracks and voids between lava tubes in the Canary Islands, providing direct evidence of dispersal of spiders through micro and meso caverns [44]. Shared haplotypes between isolated cave populations of troglomorphic species of Cicurina Menge, 1871 (Dictynidae), provide indirect evidence of gene flow between populations [5]. Furthermore, different Cicurina clade members normally only found in caves many kilometres apart have appeared sympatrically, associated with certain fractures [45]. A similar pattern of shared haplotypes was seen in troglobiont species of Neoleptoneta Brignoli, 1972 (Leptonetidae) [46]. Where subterranean pathways exist and where organisms are capable of dispersing, it is conceivable that instead of acting as isolated islands, caves can be viewed as a part of a broader, but still restricted, network of connected fissures and cracks, with ample potential dispersal pathways for fauna [47].
Cave environments are fragile ecosystems, which face a range of threats from anthropogenic sources including mining, groundwater extraction and contamination, impacts from above-ground development, and climate change [48]. Theoretical modelling predicts an increase in cave temperatures with increasing global temperatures, albeit with a lag time between surface increases and hypogean increases [49][50][51][52][53]. Having evolved in a hypogean environment, with more or less stable temperatures and humidity, it is likely that troglomorphic organisms have a substantially reduced tolerance to abiotic changes to the cave environment [5]. This, coupled with the limited dispersal potential of troglomorphic organisms [54], and the challenges associated with dispersal to other subterranean environs, means troglomorphic species are at a heightened risk of climate-change-induced extinction events [52,55]. The extinction risks facing troglomorphic mygalomorph spider lineages, such as Troglodiplura, are likely compounded by the ecological and life history traits of mygalomorph spiders more generally and such species can be considered of prime conservation concern [20,56].
Whilst surveys of the Nullarbor Plain to date have revealed a biologically intriguing, and vulnerable endemic biota (Table 1), many of the caves are yet to be surveyed. In this study we present data on surveys conducted in some South Australian caves of the Nullarbor Plain, with notes on the ecology, biology, and conservation of T. beirutpakbarai, and discuss findings in relation to the likely conservation status of other troglobiont spiders. Given that Troglodiplura is highly troglomorphic, combined with the inherent tendency towards low vagility and short-range endemism shown by mygalomorph spiders, and the barriers to dispersal presented by the harsh, arid surface environment, we hypothesise that above-ground dispersal of Troglodiplura is severely (if not entirely) restricted and dispersal in total is low, supported by the absence of any historical collections of Troglodiplura in non-karst environments. Given this, we predicted that populations of Troglodiplura from neighbouring caves would exhibit high levels of genetic divergence.  [58] Dark zone in air chamber ca. 3 km north of cave entrance [15].

Study Area
The Nullarbor Plain, located in southern Australia, is a vast, approx. 240,000 km 2 arid limestone plateau that constitutes the world's largest area of exposed karst, and hosts thousands of shallow cave and karst features [17,58]. The study area is located in the South Australian part of the IBRA 7.0 Nullarbor bioregion ( Figure 2) [59]. Caves and karsts in Australia are numbered using the Australian Speleological Federation numbering system to aid identification [60]; in this manuscript we quote the cave number name the first time a cave is mentioned and henceforth refer to the cave by the cave number. The majority, but not all of the shallow caves, blowholes and anastomosing tubes across the Nullarbor were formed underwater around 6 million years ago [61,62] and because of continental uplift, lowering of the water table, or regional uplift of limestone, are now full of air. The result is a horizontal plane of rubbly perforated limestone with a high volume of air space. Anastomosing tube development is observed in limestone all over the world [62,63] and on the Nullarbor it is an abundant and extensive form of perforation in the upper levels of Nullarbor limestone arising from epiphreatic solution (a zone of intense solution just below the top of the water table) [17,63]. In the subsurface of the plain there are many voids, ranging in size from small tubules through to large caves; the small tubules are irregular cavities, generally <20 cm in diameter, with narrow sinuous anastomosing tubes that follow bedding planes and joint surfaces [63] (Figure 3a-e). These honeycomb the limestone close to the surface of the plain, providing connectivity beneath the hard calcrete capstone and are connected to the surface via thousands of blowholes scattered across the Nullarbor [17]. The typical blowhole cave on the Nullarbor Plain is formed upwards to the surface from voids beneath by crystal weathering processes [64] and consists of a round vertical shaft (ranging from 0.5 m to 3 m diameter and up to 7 m depth), which may or may not intersect with chambers or passages of other cave types at depth; however, they consistently intersect with anastomosing tubes. Blowholes and associated shallow caves are distributed across the Nullarbor Plain in a 25-30-km-wide band located approx. 75 km inland and are evident in the study area [61]. The density of blowholes is up to 43 per 25 km 2 [61]; in the study area encompassing N-253, N-6838, and N-5896, there are 138 known shallow caves and karst features [58], and an estimated 688 blowholes [61] with an average of 520 m between two karst features if randomly distributed.

Site Selection
The geomorphic characteristics of caves in which Troglodiplura spiders from the Nullarbor Plain have been collected are diverse (Table 1), with individual caves being up to 475 km apart. Many hundreds of caves on the Nullarbor Plain share features of those where Troglodiplura have been observed. In this study, we therefore focused surveys on caves which were proximal to, or shared similar features with, N-253 Eagles Rest Cave in which T. beirutpakbarai was previously collected (Figure 1) [15]. Based on the geomorphic characteristics of N-253, nearby caves with a moist and extensive sediment floor away from the daylight zone, with ephemeral ingress of water, were selected.
In order to protect these systems, following established practice with Australian cave data, we do not intend to publicly release, or generalise, cave data. Thus, these location data are considered as Category 1 under recommendations published by the GBIF Secretariat [65].

Survey Methods
Surveys were conducted by hand, with searches focusing on the sediment floor, cave walls, inter-rock crevices, and rock surfaces. Initial surveys were conducted using a Petzl Duo 6000k at 400 Lumens (approx.), or a Scurion 3000K at 400 Lumens (approx.) caving light attached to the caving helmet; however, to reduce the possibility of disturbance of Troglodiplura caused by vibration and light, later surveys employed red LED lights and swift but gentle movements to minimise vibrations. On collection, fresh specimens were transferred to 100% ethanol; fragments of spiders were stored dry. Given the conservation significance of Troglodiplura, collection of live specimens was conducted sensitively, and minimal specimens were collected to allow analysis of genetic diversity within a cave. Two adults (females) were collected alive from each of N-6838 (SAMA NN30812-13) and N-253 (NN 30806, NN30808). Other adults were observed or caught and released. On a subsequent entry to N-253, two already dead "mummified" females were found in N-253 and collected too (NN31721-22). Immatures were also collected from each cave. N-253 was surveyed on three dates, N-6838 on two, and the remainder of caves on one date. Survey times in each cave were variable and dependent on the size and complexity of each cave. As a rule, surveys were conducted until all suitable habitats and chambers had been inspected.

Site Characterisation
Relative humidity and temperature were recorded at each cave site using a Kestrel 3000 (Nielsen-Kellerman City: Boothwyn. State: PA, USA) handheld temperature, relative humidity, and wind speed instrument. The instrument was allowed to equilibrate with the cave environment for at least 30 minutes before measurements were taken during falling atmospheric barometric pressure to be therefore representative of typical environmental conditions. Descriptions were made of the general structure of each cave site, noting evidence of water and relevant characteristics.
Refer notes in Table 1. N-253 clearly takes a stream after rainfall events leaving ephemeral pools and calcite deposits; evidence of significant water flow and historical ponding in western passage. The deep cracks in the clay of the southern chamber floor suggest some considerable time since the last flood in that chamber.  Caves N-5894, N-5895, and N-238 were within 20 km of N-253 and also received water from rainfall events. However, they differed from N-253 by not retaining a damp sediment floor as water sank deeper into unexplorable parts of the cave, or if there was some dampness or a flat floor, the cave was subject to external temperature and humidity close to the entrance, or the chamber was in the daylight zone. Troglodiplura was not observed in these caves during surveys.

Phylogenetic Analysis
Phylogenetic analysis of the 89-taxon 'expanded' dataset (Table 2) revealed that Troglodiplura specimens collected from caves N-253, N-6838 and N-5898 represent separate but unequivocally conspecific populations of T. beirutpakbarai ( Figure 6). In COI there were only slight differences among the caves, with a maximum uncorrected pairwise divergence of 1.5% between caves N-253/N-5896 and N-6838. Specimens from caves N-5896 and N-253 were more similar genetically, with an uncorrected pairwise divergence of only 0.3%, reflecting the closer geographical proximity of these two caves (Figure 4). However, the 16S haplotypes were identical for all three caves.
Insects 2023, 14, x FOR PEER REVIEW 17 of 26 0.3%, reflecting the closer geographical proximity of these two caves ( Figure 4). However, the 16S haplotypes were identical for all three caves.

Biology
Live adult spiders were observed within a few minutes of entering a chamber, and were located in the 'open' on the cave floor, or on the underside of overhanging rocks. Spiders were typically slow moving and easily caught, but moved quickly if disturbed. Smaller juvenile spiders were observed at the interface between the flat cave floor and the rock wall, and on the cave floor, close to entrances in the twilight zone. Juveniles were observed retreating into crevices and holes in the rock, or into mud holes. In N-6838, a spider was seen squeezing into a mud hole, with an opening of 4 mm × 6 mm, and another into a natural rock hole ca. 10 mm in diameter (Figure 7a,b). Fragments of mature male exoskeletons were located in small piles outside of crevices in the rock wall, or on top of rocks. No such piles of exoskeleton were found for females; however, two large, desiccated whole ('mummified') mature females and one presumably sub-adult juvenile, based on size, were observed on rock and substrate surfaces within cave N-253. Potential prey items detected during cave surveys were cave crickets, beetles and other spiders, the three most commonly observed invertebrates detected in our surveys. One juvenile T. beirutpakbarai was caught whilst consuming a ground beetle (family Carabidae).

Conservation
Fox scats and dens were detected in two caves identified as potentially suitable habitat for T. beirutpakbarai; in one cave these were extensive, extending throughout accessible passages and into the dark zone. Recent damage caused by humans was also detected in one cave, with stalagmites broken and skulls of skeletal animal remains removed.

Behaviour in Captivity
A female T. beirutpakbarai from Cave N-253 and an immature from Cave N-6838 were kept in captivity in separate Perspex containers, with a layer of silt derived from N-253. Folded cardboard and plastic and glass tubes of various diameters were provided as shelters. However, the spiders almost never entered these and did not reside in them. Silk production was only observed when moulting-the immature spider from Cave N-6838 constructed a fine silk moulting web on the substrate, situated at one end of the enclosure, in which moulting occurred. Moulting occurred 'out in the open', with the spider lying on its back. Posture when standing still varied. Sometimes the adult female would stand for long periods with her body elevated above the substrate and at other times would have her body touching the substrate. Posture did not obviously correlate with exposure to stimulation, light, time of feeding or hydration state. At times, both spiders had active periods, roaming about their enclosures, extending legs and trying to climb the smooth sides of their containers.
One adult cave cricket, Pallidotettix sp., ca 16 mm in body length, was offered to the adult female T. beirutpakbarai, which was subdued quickly and consumed. No trace of the cricket was detectable after consumption. The same spider would not accept a wild-caught Brises sp. cave beetle (a large tenebrionid) three days later despite repeated re-introduction of this item. It would also not accept adult or immature cultured crickets, Acheta domesticus, despite the offered immature crickets being only ca. 25% of the spider's body length and provided after a month of no food. The spider appeared aware of crickets moving in its enclosure and would withdraw from where a cricket was moving. On one occasion, an offered adult A. domesticus cricket moved directly towards Troglodiplura; the spider evidently detected this at a distance > 5 cm and raised its body high above the substrate, and allowed the cricket to walk completely under it three times. In contrast to the feeding behaviour of the adult, two immature spiders accepted small A. domesticus crickets whenever they were offered.
Overall, the spiders made minimal movements. When evasive movements were undertaken, they were slow and conservative. Conversely, on a couple of occasions the female was accidentally exposed to minor but sudden vibrations, and it would then lurch forward rapidly and run very fast. Gentle blowing on the spider never elicited a flight response.

Symbiotic Mite
A new and undescribed symbiotic mite species (Imparipes sp.: Scutacaridae) was found living on adult T. beirutpakbarai specimens from both caves in which adult spiders were collected. All the mites recovered from spider bodies were found in between coxae III and IV or between both coxa IV.

Discussion
Our findings of extremely low genetic divergence (≤1.5% for COI and identical haplotypes for 16S rRNA) between populations of T. beirutpakbarai from separate caves situated between 10 and 27 km apart implies recent or contemporary gene transfer between these populations. The morphology of T. beirutpakbarai-the absence of eyes, their gracile body form, and large size-in combination with observed behavioural responses suggest that the species is a troglobiont, truly restricted to the cave environment. Some spiders are known to disperse aerially, through a mechanism known as ballooning; however, ballooning has never been recorded in the Anamidae and is apparently rare in Mygalomorphae, having been recorded from only few families [67]. This, in combination with the extreme troglomorphism shown by Troglodiplura, the tendency for low vagility of mygalomorph spiders generally, the absence of any epigean records of Troglodiplura anywhere in Australia, and the stark contrast between the surface climate and the cave environs of the Nullarbor Plain, indicates that subterranean dispersal provides the best explanation for the observed genetic pattern. Previous studies have found direct evidence for dispersal by spiders through subterranean meso-or micro caverns in the Canary Islands [44] and the presence of shared haplotypes between spider species in isolated cave populations [46,68], which could provide indirect evidence of inter-population connectivity and gene flow. Nonetheless, our findings of extremely low genetic divergence and thus apparently interconnected populations over distances of up to 27 km are unexpected. The structure of the cave systems of the Nullarbor Plain, and the potential for inter-cave connectivity via anastomosing tubes and blowholes (Figure 2), may be implicated in the patterns of divergence in T. beirutpakbarai. Each karst feature provides a possible entry point for ingress of water, and energy inputs (food); this, in conjunction with extant or historic anastomosing tube connectivity between larger caves, presents a possible mechanism for genetic exchange for T. beirutpakbarai in this area. Two caves where Troglodiplura have been observed in apparently 'sealed' environments, distant from surface entry points including obvious blowholes, are T. harrisi in N-327 [57], and T. samankunani in N-49 [15]. The presence of these species in apparently 'sealed' environments provides further evidence for a cryptic but extant pathway for the flow of energy (food) into these systems, and thus potentially for genetic exchange.
Historically, the ocean retreated for the last time around 14 million years ago, after which there was a gentle uplift of the Nullarbor region [17,69]. The climate became increasingly dry with vegetation similar to that found on the Nullarbor Plain today, except that the dry eucalypt woodland extended farther inland and probably covered much of the Nullarbor Plain [17]. A warm, wet episode occurred around~5-3 million years ago, and the climate reached its present level of dryness around one million years ago [17]. The three caves in which we detected populations of T. beirutpakbarai are located in Nullarbor limestone laid down 16-14 million years ago; it was a warm sea limestone, evidenced by the large marine fossils that are embedded in the walls of each of the caves [70]. Cave development of N-253 is different from the other two caves where Troglodiplura were observed in that the cavernous void beneath the caprock has been created by the collapse of deeper voids in Oligocene limestone during the late Miocene, in a process observed elsewhere on the Nullarbor [35]. N-6838 and N-5896 appear different from N-253 in that they are shallow caves with flank margin characteristics, forming around 6 million years ago [61].
The ancestor of Troglodiplura probably entered Nullarbor cave systems sometime during or after the Miocene [15]. This entry can be accounted for by two distinct models, representing two sides of a debate in cave biology, which have implications for resolving the paradox of Troglodiplura diversity. Firstly, Troglodiplura might have evolved parapatrically based on restricted gene flow between adjacent cave-dwelling and surface-dwelling subpopulations. If so, this might have occurred at any time after caves became available to colonise (i.e., up to approx. 14 million years ago) which includes a relatively early time including a climatically subtropical period. This would correspond to the parapatric/ecological speciation model. Secondly, following the climate relict model [41], Troglodiplura might have evolved allopatrically as a climate relict [71]. If so, this would be likely to occur after aridification stranded an anamid ancestor in a cave, surrounded by a drying landscape, with an above-ground habitat matrix considered increasingly impermeable to dispersal. If so, troglomorphic Troglodiplura could have evolved later, with the pulse of aridification commencing ca. 3 million years ago a key possibility. Testing the generality of this latter model is feasible due to some highly distinct arthropods that cohabit N-253. For example, at least three mesic-adapted mite taxa in different families occur here, yet these taxa are otherwise essentially known from wet forests. They are clearly mesic-restricted and are geographically separated from known relatives by distances ranging up to 2,500 km (Geogamasus cf howardi (Ologamasidae); Castriidinychidae; Hybalicus (Lordalychidae)) [72][73][74]. While overlapping predictions can make these models difficult to disentangle, there are opportunities for future molecular work to infer the mode and tempo of Troglodiplura diversification. Outcomes of such analyses could have implications for conservation as, for under a climate relict model, intolerance of a deteriorating climate is then directly associated with Trogodiplura biogeography.
There are other distinct arthropods that co-occur with Troglodiplura which provide extra research potential for these cave systems. One unusual example that could add to our understanding of Troglodiplura spp is the scutacarid mite Imparipes sp. found upon T. beirutpakbarai females. Scutacarids frequently show high host-specificity and are clear candidates for co-phylogenetic analysis. They are regarded as harmless commensals that feed on fungi, with males living off-host in nest-like situations. Interestingly, the only other scutacarids known from spiders are three species also from nemesioid hosts and these also inhabit the intercoxal space between coxae III and IV. These latter scutacarid species are specific to three congeneric pycnothelid hosts from Argentina [75].
As hypothesised by Harvey et al. [15] based on morphology, Troglodiplura was recovered as an independent lineage within the subfamily Anaminae, here referred to as the 'Troglodiplura group' (Figure 6). Our observations taken in the natural cave environment and in captivity indicate that T. beirutpakbarai does not construct burrows, or utilise silk for shelters, even though these behaviours are typical of other Anamidae. Troglodiplura beirutpakbarai were most often first located in 'the open', on the surface of rocks or substrate and when disturbed retreated to small crevices in the cave wall, close to ground level. Our results indicate that the species is an opportunistic generalist (and likely cursorial) predator, feeding on mid-sized cave beetles (Carabidae), cave crickets (Rhaphidophoridae) and with juveniles accepting cultured small epigean A. domesticus crickets (Gryllidae) in captivity. The presence of several large, whole, desiccated mature females and a large juvenile on rock and substrate surfaces illustrates the preservative nature of the cave environment, and indicates a scarcity of scavengers (at least in areas used by large Troglodiplura).
Whilst our findings that T. beirutpakbarai is not isolated to a single cave has positive conservation implications in that the species is not an extreme short-range endemic, the known range of the species still remains highly restricted and substantial threats exist across all of its known populations. Troglomorphic organisms, which have a substantially reduced tolerance to abiotic changes to the cave environment [5], are likely to be of elevated vulnerability to fluctuations in temperatures and humidity caused by anthropogenic climate change [49][50][51][52][53], and the latter will likely present a key threat to Troglodiplura. In addition to these indirect climatic threats, humans potentially pose direct threats to troglomorphic spiders, both from damage to the cave environment e.g., [76], as evidenced during our surveys, and through the potential collection of specimens for the pet trade [77]. Collection of wild animals for the pet trade is a major threat to wildlife around the world and whilst data on the impacts on invertebrates are lacking [78], it likely represents a key threat [79,80], especially for large and charismatic organisms, such as mygalomorph spiders [78,81]. Access into the caves is sometimes difficult, so this may provide a barrier to some collectors, but the importance of keeping the location details of caves that contain Troglodiplura out of the public realm is high. Caves are fragile environments and careless visitation, or even looting, can easily cause irreparable harm to the cultural values, biological integrity, and the archaeological and palaeontological record [82].
During cave surveys, we detected a high level of evident fox activity, both in the form of scats and dens, which in one cave extended throughout accessible passages. Gut content analysis studies have shown that both foxes and feral cats consume invertebrates [83][84][85], and predation by foxes has been identified as a possible threat to mygalomorph spiders [20], indicating that foxes are likely a key threatening process for Troglodiplura, especially for cave systems that are easily accessible from the surface, such as Cave N-58.
This study provides important baseline data for these spiders. However, a large number of caves of the Nullarbor Plain are yet to have had biological surveys and the potential for them to harbor similarly rare, endemic, and vulnerable biota is high. Dedicated surveys for Troglodiplura are recommended to better elucidate the distribution of species, to detect new species, identify which are likely to be at threat and the conservation actions needed to protect them. Collection and analysis of eDNA from cave sediment has been shown to be successful in detecting arthropods in cave environs [86] and provides increased opportunity to detect these species. Additionally, the use of passive surveillance, such as camera traps, has been documented as a successful method for monitoring populations of mygalomorph spiders and for collecting ecological and behavioural data [21]. The use of non-destructive survey techniques such as these is especially important in fragile habitats, such as caves and in combination, they may facilitate the detection and subsequent longterm monitoring of Troglodiplura populations, providing important conservation-relevant data to aid in their protection.

Conclusions
The findings we have presented in this study provide insights into the biology and ecology of this enigmatic species of spider and open up new avenues for research and investigation. Our molecular findings of little to no divergence between individuals from apparently separate cave systems were unexpected and provide intriguing evidence for a possible below-ground dispersal mechanism for these spiders. However, this hypothesis requires testing in further studies. Caves of the Nullarbor Plain vary in their geomorphic characteristics, abiotic conditions, and biota. Understanding which combinations of cave environments and geological history correlate with the presence of rare troglobionts such as Troglodiplura is recommended as an important goal for both biodiscovery and conservation.

Data Availability Statement:
The data presented in this study are available in Supplementary File S1.