Revision of the spider genus Stygopholcus (Araneae, Pholcidae), endemic to the Balkan Peninsula

The genus Stygopholcus Kratochvíl, 1932 is endemic to the Balkan Peninsula and includes only four nominal species: the epigean S. photophilus Senglet, 1971 in the south (Greece to Albania) and the ‘northern clade’ consisting of three troglophile species ranging from Croatia to Albania: S. absoloni (Kulczyński, 1914); S. skotophilus Kratochvíl, 1940; and S. montenegrinus Kratochvíl, 1940 (original rank re-established). We present redescriptions of all species, including extensive data on ultrastructure, linear morphometrics of large samples, and numerous new localities. We georeference previously published localities as far as possible, correct several published misidentifi cations, and clarify nomenclatorial problems regarding the authority of Stygopholcus and the identity of the type species S. absoloni. We suggest that the ‘northern clade’ has a relict distribution, resulting from past and present geologic and climatic factors. Future work on Stygopholcus should focus on the southern Dinarides, combining dense sampling with massive use of molecular data.

Among these, the Balkan Peninsula is the only European region with an endemic pholcid genus, Stygopholcus. This is noteworthy but unsurprising. The western part of the peninsula in particular is widely recognized as one of the regions in Europe with the highest species diversity (Griffi ths et al. 2004), and for many taxa the level of endemism is the highest in Europe (e.g., Gottstein-Matočec et al. 2002;Eastwood 2004;Tzedakis 2004;Džukić & Kalezić 2004;Sket et al. 2004;Mammola et al. 2019;Bregović et al. 2019;Lukić et al. 2020). Several factors have been proposed to account for this exceptional biological richness, including historical aspects (glacial refugia, land bridges) and present geography (terrain complexity, transition zone), geology (karst, caves), and climate variability (Mediterranean, mountainous, and continental) Reed et al. 2004;Džukić & Kalezić 2004;Mihevc et al. 2010).
Some spider genera share this exceptional species diversity on the Balkan Peninsula, as for example Troglohyphantes Joseph, 1882, Harpactea Bristowe, 1939, and Dysdera Latreille, 1804(Deeleman-Reinhold 1978Deeleman-Reinhold & Deeleman 1988;Deltshev 2004;Komnenov 2009Komnenov , 2020Platania et al. 2020). By contrast, Stygopholcus is not particularly diverse. Three of the four species we recognize below have been known since Kratochvíl's (1940) diligent revision, and the fourth species was added half a century ago by Senglet (1971). Stygopholcus spiders are large and conspicuous; in the western Balkan they are among the fi rst animals any biologist will encounter upon entering a cave. Thus, we have no reason to expect that a substantial number of species remain to be discovered. Nevertheless, we deemed this revision necessary for three main reasons.
First, species limits are not always obvious. For example, the similarity between the two species described by Kratochvíl (1940) prompted Senglet (1971) to reduce them to subspecies. Numerous misidentifi cations in the literature are further testimony to the diffi culty in distinguishing some of the species, in particular females. Our revision is based on many more specimens from many more localities than available to previous authors, and we additionally support species limits with morphometric data. In addition, we point out deviating specimens from particular caves or areas that we tentatively assign to known species but that deserve further study.
13.1.1 for names published after 1930 to be available: "… be accompanied by a description or defi nition that states in words characters that are purported to differentiate the taxon." The fact that Absolon & Kratochvíl (1932) contains a bibliographic reference to such a published statement could be construed as satisfying ICZN Article 13.1.2. We do not know which 1932 paper was published fi rst. However, if Kratochvíl (1932) was published fi rst, it has priority for that reason; if it was published second, it has priority because Absolon & Kratochvíl (1932) refer to a statement that was not yet published at the time.
Confusingly, the paper to which Absolon & Kratochvíl (1932) refer was originally also not intended to be the source of the name. Kratochvíl (1932) referred to "Stygopholcus n. g. in lit., jejž popíši na jiném mistě" ["Stygopholcus n.g. in lit., that I will describe elsewhere"], suggesting that the genus name was intended to be made available in an upcoming publication (possibly Kratochvíl 1934). However, Kratochvíl (1932) provides "characters that are purported to differentiate the taxon", satisfying ICZN Article 13.1.1; and he explicitly names "H. absoloni Kulcz." as the only species to be included in the genus, satisfying the requirements of ICZN Article 13.3 regarding the fi xation of a type species. From a formal perspective, Kratochvíl's (1932) intentions (of making the name available later) are as irrelevant as the confusion between genus rank (in the Czech text) and subgenus rank (in the French summary), and the minimalist nature of his diagnosis: he mentions that (1) the female sternum process present in Holocnemus pluchei is absent in Stygopholcus; (2) the abdomen shape differs between Holocnemus and Stygopholcus; (3) the female palp of Stygopholcus resembles that of Pholcus rather than that of Holocnemus pluchei (which is enlarged); and (4) Stygopholcus has cheliceral stridulation like Holocnemus (in contrast to Pholcus).
We conclude that Stygopholcus was made available by Kratochvíl (1932). (Kulczyński, 1914) In the original description of "Holocnemus (Hoplopholcus) Absolonii", Kulczyński (1914) mentions two caves near Trebinje by name: "Mares et feminae lecta sunt prope Trebinje in speluncis: Ilijina pećina … et Zovica jama…" ["males and females were collected in caves near Trebinje: Ilijina cave … and Zovica pit …"]. He then notes that according to Absolon ("Teste Cel. C. Absolon"), this species has been found in many other caves in southeastern Herzegovina (southern region of modern Bosnia and Herzegovina). This strongly suggests that he had only seen the specimens from the two caves near Trebinje. Any specimens he had from these caves, and only these specimens, constitute the type series (syntypes).

Identity of Stygopholcus absoloni
In Kulczyński's (1914) long original text, the morphology is described in considerable detail but complex shapes are notoriously diffi cult to understand from text alone. However, he was obviously aware of differences among the specimens available to him, starting a paragraph with "Maris palpi variant paululo formâ, etiam mandibularum armatura paulo mutabilis" ["the male palps vary slightly in shape, and also the armature of the chelicerae is slightly variable"]. He clearly interpreted this as intraspecifi c variation, but without specifying to which variant(s) his four fi gures referred. Kratochvíl's (1940) error was that he felt either entitled or obliged (or both) to designate new types from a new type locality for S. absoloni, choosing a cave ~35 km NW of Trebinje, "Grabova peć kod Grabova dola". We can only speculate about his motivation, but it seems reasonable to assume that he had been postponing his detailed revision of the genus for many years (at least since 1932), hoping to eventually be able to see Kulczyński's type series. This hope was frustrated, as indicated in a footnote (Kratochvíl 1940: 16) where he explicitly stated that the specimens from Ilijina pećina and Zovica jama were not accessible to him. Unfortunately, among his rich new material there were no specimens from any of these two caves.
We made a major effort to locate Kulczyński's types, but found only one vial, deposited in the Museum & Institute of Zoology, PAS, in Warsaw, Poland. The supposed types of S. absoloni deposited in Prague and listed in Růžička et al. (2005) are Kratochvíl's erroneous 'types' from "Grabova peć kod Grabova dola". Kulczyński's true type vial contains a label in his own handwriting, but the only locality information on this label is "Trebinje". A second, very detailed label says "Ilijina pećina", but other information on this label leaves no doubt that it was added much later. Fortunately, Kulczyński's original catalogue (in Warsaw) offers a clear hint about the origin of the specimens in the vial. It says "Jaskinie koło Trebinje" ["caves near Trebinje"], suggesting that Kulczyński joined the specimens from Ilijina pećina and Zovica jama.
This type vial contains only three specimens, one male and two females. This is clearly only part of Kulczyński's material, because he obviously had at least two males available. However, both variants/ species are represented in this vial. Circumstantial evidence suggests that one female is from Ilijina pećina (it corresponds to more recently collected females from this cave). This forces us to conclude that the other specimens (one male and one female) are from Zovica jama. This is clearly a case where a lectotype should be designated for S. absoloni in order to provide stability and avoid further confusion about the basics (ICZN Recommendation 74G). However, it is not obvious which specimen would best serve as a lectotype. Selecting the male specimen has one major disadvantage: it makes Zovica jama the type locality, a cave that we were not able to locate, and that was even unknown to local cavers in Trebinje we met in 2014. On the other hand, Ilijina pećina is well known and easily accessible, providing a precise type locality. Selecting the female from Ilijina pećina would thus seem to better satisfy ICZN Recommendation 74E (Verifi cation of locality).
At the same time, selecting the male specimen as lectotype has also advantages. Most importantly, it preserves the prevailing usage of the name. This male is the same variant/species as Kratochvíl's (1940) S. absoloni ʻtypesʼ from "Grabova peć kod Grabova dola", and it also corresponds to what Senglet (1971) interpreted as S. absoloni. Thus, selecting the male satisfi es ICZN Recommendation 74A (Agreement with previous restriction). A second (minor) advantage is the fact that species in Stygopholcus are slightly easier to distinguish by males than by females (as in most Pholcidae).
Thus, we decided to select the male as lectotype, mainly for two reasons: it seems to better serve stability, one of the major guiding principles of the Code. Second, it is possible that the exact location of Zovica jama will eventually be determined, weakening the argument against selecting the male.
COLOR. In general ochre-yellow to brown, with distinct dark pattern in S. photophilus, without or with reduced dark pattern in other (troglophile) species. Carapace mostly pale, with darker ocular area and posterior triangle (Figs 1-2, 4); sternum either dark brown to black (S. photophilus), or light brown, with darker brown radial marks (other species). Legs in S. photophilus with dark rings on femora (subdistally) and tibiae (proximally and subdistally) and with short dark longitudinal lines dorsally on femora ; in other species monochromous, without dark rings and longitudinal lines. Abdomen in S. photophilus with distinct dorsal and ventral patterns; in other species with fewer dorsal and lateral dark marks restricted to posterior part and without or with indistinct ventral pattern.

Relationships
Stygopholcus is morphologically very similar and geographically very close to Hoplopholcus (Huber 2020). This led to a debate between A. Senglet and P. Brignoli in the 1970s and 80s, on whether the two genera should be synonymized or not (Senglet 1971;Brignoli 1971Brignoli , 1976Brignoli , 1983. In the end, Senglet (2001) presented a series of morphological differences, and since version 2.5 of the World Spider Catalog (Platnick 2002) Stygopholcus has been listed as a valid genus.
The most recent molecular phylogeny of Pholcidae (Eberle et al. 2018) has suggested that Stygopholcus and Hoplopholcus are not even sister taxa, placing Stygopholcus closer to Crossopriza than to Hoplopholcus (with a 'reasonable' bootstrap support of 87). Morphological characters that support a sister-group relationship between Stygopholcus and Crossopriza are the female stridulatory apparatus between prosoma and abdomen and the dark lines on the legs. However, a convincing morphological analysis will require a detailed study of Crossopriza (a revision is in preparation) and of Holocnemus. The type species of Holocnemus (H. pluchei) was not included in the molecular analysis of Eberle et al. (2018), and the two species of Holocnemus that were included (H. caudatus and H. hispanicus) may in fact be misplaced in Holocnemus and belong to Crossopriza (B.A. Huber, unpubl. data).
Within Stygopholcus, there is strong evidence that S. photophilus is sister to all other species (the 'northern clade'). Both a six genes approach (Eberle et al. 2018) and a study using hundreds of universal single-copy orthologs (L. Dietz et al., unpublished data) consistently recovered this relationship, and numerous morphological characters are shared by all species except S. photophilus. Some of these are likely to be synapomorphies: (1) spines on male femur 1 arranged in two rows rather than in just one ( Within the 'northern clade', there is equally strong support for the relationship absoloni + (skotophilus + montenegrinus). Several morphological characters shared by S. skotophilus and S. montenegrinus support this view, even though it is not clear yet if all of these shared similarities are synapomorphies: (1) procursus straight (rather than weakly curved toward femur as in S. absoloni) (Fig. 44); (2) distance between tips of dorsal bulbal process (bulbal measure D; Fig. 17) > 0.24 (< 0.24 in S. absoloni) (Fig. 159); (3) size of bulb (bulbal measure E; Fig. 15) > 0.80 (< 0.78 in S. absoloni) (Fig. 160).

Natural history
Stygopholcus photophilus can be classifi ed as an epigean species. It tolerates direct sunlight, is often found in exposed webs among the vegetation, and enters caves only occasionally. Compared to epigean relatives in the genera Holocnemus, Hoplopholcus, and Crossopriza, it does not show any obvious troglomorphism. By contrast, all other congeners are slightly troglomorphic. While eye size and leg length do not seem to indicate troglomorphism, their general coloration is paler and they have fewer dark marks on the abdomen and legs. This refl ects their strong preference for dark sheltered spaces like caves, deep fi ssures and cracks, and spaces under rocks in forests. Most records of these species are from caves, but this is probably due to the fact that access to caves is usually much easier for collectors than access to other shallow subterranean habitats.
According to Kratochvíl (1940), cave-dwelling Stygopholcus prefer dry caves, or dry areas of caves that include humid sections. We do not have quantitative data on this, but our impression was that Stygopholcus requires high humidity; specimens were occasionally found even on dripping wet rocks. They seem to prefer the twilight area, avoiding both direct sunlight at the cave entrance and complete darkness. As already noted by Kratochvíl (1940), individual specimens are occasionally found in deeper parts of caves. As usual in Pholcidae in general, they avoid caves or cave sections with strong air fl ow (Kratochvíl 1940).
Beyond basic habitat information, little is known about the biology of Stygopholcus spiders. Adult specimens of representatives of the northern clade were found throughout the year (20 of 186 collecting events between December and February); we have no records of S. photophilus from January and February. Our limited observations (MK, unpublished data) suggest that representatives of the northern clade produce egg-sacs between May and August. In a cave near Podgorica in Montenegro, the last author observed a mating of S. montenegrinus where the female kept holding on to her egg-sac while mating. Senglet (2001) observed mating in S. photophilus and studied genital mechanics by freezefi xing mating pairs. He described rhythmic palpal movements during copulation, sperm uptake with a single thread held between legs 3, and was able to ascribe specifi c functions to individual genitalic structures.

Composition
The genus includes four named species and this number is unlikely to change substantially in the future. However, some specimens assigned tentatively to S. absoloni, S. skotophilus, and S. montenegrinus, respectively, may represent separate species. A convincing resolution of species limits will probably require both a denser sampling and a massive use of molecular data.

Distribution
The genus is restricted to the Mediterranean part of the Balkan Peninsula, ranging from southern Croatia to Crete (Fig. 167). The epigean S. photophilus has a wide distribution, ranging from southern Albania to Crete. The three troglophile species of the 'northern clade', S. absoloni, S. skotophilus, and S. montenegrinus, range from southern Croatia to northern Albania (Fig. 168). They seem to be relicts that have survived further north by adapting to subterranean conditions (see Discussion).
By contrast, all caves listed in Kratochvíl (1935) probably contain S. skotophilus; this is either confi rmed by new material or very likely judging from the distribution of confi rmed specimens.
COLOR (in ethanol). Carapace pale whitish to ochre-yellow, ocular area and clypeus slightly darker (light brown); sternum light brown, with darker brown radial marks; legs ochre yellow, without dark rings, without dark lines; abdomen monochromous pale gray, only above spinnerets with darker mark. Fig. 1. Ocular area slightly raised. Carapace with deep median pit and pair of very shallow indistinct furrows diverging from pit to posterior margin of carapace. Clypeus unmodifi ed. Sternum wider than long (1.50/1.00), unmodifi ed. Abdomen oval, pointed at spinnerets. Figs 19-20, with pair of large frontal apophyses set with 2 and 3 modifi ed hairs, respectively, with 17 and 18 modifi ed hairs each on frontal face; with distinct stridulatory ridges (Fig. 35).
LEGS. With short spines ventrally on femur 1 in two rows, prolateral row ~19 spines, retrolateral row ~24 spines; without curved hairs; few vertical hairs; retrolateral trichobothrium of tibia 1 at 3%; prolateral trichobothrium absent on tibia 1, present on other tibiae; tarsus 1 with irregular platelets rather than with distinct pseudosegments. In the single male from Jama na Brezdanu (Croatia), the bulb differs in several respects from all other material examined, but it is unclear if this is teratological or not (even though it is symmetric). Most conspicuously, the heavy sclerite connecting the basal and distal bulbal sclerites (asterisk in Fig. 18) is much smaller and it is not connected to the basal bulbal sclerite; bulb measure E also deviates strongly (asterisk in Fig. 160).

Female
In general similar to male, but femur 1 without spines, with pair of indistinct processes posteriorly on carapace (arrows in Fig. 39) and pair of opposing poorly visible plates on abdomen; distance between processes/plates: 1.  Fig. 153; relatively shorter and wider than in other species), with large pore plates (Fig. 22).

Remark
According to Růžička et al. (2005), the type vial originally contained two labels saying "Stygopholcus skotophilus sp. n." and "Pećina u Ivici, č. III". This agrees with other labels in Kratochvíl's handwriting in the NMP Stygopholcus material. Both labels are lost. Růžička et al. (2005) further state that the vial contained "1♂ 3♀ (one of them without epigyne) 4 juv.", suggesting that there were two epigyna. However, the vial does not contain any epigynum. Possible record (not examined) The following record in Kratochvíl (1940) is dubious because he assigned his specimen(s) to S. absoloni, but specimens from the closest caves we examined (Debelin pećina, Kali pećina, Špilja za Gromačkom vlakom) were all unambiguous S. skotophilus. We thus assume that Kratochvíl misidentifi ed his specimen(s) from this cave.

Material assigned tentatively
The material listed here originates from caves around Sedlari in Bosnia and Herzegovina. This material has been studied before by Kratochvíl (1940) and Senglet (2001), and both authors assigned most of the specimens to S. skotophilus (only the female from Medja pećina was assigned to S. absoloni by Kratochvíl 1940). However, both males and females from these caves show some small but consistent differences to S. skotophilus (see section on variation below) so the material is here listed apart to draw attention to this potential separate species (of which no fresh material is available).

Male (variation)
Males from most other localities paler, without lateral dark marks on abdomen, with fewer median dark marks (restricted to area above spinnerets), rarely without abdominal marks. Males from near Sedlari differ slightly in bulbal measure A smaller (0.25-0.28) and bulbal measure B smaller (0.07-0.08); in addition, the smaller wart on the retrolateral bulbal process (arrow in Fig. 48) is more strongly developed.

Female
In general similar to male, but femur 1 without spines, with pair of indistinct processes posteriorly on carapace and pair of opposing poorly visible plates on abdomen; distance between processes/plates: 1.  (Figs 53-54, 147-148) with distinctive ventral arc ( Fig. 154; approximately as long as wide), with large pore plates (Fig. 54).
Females from near Sedlari differ in having the parallel sclerites posteriorly in the female internal genitalia closer together (epigynal measure F: 0.32-0.36; Figs 59-60) and slightly more distinct.

Female
In general similar to male (Fig. 3), but femur 1 without spines, with pair of indistinct processes posteriorly on carapace (Fig. 107) and pair of opposing poorly visible plates on abdomen, distance between processes/plates: 1.  (Figs 85-88, 111), with internal structures visible in uncleared specimens: round anterior structure, short sclerites originating from posterior margin and slightly converging anteriorly; posterior plate short but wide. Internal genitalia (149)(150) with distinctive ventral arc ( Fig. 155; approximately as long as wide), with large pore plates (Fig. 84).
BODY. Habitus as in Fig. 4. Ocular area slightly raised. Carapace with deep median pit and pair of very shallow indistinct furrows diverging from pit to posterior margin of carapace. Clypeus with very indistinct modifi ed area medially (arrow in Fig. 129) and sclerotized margin. Sternum wider than long (1.05/0.70), unmodifi ed. Abdomen oval to short cylindrical.
PALPS. As in Figs 9-11; coxa with prominent retrolateral hump; trochanter barely modifi ed; femur with distinct retrolateral process proximally, with stridulatory pick (modifi ed hair) proximally on prolateral side, with transversal dark line retrolaterally, distally strongly widened and with rounded ventral protrusion; femur-patella joints shifted towards prolateral; tibia very large, thickest at two thirds, with two trichobothria; tibia-tarsus joints shifted towards retrolateral; procursus dorsally with ~5 weakly curved hairs, with strong hump on prolateral side. Procursus tip (Figs 113-114 Kratochvíl, 1932. Pale blue and pale orange marks denote specimens that are assigned tentatively to the respective species. Abbreviations: Alb = Albania; BaH = Bosnia and Herzegovina; Cro = Croatia; Mon = Montenegro. sclerite connecting to tarsus, small dorsal process weakly sclerotized, and main sclerite consisting of retrolateral process and dorsal process; retrolateral process distally widening, distal margin with pair of heavily sclerotized bulges on ventral side; dorsal process with single slightly curved tip, proximally with two small apophyses.

Distribution
Widely distributed in Greece (including Crete and the Ionian Islands) and southern Albania; the species seems to be largely absent from the Aegean Islands (except Thasos) (Fig. 167).

Stygopholcus montenegrinus -species or subspecies?
Stygopholcus montenegrinus was originally described as a full species (Kratochvíl 1940) but later reduced to a subspecies of S. skotophilus by Senglet (1971). Senglet's argument focused on the small morphological differences between the two former species. While this observation is certainly correct, the conclusion is less evident. Countless species differ from the most similar species by minute differences. The most consistently applied criterion among arachnologists for separating species seems to be the presence of a gap, usually morphological, irrespective of how big the difference is. Kraus (2000) suggested that such gaps (usually in genital morphology) are almost universal in spiders, explaining the low number of subspecies in araneology. Kraus (2000) certainly played down the problem of intraspecifi c genital variation ("There is practically no variation … [in] … the taxonomically decisive genital structures"), but it is a widespread pattern that genital structures are less variable within species than other morphological traits (Eberhard et al. 1998). As a result, there is less or no overlap among species, usually producing distinct gaps that facilitate taxonomic decisions. However, the majority of spider species are known from minimal samples, often from only one locality. Many of the observed gaps may thus be artifacts, resulting from insuffi cient knowledge. As a logical consequence, it has been argued that subspecies in general "cannot be known until the species is known" (Blackwelder 1967). Subspecies are arguably easier to justify in well-studied taxa like mammals, birds, and butterfl ies than in spiders.
In the case of Stygopholcus skotophilus and S. montenegrinus, our data, even though still limited, suggest that there are several small but consistent morphological (genital) differences and that there are no intermediate forms in the region where the two taxa come close to each other (they never share a locality). In addition, a multi-gene analysis of a limited sample of specimens resolved each taxon as monophyletic (L. Dietz et al., unpubl. data). This makes both taxa diagnosable, supposedly reproductively isolated, independent evolutionary lines, satisfying the main criteria of several major species concepts (Wheeler & Meier 2000).

Remarks on distribution
The natural distribution of Smeringopinae (i.e., excluding synanthropic and anthropophilic species) ranges from South Africa to the Iberian Peninsula, the Balkan Peninsula, the Caucasus, and Central Asia (Huber 2011b). Stygopholcus thus marks the northern limit of Smeringopinae in Europe. In the absence of a reliable dated phylogeny of Stygopholcus and its putative closest relatives (Crossopriza, Holocnemus) (Eberle et al. 2018), the biogeographic history of Stygopholcus remains somewhat speculative. However, patterns paralleled in other groups give some indications about plausible scenarios.
First, the large majority of present-day Smeringopinae live in tropical and subtropical environments. Part of the present-day fl ora and fauna of the Balkan Peninsula is thought to constitute a 'Palaeo-Mediterranean' contingent that evolved under pre-Pliocene tropical climatic conditions (Scarascia-Mugnozza et al. 2000). It is thus plausible that Stygopholcus is derived from ancestors that were present in the area before the Pliocene, in climatic conditions similar to those that its closest relatives still experience today.
Second, a relatively warm climate was maintained in the Balkan Peninsula during Pleistocene glaciations, particularly in the western part: glaciations in this area affected only higher elevations; orographic barriers prevented winter intrusions of cold polar air; and the Adriatic Sea had a moderating effect on temperatures along the coast (Mihevc et al. 2010). This made the Balkan Peninsula a refugium for many taxa that recolonized neighboring regions in the north after the last glaciation (Hewitt 2000). At the same time, however, glaciations resulted in drier climate (Sinopoli et al. 2019), and epigean ancestors of some present-day troglophile and troglobite taxa may have disappeared during this time. It is thus plausible that the epigean ancestor of the present-day northern clade of Stygopholcus disappeared during this period, and only descendants that had adapted to subterranean life survived. However, earlier climaterelated phenomena could have had the same effect, i.e., Stygopholcus may have adapted to subterranean conditions earlier than the Pleistocene. For example, a dry period at the end of the Miocene (Hsü et al. 1973) that is thought to have affected the salax group of the spider genus Troglohyphantes (Deeleman-Reinhold 1978); or the disappearance of Lake Pannon and the bordering Dinarid Lake System, which is thought to have caused the extinction of all epigean species of the bivalve genus Congeria (Bilandžija et al. 2013).
Third, taxa that exhibit a so-called 'mero-Dinaric' distribution may shed some light on the northern limit of Stygopholcus. The term mero-Dinaric was introduced for taxa that show either a northwest or southeast distribution in the Dinarides or a disjunct distribution, with the gap consistently in the same region in the central Dinarides (Sket 1994). Examples of mero-Dinaric taxa are species groups of the genus Troglohyphantes (Deeleman-Reinhold 1978), several genera of the spider family Dysderidae (Deeleman-Reinhold 1993;Pavlek & Mammola 2021), and the diplopod subfamily Acherosomatinae (Antić et al. 2015). The northern limit of Stygopholcus matches closely to this gap, suggesting that the factors affecting mero-Dinaric taxa may also affect or have affected Stygopholcus. Reasons for this exclusive northwest or southeast distribution are largely unknown. One possible explanation is that in this area, much of the carbonate rocks that were exposed between ~3-6 Ma remained covered by impermeable fl ysch sediments for longer than in the neighboring areas, impeding access to the moistureretaining limestone habitat (Deeleman-Reinhold 1978;Sket 1994;Mihevc et al. 2010). This, together with decreasing temperatures towards the north may explain the northern limit of Stygopholcus, in spite of the ubiquity of suitable habitats (including caves) that would seem to allow for more northern occurrences and that have been sampled extensively.
Fourth, the Dinarides can be seen as ultimately continuing into Greece and the Greek islands, but a major distinction can be made in geological terms across the transition known as the Scutari-Peć transverse zone which runs through northern Albania, south of which the geology is dominated instead by volcanic or metamorphic rocks (Kissel et al. 1995). This geological line coincides closely with the current southern limit of the northern clade of Stygopholcus, suggesting that the presence of karst may have been a crucial factor in the evolution and survival of Stygopholcus in the western Balkan. However, Albania is relatively poorly sampled, so we cannot exclude the possibility that the southern limit of the northern clade is just a sampling artifact.
In summary, the northern clade of Stygopholcus may have been derived from an epigean ancestor that lived in the area before the Pliocene when the climate was more tropical or subtropical; during historical climatic perturbations the spiders survived because they could access exposed karst habitats in the southwest Dinarides; access to such habitats was presumably not possible further north and further south, for different reasons. The only surviving epigean congener, S. photophilus in Greece, may have largely retained the ancestral mode of life.