Faunistic and zoogeographic analysis of sac and ground spiders (Cheiracanthiidae, Clubionidae and Gnaphosidae) in mainland Greece

Abstract. The spider families Cheiracanthiidae, Clubionidae and Gnaphosidae from areas of mainland Greece are examined. In total 72 species were recorded, belonging to Cheiracanthiidae (1 species), Clubionidae (1 species) and Gnaphosidae (70 species). Marinarozelotes cumensis (Ponomarev, 1979) is recorded from Greece for the first time. Habitat preferences of the dominant species are discussed. A zoogeographic analysis focusing on those areas where sampling was more complete, reveals conspicuous similarities to both the European and the eastern spider fauna. Zusammenfassung. In diesem Artikel werden die Spinnenfamilien Cheiracanthiidae, Clubionidae und Gnaphosidae von Flächen auf dem griechischen Festlandes untersucht. Es wurden insgesamt 72 Arten gefunden, die zu den Cheiracanthiidae (1 Art), Clubionidae (1 Art) und Gnaphosidae (70 Arten) gehören. Marinarozelotes cumensis (Ponomarev, 1979) wird erstmals für Griechenland nachgewiesen. Die Habitatpräferenzen der dominanten Arten werden erfasst. Eine zoogeografische Analyse, die für die Flächen mit relativ vollständiger Erfassung durchgeführt wurde, weist auffällige Ähnlichkeiten sowohl zur europäischen als auch zur östlichen Spinnenfauna auf.

Greece is a biodiversity hotspot for Europe (Cuttelod et al. 2008). Concerning, for example, spiders 1298 species have been recorded from the country with its total surface of 132000 km 2 . This is a relatively high species number with respect to area when compared to other South European countries like, for example, Italy (301000 km 2 , 1720 species), Spain (498000 km 2 , 1441 species -without the Canary Is.), Portugal (88000 km 2 , 860 species -without Madeira and Azores), and France (544000 km 2 , 1702 species) (Nentwig et al. 2022). As part of the southern Balkan Peninsula, Greece also has a key biogeographical position, being located at the edge of two continents: Asia and Europe. The northern parts of the country in particular form a consequential zoogeographical connection between European and Eastern faunas, therefore the study of diversity in this area is essential for the understanding of dispersal paths and distributional limits of animals originating from either of the geographical parts. Because of its paleogeography, topography, climate and habitat heterogeneity, Greece is very diverse in terms of local biogeographical patterns (see Kougioumoutzis et al. 2021 and references therein). For instance, in northern areas, European and widespread spider species dominate (Schröder et al. 2011), as well as Eastern species (e.g., East Mediterranean and Ponto-East-Mediterranean) (Komnenov et al. 2016). In Thrace, the south-eastern distribution limit of many north or central European species is found (e.g. Arctosa stigmosa, Agroeca lusatica, Xysticus gallicus). Similarly, the western distribution limit of species originating from eastern parts (Asian or Pontic) is found here (e.g. Zelotes solstitialis, Xysticus xerodermus, Tita noeca turkmenia, Tegenaria angustipalpis and others) (Nentwig et al. 2022). On the other hand, in southern areas and in insular parts of the country, Mediterranean species dominate, and higher rates of endemism are evident, with the most striking case met in the arachnofauna of Crete (Bosmans et al. 2013).
However, there is still a lack of evidence in the spider catalogue and for distributions across Greece. A similar situation exists in neighboring countries with important biogeographic impact on Greece (e.g. Turkey), rendering distributional data incomplete and difficult to analyse. As a result, biogeographical studies based on spiders in this area may only be considered as preliminary estimates of biogeographic trends along the main axis of the country, i.e. north-south, east-west, or between mainland and insular parts.
Notwithstanding the extensive surveys on the arachnofauna of some of the insular parts of Greece, e.g., Crete (Chatzaki et al. 2002a, 2002b, 2003, Bosmans et al. 2013, Lesbos (Bosmans et al. 2009), Chios (Russell-Smith et al. 2011) and the Dodekanese (Chatzaki & Van Keer 2019)), the spiders of mainland Greece, and especially the northern parts of the country, have largely escaped the interest of arachnologists. To this end, an effort was made to focus on spiders of the least known areas of Greece and hence an extensive sampling survey was organized in previous years including pitfall trapping and hand collecting in many parts of the country's mainland. This paper represents part of a series of taxonomic and ecological studies (i.e., Komnenov et al. 2016, Chatzaki 2018, Zografou et al. 2017, Pitta et al. 2019) dealing with the large material collected in this framework. It focuses on the sac and ground spiders of the families Cheiracanthiidae, Clubionidae and Gnaphosidae and results in 72 species records, one of which is a new record for Greece. These species are further analyzed in order to draw some preliminary results about their ecological and biogeographical patterns.

Material and methods
Material examined here was obtained between 2014 and 2015 by pitfall trapping and hand collecting from 88 sampling sites, in the framework of a biodiversity project (SPIDOnetGR) centered in eastern Thrace (Greece) and extending to many other parts of mainland Greece (i.e. Makedonia, Ipeiros, Thessalia and Sterea Ellada). Coordinates, habitat details and dates of sampling at the 88 sites are given in Tab. 1 and are shown in Fig. 1 The choice of sampling plots depended on habitat dominance in the area. More specifically, in the Evros district pitfall traps were set in grasslands (mostly natural, but also some abandoned old arable lands), maquis and mixed forests (Quercus, Ostria, Juniperus and Pinus) of variable openness. In the Evros Delta, pitfall traps were set in sand dunes, on riverbanks, in rushy marshes and meadows. At the Prespes lake, the traps were set in agricultural areas as well as in open oak, beech and juniper forests and dry grasslands, while litter sieving and hand collecting also took place at ruderal stripes of similar habitats. In Magnisia, pitfall traps were set in phrygana and maquis of variable density and in oak, pine and beech forests. In Tab.1 the time period during which pitfall traps were active is indicated, while all other collecting methods are indicated by a single date.
Species identification took place at the Natural History Museum of Basel, Switzerland and was finalized in the Department of Molecular Biology & Genetics of the Democritus University of Thrace. The nomenclature and distributional data used for the biogeographical analysis follows the WSC (2022). All material of species here presented, are deposited at the Natural History Museum of Crete (NHMC hereafter), curator Maria Chatzaki.
The analysis included number and dominance of spider species as well as comparison of patterns of different geographical ranges across the study area. Accumulative data from the whole inventory were initially used. Furthermore, data obtained from areas which were considered more thoroughly surveyed because of the higher number of sites sampled in them, were analysed separately. The latter case involves the Evros district (divided into the Delta of Evros river -Sites 23-28, six sites in total -and in the rest of the Evros district -Sites 1-22, twenty two sites in total -in Thrace), the Magnisia district (in Thessalia) -Sites 75-83, nine sites in total -and the area surrounding the Greek part of Prespes Lake (in Makedonia) -Sites 29-39, eleven sites in total. Because the area of the Evros Delta was extensively sampled and its habitats are distinct from the rest of sampling sites in the district of Evros (see habitat details in Tab. 1), it was considered as a different "area". To better illustrate zoogeographic trends, zoogeographic elements (as given by Komnenov (2014)) were attributed to five general categories as follows (see also Fig. 3a): WIDE (i.e. species with widespread distribution): EMA = Euro-Middle Asiatic, EMMA = Euro-Mediterranean-Middle Asiatic, EUA = Eurasian, HOL = Holarctic, PAL = Palearctic, SEM = South-European-Middle Asiatic; EU (i.e. species with mainly European distribution): EUR = European, EUS = Euro-Siberian, EUC = Euro-Caucasian, ECA = Euro-Carpathian, SEU = South-European; MED (i.e. species with mainly Mediterranean distribution): MED = Mediterranean, MMA = Mediterranean-Middle Asiatic; EAST (i.e. species with mainly Anatolian distribution or limited to the eastern parts of Europe): EME = East-Mediterranean, EMI = East-Mediterranean-Iranian; STENO (i.e. species with limited distribution within the national borders or in the nearby areas): GRE = Greek, AEG = Aegean, BALK = Balkanic, CBAL = Carpatho-Balkanic, PON = Pontic, PEM = Ponto-East-Mediterranean. Furthermore, recent faunistic data from Komnenov et al. (2016) and Chatzaki (2021) were added in the original data provided here, as they both origi-nate from the same sampling areas and inventories. This resulted in an overall dataset of 83 species used in the analysis. Chatzaki, 2002 Material. Site 2 (a: 2 )), 2 ((; b: 1 ), 3 ((; c: 1 (), Site 19 (a: 1 ); b: 3 ))), Site 22 (a: 2 )); b: 1 )).

Ecology and zoogeography
Gnaphosidae represent ca. half of the total spider samples collected in this inventory, resulting in 3795 adult individuals (of which, 1748 were in the Evros gradient, 1147 in the Evros Delta, 226 in Magnisia and 166 in Prespes, and the rest at all other sites). The genera and species recorded in each area, as well as the total number of corresponding specimens caught are shown in Tab. 2. In Fig. 2 the dominant Gnaphosidae species in terms of numbers of specimens (four to five and up to nine species in the Evros district), each covering 4-29% of the local Gnaphosidae fauna, are shown. Additionally, Dras syllus villicus is present in very high numbers (340 individuals, 9% of the total number of specimens collected), mainly at the sites where oak or mixed deciduous forests prevail (e.g. sites 53, 61, 65 and others). Trachyzelotes pedestris is a dominant species at three of the four mentioned areas and Civizelotes caucasius is dominant in all of them (except for the area of the Evros Delta where very specific habitats are found -see Materials and methods). Anagraphis ochracea, Berinda infumata and Drassyllus centrohellenicus were exclusively recorded from Magnisia. Zelotes olympi, Z. apricorum, Micaria pulicaria and Haplodrassus sylvestris were sampled only in Prespes. Fortyfive species were found only in Evros, of which nine are exclusively recorded from the Evros Delta.
The zoogeographic characterization attributed to the species recorded is shown in Tab. 2 and their representation is illustrated in Figs 3 (cumulative data), and 4-5 (patterns and percentages presented per focal region).
In the category STENO (30% of the total inventory), nine species are Greek endemics and three are endemics of the specific Aegean region, while the rest of the species in this category (twelve) are endemics of the wider Balkan Peninsula and Pontic area (Fig. 3). In the category WIDE (28% of the total inventory), nine species present a Euro-Middle Asiatic distribution, four species present a Euro-Mediterranean-Middle Asiatic distribution, and another ten species are of wider distribution. The EU category (14% of the total inventory) is dominated by South-European (four) and Euro-Caucasian (four) species. The EAST category (14% of the total inventory) is composed of seven East-Mediterranean and five East-Mediterranean-Iranian species. Finally, the MED category (14% of the total inventory) is dominated by Mediterranean-Middle Asiatic (seven) and Mediterranean (five) species.
Widespread elements are the dominant category in all four regions examined separately, while Eastern elements are equally dominant in Evros (24%) and Magnisia (27%) (Fig.  5). Although dominant in the overall inventory, stenoendemics present a high percentage only in the two areas of the Evros district (19% and 20%), but they are the least represented category in the other two regions (9% and 14%). European elements range between 13% (in the Magnisia district) and 26% (in Prespes) and Mediterranean elements range between 9% (in Prespes) and 17% (in the Evros Delta).

Discussion
This paper presents the results of a taxonomic and faunistic work focusing on ground spiders of the least studied areas in mainland Greece, e.g., Thrace, Makedonia, Ipeiros, Thessalia and others. In total, 72 species were recorded. Since the publication of earlier papers dealing with the spiders of Thrace (Komnenov et al. 2016, Chatzaki 2018) and other areas of northern Greece (Chatzaki 2021), 21 Gnaphosidae species have been newly reported for the region (see Tab. 2), revealing that knowledge on spider faunas of these areas is still incomplete.
Marinarozelotes cumensis is the single new record for Greece. This species was known to occur in Ukraine, Russia (European part), Azerbaijan and Kazakhstan (Nentwig et al. 2022), and hence its presence in Greece also represents the first record from southern Europe. Further evidence from possible new recordings in the other Balkan countries could confirm its distributional expansion to SE Europe and would render its chorological characterization from Pontic to Ponto-East Mediterranean. A full revision of the species is required (particularly due to the current geographical gap in the range of occurrence), including comparative taxonomic notes, and will represent part of another paper (Chatzaki in prep.).
The number of species recorded exclusively from each of the studied areas demonstrates the diversity of the arachnofauna of the area, mainly correlated with the habitat hetero-  common throughout the country, both insular and mainland. All have wide distributions, either ranging from Europe to Central Asia (i.e. D. praeficus, D. villicus, T. pedestris) or presenting a more Mediterranean distribution (i.e. C. caucasius, M. malkini and M. lyonneti, see Tab. 2). Usually they occur in large numbers when present.
Civizelotes caucasius is a dominant species in grasslands, shrublands and generally open areas (Chatzaki et al. 2003, Zografou et al. 2017, Pitta et al. 2019, presenting a peak of activity in mid-summer (Chatzaki et al. 2005, Zografou et al. 2017. The combination of this activity pattern and preference to open habitats clearly demonstrates the xerophilic character of this species. For this reason Zografou et al. (2017) concluded that changes in agricultural land use, such as afforestation of abandoned fields, might limit its suitable habitats, therefore leading to a plausible threat for its regional populations.
Trachyzelotes pedestris dominated sites in oak and mixed forests (see also Pitta et al. 2019) in the whole range of this study, but it was also found in high numbers in the marshes of the Evros Delta as well as in grasslands of the Evros district. It was recorded from about half of the sampling sites of this study. This species was found to be negatively associated with forested areas in other parts of mainland Greece (Zakkak et al. 2014) or Europe (Bell et al. 2001, Piterkina & Ovtcharenko 2008, which contradicts our results. This might be a result of different definition of sites' characteristics or a sampling artefact, i.e. the degree of open space between the trees in a forest and the exact position of the traps in the forested sites. In this study, most of investigated forests had openings and bushy understory (see Table 1). Since in the analysis of Zakkak et al. (2014) the cover of woody plants was calculated with higher accuracy, the negative association of T. pedestris with closed forests might be true and does not necessarily contradict our results.
Drassyllus villicus was another species found almost exclusively in oak and mixed forests of the study area and in relatively very high numbers. This species is thought to occur in warm steppic areas of central Europe (e.g. Grimm 1985, Roberts 1995 (2014) found that, together with Marinarozelotes cumensis and Civizelotes gracilis, they are mainly responsible for differences in community structure among sites of variable forest encroachment, with 48% of the abundance of D. villicus being recorded in areas of 75-100% woody vegetation cover. The differences in habitat preferences of this species along its range of occurrence probably reflect its adaptation to various climatic zones and/or different competitive interactions in the corresponding communities.
Drassyllus praeficus was found in high numbers at the sites of the Evros Delta, in open grasslands and in grazed areas. This species was recorded frequently on Crete from sea level to 1650 m a.s.l., but it preferred maquis of middle altitude mainly in central Crete (Chatzaki et al. 2005). Kaltsas et al. (2019) found that D. praeficus is an indicator of intensive grazing in Crete, corroborating its preference for open habi-tats and its tolerance to disturbance. As a common species in Europe (Batáry et al. 2008, Grimm 1985, it is considered typical of sites near woods (Isaia et al. 2007), but it is also positively associated with agricultural landscapes (Samu & Szinetar 2002).
Both Marinozelotes lyonneti and M. malkini are mainly found in grassy openings and marshes. Marinozelotes lyonneti was also found to be a significant indicator of overgrazing in Crete (Kaltsas et al. 2019) where it is very common.
Overall, the present study is in line with previous data showing that increasing habitat heterogeneity (Kati et al. 2010), and -especially for generalist ground spider species -increasing distance from forests (Stenchly et al. 2012) favour species richness and diversity, presumably related to the corresponding high diversity of herbivorous insects, which constitute the main food resources for predatory arachnids (Zografou et al. 2009).
Greece has the most diversified habitats on the Balkan Peninsula, thus constituting a major biodiversity hotspot for the Mediterranean and Europe as a whole (Cuttelod et al. 2008, Kougioumoutzis et al. 2021. Particularly its northern area, forms the merging point of Caucasian and Irano-Anatolian elements originating from the NE side, and the European fauna from the NW side. This is mirrored in our zoogeographical analysis. The Evros district and the area of Prespes Lake represent the easternmost and westernmost areas of northern Greece respectively, and the Magnisia district represents an area in central Greece. Apart from the highest number of the Greek endemics (9 species, or 10.8% of the total number of species included in the analysis) due to the six recently described species by the first author (i.e. Civizelotes akmon, Drassyllus centrohelleni cus, Drassyllus covid, Drassyllus dadianus, Phaeocedus similaris and Zelotes histius, see Chatzaki 2021), Euro-Middle Asiatic (EMA, 9 species or 10.8%), Mediterranean-Middle Asiatic (MMA, 7 species or 8.4%) and East-Mediterranean (EME, 7 species or 8.4%) elements form the largest part of the total species inventory. In the general zoogeographical categories analysis species with very narrow distribution (i.e. Steno) or very wide distribution (i.e. Wide and Med) -reaching from the Middle Asia to either Europe (e.g. Drassyllus praeficus, D. lutetianus, D. villicus, Gnaphosa lucifuga, Zelotes apricorum, Z. atrocaeruleus, Z. babunaensis, Z. electus) or only the Mediterranean zone (Berlandina plumalis, Drassodes lutescens, Marin arozelotes barbatus, Micaria coarctata, M. pallipes, M. pulicaria and Zelotes tenuis) respectively -dominate, mostly highlighting the aforementioned intermixing between European and Eastern elements. This overall pattern deviates regionally, where different elements seem to outnumber the others. The most striking example is between the areas of Prespes (the westernmost area) and the Evros Delta (the easternmost), where not only the percentages of the four biogeographic categories form a different pattern, but also, the species composition is distinct from 43% (the EU elements) to 100% (the STENO elements).
In the Evros district, except for the endemic and East-Mediterranean elements (represented by six species each), there are also many East-Mediterranean-Iranian (Drassyllus villicoides, Haplodrassus invalidus, Marinarozelotes malkini, Nomisia exornata, N. ripariensis), Mediterranean (Marinaro zelotes fuscipes, M. lyonneti, Setaphis carmeli, S. parvula, Zelotes fulvaster) and Euro-Middle Asiatic (Drassyllus praeficus, D. villicus, Gnaphosa lucifuga, Scotophaeus scutulatus, Zelotes babunaensis) species, as well as Mediterranean-Middle Asiatic species (Drassodes lutescens, Marinarozelotes barbatus, Mi caria coarctata, Zelotes tenuis) leading to a large percentage of eastern elements (24%) in the general zoogeographic analysis, similar to that of widespread species (27%). In the small area of the Evros Delta, there are less stenoendemics, making the more widespread species (but not the European ones) the dominant element (for example, the Palearctic D. lapi dosus, Micaria albovittata, M. dives and Phaeocedus braccatus, the Euro-Middle Asiatic D. praeficus, D. lutetianus, G. lu cifuga and Z. electus and the Mediterranean-Middle Asiatic B. plumalis, M. barbatus, M. pallipes and Z. tenuis), thus only partially following the overall pattern. When compared with previous zoogeographic analyses in the Evros area where all ground spiders were analysed (i.e. Komnenov et al. 2016), similar results are obtained, with the only difference being that when all spider species are analysed, the European elements (e.g. European, Eurasian, Euro-Siberian etc.) are more numerous and the Eastern elements are less evident. This is related to the fact that the spider family Gnaphosidae is mostly composed by xerophilic species that prefer (or withstand?) warmer dry habitats (Cloudsley-Thompson 1975). As a result, they are more diverse in the southern parts of Europe and in the Eastern areas of the Mediterranean and the Middle East, where they also present higher rates of endemism.
Finally, in the area of Magnisia the general zoogeographic picture is composed equally by Widespread (such as Z. atro caeruleus and Z. babunaensis) and Eastern elements (such as the East Mediterranean-Iranian D. villicoides, N. exornata and N. ripariensis), while fewer steno-endemics occur here, as is the case for Prespes. The general pattern in Magnisia resembles that of the Evros district and, accordingly, the composition of species in all categories does not present significant deviations, with the exception of the STENO category where the only common endemic species the two areas share is Zelotes cingarus; all others being distinct. Two more species (the East Mediterranean Berinda infumata and the Euro-Middle Asiatic Zelotes atrocaeruleus) are distinct in the Gnaphosidae fauna of Magnisia (the latter being present in Prespes too). Overall, the pattern of Magnisia probably represents a more realistic zoogeographic distribution of chorological elements for the central parts of mainland Greece, since the high percentages of STENO elements in Evros reflect an artefact due to the newly described species from that area, the distribution of which will possibly change in the near future.
This study demonstrates the value of taxonomic works in the less surveyed areas of Greece, still revealing new taxa, and certainly adding ecological information on a species (autecology) and at a community level. Biogeographical studies based on spiders are still preliminary, but still they are valuable in denoting evidence of species occurrence and routes by which they expand their distributions.