Lycosa Latreille, 1804 (Araneae, Lycosidae) of Israel, with a note on Geolycosa Montgomery, 1904

Lycosa (Araneae, Lycosidae) is a wolf spider genus typical of subtropical latitudes in the western Palearctic. Despite being erected over 200 years ago, the taxonomy of Lycosa is still unclear. Many species formerly ascribed to it are currently being moved to other genera, while new species are still being described. The species of Lycosa of the western Mediterranean basin are relatively well known, yet the Levantine region, the easternmost part of the Mediterranean basin, has not received much attention since the early 20th century. Here, we study Lycosa from the southern Levant using morphological, molecular and behavioral characteristics, to delimit the species found in this region. We describe two new species: L. hyraculus sp. nov. and L. gesserit sp. nov. We re-describe the widespread and polymorphic species, Lycosa piochardi Simon, 1876. Lycosa piochardi infraclara Strand, 1913 is synonymized with Lycosa piochardi. By adding novel data to the molecular phylogeny of Lycosa created by Planas et al. (2013) and re-analyzing it, we explore the relationship of the Levantine species to other Mediterranean species of Lycosa. We discuss habitat preferences of the two species of Lycosa. Additionally, we report the burrowing species Geolycosa vultuosa (C.L. Koch, 1838) as a new record to Israel, thus extending the distribution of this species and genus into the Levant.


Introduction
by us led us to hypothesize that there are only two species of Lycosa in the southern Levant: a doorbuilding and a turret-building Lycosa (Table 1). Integrating these data could indicate the existence of several sympatric, or vicariant species, or, conversely, a few wide-ranging species with variability in their genital and somatic morphology -or a combination of the two. We tested these hypotheses using a third set of characters -molecular sequence data, for those populations for which we had freshly collected material. Regarding sympatry, we aimed to investigate possible mechanisms of coexistence, such as niche partitioning.
In addition, we aimed at testing the phylogenetic relationships of these putative species, especially in relation to the western Mediterranean species. There are several possible scenarios for the origin of the southern Levantine Lycosa in relation to the previously studied western Mediterranean species. The species of Lycosa of the southern Levant, studied here, could have originated within one or more of the western north African Lycosa lineages reported in Planas et al. (2013) and migrated eastward. In this case, the migration of Lycosa to the Levant could have taken a southern route, directly from Africa, or a northern route, originating in the groups that had colonized Europe prior to the dispersal eastward. If the split is older, and the southern Levantine species are nested deep in the western Mediterranean group before its radiation into the four current lineages, dispersal would be expected to take place directly from north Africa, where the taxon is supposed to have originally diversifi ed. An additional scenario is that the southern Levantine species of Lycosa do not belong to the western Mediterranean clade at all and would be recovered in the phylogeny as its sister taxa. Under this scenario, no information could be gleaned on the origin of these species, as genetic material of most of the non-Mediterranean species has not been sequenced yet. More elaborate dispersal scenarios can be suggested beyond these routes, but the phylogenetic position of the Levantine clades should provide a strong hint as to the most parsimonious possibility.

Character set Number of species Character Distribution
Somatic morphology four to six Ventral abdomen coloration and patch size (Fig. 15) One species throughout the region, one to three in northern Israel, one throughout southern Israel, and one in the Negev high mountains and the Arava valley. Genital morphology three (according to each character separately) to fi ve (if combined) male: Tegular apophysis tip bent proximally or not bent proximally. Crest of tegular apophysis serrated or smooth (Figs 13,(19)(20) female: Septal pedicel present or absent. Septum triangular or trapezoidal ( Fig. 9). Spermatheca spiral-shaped or bent dorso-ventrally. Head of spermatheca narrow or bulbous (Fig. 10).
Two species along Israel from the Hermon to the Negev, one in the loess plains of the Negev, one in the Negev high mountains, and one in the Arava valley.

Burrow building behavior two
Door-or turret-building (Fig. 22) One species throughout the southern Levant, in all habitat types. Another species in the desert habitats of the south of Israel. In order to differentiate among these scenarios, and test our species hypotheses, we used fresh material from Israel and Palestine as well as historical material collected from Israel, Jordan, Lebanon, Palestine and the Sinai Peninsula, and examined the identities, distributions and relatedness of species of Lycosa and Geolycosa of the southern Levant. A short ecological survey was performed, to assess the habitat preferences and niche partitioning of two of the species found to be distributed sympatrically in the Negev desert.

Material and methods
This synopsis is based on material deposited at the Israel Arachnid National Collection, the National Natural History Collections (NNHC), the Hebrew University of Jerusalem (HUJ), Senckenberg Naturmuseum (SMF), National Museum of Natural History, Paris (MNHN) and Natural History Museum in London (NHM). Four hundred and forty-one specimens from ~160 localities were examined ( Fig. 1), 348 of which were freshly collected by the authors, or by collaborators. The samples were preserved in 70% ethanol in room temperature (for morphology only), or absolute ethanol in -80°C (for DNA extraction) and identifi ed using relevant literature (Koch 1836;Simon 1876;Strand 1915), and comparison with paratype specimens. Specimen lists created according to . Measurements are in mm. Epigynes were detached and temporarily cleared with clove oil. Specimens were examined and measured using a Nikon SMZ25 stereo microscope. Digital microscope images were taken using NIS-elements imaging system with Nikon DS Fi2 digital camera mounted on a Nikon SMZ25 stereo microscope. Habitus images were taken with Nikon B500. Images were edited using GIMP ver. 2.10. 24.

Molecular analysis
DNA was extracted from the legs of 16 specimens (Table 2) of three species of Lycosa. The method chosen was Minicolumn Purifi cation, using the BioVision Insect Genomic DNA Kit (Catalog #: K1412), following the protocol provided with the product (except that the samples were incubated in proteinase K overnight, rather than 30 minutes). Two genetic markers were amplifi ed by PCR using general primers (Table 3) and sequenced bidirectionally with Sanger sequencing. We retrieved a ~650 bp long fragment of mitochondrial cytochrome c oxidase I (COI) and a ~650 bp long fragment of nuclear 28S ribosomal RNA (28S).
All sequences are deposited in GenBank (Table 2). To these, we added the corresponding COI and 28S fragments from 18 species retrieved from GenBank (Table 2), mostly from the material used by Planas et al. (2013). These two markers were selected for their proven utility in lycosid phylogeny (Piacentini & Ramírez 2019). The sequences were trimmed and aligned using MEGA 10 (Kumar et al. 2018). We performed a Bayesian analysis, using BEAUti ver. 1.10.4 and BEAST ver. 1.10.4 (Suchard et al. 2018) and constructed a phylogeny using TreeAnnotator ver. 1.10.4 (Drummond & Rambaut 2007). Tree prior used in BEAUti was 'Speciation: Yule Process' (Yule 1925;Gernhard 2008 In each locality the habitat was recorded (plain, hill), the incline of the surface (> 10º, < 10º), the direction of the incline (north, east, south, west), relative vegetation cover (negligible, sparse, abundant), percentage of stones on surface, and the distance to the nearest shrub were measured. Percentage of

Results
We reject our hypotheses of four to six (morphology) or two (burrow construction behavior) species of Lycosa in Israel and Palestine. By using integrative taxonomy, we conclude that three species can be identifi ed using morphology, behavior and molecular data, from Israel and Palestine. We therefore describe (see below) two new species of Lycosa from Israel and redescribe the widespread species Lycosa piochardi Simon, 1876. We additionally document one female specimen that might represent an additional species of Lycosa, yet we did not have males or material for molecular analysis to test this hypothesis. We here present a key to the species of large lycosids (Geolycosa, Hogna, Lycosa) in Israel and Palestine. We added to the key the northern Levant species, Lycosa praegrandis C.L. Koch, 1836, that might exist in the north of Israel.    Kovblyuk et al. 2012: 246, fi gs 8, 11, 14, 18, 21, 27, 30, 36, 39, 43 (♂♀).

Natural history
This is a burrowing species, found mainly in open, herbaceous habitats (Mcheidze 1997). Hibernation takes place in the burrow (at least in the north of its range) (Mcheidze 1997), juveniles emerge in springtime. Maturation takes about a year (Mcheidze 1997). The species is preyed upon by pompiliids, scorpions and centipedes, and the eggs are subject to mite infestations (Mcheidze 1997). The specimens examined in this study were found active in daytime in an urban meadow: one on the surface, the other in a burrow with a turret made of soil (Figs 7A, 19A, 21A, 22, Table 4).

Relationships
Geolycosa vultuosa is related to Geolycosa charitonovi (Mcheidze, 1997) and Geolycosa dunini Zyuzin & Logunov, 2000(Zyuzin & Logunov 2000. The type species of Geolycosa (G. latifrons Montgomery, 1904) is of Nearctic distribution. It may not be congeneric with the Palearctic species ascribed to Geolycosa. As the revision work needed to fi nd the correct phylogenetic placement of G. vultuosa is widely beyond the scope of this work, we refrained from moving G. vultuosa to a different genus.

Diagnosis
Large lycosids; anterior eye row clearly shorter than PME; epigyne septum tongue-shaped or rhomboid, confi ned to posterior half of epigyne; septal pedicel reduced or absent; TA wide and fl at, ending with a distal process (TAT) directed retrolaterally to posteriorly (Zyuzin & Logunov 2000).

Description
Large lycosids (body length over 12 mm). Carapace sloping posterior to ocular area. Carapace margins with thick pubescence (Zyuzin 1990). Chelicerae with three retromarginal teeth. AER recurved. PME wider than AER. Abdomen venter usually with black markings. Sexual dimorphism weak. Posterior part of epigynal septum widened and clearly outlined. Septal pedicle fused with genital groove or wholly absent. This contrasts with other large Lycosinae (Hogna, Geolycosa), having a clearly outlined pedicle. Cymbium asymmetrical. Tegular apophysis wide and fl at, with an elongated distal process, directed retrolaterally to posteriorly. Embolus thin, long, with narrow epiconductor.

Natural history
Generalist predators, typical of open habitats. Most species are burrowing, but some seem to be vagrant (Planas et al. 2013). The burrow entrance is usually modifi ed with a door or a turret (an elevated structure made of silk and vegetation, surrounding the entrance). Activity is mostly nocturnal. Reproductive season recorded in the summer and autumn.

Distribution
Known from the Mediterranean Basin, eastern Europe, Middle East, central Asia and Japan (World Spider Catalog 2022).

Natural history
This species is nocturnal. A single juvenile male was collected at night, in September, at 650 m a.s.l., in a stony plain near Har Karkom, a secluded tableland in the hyperarid desert (annual precipitation < 75 mm) (Fig. 1, Table 4). The specimen was mature after two molts in captivity, in August 2018. Like the closely related L. hyraculus sp. nov., it makes a burrow with a trapdoor. Shares habitat with L. piochardi. Reproductive season unknown.

Distribution
ISRAEL. Possibly endemic to the Har Karkom tableland or found throughout hyperarid zone.

Relationships
According to our molecular phylogeny, this species is the sister taxon of Lycosa hyraculus sp. nov., and closely related to Lycosa oculata, Lycosa aff. oculata 1 and Lycosa aff. oculata 2 (Figs 21-24). Cephalothorax is similar to Lycosa macrophthalma Nadolny & Zamani, 2020 and it might be related to it as well. The single known male specimen was raised in the laboratory for two molts. This has probably affected the size of the adult spider. Even though we have observed an aberrant morphology in the genitalia of females grown in lab conditions, we have not observed the same in males, either of L. piochardi or of L. hyraculus sp. nov. We consider it highly probable that the diagnostic characters in the male's pedipalp were not affected by the laboratory conditions.
Although genitalia of other males of Lycosa grown in our laboratory did not show distortions (see discussion about distortions of female genitalia), we cannot dismiss the possibility that the pedipalps of specimens in the wild are somewhat different to the type specimen, which was raised in the laboratory.

Notes
The locality where the type was found is a remote and not easily reachable nature reserve, encircled by army training areas, that are closed to the public most of the year. There are only few occasions every year that access to the public is possible. As of the writing of this paper, we have not succeeded in collecting additional specimens.

Female
Recognized by combination of characters: ocular area: > ⅓ length of carapace, distinctly raised, making the carapace between PLE and fovea appear slightly concave in profi le. Epigyne: septal pedicel reduced, septum trapezoidal, broader proximally, wider than long. Copulatory openings narrow, at anterior end of septum. Carapace resembles L. oculata, L. suboculata, and L. macrophthalma, but in these species the septum is longer than wide. Carapace also resembles Lycosa sp., but in this species the copulatory openings are not visible in ventral view. Epigyne similar to L. piochardi and L. baulnyi, but in these species the ocular area is not distinctly raised, and is < ⅓ length of carapace (Figs 6C, F, 7C, F, 9B, 10B, 11C-D).
CARAPACE. Ocular area > ⅓ length of carapace, distinctly raised, making the carapace between PLE and fovea appear slightly concave in profi le.
COLOR. Carapace cream colored, with light brown median bands, radiated and darker on margins, connecting to dark ocular area. Perimeter light brown. Sternum black. Chelicerae reddish-brown.

Natural history
This species is nocturnal. Females were collected March, June, August and September. Males collected June, August and September. Subadult females were collected April and June. Subadult males were collected May, July and September. Juveniles were collected January, June, July and October (Table 4). Reproductive season unknown. This species inhabits loess soils in arid desert environments (annual precipitation 75-200 mm). The burrow usually with a modifi cation in the form of a thin, fl exible door made of silk and soil, and hinged to the entrance with silk. Rarely an object such as a piece of soil crust or a fl at stone can be used as a door. The door is often left open. The spiders are usually found at night, standing near the burrow, rarely on rocks or vegetation. Captive specimens accept a wide variety of insects, similar to other species of Lycosa (Steves et al. 2017) (Figs 8C-D, 21B, 22B).

Relationships
According to our molecular phylogey, this species is the sister taxon of Lycosa gesserit sp. nov., and closely related to Lycosa oculata, Lycosa aff. oculata 1 and Lycosa aff. oculata 2. The cephalothorax is morphologically similar to Lycosa macrophthalma Nadolny & Zamani, 2020 and the species might be related to it as well (Figs 23-24).

ISRAEL -Arava Valley
COLOR. Carapace cream colored, darker on margins, with chocolate brown median bands, connecting to dark ocular area. Sternum yellow. Chelicerae reddish-brown. Abdomen cream colored with light brown chevrons dorsad. Abdomen venter with black patch posterior to epigastric fold, connecting to two parallel black bars stretching towards spinnerets (Π shaped patch). Spinnerets yellow. Legs yellow, infused with brown dorsad, yellow ventrad with two black bands on tibia, to reddish-brown distad, Coxae and trochanters yellow. Pedipalps yellow, to dark brown distad (Figs 6E, 7E).

Natural history
A single female was collected in the hyperarid desert (annual precipitation < 75 mm) of the 'Arava valley, ~ 200 m below sea level (Fig. 1, Table 4). No ecological data is available.

Relationships
As of the writing of this paper, we have not succeded in collecting fresh specimens, and therefore we could not test its relationship using molecular methods. Somatically, it resembles L. gesserit sp. nov. and L. hyraculus sp. nov. Additionally, it is the only other specimen in our dataset, apart from the L. gesserit male, to exhibit a ventral patch on the abdomen shaped like the letter Π. As we could not know whether this is a new species or the female of L. gesserit, it is described here as Lycosa sp. only.
COLOR VARIATION IN MALES (n =52). Coloration varies greatly. General coloration may be sand-yellow (almost as light as Lycosa hyraculus sp. nov.), grey, light brown, reddish-brown and dark brown (Fig. 17). The lighter specimens usually found in arid environments. Black patch on abdomen may be covering the whole ventral side, a part of it (this is the most common condition) or, in rare cases, be altogether absent.
VARIATION IN FEMALE GENITALIA (n = 241). Epigyne: proportions vary greatly! General shape of epigyne oval to pentagonal; as long as wide to much longer than wide; sometimes asymmetrical in relation to septum. Septum triangular to trapezoidal, almost rectangular, wider than long to longer than wide. Copulatory openings, usually reduced to slits, but rarely a little wider. Wide copulatory openings sometimes appear in laboratory-raised specimens (not found in the wild, Fig. 27). Anterior edges of atria aligned or prolaterally slanting, wider or narrower than septal pedicel. Ridges anterior to copulatory openings sometimes accentuated with dark color. Vulva: head of spermatheca often with protrusions ('warts') in random places. Distal part of spermatheca sometimes slightly undulating (Figs 9-10).

Natural history
The species is mostly nocturnal, but diurnal activity in winter is suspected. Females were collected throughout the year. Females carrying eggs or juveniles were collected September through January, with a single specimen from June. 79% of adult females (167 specimens) were collected July-October. Males were collected in April, July-October and December. 86% of the males (45 specimens) were collected in July-September. Subadult females were collected in March-July and October. Subadult males were collected in January and March-August. Juveniles were collected throughout the year (Table  4). Twenty seven of the specimens kept in the laboratory lived for more than a year. Specimen HUJ INV- AR20813 was collected as a young juvenile, lived for 26 months and moulted ten times, before dying as a subadult (Fig. 26). Lycosa piochardi is found in a wide range of climates: dry summer continental climate, Mediterranean, steppe, arid desert and hyperarid desert, spanning altitudes of 380 m b.s.l. to 2400 m a.s.l., and average precipitations of 21 mm to 1150 mm. It is most commonly found in low scrub, grasslands and habitats with bare ground. Lycosa piochardi is abundant in natural and agricultural areas, but is very uncommon in urban environments. The burrow is built in open ground, in a great variety of soil types: from stabilized sand to alluvial clay soils and rocky regosols. The burrow opening is usually modifi ed, with a short turret, made of vegetation held together with silk. Yet, it is not rare to fi nd turretless entrances. Some turrets are found sealed with silk in autumn. One burrow (specimen HUJ INV-AR20912) found in a sandy substrate had a full silk lining. In rare cases L. piochardi is found in a retreat under a stone. Lycosa piochardi is a generalist predator, feeding on a wide variety of prey species. Captive specimens readily accept crickets, cockroaches, honeybees, fl ies, butterfl ies, pholcid spiders and other soft-bodied arthropods (personal observation). Beetles and bugs were usually rejected, yet wild specimens were observed feeding on Pyrrhocoris apterus (Linnaeus, 1758) (Figs 16-17, 21C-D, 22C-D, 25-26).

Notes
We could not locate the holotype of Lycosa piochardi Simon, 1876, either in its original repository in the MNHN or in other collections. For the identifi cation we used a paratype (MNHN 1266, Fig. 18) that was collected with the holotype and identifi ed by Simon.
We examined the holotype of Lycosa piochardi infraclara Strand, 1915. We synonymize it with L. piochardi, as its morphology (Figs 9F, 10F) falls within the normal range of L. piochardi morphology and is described from the typical area of distribution for this species.
We could not locate the holotype of Allocosa olivieri (Simon, 1876), but judging from the locality (the Jordan Valley) and the shape of the epigyne in the illustration (Simon 1876: pl. 3 fi g. 10), it should probably be synonymized with L. piochardi.
A male specimen (BMNH 742) collected by Koch in Sardarapat (Armenia) was originally identifi ed as L. piochardi. Our examination shows it to be L. praegrandis, with a characteristic semicircular conductor, rather than a triangular one. Lycosa piochardi was recently recorded by Zamani et al. (2021) from Iran. The epigynes and spermathecae presented in the paper (Zamani et al. 2021: fi g. 8) bear great resemblance to the material examined by us. However, the Iranian specimens differ by having stouter spermathecae with bulbous heads (unlike the elongated heads in our material). A molecular phylogeny may help to decide whether the specimens are indeed L. piochardi, or represent a closely related, but separate species.

Ecological survey
A total of 47 individuals of the genus Lycosa were recorded 17 in Yeruham park), of these 31 were L. hyraculus sp. nov. and 16 L. piochardi. The micro-habitats of the two species were found to be signifi cantly different in all parameters (Table 5). The abundance of Lycosa hyraculus sp. nov. was positively correlated with a plain habitat, with a surface incline smaller than 10º, east and north facing slopes, distance from shrubs (0-9 m), low percentage of stones on surface ( 0.3-77%) and localities with sparse vegetation. The abundance of Lycosa piochardi was positively correlated with a hill habitat with a surface incline greater than 10º, south facing slopes, short distance from shrubs (0-1.06 m) and high percentage of stones on the surface (8.9-99.8%). But the habitat category 'hill' was signifi cantly positively correlated with the degree of incline (likelihood ratio = 0.0002, Pearson = 0.0002) and distance from nearest shrub (two-tailed T test, p = 0.03), but not with direction of incline (likelihood ratio = 0.3, Pearson = 0.5) and amount of vegetation (two-tailed T test, p = 0.6) ( Table 5).

Discussion
In this study we aimed at exploring the identity and diversity of the large lycosids of Israel and Palestine. We used genital and somatic morphology as well as burrow building behavior to hypothesize how many species of Lycosa are found in our material. Our initial hypothesis using morphology suggested three to fi ve species, while using burrow building behavior our hypothesis suggested only two species: a door-and a turret-building species. Testing these hypotheses with molecular analyses recovered three species of Lycosa, of which two are new to science. We found one additional specimen that may be a fourth species also new to science; however, we could not test this species hypothesis with either burrow building behavior or molecular analysis as we had only one female specimen collected more than 30 years ago with no information on its burrow construction. This specimen differs in its genital structures from all the other material we had; however, because we had no male and only one specimen, we decided to document it but not formally describe it. In addition, we report here one species of Geolycosa, a genus new to the region.
We also aimed to differentiate between a number of possible hypotheses as to the origins of southern Levantine Lycosa: the local species could be nested in the western north African clade and therefore would have originated from the aforementioned region either relatively recently or before the radiation into the current lineages; conversely, they could belong to a sister clade to the African Lycosa. Our results point to a relatively recent radiation within the western north African clades and at least two waves of eastward migration. Furthermore, we aimed to understand the ecological relationships between the southern Levantine species and to map their distribution to the different available habitats. We found that the seeming sympatry of two species at low resolution is resolved into apparent niche partitioning at high resolution, with Lycosa hyraculus sp. nov. using the loess plains as a preferable habitat, and  Table 2. Photos by I. Armiach Steinpress.
L. piochardi occupying the hill habitat when in sympatry with L. hyraculus, and both habitats when L. hyraculus is absent.

Intraspecifi c variation
Lycosa piochardi is a species with a high variation in genital and somatic morphology. Size (Fig. 14), dorsal coloration (Fig. 17), ventral pattern (Fig. 15) and epigyne shape (Fig. 9) are all found in several variants. While the shade of the dorsal coloration could be tied to the color of the surface (lighter on desert soils, darker on Mediterranean soils), the other characters do not seem to be closely tied to environmental factors and tend to vary within local populations. Because of this variation, L. piochardi was historically identifi ed as several taxa, such as L. tarantula, Allocosa olivieri (Simon, 1876) and even  Table 2. in the dataset), up to 14.29 mm. It is unknown what factors infl uence the size of the spiders in this study and whether 'gigantism' in females is of any signifi cance.

Abnormal variation in lab grown female specimens
Describing new species from individuals raised in captivity, without seeing wild adult might be problematic. Many of the specimens in this dataset were collected in the wild and raised to maturity in captivity. We have noticed that in both L. piochardi and L. hyraculus sp. nov., females that took more than one molt to mature in captivity sometimes developed epigyne shapes different from those of specimens that were collected mature or molted to maturity shortly after collection (Fig. 27). Some of these epigynes were paedomorphic or weakly sclerotized, but others did not appear unusual, except (apparently) not being found in nature. Compared to the symmetrical, almost triangular epigynes of specimens from the fi eld, the epigynes of captive-raised specimens of Lycosa tended to be asymmetrical and almost rectangular, with wide atria (wild specimens usually have narrow atria). We assume all of these to be abnormal, resulting from some discrepancy between the natural and laboratory conditions. Male genitalia were seemingly unaffected and were identical in shape to the genitalia of specimens collected from the wild. We have decided to describe the laboratory raised specimen, HUJ INV-AR20631, as a new species (Lycosa gesserit sp. nov.) only after both seeing that the male pedipalp differs from other species (Figs 2,20), and receiving supporting data in the molecular phylogeny .

Biogeography
This study concentrated on Israel and Palestine, whereas only a few specimens from other countries were available to us. Therefore, the species list presented in this paper should not be seen as exhaustive for the southern Levant, as additional species may be found in Jordan and Egypt (Sinai Peninsula). Lycosa piochardi can be assumed to be distributed throughout the region, but the other species of Lycosa appear to have more restricted local distributions. Lycosa hyraculus sp. nov. is confi ned in Israel to the central Negev Desert and is also found in the central Sinai Peninsula, which is a continuation of the aforementioned geographical region. Lycosa gesserit sp. nov. and Lycosa sp. were recorded from Israel and are likely found in hyperarid habitats in Jordan and the Sinai Peninsula. Planas et al. (2013) constructed a phylogeny of western Mediterranean Lycosa and suggested that all western European species originated in a few expansion events from western north Africa. Our additions to this phylogeny do not alter the topology presented by Planas et al. and indeed support their conclusions. In our phylogeny L. piochardi is a sister species of L. baulnyi from the southern Atlas Mountains. The phylogeny supports L. piochardi as monophyletic, but lacks the resolution to show any topology within the clade. Lycosa hyraculus sp. nov. and L. gesserit sp. nov. appear as sister species in the L. oculata clade. These fi ndings support the hypothesis that the southern Levantine species originated in separate expansion events, possibly from Africa. Despite the similarity to Lycosa praegrandis, the hypothesis that L. piochardi represents a northern lineage related to L. praegrandis was not supported, as the two species are not closely related in our phylogeny. A north African origin of the genus could be suggested, as all the non-African species (L. hyraculus, L. gesserit, L. piochardi, L. praegrandis, L. tarantula + L. hispanica) in our phylogeny have African sister clades. Adding genetic material from other west Asian and north African countries (e.g., Iran, Egypt, respectively) should further clear up the delimitation and historical biogeography of this genus.
It is interesting to note that according to Planas et al. (2013), species of Lycosa are sympatric only when they belong to different lineages (of the four main lineages outlined by those authors). This is also the case in our study area. Additionally, it is consistent with the apparent lack of sympatry between L. praegrandis and L. piochardi, despite the geographical proximity (as both belong to the same main lineage -the 'baulnyi group'). If we take this to be a general rule, it would be more reasonable to assume that L. piochardi is indeed a species with high variation in genital and somatic morphology, rather than a number of closely related sympatric species. Identifying the region where there is a replacement between L. praegrandis and L. piochardi may contribute to understanding their ecology and biogeographic history.
Another interesting observation, in apparent contradiction with the lack of sympatry between related lineages mentioned above, is the phylogenetic and geographical proximity of L. hyraculus sp. nov. and L. gesserit sp. nov. (Figs 1, 24). They appear to maintain separate populations in proximity to one another, suggesting a reduced mobility. These species belong to the L. oculata group, which in Planas et al. (2013) is the most speciose clade, most species of which occupy small, closely located distributions. The species of the oculata group are also very similar in morphology, a fact that is accentuated by the discovery of three of them during the study of Planas et al. (2013), based on molecular phylogeny. All this is suggestive of the tendency to become isolated and to form species with restricted distributions. While isolation due to low dispersibility may be the reason for the existence of the two species in geographical proximity, another possible reason for the proximity of L. hyraculus and L. gesserit may be the ecological vicariance. Lycosa hyraculus inhabits arid desert habitats, where the annual precipitation is between 200 and 75 mm, and at an elevation lower than 500 m a.s.l., while the only specimen of L. gesserit was found in the hyperarid (precipitation under 75 mm) Har Karkom, at an elevation over 600 m a.s.l. Each species could be adapted to very specifi c climatic conditions, such as levels of humidity (DeVito et al. 2004). Yet, there seems to be a gap between the distributions of the two species, where despite apparently suitable conditions and repeated searches, only L. piochardi was found. Due to this restricted distribution and apparent low mobility, L. hyraculus might come under danger of extinction, as loess plains in Israel are being developed for housing and agriculture and as the current climate change may further aridify this habitat.

Ecological survey
The coexistence of two similarly-sized, congeneric species always raises the question of resource partitioning. In our preliminary survey in the central Negev Desert, the distribution of the two species of Lycosa across the two main habitat types (plain and hill) was found to be signifi cantly differential. The different association of the species with magnitude of incline, percentage of stones on surface and distance to shrubs may be explained by the attributes of the different habitats, but even though most specimens of L. piochardi were found in the plain habitat, they inhabited steeper surfaces compared to L. hyraculus sp. nov. The differential occupation of incline directions may be due to microhabitat preference. South and north slopes differ greatly in evaporation, due to different exposure to the sun (Pavlícek et al. 2003) and are known to have differences in biomass and species composition. The difference in amount of vegetation in the vicinity of the two species of Lycosa may suggest that L. piochardi in the survey area has a tendency to be found in patches of ground more densely vegetated than those typical of L. hyraculus. Based on these preliminary fi ndings, we would like to suggest that in the survey area some specialization of L. hyraculus and L. piochardi occurs both at the habitat and microhabitat level. Lycosa piochardi is dominant in steep hill habitats, while L. hyraculus is dominant in fl at plain habitats. Yet, in some habitats these species coexist, and there may exist a fi ner niche partitioning, in which L. piochardi is found in the steeper, more densely vegetated patches, while L. hyraculus is found in the fl atter, more sparsely vegetated patches. At localities where only L. piochardi is found, it was observed occupying all habitats. Further collection of data will be needed to explore the behavioral / ecological interactions of L. hyraculus and L. piochardi.