On the morphological separation of two sibling species: Pardosa proxima (P. vlijmi syn. nov.) and P. tenuipes (Araneae: Lycosidae)

Abstract. Morphological descriptions of P. proxima (C. L. Koch, 1847) (= P. vlijmi den Hollander & Dijkstra, 1974, syn. nov.) and its sister species P. tenuipes L. Koch, 1882 (= P. proxima auct.), a pair of species hardly distinguishable on a morphological basis but easily separable through behavioural characters, are given. The identification on an ethological basis allowed us to study the morphology of the males and to point out the morphological characteristics that can be used to discriminate the two species. Moreover, the examination of type material of Pardosa proxima and P. tenuipes, together with details given in the original descriptions, led us to conclude that P. vlijmi is a junior synonym of P. proxima and that P. tenuipes has been long overlooked. However, females remain hard to distinguish due to high intraspecific variability in the shape of the epigyne, vulva and habitus. According to the examination of material from different parts of Europe, P. proxima seems to be quite common in Italy and in the south of the Balkan Peninsula (mainly Greece and Bulgaria) while P. tenuipes is more widely distributed in western Europe, reaching central Europe. Contact zones between the two species were found in northwestern Italy and France. Far from detailing the precise distribution of the two species, we suggest that material previously identified as “Pardosa proxima” should be checked for establishing the occurrence of one or both species in different countries.

rope differ in a very specific way. Males of Lyc. proxima C. L. Koch have the femora of the first leg pair much darker than the others, almost unicolour, while in the form from southwestern Europe, which L. Koch named Pard. tenuipes (cf. 1881 [should be 1882]), the first femora are not darker and annulated like the others].
In addition, it is worth noting that in the original description of P. proxima, C. L. Koch (1847: 53) mentioned and illustrated the dark femur of leg I of males (see Fig. 24, arrow). The same characteristic was also emphasized by de Lessert (1910: 516, footnote).
Den Hollander et al. (1972) published a paper on the occurrence of wolf spiders of the genus Pardosa in southern France. In this work, they observed that specimens previously identified as P. proxima showed different courtship behaviour compared to the typical form. They found these "aberrant specimens" in two localities (Pas d'Esculette, near Millau and Le Pin, near Auxerre), occurring together with P. proxima auct. and P. hortensis. According to den Hollander et al. (1972), the preliminary examination of the external genital structures of the "aberrant specimens" showed intermediate features between P. proxima auct. and P. hortensis. A few years later, the same material was examined by den Hollander & Dijkstra (1974) who described the "aberrant specimens" as a new ethospecies, P. vlijmi den Hollander & Dijkstra, 1974. Ethospecies are defined as pair or complex of species that can be hardly distinguished on a morphological basis, if at all, but are easily separable through behavioural characters (O'Connor et al. 2011, see also Vlijm 1986). The concept of ethospecies was defined by Emerson (1956) to describe species belonging to the genus Apicotermes (Isoptera) that could only be distinguished by the different construction of the nests. Among spiders, ethospecies are particularly known in wolf spiders (e. g., Uetz & Denterlein 1979, Cordes & von Helversen 1990, Töpfer-Hofmann et al. 2000, Roberts & Uetz 2004.  confirmed that courtship behaviour is a valuable tool to discriminate P. proxima auct. from P. vlijmi. However, it seems likely that the concept of ethospecies only reflects the inability of the researcher to detect useful morphological features. The fact that the description of P. vlijmi as a new species was only based on courtship behaviour created nomenclature problems (see also Vlijm 1986). In particular, den Hollander & Dijkstra (1974) considered P. proxima auct. and P. vlijmi "morphologically indistinguishable", implying that it was not necessary to see any type material of P. proxima C. L. Koch, 1847. In their work, they assigned the new name vlijmi to the less common species: "It therefore seems plausible that the widespread P. "proxima" refers to Pardosa proxima (Koch, 1848) [sic!] and that the aberrant specimens belong to a new species". This kind of conclusion is rather arbitrary, given that there is no evidence that Carl Ludwig Koch had described P. proxima on specimens that performs the "normal" or the "aberrant" courtship behaviour observed by den Hollander & Dijkstra (1974). Although the authors provided some preliminary measurements of the body and the external genital structures, the drawings of palps and epigynes are poor and lack details. Moreover, given the apparent lack of documented morphological characters useful to distinguish the new species, it is likely that P. vlijmi has been routinely overlooked and often confused with its sibling species, P. proxima auct.
The results obtained in our previous studies on the courtship behaviour of these two species  concur with those reported by den Hollander & Dijkstra (1974). A closer examination of the specimens used for the behavioural analysis, together with the examination of material from different parts of Europe, revealed that the two species were in fact mostly confused in the past, and that a revision of the current nomenclature was needed. Further support for our assumption, linking morphological and ethological traits, is found in the molecular analysis previously conducted (Chiarle 2013), providing evidence for a clear separation of the two species, and justifying our use of morphological characters for species identification.
Here we present the results obtained from the morphological examination, we point out the characters useful for separating the two species and we revise their nomenclatural status.

Material and methods
Samples were photographed using an Olympus E-520 camera attached on an Olympus SZX16 stereomicroscope at the Zoological Museum, University of Turku and a Leica EC3 camera attached on an Leica MS5 stereomicroscope at the Department of Life Sciences and Systems Biology of the University of Torino. Dishes of different size with paraffin at the bottom were used to photograph the specimens in the correct position. Images have been subsequently fixed using "Com-bineZP" image stacking software. SEM micrographs were taken with a Hitachi S-4300 scanning electron microscope at the Swedish Museum of Natural History in Stockholm. The digital photo (stacked) in Fig. 26 was taken using an Infini-tyX camera on an Olympus SZX12 stereomicroscope at the Swedish Museum of Natural History in Stockholm.
Part of the examined material (marked with an asterisk*) has been formerly identified on an ethological basis (see . For both species, total body length, prosoma length and width, leg I length is reported (minimum and maximum). For males, we also measured palp tibia length and width. Description and measurements of females are based on presumed "pure" populations (i.e. populations where we only found males of one of the two species). All measurements are given in millimetres.
Apart from types, all studied materials are preserved at: Brief description of the courtship behaviour. The male quickly raises and lowers the whole body on the spot, with a series of small jumps. The vibration turns into a conspicuous hopping, characterized by up and down movements of the whole body toward the female. At the same time, the male performs some very rapid movements of the opisthosoma, kept parallel to the substrate. During hopping, the palps and opisthosoma scrape on the substrate. Type material. Pardosa proxima: Lectotype ( with old labels "Pardosa proxima" and "Griechenland Type" in Natural History Museum, London, here examined and designated: GREE-CE, possibly near Nafplio (cf. "Historical background" above). Another female stored in a separate tube labelled "Lyc. proxima type" (not old label) turned out to be P. hortensis. Pardosa vlijmi: Holotype ) and allotype ( from FRANCE, Lozère, Pas d'Esculette and 2( paratypes from FRANCE, Yonne, Le Pin in Naturalis Biodiversity Center, Leiden, examined.  Prosoma dark brown with darker eye region, with a narrow yellowish median band, lateral bands of the same colour, broken into three distinct parts (Fig. 5). Eye region with short hairs. Clypeus brownish, chelicerae brown with yellow internal side. Sternum brown. Opisthosoma dorsally dark brown with a distinct lighter cardiac mark surrounded and followed by a couple of spots of the same colour, spots fused near the spinnerets (Fig.  5). Ventral side of the opisthosoma yellowish with short, stumpy dark hairs (Figs 7, 19, 21-22). Legs uniformly yellowish, femora with brown annulations (Fig. 5). Leg I with femur brown (Fig. 5)  Prosoma dark brown with darker eye field. Median band yellowish, lateral bands broken into three distinct spots, same colour as median band. Rarely, the lateral bands are unbroken. Clypeus and cephalic flanks yellow-brownish, chelicerae of the same colour. Sternum brownish sometimes with a lighter central area. Dorsal side of the opisthosoma dark brown with a distinct narrow cardiac mark, flanked and followed by 4-5 couples of yellow-brownish spots, which are fused near the spinnerets. The whole pattern is quite variable and, in some specimens, it is faint and not clearly visible. Ventral side of the opisthosoma light brown with two lighter V shaped strips. Legs uniformly yellowish brown with few faint brownish marks on femora and patella. Leg I length: Femur 1.54, Patella 0.76, Tibia 1.22, Metatarsus 1.24, Tarsus 0.89. Epigyne as in Fig. 10. Habitat. Similar to P. tenuipes. The two species may co-occur in the same habitat. Remarks. The stumpy (peg-like) dark hairs on the venter of the male opisthosoma show some variation in density, being somewhat less dense in males from Morocco and Tenerife in    Description. The general description is based on specimens in which we observed courtship behaviour (marked with an asterisk in the material section). Specimens from other localities were considered for comparison. Male. Total length: 4.32-5.70. Prosoma: 2.20-2.97 long, 1.86-2.32 wide.
Prosoma dark brown, blackish in eye region, with narrow yellowish brown median band, and lateral light brown bands broken into three spots (Fig. 11). Eye region with long hairs. Clypeus yellowish, chelicerae brown with a light longitudinal strip on the internal side. Sternum dark brown. Opisthosoma dorsally grey-brownish with a remarkable brownish cardiac mark followed by 4-5 faint spots of the same colour (Fig. 11). Ventral side greyish with a wide yellow central area covered with hairs, normally developed (Figs 13, 20, 23). Leg I and all other legs uniformly yellow, femora with dark markings (Fig. 11). Leg I with numerous scattered long hairs on tibia and metatarsus. Leg I length: Femur 1.98, Patella 0.94, Tibia 1.61, Metatarsus 1.94, Tarsus 1.23. Palp as in Figs 12, 14, 15, brown with some yellowish areas on patella and femur, cymbium brown with lighter distal part. Embolus bent at approximately 90°, with the distal part almost equal in length to the proximal one (Figs 3-4). Female. Total length: 5.28-5.91. Prosoma: 2.56-2.76 long, 2.02-2.12 wide.
Prosoma dark brown with darker eye region. Median band light brown, lateral bands of the same colour broken into three different parts, sometimes wide and unbroken with few  (25); 26. epigyne (lectotype). Scale line = 0.1 mm small brownish marks. Clypeus, cephalic flanks and chelicerae yellow-brownish. Sternum uniformly light brown. Opisthosoma dorsally dark brown with a distinct lighter cardiac mark usually bordered by black dots. Pairs of light brown spots, often fused, follow the cardiac mark until the spinnerets. Ventral side of the opisthosoma uniformly light brown, sometimes a lighter V shaped strip is present. Legs uniformly yellowish brown with brownish marks clearly visible on femora. Leg I length: Femur 1.74, Patella 0.78, Tibia 1.48, Metatarsus 1.40, Tarsus 0.92. Epigyne as in Fig. 16. Habitat. Meadows, wet meadows, cultivated fields, swampy areas, edge of ponds and lakes.

Comparative remarks
Males of P. proxima and P. tenuipes show differences in the shape of the embolus (Figs 1-4), yet other morphological features permit a clear separation of the two species (Tab.1). In addition, the two species can be also well clustered comparing leg I metatarsus length with prosoma length (Fig. 17) and comparing palpal tibia width/length ratio with palpal tibia length (Fig. 18). However, the easiest way to discriminate males of the two species is looking at the hairiness ventrally on the opisthosoma (with numerous short modified hairs in P. proxima and normally developed in P. tenuipes, Figs 19 and 20, respectively) and at femur of leg I (same as other legs in P. tenuipes and darker in P. proxima; Figs 5 and 11, respectively). Photos with SEM (Figs 19-23) highlight striking differences in the length and the shape of the ventral hairs. As previously observed (Kronestedt 1996, modified hairs ventrally on the opisthosoma are found in other lycosid species in which the opisthosoma hits the substrate during courtship. Thus, the modified hairs present in P. proxima may be associated with a similar behaviour in this species.
On the other hand, females remain hard to distinguish on a morphological basis due to the high intraspecific variability and high overlap with respect to the shape of epigyne and vulva. Although some slight differences in shape and in the proportion between length and width of the epigyne could be considered, we argue that females of the two species cannot be distinguished on a morphological basis. Remarks on Pardosa proxima poetica. Simon (1876) described Pardosa proxima poetica as a small sized ('minima') variety of P. tenuipes (sub P. proxima). According to the origi-nal description, this variety was characterized i.a. by having very clear unbroken lateral bands on the carapace, and the male palp yellowish brown with the cymbium distally of bulbus much shorter than the bulbus. A sample with material fitting Simon's description was available from Spain, Andalucia, Fuengirola, ruderal ground, 18.V.1977 (T. Kronestedt, NHRS), 4) 5(, together with 1) and 2( of P. tenuipes. We think that the characteristics mentioned by Simon (1876) for P. proxima poetica should be further investigated, and we have therefore not placed Pardosa poetica as a senior synonym of P. tenuipes. It should be mentioned that Simon (1937) regarded P. proxima poetica as an "espèce dominante" in Spain and Portugal. Regrettably, a loan of the type material of P. proxima poetica, probably present in the Muséum national d'Histoire naturelle in Paris, was not possible. Remarks on WSC entries for P. proxima. In an attempt to assign the correct names to each of the WSC entries referring to P. proxima, several cases remained doubtful (Tab. 2). Despite the fact that, in a few cases, descriptions were matching some of the diagnostic features of P. tenuipes or P. proxima, we could not objectively establish whether they were just reporting Koch's original description or whether they were referring to multiple specimens from different countries.

Distribution
According to our data, P. proxima seems to be quite common in Italy and in the south of the Balkan Peninsula (including Macedonia: Komnenov pers. comm.). In Greece, only P. proxima has been found (e.g. Bosmans et al. 2013, Bosmans pers. comm.). Contact zones with P. tenuipes are found in northwestern Italy and in France (original records by den Hollander & Dijkstra 1974). The countries in which the presence of P. proxima is certain are: Greece, Macedonia, France, Bulgaria, Italy, Morocco, Turkey, Canary Islands.
Although P. tenuipes is considered widespread in Europe, we examined only a few specimens from Western Europe and the Iberian Peninsula. According to our data, P. tenuipes occurs in Spain, Belgium, Great Britain, France, Italy and Portugal.
It seems likely that P. tenuipes occurs mostly in western Europe, while P. proxima seems more common in southern and eastern Europe.
Most illustrations and or descriptions available in literature do not allow a clear understanding about how names were used by previous authors (see Tab. 2). Far from detailing the precise distribution, we suggest that material previously identified as "Pardosa proxima" should be checked for establishing the occurrence of one or both species in different countries.
With a certain degree of uncertainty, illustrations and descriptions available in literature seems to confirm the occurrence of P. proxima in France (Simon 1876, Tongiorgi 1966, former Yugoslavia, Italy, Macedonia, Albania, Greece and Austria (Tongiorgi 1966).