A redescription of Harpactea dufouri ( Thorell , 1873 ) ( Araneae , Dysderidae ) , its occurrence outside the Balearic Islands , and some notes on the corticalis group of the genus

Harpactea dufouri (Thorell, 1873) was collected in the Gavarres protected natural area in Catalonia, Spain. The specimens were compared with specimens from Mallorca, Balearic Islands, and found to be conspecific. The female of the species is described here for the first time. The new finding proves that Harpactea dufouri occurs outside the Balearic Islands. The species, however, may be endemic to Catalonia.

The authors made a one week field trip to the Gavarres protected natural area in Catalonia, a 300 km 2 hill massif in the northernmost part of the Catalonian coastal range, characterised by an amazing biodiversity (Bosselaers 2004, Bosselaers in prep.).On this occasion, the junior author collected six specimens of Harpactea dufouri, formerly only known from the Balearic Islands.This rare species, including the hitherto unknown female, is redescribed and discussed below.

Material and methods
Specimens were observed, photographed and drawn using Euromex MIC465 and Olympus SZX9 stereo microscopes.For stereo microscope photographs, a Praktica DC440 digital camera was used.Details of the male palp and the vulva were studied with a Zeiss Axio Imager A1 microscope equipped with a Canon Powershot G6 digital camera.The image series obtained were stacked with Zerene Stacker version 1.04 build T201411272115.The vulva was cleaned for microscopy with trypsin (Sigma) for 24 h at room temperature an subsequently cleared in methyl salicylate for observation.The male palp was immobilized in Schwarzkopf "Freezing Gel Extreme 5", a transparent vinylpyrrolidone-vinylacetate copolymer (CAS 25086-89-9) / carbomer gel (Schröder et al. 2000;Schulze zur Wiesche 2006).All measurements are in millimetres.The format for leg spination follows Platnick & Shadab (1975), amended for ventral spine pairs according to Bosselaers & Jocqué (2000: 307).Leg spination is also illustrated in a schematic representation (Fig. 1E-G) where pl, do, rl and ve sides of leg articles are flattened as a folding net (Dürer 1525).

Diagnosis
Harpactea dufouri differs from H. corticalis (Simon, 1882) by the more slender bulbus of the male palp, the more straight conductor and embolus and the larger number of spines on femora III and IV.
Harpactea dufouri somewhat resembles H. heizerensis Bosmans & Beladjal, 1991, but differs from it by the knife-shaped, pointed conductor with very sharp teeth, the posterior diverticulum of the vulva consisting of a small ventral and a larger dorsal lobe and by the larger number of spines on femorae III and IV.

Description
Male Largest specimen Mont Ras, total length 5.50 (avg (n = 4) 4.48, sd 1.08).Carapace length 2.16, width 1.60, yellowish brown with faint grey radial striae, more reddish brown in cephalic region, smooth.Cephalic region slightly wider than half the carapace width, three times as wide as the eye group (Fig. 2A-B).Fovea weak and thin, length 0.32, anterior end 1.42 from front end of carapace.MOQ length 0.22, anterior width 0.26, posterior width 0.16, AER width 0.26, PER width 0.34.All six eyes ringed with black, PE clear pearly white, almost touching, diameter of PLE 1.5 times PME.AE as large as PLE, pearly white, subquadratic, separated by half their diameter.PER procurved in both do and fr view.Clypeus vertical, 0.05, equal to half the diameter of AE.
Chilum single, triangular, brown and sclerotised.Chelicerae brown, slightly rugose, anterior rim with two teeth close to fang tip, the bigger one furthest from tip.Posterior rim with two small, widely spaced teeth.Labium twice as long as wide, with small anterior notch.Endites long and parallel, with serrula and bluntly pointed anterior end.Exterior margin notched.
Male palp as illustrated (Figs 1A-B, 2C-D, F), with spindle-shaped bulbus, straight, black, simple, pointed embolus and flat, hyaline conductor with a broadened, sharply toothed tip.Both male specimens captured in Mont Ras had the right palp missing.The loss of the palp occurred quite some time before capture, as the scar had completely healed and was sclerotised.

Distribution
Only known from the Balearic islands and the Northern part of the Catalonian coastal range (Fig. 1H).

Discussion
The genus Harpactea has a quite homogeneous somatic morphology, but there is considerable genitalic diversity, suggesting the existence of a number of subgenera, or even the possibility that several related genera are involved.Simon (1893: 318) was the first to suggest a division of Harpactes Templeton, 1835 (preoccupied as the bird name Harpactes Swainson, 1833 (Trogonidae) and replaced by Harpactea by Bristowe 1939: 5) in three subgroups.However, only one of these groups concerns species presently considered as true Harpactea, the other two being composed of species belonging to Dasumia Thorell, 1875, Harpactocrates Simon, 1914and Parachtes Alicata, 1964.
The first to suggest a subdivision of Harpactea as presently delimited is Alicata (1966a: 192).He bases his subdivision on genitalic characteristics and recognises three large groups: one consisting of species reminiscent of Dasumia; another, consisting of two subgroups, having a male bulbus with a complex set of apophyses and including H. hombergi; and a third group, characterised by a simple, elongated bulbus with 1-3 apophyses (embolus, conductor, pseudoconductor (Brignoli 1978b: 481)), consisting of four subgroups and including H. corticalis (Simon, 1882).Brignoli (1978b: 481) recognises two groups: a corticalis group with a simple set of apophyses on the bulbus (consisting of six subgroups), and a hombergi group with complex apophyses (consisting of four subgroups).Deeleman-Reinhold (1993: 130) recognises four groups, based on a combination of genitalic and leg spination characters: a corticalis group, largely similar to Alicata's and Brignoli's group with a simple male bulbus, a hombergi group characterised by a complex male bulbus, a lepida group with a lamellar conductor and a wide PD, and a rubicunda group with a large, entirely membranous PD.Beladjal & Bosmans (1997: 24) follow the subdivision of Brignoli (1978b), but add a third group, the auriga group, characterised by a globular bulbus and a long, filiform embolus.
none of these subdivisions is entirely satisfactory, as they are purely phenetic and no cladistic analysis was performed.Although the hypothesis sounds plausible, there is no proof that the corticalis group is "le groupe le plus primitif" (Beladjal & Bosmans 1997: 24), as no phylogenetically valid outgroup comparison (Watrous & Wheeler 1981;Maddison et al. 1984) is available.There is even no solid proof that this group actually exists, id est, that it is not polyphyletic.
nevertheless, provisory as they are, the subdivisions proposed to date have some practical advantage, facilitating an overview of the genitalic diversity in the large genus Harpactea.Moreover, it has to be admitted that the corticalis group is indeed "abbastanza omogeneo" and "abbastanza ben delimitato dagli altri" (Brignoli 1978b: 481), so it can be retained for the time being, until the enormous work of a cladistic analysis of Harpactea has been undertaken.