ON RONCUS ALMISSAE N. SP., R. KRUPANJENSIS N. SP., AND R. RADJI N. SP., THREE NEW PSEUDOSCORPIONS (PSEUDOSCORPIONES, NEOBISIIDAE) FROM CROATIA AND SERBIA, RESPECTIVELY

– Three new species of the pseudoscorpion genus Roncus L. Koch (Neobisiidae) are described from Croatia (from nr. Omiš, Dalmatia: R. almissae n. sp.) and Serbia (near the town of Krupanj, north-western Serbia, Lukića Pećina Cave and nr. Izvor: R. krupanjensis n. sp., and R. radji n. sp.), and their diagnostic characteristics are illustrated. Their interrelations with phenetically close congeners are analyzed; in addition, the presence/absence of microsetae proximal to the trichobothria eb and esb is established as an important taxonomic characteristic.


INTRODUCTION
Over the past four decades there has been a marked increase in our knowledge of the Neobisiidae of south-eastern Europe (the Balkan Peninsula), and especially of the representatives of the genus Roncus L. Koch, 1873 which occur in leaf litters, soil and caves (Ćurčić, 1988;Ćurčić et al., 2004;Harvey, 1990).Increased interest in the soil/litter and cave ecosystems and improved sampling techniques have contributed to this knowledge.During a study of postembryonic development and teratology of the pseudoscorpions in Dalmatia and Serbia, three hitherto undescribed species of Roncus were found.
This paper provides descriptions of Roncus almissae n. sp., R. krupanjensis n. sp., and R. radji n. sp., with some details on the comparative morphology of both sexes.
All specimens are mounted on slides in Swan's fluid (gum chloral medium) and all are deposited in the Institute of Zoology, Faculty of Biology (IZB), University of Belgrade, Belgrade, Serbia.

SYSTEMATIC PART RONCUS ALMISSAE, NEW SPECIES
Etymology.-After Almissa, the old Latin name of Omiš Material.-Holotype male and allotype female samples residing under stone were collected by Tonći Rađa in the village of Podašpilje, nr.Omiš, on the northern slopes of Mt.Omiška Dinara, Dalmatia, Croatia, 22 September.
Galea (cheliceral spinneret) low (Figs. 8 and 15).Cheliceral palm with six setae, movable finger with one seta (both in male and female).Cheliceral dentition as in Figs. 8 (male) and 15 (female).Eight-bladed flagellum (Figs. 5 and 14); one short proximal blade and 7 longer blades distally, all blades denticulate.Apex of pedipalpal coxa with 4 long acuminate setae.Pedipalpal trochanter with a small lateral tubercle and some rare tiny and inconspicuous denticulations dorsally.Pedipalpal femur with a small exterior and lateral tubercle and with interior and dorsal granulations as in Figs. 5 (male) and 12 (female).Tibia smooth; chelal palm with tiny interior granulations or smooth; exteriorly palm with some rare and inconspicuous surface irregularities (Figs. 5 and 12).No microsetae proximal to eb and esb (Figs. 2 and 9); however, 4-6 microsetae present distally or laterodistally to eb and esb.In both sexes, sensillum located between the 10 th and 17 th teeth.The trichobothrium ist slightly closer to isb than est, or equidistant from them (Figs. 2 and 9).
Chelal fingers generally as long as the chelal palm and shorter than the pedipalpal femur (Table 1).Pedipalpal femur shorter (male) or slightly longer than carapace (female) (Table 1).Trichobothriotaxy as in Figs. 2 and 9. Tibia IV, basitarsus IV and telotarsus IV each with a single tactile seta (Figs. 6 and 13).Tactile seta ratios are presented in Table 1.
Based on present knowledge, R. almissae n. sp. is known from its type locality only.
Etymology.-After Krupanj, a town near the type-locality of R. krupanjensis n. sp.
Trichobothriotaxy: eb, esb, ib, and isb on finger base; it, et, and est in proximal half of finger; ist slightly closer to est than to isb (or equidistant from these).Seta sb only slightly closer to b than to st, st closer to t than to sb.For trichobothrial pattern, see Figs. 17 and 25.Tibia IV, basitarsus IV, and telotarsus IV each with a long tactile seta (Figs.19 and 27; Table 1).For morphometric ratios and linear measurements, see Table 1.
Distribution.-Western Serbia, epigean, under stones, and in humus and leaf-litter.Probably endemic to Serbia and the Balkan Peninsula.
Remarks.-The present species is distinct from the phenetically close congener R. tintilin Ćurčić, 1993, in many important respects: body size (2.28 mm in the male of R. krupanjensis n .sp. vs. 2.84-3.58mm in R. tintilin), in the pedipalpal length (3.37-3.85mm in krupanjensis vs. 3.97-4.57mm in R. tintilin), in the form of the epistome (apically blunt in krupanjensis vs. triangular i tintilin), in the form of the pedipalpal articles, in the cheliceral dentition, and the presence (in krupanjensis) vs. absence of small setae proximal to eb and esb (in tintilin).
Etymology.-The town of Krupanj is the center of the Radjevina region, which was named after Radj, a great knight of Prince Lazar, who defended it from Hungarian and Ottoman conquerors.The new species is therefore named after this nobleman.
Fixed chelal finger with 79-81 teeth; distal teeth pointed and asymmetrical, followed by small, closely-set, and square-tapped or rounded teeth proximally.Movable chelal finger with 72-77 teeth; only distal teeth pointed and retroconical, other teeth square-cusped or rounded.Chelal fingers longer than chelal palm and considerably longer than carapace (Table 1).Tiny microsetae proximal to eb and esb absent; chelal palm with 4 microsetae distal to these trichobothria (Fig. 40).
Trichobothriotaxy: eb, esb, ib, and isb on finger base, it, et, and est on proximal finger half; ist slightly closer to isb than to est.Seta sb equidistant from b and st, respectively, seta st closer to b than to sb, respectively.For trichobothrial ratios and linear measurements, see Fig. 40 and Table 1.
Distribution.-Western Serbia, cave-dwelling.Probably an endemic and relict species.
Remarks.-R.radji n. sp. is easily distinguished from R. trojan Ćurčić, 1993 (its phenetically most similar species), from southeastern Serbia, by the (male) body size (3.395-3.74mm vs. 2.415-3.07),by the number of setae on sternite II (20-21 vs. 12-13), by the presence/absence of microsetae proximal to eb and esb (absent vs. present), by the number of teeth on the fixed (79-81 vs. 51-63) and movable chelal fingers (72-77 vs. 56-63), by the pedipalpal femurs length-to-breadth ratio (4.52-5.54 vs. 3.27-3.55),by the pedipalpal chela length-to-breadth ratio (4.125-4.285 vs. 3.27-3.705),etc.The discovery of the described representatives of Roncus in Serbia (and Dalmatia) supports the fact that the taxonomy of this genus is still far from being complete (Ćurčić, 1991, 1992a, 1992b; Ćurčić, et al., 1992a, 2004).The variety of cave-dwelling species of Roncus described elsewhere by Ćurčić (1984, 1991) and by Ćurčić et al. (1981, 1988, 2004), offers further proof that this genus is presently subjected to intensive radiation or divergent differentiation into new species.Furthermore, the diversity of Roncus representatives in the Balkan regions bordering on Serbia (Ćurčić, 1984;Ćurčić and Beron, 1981;Ćurčić et al., 2004) compared to the same feature in other areas, points to the Balkan Peninsula as a center of origin and genesis of numerous forms of the genus.In addition, the occurrence of numerous Roncus species with extremely limited distribution areas demonstrates their endemic nature.

NOTE
With regard to a single diagnostic characteristic (presence/absence of microsetae proximal to eb and esb), it should be noted that this feature is present in R. krupanjensis n. sp.(as well as R. pannonius Ćurčić, Dimitrijevic and Karamata, 1992 in R. trojan Ćurčić, 1993, andin R. lubricus L. Koch, 1873. However, these microsetae are missing in other epigean and cave species of the genus which inhabit the Balkan Peninsula (R. onaemi n. sp., R. radjai n. sp., R. parablothroides Hadži, 1937, etc).Therefore, it is possible that the presence or absence of this characteristic could be useful in distinguishing representatives of two species groups, which we have described as the "Roncus lubricus" (microsetae present) and "Roncus parablothroides" (microsetae absent), respectively.It seems that both groups are widespread in Europe (Ćurčić, 1992, 1992b); however, their precise taxonomic and biogeographic features are insufficiently known.Therefore, this problem remains one of the main goals for future research.