Further notes on spiders from the Special Nature Reserve Zasavica (Serbia)

doi: 10.5431/aramit5107 Abstract. Zasavica is a wetland in western central Serbia with very little information on its spider fauna. During 2011 faunistic research was carried out; the material collected mostly using pitfall traps, but also sweep netting, beating and hand-collecting. A total of 3053 individuals were caught, and 107 species from 21 families were identified. Among these species, seven are first records for the Serbian fauna: Holocnemus pluchei (Scopoli, 1763), Dactylopisthes digiticeps (Simon, 1881), Walckenaeria alticeps (Denis, 1952), Pachygnatha listeri Sundevall 1830, Liocranoeca striata (Kulczyński, 1882), Phrurolithus minimus C. L. Koch, 1839 and Tibellus maritimus (Menge, 1875). Additionally, 59 species are new for the reserve. Beside a new species list for the reserve, some notes on these national records were made.

Research on spiders in Serbia has been neglected for years. From the first published data (Spasojević 1891) to the latest (Grbić et al. 2015), only 696 species were reported. As Serbia is a part of the Balkans, it could be marked as one of the most significant biodiversity regions in Europe (Savić 2008). The great number (379) of Balkans endemic spider species (Deltshev 2004), and Serbian endemic spiders (21) (Deltshev et al. 2003) support this hypothesis, but an understanding of its importance in support of more faunistic research is lacking. A rare positive example is the management of the Special Nature Reserve (SNR) Zasavica.
SNR Zasavica is a wetland in the western central Serbian region of Mačva. The majority of the protected zone includes the water surface of the Zasavica river and the Jovača, Prekopac and Batar canals. The second largest protected area is the Valjevac pasture with its mixture of semi-aquatic and hydrophilic vegetation and a dry pasture. Also under legal protection is the forest vegetation of the Reserve that contains various hydrophilic forests of European ash, poplar, willow and black alder (Obratov-Petković et al. 2007).
The first historical data reveal that only five spider species were recorded at the locality of Crna bara by Stoićević (1929): Araneus diadematus Clerck, 1757, Araniella cucurbitina (Clerck, 1757), Dipoena braccata (C. L. Koch, 1841), Heliophanus cupreus (Walckenaer, 1802) and Xysticus lanio C. L. Koch, 1835. Several years later, Drensky (1936 added Argyroneta aquatica (Clerck, 1757) to the list, found at the same locality (in Deltshev et al. 2003: 173), but after that no effort was invested in any kind of spider research in this area. Finally Grbić et al. (2011) created the first significant list of 104 spider species based on material collected during student scientific research camps that were organized only in August 2008August , 2009August and 2010 In this study, five species were recorded for the first time in Serbia and, except for D. braccata and X. lanio, the historical records could be confirmed. This preliminary list of species formed a basis for future faunistic and ecological studies.
In 2011, further collection was performed at the localities of Valjevac and Turske livade, with an emphasis on the faunistic data and species composition in flooded and non-flooded forest areas. The project was a part of the Master's thesis of the first author. The results presented here are envisaged as additions to the faunal list and a short presentation of spider species new to Serbia. The second purpose of the manuscript is to draw attention to ongoing habitat devastation of the area that is out of control, and which could affect survival of the species.

Material and methods
The Turske livade locality (44° 57' 32", 19° 31' 37") is situated on the left side of the Zasavica river, at an altitude of 78-80 m. It consists of small forest fragments and arable land (Obratov-Petković et al. 2007). About 80% of these forests are private property (Stanković pers. comm.). Wood harvesting happens often, with no supervision, so human influence is very intensive on these forest habitats. For the purpose of this research, two non-flooded forests (coded as non-flooded forests 1 and 2) and one flooded forest were chosen. Nonflooded forests 1 and 2 are deciduous forest fragments with dry vegetation, a thick duff layer and many bushes. The flooded forest has dense vegetation with lots of shade, and becomes flooded during spring. More details on the vegetation of the habitats are given in Tab. 1.
The Valjevac locality (44° 56' 10", 19° 31' 11") is also situated on the left side of the river and at the same altitude as the previous locality, and it consists of a large pasture, forest fragments and arable land. The pasture is property of the Reserve, while 80% of the woods and all arable land are private property (Stanković pers. comm.). On this locality, one nonflooded forest, one flooded forest, the riverside near the Visitors' Centre and a pasture site were chosen for fieldwork. The non-flooded forest has a dense canopy, which provides lots of shadow. The flooded forest shows a mosaic pattern of trees, shrubs and open spaces with tall herbs (Obratov-Petković et al. 2007). The vegetation on the Valjevac pasture is heavily influenced by the Podolian cow and Mangulica pig population. The riverside (Fig. 1) is under pressure from intensive tourism in the Reserve.
Collecting took place in 2011, from 19.05. to 19.11. The pitfall traps were made from two litre plastic bottles cut at 2/3 of their height, thus the dimensions of the traps were 15 cm in height and 11 cm in diameter. Plastic plates were fixed above the traps as a roof, using wooden sticks (Fig. 2). They were filled up to 2/3 with a fixative. At the beginning, vinegar (9%) was used as a fixative, but it was later replaced with 4% formaldehyde with liquid detergent. The change of fixative was made due to the bad condition of the material. A total number of 35 traps were set at two localities in six chosen habitats (Tab. 1). The traps were emptied once a month. Additionally, sweeping, beating and hand collecting methods were applied, but only at the Valjevac pasture.
Collected material was rinsed in water using a sieve with 0.05 mm diameter slits and then preserved in 70% alcohol.
There was a clear difference between the material collected in vinegar and in formaldehyde. The vinegar material was in a very bad condition, so identification was problematic; on the other hand, the formaldehyde species were well preserved.
Tab. 1: Description of habitats and number of traps during the 2011 campaign for spider research at SNr Zasavica (information taken from Kiš, 2012, Perić, 2012, Stojšić, 2012. Plant communities are given in order they appear -from water to land.
It is important to point out certain problems that occurred during the fieldwork which influenced the results. Some traps were removed earlier than others because of forest harvesting. From the flooded forest at the Valjevac locality they were removed on 16.09.2011 and from the non-flooded forest 2 at the Turske livade locality they were removed on 15.09.2011. As mentioned above, the forest fragments in the reserve are private property, so it was impossible to predict this kind of influence. At the riverside of the Valjevac locality, the traps were also removed earlier (30.06.2011) due to seasonal cane harvesting.

Results
A total number of 3053 adult spiders were caught (2109 males and 944 females) (Tab. 2) and 107 species from 21 families were recorded. From the total list of species, 7 represent the first records for the Serbian fauna (marked with * in Tab. 2), and 59 species are new records for the Reserve fauna (marked with x in Tab. 2). The most diverse family in the material was Lycosidae with 19 species, while the most dominant species in the material was Diplostyla concolor (562 individuals), followed by Urocoras longispinus (300) and Histopona torpida (267).
The majority of the material was collected by traps, while the following species were collected only by sweep netting, beating and hand-collecting and thus only on pasture at the   Species  46  19  36  17  44  23  33  Families  13  12  10  7  16  13  13  Individuals  699  134  766  36  609  410  400  3053  Males  348  90  563  21  490  356  242  Females  351  44  203  15  119  54 158 Species information ))/(( ))/(( ))/(( ))/(( ))/(( ))/(( ))/(( can easily be distinguished by the slender body and legs. Differences between the species in the genera are more or less solid, so determination is not so complicated (Efimik 1999). This species has a Holarctic distribution (World Spider Catalog 2014), and with records from almost all European countries (Nentwig et al. 2014) it was also expected in Serbia. Usually T. maritimus can be found both on humid and dry but also sunny places (Nentwig et al. 2014, Cera et al. 2010, and the riverside Visitor's centre in our study is just like that. But like in the case of Dactylopisthes digiticeps described earlier, the future of the species is uncertain since the riverside is under constant threat. If we add the fact that this species completely disappeared from Poland after the drying up of flooded areas (Kajak 1993), we have to suggest that this species should also be included in the list of protected species in Serbia to prevent the possibility of a loss scenario. Although they normally differ markedly in size, occasionally T. pinicola may superficially resemble young T. extensa and stayed unnoticed. This specimen was found by chance and caught by sweep netting. Apparently it has a Palaearctic distribution (World Spider Catalog 2014), but it is also considered very rare in Northern Europe (Roberts 1995). In Serbia this species was last recorded in 1985 at the Sokolica locality on the Veliki Jastrebac Mountain (Deltshev et al. 2003). Since then no recent records were made and its current status in the country cannot be determined, therefore its rediscovery in the protected area of the SNR Zasavica could be highly relevant for the species and for the future habitat protection plans of the Reserve.

Trochosa hispanica
Records. 3))/2(( 03.08.2011.Valjevac -non-flood forest; 2)) 03.08.2011.Valjevac -flood forest; 8))/3(( 12.10.2011. Turske livade -flood-forest Note. According to literature data T. hispanica was very recently found in Serbia by Ćurčić et al. (2007) in Čačak. But as Hepner & Milasowszky (2006) noted, many misidentifications of Trochosa females occur in museum and private collections. Considering also that no specimens from Serbia have been included in their revision or rechecked yet, there is a possibility that T. hispanica could be more widely present in Serbia, not only at the two recently discovered localities. Unfortunately, most of the old collection records are lost or could not be re-checked, so the current status and distribution of the T. hispanica as a Mediterranean species (Nentwig et al. 2014) in Serbia still needs to be confirmed.

Conclusions
Based on our first impressions, the number of new records in Serbia appears to be high for the small sampled area, but most of the species were already known in the surrounding countries (Nentwig et al. 2014), thus their presence at Zasavica was not unexpected. This research increased the total number of species in Serbia to 703, although this is only a fraction of what would be expected for this region. More important is that this is a one more step towards a future comprehensive species list. Spider research in the Special Nature Reserve Zasavica is also far from complete. Together with the current results, a total number rose to 163 species, but we could roughly estimate that at least the same number still awaits detection. In future studies, it should be investigated if there is any connection between species compositions and human devastation in the area, since unsupervised harvesting is still taking place.