Spider (Arachnida: Araneae) distribution across the timberline in the Swiss Central Alps (Alp Flix, Grisons) and three morphologically remarkable species

We collected 6251 adult epigeic spiders from the dwarf-shrub heath to subalpine coniferous forest on Alp Flix (CH, canton Grisons, 1950 m) between May 2005 and May 2006 using pitfall traps. Total species richness and activity density of all species decreased from the open land to the forest, although this pattern varied according to family. The distribution of the 102 species found indicates that the small area around a single tree at the timberline provides habitats for both open land and forest spider species as well as some possible timberline specialists. Five species were new to the canton Grisons: Centromerita bicolor, Centromerita concinna, Hilaira excisa, Meioneta alpica and Tallusia experta. Three species showed remarkable morphological characteristics and were analysed in more detail. We found males of Pelecopsis radicicola without the characteristic longitudinal depression on the raised carapace. It is shown that the males of Meioneta alpica have a considerably variable lamella characteristica, which is nevertheless distinct from the sister species Meioneta ressli. Because we found intermediate forms of the head region described for Metopobactrus prominulus and M. schenkeli, respectively, M. schenkeli is considered a syn. nov. of M. prominulus. This study shows that the known distribution and taxonomic status of various spider taxa in the Central Alps are still incomplete and further work on arthropods in remote areas should be strongly encouraged.

Knowledge about species diversity and activity across landscapes is indispensable to understand how landscape structure influences habitat quality. Since landscapes exist as mosaics of numerous different patch types, knowledge about the ecology of habitat edges is of particular importance (e.g. RIES et al. 2004). Spiders are abundant, species-rich and known to respond sensitively to environmental and structural conditions, which makes them suitable for studying organism-habitat relationships (e.g. WISE 1993, FOELIX 1996. However, despite extensive work, our knowledge about the distribution pattern of spiders across the Central Alps in general and Switzerland in particular is still limited (MAURER & HÄNGGI 1990, HÄNGGI 1993, THALER 1995, HÄNGGI 1999, HÄNGGI 2003. A recent study analysing the influence of environment and space on the differentiation of spider communities across an alpine timberline in Switzerland (MUFF et al. in prep.) provided faunistic data, which are reported here. Besides a short description of the distribution pattern of species from the open land to the forest, we focus on three morphologically remarkable species, two of them causing problems of identification since their first description. A discussion about other faunistically notable species of the study site can be found in FRICK et al. (2006).

Study site
Alp Flix (WGS84: 9°38'E, 46°31'N) is part of the Swiss Central Alps and belongs to the village of Sur in the canton Grisons, Switzerland. The alp is a southwest-exposed terrace of 15 km 2 at 1950 m above sea level. It is surrounded by 3000 m mountain peaks and a valley. Sampling was conducted in a 300 m long stretch of timberline plus fragments of the adjoining Norway spruce forest (Vaccinio-Piceion) to the northwest and the dwarf-shrub heath ( Juniperion nanae) to the southeast. Each of the three parts covered approximately 3 ha. The site is located on a small slope inclined slightly towards the forest and is used for occasional cattle grazing throughout the vegetation period. For a more detailed description of the study area see HÄNGGI & MÜLLER (2001).

Study design and spider sampling
We differentiated between five habitat zones which represented the whole gradient of habitat structures: the open land (dwarf-shrub heath, O), three microhabitats linked to a single spruce tree at the timberline and the forest (F). The three areas at the timberline were defined by their location relative to the tree as: next to the trunk (TT), at the end of branch cover (TB) and in the adjoining open area outside of branch cover (TO). In each of the five habitat zones we placed 15 pitfall traps. In O, F and TO the traps were randomly positioned at least 15 m apart from each other. For placing the traps in TB and TT, the tree nearest to the trap in TO was chosen. The mean distances between the traps in these three microhabitats were 4.1 m (TO -TB), 4.5 m (TO -TT) and 1.5 m (TB -TT). The traps consisted of white plastic cups with an upper dia-meter of 6.9 cm and a depth of 7.5 cm filled with a solution of 4% formaldehyde in water plus detergent (0.05% sodium dodecyl sulphate, SDS). Each trap was covered with a quadrangular transparent plastic roof (15 x 15 cm) fixed by three wooden rods 8 cm above ground. Due to the proximity of cattle and the toxicity of the trapping liquid we fenced off each trap with three plastic poles connected with ribbons. The traps were emptied monthly during the snow-free period (May 2005to October 2005 and then left under the snow layer until May 2006, when they were emptied a final time.

Identification
Only adult spiders were identified to species level, juveniles were excluded. Identification was mainly carried out using NENTWIG et al. (2003), ROBERTS (1985ROBERTS ( , 1987 and WIEHLE (1956WIEHLE ( , 1960 TO TB TT  F  F  TT  TB  TO  O   O  TO TB TT  F  F  TT  TB  TO  O   O  TO TB TT  F  F  TT  TB  TO  O all species N = 6251 Alopecosa pulverulenta N = 639 Caracladus avicula N = 196 Alopecosa taeniata N = 429 Scotinotylus alpigena N = 146 Cryphoeca silvicola N = 134 of individuals in open areas, however, must be qualified when looking at single families. While Lycosidae, Gnaphosidae, Thomisidae and Philodromidae had higher densities in open zones, Linyphiidae, Theridiidae and Hahniidae preferred more shaded habitats. Clearly, in the three habitat zones at the timberline the species mostly followed these distinct trends in activity density across the open land and forest in a very gradual manner (Tab. 2). It is notable that all five habitat zones were dominated by certain species, but no species (with N ≥ 15 individuals) was found exclusively in only one habitat. Nine species were found with at least 50% of the individuals in the open land, six species in the forest and none, four and six species, respectively, in the three zones at the timberline (TO, TB and TT) (Fig. 1). Fig. 1 illustrates these species-specific patterns in activity density. For example, the lycosid Alopecosa pulverulenta clearly preferred the open land, while A. taeniata was found mainly around single trees in the timberline. Other species (e.g. Scotinotylus alpigena, Cryphoeca silvicola) preferred shaded habitats and were found almost exclusively close to the tree trunks in the timberline or inside the forest, respectively. Caracladus avicula has been discussed as a habitat-specialist of the timberline by FRICK et al. (2007). Our results are partly in accordance the alpine timberline with its stand-alone trees provides habitats for both open land and forest spider species as well as some possible timberline specialists. We thereby demonstrate the particular value of the timberline and of heterogeneous, spa-  with this, as most individuals of this species were found there. However, its status as a timberline specialist must be questioned, since in this study it also occurred in considerable numbers in the open land and inside the subalpine forest (40 and 33 individuals, respectively) (Tab. 2).
Our findings indicate that we show the distinction of the two species based on the lamella characteristica (Fig. 3)

Conclusion
Our study demonstrates that the known distribution and taxonomic status of various spider taxa in tially limited structures in general in maintaining divergent invertebrate communities. For a more detailed description of the habitat zones and a full analytical discussion regarding the results see MUFF et al. (in prep.).

Some remarkable species
Our study revealed one species new to science (Caracladus sp. A, Tab. 2) which will be described elsewhere. Five species were new to the Canton Grisons: Centromerita bicolor, Centromerita concinna, Hilaira excisa, Meioneta alpica and Tallusia experta.
Other remarkable species included Meioneta orites and Panamomops palmgreni, both endemic to the Alps. The following three species showed remarkable morphological characteristics and were analysed in more detail.
Pelecopsis radicicola (L. Koch, 1872) THALER (1978) considered Brachycentrum delesserti Schenkel, 1925 a synonym of Pelecopsis radicicola since the characteristics of the first species were within the variation range of the latter. The only difference was the absence of the median longitudinal depression on the elevated male head region in B. delesserti. Our material from Alp Flix comprises individuals showing practically identical male and female genital organs, but a great and gradual variability regarding size and shape of the raised male head region (Fig. 2). This clearly emphasises the retention of only one species as suggested by Thaler.