MOLLUSCAN ASSOCIATIONS WITH THE CLAUSILIID ALINDA BIPLICATA ( MONTAGU , 1803 ) ( GASTROPODA : PULMONATA : CLAUSILIIDAE ) IN FLOODPLAIN FORESTS AT THE NORTH-EASTERN EDGE OF ITS RANGE

We quantified the mollusc assemblage composition in eight riverine forests inhabited by the door snail Alinda biplicata (Montagu) in Central Poland where it occurs at the north-eastern border of its distribution range. In each location, eight random 0.25 m2 plots were quantitatively sampled from a 400 m2 core area that was searched for additional species. A total of 54 species were found, composed of 46 terrestrial snails and slugs, six freshwater gastropod and two clam species. Abundances ranged from 220 to 4,400 ind.m–2 per location, with a maximum of 2,200 individuals per plot. The number of taxa ranged from 17 to 34 per location and from 3 to 23 per plot. A. biplicata occurred in each randomly sampled plot. The highest number of co-occurrences with Alinda was found for Carychium tridentatum and Nesovitrea hammonis. Additionally, forest-specific, wetland-specific and even dry habitat-specific snails can use the same patch of microhabitat. The lack of narrow ecological specialisation in A. biplicata may favour its presence in man-made habitats where it is sometimes considered an invasive native gastropod. The species list included invasive Arion vulgaris and some other species with a mainly West European distribution, such as Oxychilus draparnaudi, O. alliarius, Arion rufus, Cepaea nemoralis and C. hortensis. The eastern borders of the native ranges of these species reach into the western part of Poland but in the sampled region they seem to be recent arrivals. Our results indicate that “western” (Atlantic) faunal elements benefit from more balanced microclimate in wet floodplain forests, which possibly reduces macroclimatic constraints at their distribution border. key worDS: alder forest, Clausiliidae, diversity, Gastropoda, land snails, Mollusca, stream


INTRODUCTION
Alinda biplicata (Montagu, 1803) is a clausiliid inhabiting humid floodplain forests, woody successions, sometimes parks and gardens, and other shaded habitats (kerney et al. 1983).It usually lives on the ground, but often climbs tree trunks and occasionally can be found on exposed rocks and old walls (wiktor 2004).The species reaches its north-eastern distribution border in Poland: in the lowlands it does not cross the Vistula river, while in the Carpathians it reaches to the Wisłoka river valley (rieDel 1988).The limited distribution of the species to the east probably is linked to some climatic factor such as the January isotherm of -2°C, which crosses Poland almost longitudinally and runs through its central part (vicinity of Łódź).
In the region of Łódź, A. biplicata was previously known from only two sites (piecHocki 1963(piecHocki , 1966)), but it recently was found in abundant populations in several floodplain forests (SulikowSkA-DrozD 2010, and unpublished).
The mollusc assemblage composition of some of these sites has not yet been studied quantitatively, thus we aimed at quantifying the malacocoenoses and intended to analyse species co-occurring with A. biplicata at the small spatial scale.Additionally, we wished to define the ecological requirements of the species found in riverine forest in Central Poland.

METHODS STUDY AREA
The study was conducted in the vicinity of Łódź (Łódź Upland), in the area adjacent to the main watershed between the Vistula and the Odra rivers (Fig. 1).We selected floodplain forests inhabited by Alinda biplicata.European alder (Alnus glutinosa) was the characteristic tree species in all forests, with admixture of Acer platanoides, A. pseudoplatanus, A. negundo, Quercus robur, Betula pendula, Carpinus betulus, Ulmus laevis and Picea abies at some locations.All sites were located at the upper courses of streams (0.5-3 m wide) that are tributaries of the Bzura River (Wrząca, Czarnawka, Moszczenica) and the Ner River (Łódka).The selected riparian forests are irregularly inundated, usually after local rainfall in the summer.Additionally, they tend to have moist areas from local groundwater discharge.Two of the locations are protected as nature reserves, namely "Grądy nad Moszczenicą" and "Polesie Konstantynowskie".
We recorded the local habitat characteristics in terms of the average breast height diameter of trees, the amount of deadwood per hectare and soil pH.We additionally estimated the cover of the tree, shrub, herb and moss layer (to the nearest 5%) and identified tree species.The habitat characteristics are presented in Tables 1-4.

SAMPLING
Sampling was performed in spring and early summer 2010.We considered one quadrat of 20 m × 20 m per location, as suggested for land snail forest sampling by cAmeron & pokrySzko (2005).In each of the eight locations, we took substrate samples, such as decaying herbs and leaf litter, from eight random plots to quantify alpha diversity and density of the molluscs.Each plot covered an area of 50 cm × 50 cm (0.25 m²), thus a total of 2 m² were sampled at each site.Only living specimens were counted.We additionally collected molluscs from tree trunks and fallen logs within the location.These species were added to the species list of the location (the beta diversity), but not to assemblage density calculations.
The identification of molluscs followed kerney et al. (1983), wiktor (2004), Glöer &meier-brook (2003) andpinceel et al. (2004).Taxonomic placement followed welter-ScHulteS (2012) but based on norDSieck (2007) we used for our focal clausiliid the generic name Alinda instead of Balea.Authorities for species names are given in Tables 1-4.The distribution ranges of some species were classified as "western" according to the maps in kerney et al. (1983), wiktor (2004) and welter-ScHulteS (2012).We also adopted the classification of mollusc species into ecological groups proposed by Ložek (1964) and       Species relative abundance and frequency of occurrence were also calculated from the plot data.At each site four most abundant terrestrial species (A.biplicata excluded) were classified as four top dominants, then Pearson correlation coefficients of their abundance and abundance of A. biplicata per plot were calculated in Statistica, Version 10.

SPECIES RICHNESS
At the eight studied localities, we found a total of 54 species.Of these, 39 were terrestrial snails and seven species were slugs.Freshwater molluscs were represented by six gastropod species and two clam species.At a single locality the total number of mollusc taxa varied two-folds, from 17 to 34 species (number of land snails ranged from 14 to 25; Tables 1-4).Within the 0.25 m 2 plots, a total of 52 species were recorded.The number of taxa per plot ranged from 3 to 23 (Fig. 2).Additional sampling by visual search in each locality revealed a few species which were not found within the eight plots.These species were added to the data in Tables 1-4.The values for Chao1 and Chao2 (Table 5) suggest that the species richness in some locations might be substantially higher than recorded in the eight forest floor plots.

ABSOLUTE AND RELATIVE ABUNDANCE OF SPECIES
The abundance of mollusc assemblages revealed by quantitative sampling ranged from 222 to 4,407 ind.m -2 per location (only live individuals were counted).The highest mollusc density per plot (0.25 m 2 ) was 2,238 individuals, the lowest density was 22 (Fig. 2).The abundance of A. biplicata per plot ranged from 2 to 475 individuals.The average density of its population estimated from plots ranged from 70 ind.m - to 1,000 ind.m -2 at a location.
In each location, A. biplicata was among most abundant species with the relative abundance above 10%.The extreme values of its relative abundance reached 28% at location 1 (Szczecińska).Other terrestrial species predominating in the assemblages (four top dominants) are listed in Table 6.No species was found in more than five out of eight sites as one of the top dominants.Carychium tridentatum and N. hammonis were the only species in top four in sites and present in all remaining localities.Considering all mollusc taxa also Anisus leucostoma and Pisidium personatum reached relative abundance above 10% at a single location (Helenówek).Abundance of A. biplicata per plot was positively correlated with the number of individuals of Carychium tridentatum, C. minimum, Zonitoides nitidus, Vitrea crystallina, Nesovitrea hammonis, Succinea putris, Fruticicola fruticum, but not with the number of Aegopinella pura, Discus rotundatus, Punctum pygmaeum, Vertigo pusilla and Perforatella bidentata (Table 6).

HABITAT PREFERENCES
The assemblages were composed of species that were classified as indicator species of several different habitats (Fig. 3).Typical forest dwellers (group 1), woodland species (group 2) and wetland species (group 9) were represented by several taxa, while eurytopic species (group 7) predominated in terms of species richness (Fig. 3A).The ecological groups had a similar distribution in terms of the numbers of individuals (Fig. 3B), but here, clams such as P. personatum (lotic habitats, group 12), were additionally highly abundant.

WEST EUROPEAN ZOOGEOGRAPHICAL ELEMENTS
We found six species that have their distribution ranges confined to Western or Western and Central Europe.These occurred in species pairs from the same genus, namely: Oxychilus draparnaudi, O. alliarius, Arion rufus, A. vulgaris, Cepaea hortensis and C. nemoralis.Of these species, O. alliarius and C. nemoralis occurred most frequently (at 6 and 4 locations, respectively) whereas C. hortensis was only found in a single location.The number of "western" species per location ranged from one to five.Three species or more were found at Szczecińska, Polesie and Helenówek localities.The relative abundance of the "western" species was higher than 5% at Szczecińska, Polesie and Dzierżązna, mostly because O. alliarius reached high densities (see Tables 1-4).

DISCUSSION SPECIES RICHNESS AND COMPOSITION
The riparian alder forests harbouring A. biplicata appear to be rich in terms of species numbers (17-34 species per 400 m 2 ) but it is partly the effect of including the freshwater gastropods and clams cooccurring with land snails in swamps.The number of terrestrial shelled species, here 14-25, is below the richness reported for Polish lowland forests (17-35 species) (cAmeron et al. 2010).Regionally rare species such as Platyla polita, Cochlicopa nitens, Nesovitrea petronella and Ruthenica filograna were recorded, even though forest habitats are fragmented in the vicinity of Łódź (Fig. 1).
Our sampling design probably underestimated the species richness in some locations.For example, the number of terrestrial species found at Grądy nad Moszczenicą (23 species) was below values previously recorded by SulikowSkA-DrozD (2010) from the whole area of this nature reserve (34 species).The discrepancy may be at least partly associated with the heterogeneity of mesohabitat conditions within the reserve area, but diversity estimators also indicate that the randomly sampled total of 2 m² of forest floor did not cover total species richness within the location.
We found a surprisingly high number of species that have the cores of their distribution ranges in Western Europe and reaching as far as the western part of Poland with their native ranges (distribution maps published in kerney et al. (1983)

HABITAT PREFERENCES OF A. BIPLICATA
The 100% frequency of occurrence of A. biplicata within the studied plots suggests a wide tolerance of this clausiliid for light and moisture conditions, at least in the studied type of forest habitat.The study shows that A. biplicata has viable populations that can reach high densities (up to 1,000 ind.m -2 ), especially if sampled in late spring/early summer when the reproduction peak of the species occurs (SulikowSkA-DrozD et al. 2013).Alinda biplica-ta is accompanied by eurytopic species, but also by forest-specific, wetland-specific and occasionally even by dry habitat-specific species.High abundance of A. biplicata in the riverine forests may favour its spreading not only along rivers but also to adjacent drier habitats.
The lack of narrow microhabitat specialisation may favour the presence of A. biplicata in man-made habitats.According to kerney (1999) and welter-ScHulteS (2012), A. biplicata was introduced several hundred years ago in southern England, where it lived in ground litter at the river banks, commonly associated with human rubbish, and has also been successfully introduced to gardens.Spreading of A. biplicata on a local scale was recently recorded in NW.Germany (kobiAlkA et al. 2009).The number of occupied grids increased in Northrhine-Westphalia during the last decade(s) so it has been suggested that the species is an invasive native gastropod.Laboratory observation of its breeding biology revealed that A. biplicata overcomes environmental fluctuations by intrauterine egg retention and viviparity and possesses other life history traits of a good coloniser (mAltz & SulikowSkA-DrozD 2012).
Assemblages of land snails and slugs occurring in riparian forests have been sampled by different authors, both from montane (DyDucH-FAlniowSkA & tobiS 1989, SulikowSkA-DrozD & HorSák 2007) and lowland rivers (Čejka 1999(Čejka , 2005(Čejka , Čejka et al. 2008(Čejka , kAppeS et al. 2007).According to Čejka et al. (2008), the ecological factor with the highest influence on mollusc species composition in floodplain forests is humidity, which is partly linked to flood frequency.At the scale of microhabitats, however, the co-occurrence of floodplain molluscs is still poorly understood and needs to be studied in more detail.

Fig. 1 .
Fig. 1.Geographic position of the studied locations: 1 -Szczecińska, 2 -Polesie Konstantynowskie, 3 -Grądy nad Moszczenicą, 4 -Rosanów, 5 -Dzierżązna, 6 -Wola Branicka, 7 -Helenówek, 8 -Czarnawka 24 23 21 19 13 12 16 15 13 17 15 Number of individuals per subsample 722 1,192 2,238 963 1,248 823 862 767 333 251 194 285 138 195 203 293 Number of species (total) 34 27 Number of individuals per 1 m 2 4,407.5 948 modified by AlexAnDrowicz (1987).This was restricted to snails and clams, because slugs were not always classified on the species level.STATISTICS ANOVA with Tukey HSD post-hoc test was used to compare the number of mollusc species and densities per plot among locations.A further comparison between locations was performed using the diversity estimators Chao1 and Chao2.When compared to the actual number of species found, these two indices provide an estimation of the completeness of sampling based on the number of species found only as one or two individuals, or in only one or two plots, respectively.Chao1 and Chao2 were recently suggested for forest gastropods by cAmeron & pokrySzko (2005).Chao1 and Chao2 values were calculated for each location separately in EstimateS, Version 8.2.0 (colwell 2006).

Fig. 2 .
Fig. 2. Comparison of mollusc species numbers and densities per plot (n=8 per location) at different sites.Shared letters indicate no significant difference between locations (ANOVA with the Tukey post-hoc test) the species occurs in several localities in Łódź and its vicinity(piecHocki 1963, SulikowSkA-DrozD 2011,  this study).We found also the major invasive slug in Europe -A.vulgaris, previously mistakenly recognised as A. lusitanicus (welter-ScHulteS 2012).The first record of this slug from Poland dates back to the 1990s (kozłowski & kornobis 1995) and from Łódź to 2007(SorokA & kałuski 2011).Here, we found these alien species in semi-natural floodplain forests indicating that they are able to establish populations outside cities in Central Poland and may potentially be spread further along truck roads and rivers.All these species might be regarded in Central Poland as recent arrivals from the west.It is likely that the mild microclimatic conditions in the moist floodplain forests mimic the lack of extreme cold conditions in the Atlantic Bioregion of Europe (e.g.metzGer et al. 2005) and thus support the presence of these snails and slugs.The microclimate of all the locations is influenced by both the river and groundwater discharge.It is known that high soil moisture and rich vegetation can buffer daily and annual temperature amplitudes (GeiGer et al. 1995, and references therein).We suppose that the reduced cold and heat stress in floodplain forest might favour the presence of A. biplicata at the border of its distribution area.Similarly, Cepaea occurred in ground water discharge areas in post-glacial Northern Europe (JAeckel 1949), indicating that microhabitat use can ameliorate macroclimatic constraints.Our results indicate the existence of some zoogeographical gradient in Central Poland where several species reach the easternmost distribution limit, probably in connection to the increasing continentality of climate.This was suggested by pokrySzko & cAmeron (2005: p. 123, fig.5) who calculated the similarity of land snail faunas across Europe.The malacocoenoses of Central Poland have ca.40% similarity (Nei index) to westernmost localities in Ireland and 20-30% similarity to malacocoenoses of the Carpathians, while Ireland is more distant.Malacocoenoses located more to the east at the same latitude seem to have more affinity to the Carpathian fauna.

Table 1 .
Habitat characteristics, list of molluscs and their abundance in localities 1 (Szczecińska) and 2 (Polesie Konstantynowskie).Grey fields indicate the presence of the species at the locality but not in the quantitative samples

Table 2 .
Habitat characteristics, list of molluscs and their abundance in localities 3 (Grądy nad Moszczenicą) and 4 (Rosanów).Grey fields indicate the presence of the species at the locality but not in the quantitative samples Nesovitrea petronella(L.Pfeiffer, 1853)

Table 4 .
Habitat characteristics, list of molluscs and their abundance in localities 7 (Helenówek) and 8 (Czarnawka).Grey fields indicate the presence of the species at the locality but not in the quantitative samples -Vallonia pulchella (O.F. Müller, 1774)

Table 5 .
Species richness recorded at the eight localities and species richness estimation (Chao1 and Chao2 bias-corrected formula) based on the eight replicates covering a total area of 2 m² per location

Table 6 .
Fauna associated with A. biplicata.Relative abundances calculated for terrestrial gastropods alone , wiktor (2004) and welter-ScHulteS (2012)).For example O. draparnaudi, C. nemoralis and C. hortensis were regarded as the Atlantic zoogeographic elements and in Central Poland inhabit only man-made habitats (rieDel 1988).According to rieDel & wiktor (1974), Arion rufus occurs in natural habitats in Western Poland and its native range should not reach the area of our study.Its population in Łódź was probably established by introduction in the 1960s (SulikowSkA-DrozD 2007).Similarly, the natural distribution of O. alliarius is confined to the influence of oceanic climate (rieDel 1988).At present,