Terrestrial snail fauna in the Someg/SzamosrRiver valley from the spring region to the inflow into the river Tisza

Kdroly Bdba and Andrei Sdrktinv-Kiss

Abstract

The sampling sites included montaneand lowland riparian plant communitiesof

ecologicaltolerance range can be found in the riparianvegetation of the Someg/Szamos Valleyand alsoof minor watercoursesarriving from neighbouringhills and mountains. The fauna dispersalfrom the spring region is hamperedby water reservoirs,shrinkage of foresthabitats to single row of treesalong the river and the lack of riparian gallery forests in the flood area. The biotopesare subjectedto the strongesthuman impaCt resultedfrom the forestmanagement, or loggingthe flood plain into agriculturalfietds. The garbagedisposal in/or nearthe riparianvegetation also cause major disturbancesfor the snailassemblages. Keywords:, human impact,someg/Szamos River valley

Introduction

A scientificexpedition was organisedin the valley of river Someg/Szamosbetween I - I 8. 06. 1992with the aim of assessingthe currentbiotic statusof the river. During the field investigationsspecial attention was paid tothe influenceof humanactivities on the microfloraand invertebratefauna of the river,and on its naturalself-purification. It was also followed how the prevailingwater quality and the environmentalconditions affects the capacityof the river for transportingfauna elements along its course.This processis decisive in settling of montane invertebrates(e.g. snails) of more or less narrow ecologicaltolerance range in plantcommunities on the GreatHungarian Plain, even far beyondthe bordersof Romania.The invertebratefauna populating the environsof the river Szamoswere studiedby Erd6s(1935) and by B6ba(1968-69, 1970, 1972, 1973, 1974,1975,1977,1978,1979,1980-81,1983a, b, 1986,1991,l9g2a), studying beetles and snailsrespectively.

I Thefirst nameis Romanian,and the second Hungarian 279 Msteriuls and metlrods

Sixteenmajor samplingsites were appointedfor the expedition,but not all of them were appropriatefor collectingsnails. Snail data were gatheredfrom further localities (lb, 16, la, 19-25)during rest periods.Mostly the quadratemethod (with plots of 10x25x25cm) was usedfor samplingsnail assemblages, but on sites19-25 snails were collectedby thinning. For comparison,two additionallocalities (17-18 quadrates) situatedclose to the country borderswere also analysed.The samplingsites, the snail speciesfound and severalcalculated parameters are listed in Table.The field data(from both quadratesand thinning) were analysedby ecological and zoogeographical techniques. During the data analysiswe focusedon two main points,namely on the dispersal ability of speciesand on theecological status of snailassemblages in plantcommunities studiedby the quadratemethod.

An area-analyticalzoogeographical technique (Bdba, 1982) was usedto study the dispersalcharacteristics of snailspecies. The following faunagroups were distinguished:climatically continental groups: l. Siberian-Asian(1.1. East Siberian, 1.2. West Siberian, 1.3 Euro-Siberian, 1.4. Holarctic), 3. Kaspi-Sarmatian,4. Ponto-Pannonic,9.5. Daco-Podolian,10.1. Boreo-alpine; climaticallysubatlantic groups 2.1. CentralAsian xeromontane,5. Ponto-Mediterranean (5.2.1. Quercion frainetto, 5.2.2. Fagion illyricum-moesiacum),6. Adriato- Mediterranean,7. Atlanto-Mediterranean,8. Holomediterranean,9. Central European montane(9.1. Carpathian,9.2. Carpatho-sudetian,9.3. Carpatho-Baltic,9.4. Alpo- Carpathian),10.2. Boreo-montane.

The speciesdistribution data reported by Grossu(1981, 1983,1987), So6s (1943) and Lozek (1964) were also considered in bielzi, Carpathica calophana, Cochlodina marisi, Balea stabilis, Vestiaelata, Vestiagulo and Trichio bielzi in groups 9.1. or 9.2.. The occurrenceof Vitrea transsylvanicaand Balea fallax in Bulgaria (Serafim-Liharev,1975)justified to put them into the group5.2.2.

In order to observechanges in the structuralcomposition of snail assemblagesthe following variableswere used:abundance (A/m2), speciesdensity (mean number of speciesin ten quadrates),percentage proportion of juvenile individuals,percentage of mortality, Shannon-Wienerdiversity (H') and the habitat typology system of Lozek

(S, createdfrom the groupsof O, X, Sf and S)' The ecologicalgroups obtained by the block clusteringmethod of Feoli & Orl6czi (lg7g) (B6ba& Podani,1992b) were used in accordancewith the dataof Grossu(1981, 1983, 1987), So6s (1943), Lisicky (1991) and Kerney et al. (1983) for the

280 zoogeographicalgroup assignmentsof species listed before. Five categorieswere recognised:l. hydrophilousmontane-submontane species, 2. subhydrophilousspecies, 3. riparian hydrophilousspecies, 4. mesophilous-mesoxerophilouselements, 5. ubiquitariancomponents. The relationshipsamong snail assemblageswere analysedby the Sokal-Michener group averageclustering method and by PrincipalCoordinates Analysis (PRINCOOR, Podani1988). The characteristicsused for the analysis of structural composition of snail assemblagesare listed in Table. Acknowledgements:We would like to thankto Gilbor Majoros and Mikl6s Szekeres for the identificationof slugs and revisingthe Clausilidaegroup; to M6rta Bdresand Lorincz Istviln for providing their own field collections,to GheocaVoichi{a for his help during the fieldwork, to botanistConstantin Drdgulescu for his assistancein identifying plantcommunities in the field,and to JakabS6muel for the localisationof samplingsites and supplyingbasic data on river fall, on altitudesand on bedrock.

Sampling sites and plant communities

The sampling sites listed in Table L are of two types dependingon the sampling techniqueused (thinning or quadratemethod), which is reflectedalso in theirnumbering. For each plant community the altitude above sea-level,annual precipitation (Tufescu, 1965)and magnitudeof river fall (m/km) aregiven in this order.The numberssigning siteswere inappropriatefor snail collectionare omitted.

Gildu Mountains;Somegul Cald

l.a. Cliffs abovethe springBazarul-Someqului,03.06. 1992 (thinning) by Gheoca V., 1250-1300m a.s.l. 1.b. Ic Ponor,150 m downstreamof the spring,a Petasitetum albae(Klika, 1954)stand of about20 cm height(1200 m a.s.l.,limestone bedrock, 1500 mm, 5 m/km). l9l1. On site l.b. thinningfrom rocksalong a path escortingthe creek. 20l1.ln the creekwave area in the neighbourhoodof the ,,RumenArs" forester'slodge. 1.c.2,5km downstreamof Lb., in a primordial standof Alnetum incanae-Petasitetum albae (Rumex obtusifolius, Caltha pelta, Salix eleagnos, Salix caprea, about ll00 m a.s.l.,metamorphic rocks, biotitic paragneis, 1500 mm, 5 m/km).

Muntele Mare: SomegulRece

2. Bldjoai4 04. 06. 1992.Due to deforestationand heavy grazingby sheep,the soil is substantiallyacidified, on which peatbogs and swamps of SphagnetaliaPawlowski, 1928, andNardetalia Passarage, 1949 plantcommunities developed. Especially the Polytrichum iuncetum plant community forms large stands (1350 m a.s.l., granite, 1200 mm). Accordingto measurementson thesite, the soil pH was 5,5on the bankof the sffeam.2ll2: Near the water reservoirat the confluenceof the two Someg/Szamosbranches upstream Gil6u. 05. 06. 1992.On the bank in a Quercopetreae Carpineturz stand.

281 SomegulMic. 3. UpstreamCluj, 07.06. 1992.Salicetum albae fragilis Isser1926 (Galium aparine, Urtica dioica, Rubussp., someAgrostis stolonifera,Aegopodium podagraria; disturbed site,350 m a.s.l.,sand, silt, Eocenemarl and sandstone,700 mm, 3m/km). 4. DownstreamCluj, 07. 06. 1992. Salicetumalbae fragilis lsser 1926 (Galium aparine, Rubussp., Urtica dioica, someAegopodium podagraria; disturbedsite, 330 m a.s.l.,sand on Sarmatianmarl, silt, 700 mm, 1,7m/km). 5. DownstreamGherla, 2215 upstream the village Nima, 08. 06. 1992.Thinning in tall vegetationalong the stream.

Rodna Mountains: SomegulMare River

6.a. 500 m downstreamof the spring, upstreamValea Mare, 08. 06. 1992, initial phaseof Petasitetumkablikiana Paul et. Wales(1936) 1946 (Coldea 1990),900 m a.s.l., gneiss,1400 mm, 5 m/km. 6.b. Downstreamof ValeaMare, beforeArieg. 09. 06. 1992,Petasitetum kablikiana Paul et. Wales (1936) 1946 (Coldea, 1990) (Athyrium filix-femina, Rumex obtusus, Abieto-Fagetumzone,700 m a.s.l.,gneiss, 1400 mm, 5 m/km). 7. DownstreamSdngeorgi Bii,09.06. 1992.Primordial stand of Telekio-Alnetum incanae Coldea, 1990, with Aegopodiumpodagraria, Mentha aquatica, Athyrium filix- femina. In a narrow strip along the bank of the streamclose to a plough-land,480 m a.s.l.,sandy silt, 1000mm,4,5 m/km.

Someg/Szamos

9. DownstreamBeclean, 10. 06. 1992.Salicetum olbae-fragflri Isser 1926. Single row of trees on a steepbank(Galium aparine, Rubussp., Scrophularia nodosa),near a cornfield,250 m a.s.l.,clay on volcanicbedrock, 800 mm, 2 mlkm.2319Downstream Beclean,10. 06. 1992. 10. Cdgeiu,10. 06. 1992.Salicetum albae-fragilri Isser,1926, consotiation Populus nigra / Urtica dioica, mass occurrence of the adventive Helianthetum decapitatus. Singlerow of treesamong agricultural fields,225 m a.s.l.,marl and clay on sandstone bedrock,700mm, 1,5m/km. 12.S6lsig.24l12 Benesat, l l. 06. 1992.Upstream of Silsig at a railwayviaduct, 70 m away from the river on the edge of a Quercopetreae-Carpinetum Soo 1957 stand. Thinning,164 m a.s.l.,gravely soil, 700 mm, I m/km. 13. Pomi, ll. 06. 1992.Quercetum petreae-Carpinetum 5o6 1957 standextending down to the riverbank.Hilly regionin the environsof Gutin Mountains.A small creek reachesthe Someg/Szamosin the neighbourhoodof the samplingsite at Pomi (Lamium galeobdolon, Chrysantemumcorymbosum, Echinocystis echinata). 142 m a.s.l., sandy clay on andesitebedrock, 800 mm, I m/km. 25112.A foreststand owned by the village Agrigude Josin the environsof Fersig-Miregu-Mare.Canals go throughand around the oak forest.Collection by thinning. 14.Upstream Satu Mare, 12.06. 1992.Salicetum albae-fragilrs Isser 1926. On a high bankof 50-100m width, some30 m awayfrom the river in a muchthinned vegetation

282 (Rubussp., Aristolochia clematitis,Agrostis stolonifera). ll5 m a.s.l.,alluvial sandand clay,700 mm, I m/km. 15. DownstreamSatu Mare, 12.06. 1992.Salicetum triandrae Mtiller-Gors 1958on aflat river bank (Phragmites australis, Lycopusexaltatus, Agrostis stolonifera). ll5 m a.s.l.,alluvial sand, 700 mm, I m/km. 16.The confluencewith river Tisza,at Viis6rosnamdny.Salicetum triandrae MUller- Gdrs 1958on a steepbank (Rzbus sp., Calystegia sepium, 100 m a.s.l.,alluvial sand,700 mm, I m/km). 17. S6rk6nykertat the Szamos/Some$confluence with river Tisza, 30. 08. 1966. Salicetumalbae-fragills Isser 1926.Rubus sp., Echinocystislobata, alluvial sand. 18. V6s6rosnameny,steep bank of the Tisza, 31.07. 1967. Salicetum triandrae Mtiller-Gors1958, alluvial sand. The geographicallocation of eachsampling site is shownin Figurel. HUNGARY

VasarcsamOny 16 '-''' i

CLUJ-NAPOCA ROMANIA

2

Figure l. Map of the geographicaldistribution of samplingsites along the Someg/Szamosriver

Environmentalstatus of sampling sites

Undergiven climatic conditions the (snails inclusively) will developa homogenouszoocoenotic character within a geographicregion with freemigration. The forests(Bdba 1992 a) andrunning waters serve as routes for faunadispersal, provided thatthese ecological corridors are undisturbed (Erd6si, 1935; B6ba, 1978,1979; B6ba et

283 al. 1982-83;B6ba, 1992a). Examiningthe entirelength of the river Someg/Szamosin this respect,we found everywhereevidences of inhibition of fauna migration and biodiversity developmentby human activities.The fauna movementsgenerally start from the directionof mountains.The humaneffects hindering fauna dispersal occur up to the montanezone of all threemountains considered. These effects are of variouskind. In orderto slow waterrunning people often build wood weirs in streambeds,which lead to the deathof ripariantall communities(at Mun{ii Gildu and Rodna).The Petasitetum stand emergingafter the disturbanceare not taller than 20 cm. The cover of riparian vegetation is disrupted by man-made constructions,supporting roads along the watercourse.The reservoirs(at Mun{ii Gildului) or local water bariers hamper the continuousfauna movementand with high precipitationlead to soil acidificationand degradationover largeareas. The intensivedeforestation and grazing,both retardingthe regenerationof the original vegetation(Muntele Mare) further amplify this latter process.In this mountainonly one specimenof Arion subfuscuswas found after several kilometresof thoroughinvestigation. At lower altitudesthe crop productionand husbandryconfines the natural riparian vegetationto a narrow band (samplingsites 9 and l0). The plastic and metal wastesthrown awayby the local populationalong the river are carriedover severaltens of kilometreswhen the water level is high (site 7). A municipal dumping ground has beensettled near the samplingsite 14.On the lowland,the river flood plain is completely treeless,thus there are no refugeesfor the fauna,which existedin the previouscentury to controlthe Someg/Szamoswaterway. In contrastwith these impedimentson the lowland (Great Hungarian Plain), the fauna migration is facilitated by numerous smaller watercoursescoming from the neighbouringmountains and hilly regionsand reachingthe Szamos/Someg(sampling sitesl0 and l3).

Results und discusslorns

l. Speciesrecorded and their distribution

The expeditionin the Someg/SzamosValley resultedin a collection of 763 living individualsand 168 shellsbelonging to 58 species(data of sites 17 and 18 are not included).Some species are montaneforest dwellers requiring high moisturecontent in biotopes(ecological species group l). Zoogeographicallythese are (East) Carpathian (9.1), Fagion illyricum-moesiacum(5.2.2.) or xeromontane(2.1.) elements. Their distributionto lowland till the river could not be proved.This is not probablefor the xeromontanerock dwellers(Table ; species4,5,l4), the Boreo-montanespecies (10.2.; Ena montana) and the Carpathianand Balkan montane species(Argna, Semilimax, Vitrea transylvanica, V diaphana, Carpathica, Cochlodina marisi, Balea fallax, B. stabilis, Vestia,Bulgarica, Trichia bielzi). The largewater surfaceof the reservoirat Gildului markedlyreduces the temperature fluctuations in its environs. This altered mesoclimatepromotes settling of the

284 Carpathicaspecies of high moisturerequirements even at the foothills. The occurrence of Ruthenicafilograna and Trichia hispida at the edgeof a Quercop. Carpinetum forest stand is due to the close proximity of a hilly region and the fauna transportationby countlessunnamed minor watercourses.These two speciesturn up in the flood plain of the Someg/Szamos,as Querco petreae-Carpinetumstands proceedingdown to the riverbank(site l5). Similar factors promote Cochlodina laminata and Perforatella vicina to reach lowlandareas with annualprecipitation above 600 mm (samplingsites l3 and 17).These two speciesare widespreadon the northern part of the Great Hungarian Plane in Hungary, where large forestedareas were and still are. Hungarian researches(B6ba 1977,1983a, 1986) showedthat the riparianforests (Fraxino-Ulmetum) enable the establishmentof river-carriedmontane snail speciesneed high moisture.There are not such foresttypes in the flood areaon the Romanianpart of the Someg/SzamosValley. Data from sites2219 and 2412show that Perforatella vicina is able to inhabit thickets growing on the riverbank. The absence of the Daco-Podolian Hygromeria transsylvanica and Chilostoma banaticum- both requiringcontinental climatic conditions- is due to the lack of gallery forestsalong the lowland reachesof the Someg/Szamos.This is in contrastwith the paffernfound earlier in the flood areaof the river Mureg(So6s, 1943),where thesetwo speciescoexisted with Arianta arbustorumin extensivegallery forests. Their occulrence in the Hungarianpart of the Muregwas provedby the author'sown collectionin 1972. Cochtodina laminat4 was found to be associatedwith Perforatella vicina at several localitiesnear V6siirosnam6nydownstream the confluenceof the Someg/Szamosand Tisza rivers (B6ba, 1992d). The co-occurrenceof Chilostoma and Hygromia at Benesatin a thicket growing underneatha railway viaduct(site 24112)is supposedlydue to the faunadistribution by waterways from the nearby Jibleg Mountains,whence both snail groups have been reportedearlier (Grossu 1983).Probably the sameis true for Trichiq bielzi. The water reservoirsand the treelessconditions of the flood areasat foothills might hampertheir further dispersal downstream of Gilau Mountains and Rodna Mountains. The appearance of Chilostomain the high flood areaon site3 may be the remnantof a former population. In three snail speciesan unhindereddistribution was observedin the riparian zone

observedearlier (So6s, 1943).Both specieswere also found at V6s6rosnam6nyeven in sitesnot listed in TableNumerous individuals of the amphibian(continental) Succinea putris were observed downstream Beclean on rocks emerging from the water of Someg/Szamos. The third freely distributedspecies is the ubiquitous Deroceras rodnae with an Alpino-Carpathianarea. Due to its high moisture requirementsthis speciesfollows rivers closelv.

285 2. Study of speciesdistribution by zoogeographicalmethods

By the appliedzoogeographical methods it was possibleto follow geographicallythe distribution of fauna group elements.In Figure 2. the distribution of the four most important fauna groups distinguishingmontane and lowland areas are shown. The CentralAsian xeromontane,petrophilous elements (2.1) occur only at the upperreaches of the Someg/Szamos.Having high moisturerequirements, the Boreo-alpine(10.1) and Boreo-montane(10.2) species occur in the montanezone or occasionallyat lower places in humid gorgesand river valleys.This is also true for Arianta arbustorum,which is similarly spreadin galleryforests of the DanubeValley. The Europeanmontane elements (9.1, 9.2, 9.3, 9.4,9.5), some continental(Hygromia, Chilostoma) and hydrophilous species of wide ecological tolerance range can also populate river valleys (e.g. Perforatellabidentata).The members of the Siberian-Asianfauna groups (1.1, I .2,1.j, 1.4)typically inhabitareas of lower altitude,although exceptionally they may occur in montanealder forestsor in forestsand grasslandsunderwent to human disturbances, where their proportion increases(Bilba, 1992 c, e). They can also be consideredas representativesof lowland fauna in contrastwith the rest of the groups in Figure 2. having a montanecharacter.

I1 z9 / 210.1Jl0.2 2'1r / 2.1

/.ti".,.r* 'r'-.. n N \.-.- .,r'''t't.rr 'l 1o 1b 1c 6o 6b 1l+ 1 17 18

Figure2. Percentageproportion of Siberian-Asian(l: 1.1, 1.2, 1.3, 1.4),Central Europeanmontane (9: 9.1,9.2,9.3,9.4,9.5), xeromontane (2.1), and Boreo-alpine / Boreo-montane(10: 10.1,10.2) fauna groups of eachsampling site

286 The faunagroups 2.1, 10/10.I and l0.2 do not extendbeyond the altitudeof 400-600 m a.s.l. (Figure 2., Pyramidula, sphyradium, Arianta, Ena: Table l.). Among the European montane group, the hydrophilousCarpathian species of narrow tolerance range (Cochlodina marisi, Balea stabilis, Vestia elata, Argna, Oxychilus orientalis, Trichiabielzi) are similarly confinedto the montanelocalities (the occurrenceof Trichia bielzi at Benesatmay resultsfrom immigrationfrom neighbouringmountains). The Carpatho-Sudetian(9.2) Vestiagulo and the East Baltic (9.3) Ruthenica filograna found in samplingsite l3 is probablydue to the influenceof a local oak forest and creek.The samecan be true for the snail assemblageon site 24112. The Alpo-Carpathian ubiquitous Deroceras rodnae (sites 15 and 16) and the Carpatho-Sudetian(9.2) Perforatella vicina (sites 2215,25, 12 and l7) speciesmay extendtheir areamuch further.The high proportionof Siberian-Asianfauna groups on siteslb, Lc,4,9, and l0 indicatehuman disturbances, while on sites14, 16 and l8 it reflectsthe continentalcharacter of snail assemblagesin primordialSalicetum triandrae stands(Bilba, 1980 - 8l). Explantationof Table: Samplingsites (sampled by the quadratemethod) (lb, lc), numberof individuals, speciesdiversity, abundance and diversity data,and abbreviationsof zoogeographical(l) andecological (ll) speciesgroups, and of Lozek'shabitattypes (lll). Sitesl7 and l8 are locatedin the Hungariansection of the rivers Someg/Szamosand Tisza, and serveas conffolsof collectionduring this expedition.Sites l9-2511-12 are additional localities for snail collection,which were pointedout in the vicinity of major sites(19-25) and were sampledby thinning. The crosssign indicatesspecies represented by shells on a givensite.

Environmentol stutus of sampling sites

l. Zoogeoeraphicalapproach

The collective values of continental and subatlanticfauna groups - based on percentageabundance data - indicate the predominanceof continental climate in lowlands,and subatlanticclimate in the mountains(Bilba, 1982,1983, 1992 c, Bilbaet al. 1982-83). The ratio of continental fauna groups increasessignificantly as altitude and precipitation(Tufescu, 1965) simultaneously decrease (Figure 3.). The samplingsites at Gil6u Mountains(la), RodnaMountains (6a, 6b) and thoseof hilly regionscovered by typical forests( I 3) are dominatedby subatlanticfauna elements. Lacking of continental patternsreflects human disturbances (sites lb, 1c,3,4 and l0).

287

herghtobove seo Ievel Precipitotion

1b1 15 16 Figure3.

2. Ecologicalapproach; the characteristicsof snail assemblages

The structural characteristicsof snail assemblagesmay also reflect their environmentalstatus. The variationin diversityand speciesnumber of the assemblages studiedare shown in Figure4. Thesevalues depend on many factors,such as the position of the assemblagein a successionalseries. The number of speciesand diversity are generally low in an initial successionalstage (Petasitetum, Salicetum triandrae). These parameterswere indeedlow on siteslb, lc (bothare at Gil6u Mountains),4and 9. The abundanceand mortality percentage is comparedin Figure5. In contrastwith the very high abundancevalue of an undisturbedwillow-poplar forest on site 17 (the confluenceof the Someg/Szamosand Tiszarivers), rather low valueswere found in the sameforest type on sites3, 4, 9, l0 and 15.This is most probablybecause of a gradual shrink of treesor bushesof the woodedhabitat to a sing

I H 4 drversrtot Number of specres

-\ t- \ ..f' \ t'/ :.\ t 2 \, -'_t

46s6b 10 17

289 ttt --\' - - 1b lc 3 4 6o 6b 1 9 t0 13 11 15 16 1't tB Figure5. Variationin abundance,mortality percentage and species number among sampling sites

The high proportionof deadindividuals on sites lc, 3, 4 and l5 indicatean increased human impact.Indeed it could be seenduring our expeditionupstream and downstream Cluj, and in a heavily thinnedforest in the vicinity of a dumping ground upstreamSatu Mare. The distributionfrequency in Lozek's habitattypes showsthe following trend (Fig. 6). The riparian speciespredominate mostly on the lowland sites.The steppedwellers (Phenicolimax,Pyramidula) aredominant components on rocks nearthe spring (site la). The occurence of steppe dwellers (Vallonia, Monacha, Helix lutescens) in willow- poplar forestson sites3 and 4 indicateshuman impact.

.h 100 RU BW S w

S 1 .;.. \j \ I \.j.\ \: \'. I \ -.'\r / w,' \. --- \./

1o 1b 1c l 910 13 14 1 16 11

Figure 6. Percentageproportions of Lozek's habitattypes at the samplingsites W: forestdwellers, BW: bushforest dwellers, RU: riparianhydrophilous elements, S: steppedwellers.

290 The forest dweller speciesreach a dominant status in the montanezone (Rodna Mountains),in undisturbedinitial communitiesand in oak-hornbeamforests (site l3). However, in disturbedprimordial Petasitetumvegetation (lb and lc) and in willow- poplar forests(9 and l5) bushforest dweller speciesare the most abundant. In additionto riparianhydrophilous species, bush dwellers (and occasionally steppe dwellers)appear in initial Salicetumtriandrae habitats (sites 14, 16 and l8). The water conditionsof samplingsites are comparedto differencesin the moisture requirementsof snail assemblagesand the amountof annualprecipitation on eachsite in Figure7. ln accordancewith Lozek'shabitat types, subhydrophilous components (l-2) follow a trend similarto that of the riparianubiquitous (RU) and forestdwellers (W). Xeromesophilouselements (4) reachhigh abundanceon samplingsites lb, 1c and 10, while moderatelyhigh on site4. The ubiquitouselements (5) tend to increase,but still remainin relativelylow valuesin siteslb and lc, whereasoutstandingly high onesin sites4 and 9.

nm Precipt to tron 500r

500 "h 100

Figure7. Percentageproportions of ecologicalspecies groups at thesampling sites l: hydrophilousshade species, 2:subhydrophilous species, 3:rlparian ubiquitous, 4' xero-mesophilouselements, 5:ubrquitous species

The characterisationrevealed here indicates different aspects of structuralchanges in snail assemblages.A slight human influence is reflected in diversity and abundance figures in the low ratio of forest dweller speciesand in the ecologicalspecies group compositionof the initial Petasiteturnstands in siteslb and lc. In sites4,9, l0 and l5 the numberof species,diversity, low abundanceand high mortality values, increasing ratio of bush forest dwellers, steppe dwellers, xeromesophilousand ubiquitous species indicate various degree of humanimpact. These characteristicsare more homogenouson sampling sites with successionallyinitial

29r Salicetumtriandrae vegetation,than thosewith more advancedwillow-poplar forests. The backgroundof this phenomenonmust be a habitatreduction and harmful effectsof agriculturalpractices on adjacentfields. We could not detect any direct influence of water pollution on the terrestrialsnail assemblagesof riparianvegetation.

3. Relationshipsamong snail assemblages

Four clustercores emerge in the dendrogramin Figure 8. The first two includethe rock dwellers collectedby thinning in two mountains(Gilau and Rodna), and in Petasitetutn communities (Petasitetum albae, P. kabikliand. The next cluster core contains snail assemblagesoccurring in Salicetum albae-fragilis standsthat underwent to human influenceon sites3, 4,9 and 10. The last core containsSalicetum triandrae, Salicetum albae-fragilis and Telekio-Alnetum stands. The Querco petreae-Carpinetum plant associationdiffers from thesefour clustercores.

60 6b 1o 1b lc 3 9 4 10 ? 15 14 18 16 1't1J Figure8..Dendrograrn of the Sokal-Michener group average cluster analysis with the indication of samplrngsrtes

The groups obtainedby PrincipalCoordinates Analysis (Figure 9.) agreewith the clustercores detailed above. The snailassemblages in Petasitetumplant communities of the two mountainsdiffer from those in Telekio-Alnetum(7) and Querco-Carpinetum(13) plant communities.Also, the snail assemblagesof disturbedSalicetum albae-fragilis standson sites3,4,9 and l0 separatefrom thosein seminaturalSalicetum albae-fragilis forestsoccurring on sites14-18.

292 Figure9. Resultsof PrincipalCoordinates Analysis (PRINCOOR). The various plant communities andthe willow-poplar forest underwent human activities differ clearly from each other in the compositionof theirsnail assemblages.

Conclusion

The expedition in the Someg/SzamosRiver Valley resultedthe collection of 763 living and 168shell snail individuals belonging to 58 species.The quadrateand thinning methodswere used during collection.The vegetationin the sampling sites included montane and lowland riparian plant communities of early (Petasitetumalbae, P. kitaibeliana, Telekio-Alnetum and Salicetum triandroe) or advanced (Salicetum albae- fragilis) successionalstages, and Querco p. Carpinetum forests of hilly region often extendingdown to riverbanks(5o6 1964,Coldea 1990). The plant communities differ from each other also in their snail assemblages composition,as it is shownby the resultsof clusterand principalcomponent analyses (Figure7. and 8.).

293 It was concluded,that ubiquitousand continentalspecies (Bradybaena, Deroceras rodnae, Succinea putris, Helix lutescens)dominate among snails transported by the Someg/Szamosover long distances.Species of narrow ecologicaltolerance range can find their way to the riparian vegetationof the Someq/SzamosValley through minor watercoursesarriving from neighbouringhills and mountains.The fauna distribution from the spring region is hinderedby water reservoirs,shrinkage of forest habitatsto single row of treesalong the river and the lack of ripariangallery forestsin the flood area.In theseextremely constrained forest habitatssettling of snail individualstaking placeduring high waterlevels is not probable. We could not detectany direct influenceof water pollution on the terrestrialsnail assemblagesof the riparian vegetation.However, disturbancein the flood area was proved by changes in the zoogeographicaland ecological characteristicsof snail assemblages(Figures 2. - 8.). The biotopesare under the strongesthuman impact resultingfrom forestmanagement (site lb and lc at Gildu mountains),or loggingin the flood plain to convertthese lands into agriculturalfields. The garbagedischarge in the neighbourhoodof the riparianvegetation also causes major disturbances.These human interventionswere seen upstreamand downstreamCluj (sites 3 and 4), downstream Beclean(site 9), Cdgeiu(10) and upstreamSatu Mare (15).

References

B6ba, K. (968): Einige continentaleSchneckenzdnosen im -Tisza Tal. M6ra Ferenc Mfzeum Evkdnyve,Szeged, 269 - 202. B6ba,K. (1969):Zdnologische Untersuchungen der an der Flussbettkanteder Tisza und ihrerNebenfltisse lebenden Schnecken. -Tiscia (Szeged), 107-l19. B6ba,K.(1970): OkologischeBeobachtungen beziiglich der Schneckenartenim Tisza Tal. Die besidlungdes Inundationsraumes.-M6ra Ferenc Mrizeum Evk0nyve, Szeged,93- 100. B6ba,K. (1972):The snailcoenoses of the willow grovesin the Middle Tiszaregion. - Tiscia(Szeged), VIII, l0l - 103. B6ba, K. (1973): Peopling of the innundation area of the Tisza by . Regenerationof the snailpopulations in the areaof river barrage,,Tisza II". -Tiscia (Szeged),VIII, 98 - 99. B6ba,K. (1974):Mollusca communities in theTisza bed in the regionof Szeged.-Tiscia (Szeged),IX, 99 - 104. B6ba,K. (1975):From where the woods of theTisza valley are populated by snailsfrom the GreatHungarian Plain. -Tiscia (Szeged), X, 109- ll0. B6ba, K. (1977): Die kontinentalenSchnecken bestande der Eichen Ulmen-Eschen Auwiildern (Fraxino-pannonicae,Ulmetum pannonicum5o6) in der Ungarischen Tiefebene.-Malakologia, l6 (l), 5l - 57. B6ba,K. (1978):The watercarried Mollusks of our river and the analysisof the fauna of the deposit.-Tiscia (Szeged), XIII, 186.

294 B6ba,K. (1979):Die Succesionder Schneckenzcinosenin den WiilderndesAlfoldund die methodenzum Studiumder Succesion.-Malakologia, l8 (l-2), 203 - 210. Birba,K. (1980- 8l): Investigationinto the succesionof snailassociations in the flood pfain of the river.-Atti IV congressoSMl. Siena6 -9 oktobre 1978.Atti Academia FisiositiciSiena, 177 - 192. B6ba, K., varga, A., wagneq M., zseni, L. (1982 -83): Daten zu den verbreitungder Land-Schneckenim Bi.ikk-Gebirgebeeinfl iissenden biotischen Faktoren. -So6siana, l0 - 11,25-30. B6ba, K. (1982) Eine neue ZoogeographischeGruppierung der Ungarischen Landmolluskenund die Wertungdes Faunabildes. -Malakologia,22 (l-2),441- 454. Bitba,K. (1983 a): Effect of the regionsof the TiszaValleyon the malaco-fauna.-Tiscia (Szeged),XVIII, 97 - 102. B6ba,K. (1983 b): A Szatm6r-Beregisikszirazfdldi csig6i es k6rnyezetUkrelevonhat6 kovetkeztetdsek.-Acta Acad. Paed.Szeged, Ser. Biol. Geogr.,27 - 42. Birba,K. (1986)Uber die Sukzessionder Landschneckenbestandein den verschiedenen Waldassoziationender UngarischenTiefebene. Procc. of the 8th Internat. MalacologicalCongress, Budapest, 1983,7 - 12. B6ba,K.(1991) Die Verbreitungder Landschneckenin UngarischenTeil derAlfold. - So6siana,18,25 - 59. Bilba, K. (1992a):DynamicalZoogeography of Mollusks in the HungarianGreat Plain. Abstractsof the EleventhInternat. -Malac. Congress, Siena, 380 - 382. BAba,K., Podani,J. (1992b): A MultivariateAnalysis of snaildistribution in the BUkk Mountains,Hungary. -Proc. Ninth Int. Malak.Congress,35 - 40. B6ba,K.(1992c): The influenceof sylvicultureon the structureof snailassemblage. - Proc.Ninth Intern.Malac. Congress, 27 - 34. Bdba,K., Domokos,T. (1992d):The occurrenceand ecologyof Chilostomabanaticum Rossmiissler1838 in Hungary.-Abstr. llth Intern.Malak. Congr., Siena,383 -385. B6ba,K. (1992e):Zdnologisch-zoogeographische Untersuchungen der Schneckenz6nosen im ungarischenBtikk-Gebirge. -Proc. Tenth Intern.Malak. Congr.(Tiibingen 1989), 463 - 468. Coldea,G. (1990):Mun{ii Rodna;Studiu geobotanic. -Acad. Romdnd, Bucuregti, I - 813. Erdos,J. (1935):Maros torkolat6nakirvizi ds drtdribog6rvil6ga biol6giai szempontb6l (Elenchusbrevis totius dissertationisenamatur). -Arp6d-Nyomda, Szeged , l-87. Feoli, 8., Orl6czi, L. (1979): Analysis of concentrationand delection of underlying factorsin structuredtable. -Vegetatio,40,49 - 54. Grossu,A.V. (1981):Gastropoda Romaniae, 3, -Ed. Litera,Bucuregti, | -268. Grossu,A.V. ( 198I ): GastropodaRomaniae, 4, -Ed.Litera, Bcuregti, 1-563. Grossu,A.V. (1987):Gastropoda Romaniae,2, -Ed. Litera, Bucuregti, l-443. Kerney,M.P., Cameron, R.A.D., Jungbluth, J.H. (1983):Die LandschenckenNord und Mitteleuropas.-Paul Parey, Hambrug-Berlin, l -384. Lisicky,M. (1991):Mollusca slovenska. -Slovenskij akademia vied Bratislava,l-341. Lozek, V. (1964): Quartarmolluskender Tschechoslowakei.-Tschechoslowakischen Akademieder Wissenschaften.Praha. l-374.

295 Lozek, V. (1965): Entwicklung der Molluskenfaunader Slowakei in der Nacheiszeit.- Informationsberichtder LandwirtschaftlichenHochschule, Nitra, t (l-4), 9-24. Podani,J. (1988):Syn-Tax III. AbstractaBotanica, Budapest, l2 (l),1-183. 5o6, R. (1964) Synopsissystematico-Geobotanica florea vegetationisqueHungariae. T. -Budapest,l-589. Szerafim,G.D., Liharev,l.M. ( 1975)Fauna na Bulgarica(Gastropoda terrestri). -Acad. Sci. Bulgaricae,Sofia, l-426. So6s,L. (19a3):A K6rprit-medenceMollusca faun6ja.-Magyar Tud. Akad., Budapest, t-478. Tufescu,V. (1965):Atlas Geografic.-Bucuregti, l-l10.

Kdroly Bdba Andrei SdrkdnY-Kiss Juhdsz Gyula Teacher Training College University BabeS-Bolyai Department of Biology Department of Ecology and genetics 6 Boldogasszony St. 5 - 7 Clinicilor St- 6701 Szeged 3400 Clui Hungary Romania

296