Malakologický bulletin, December 14, 2006 http://www.mal‐bull.blogspot.com/

USE OF TERRESTRIAL MOLLUSCS FOR BIOINDICATION OF THE IMPACT OF THE GABČÍKOVO HYDRAULIC STRUCTURES (THE DANUBE RIVER, SLOVAKIA)

TOMÁŠ ČEJKA Institute of Zoology of the Slovak Academy of Sciences, Dúbravská cesta 9, SK‐84506 Bratislava, Slovak Republic; e‐mail: [email protected]

The aim of this paper is (1) describe main methods using within the biomonitoring programme of the Gabčíkovo hydraulic structures along the Danube River, SW Slovakia; (2) present main results of the monitored long‐term structural changes in alluvial molluscan fauna over the past ten years; (3) describe changes to surface and ground water dynamics on communities of terrestrial molluscs and (4) compare structure of molluscan faunas before the Danube damming and ten years on.

Bioindication abilities of terrestrial molluscs • With long‐term monitoring of the changes in The advantages of molluscs as a model group the natural environment, the changes in the can be characterized as follows: mollusc community structure reflect rather • Molluscs belong, within the scope of Central long‐term trends than short‐term fluctuations in Europe, to well investigated groups environmental parameters. For this reason, they from the viewpoint of community structure, are an ideal group for biomonitoring or sur‐ distribution and habitat requirements, veillance, • The form well‐defined communities • In river sediments rich in calcium, the mol‐ and have strong ecological relationships to spe‐ lusc conchs easily fossilize and form thanato‐ cific habitats and reliably react (although more coenoses in situ and, in this way, make it possi‐ slowly) to changes in soil moisture and the ab‐ ble to study the development of natural envi‐ sence or presence of floods. In the Slovak part of ronment at sites where they occur from a very the iuxta‐Danubian area, the occurrence of 58 long‐term perspective. species of terrestrial molluscs has been con‐ firmed, including species indicating relatively Calibration of indication properties natural habitats, changes in soil moisture and It is known that , hence also molluscs, different degrees of habitat degradation, can often have relatively different ecological re‐ • Most Central European land snails are small quirements in different zoogeographical units. (several millimetres) and little mobile animals Thus, knowledge of habitat requirements of the with a strong dependence on habitat, unlike iuxta‐Danubian (close, at, beside the Danube) some insect groups and spiders. Small species molluscs cannot be automatically applied for penetrate into the adjacent habitats very slowly, example to conditions of the Borská nížina

1 lowland and not at all to more remote parts of Principles for selection of investigation plots and Europe under the influence, for example, of the problems of data interpretation Atlantic climate. For this reason, results of Within the scope of the monitoring plots set by monitoring from geographically and climati‐ the monitoring coordinator, the structurally cally distant areas are not fully comparable. most homogeneous parts are to be selected, in Prior to a monitoring investigation, a calibrating which specific investigation plots are then es‐ indication investigation must be carried out in tablished. Investigation plots of 50×50 m are each area where the local ecological require‐ suitable. The investigation plots should be situ‐ ments of individual species are not known in ated as far from the stand margin as possible to detail in order to determine (calibrate) them by eliminate the edge effects. It is important to en‐ means of the methods given below. sure that the monitoring plots remain unaf‐ fected by silvicultural interventions (e. g. thin‐ ning or even timber exploitation) during the A brief description of methods and interpreta‐ whole period of monitoring. Otherwise, we tion of results would be monitor the influence of silvicultural, Field and laboratory work but not the influence of changes to hydrological When evaluating changes in the natural envi‐ conditions. ronment by means of molluscan communities, Problems with sampling arise on those plots the quantitative and semi‐quantitative methods that are flooded by surface water several times a are used. The most applied method is the vol‐ year. After a flood most molluscs are washed ume method, with which a standardised vol‐ away from the plot or the local patchy structure ume (mostly 5 litres) of litter and soil upper lay‐ of communities is mixed by water turbulence. ers are taken from several places in a homoge‐ For this reason the samples obtained 5‐8 weeks nous plot, approximately up to a depth of 3‐5 after the flood are hardly interpretable from the cm. After washing, drying and subsequent viewpoint of moisture changes. Another prob‐ sieving, the mollusc conchs are sorted from in‐ lem of data interpretation arises in the habitats dividual fractions, identified and counted. The situated at a considerable distance from the proportion of individual species in a sample of flood zone and which remain for a long time the community determines its character (mutual out of the influence of floods, especially in proportion of hygrophilous, shadow‐preferring, habitats situated on elevated places (terrace hygrotolerant or euryecious species). Another, platforms, fluvial levees), where the ground less frequently used method is the qaudrat water table is deep and does not influence the method, which differs just by collecting litter hydropedological regime in the soil upper lay‐ and soil substrate over a specific surface area ers. In such a habitat it is mostly problematic to (e.g. 50×50 cm). In this case the number of mol‐ interpret changes in molluscan communities as luscs is related to the area unit. a direct consequence of water regime changes.

Analysis and synthesis of data Influence of changes to surface and ground Counting individuals in samples gives us ab‐ water dynamics on communities of terrestrial solute data, which should firstly be transformed molluscs (converted to a scale determined by experts, transformed logarithmically, etc.), to be pre‐ Molluscs, and their communities, reliably react pared for statistical analysis. In addition, to to changes in soil moisture and the absence or usual descriptive quantitative characteristics presence of floods. like population density, relative abundance (a) Habitats frequently flooded by more or less (dominance) and structural characteristics like destructive (streaming) surface floods are char‐ diversity and equitability, data can be subjected acterised by the exclusive presence of polyhy‐ also to different methods of multivariate analy‐ grophilous species and by a naturally low al‐ sis (cluster analysis, ordination), which helps us pha‐diversity of communities (in average 12 visualise the data structure and makes data in‐ species). The poorest in the number of species terpretation easier. Results obtained in individ‐ are molluscan communities in the initial stages ual years can be continuously compared and the of floodplain forest (in average 6 species) [3, 4]. trends estimated. Only those species that are adapted to flood

2 conditions survive there. They are represented Influence of silviculture on molluscs especially by larger species that are able to es‐ From the 48 species of terrestrial molluscs cape from the rising water level by climbing regularly occurring in the iuxta‐Danubian onto trees (Arianta arbustorum, Cepaea hortensis, floodplain forests, as many as 32 species also Succinea putris, Oxyloma elegans, Trochulus his‐ occur in older poplar lignicultures [1, 3]. Ac‐ pidus, Trochulus striolatus danubialis) or that are cording to existing data, the most constant spe‐ able to survive for a time directly in the water or cies inhabiting the stands of poplar cultivars are in different water‐bearing substrates, especially incarnatus, an eurytopic forest spe‐ in deposits of plant material (Zonitoides nitidus, cies, which reaches a dominant to subdominant Pseudotrichia rubiginosa, Carychium minimum, position here, and Fruticicola fruticum having and partly also probably Trochulus striolatus da‐ similar ecological requirements as Monachoides nubialis and Trochulus hispidus). incarnatus, but only managing a subdominant to In recently flooded habitats, which are covered receding position here. Among the constant after an intensive flood by a silt crust, the poly‐ species, the forest mesohygrophilous Aegopinella hygrophilous species Oxyloma elegans, Succinea nitens, forest hygrophilous Arianta arbustorum putris, Zonitoides nitidus and Pseudotrichia rubigi‐ and forest eurytopic Alinda biplicata also should nosa appear as pioneer species. Other strongly be named. Cochlicopa lubrica, Cepaea hortensis hygrophilous species, Carychium minimum, ap‐ and Trochulus striolatus danubialis, having opti‐ pears later, because it is a small species having a mum conditions in moister sites in willow‐ small dispersal power, and the oligotrophic poplar forests, and forest hygrophilous species type of substrate does not satisfy its trophic re‐ Vitrea crystallina, Urticicola umbrosus and Clausi‐ quirements. lia pumila, as well as forest mesohygrophilous (b) In sporadically flooded parts of the flood‐ Cochlodina laminata occur here mostly as sub‐ plain, which are influenced by surface waters dominant species [3]. Richterová [11] found 28 only at high or very high water levels in the species in poplar ligniculture of the within‐dike Danube, but which have a high level of ground zone. The most frequent were forest hygrophi‐ water and a moist to waterlogged substrate, the lous species Arianta arbustorum, Trochulus dominant species are Carychium minimum and striolatus danubialis, Vitrea crystallina, Cepaea Zonitoides nitidus, while Succinea putris, Arianta hortensis and Fruticicola fruticum, mesohy‐ arbustorum, Trochulus striolatus danubialis, Tro‐ grophilous forest species (Aegopinella nitens, chulus hispidus, Deroceras laeve or also Euconulus Monachoides incarnatus) and polyhygrophilous praticola occur as subdominant species. Carychium minimum. As dominate species she (c) Habitats with the natural or semi‐natural gives Aegopinella nitens, Punctum pygmaeum and hydrological regime mentioned above, which Carychium minimum, but the largest portion of see a sudden drop in the ground water level or the community consisted of euryecious species the halting of the influence of floods, are at first (22%) and eurytopic forests species (17%). penetrated by moderately hygrophilous species The above enumeration shows that the forest (Semilimax semilimax, Cochlicopa lubrica, Suc‐ species of the Illyrian regiotype (sensu Lisický cinella oblonga, Vitrea crystallina), while the poly‐ [6]) were able to successfully adapt to condi‐ hygrophilous species start to decline and free tions of economic forest stands and to cyclical up niches for mesohygrophilous or eurytopic anthropogenous disturbances (clear‐cuttings, forest species (Alinda biplicata, Monachoides in‐ overall soil preparation etc.). Even species of the carnatus, Fruticicola fruticum, Aegopinella nitens) regiotype of forest permanent dwellers do not or even for euryecious or anthropotolerant spe‐ represent an exception. The poplar cultivars cies (Punctum pygmaeum, Limax maximus, Helix have their moisture optimum in dry types of pomatia). According to the degree of moisture softwood floodplain forests (especially at sites the new community gradually obtains a struc‐ of the Salici‐Populetum typicum variety with ture corresponding to a specifically drier habitat Swida sanguinea), in transitional floodplain for‐ type, mostly characteristic of a degraded com‐ ests (Fraxino‐Populetum) and in the surround‐ munity with a certain portion of xenocenous ings of Bratislava also in sites of moister hard‐ and anthropotolerant species (Limax maximus, wood floodplain forests (Ulmo‐Fraxineta). Arion fasciatus, Arion lusitanicus). Hence, individual habitat types have a corre‐

3 sponding structure of molluscan communities. in the within‐dike zone is just a reduced relict of We do not exclude the potential occurrence of them. Cochlicopa nitens, Succinella oblonga and Euconu‐ lus praticola in moist habitats. They are present In the Danube calciferous sediments, mollusc in this territory, but occur rarely. As to Colu‐ conchs preserve well and often form complete mella edentula, to date it seems that it escapes thanatocoenoses (death assemblages). Accord‐ medium‐moist habitats (transitional floodplain ing to thanatocoenoses, which can be dated, in forests and drier varieties of willow‐poplar addition to radiocarbon methods, also on the floodplain forests) and very moist habitats (it basis of old maps, the habitat conditions of spe‐ probably cannot tolerate floods) [3]. cific localities can be relatively accurately recon‐ structed. Mostly a complete successive series of In young stands of poplar cultivars, in depend‐ the habitat can be reconstructed. ence on the habitat, usually Trochulus hispidus (in moister habitats) Aegopinella nitens, Mona‐ According to existing findings of Holocene’s choides incarnatus and Urticicola umbrosus (in mollusc fauna, it is probable that a border sepa‐ drier habitats) predominate. Remarkable is also rating the wide zone of flooded floodplain for‐ the occurrence of other forest species, for exam‐ ests from the elevated, drier and partly steppe‐ ple Cochlodina laminata and Trochulus striolatus like interior of the Žitný Ostrov Island run ap‐ danubialis, whose occurrence in this type of proximately along the area of the present village stands was not originally expected. Insufficient of Vrakúň. Hence in the past (in younger Holo‐ shadowing by the tree canopy in the case of cene), the floodplain on the present Slovak ter‐ young stands is substituted by high and dense ritory had a width of about 10 km (today in the ruderal vegetation and, if stumps remained at widest part at the village of Baka it is only 3 km) the place of clear‐cut plots, also “sprout en‐ [8]. The question of the steppe‐like enclaves claves”, in which even some forest hygrophi‐ within the floodplain is dealt with in [8, 9]. As lous species are able to survive due to a favour‐ yet the dating of individual processes is miss‐ able microclimate. ing, together with knowledge of the topog‐ raphical distribution of individual geobiocoe‐ noses, which can be made more accurate only Long‐term changes in the structure of iuxta‐ by further investigation of recent and fossil Danubian fauna mollusc fauna, prehistoric settlements and by Structure of iuxta‐Danubian mollusc fauna in the searching for data in archive documents from past (evaluated based on thanatocoenoses) this area. A knowledge of the history of development of hydrological and habitat conditions in individ‐ From the perspective of the changes caused by ual plots is necessary foremost in order to pro‐ the Gabčíkovo hydraulic structures, the period duce an objective evaluation of the so‐called before the so‐called Lanfranconi regulation of zero state of the natural environment, but also the Danube for the medium water (1886‐1896) is for evaluation of the subsequent sequence of of particular interest. The mollusc thanatocoe‐ changes caused by the changed hydrological noses in the iuxta‐Danubian area were studied regime. by Pišút & Čejka [10]. According to historic maps, they dated the absolute age of sediments In spite of existing local investigations, the gen‐ and by means of mollusc thanatocoenoses re‐ eral question still remains as to when and how constructed the complete succession series of the present iuxta‐Danubian geobiocoenoses the floodplain forest from its initial stage to the originated? There is a relatively clear answer in present transitional floodplain forest (period of the case of floodplain forests, especially for the 1791–1999, i.e. 208 years). They confirmed the strip of floodplain forests in the within‐dike occurrence of similar types of floodplain forests zone. The floodplain forests doubtlessly existed as those occurring in the iuxta‐Danubian area at there already in the prehistoric period (younger present and the occurrence of thanatocoenoses, Holocene) [5]. They were, however, probably whose structure approximately corresponds to much more extensive, and so the present strip the structure of the most preserved forest mol‐

4 luscan communities occurring at present in the along the Danube old main channel. The area between the villages Dobrohošť and Sap. changes appeared first of all in the representa‐ The total area of the moistest types of softwood tion of individual subtypes of softwood flood‐ floodplain forests with corresponding malaco‐ plain forests with characteristic structure of coenoses was, of course, much larger in the past molluscan communities. In the by‐passed area than at present and showed a more natural molluscan communities of the moistest types of structure of stands. This is obvious even upon floodplain forest are disappearing, especially comparison of phytocoenological relevés from those bound to the subassociations Salici‐Popu‐ the 1940‐1950‐s with the present ones [12, 13]. letum phragmito‐caricetosum and Salici‐Populetum Unfortunately, there are no quantitative or myosotidetosum. After the damming, most of semi‐quantitative data available on molluscan these types of molluscan communities started to communities in the iuxta‐Danubian area from restructure on “drier” types bound to various the first half of the 20th century. Therefore the degraded types of willow‐poplar floodplain only possibility of comparison is an extensive forests (Salici‐Populetum typicum). paleoecological investigation of recent Danube sediments. In the third stretch, the Danube riverbed bottom deepened, but the water levels were not signifi‐ cantly changed. For the moment this has not re‐ Mollusc fauna structure before the Danube flected on the communities of terrestrial mol‐ damming and ten years on luscs, but in future we cannot exclude adaptive In the beginning it must be noted that before the changes caused secondarily by vegetation Danube damming there were two principal changes resulting from the eventual changing of types of molluscan communities: (1) molluscan the ground water level. communities of the softwood, willow poplar floodplain forests and (2) those of hardwood The fourth stretch at the village of Čičov is out floodplain forests. of the direct influence of the Gabčíkovo hy‐ The iuxta‐Danubian area can be divided from draulic structures. the viewpoint of molluscan communities into the following parts: Expected development of the molluscan com‐ munities structure while saving the present hy‐ (1) the surroundings of the Čunovo reservoir on the drological regime. stretch Bratislava – Dobrohošť, (2) the by‐passed area of the Gabčíkovo hydraulic The existing hydrological regime is relatively structures on the stretch Dobrohošť – Sap, satisfying only from the viewpoint of revitali‐ (3) the stretch downstream from the mouthing of the tail‐race canal (Sap – Klúčovec), zation of ground water levels at the Čunovo (4) the area of Čičov village out of the direct influence reservoir (the floodplain stretch downstream of the Gabčíkovo hydraulic structures. from Bratislava). However, with some excep‐ tions, this area is not regularly flooded. Upon In the first stretch, the level of ground water in‐ high discharges in the Danube, some parts of creased after the Čunovo reservoir was filled the area in the stretch between the Prístavný and the hydropedological regime became re‐ Most Bridge and Čunovo are flooded (for ex‐ naturalized. The habitats in this area were, ample at the Rusovské Ostrovy Islands). Areas however, void of any larger influence of the situated behind the flood protective dikes are groundwater level and relied exclusively on not flooded. Accounting for the relatively high rainwater. The original hygrophilous commu‐ level of ground water, the conditions for terres‐ nities disappeared and the habitats in the course trial mollusc are potentially satisfying. How‐ of revitalization are lacking in immigration ever, the polyhygrophilous species have almost sources of autochthonous mollusc populations disappeared due to insufficient moisture and due to the absence of the possibility of passive the absence of floods. These species are already (drifting by floods) or active spreading [2, 3]. unable to colonize these habitats either pas‐ In the second stretch, after the damming, the sively or actively. Some chance for the revitali‐ groundwater level declined in the zone running sation of molluscan communities might be the

5 relocation of polyhygrophilous species from ar‐ [4] ČEJKA T. 2003: Expertízne vyjadrenie k optimal‐ eas situated more downstream (Kráľovská lúka, izácii vodného režimu v inundácii. In: Lisický, M.J. & Erčéd). Mucha I. (eds) Optimalizácia vodného režimu ramen‐ nej sústavy v úseku Dunaja Dobrohošť – Sap z A worse situation exists on the stretch Dobro‐ hľadiska prírodného prostredia. Prírodovedecká fa‐ hošť – Sap, where the absence of natural floods kulta Univerzity Komenského v Bratislave, 205 pp. and results in negative structural changes in the (Monografia). molluscan communities. The floods can be [5] CHEBEN I., HAJNALOVÁ E. & ČEJKA T., 2001: Wald‐ partly simulated by leading water into the arm vegetation des Unterlaufes der Žitava im Neolithi‐ system from the intake structure at Dobrohošť, kum. Pravěk, suppl. 8: 40‐52. but there is not a sufficient will to carry out [6] LISICKÝ M. J., 1991: Slovenska, Veda, such flooding because of conflicts of interests. Bratislava, 344 pp. The disadvantages of simulated floods are dealt [7] LISICKÝ M. J. & MUCHA I. 2003: Optimalizácia vod‐ with in the monograph [7], which presents ideas ného režimu ramennej sústavy v úseku Dunaja Do‐ brohošť – Sap z hľadiska prírodného prostredia. about the revitalisation of the area. The present Prírodovedecká fakulta UK, Bratislava, 205 pp. floodless situation is bad and the succession of (Monografia). most of the originally strongly moist habitats [8] LOŽEK V. 1955: Zpráva o malakozoologickém will stop in the stage of dry types of willows výzkumu Velkého Žitného ostrova v roce 1953. Práce poplar floodplain forests, or even drier succes‐ II. sekcie SAV, séria biol., zv. I., zošit 6, 31 pp. sion stages, which will concurrently influence [9] LOŽEK V. 1964: Quartärmollusken der Tschecho‐ the succession of respective molluscan commu‐ slowakei. Rozpr. Ústř. Úst. Geol., Vol. 31, 374 pp. nities. [10] PIŠÚT P. & ČEJKA T. 2002: Historical development of floodplain site using Mollusca and cartographic evidence. Ekológia, Bratislava, 21/4: 378‐396. References [11] RICHTEROVÁ P. 2004: Malakocenózy topoľových [1] ČEJKA T. 1999: The terrestrial molluscan fauna of lignikultúr vybraných lokalít v derivačnom území the Danubian floodplain (Slovakia). Biologia, Brati‐ vodného diela Gabčíkovo. Diplomová práca, Ms. de‐ slava, 54: 489‐500. pon. in: PriFUK, Bratislava, 66 pp. [2] ČEJKA T. & FALŤAN V. 2001: Hodnotenie stano‐ [12] UHERČÍKOVÁ E. 1995: Fytocenologické a eko‐ vištných pomerov podunajských lužných lesov pri logické pomery lesov inundácie Dunaja. ‐ Kand. diz. Bratislave na základe štruktúry fytocenóz a malako‐ práca, Ms. depon. in PríFUK a ÚZ SAV, Bratislava, cenóz (prípadová štúdia). Sborník Přír. klubu v Uher‐ 221 pp. ském Hradišti 6: 38–52. [13] UHERČÍKOVÁ E. 1998: Transformation changes [3] ČEJKA T. 2003: Ekologické väzby ulitníkov (Gas‐ in the association Salici‐Populetum in the inundation tropoda) v podunajských lužných lesoch. Kand. diz. area of the Danube river (Slovakia). Biologia (Brati‐ práca, Ms. depon in: PriFUK, Bratislava, 97 pp. slava), 53: 53‐63.

6