RMZ - Materials and Geoenvironment, Vol. 52, No. 1, 45-49, 2005 45

Hydrological changes of the Mura River in , accompanied with habitat deterioration in riverine space

LIDIJA GLOBEVNIK1, MITJA KALIGARIÈ2

1University of Ljubljana, Faculty of Civil Engineering and Geodesy, Hydrotechnical Department, Jamova 2, Ljubljana, Slovenia; [email protected] 2University of Maribor, Faculty for Education, Department for Biology; Koroška 160, Maribor, Slovenia; [email protected]

INTRODUCTION sity of wetland habitats and species in the ri- parian forest of the inner Mura. We evaluate The Mura River (380 km), the tributary of hydrological changes of the section (1); give the River (, ) is a the draft historical reconstruction of the river transboundary river, flowing from to flow and riparian space (2); show the habitat Slovenia and . Historical development types and their degradation rate on a small of hydro-electric power plants in Austria, pilot area (3). channel straightening and other engineering of the channel for flood protection have al- tered the hydrological and sediment transport Description of the area regimes in the Slovenian section of the river. The channel bed of the upper part of the Mura The Mura River is the largest tributary to the section in Slovenia has deepened and this has Drava River, the largest Danube tributary. It affected low to medium flow levels. Enlarge- springs 1898 m above see level in the east- ment of the main channel and reduction in ern edge of the Lower Tauren Alps in Aus- connectivity with the floodplain area between tria and inflows to the Drava River on the flood protection dykes were additional Croatian – Hungarian border. It catchment causes for reduced flood frequency and is 14304 km2 large. More then half of the groundwater level lowering. However, the catchment belongs to Austria, while the main channel in Slovenia flows within lower part of the Mura lies in Slovenia a floodplain up to 1 km wide bounded by flood (1393 km2), Croatia (987 km2) and Hungary protection banks. This flood plain comprises (1911 km2). The length of the Mura is many old meanders, oxbow lakes and flood 444 km. The hydrological regime depends forest areas. Reduced flood frequency and on snow melting season in Austria that causes groundwater level lowering is indicated in a usual higher water flow in spring (March – reduction in natural tree species in flood for- May). In the winter period (October – March) est areas and an increase in the number of the low water regime prevails. The average alien species. The above problems coupled water discharge in the lower section is with the intensive drainage of wetland areas 170 m3/s (Cmurek, Slovenia). The highest for agriculture that has been carried out in the water peak (1357 m3/s, Cmurek, Slovenia) past has caused the general decline in diver- can be more than 30 times larger than aver-

Short scientific paper 46 GLOBEVNIK, L., KALIGARIÈ, M. age minimal flow (Halcrow Water and Wa- tial periods (25 year period from 1926-2000) ter Management Institute, 2000). While large plotted against the average annual precipita- parts of the upper river section in Austria tion values for Murska Sobota and Gornja (315 km) were channelised for hydropower Radgona to find any significant trend for production, the middle and the lower sec- change in the relationship between rainfall tions (130 km) starting upstream of the Aus- and river flows. On the Mura River main- trian border at Šentilj (Spieldfeld) are still stream within Slovenia there are two gaug- largely intact. The river banks and bed of the ing sites with virtually continuous records river section along the Slovenian-Austrian of mean daily discharge from 1926. The border (33.28 km) is completely regulated Petanjci gauge with available record period with a 60 to 80 m wide river bed. Flood pro- 1926 - 2000, is situated within the border tection dikes (embankments) were con- section and upstream of that part of the Mura structed along the flood plain of the river wetland complex that lies within the coun- section in Slovenia (the Inner Mura, 27 km) try. The Mursko Središèe gauge with avail- in the period 1972-1990 with a level of ser- able record 1926 – 1990, is sited at the down- vice against flooding of 1:100 years. The stream boundary of the system. The only Mura River has a variety of typical plant and other river of any significance that could be animal communities, ranging from pioneer considered as part of the wetland complex is to mature stages, including pannonian- the Šèavnica, which enters the Mura between dinaric and pontic-caspian elements, with the two gauging sites but which is ungauged. large floodplain forests and side arm systems. Apart from natural landscapes, there is a For the purposes of consistency the contem- highly value cultural landscape. poraneous records (from 1926 to 1990) at each site are emphasised and the later data at Petanjci considered independently. Linear METHODS trends have been analysed on the discharge data (annual, frequency distribution) to find Hydrology: Existing national monitoring any long term hydrological changes. data (HMZ, 1995a; HMZ 1995b; HMZ 1995- 2004a; HMZ 1995-2004b; ARSO 2000) have The contribution of the wetland system to been used to investigate whether there have the overall hydrology is referred to as the been any long-term changes in rainfall or ‘residual’ hydrology that is the difference river flow which may have had adverse im- between the inflows (at Petanjci) and the pacts on the River Mura wetlands. First the outflows (at Mursko Središèe). The storage average values for annual precipitation for or attenuation effect of the residual area (wet- rain gauge stations at Gornja Radgona, lands in Slovenia) is shown by the flow du- Murska Sobota and Lendava for the period ration curve (FDC) of the daily flows from 1961 to 2000 are defined and linear trend the residual area and was obtained by sub- estimated. Then the average discharge val- tracting the quantiles of the FDC at Mursko ues at 95 %, 90 %, 80 %, 50 % and 1 % Središèe (downstream) from that at Petanjci frequency for Petanjci station for six sequen- (upstream).

RMZ-M&G 2005, 52 Hydrological changes of the Mura River in Slovenia, accompanied with habitat... 47

Based upon what data appears to be avail- program. The present-day land-cover were able from hydrological studies from 1981 and taken from the data base of agriculture land 1990 and field inspection in 2001 the avail- use, a GIS-supported database, owned by ability of transportable material is assessed Ministry of agriculture, forestry and food of and historic changes in river cross-sections the Republic of Slovenia (Anon., 2002). It is and bed levels defined. The linear trends of based on digital aerial photographs. The cat- groundwater level changes of the Mura river egories were aggregated and simplified to aquifer are studied for 18 locations for the make data comparable with the older maps. period 1961-2000 and 1981-2000. Param- eters investigated are average levels, maxi- Habitat evaluation: Habitats were mapped mum levels, minimum levels, average depth on a pilot area Veržej – Mota (LESKOVAR and and amplitudes. JAKOPIÈ, 2000) and later adapted to the “PHYSIS” typology, based on Palearctic clas- Draft vegetation history: The vegetation sification (DEVILLERS ET AL., 1993; DEVILLIERS cover of the pilot area of 8.2 km2 from 18th and DEVILLIERS-TERSCHUREN, 1996), which Century were interpreted from Austro-Hun- was chosen as one of the most accurate one. garian military maps, made between 1763 Hierarchical classification is based on vegeta- and 1787, which were exclusively used for tion typology. This classification was adopted military purposes and formed a part of “Em- and modified for Slovenian conditions. (JOGAN peror Joseph II Land Survey” (RAJŠP and ET AL., 2004) Due to various transitional stages FICKO, 1996). The maps were drawn at a scale and mixed habitat types also “hybrids” (tran- of ca. 1:28 000. They are not sufficiently ac- sitional forms, mixture or mosaic) between 2 curate to be processed directly by GIS (ÈARNI or 3 habitat types were used. ET AL., 1998), but by positioning the settle- ments were possible to transfer categories, Habitat evaluation was attributed due to spe- clearly visible from the maps into GIS, us- cies-richness, naturalness, species-rarity and ing Arc-View 3.2 computer program. The rate of invasion of allien species. following categories were distinguished: for- est (including scrub and wooded grasslands), grassland (pastures and meadows), fields, RESULTS vineyards and orchards included and settle- ments. The first areal photographs available Hydrology: Linear trend for annual precipi- are 40 years old and the following catego- tation values at three locations, Gornja ries are visible: forest (including scrub and Radgona, Murska Sobota and Lendava for the wooded grasslands), grassland (pastures and period 1966-1990 are estimated to be slightly meadows), fields, vineyards, orchards, settle- positive, whereas for the period 1966- 1995, ments and river channel with larger gravel precipitation values for Ledava have a slight bars. The photographs (accurancy of negative trend. Regarding the average dis- 1: 17500) are not geo-oriented. With geo- charge values at 95 %, 90 %, 80 %, 50 % and referrencing to settlements we transferred 1 % frequency for Petanjci station for six se- categories, clearly visible on the photographs quential periods compared against the aver- into GIS, using AUTODESK MAP computer age annual precipitation values for Murska

RMZ-M&G 2005, 52 48 GLOBEVNIK, L., KALIGARIÈ, M.

Sobota and Gornja Radgona we conclude that the same is true for the river bed, while sur- there is no discernible or significant trend for veys in areas further away from the river change in the relationship between rainfall show that the groundwater tables have de- and river flows, and that any change in hy- creased more. The river bed of the Mura at drology that may exist is apparently statisti- Petanjci lowered for 27 cm during the pe- cally independent of climate. riod 1970-80. Further lowering of the bed profile was apparently controlled with tech- The linear trends of annual average dis- nical measures, the effects of which were charges and minimal discharges for Petanjci shown in lowering of the river bed profile and Mursko Središèe are negative. The same by 16 cm in the following ten years. The is true for 1 %, 50 % and 80 % frequency amount of bed degradation decreases down- distribution values, but positive for 95 % and stream in the Inner Reach. 90% frequencies. The FDC of the flows in the Mura mainstream indicates a continuous Draft vegetation history and habitat evalu- distribution, but the flows from the wetland ation: A decline of both water surfaces and fall into three distinct component parts, rep- riparian forests was percieved. Water surface resenting floods and low flows and a defini- has declined for 9% in 200 years, the big- tive and virtually constant outflow discharge gest change happend in tha last 40years. of 7 cumecs which typifies flows for virtu- Area of riparian forest has declined for 30%. ally 40 % of the year. Today 4 % of the area is covered by running water surface, the main stream of the Mura The availability of transportable material in river. At this part a great proportion is still mid 70’ has been much lower than the river forested (65%), but among forest 10% is transporting capacity, which was estimated degraded by allien Robinia pseudoacacia. to be from 20,000 to 70,000 cubic metres Only 35% of are is completely natural or per year. Typical d50 sizes quoted from the nature-close stage. 1970’s were of the order of 5-10 cm at Mursko Središèe, and estimated transport has been of the order of 30,000 tonnes or 17,000 DISCUSSION AND CONCLUSIONS cubic metres for an average year. Changes in water and sediment flow dynam- The results of the surveys in the period be- ics have been influenced by the construction tween 1971-89 indicate that the average low- of hydropower plants in the upper parts of ering of the river bed at Austria- Slovenia the Mura River in Austria. Coupled with border is 33 cm. The maximum values ex- extrensive water abstrations and controlled tend to 88 cm and were observed below the drainage of water in tributary stream network Cmurek bridge, while in the profile of the have stimulated extensive lowering of the Radenci bridge the elevation of the river bed river bed, groundwater level and increased in the range of 11 cm was recorded. Surveys rate of bed erosion. It is estimated that the of the ground water levels in the wells closer level of river bed in the border section of the to the river indicate the average level of low- Mura River (Slovenia/Austria) has decreased ering of groundwater tables for 30 cm and by of the order of 1.0 m, and decreased river

RMZ-M&G 2005, 52 Hydrological changes of the Mura River in Slovenia, accompanied with habitat... 49 bed levels can be observed well into the In- Dynamics of channel and flooding patterns ner Mura Section. A substantial (and signifi- are, together with soil moisture and human cant) difference is evident for the residual impacts, the main factors influencing ripar- flows which reflect the hydrology of the ian zones. In these areas the direction of flow wetland (residual area between Petanjci and is from the soil water and groundwater into Mursko Središèe) complex itself. Flood out- the channel. Nowadays water from the main flows have almost doubled there in the latter channel of the Mura fills the soil only when period while the median and low flow re- a flood pulse raises the level of the river gime has fallen to less than a half of the dis- above the hydraulic base level. Hence, the charges that typified the pre 1965 hydrology. level of groundwater and water in the chan- This pattern points to a very large reduction nel are mutually dependent. Numerous back- to the natural storage available within the waters and oxbows, once typical for the wetland complex. floodplain areas along the Mura River, now have very weak or no connection at all with The situation is especially visible during the the river waters. periods of low and average water levels.

REFERENCES

ANONYMUS (2002): Kmetijska raba tal Slovenije. - HMZ (Ministry for Environment and Physical Plan- Ministrstvo za kmetijstvo, gozdarstvo in ning - National Hydrometeorological Institute). prehrano RS, Ljubljana. (1995a): Klimatografija Slovenije-Padavine ARSO (Agency for environment of the republic of 1951-1980. Ljubljana. Slovenia) (1999): Hydrological data 1926 – HMZ (Ministry for Environment and Physical Plan- 1990. Archive. ning - National Hydrometeorological Institute). DEVILLERS, P., DEVILLERS-TERSCHUREN, J. (1996): A clas- (1995b): Klimatografija Slovenije - Padavine sification of Palaearctic habitats. - Council of 1961-1990. Ljubljana. Europe. HMZ (Ministry for Environment and Physical Plan- DEVILLERS, P., DEVILLERS-TERSCHUREN, J., VANDER LIN- ning - National Hydrometeorological Institute). DEN, C. (1993): Palaearctic Habitats: a (1995_2004a): Meteorološki letopisi Slovenije DatabaseProvisional draft (PHYSIS Database). 1991-2000. Ljubljana. - Institute Royal des Sciences Naturelles de HMZ (Ministry for Environment and Physical Plan- Belgique. ning - National Hydrometeorological Institute). ÈARNI, A., JARNJAK, M., OSTIR-SEDEJ, K. (1998): Past (1995 -2004b): Hidrološki letopisi Slovenije and present forest vegetation in NE Slovenia 1991-2000. Ljubljana. derived from old maps. Applied Vegetation Sci- JOGAN, N., KALIGARIÈ, M., LESKOVAR, I., SELIŠKAR, A., ence 1, 253-258. DOBRAVEC, J. (2004): Habitatni tipi Slovenije HALCROW WATER AND WATER MANAGEMENT INSTITUTE HTS 2004. Ministrstvo za okolje, prostor in (2000): Transboundary Co-operation in Nature energijo, Ljubljana. 63 Conservation and Wetland Management for the RAJŠP, V., FICKO, M. (1996): Slovenija na vojaškem Danube River Basin. Component 1: Sustainable zemljevidu (Josephinische Landesaufnahme River Basin Development and Wetland Manage- 1763-1787 fuer das Gebiet der Republik ment in the Lower Mura River. Final Report, Slowenien). - ZRC SAZU and Arhiv Republike London. Slovenije, Ljubljana.

RMZ-M&G 2005, 52