LITHUANIAN UNIVERSITY OF AGRICULTURE
SYSTEMATIC APPROACH INWATER QUALITY ASSESSMENT OF LITHUANIAN RIVERS IN THE CONTEXT OF PHYSICAL, CHEMICAL AND HYD-ROBIOLOGICAL PARAMETERS Summary of Doctoral Dissertation Technological Sciences (T 000) Environmental Engineering and Land Management (04T) Kaunas, 2010
Doctoral dissertation was accomplished at the Lithuanian University of Agriculture in 2005-2010. Academic supervisor: Prof. dr. Arvydas POVILAITIS (Lithuanian University of Agriculture, Technological Sciences, Environmental Engineering and Landscape Management 04T) The dissertation will be defended at the Council of Environmental Engineering and Landscape Management of the Lithuanian University of Agriculture: Chairman: Prof. dr. Petras PUNYS(Lithuanian University of Agriculture, Technological Sciences, Environmental Engineering and Landscape Management 04T) Members: Dr. Jrat KRIAUČINIEN (Lithuanian Energy Institute, Technological Sciences, Environmental Engineering and Landscape Management 04T) Assoc. prof. dr. Valdas PAULAUSKAS(Lithuanian University of Agriculture, Te-chnological Sciences, Environmental Engineering and Landscape Management 04T) Assoc. prof. dr. Egidijus ARAUSKIS(Lithuanian University of Agriculture, Techno-logical Sciences, Environmental Engineering and Landscape Management 04T) Assoc. prof. dr. Julius TAMINSKAS(Institute of Geology and Geography at the the Nature Research Centre, Physical Sciences, Geography 06P) Opponents: Prof. habil. dr. Vida STRAVINSKIEN(Vytautas Magnus University, Biomedical Sciences, Ecology and Environmental Research 03B) Dr. Antanas Sigitas ILEIKA University of Agriculture, Technological (Lithuanian Sciences, Environmental Engineering and Landscape Management 04T) The official defense of the dissertation will be held at the public session of the Council of Environmental Engineering and Landscape Management Sciences at 10 a.m. on July 2, 2010 in audience 261 of the Main building of the Lithuanian University of Agriculture, Student11, LT-53361 Akademija, Lithuania. Summary of doctoral dissertation has been sent out on the 1stof June, 2010. The dissertation is available at the library of the Lithuanian University of Agriculture. 2
LIETUVOS EMSKIO UNIVERSITETAS
LIETUVOS UPIVANDENS BKLS SISTEMINIS VERTINIMAS FIZIKINI, CHEMINIIR HIDROBIOLOGINIPARAMETRKONTEKSTE
Daktaro disertacijos santrauka Technologijos mokslai (T 000) Aplinkos ininerija ir kratotvarka (04T) Kaunas, 2010
Disertacija rengta 2005-2010 metais Lietuvos emskio universiteto Vandentvarkos katedroje. Mokslinis vadovas: Prof. dr. Arvydas POVILAITIS em (Lietuvosskio universitetas, technologijos mokslai, aplinkos ininerija ir kratotvarka 04T) Disertacija ginama Lietuvos emskio universiteto Aplinkos ininerijos ir kratotvarkos mokslo krypties taryboje: Pirmininkas: Prof. dr. Petras PUNYS(Lietuvos emskio universitetas, technologijos mok-slai, aplinkos ininerija ir kratotvarka 04T) Nariai: Dr. Jrat KRIAUČINIEN (Lietuvos Energetikos Institutas, technologijos mokslai, aplinkos ininerija ir kratotvarka 04T) Doc. dr. Valdas PAULAUSKAS(Lietuvos emskio universitetas, technologi-jos mokslai, aplinkos ininerija ir kratotvarka 04T) Doc. dr. Egidijus ARAUSKIS(Lietuvos emskio universitetas, technologijos mokslai, aplinkos ininerija ir kratotvarka 04T) Doc. dr. Julius TAMINSKAS(Gamtos tyrim geologijos ir geografijos centro institutas, fiziniai mokslai, geografija - 06P) Oponentai: Prof. habil. dr. Vida STRAVINSKIEN(Vytauto Didiojo universitetas, biomedicinos mokslai, ekologija ir aplinkotyra - 03 B) Dr. Antanas Sigitas ILEIKA(Lietuvos emskio universitetas, technologijos mokslai, aplinkos ininerija ir kratotvarka 04T) Disertacija bus ginama vieame Aplinkos ininerijos ir kratotvarkos mokslo kryp-ties tarybos posdyje 2010 m. liepos m1la.0.002.vdnvuso.iLteemskio uni-versiteto centrinirm261 auditorijoje. Adresas: Lietuvos emskio universitetas, Studentg. 11, LT-53361 Akademija, Kauno r., Lietuva Disertacijos santrauka isista 2010 m. birelio 1 d. Disertacijągalima perirti Lietuvos emskio universiteto bibliotekoje.
Significance of the topicFrom the ancient times river water quality has been the subject causing re-searchers, water suppliers, water managers and ordinary people a great interest con-sidering it as a research object as well as the essential resource to sustain life and economy. Due to the fact that different water uses require different levels of water quality, a variety of water quality standards targeting relevant uses have been elaborated. In addition to these anthropocentric norms, the derivation of water qual-ity criteria meeting the requirements for healthy functioning of water ecosystems has become increasingly common. Biological criteria are especially relevant in this context, as the development of biological water quality standards is one of the main tasks for current water management. Water quality criteria development is not the only theme that is important in water management. After water quality problem is identified the need to find out its causes follows. Since human activities are programmed to influence environment, the search for causes for failures in water status will always constitute a relevant ac-tivity, fostering the development of appropriate tools and methods for this purpose. It is to be emphasized that water quality research often targets only some small subset of parameters of the same type (physical, chemical, biological etc.). This re-sults in only partial description of water ecosystem status without accounting for in-teraction among different types of water quality elements and other interacting processes, which makes it difficult to spot reasons for measured values of the vari-ables. It is therefore very important to analyze different types of data and do it in a systematic way. When interaction and causality comes into focus, it commonly bears a narrow scope. For instance, when relationships between river basin landuse characteristics and concentrations of physico-chemical elements in river water are being investigated, their effects on biology are usually not analyzed. Conversely, when a study searches for how chemical parameters affect biology, landuse im-pacts on the same chemistry are usually out of scope. The narrow scope prevents from holistic understanding of the internal and external factors affecting the state of aquatic systems. Considering the aforementioned problems and needs, this work aims to un-cover external landuse, landcover and internal ecosystem processes and their af-fects on water physico-chemical parameters as well as the effects the latter have on diatoms a biological quality element. As a result, the established links between river chemistry and diatoms are to be used in this work to assess the possibilities to apply diatoms in the assessment of status of the Lithuanian rivers. Relevance of the work In the year 2000 the EU Directive 2000/60/EC (further referred to as WFD) came into force obliging all member states to carefully assess the status of all their water bodies, to identify water status problems, the reasons behind them and to elaborate programs of measures to solve those problems. Implementation of measures is a very costly activity, therefore it is very important to have a correct identification of problems and their causes and thus avoid unnecessary costly ac-
tions. Ideal reasoning for water status problems could be achieved by the use of good accounting data on point source pollution, diffuse pollution inputs and de-tailed process-based water quality models. Unfortunately, in many cases and in Lithuania in particular the aforementioned information is far from complete and the run of detailed models is impossible or not feasible in terms of precision of the results. One of the solutions to help find causes with limited information could be the use of proper statistical methods, applied on water monitoring data. The meth-ods could be employed in conjunction with modeling to create a broader picture on possible problems and the reasons behind them. However, monitoring data contains a huge array of parameters (in this work 27), some of which are very much intercorrelated. Consequently, conventional univariate statistical methods are unfit to analyze their complexity and detect meaningful structures in data, which could shed light on processes affecting water status. Luckily, multivariate statistical methods are perfectly designed to solve such kind of problems. Among them Factor analysis is on of the most popular and most appropriate statistical tools to deal with such tasks. It has almost not been used in Lithuania in the con-text of the analysis of physico-chemical data. Water quality assessment according to biological quality elements is another relevant topic facing water managers these days. The WFD introduced require-ments to elaborate new type of water classification system of 5 classes where bio-logical parameters play a major role. Biological parameters or indicators created for the purposes of water status classification must be sensitive to different types of human impacts, including those resulting in elevated levels in the riverine values of physico-chemical parameters. Phytobenthos, represented by diatoms, is one of bio-logical quality elements that falls under the WFD requirement to elaborate criteria for water status classification. This element is very useful in water quality assess-ments since it is present practically in all rivers regardless of climatic conditions and habitat type. Moreover, diatoms are known for quick response to changes in river water chemistry, therefore it suits well to indicate fresh pollution. Unfortu-nately, diatom criteria for river water quality assessment that would fit Lithuanian conditions are lacking. Thetarget subject of the research The target subject of the research encompasses Lithuanian rivers where the State water quality monitoring is performed as well as factors determining physico-chemical status of the rivers and elaboration of biology-based river water status classification system on the basis ofdata from 107 State monitoring sites. General objective The objective of Dissertation is to identify interaction patterns among landuse, soils and river water physico-chemical as well as biological parameters when as-sessing river water status. Research goals: 1. Using system-oriented approach to assess the impacts different factors have on the state of Lithuanian rivers water and identify their spatial patterns.
2. To assess the suitability of diatom-based water status assessment methods to apply under Lithuanian conditions. The defense statements: 1. Spatial patterns of the state of the Lithuanian rivers water are determined by interacting effects of anthropogenic pollution (agriculture and wastewater from set-tlements), geological-soil structure of river basin surface, vegetation periods, accu-mulation and breakdown of organic matter in the river bed and the aeration proc-esses. 2. widely applied abroad have a reasonable level ofDiatom indices that are sensitivity to changes in Lithuanian river water physico-chemical conditions, there-fore some of them could be potentially used for the assessment of water status of Lithuanian rivers. Scientific novelty and practical value By this study the relationships among the river basin and the buffer strip lan-duse, soils and river water quality physico-chemical parameters in Lithuania have been identified. One of multivariate statistical methods - the spatially based Factor analysis has been tested on Lithuanian river hydrochemical data for the first time. The Factor analysis tool enabled the identification of the main natural and anthro-pogenic processes (factors) determining water quality during each season. As a re-sult, monitoring stations were grouped into clusters each representing a group of stations mostly affected by a relevant factor. The results of established water qual-ity determining factors and the tested multivariate statistical method can be applied in practice when the reasons for water quality impairments are to be investigated or river monitoring network is to be optimized. The results can also be of value for the elaboration of water quality enhancing programs of measures and river basin man-agement plans. There are few studies targeting diatoms in rivers in Lithuania. This study has investigated interactions between groups of physico-chemical parameters and the diatom indices as well as the possibilities to derive criteria for water quality as-sessment using diatoms. Preliminary boundary values separating certain quality classes have been proposed for some diatom indices. By this action the example was set for the methodology that could be employed for further work, which would deal with diatom criteria development after an influx of new data. When the system for diatom-based water status classification will be fully developed, the WFD obli-gation regarding phytobenthos will be fulfilled. METHODOLOGY For the achievement of the 1st goal, the State monitoring physico-chemical data from 107 sites have been used (Table 1, Fig. 1). The measured parameters in-clude flow velocity (V), discharge (Q), suspended matter (SM), transparency (SK), pH, dissolved oxygen (O2), oxygen saturation (O2%), biochemical oxygen demand in 7 days (BOD7), chemical oxygen demand using dichromate (CODCr) or perman-
ganate (CODMn), total organic carbon (TOA), ammonia nitrogen (NH4-N), nitrite nitrogen (NO2-N), nitrate nitrogen (NO3-N), organic nitrogen (Norg), phosphate phosphorus (PO4-P), organic and adsorbed phosphorus (Porg), calcium (Ca), magne-sium (Mg), sodium (Na), potassium (K), silica (Si), bicarbonate (HCO3), sulphate (SO4), chlorine (Cl) and total iron (Fe). In addition, GIS layers of CORINE2000, Lithuanian soils (1:300 000 scale), river basin boundaries and rivers (Lithuanian georeference database and the data of Lithuanian Environmental Protection Agency) have been used. The aforementioned data were subject to different statistical techniques. Pear-son and Spearmans rank correlation were used to find out possible relationships among riverine concentrations of physico-chemical elements at the monitoring sta-tion and the percentage expression of different categories of monitoring station river basin landuse and soil. The analysis was also applied for landuse and soil characteristic in 110 310 m wide buffer strips. Physico-chemical data in those analyses were represented as averages for the whole period of 1999-2004. Afterwards the so-called spatial Factor analysis (FA) was employed, which was based on seasonal averages of physico-chemical data for the whole period of 1999-2004. Winter was represented by the period from September to February months, spring by March to April and summer by May to August. Spatial FA was applied to identify main factors affecting river water quality and spatially group them according to the affecting factor and its strength. The Mann-Whitney U test was used to confirm significant differences in relevant parameters values in factor groups. The so-called local FA was employed for different groups represent-ing monitoring stations, using the dataset of all data in particular selected station, covering the period of 1992-2004. Local FA results were utilized for more detailed description of processes affecting the aforementioned groups of stations. The idea of factor analysis is to reduce intercorrelating variables into a few new representative uncorrelated integrated variables or the so-called principal com-ponents, which take a form of the 1st-order equation. The components then are or-thogonally rotated in space to transform into latent factors that are explained by variables that correlate with them. The FA mathematical model is based on the as-sumption that the behaviour of each variableXiis influenced bymcommon factors (F1, F2, ..., Fm). The interrelationship between factors and variables (X1, X2, ..., Xk) is expressed by the 1st-order equations: m X= λiF (1) , ij∑=1j j m Xk∑λF, (2) =j=1kj j whereλij(i=1, , k; j=1, , m) =cov(Xi,Fj) factor loadings (the higher the loading, the more a variable is related to the factor).
For the achievement of the 2nd goal, the State monitoring physico-chemical and diatom data of 2007 from 60 stations have been used. The measured physico-chemical parameters include biochemical oxygen demand in 7 days (BOD7), am-monia nitrogen (NH4-N), nitrate nitrogen (NO3-N), total nitrogen (Ntot), phosphate phosphorus (PO4-P) and total phosphorus (Ptot), while for diatoms taxa and their abundance were measured. Diatom taxa data was converted to 17 indices which are most commonly used in the world. Special OMNIDIA software for conversion was applied. Shapiro-Wilks and Kolmogorov-Smirnov tests were employed to check compatibility of indices with normality condition. In cases of non-conformity data transformations were made. Data preparation was followed by ANOVA and F test to detect indices that distinguish between groups of adjacent physico-chemical water quality classes. Hydrochemical water quality classes were determined using values for the afore-mentioned 6 parameters judging in accordance to water quality classification crite-ria, confirmed by the Order of the Minster of the Environment No.D1-176, signed on March the 4thwas assigned judging according to the, 2010. Water quality class parameter of the lowest value. Finally, boundary values for sensitive indices have been elaborated, distin-guishing between adjacent groups of water quality classes according to diatoms. The boundary values reflect averages of the 75thand 25thpercentiles of indices dis-tributions in corresponding adjacent groups of physico-chemical quality classes. Table 1. River monitoring stations from which the hydrochemical, biological and basin characteristic data have been used in the research Code Monitoring station Code Monitoring station Code Monitoring station R1NeDmruusnkainsianbkaoive at the mouR61 Vilnia thR151*la-Pelesa at Katos R3NeDmruusnkaisnibneklaoiw at KaltanR62 eimena Mera-Knai R175*na at PaeimenR4NemunastuasboveAly-R63eimenčiabe-nevaililowR176* Kena at Rukainiai on R5 Nemunas below Aly- R64 eimena above Pabrad at TaubuR217* euvisčiai tus R7NemPurnieansaaibove below PabradR65 eimena iemara R219* at Paparčiai R8 Nemunas above Kau- R66 Bka above BaluoasR265* Jra at Mocikiai nas R9NemunasnbelowKau-R67Strva below SemelikssaehmlVilliusKara267*RDrterev as nachannel a R11NemunlaisnibneklaoiwSma-R68 Strva at Litonys at GudaiR268* Vilka R12 Nemunas above RusnR69 Merkys above Varna at NasrR269* Salantasnai R14 Minija above Plung below PuvoR70 Merkysčiai R270* Graumena at Pakalnikiai R15 Minija below PlungobluSkrR71-nniuDibolwsebikasitnameAk*71R2-namargaPevoba
R16 Minija below Gargdai R73 alčia below alčininkaiR325* Dysna at KačergikR17 Minija below Priekulsiai-anenaD7RmkA4 ventoji at Sabaliat Tubau- R327*nai R18VVeeiivviirraisantagteniemkA57Ranna-DKrawbelo357* Nem R unlis at Tabokinena-Danabove R19 ya above ilut R76 Akm KlaipdaR358* K-5 at Pakriauniai R20 ya below ilutD-naAk77namethR ji at Bindzelikiai R359* venat the to mou R21 Jra above TauragGeivandiaiitonR78BartvuabavoekSouad36sRP0*esyvbeaeewtankiediai R22 Jra below Taurag Josvainis R361* at Oreliai below SkuodasR79 Bartuva R23 euvis at Skirgailiai R80 Venta above KurnaiR401* Rausvat Nadrausv altuona below R81 Venta below KurnaiR402* Viakis at Pilvikiai R24RaseiniaiR25 Lokysta below ilalR82 Venta R430* below Maeikiai at Grie Varduva R26 euptorderndPolaaeobthtR83 Virvytbelow Pateklala*P31R4dertoLatviaotinstatehbro R27 eup M R84below Kal-a above Kulpia-vatLtordeorehbtatva23A*R4 varija eupabove R28jampol MMari- R85a below KulpR433* Virvytat JanapolR29 eupjwollebMari-R86 Ma below Saloč Striai R566*va at Tadarava ampo R30 Siesartis below akiai R87 Sidabra below JonikisR612* Nemunas at Paggiai eimena below Vi - 88 S R31kavikislRidabraaLtatthveiabordertotulasbe1106*Skuitefi"kaanolwM"aRR32 elmenta at the border unlis below Pan to Poland R89 Nem munis e-bysa above R33DuSerediusR90Juodupbelow JuodupR34 Kraantabove KelmR91 Laukupbelow Rokikis R35 Kraantbelow Kelm above BiraiR92 TatulaR36NePvavensisy39RutaTbalowelirBiaevoba R37NePvanevistaaluterTebolwysR94Tačionys R38 NeveKbavoiniadiasi-R95 Lvuo above KupikisR39 NevainiaiwoldKiebs96L-Rvuo below Kupikis R40 Nevabsiiravdnod-uaRevo7LsR9vuo above Pasvalys R41 uvat the mouthR98 Lvuo at the mouth Juosta below Jacka-R99 Daugyvnat the mouthR42galisR43 Neris at Buivydiai at the mouthR100 Kruoja R44 Neris above Vilnius R101 Obelbelow Radvilikis 10
R45 Neris below Vilnius R102 Obelat the mouth R48 Neris above Jonava R103 Kulpbelow iauliai R49 Neris below Jonava R104 Kulpat the mouth R50 Neris above Kaunas R105 BirvattaehtbordertoBelaruslow R51LoKmaieniaadboerysR106 Laukesa below ZarasaiR52veAnntyokjič721RrikSiaieabovtvyabove RusnR53veAnntojkičiaibelowR133 ventoji at the mouthy -nas below Kaunas R54ventojmiearbgokUveatuKaltuumveuN6a31Reupat the border to R55ventojmiebreglow Uk-alinheKtR731noigerdargniR56irvintaaboveirvin-R138venttohjeiaBtatlhieSmeoauthattos t c rvinta below irvin- ina-eimena above R57 i tos R141 vog Vaininai R58Siesarltoisw-MMalolkiR1tarvin42ithtatatuheombsats-e R59 Vyuona below Utena R143 Siesartis at the mouth R60 Vilnia above N.Vilnia R150 Jiesia at Jiestrakis Here:Italic monitoring stations, where phytobenthos (diatoms) sampling data available; underlinedfont monitoring stations hydromorphologically altered river sections; * - monitoring stations, used only for the 2ndgoal of the dissertation (for the assessment of the suitability of diatom-based water status assess-ment met
Fig. 1.State river monitoring stations, the data from which were used in the dissertation 11