Intra-basin spatial approach on pollution load estimation in a large Mediterranean river

Desalination 250 (2010) 118–129

Contents lists available at ScienceDirect

Desalination j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / d e s a l

Intra-basin spatial approach on pollution load estimation in a large Mediterranean river Yorgos Chatzinikolaou a,b,⁎, Alexis Ioannou a, Maria Lazaridou a a b

Department of Biology, Aristotle University of Thessaloniki, GR-54124, Thessaloniki, Greece Institute of Inland Waters, Hellenic Center for Marine Research, P.O. Box 712, GR-19013, Anavyssos, Attica, Greece

a r t i c l e

i n f o

Article history: Accepted 29 December 2008 Available online 29 October 2009 Keywords: Biochemical oxygen demand Nutrient retention Organic pollution load Nutrient pollution load Pinios River

a b s t r a c t In order to find the segments of Pinios River which lack the retention capacity of the BOD and nutrient input, the difference of the estimated input and output pollution loads was compared at upstream and downstream clustered areas of a total of 73 segments. Catchment areas ranged from 1 to 11,300 km2. Emissions were always higher than the actual transport, therefore retention was assumed to take place. Specific runoff, percentage of the surface water area and the calculated input of pollution loads (BOD, P, N) varied between the different Pinios River basin catchment areas. The transport to the emission load ratio was different between large and small catchments. The rate of retention among consecutive segments revealed that four lowland segments lacked in their relative retention capacity. © 2009 Published by Elsevier B.V.

1. Introduction Pollutants from agricultural and farming activities, untreated urban sewage and industrial effluents when driven into the river by direct input, wastewater or surface runoff can cause severe pollution [1]. According to the European Union (EU) Water Framework Directive (WFD) 2000/60/EC [2] Member States must report on the type and magnitude of the significant anthropogenic pressures to which the surface water bodies in each River Basin are liable to be subject. This includes an estimation of significant point source of pollution, diffuse source of pollution, water abstraction, water flow regulations and morphological alterations. In order to identify whether a pressure on a water body is significant or not, a thorough knowledge of the pressures within the catchment area, together with a conceptual understanding of the water flow, chemical transfers and biological functioning of the water body within the catchments is needed [3]. Since organic pollution (nutrients included) is such a pressure, there is a need to identify in every basin the relevant importance of its driving forces (agricultural, urban, industrial etc.) in order to address the problems and take the appropriate measures. Point and diffuse source pollutants are becoming more of an issue in Greece due to an increase in developmental activities and the need to comply with the European Union environmental standards. While the point sources from municipal and industrial wastewater discharges can be monitored, quantified and treated before they are discharged into surface water bodies, the non-point pollution from agricultural runoff, landfill sites ⁎ Corresponding author. P.O. Box 712, Institute of Inland Waters, Hellenic Centre of Marine Research, GR 19013, Anavyssos, Greece. Tel.: +30 2291076395. E-mail address: [email protected] (Y. Chatzinikolaou). 0011-9164/$ – see front matter © 2009 Published by Elsevier B.V. doi:10.1016/j.desal.2008.12.062

and hazardous waste dumping sites is difficult to control [4]. The simplest method of estimating diffuse load is to estimate unit area loads or yields i.e. mass per unit area per unit time. It is based on the principle that, under the average hydrologic conditions, the different categories of land uses, such as agricultural, etc. will yield a relatively constant quantity of pollutant over the annual cycle [5]. To maintain pollution at an affordable level, the self-purification capacity of the river, or the assimilative capacity when referring to the riverine system, should remain sufficient to comply with the current pollution load all along the river [1]. According to Ostrumov [6] the assimilative capacity is based on a number of interconnected processes among which are: the export of pollutants to the adjacent land areas, sedimentation, evaporation, hydrolysis, photochemical transformations, chemical oxidation, biological sorption, biotransformations, water filtering by filter-feeder organisms, uptake of nutrients (including phosphate, nitrate ions and organic molecules) by organisms, biotransformation and sorption of pollutants in soil. A prevention measure of river water pollution based only on the establishment of maximum permitted pollution load limits cannot guarantee protection from pollution since the mechanisms (pollutants' transport, chemical reactions, biological uptake) which determine water quality are complicated. In the present study we used emission factors and exchange coefficients for point and diffuse sources of pollution to estimate emitted pollution loads. By the use of mass balance equations we compared emissions with the actual (transported) loads measured in the river. The aims were i) to assess the reliability of pollution load emission calculations, ii) to identify the relevant importance of point and diffuse pollution loads on transported loads at the studied sites and iii) to identify assimilation inefficient areas of the Pinios River.

Y. Chatzinikolaou et al. / Desalination 250 (2010) 118–129

2. Area description

3. Materials and methods

Alluvial deposits occupy 34% of the drainage basin surface, calcareous rocks occupy 18%, crystal schist rocks 16%, flysch with tertiary sediments 13%, and ophiolites 6% [7]. The Pinios River basin is 11,318 km2 (including the former Karla Lake basin), has a mean altitude 589 m and a basin mean slope of 6.92% (Fig. 1). Annual rainfall is about 700 mm but varies across the basin from approximately 500 mm in the vast alluvial plain to 1100 mm in the west mountainous area; the wettest months occur in the winter–spring period [8]. Three soil types were identified according to Fytianos et al. [9], recent alluvial Entisols and Inceptisols and hilly region Entisols. The mean water discharge at the river mouth is 86.1 m3/s according to the Greek National Bank of Hydrological and Meteorological Information. Under special circumstances discharge drops down to 3 m3/s [7]. The area with a population of 529,244 inhabitants [10] is mainly agricultural with localised industrial activity. Of the amount of the river water used 96% accounts for irrigation purposes and about 3.3% for water supply [11]. Irrigation needs are covered by surface water from river Pinios and its tributaries, which is distributed via a dense network of earthed irrigation canals. The main urban centres of the area are Larisa (126,076 inh.), Trikala (51,862) and Karditsa (37,768). Sewage treatment plants serve the main urban centres (40.6% of the basin population). In addition to the domestic sources of pollution, the localised industries situated close to the cities and the livestock wastes contribute to the total point pollution load. During summer, high temperatures and intense use of water result in the dramatic restriction of the river's water flow [7,12].

3.1. Preliminary water bodies

119

Since there is a lack of a national water body designation in Greece we used a modified version of the fluvial hydrosystems concept [13] in order to distinguish preliminary water bodies. Initially the main channel of Pinios River was divided in 500 m reaches, numbered from the river mouth to the far most source and coded according to a previous study in Axios-Vardar River by Chatzinikolaou et al. [14]. Next, the 500 m reaches were clustered into homogenous segments. The criteria and the means used to cluster the successive reaches are presented in Table 1 and were based on an extension of the “fluvial hydrosystems” [13] which in turn correspond to the “stream segments” [15]. Segments were not fixed according to administrative boundaries. Thus, a single segment may flow through one or several communes or municipalities. In such a segment there is only one geological underlying type, there is no tributary confluences or draining channels, the land use does not change significantly, no dramatic slope gradient occurs and no discontinuities from human induced alterations exist. 3.2. Transported pollution load measurement Sampling sites were located at the upstream and downstream end of each segment (preliminary water body) and at tributaries at the confluence point. The ‘snapshot’ methodology [16] was applied for the spatial analysis of organic and nutrient pollution inputs at an intra-

Fig. 1. The river basin of Pinios and the 89 sampling sites for the physicochemical parameters located at the main river and at the confluence of its main tributaries.

120

Y. Chatzinikolaou et al. / Desalination 250 (2010) 118–129

Table 1 The criteria used to define the homogenous segments of Pinios River. Criteria

Sources

Geology

Hydrolithological map. 1:1,000,000 Hellenic Ministry of Development, 1996 Input from tributaries and draining channels Topographic maps. 1: 50,000 Hellenic Military Geographical Service, 1970 Slope gradient dramatic shifts Topographic maps. 1: 50,000 Hellenic Military Geographical Service, 1970 Land uses: Natural (coniferous and broadleaf forests, marshes, and dunes), semi-natural (maquis, pastureland, not irrigated olive Air-photographs. 1:30,000 tree plantation), and non-natural (urban areas, intensive agricultural areas, industrial sites, roads etc.) Hellenic Military Geographical Service, 1995 Man induced discontinuities: Works that produce erosion in the immediate downstream, such as bridges with in-river supports, Pre-sampling field survey, 2002 fords, sluices, other construction works

basin-wide assessment. All water samples were taken during the beginning of autumn in 2002 (from 1/10/2002 to 5/10/2002). This time of the year responds to the low flow period when the pollution problems and the water abstraction for irrigation are at their highest levels. The weather conditions were fine-fair during the sampling days as well as the two prior days and thus no water quality alteration was produced from raining conditions in the catchment. The almost simultaneous measurement of water discharge, organic load (BOD5) and nutrient concentrations [orthophosphoric (PO4-P), nitrate (NO3-N), nitrite (NO2-N) and ammonium (NH3/NH4-N) ions] was used to provide the actual (transported) pollution load. When compared the transported load to the emitted (estimated) one a quantitative evaluation of the respectively input from the point and the diffuse sources can be done. Water was sampled and preserved in polypropylene bottles at 4 °C for further laboratory analysis of PO4-P, NO3-N, NO2-N, NH3/NH4-N according to the APHA standard methods for the examination of water and wastewater [17]. Because total phosphorus (TP) was not

measured the transported phosphorus load is represented by the soluble reactive phosphorus (SRP = PO4-P) [16]. Total nitrogen (TN) was not measured, therefore, the transported nitrogen load is represented by the dissolved inorganic nitrogen load (DIN = NH3/ NH4-N + NO2-N + NO3-N) [18]. Additionally, two water samples were collected in 300 ml BOD bottles and sealed in the dark at 20 °C. BOD5 was determined as the difference in dissolved oxygen measured at the time of collection and five days later with an oxygen meter (YSI 55). Discharge (Q) was calculated from channel dimensions and flow measurements [19]. Channel dimensions and flow were measured in situ with a tape-meter and a Swoffer 2000 flow/depth meter. Where not accessible, discharge was extracted from upstream and downstream measurements corrected for the catchment area with the formula:

Qi =

Qi−1 ⁎Abi−1 + Qi 2Abi

+ 1 ⁎Abi + 1

−QIr

ð1Þ

Fig. 2. The GIS extrapolation method for the calculation of point and non-point pollution loads in each catchments produced by the sampling sites in Pinios river basin. I) 89 sampling sites and the relief of Pinios River basin producing II) the 89 overlapped catchment areas, III) point pollution inventory calculated at commune level (NUTS5), and (IV) diffuse pollution load inventory calculated from Corine land-uses.

Y. Chatzinikolaou et al. / Desalination 250 (2010) 118–129

121

Table 2 Emission factors from Ioannou et al. [20] for the calculation of domestic, animal husbandry and diffuse pollution emissions at Pinios River basin. Pollution source

Pollution input

Point Point Point Point Point Point Point Point Point Diffuse Diffuse Diffuse Diffuse Diffuse

WWTP Network Tanks Bovine Horses Sheep Goats Pigs Poultry Surface waters Forested areas Grasslands Agricultural area Urban area

(g/p.e./day) (g/p.e./day) (g/p.e./day) (g/ind./day) (g/ind./day) (g/ind./day) (g/ind./day) (g/ind./day) (g/ind./day) (kg/ha/year) (kg/ha/year) (kg/ha/year) (kg/ha/year) (kg/ha/year)

BOD

TP

TN

As defined by the local agency 63 36 675 381.5 100.2 100.2 80 3.825

As defined by the local agency 4.5 2.4 22.5 7 4.2 4.2 11.2 0.55 0.01 0.1 0.5 0.5 1

As defined by the local agency 11 10.8 202.5 31.5 24.6 24.6 38.4 0.825 1.2 3 5 30 5

WWTP is waste water treatment plantation.

where Qi and Abi are the discharge and the catchment area of site i respectively. QIr is the discharge taken by abstraction at the segment between sites i and i − 1. We based the above formula on the assumption that neighbouring sites located at distances less than 20 km have a proportional to their catchment area discharge. Transport load (LBOD;P;N) was calculated from the BOD5; SRP; DIN and the discharge. For reasons of comparability LBOD;P;N was projected to annual loads.

3.4. Pollution load retention The mean annual transport or pollution loads of organic inputs and nutrients in a river system (LBOD;P;N) within a certain time period is the result of the sum of all point (Epoint) and diffuse (Ediffuse) emissions (EBOD;P;N) reduced by the sum of all load retention and losses due to retention (RBOD;P;N) [18]:

LBOD;P;N = EBOD;P;N −RBOD;P;N

ð2Þ

= ΣEpointBOD;P;N + ΣEdiffuseBOD;P;N −ΣRBOD;P;N :

3.3. Emitted pollution load estimation An inventory of pollution loads for each and every catchment was created. Catchment areas (Ab) for the 89 sampling points were extracted from the relief of Pinios River basin with the ArcGIS 9.1 watershed tool (Fig. 2). Point pollution loads were extracted from data (animal husbandry, urban population) provided by the National Greek Statistics Service, the industrial pollution load inventory of the Greek Ministry of Development and the output water quality of the water treatment plantations by the local waste water agencies (Trikala and Larisa). Data concerning domestic, animal husbandry and industry emissions were attributed to a single commune (administrative level NUTS 5). For the calculation of the emission loads (EBOD;P;N) of Pinios River we used emission factors for BOD, TP and TN [20] (Table 2). Each catchment area load was calculated with the help of GIS from the total sum of the pollution loads produced by the communities that are found in the catchments (Fig. 2). Although the administrative boundaries do not coincide with the catchment areas, due to the small area a community occupies, we accepted that the expected mistake is negligible. Specifically, the calculated amount of inhabitants, animal husbandry and industrial wastes for each commune was designated to the total emitted point pollution load of each sub-basin only if the sub-basin occupied more than 50% of the commune's area. The diffuse pollution emissions were calculated from the Corine (2007 released version) land-uses (Fig. 2).

Because the river basins differ in size, it is essential to eliminate the influence of catchment area when comparing river basins. Behrendt [21] has shown that the use of the ratio of load to the sum of emissions is one possible way of normalization. Solving Eq. (2): RLBOD;P;N =

EBOD;P;N −1 LBOD;P;N

ð3Þ

where RLBOD;N;P is the weighted retention load (RBOD;P;N divided by LBOD;P;N). To avoid zero values of the denominator, Eq. (3) was modified: RLBOD;P;N =

EBOD;P;N −1: LBOD;P;N + 1

ð4Þ

In order to identify problematic areas, where the rate of pollution load losses decelerates, we calculated the relative difference within each preliminary water body (i.e. RLi − 1 input minus the RLi output value). Nevertheless, the most impacting pollution load species (i.e. BOD, P and N) to the river health is unknown. Therefore, assimilation inefficient was considered the areas where decrease in the RL rate of BOD, P and N (dRBOD, P, N) occurred concurrently. dR was calculated from the difference in the transported to emission load ratio between the consecutive catchments including the input from

Table 3 BOD5, soluble reactive phosphorus (SRP) and dissolved inorganic nitrogen concentration among different large northern Greek rivers (river basin > 10,000 km2). River

Axios-Vardar (21) Pinios (89) Aliakmon (20) Aoos-Vjose (17)

BOD

SRP

DIN

Reference

Maximum

Average

Maximum

Average

Maximum

Average

5.73 mg/L 6.79 mg/L – 5.17 mg/L

2.48 mg/L 2.42 mg/L – 2.41 mg/L

1474 μg/L 231 μg/L 1296 μg/L 126 μg/L

789 μg/L 31 μg/L 173 μg/L 19 μg/L

4.51 mg/L 17.17 mg/L 3.20 mg/L 1.48 mg/L

2.05 mg/L 2.24 mg/L 1.17 mg/L 0.80 mg/L

[14] Current study [26] [27]

Sites were located across the river from the furthest source to the mouth. The number in brackets next to the river name concern the number of sites for which the average values were calculated. Samples were taken in different years but always during the end of the low flow period.

122

Y. Chatzinikolaou et al. / Desalination 250 (2010) 118–129

Fig. 3. I) Cumulative transported equivalent pollution load (L) (t/year) measured during the low flow period (2002).Composition of the emitted load (E) (t/year) of: II) phosphorus and III) nitrate loads according to their sources.

Y. Chatzinikolaou et al. / Desalination 250 (2010) 118–129

tributary catchments. It was studied only for the main river of Pinios and not for its tributaries: + 1LBOD;N;P −RiLBOD;N;P

dRiBOD;P;N = Ri

123

The parameter specific runoff (q) was calculated from the discharge for the studied period. Using q to denote specific runoff, the water input is [24]:

ð5Þ q=

where Ri + 1LBOD;N;P is the weighted dR for the immediate upstream site of site i. 3.5. Reliability of emitted pollution load estimation In order to identify significant effects that the river basin characteristics and the emitted pollution loads have on the calculated retention capacity, specific relevant catchment characteristics were calculated. The percentage of the catchment area occupied by surface water (W) was calculated using the following formula [18,22,23]:

Q : Ab−W

ð7Þ

3.6. Statistical analysis In order to find out the significance of the river basin and emitted pollution load on the depended value of the transported pollution load, the relation of the catchment area size was examined on the transported to emitted pollution loads ratio. We then applied a separate stepwise linear regression model for the different size group of catchment areas. 4. Results and discussion 4.1. Transported load

0:185

W = 0; 1*Ab

½%

where Ab denotes the river catchment area.

ð6Þ

In comparison to other Greek large rivers monitored during the low flow from source to mouth, Pinios River has the highest concentration of dissolved inorganic nitrates (Table 3). This

Fig. 4. I) The measured transported BOD load at Pinios River during the low flow season (2002) and the estimated total emitted BOD load that was calculated from the animal production (II), the industry (III) and the domestic wastes (IV). Area A is located at the Amygdalea straits.

124

Y. Chatzinikolaou et al. / Desalination 250 (2010) 118–129

confirms the characterization of the Pinios plain as a nitrate vulnerable zone according to the EU Directive 91/676/EC criteria [25].

4.2. Pollution load emissions proportion The total transported pollution loads from the source up to the confluence of Trikala canal were of minor quantity (Fig. 3I). From the canal confluence to the Koutsohero drainage canal non-point source nitrate pollution loads by agricultural origin prevailed (Fig. 3I and III). The largest quantity of the sum of transported pollution loads was located at the area from the industrial zone of Larisa (P122) and upstream the Tempi Gorge (P044). Considering the relevant importance of point and diffuse sources of P pollution load, the diffuse sources were the most important source of pollution load except for the water bodies downstream Trikala tributary (Fig. 3II). The Trikala canal receives the WWTP wastes of Trikala and therefore this increment of point source load was attributed to domestic waste. The mean overall participation of diffuse pollution load for P was 69%. The diffuse sources comprised 80% of the original N load. This load originated from farming activities and confirms the characterization of Pinios Plain as a vulnerable zone to nitrate pollution of agricultural origin [25].

The transported BOD load (LBOD) increased downstream of Trikala (Ρ328), whereas downstream of Larisa (Ρ122) a further increase occurred. Raised values were retained until the river mouth (Fig. 4I). In order to identify the driving force (domestic sewage, livestock waste, industrial waste, farming etc.) triggering the increments in the BOD values the different source emitted loads were examined separately as to the transported load. For the raised LBOD values at Amygdalea straits (Fig. 4II) only the livestock waste was concurrently increased and therefore the animal husbandry was held responsible. Regarding the second raise, downstream of Larisa at the industrial area, based on the current estimations of loads domestic, livestock and industrial production all are increasing and therefore no safe conclusion on the origin of it can be drawn. The transported SRP load (LP) values were low upstream the city of Trikala (P339) (Fig. 5). Downstream Trikala the LP values gradually increased. At the confluence of Koutsohero canal the rate of LP increment accelerated. The LP was further increased at the industrial area of Larisa and reached a peak at the confluence of Titarisios (Fig. 5). LP values decreased at the gorge of Tempi (Fig. 5). The increase that occurred at the Koutsohero canal was due to agricultural farming (Figs. 5 and 3II). The peaks at the industrial area of Larisa and Titarisios confluence were due to industrial wastes and domestic sewage.

Fig. 5. Ι) The transported P load at Pinios River during the low flow season (2002) and the estimated diffuse pollution P load and II) the estimated point pollution P load. Area A is located at the confluence of Koutsohero, B is at the industrial area of Larisa and C is at the downstream end of Tempi gorge.

Y. Chatzinikolaou et al. / Desalination 250 (2010) 118–129

125

Fig. 6. Ι) The transported Ν load at Pinios River during the low flow season (2002) and the estimated diffuse pollution Ν load, II) the estimated point pollution Ν load. Area A is located upstream of the Trikala canal confluence, B is at the industrial area of Larisa and C is the furthest downstream site.

Fig. 7. Dependency of the transport to emission ratio (R) of the BOD and nutrients on the catchment size for the 89 different catchments within Pinios River basin.

126

Y. Chatzinikolaou et al. / Desalination 250 (2010) 118–129

Table 4 Stepwise linear regression model for transported loads of BOD, phosphorus and nitrogen. Catchment size > 1000 km2

Average ± S.E. (t/year) Model efficiency R2 Std. error of the estimate ANOVA F Sig. Model standardized coefficients W Q q Area Point pollution load Diffuse pollution load Total pollution load

Catchment size < 1000 km2

BOD

P

N

BOD

P

N

427 ± 40.3

7.5 ± 1.01

516 ± 72.9

23 ± 3.6

0.17 ± 0.05

34.6 ± 22.97

0.682 171.678

0.356 6.119

0.382 432.941

0.542 14.268

0.809 0.129

0.779 64.101

115.830 0

29.791 0

33.448 0

36.627 0

63.705 0

52.816 0

− 1.107

− 0.525 1.134

0.736

0.826 0.596

The DIN transported pollution load (LN) values increased upstream the confluence of Trikala canal, the tributary that carries the domestic sewage wastes from the city of Trikala (Fig. 6). Further downstream LN was high due to the non-point pollution loads originated from farming. At the industrial area of Larisa LN increased (Fig. 6) due to point source pollution loads originating from livestock, domestic and industrial wastes. LN values decreased at the Tempi Gorge (Fig. 6). Site P003, located very close to the river mouth (1.5 km from the sea), had the highest LN value due to the impact of sea-water intrusion. 4.3. Reliability of emission loads' estimation As shown in Fig. 7 the catchment area affects the ratio of transported to emitted pollution loads. By examining the behaviour of the first derivative of the best fitted curve, we assumed that at basins

0.618

1.878

smaller than 1000 km2 the variation of the R values is disproportional regarding the basins larger. Thus, it is sound to suppose a limitation to this control at catchment area smaller than 1000 km2. The results confirm the assumption made by Behrendt and Opitz [18] concerning the separation of river basins to 1–1000 km2 and 1000–10,000 km2. The influencing factors produced by the regression analysis at the transported load differed according to the catchment area size class, as was found by Behrendt and Opitz [18]. During the first days of October, the low flow period, in catchments larger than 1000 km2, BOD, P and N transported pollution load can be predicted by the point, diffuse and total pollution emission loads respectively (Table 4). In the Mediterranean region environmental damage arising from pollution and water abstraction is most likely to happen during the low flow period [28]. Consequently, during October, at Pinios lowland sites with large catchment area (Ab > 1000 km2) the observed

Fig. 8. The standardized pollution load retention (RL) at Pinios River sub-basins (>1000 km2). The vertical cross lines refer to the segments that showed a lack in RL of BOD P and N, in comparison to the previous segment.

Y. Chatzinikolaou et al. / Desalination 250 (2010) 118–129

pollution loads are influenced by the pollution emissions and are independent of the Pinios River basin local characteristics. On the contrast, in catchments smaller than 1000 km2, BOD and N transported pollution load cannot be predicted solely by the emission loads. Consequently, the emitted BOD and nitrate pollution load estimation is not reliable. 4.4. Assimilation inefficient segments At the 89 sampling stations the emitted loads were always higher than the transported ones indicating pollution assimilation processes were under way. Phosphorus weighted retention values were higher at the lowland close to the river delta catchments, whereas, nitrate retention values were higher at the upstream part of the floodplain (Fig. 8). All three weighted retention loads (RLBOD;N;P) varied greatly throughout the sampling sites. BOD and phosphorus weighted retention values (RLBOD,P) were greater for small tributaries with negligible discharge, whereas some negative values at catchment areas smaller than 10 km2 could be the bias of the way the pollution load was estimated. Specifically, the assumption made for the point pollution load estimation that each commune's load contributed to the total emitted load of each sub-basin only if the sub-basin occupied more than 50% of the commune's area cannot be met at small catchments. None of the nitrate weighted retention (RLN) values was negative. According to Eq. (5) four preliminary water bodies of Pinios had a negative relative retention rate value (Fig. 8). These segments that presented a simultaneous drop in the rate of pollution load losses for BOD, P, N were the P350–376 at the area of Trikala, the P149–191 and P122–143 upstream and downstream Larisa respectively, and the P069–073. Nevertheless, Eq. (5) according to Eq. (4) is the ratio of emitted to transported pollution load difference of the upstream from the downstream site:

dRiBOD;P;N =

EiBOD;P;N LiBOD;P;N + 1

−

Ei1BOD;P;N Li1BOD;P;N + 1

:

ð8Þ

Considering that the sub-basins are clustered (each upstream site's sub-basin is included to the downstream's one), the emitted loads are progressively increasing downstream (EiBOD;P;N ≥ Ei − 1BOD;P;N + EiTrBOD;P;N). Thus, in Eq. (8) the relative lack of retention capacity (dRi < 0) between two consecutive sites occurs only if the transported pollution loads at the downstream site were bigger (Li < Li − 1) or if the emitted loads were not appropriately calculated due to undue input loads. Therefore, the relative lack of retention may be attributed to the insufficient natural biological and morphological functioning of the segment or the illegal input of pollution loads (e.g. from industries) that could not be calculated. Therefore, at these four segments an investigational monitoring for all quality elements (third monitoring level according to the WFD) should be established in order to determine the reason of inefficiency in the assimilation capacity and either impose necessary restoration measures or enforce the current legislation. 5. Conclusions In conclusion, the division of the Pinios River basin into subbasins proved that differences in the pollution exerted among the sub-basins exist in both quality and quantity. The estimates showed that greatest portion of pollution emissions originates at the vast alluvial Pinios plain and were confirmed by the transported BOD and nutrient load measurements. In comparison to other northern Greek rivers the Pinios dissolved inorganic nitrates average concentration was the highest. Roughly, about 80% of that nitrate originates from diffuse sources that come from the intensive agricultural land-uses.

127

The soluble reactive phosphorus transported load varied as to its origins depending on the river segment location in the basin. According to Karaer and Küçükballi [29] in cases where the pollution loads are controlled solely at a local administrative scale, as was the case for Pinios, the state of a river may deteriorate dramatically. This is due to the fact that a local municipality authorities can use the thresholds values for WWTP's wastewater inputs to the river without the consideration of the cumulative impact of all the inputs. A unique competent authority is necessary, according to article 5 at the WFD, in order to control and manage effectively the discharged pollution load inputs at a river basin scale. Additionally, an integrated management at the river basin scale must also include an extensive monitoring network. Tools that can facilitate the pollution emissions estimation and value their impact on water quality can be based on commune census data, specifically for catchment areas larger than 1000 km2. Weighted retention relative rate among consecutive segments may also assist the identification of their assimilative capacity efficiency. 6. Annexes Annex I The biological oxygen demand (BOD5) and the concentration of nutrients at 89 sites along Pinios River during the low flow period (2002). SRP is the dissolved reactive phosphorus and DIN is the sum of dissolved inorganic nitrate ions. Site

SRP BOD5 (mg/L) (μg/L)

P003 P004 P009 PML01 P027 P028 P035 P036 P044 P045 P061 P062 P068 P069 P073 P074 P085 P086 Titarisios P088 P122 P143 P144 P148 P149 P191 P192 P198 P199 P201 P202 P205 P206 P222 Canal Koutsohero P223 P262 Enippeas P263 P266 P275 Canal Klokotos P276 P300 P301

2.81 3.56 3.51 0 3.63 3.26 2.07 1.64 2.97 4.22 5.67 2.79 1.87 2.99 2.91 2.02 4.24 4.08 7.09 3.43 5.4 3.1 2.55 1.76 2.29 1.96 4.06 2.03 1.82 2.28 1.73 2.68 2.22 2.67 2.15

DIN Site (mg/L)

0 11.44 14.11 1.88 0 3.12 0 0 20.65 2.04 0 1.66 40.26 1.99 51.93 2.14 83.28 2.68 109.28 3.96 86.48 2.24 64.96 2.26 65.02 4.1 114.23 9.16 79.97 7.37 46.96 7.05 106.3 7.59 71.5 2.68 231.57 3.35 129.21 6.7 97.83 3.03 51.86 1.8 62.01 1.88 79.66 3.6 42.11 2.26 36.99 1.15 9.25 1.68 46.9 1.08 46.99 2.53 7.54 1.53 48.81 1.2 25.57 1.76 40.23 1.7 56.81 1.71 0 3.05

BOD5 SRP DIN (mg/L) (μg/L) (mg/L)

P326 P327 Pamisos P328 P339 Canal Trikala P340 P349 P350 P376 P377 P388 P389 P410 Tranos Lakos P411 P420 PUU01 P421 P424 PKL01 P425 P435 Kastaniotiko PMB01 PMB15 PTY01 PMB16 PMB19 Kriorema PMB20 PMB25 Malakasiotiko PKS01 PKS04

3.58 2.26 0.22 2.13 0.95 0.92 0.93 0.63 0.46 0.47 0.75 0.88 0.71 1.58 0 0.22 0.4 1.03 0.64 0.63 0.84 0.38 0.47 0.53 0.37 1.41 3.62 1.84 0.78 0.63 1.28 1.86 2.1 1.17 1.24

34.23 43.55 13.33 22.55 29.01 27.43 12.85 32.76 33.66 0 20.97 1.31 29.18 18.9 0 0 14.59 13.73 9.27 7.41 11.45 0 0 0 6.67 11.33 0 0 0 0 0 6.94 4.45 0 2.49

3.52 4.19 6.37 6.41 4.61 17.16 1.33 0.41 0.54 0.01 0.01 0.01 0.01 0.01 0 0 0.01 0 0.01 0.02 0.65 0 0 0 0 0.66 0 0.5 0.09 0 0 0 0 0 0

1.98 3.57 4.37 3.29 1.77 2.12 0.7

43.55 20.9 35.14 0.83 30.77 35.17 16.08

5.69 1.04 5.48 0.71 3.01 4.84 0.93

PKS05 PKS10 Anilio PKS11 PKS15 Kriasa PKS16

1.92 1.41 0 1.19 1.22 0.94 2.03

0 10.1 0 10.14 0 12.48 12.78

0 0.01 0 0.01 0 0 0

1.68 2.53 1.59

2.73 16 4.71

3.62 5.48 5.18

PKS23 PKS24

1.56 1.36

11.79 19.65

0 0.01

128

Y. Chatzinikolaou et al. / Desalination 250 (2010) 118–129

Annex II Basin, river and pollution load characteristics at the 89 sampling sites of the Pinios River basin. Site

P003 P004 P009 PML01 P027 P028 P035 P036 P044 P045 P061 P062 P068 P069 P073 P074 P085 P086 Titarisios P088 P122 P143 P144 P148 P149 P191 P192 P198 P199 P201 P202 P205 P206 P222 Canal Koutsohero P223 P262 Enippeas P263 P266 P275 Canal Klokotos P276 P300 P301 P326 P327 Pamisos P328 P339 Canal Trikala P340 P349 P350 P376 P377 P388 P389 P410 Tranos Lakos P411 P420 PUU01 P421 P424 PKL01 P425 P435 Kastaniotiko PMB01 PMB15 PTY01 PMB16 PMB19

Area

W

Q

q

LBOD

EBOD

LP

Diffuse EP

Point EP

LN

Diffuse EN

km2

%

m3/s

l/s km2

t/year

t/year

t/year

t/year

t/year

t/year

t/year

Point EN t/year

11,289 11,288 11,221 15 11,154 11,141 11,129 11,127 11,112 11,111 11,085 11,084 11,003 10,987 10,985 10,955 10,937 10,936 1916 9011 7841 7824 7053 7052 7051 6929 6924 6913 6898 6897 6886 6883 6859 6825 247 6578 6460 3235 3226 3216 3179 417 2755 2582 2581 2269 2242 255 1988 1985 336 1647 1640 1267 1230 1226 1220 1185 1125 21 1099 1056 493 560 560 186 372 346 88 255 225 1 223 211

0.56 0.56 0.56 0.16 0.56 0.56 0.56 0.56 0.56 0.56 0.56 0.56 0.56 0.56 0.56 0.56 0.56 0.56 0.40 0.54 0.53 0.53 0.52 0.52 0.52 0.51 0.51 0.51 0.51 0.51 0.51 0.51 0.51 0.51 0.28 0.51 0.51 0.45 0.45 0.45 0.44 0.31 0.43 0.43 0.43 0.42 0.42 0.28 0.41 0.41 0.29 0.39 0.39 0.38 0.37 0.37 0.37 0.37 0.37 0.18 0.37 0.36 0.31 0.32 0.32 0.26 0.30 0.29 0.23 0.28 0.27 0.09 0.27 0.27

7.689 7.688 7.655 0.001 7.621 7.615 7.258 7.258 7.250 7.249 7.237 7.236 7.196 7.188 7.186 7.171 7.162 7.162 2.652 6.199 5.614 5.606 5.221 5.220 5.220 5.159 5.156 5.151 5.143 5.142 5.137 5.136 5.124 5.107 0.995 4.983 4.924 3.312 3.307 3.302 3.284 1.690 3.071 2.985 2.985 2.829 2.815 1.028 3.288 3.287 1.360 3.118 3.114 2.928 2.309 2.307 2.304 2.287 2.257 0.001 1.866 2.048 2.003 2.233 2.337 0.745 1.505 1.317 0.343 1.244 0.903 0.007 0.895 1.321

1.5550 1.5550 1.5553 0.0807 1.5555 1.5556 1.4841 1.4841 1.4841 1.4841 1.4841 1.4841 1.4841 1.4841 1.4841 1.4841 1.4841 1.4841 2.3257 1.4926 1.5086 1.5089 1.5267 1.5267 1.5268 1.5302 1.5304 1.5307 1.5312 1.5312 1.5315 1.5316 1.5323 1.5334 5.5644 1.5415 1.5457 1.8478 1.8496 1.8517 1.8595 5.8350 1.9662 2.0205 2.0206 2.1406 2.1528 5.5806 2.7917 2.7944 5.7211 3.1217 3.1300 3.6961 2.9939 2.9999 3.0085 3.0649 3.1682 0.0519 2.6737 3.0422 5.9278 5.8873 6.1591 5.4227 5.7737 5.3954 5.0351 6.7617 5.5169 12.3828 5.5125 8.5607

681.3 863.2 847.3 0.0 872.4 782.9 473.8 375.4 679.0 964.8 1294.0 636.7 424.3 677.7 659.5 456.8 957.7 921.6 593.0 670.6 956.1 548.1 419.8 289.7 377.0 318.9 660.1 329.7 295.2 369.8 280.3 434.1 358.7 430.0 67.5 311.1 554.4 456.4 343.1 184.3 219.5 37.3 162.7 238.2 149.7 319.4 200.6 7.1 220.9 98.5 39.4 91.4 61.9 42.5 34.2 54.6 63.9 51.2 112.4 0.0 12.9 25.8 65.1 45.1 46.4 19.7 18.0 19.5 5.7 14.5 40.2 0.8 51.9 32.5

24,345.1 24,345.1 24,231.0 25.9 24,130.2 24,081.8 24,081.8 24,081.8 24,081.8 24,057.2 24,057.2 24,057.2 23,718.5 23,621.3 23,621.3 23,421.6 23,421.6 23,421.6 5763.7 17,657.9 13,787.2 13,787.2 12,148.1 12,148.1 12,148.1 12,030.2 12,030.2 12,030.2 12,030.2 12,030.2 12,030.2 12,030.2 12,030.2 11,997.7 295.9 11,701.7 11,543.8 4901.3 6642.5 6509.2 6451.7 828.9 5622.8 5074.4 5074.4 4671.3 4660.4 653.4 4007.0 4007.0 1373.5 2633.4 2620.3 1482.7 1340.1 1340.1 1340.1 1340.1 1115.2 61.4 1053.8 1033.3 634.4 388.7 388.7 204.5 184.2 184.2 67.9 116.2 94.6 12.0 63.3 63.3

0.0 3.4 0.0 0.0 5.0 0.0 9.3 11.9 19.0 25.0 19.7 14.9 14.8 25.9 18.1 10.6 24.0 16.2 19.3 25.3 17.3 9.2 10.2 13.1 6.9 6.0 1.5 7.6 7.6 1.2 7.9 4.2 6.5 9.2 0.0 6.9 3.2 3.7 0.1 3.2 3.7 0.8 0.3 1.5 0.4 3.0 3.9 0.4 2.3 3.0 1.2 1.2 3.2 3.1 0.0 1.5 0.1 2.1 1.4 0.0 0.0 0.9 0.6 0.6 0.6 0.3 0.0 0.0 0.0 0.3 0.3 0.0 0.0 0.0

447.8 447.7 445.0 0.5 442.2 441.6 441.0 440.9 440.2 440.2 438.9 438.9 435.1 434.3 434.2 432.8 431.9 431.9 90.4 341.1 285.0 284.0 243.4 243.4 243.3 237.1 237.0 236.4 235.7 235.6 235.0 234.9 234.1 232.3 12.5 219.8 214.3 103.7 129.0 117.9 126.5 20.3 106.0 97.0 97.0 86.5 85.2 10.5 74.6 74.4 16.9 57.5 57.2 43.8 42.0 41.9 41.5 40.1 37.5 1.1 36.3 34.7 18.2 16.4 16.4 5.6 10.7 10.0 2.2 7.7 6.4 0.0 6.3 5.8

230.8 230.8 230.4 0.2 229.9 229.7 229.7 229.7 229.7 229.6 229.6 229.6 227.7 227.2 227.2 226.0 226.0 226.0 33.8 192.2 174.3 174.3 165.3 165.3 165.3 164.8 164.8 164.8 164.8 164.8 164.8 164.8 164.8 164.7 2.0 162.7 161.8 45.4 116.4 115.7 115.3 4.4 110.9 107.6 107.6 105.7 105.6 3.6 102.1 102.1 83.7 18.4 18.4 8.2 7.4 7.4 7.4 7.4 6.3 0.3 6.0 5.9 3.6 2.2 2.2 1.2 1.0 1.0 0.4 0.7 0.5 0.1 0.3 0.3

2775.1 455.8 753.7 0.0 490.7 398.5 455.2 488.8 612.8 906.2 511.6 515.0 930.4 2075.3 1669.6 1595.5 1713.6 605.5 280.4 1310.0 536.2 318.6 310.2 592.4 371.8 187.4 273.3 176.0 410.1 248.3 194.4 285.6 274.8 275.4 95.8 894.1 161.6 572.0 73.9 313.4 501.5 49.7 350.5 516.0 487.3 313.7 371.7 206.5 664.8 478.2 736.1 130.7 40.1 49.7 0.5 0.9 1.0 0.9 1.1 0.0 0.1 0.4 0.0 0.4 1.5 15.1 0.2 0.0 0.0 0.2 18.6 0.0 14.2 3.9

15,299.5 15,298.1 15,186.9 6.1 15,094.8 15,087.8 15,078.9 15,077.3 15,058.8 15,057.1 15,010.3 15,008.7 14,911.6 14,897.4 14,896.2 14,881.1 14,865.1 14,864.7 2586.3 12,268.0 10,001.8 9964.6 8015.9 8011.8 8010.8 7668.8 7665.5 7656.2 7641.8 7637.7 7626.2 7625.8 7604.4 7577.5 639.3 6930.6 6781.2 3608.1 4275.2 3879.6 4150.9 459.9 3685.5 3222.9 3223.2 2804.2 2747.0 436.0 2304.1 2295.9 631.6 1663.6 1646.5 1234.2 1143.3 1140.8 1131.9 1101.1 1024.6 19.9 1002.3 953.3 544.4 406.5 405.3 146.4 257.3 233.5 46.9 183.0 157.3 0.3 154.4 141.7

3725.0 3725.0 3713.7 5.1 3702.1 3695.6 3695.6 3695.6 3695.6 3692.1 3692.1 3692.1 3633.1 3615.7 3615.7 3581.1 3581.1 3581.1 999.5 2581.6 2191.8 2191.8 1816.7 1816.7 1816.7 1802.3 1802.3 1802.3 1802.3 1802.3 1802.3 1802.3 1802.3 1797.1 51.5 1745.5 1720.3 643.0 1077.3 1058.4 1049.3 135.0 914.3 834.9 834.9 782.1 781.1 93.4 687.7 687.7 218.0 469.7 468.4 239.1 220.0 220.0 220.0 220.0 185.1 8.8 176.3 173.0 108.0 63.4 63.4 34.4 29.1 29.1 11.1 18.0 14.1 1.9 9.0 9.0

Y. Chatzinikolaou et al. / Desalination 250 (2010) 118–129

129

Annex II (continued) Site

Kriorema PMB20 PMB25 Malakasiotiko PKS01 PKS04 PKS05 PKS10 Anilio PKS11 PKS15 Kriasa PKS16 PKS23 PKS24

Area

W

Q

q

LBOD

EBOD

LP

Diffuse EP

Point EP

LN

Diffuse EN

Point EN

km2

%

m3/s

l/s km2

t/year

t/year

t/year

t/year

t/year

t/year

t/year

t/year

0.14 0.27 0.27 0.23 0.23 0.23 0.23 0.22 0.13 0.22 0.21 0.14 0.21 0.16 0.15

0.005 0.820 1.027 0.712 0.408 0.443 0.354 0.424 0.003 0.264 0.375 0.003 0.209 0.027 0.012

0.1 5.6 5.2 2.3 2.8 2.5 2.5 1.7 0.1 1.5 1.4 0.1 1.2 0.2 0.2

0.0 0.3 0.3 0.1 0.2 0.2 0.2 0.1 0.0 0.1 0.1 0.0 0.1 0.1 0.0

6 205 195 96 97 91 90 72 4 66 62 5 56 11 8

0.9431 5.4696 7.1801 9.6622 5.4621 6.3046 5.0874 7.5828 0.8419 5.0837 7.6531 0.6393 4.7538 2.8556 1.8255

0.1 33.1 60.2 47.1 15.0 17.3 21.4 18.9 0.0 9.9 14.4 0.1 13.4 1.4 0.5

0.0 63.3 58.5 13.5 45.0 32.8 32.8 23.0 0.0 23.0 23.0 0.0 23.0 18.1 0.0

Acknowledgments This study is funded by a grant from the Hellenic Ministry of Education as part of the PhD thesis “HRAKLEITOS: Effect of management practices on the water quality and ecology of rivers in Greece”. Contract number: MIS 88740 (8/11/2002).

[15]

[16]

[17]

References [1] M. Campolo, P. Andreussia, A. Soldati, Water quality control in the river Arno, Water Res. 36 (2002) 2673–2680. [2] European Union, Directive 2000/60/EC establishing a framework for community action in the field of water policy, Water Framework Directive, 2002. [3] A. Andreadakis, E. Gavalakis, L. Kaliakatsos, C. Noutsopoulos, A. Tzimas, The implementation of the Water Framework Directive (WFD) at the river basin of Anthemountas with emphasis on the pressures and impacts analysis, Desalination 210 (2007) 1–15. [4] A. Sargaonkar, Estimation of land use specific runoff and pollutant concentration for Tapi River basin in India, Environ. Monit. Assess. 117 (2006) 491–503. [5] Ryding, S.-O. and Rast, W. (1989) ‘The control of eutrophication of lakes and reservoir’, Man and the Bioshepre Series Vol. 1. United Nations Educational, Scientific and Cultural Organization (UNESCO), Parthenon, Carnforth, Lancashire, pp. 314, ISBN 0-929858-13-1. [6] S.A. Ostroumov, Polyfunctional role of biodiversity in processes leading to water purification: current conceptualizations and concluding remarks, Hydrobiologia 469 (2002) 203–204. [7] G. Stamatis, The chemical composition of the surface system of Peneos river, Thessaly/Central Greece, Environ. Geol. 38 (1999) 126–140. [8] D. Koutsogiannis, The Report of the Hydrological Study of the Thessaly Hydrological District, Metsovian National Polytechnical Scool, Athens, 1988 in Greek. [9] K. Fytianos, A. Siumka, G.A. Zachariadis, S. Beltsios, Assessment of the quality characteristics of Pinios River, Greece, Water Air Soil Pollut. 136 (2002) 317–329. [10] General Secretariat of National Statistics Service of Greece, The real population report of prefectures, municipalities and communes of Greece at 2001, General Secretariat of National Statistics Service of Greece, Athens, 2002. [11] Hellenic Ministry of Development, Master Management Plan for National Water Resources, Directorate of Hydrology and Natural Resources, Hellenic Ministry of Development, Athens, 2003. [12] D. Bellos, T. Sawidis, Chemical pollution monitoring of the River Pinios (Thessalia— Greece), J. Environ. Manag. 76 (2005) 282–292. [13] C. Amoros, A.L. Roux, J.L. Reygrobellet, J.P. Bravard, G. Pautou, A method for applied ecological studies of fluvial hydrosystems, Regul. Rivers 1 (1987) 17–36. [14] Y. Chatzinikolaou, V. Dakos, M. Lazaridou, Longitudinal impacts of anthropogenic pressures on benthic macroinvertebrate assemblages in a large transboundary

[18] [19] [20]

[21]

[22]

[23]

[24] [25]

[26]

[27]

[28]

[29]

0.0 0.0 0.2 0.1 0.0 0.0 0.0 0.1 0.0 0.1 0.0 0.0 0.1 0.0 0.0

0.0 0.1 0.2 0.1 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0

1.8 139.2 131.7 50.7 80.2 70.2 68.5 43.3 2.1 38.5 34.1 2.1 31.4 4.1 2.7

0.0 9.0 8.3 1.8 6.6 4.7 4.7 3.3 0.0 3.3 3.3 0.0 3.3 3.3 0.0

Mediterranean river during the low flow period, Acta Hydrochim. Hydrobiol. 34 (2006) 453–463. C.A. Frissell, W.J. Liss, C.E. Warren, M.D. Hurley, A hierarchical framework for stream habitat classification: viewing streams in a watershed context, Environ. Manage. 10 (1986) 199–214. M. Salvia, J.F. Iffly, P. Vander Borght, M. Sary, L. Hoffmann, Application of the ‘snapshot’ methodology to a basin-wide analysis of phosphorus and nitrogen at stable low flow, Hydrobiologia 410 (1999) 97–102. APHA, Standard Methods for the Examination of Water and Wastewater16th Edition, American Public Health Association Inc, Washington D.C, 1985. H. Behrendt, D. Opitz, Retention of nutrients in river systems: dependence on specific runoff and hydraulic load, Hydrobiologia 410 (2000) 111–122. A.J. Horne, C.R. Goldman, Limnology. McGraw-Hill International Editions, New York, 1983. A. Ioannou, Y. Chatzinikolaou, M. Lazaridou, A preliminary pressure-impact analysis applied in the Pinios River basin (Thessaly, central Greece), Water and Environment Journal 23 (2009) 200–209. H. Behrendt, Inventories of point and diffuse sources and estimated nutrient loads — a comparison for different river basins in Central Europe, Water Sci. Technol. 33 (1996) 99–107. G. Billen, The PHISON river system: a conceptual model of C, N and P transformations in the aquatic continuum from land to sea, in: R. Wollast, F.T. Mackenzie, L. Chou (Eds.), Interactions of C, N,P and Si Biogeochemical Cycles and Global Change. NATO ASI Series, vol. 14, 1993, pp. 141–161. G. Billen, J. Garnier, C. Billen, E. Hannon, Global change in nutrient transfer from land to sea: biogeochemical processes in river systems, GMMA, Free University of Brussels, 1995. L. Bengtsson, J. Malm, Using rainfall-runoff modeling to interpret lake level data, J. Paleolimnol. 18 (1997) 235–248. EC (European Council, Council Directive of 12 December 1991 concerning the protection of waters against pollution caused by nitrates from agricultural sources (91/676/EEC), Official Journal of European Communities L 375 (2002) 1–12. M. Lazaridou-Dimitriadou, V. Artemiadou, G. Yfantis, I. Mourelatos, Y. Mylopoulos, Contribution to the ecological quality of Aliakmon river (Macedonia, Greece): a multivariate approach, Hydrobiologia 410 (1999) 47–58. Y. Chatzinikolaou, V. Dakos, M. Lazaridou, Assessing the Ecological Integrity of a Major Transboundary Mediterranean River Based on Environmental Habitat Variables and Benthic Macroinvertebrates (Aoos-Vjose River, Greece-Albania), Int. Rev. Hydrobiol. 93 (2008) 73–87. A. Gasith, V.H. Resh, Stream in Mediterranean climate regions: abiotic influences and biotic responses to predictable seasonal effects, Annu. Rev. Ecol. Syst. 30 (1999) 51–81. F. Karaer, A. Küçükballi, Monitoring of water quality and assessment of organic pollution in the Níllüfer stream, Turkey, Environ. Monit. Assess. 114 (2006) 391–417.