**5. Results of the performed surveillance monitoring in Prespa Lake watershed**

According to the historical hydrology data and observations, there are four major tributaries in the Macedonian part of Macro Prespa Lake which can be considered as separate rivers of significance and should be subsequently subdivided into the Istočka, Golema, Kranska and Brajčinska Rivers. Although small in it watershed, Kurbinska River was also taken into consideration because of it relatively significant quantity of water.

All of the other temporal water courses have been proven to be torrent carriers which usually drain forested areas and have little significance in the overall water quality analysis; although they may have a significant role in the water balance of the lake at specific times, as well as in flood hazard management because of their character.

Considering the Prespa Lake itself, there is a littoral plateau of approximately 14-16 metres depth that completely surrounds the lake and two major depressions – one near the village

Environmental Changes in Lakes Catchments

**Nutrient status** 

Golema Reka 1 Golema Reka 2 Golema Reka 3 Golema Reka 4 Golema Reka 5

Golema Reka 6

0

5

10

15

20

25

as a Trigger for Rapid Eutrophication – A Prespa Lake Case Study 87

Fig. 13. Basic physical parameters of delineated rivers in Prespa Lake watershed.

Kurbinska Reka

Istocka Reka 1 Istocka Reka 2 Kranska 1 Kranska 2 Brajcinska

1

Brajcinska 2

pH DO mg/l Т (оC) TDS (ppm) Conduc.(µS)

TDS Cond.

Cheshinska Reka

caused by prolonged deposition of acidifying chemicals.

experience intensive pressure from nutrients.

Measured major nutrients in delineated river water bodies in Prespa Lake watershed (Fig.14*)*  were mostly based on *nitrogen* and *phosphorus* compounds, with the addition of *sulphates* as an indicator of the direct human influence. All of the examined water bodies show highly increased nutrient composition, especially regarding the *total N* and *P* compounds. In water bodies of the Golema and Istočka Rivers, *ammonia* and especially *sulphates* were very high, but their presence is also remarkable in the rest of the examined rivers. *Nitrites* were present in all examined water bodies, except in Kurbinska River, in low quantities, pointing to continual wastewater pressure. Detected high amounts of nutrients in examined river water bodies points to a complete lack of wastewater treatment, significant diffuse source pollution run-off and vast amounts of waste and agricultural and industrial discharge (refer to Fig. 15 for an estimation of the total annual load of nutrients by different river water bodies). These results might explain, in broad terms, the detected variations of *acidity* in examined waters which were most probably directly connected to the discharged nutrient quantities and the season of fertiliser application; all of the examined nutrients, especially sulphates, have high potential of forming acids which have rapid or prolonged impact on water biota and chemistry. In other words, river water bodies in Prespa Lake watershed have very low acid neutralising capacity

Amid the lack of categorisation of surface waters according to nutrients in the Macedonian legislation (only the values for *total N* and *ammonia* are considered, but with wrongly stated units of measurement; *Official Gazette of RM 18/99*), the detected values for *phosphates, sulphates, total N* and *ammonia* (Fig.14) place the lower parts of the examined water bodies highly in the V water quality class. In this respect, Golema River and Istočka River have the maximum detected levels, but even Kranska River and Brajčinska River have been shown to

According to results presented in Figure 15, *sulphates* are the major nutrient carried to Prespa Lake mostly by Brajčinska, Istočka and Golema Rivers. Significant amounts of *nitrogen* and

of Stenje (approximately 47 metres depth) and the other in the vicinity of Agios Germanos in Greece (approximately 58 metres depth) that represent the true profundal of the ecosystem. It is therefore a logical solution to delineate the water bodies in the Macedonian part of the Prespa Lake according to sampling sites presented on Figure 12, and to consider the littoral and profundal part of the lake as a one water body.

This delineation of the surface water bodies in the Macedonian part of Macro Prespa Lake has been used as a basis for a 12 month surveillance monitoring programme conducted in the period April 2010 - June 2011.

## **5.1 Rivers**

All rivers in Prespa Lake watershed belong to the same river type (type 1 – siliceous rivers of mid-altitude and size) in the one eco-region 6 (Hellenic Western Balkan). They are mostly mountain type rivers with steep slopes and a stony bottom, normally with well-developed riparian vegetation; only the small part of the watercourses prior to their mouth in Prespa Lake may be slow flowing and on sandy or muddy substrata. In combination with the very small river lengths (and also catchments - <100 km2), the only important driving force that might impact the natural ecological conditions in the rivers are human activities.

Due to the lack of a continual monitoring system in the area, the short timeframe of the project and the need for obtaining the most reliable data relevant for the development of the Management Plan, delineated river water bodies in Prespa Lake watershed were monitored for the basic physical, chemical (including major nutrients, priority substances and heavy metals) and biological parameters. Out of the WFD proposed biological quality elements, the categorization of rivers is chosen to be based on phytobenthos and macroinvertebrates. The reference conditions are also based on these parameters.

#### **Basic physical parameters**

The basic physical parameters measured during the established monitoring system in the course of the project (Fig. 13) reflect the overall natural conditions of the selected river water bodies. The only obvious sign of human impact are the values for *conductivity* and *total dissolved solids* which are much higher in Golema River and Istočka River. Values for *pH* are also worth notifying since in July they were also significantly lower than expected in Golema Reka River water bodies (3.92-5.17 respectively), but also in Kurbinska River (3.9!). This finding may be attributed to some specific pollution impact at the moment of sampling, but also to some other origin and therefore deserve investigative monitoring in future.

The values for dissolved oxygen generally decreased in the water bodies under intensive human influence (Golema River), but were still in the realm of a good ecological status; Kurbinska River is again an exception since there is no major human impact (apart of the water abstraction) recorded for this water body, but its DO value was only 5.4 mg O2. L-1. Nevertheless, the performed monitoring period was far too limited to enable any firm conclusions on the oxygen dynamics or the underlying causes; it should be substantially extended in future.

Fig. 13. Basic physical parameters of delineated rivers in Prespa Lake watershed.

#### **Nutrient status**

86 Studies on Environmental and Applied Geomorphology

of Stenje (approximately 47 metres depth) and the other in the vicinity of Agios Germanos in Greece (approximately 58 metres depth) that represent the true profundal of the ecosystem. It is therefore a logical solution to delineate the water bodies in the Macedonian part of the Prespa Lake according to sampling sites presented on Figure 12, and to consider the littoral

This delineation of the surface water bodies in the Macedonian part of Macro Prespa Lake has been used as a basis for a 12 month surveillance monitoring programme conducted in

All rivers in Prespa Lake watershed belong to the same river type (type 1 – siliceous rivers of mid-altitude and size) in the one eco-region 6 (Hellenic Western Balkan). They are mostly mountain type rivers with steep slopes and a stony bottom, normally with well-developed riparian vegetation; only the small part of the watercourses prior to their mouth in Prespa Lake may be slow flowing and on sandy or muddy substrata. In combination with the very small river lengths (and also catchments - <100 km2), the only important driving force that might impact the natural ecological conditions in the rivers are human

Due to the lack of a continual monitoring system in the area, the short timeframe of the project and the need for obtaining the most reliable data relevant for the development of the Management Plan, delineated river water bodies in Prespa Lake watershed were monitored for the basic physical, chemical (including major nutrients, priority substances and heavy metals) and biological parameters. Out of the WFD proposed biological quality elements, the categorization of rivers is chosen to be based on phytobenthos and macroinvertebrates.

The basic physical parameters measured during the established monitoring system in the course of the project (Fig. 13) reflect the overall natural conditions of the selected river water bodies. The only obvious sign of human impact are the values for *conductivity* and *total dissolved solids* which are much higher in Golema River and Istočka River. Values for *pH* are also worth notifying since in July they were also significantly lower than expected in Golema Reka River water bodies (3.92-5.17 respectively), but also in Kurbinska River (3.9!). This finding may be attributed to some specific pollution impact at the moment of sampling, but also to some other origin and therefore deserve investigative monitoring in

The values for dissolved oxygen generally decreased in the water bodies under intensive human influence (Golema River), but were still in the realm of a good ecological status; Kurbinska River is again an exception since there is no major human impact (apart of the water abstraction) recorded for this water body, but its DO value was only 5.4 mg O2.

Nevertheless, the performed monitoring period was far too limited to enable any firm conclusions on the oxygen dynamics or the underlying causes; it should be substantially

L-1.

and profundal part of the lake as a one water body.

The reference conditions are also based on these parameters.

the period April 2010 - June 2011.

**5.1 Rivers** 

activities.

future.

extended in future.

**Basic physical parameters** 

Measured major nutrients in delineated river water bodies in Prespa Lake watershed (Fig.14*)*  were mostly based on *nitrogen* and *phosphorus* compounds, with the addition of *sulphates* as an indicator of the direct human influence. All of the examined water bodies show highly increased nutrient composition, especially regarding the *total N* and *P* compounds. In water bodies of the Golema and Istočka Rivers, *ammonia* and especially *sulphates* were very high, but their presence is also remarkable in the rest of the examined rivers. *Nitrites* were present in all examined water bodies, except in Kurbinska River, in low quantities, pointing to continual wastewater pressure. Detected high amounts of nutrients in examined river water bodies points to a complete lack of wastewater treatment, significant diffuse source pollution run-off and vast amounts of waste and agricultural and industrial discharge (refer to Fig. 15 for an estimation of the total annual load of nutrients by different river water bodies). These results might explain, in broad terms, the detected variations of *acidity* in examined waters which were most probably directly connected to the discharged nutrient quantities and the season of fertiliser application; all of the examined nutrients, especially sulphates, have high potential of forming acids which have rapid or prolonged impact on water biota and chemistry. In other words, river water bodies in Prespa Lake watershed have very low acid neutralising capacity caused by prolonged deposition of acidifying chemicals.

Amid the lack of categorisation of surface waters according to nutrients in the Macedonian legislation (only the values for *total N* and *ammonia* are considered, but with wrongly stated units of measurement; *Official Gazette of RM 18/99*), the detected values for *phosphates, sulphates, total N* and *ammonia* (Fig.14) place the lower parts of the examined water bodies highly in the V water quality class. In this respect, Golema River and Istočka River have the maximum detected levels, but even Kranska River and Brajčinska River have been shown to experience intensive pressure from nutrients.

According to results presented in Figure 15, *sulphates* are the major nutrient carried to Prespa Lake mostly by Brajčinska, Istočka and Golema Rivers. Significant amounts of *nitrogen* and

Environmental Changes in Lakes Catchments

161.9

into Prespa Lake.

NO3-N (mg/l) NO2-N (mg/l) NH3-N (mg/l) Total N (mg/l) P2O5-P(mg/l) PO4 -P(mg/l) Sulphates (mg/l) Total P compounds Total N compounds

**Heavy metals** 

water habitats.

watershed.

400 500 600

as a Trigger for Rapid Eutrophication – A Prespa Lake Case Study 89

**Total Load of nutriens from river water bodies into Prespa Lake**

1286.3

Fig. 16. Estimation of the cumulative load of nutrients coming from all river water bodies

0 500 1000 1500 2000 2500 3000 3500

t/year

Three heavy metals dominate the river water bodies of Prespa Lake watershed: *manganese, iron* and *aluminium*. Their recorded concentrations are usually way beyond the permissible levels for natural conditions (III-IV water quality class) and therefore denote an intensive human origin (Fig. 17). Again, Golema and Istočka Rivers express the highest concentration levels, while the lower segment of Kranska River recorded very high *manganese* and slightly lower *iron* concentrations. Upper segments of all river water bodies (the reference conditions) have been found with heavy metals concentrations several times lower than the lower parts, thus reflecting the natural background emissions to the

**Heavy metals in all river water bodies** 

Mn; Hg; Cu Hg (>1 μg/l) - V class Fe (>300-1000 μg/l) - III-IV class Cu (>50 μg/l) - V class Mn (>50-1000 μg/l) - III-IV class

Fig. 17. Detected concentrations of heavy metals in river water bodies of Prespa Lake

March April July July July July March April July July March March April March March April March March April

1

Kranska 2 Brajcinska 1

Brajcinska 2

As [μg/L] Al [μg/L] Cd [μg/L] Cr [μg/L] Fe [μg/L] Ni [μg/L] Pb [μg/L] Zn [μg/L] Cu [μg/L] Hg [μg/L] Mn [μg/L]

Golema 1 Golema 2 Golema 3 Golema 4 Golema 5 Golema 6 Cheshinska Kurbinska Istocka 1 Istocka 2 Kranska

*phosphorus* compounds are also recorded in all major water bodies, except Kurbinska River which has been found not to be subject to intensive human impacts regarding nutrients.

Fig. 14. Detected nutrients in delineated rivers in Prespa Lake watershed.

Conducted surveillance monitoring is far from sufficient to enable reliable quantification of the total nutrient pressure in the Prespa Lake system. Nevertheless, the obtained estimations of cumulative nutrient loads (Fig. 16) are in concordance with the estimations done for the diffuse source of pollution coming from agriculture. Furthermore, these results disprove the hitherto reported estimations (Grupce, 1997; Naumovski et al., 1997) of the phosphorus load in Prespa Lake of 84 tons per year (of which 41 tons per year are coming from natural processes and 43 tons per year are due to anthropogenic activities) by almost double amount (Fig. 16), without including all of the possible emissions (diffuse sources, natural run-off, etc.).

Fig. 15. Estimated total nutrient load detected in investigated rivers of Prespa Lake watershed.

Fig. 16. Estimation of the cumulative load of nutrients coming from all river water bodies into Prespa Lake.

#### **Heavy metals**

88 Studies on Environmental and Applied Geomorphology

*phosphorus* compounds are also recorded in all major water bodies, except Kurbinska River which has been found not to be subject to intensive human impacts regarding nutrients.

Fig. 14. Detected nutrients in delineated rivers in Prespa Lake watershed.

**July**

**July**

**M arch**

**Golema 1 Golema 2 Golema 3 Golema 4 Golema 5 Golema 6 Cheshinska Kurbinska Istocka 1 Istocka 2 Kranska 1 Kranska 2 Brajcinska 1 Brajcinska 2**

**M arch** **A pril** **M ay** **M arch** **M arch** **A pril** **M ay** **M arch** **M arch** **A pril** **M ay**

PO4 -P(mg/l) Sulphates (mg/l) P2O5-P(mg/l) Total N (mg/l) NH3-N (mg/l) NO2-N (mg/l) NO3-N (mg/l)

t/year

**NO2-N (mg/l) NH3-N (mg/l) P2O5-P (mg/l) PO4 -P (mg/l) NO3-N (mg/l) Total N (mg/l) Sulphates (mg/l)**

**M ay**

Conducted surveillance monitoring is far from sufficient to enable reliable quantification of the total nutrient pressure in the Prespa Lake system. Nevertheless, the obtained estimations of cumulative nutrient loads (Fig. 16) are in concordance with the estimations done for the diffuse source of pollution coming from agriculture. Furthermore, these results disprove the hitherto reported estimations (Grupce, 1997; Naumovski et al., 1997) of the phosphorus load in Prespa Lake of 84 tons per year (of which 41 tons per year are coming from natural processes and 43 tons per year are due to anthropogenic activities) by almost double amount (Fig. 16),

LOAD - NUTRIENTS

without including all of the possible emissions (diffuse sources, natural run-off, etc.).

Fig. 15. Estimated total nutrient load detected in investigated rivers of Prespa Lake

0 500 1000 1500 2000 2500

watershed.

Istočka Reka

Golema Reka

Kurbinska Reka

Kranska Reka

Brajčinska Reka

0 0.5 1 1.5 2 2.5 3 3.5

> **M arch**

**A pril** **July**

**July**

**July**

**July**

**M arch** **A pril**

> Three heavy metals dominate the river water bodies of Prespa Lake watershed: *manganese, iron* and *aluminium*. Their recorded concentrations are usually way beyond the permissible levels for natural conditions (III-IV water quality class) and therefore denote an intensive human origin (Fig. 17). Again, Golema and Istočka Rivers express the highest concentration levels, while the lower segment of Kranska River recorded very high *manganese* and slightly lower *iron* concentrations. Upper segments of all river water bodies (the reference conditions) have been found with heavy metals concentrations several times lower than the lower parts, thus reflecting the natural background emissions to the water habitats.

Fig. 17. Detected concentrations of heavy metals in river water bodies of Prespa Lake watershed.

Environmental Changes in Lakes Catchments

4605.7

3508.4

Prespa Lake watershed.

*Phthalates* and *Organochlorine pesticides*.

also have to be included in future.

**Priority substances** 

563.0

56.5 132.0 1538.7

As

Al

Cd

Cr

Cu

Fe

Hg

Mn

Ni

Pb

Zn

117.9

545.5 504.3

as a Trigger for Rapid Eutrophication – A Prespa Lake Case Study 91

Total heavy metal load from river water bodies in Prespa Lake watershed

Fig. 19. Estimation of the overall heavy metal load originating from the river water bodies in

0.0 5000.0 10000.0 15000.0 20000.0 25000.0 30000.0

25887.5

27192.2

kg/year

The presented results for WFD priority substances are the first record of this kind for the Prespa Lake watershed (Fig. 20). Out of the proposed priority substances for the surveillance and operational monitoring purposes (WFD 2008), the comprehensive analyses performed so far in Prespa Lake watershed embrace *Chlorinated aromatic hydrocarbons, Poly-aromatic hydrocarbons (PAHs), Poly-chlorinated biphenyls (PCBs), Organophosphate pesticides, Phenols,* 

A total of 18 priority substances (out of the whole set of 80 analysed chemical compounds) were detected in river water bodies in Prespa Lake watershed (Fig. 16). *Bis(2- Ethylhexyl)phthalate* was present in almost all samples, the highest values recorded in Golema and Brajčinska Rivers. *Dibutilphthalate* was also found in all river water bodies, except Kurbinska River, but in slightly lower concentrations. O*rganochlorine pesticides* were recorded in different concentrations and were dominant in investigated water bodies. *Gama-HCH (Lindan)*, *Alpha HCH* and *Alpha Endosulfan* were the most commonly present in all water bodies, but the very high values for *Heptachlor* in Golema River 6 and especially in Kranska River. In summary, it is very clear that all of the examined river water bodies in Prespa Lake watershed are under prolonged influence of a significant pollution pressure coming from excess utilisation of various pesticides groups, plastic materials and industry. Even the uphill mountain river parts not subjected to any significant (visible) human impact, that should be used as reference conditions in the watershed, are also under obvious pressure; DDE and DDD although banned from utilisation are still present in Golema 1, Kranska 1 and Brajčinska 2 water bodies. These results point out that the surface water bodies in Prespa Lake watershed have been, and still are, subjected to intensive pressure coming from agriculture and irregular waste disposal. Apart from an *in situ* pollution of the surface waters, aero deposition might play a significant role especially in polluting the upstream mountain river segments with pesticides or their residues. Although in domestic legislation only *Heptachlor* and *PCB* are listed in water categorisation tables (both found in concentrations stated for V or III-IV category respectively), other detected substances will

Two other heavy metals, *copper* and *zinc*, have also been detected in significant quantities in the lower river parts of the Prespa Lake watershed, being most pronounced in the Istočka River. Their increased presence is the inevitable result of adverse human impact in the region.

The most severely influenced river water bodies, Golema and Istočka Rivers, are also characterised with a marked presence of *mercury, lead* and *arsenic*. Their toxicity and harmful effects on the environment and humans are well known and do not need any further elaboration.

Figure 18 gives an overview of the total intensity of the heavy metal load detected in the mouth waters of river water bodies prior to entering the Macro Prespa Lake. It is very clear that Golema and Istočka Rivers are the major source for all measured heavy metals, carrying more than 8 tons of *iron* and *aluminium* annually into the lake, but also significant amounts of *manganese, zinc, copper* and *lead* or *mercury*. Rivers Brajčinska and Kranska also add to the overall heavy metal load (Fig. 19), but to a much lesser extent.

Fig. 18. Estimation of heavy metal load from river water bodies in the Prespa Lake watershed.

In summary, according to calculated estimations, Macro Prespa Lake receives more than 27 tons of *iron* and almost 26 tons of *aluminium* per year, coming from its major tributaries. It is also loaded with 4.6 tons of *manganese*, 3.5 tons of *zinc* and more than 1.5 tons of *copper* per year. Toxic metals are less abundant (563, 504, 132 and 118 kg per year for *arsenic, lead, chromium* and *mercury* respectively), but they do represent a significant load and a dangerous hazard to water biota (through processes of bioaccumulation) and humans. At this point, we do not have any information on the intensity of accumulation of the various heavy metals in water biota, specifically in fish; therefore these investigations will be proven invaluable for the proposed WFD monitoring system to be established in Prespa Lake watershed.

Fig. 19. Estimation of the overall heavy metal load originating from the river water bodies in Prespa Lake watershed.

#### **Priority substances**

90 Studies on Environmental and Applied Geomorphology

Two other heavy metals, *copper* and *zinc*, have also been detected in significant quantities in the lower river parts of the Prespa Lake watershed, being most pronounced in the Istočka River. Their increased presence is the inevitable result of adverse human impact in the

The most severely influenced river water bodies, Golema and Istočka Rivers, are also characterised with a marked presence of *mercury, lead* and *arsenic*. Their toxicity and harmful effects on the environment and humans are well known and do not need any further

Figure 18 gives an overview of the total intensity of the heavy metal load detected in the mouth waters of river water bodies prior to entering the Macro Prespa Lake. It is very clear that Golema and Istočka Rivers are the major source for all measured heavy metals, carrying more than 8 tons of *iron* and *aluminium* annually into the lake, but also significant amounts of *manganese, zinc, copper* and *lead* or *mercury*. Rivers Brajčinska and Kranska also add to the

Fig. 18. Estimation of heavy metal load from river water bodies in the Prespa Lake

In summary, according to calculated estimations, Macro Prespa Lake receives more than 27 tons of *iron* and almost 26 tons of *aluminium* per year, coming from its major tributaries. It is also loaded with 4.6 tons of *manganese*, 3.5 tons of *zinc* and more than 1.5 tons of *copper* per year. Toxic metals are less abundant (563, 504, 132 and 118 kg per year for *arsenic, lead, chromium* and *mercury* respectively), but they do represent a significant load and a dangerous hazard to water biota (through processes of bioaccumulation) and humans. At this point, we do not have any information on the intensity of accumulation of the various heavy metals in water biota, specifically in fish; therefore these investigations will be proven invaluable for the proposed WFD monitoring system to be established in Prespa Lake

overall heavy metal load (Fig. 19), but to a much lesser extent.

region.

elaboration.

watershed.

watershed.

The presented results for WFD priority substances are the first record of this kind for the Prespa Lake watershed (Fig. 20). Out of the proposed priority substances for the surveillance and operational monitoring purposes (WFD 2008), the comprehensive analyses performed so far in Prespa Lake watershed embrace *Chlorinated aromatic hydrocarbons, Poly-aromatic hydrocarbons (PAHs), Poly-chlorinated biphenyls (PCBs), Organophosphate pesticides, Phenols, Phthalates* and *Organochlorine pesticides*.

A total of 18 priority substances (out of the whole set of 80 analysed chemical compounds) were detected in river water bodies in Prespa Lake watershed (Fig. 16). *Bis(2- Ethylhexyl)phthalate* was present in almost all samples, the highest values recorded in Golema and Brajčinska Rivers. *Dibutilphthalate* was also found in all river water bodies, except Kurbinska River, but in slightly lower concentrations. O*rganochlorine pesticides* were recorded in different concentrations and were dominant in investigated water bodies. *Gama-HCH (Lindan)*, *Alpha HCH* and *Alpha Endosulfan* were the most commonly present in all water bodies, but the very high values for *Heptachlor* in Golema River 6 and especially in Kranska River. In summary, it is very clear that all of the examined river water bodies in Prespa Lake watershed are under prolonged influence of a significant pollution pressure coming from excess utilisation of various pesticides groups, plastic materials and industry. Even the uphill mountain river parts not subjected to any significant (visible) human impact, that should be used as reference conditions in the watershed, are also under obvious pressure; DDE and DDD although banned from utilisation are still present in Golema 1, Kranska 1 and Brajčinska 2 water bodies. These results point out that the surface water bodies in Prespa Lake watershed have been, and still are, subjected to intensive pressure coming from agriculture and irregular waste disposal. Apart from an *in situ* pollution of the surface waters, aero deposition might play a significant role especially in polluting the upstream mountain river segments with pesticides or their residues. Although in domestic legislation only *Heptachlor* and *PCB* are listed in water categorisation tables (both found in concentrations stated for V or III-IV category respectively), other detected substances will also have to be included in future.

Environmental Changes in Lakes Catchments

watershed (in ppm).

watershed (in ppb).

Dibutilphthalate

Bis(2-Ethylhexyl) phthalate

PCB-52

as a Trigger for Rapid Eutrophication – A Prespa Lake Case Study 93

Fig. 21. Priority substances load to Prespa Lake coming from river water bodies in the

Fig. 22. Priority substances load to Prespa Lake coming from river water bodies in the

**Total Load of Priority Substances from rivers (in ppm)**

0 50 100 150 200 250 300 350 400

kg/year

Fig. 20. Detected priority substances (without heavy metals) in river water bodies of Prespa Lake watershed.

*The final conclusion on detected priority substances* in the river water bodies in Prespa Lake watershed is that numerous different compounds were detected, some of them with very high concentrations (III-IV or V water quality class) and that all of them represent an elevated risk for the environment, water biota and humans. Toxic and already forbidden chemicals like *DDD* or *DDE* are still present in the waters what underlines their constant utilisation. Apart from detected pesticides or their derivatives, detected *phtalathes* as main chemicals in the production of plastics (Fig. 21) directly emphasise irregular solid waste treatment or disposal, and insufficient industrial and domestic wastewater treatment.

It is very critical to note that the spreading of detected priority substances also affects the remote regions not subject to any visible human influence in the Prespa Lake watershed thus endangering the environment and biota in already declared protected areas, like NP "Pelister" and "Galichica". Their potential for bioaccumulation and prolonged devastating impact should be the focus of the Management Plan and consequent reduction measures.

**Priority substances in river water bodies in ppm**

March March March July July July March July July March March March July March Istočka 1 Istočka 2 Golema 1 Golema 2 Golema 3 Golema 4 Golema 6 Kurbinska Kranska 1 Kranska 2 Brajčinska 1 Brajčinska

Anthracene Benzo (a )anthracene Benzo (a) pyrene Fluorene Alachlor Atrazine Dibutilphthalate Bis(2-Ethylhexyl)phthalate

**Priority substances in river water bodies in ppb** 

PCB (III-IV class) 1-10 ng/l Heptachlor (V class) > 1 ng/l

March March March July July July March July July March March March July March Istočka 1 Istočka 2 Golema 1 Golema 2 Golema 3 Golema 4 Golema 6 Kurbinska Kranska 1 Kranska 2 Brajčinska 1 Brajčinska

> PCB-52 PCB-101 PCB-156 PCB-180 trans-Chlordane 2,4'-DDD 4,4'-DDD 2,4'-DDE 4,4'-DDE alpha-Endosulfan alpha-HCH beta-HCH gamma-HCH (Lindan) delta-HCH Heptachlor

2

2

Fig. 20. Detected priority substances (without heavy metals) in river water bodies of Prespa

*The final conclusion on detected priority substances* in the river water bodies in Prespa Lake watershed is that numerous different compounds were detected, some of them with very high concentrations (III-IV or V water quality class) and that all of them represent an elevated risk for the environment, water biota and humans. Toxic and already forbidden chemicals like *DDD* or *DDE* are still present in the waters what underlines their constant utilisation. Apart from detected pesticides or their derivatives, detected *phtalathes* as main chemicals in the production of plastics (Fig. 21) directly emphasise irregular solid waste treatment or disposal, and insufficient industrial and domestic wastewater treatment.

It is very critical to note that the spreading of detected priority substances also affects the remote regions not subject to any visible human influence in the Prespa Lake watershed thus endangering the environment and biota in already declared protected areas, like NP "Pelister" and "Galichica". Their potential for bioaccumulation and prolonged devastating impact should be the focus of the Management Plan and consequent

Lake watershed.

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

ng/l

μg/l

reduction measures.

Fig. 21. Priority substances load to Prespa Lake coming from river water bodies in the watershed (in ppm).

Fig. 22. Priority substances load to Prespa Lake coming from river water bodies in the watershed (in ppb).

Environmental Changes in Lakes Catchments

highly inapplicable.

significantly extended.

as a Trigger for Rapid Eutrophication – A Prespa Lake Case Study 95

ii. Using biotic indices developed for any organism in a region (country, continent) other than the region (river, lake, watershed) in question has already been proven erroneous on multiple occasions (Van Dam et al., 2005; Krstic et al., 2007). This is fundamentally based on differences in the *total capacity* of every ecosystem, numerous (not accountable) sources of variation of the ecological parameters and direct human influence. Since all of the indices are derived via assigning specific numbers to organisms (algae in particular) to be used in formulae calculations, the only scientifically sound approach is to develop a specific index for every water body; iii. The last (and not least) reason against using bio-indices other than a basic community similarity (or dissimilarity) index is their gross oversimplification of the biological (or genetic) responses of living biota to different environmental conditions. The same organism (if we exclude taxonomical errors as yet another frequently very misleading source) cannot react in the same way in different environments (habitats); therefore, using and comparing biotic indices based on species variations (numbers in the population or different biological treats) between habitats seems (and has been proven)

With all the stated limitations in focus, class boundaries among different water quality classes and the reference conditions were based on benthic flora and fauna community structure in which dominant taxa characteristics, and therefore different classes, are recorded. In the case of benthic algae, clear distinction among class boundaries are based on the occurrence of taxa indicative for higher eutrophication or saprobity levels, or mass occurrence of specific cyanobacteria indicating very high pollution. Macrozoobenthos communities were subjected to evaluation according to the Danish Stream Fauna Index (DSFI – Skriver et al., 2000) and EPT richness (Bode et al., 1997), but we still express our reservations regarding the above mentioned restrictions for application of the biological indices. To achieve the best possible level of certainty regarding the biological quality elements in Prespa Lake watershed, the surveillance and investigative monitoring must be

In general, algal assemblages in river water bodies of Prespa Lake watershed are quite distinctly separated along the eutrophication or pollution gradient detected in examined watercourses. Namely, the upper river segments of all rivers are dominated by diatom flora characteristics for clean, slightly acidophilic, oligotrophic waters or **good ecological conditions** composed of *Diatoma hyemalis, Diatoma mesodon, Hannea arcus, Meridion circulare, Meridion circulare var. constricta, Eunotia minor, Achnanthidium jackii, Decussata hexagona, Encyonema mesianum, Krsticiella ohridana, Pinnularia eifelana, Pinnularia sudetica, Psammothidium daonense*. The rocky bottom of these river parts is usually not covered by any visible flora, but occasionally fragments of water mosses, typical chrysophyte (golden algae) *Hydrurus foetidus* colonies or the red alga *Lemanea fluviatilis* which are usually deprived of any rich epiphytic diatom flora. Therefore, this algae assemblage is chosen to be the

The second type of algal assemblage is found in lower river stretches prior to or after the major pollution events indicating **moderate ecological conditions**; Golema Reka 2, 3, 4, Češinska, Kurbinska, Kranska 2 and Brajčinska 2. This assemblage is clearly dominated by diatom taxa indicative for higher nutrient concentrations in the water and more intensive

indicator for the *reference conditions of all rivers* in Prespa Lake watershed.

Fig. 23. Estimated total load of priority substances to Prespa Lake coming from river water bodies in the watershed (in ppm and ppb).

#### **Ecological quality elements**

Having in mind that there are practically no useful data (apart from the one project report – Krstic, 2007) on WFD ecological quality elements for the rivers in Prespa Lake watershed, the only feasible and promising approach was to use benthic organisms, algae and macrozoobenthos, to evaluate the ecological conditions in rivers. For the same reasons, there was no firm scientific opportunity to run any statistical calculations on the obtained data without significant bias and/or error.

The principal approach used in this situation was to determine the benthic algal and macrozoobenthic assemblages in upper river segments (river water bodies not subject to any significant human influence) to parts of the rivers where increased human interference in water physical or chemical properties was detected (usually lower river stretches after the point or diffuse wastewater input or leaching coming from cultivated land). Using this approach, we were able to detect the benthic assemblages in natural conditions (or reference conditions) for the rivers in Prespa Lake watershed, and also to reveal to the highest possible extent the succession of benthic communities under the detected human pressure. All of the methods for sampling, handling and analysing the samples were performed according to WFD recommended ISO or CEN standards (ISO, 2003; CEN, 2003).

Regarding the WFD proposal for using biotic indices in the process of describing the ecological status of water bodies, our approach was ruled by the following facts and postulates:

i. The biotic indices in relation to water quality monitoring (for both algae and macrozoobenthos) have been developed in different countries based on long-term monitoring data and detected autecological preferences of different taxa. In the case of Prespa Lake and the water bodies in its watershed, neither of those lines of data are available. Even more so since there were numerous diatoms (a group of algae) described new to science in the region (Levkov et al., 2006) for which there are no data in the literature;

**Total Load of Priority Substances from rivers (in ppb)**

Fig. 23. Estimated total load of priority substances to Prespa Lake coming from river water

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35

kg/year

Having in mind that there are practically no useful data (apart from the one project report – Krstic, 2007) on WFD ecological quality elements for the rivers in Prespa Lake watershed, the only feasible and promising approach was to use benthic organisms, algae and macrozoobenthos, to evaluate the ecological conditions in rivers. For the same reasons, there was no firm scientific opportunity to run any statistical calculations on the obtained data

The principal approach used in this situation was to determine the benthic algal and macrozoobenthic assemblages in upper river segments (river water bodies not subject to any significant human influence) to parts of the rivers where increased human interference in water physical or chemical properties was detected (usually lower river stretches after the point or diffuse wastewater input or leaching coming from cultivated land). Using this approach, we were able to detect the benthic assemblages in natural conditions (or reference conditions) for the rivers in Prespa Lake watershed, and also to reveal to the highest possible extent the succession of benthic communities under the detected human pressure. All of the methods for sampling, handling and analysing the samples were performed

Regarding the WFD proposal for using biotic indices in the process of describing the ecological status of water bodies, our approach was ruled by the following facts and

i. The biotic indices in relation to water quality monitoring (for both algae and macrozoobenthos) have been developed in different countries based on long-term monitoring data and detected autecological preferences of different taxa. In the case of Prespa Lake and the water bodies in its watershed, neither of those lines of data are available. Even more so since there were numerous diatoms (a group of algae) described new to science in the region (Levkov et al., 2006) for which there are no data

according to WFD recommended ISO or CEN standards (ISO, 2003; CEN, 2003).

bodies in the watershed (in ppm and ppb).

2,4'-DDD

4,4'-DDE (TL=2) alpha-Endosulfan

alpha-HCH

gamma-HCH (Lindan) Heptachlor (TL=2)

without significant bias and/or error.

**Ecological quality elements** 

postulates:

in the literature;


With all the stated limitations in focus, class boundaries among different water quality classes and the reference conditions were based on benthic flora and fauna community structure in which dominant taxa characteristics, and therefore different classes, are recorded. In the case of benthic algae, clear distinction among class boundaries are based on the occurrence of taxa indicative for higher eutrophication or saprobity levels, or mass occurrence of specific cyanobacteria indicating very high pollution. Macrozoobenthos communities were subjected to evaluation according to the Danish Stream Fauna Index (DSFI – Skriver et al., 2000) and EPT richness (Bode et al., 1997), but we still express our reservations regarding the above mentioned restrictions for application of the biological indices. To achieve the best possible level of certainty regarding the biological quality elements in Prespa Lake watershed, the surveillance and investigative monitoring must be significantly extended.

In general, algal assemblages in river water bodies of Prespa Lake watershed are quite distinctly separated along the eutrophication or pollution gradient detected in examined watercourses. Namely, the upper river segments of all rivers are dominated by diatom flora characteristics for clean, slightly acidophilic, oligotrophic waters or **good ecological conditions** composed of *Diatoma hyemalis, Diatoma mesodon, Hannea arcus, Meridion circulare, Meridion circulare var. constricta, Eunotia minor, Achnanthidium jackii, Decussata hexagona, Encyonema mesianum, Krsticiella ohridana, Pinnularia eifelana, Pinnularia sudetica, Psammothidium daonense*. The rocky bottom of these river parts is usually not covered by any visible flora, but occasionally fragments of water mosses, typical chrysophyte (golden algae) *Hydrurus foetidus* colonies or the red alga *Lemanea fluviatilis* which are usually deprived of any rich epiphytic diatom flora. Therefore, this algae assemblage is chosen to be the indicator for the *reference conditions of all rivers* in Prespa Lake watershed.

The second type of algal assemblage is found in lower river stretches prior to or after the major pollution events indicating **moderate ecological conditions**; Golema Reka 2, 3, 4, Češinska, Kurbinska, Kranska 2 and Brajčinska 2. This assemblage is clearly dominated by diatom taxa indicative for higher nutrient concentrations in the water and more intensive

Environmental Changes in Lakes Catchments

community, and used to derive the EPT and DSFI indices.

Table 7. Water classification based on EPT and DSFI.

profundal) for the macrozoobenthos analyses.

regarding the total phosphorus content and diatom composition.

collected during June 2010 and according to WFD sampling guides.

**5.2 Prespa Lake** 

**Basic physical parameters** 

nutrient supply of the epilimnion layer.

as a Trigger for Rapid Eutrophication – A Prespa Lake Case Study 97

*Brachyptera risi* and *Potamophylax latipennis, Crenobia alpina* and even crayfish *Austropotamobius torrentium* were compared for the percentage abundance within the

**>10 7 high (reference condition)** 

Out of the WFD's biology quality elements, phytoplankton, zoobenthos, macrophytes and fish were in the focus of investigations, supported by a full range of physic-chemical analyses including heavy metals and priority substances. Water for the chemical analyses was sampled as a collective sample from the full water column on the site or as sediment, while the basic physical parameters were measured at every sample depth. Special attention was paid to the sampling of plankton for algae and benthic habitats (littoral, sub-littoral and

In order to determine the *reference conditions* for Prespa Lake, analyses of core samples dated 10 ka (500, 1,000, 2,000, 5,000 and 10,000 years respectively; the deepest analysed core sample from approximately 30 metres of the core depth) were performed for the first time

Macrophytes and fish samples from the selected sampling sites on Prespa Lake were

The basic physical parameters detected in the waters of Prespa Lake (Fig.24) revealed some interesting features of this unique ecosystem. For example, recorded temperatures show a normal and gradual increase towards warmer months, but there was no sharp and rapid decrease in one water layer (thermocline) during the warmest month (July 2010) although the water temperatures between the deepest and the shallowest parts differ by more than 10ºC. This may be a result of a very turbulent climate in the sampling period with constant mixing of the water layers or as a consequence of intensive discharge of the deep water sources (sub-lacustrine water sources) again related to the rainy season. High deep water temperatures of 14-15ºC also indicate the full intensity of thermal insulation and possible full scale mixing of the entire water column during storms, which results in a constant

**1-3 bad** 

EPT richness **DSFI value Water quality** 

**6-10 6 good** 

**2-5 5 moderate** 

**<2 4 poor** 

decomposition processes, such as: *Melosira varians, Fragilaria capucina, Ulnaria ulna, Achnanthidium lanceolatum, Cocconeis placentula* var*. euglypta, Navicula phyllepta, Navicula cryptotenella, Navicula halophila, Navicula lanceolata, Navicula tripunctata, Navicula cryptocephala, Frustulia vulgaris, Reimeria sinuata, Gomphonema olivaceum, Gomphonema angustatum, Gomphonema micropus, Gomphonema aff. olivaceoides, Encyonema silesiacum, Amphora pediculus, Nitzschia linearis, Nitzschia palea, Surirella minuta*. Usually, the green branched alga *Cladophora glomerata* is also markedly present in these water bodies, bearing a rich epiphytic growth usually by *Cocconeis placentula* var*. euglypta*, but sometimes blue-green cyanobacteria *Heteroleiblenia kossinskajae* or *Pseudoanabaena limnetica* were observed as significant epiphytes.

Finally, the last detected algal assemblage was found in the most severely polluted river bodies of Prespa Lake watershed, like Golema Reka 5, 6, 7, and Istočka 2, and thus denoting **bad or poor ecological conditions**. The typical example of this algal assemblage was found at Golema Reka 5 sampling site where the diatom flora is reduced to as much as only 3 - 4 taxa, like *Nitzschia palea, Navicula cryptotenella* and *Ulnaria ulna* representing more than 98% of all detected cells. Highly decreased algal biodiversity is replaced by a mass development of two cyanobacterial species *Pseudoanabaena limnetica* and *Phormidium limosum* which completely cover the rocks on the bottom.

The WFD requires classification, in terms of ecological status, for all European surface waters. The classification should be based on reference conditions, which are intended to represent minimal anthropogenic impact and observed deviation from these conditions (Andersen et al., 2004). For each surface water body type, type-specific biological reference conditions were established, representing the values of the biological quality elements for that surface water body type at high ecological status.

Among the biological communities, the macrozoobenthos is by far the most frequently used bioindicator group in standard water management (Hering et al., 2004). Numerous biotic index and score systems have used macrozoobenthos in the assessment of running waters (Rosenberg & Resh, 1993). The most represented biotic index or score methods are: taxa richness, number of EPT taxa, Saprobic Index (SI), Biological Monitoring Working Party (BMWP) Score, Average Score Per Taxon (ASPT), Danish Stream Fauna Index (DSFI). All indices were part of the respective national method planned for biological monitoring in the context of the Water Framework Directive (Birk & Hering, 2006).

Thus, in the frame of the project "Development of Prespa Lake Watershed Management Plan" and according to WFD requirements, categorisation of the delineated water bodies in the Prespa Lake watershed based on macrozoobenthos was done. Two metrics (EPT richness and DSFI) in assessment of the ecological health of the rivers were used. These metrics were selected because statistical power to detect a difference between the nutrient enriched and non-impacted sites was >0.99 for total taxa richness and number of EPT taxa. DSFI also had relatively high power >0.95 (Sandin & Johnson, 2000) (Table 7).

In order to assess the ecological conditions of the river water bodies and related macrozoobenthos assemblages, the analyses of collected samples were performed to detect the presence of so called *positive* (like Ephemeroptera, Plecoptera, Trichoptera, Diptera, Gammaridae and even Astacidae) versus *negative* taxa (usually Oligochaeta – Chironomidae or Tubificidae). Pollution sensitive taxa like *Ecdyonurus venosus*, *Baetis alpinus*, *Capnia vidua*, *Brachyptera risi* and *Potamophylax latipennis, Crenobia alpina* and even crayfish *Austropotamobius torrentium* were compared for the percentage abundance within the community, and used to derive the EPT and DSFI indices.


Table 7. Water classification based on EPT and DSFI.
