Hydrology and Nutrient Enrichment at Two Coastal Lagoon Systems in ...

Water Resources Management 16: 171–196, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.

171

Hydrology and Nutrient Enrichment at Two Coastal Lagoon Systems in Northern Greece GEORGIOS SYLAIOS1∗ and VASSILIKI THEOCHARIS2

1 National Agricultural Research Foundation, Fisheries Research Institute, 640 07 Nea Peramos, Kavala, Greece; 2 National Agricultural Research Foundation, Bureau of Applied Hydroecology, P.O. Box 1103, 45110 Ioannina, Greece (∗ author for correspondence, e-mail: [email protected].)

(Received: 4 December 2000; accepted: 3 April 2002) Abstract. The lagoons of N.E. Greece, located on the western side of Nestos River, and of N.W. Greece, located at the lower reaches of Kalamas River, are among the most important shallow, semi-enclosed ecosystems in Northern Greece. The temporal variability of nutrients at both lagoonal systems shows the strong influence of fresh water discharge on water quality. Nutrient enrichment factors showed that nitrites and ammonium were six times higher at the lagoons of N.W. Greece than those observed at N.E. Greece, while phosphates were forty times higher at Nestos River lagoons. The flushing half-life was calculated based on a combination of hydrological and tidal processes, for each lagoon of these two systems, allowing for the assessment of water quality changes. Proper management measures for both systems should focus on the control of fresh water quality entering the lagoons, the reduction of phosphoric fertilizers used by agriculture and the better oxygenation of the water column. One way to eliminate massive fish deaths during the winter in N.E. Greece is also the transfer of fresh, warm groundwater, while bathymetric modifications and channel widening are needed at the lagoons of N.W. Greece. Key words: ecosystem improvement, fish production, Kalamas River, lagoon, Nestos River, nutrient enrichment, water quality parameters

1. Introduction Coastal lagoons have been considered as complex natural ecosystems being in a delicate balance, from which they easily deviate, mainly due to environmental degradation, caused by alterations in hydrology, pollution and human activities (Colombo, 1977; Mee, 1978; Sikora and Kjerfve, 1985; Miller et al., 1990). According to the local dominating climatic and hydrographic conditions, coastal lagoons exhibit salinities that range from completely fresh to hypersaline (Kjerfve, 1986; Kjerfve and Magill, 1989; Knoppers et al., 1991). Physical processes in these coastal ecosystems include meteorological influence, tidal effect, wave action and spatial and temporal variations in water temperature and salinity, serving as the driving force for water and nutrient exchanges with the adjacent coastal sea (Dyer, 1973).

172

G. SYLAIOS AND V. THEOCHARIS

Coastal lagoons are providing significant food resources, since fishing and aquaculture have constituted one of the oldest forms of coastal resource exploitation (Vallejo, 1982). The importance of coastal lagoons for fisheries and aquaculture at the global level, and for Greece in particular, directs research towards the study of the relations between the various environmental factors and biological productivity of these ecosystems (Theocharis et al., 2000). Environmental parameters (physical, chemical and biological) may directly influence fish production dynamics of the ecosystem (Corsi and Ardizzone, 1985). Integrated management approach in coastal lagoons is essential for the study of their environmental characteristics, aiming at understanding the patterns and functioning of the system (OECD, 1992). Nutrient availability, in particular, plays an important role in the functioning of lagoon communities. The availability of nitrogen and phosphorus in the lagoon, may affect plant and algal production, biomass, tissue nitrogen and phosphorus content. Potential sources of nitrogen and phosphorus to lagoons include nitrogen fixation and phosphorus remineralization within the lagoon, tidal flows, rainfall, streamflow and groundwater (Nixon, 1982). The latter two sources may be dramatically affected by agricultural and urban development in the adjoining watershed and carry significant loadings of nutrients to coastal waters and lagoons (Correll et al., 1992; Valiela et al., 1992). Comprehensive studies on the lagoons developed by the sediment load of River Nestos (N.E. Greece) and River Kalamas (N.W. Greece) have not be carried out so far, in terms of environmental factors or fish production management. The general scope of this paper is to present a comparative approach on the hydrographic and water quality characteristics of two coastal lagoon ecosystems in Northern Greece (the lagoons of N.E. Greece at the delta region of Nestos River and the lagoons of N.W. Greece at the delta region of Kalamas River). The detailed objectives of the present work are: (1) to report on the seasonal fluctuation and scaling of current meteorological and hydrological patterns that prevail at the small, shallow, fishery-developed lagoons Vassova and Eratino at N.E. Greece; (2) to describe their eutrophic character and the degree of differentiation from the lagoons at N.W. Greece and (3) to evaluate the terms of water balance and estimate the flushing time for each lagoon basin.

2. Study Sites The spatial and temporal variability of the environmental characteristics of eight in total lagoons, belonging to two different ecosystems, in northeast and northwest Greece, were examined. All lagoons are small in size (areas range between 0.7– 3.5 km2 ), shallow (mean depths ranges between 0.7–3.5 m) and forced by similar tidal influence (mean tidal range 0.20 m) at their mouths. However, they belong to different climatic zones, e.g., evaporation exceeds precipitation in N.E. Greece and the reverse occurs in N.W. Greece. The geometric, hydrologic and climatic parameters for lagoons Vassova and Eratino (N.E. Greece) and lagoons Loutsa-

173

HYDROLOGY AND NUTRIENT ENRICHMENT AT TWO COASTAL LAGOON SYSTEMS

Table I. Geometrical, climatical and hydrological characteristics for the lagoons of N.E. and N.W. Greece Lagoon

Northing

Easting

A (km2 )

h (m)

V (km3 )

h (m)

P (mm)

E (mm)

N.E. Greece Vassova Eratino

40◦ 57 N 40◦ 55 N

24◦ 34 E 24◦ 35 E

2.7 3.5

1.0 1.0

0.003 0.004

0.20 0.20

490 490

990 990

39◦ 33 N

20◦ 09 E

2.0

1.0

0.002

0.20

1100

250

39◦ 32 N 39◦ 32 N 39◦ 31 N 39◦ 34 N 39◦ 30 N

20◦ 11 E 20◦ 09 E 20◦ 13 E 20◦ 09 E 20◦ 16 E

0.7 0.8 1.0 3.0 2.5

0.8 1.2 0.6 0.8 1.5

0.001 0.001 0.001 0.002 0.004

0.20 0.20 0.20 0.20 0.20

1100 1100 1100 1100 1100

250 250 250 250 250

N.W. Greece Loutsa Papadia Kalaga Vatatsa Richo Alykes Vontas

where A = lagoon surface area; h = mean lagoon water depth; V = lagoon water volume; h = mean lagoon tidal range; P = annual precipitation; E = annual evaporation.

Papadia, Kalaga, Vatatsa, Richo, Alykes and Vontas (N.W. Greece) are summarized at Table I.

2.1. SITE 1 : LAGOONS OF N . E . GREECE A series of nine choked, shallow and elongated coastal lagoons are located west of Nestos River Delta (N.E. Greece), having a total area of 1,700 ha. Vassova and Eratino are very important ecosystems in terms of fish production (Figure 1). Vassova lagoon has a mean depth of 1 m and a maximum depth of 3 m at the wintering canals. The lagoon consists of a central main basin for extensive aquaculture, and 13 artificially made wintering canals, 50 m long and 0.5 m deep each. A shallow and narrow channel, 30 m wide and 0.8 m deep, connects the lagoon to the open sea (North Aegean Sea). A number of small entrances at the northern part connect the lagoon with the adjacent surrounding drainage channels, considered as the only irregular source of fresh water entering the ecosystem. Eratino lagoon has a length of 6 km, a maximum width of 1.5 km, a perimeter of 17 km and a mean depth of 1.0 m at the main basin. Wintering canals have a depth of 3 m, located at the eastern side of the lagoon entrance. Water exchange between the lagoon and the adjacent sea exists through a narrow mouth, 40 m wide and 3 m deep. The northern part of the basin is connected to a natural, shallow, 2 km long channel, part of the old deltaic tributary of Nestos River, which supplies the lagoon with significant quantities of fresh water.

174


Figure 1. Sampling stations at the lagoons of Nestos River (N.E. Greece).


175

Nestos River has a discharge that ranges from 8 to 106 m3 s−1 , with an annual mean value of 57 m3 s−1 (Therianos, 1974). Annual precipitation in the area for years 1995–1997 ranged between 449.8 and 530.2 mm, according to data recorded by the Hydrological Station of Chrisoupolis, Ministry of Agriculture. The same station recorded annual evaporation of the area for the years 1995–1997 ranging between 968.2 to 1012.4 mm. 2.2. SITE 2: LAGOONS OF N . W. GREECE The lagoons of N.W. Greece, located 405 km away from the lagoons of N.E. Greece, cover a total area of 140 ha with a mean depth of 1.5 m (Figure 2). Lagoons Loutsa-Papadia and Kalaga were formed from the sediment deposition of Kalamas River. Lagoons Vatatsa and Richo developed a sand barrier as a result of the combined action of waves and currents. Alykes and Vontas are newly developed lagoons from the sediment deposition of Kalamas River. Water exchange between these lagoons and the open Ionion Sea is limited due to the shallow and narrow channels that exist in each lagoon. The water of Kalamas River, flowing through a network of irrigation and drainage canals, supplies these lagoons with fresh water. Kalamas River has a mean annual discharge of 15.5 m3 s−1 (Xanthopoulos, 1985). These two lagoon systems appear to be similar in size, shape, depth, origin, tidal regime and fresh water input, but they belong to different climatic zones and are affected by different human activities and degrees of exploitation. A comparative approach in terms of the seasonal patterns of their physical and chemical aquatic parameters may be helpful in proposing management measures for ecosystem restoration. 3. Materials and Methods Meteorological and hydrological influences mainly determine the distribution of environmental parameters in time and space within these lagoons. Data sets of physical and chemical (nutrients) aquatic parameters were collected from surface water of the above lagoons at representative stations during various sampling periods (1993–94 for the lagoons of N.W. Greece and 1996–97 for the lagoons of N.E. Greece). Although simultaneous data from both lagoon ecosystems would be more preferable for this analysis, such data were not available. However, the data used in this paper were collected during periods representative for each site, taking into consideration the prevailing climatic factors. For the lagoons of N.E. Greece, stations B1, B2 and E1 were located at the main basins of these two lagoons, stations B3 and E2-E5 at the surrounding drainage canals and stations B4, E6 at the adjacent coastal area (Figure 1). For the lagoon system of N.W. Greece, only one station was positioned at the main basin of each small lagoon (Alykes, Bastia, Loutsa-Papadia, Vatatsa, Richo and Igoumenitsa harbor), while two stations were located at the larger lagoons (Vontas and Kalaga).

176


Figure 2. Sampling stations at the lagoons of Kalamas River (N.W. Greece).

Station K1 at the lower reaches of Kalamas River, represents the fresh water flowing into the lagoons through a series of drainage canals, and stations S1, S2 and S3 the seawater of the Ionion Sea (Figure 2). Data sampling of water temperature, salinity (as computed from electrical conductivity), pH, dissolved oxygen (always measured during the day period), nitrates, nitrites, phosphates, ammonium, wet and dry biomass were always taking place on


177

the day of maximum tidal range (spring tide), during the entrance of saline water into the lagoons. This sampling strategy eliminates the effect of tidal circulation on the temporal variability of physical and chemical parameters of the system (Kjerfve, 1986). Water temperature was measured using a SYLAND probe (accuracy: ± 0.2 ◦ C), water electrical conductivity with a HANNA HI8733 probe (accuracy: ± 1 psu), dissolved oxygen by a galvanised SYLAND electrode (accuracy: ± 1%) and pH by a HANNA type ion-determining electrode (accuracy: ± 0.01). Salinity was determined using temperature and electrical conductivity values based on the equation of state and UNESCO tables (UNESCO, 1981). Nutrient concentrations were determined by spectrophotometric methods according to Strickland and Parsons, (1972). Principal component analysis (PCA) uses linear combinations of data to provide a simplified summary based on two or three linear combinations in order to reduce the number of variables of the system and to detect the system structure by inter-relating variables (Dunteman, 1989). The mathematical procedure for principal component derivation involves the use of least-squares approximation for the computation of eigen-values from an initial vector (n x k) with n the number of measurements at each location and k the number of variables measured. It occurs from this analysis that the sum of eigen-values is equal to the sum of the diagonal elements of the correlation matrix, as: λj = si,j , where λj is the jth eigen-value and si,j the diagonal elements of the correlation matrix. These results are arranged along principal axes (in numerical order 1,2,3 etc.) according to their ability to explain the variability of the system. In this study, PCA was used in order to reveal the mechanisms responsible for water quality variability of each system and the degree of their differentiation (Linn et al., 1975). Furthermore, in order to estimate the amount of total nutrients held in dissolved form in each lagoon, the monthly values of nutrient concentrations were multiplied by the appropriate lagoon volumes. These results were divided by the nutrient background concentrations of a reference North Aegean station, in order to obtain relative nutrient enrichment factors and to allow inter-comparisons between the two systems (Friligos, 1988). Water fluxes and flushing half-life for each lagoon were estimated following simple tidal prism models (Officer, 1976; Luketina, 1998; Kjerfve et al., 1996). 4. Results and Discussion 4.1. SEASONAL PATTERNS IN THE LAGOONS OF N . E . GREECE Table II presents the mean annual values of the water parameters sampled at the two lagoon systems. Figure 3 presents the mean daily wind variability, atmospheric pressure and air temperature, as recorded by the National Meteorological Service at Chrisoupolis Airport, near the lagoons of N.E. Greece, during the period 1993– 1997. In the same figure, the variation of tidal elevation during the same period, as recorded by the National Hydrographic Service at Kavala harbor, is also presented.

178


Table II. Mean annual values of sampled water parameters at the lagoon systems of N.E. and N.W. Greece Sampling Location

N.E. Greece Vassova Eratino Fresh-water discharge Sea-water N.W. Greece Alykes Vontas LoutsaPapadia Kalaga Vatatsa Richo River Kalamas Sea-water

Temperature (◦ C)

Salinity Diss. pH (psu) Oxygen (mg L−1 )

N-NO3 (µM)

N-NO2 (µM)

N-NO4 (µM)

P-PO4 (µM)

16.63 17.04 16.38

32.26 30.42 5.44

5.73 7.32 5.95

8.06 8.30 7.71

0.29 0.30 1.96

0.01 0.01 0.10

0.02 0.02 0.04

0.24 0.20 0.27

17.68

34.20

8.94

8.14

0.21

0.03

0.017

0.20

19.80 19.58 20.03

39.74 31.90 32.80

8.48 8.08 9.78

8.26 8.13 8.06

0.02 0.02 0.05

0.007 0.002 0.007

0.008 0.005 0.030

0.006 0.002 0.007

20.68 19.78 20.08 16.48

39.80 32.42 30.46 24.74

7.80 7.64 7.73 7.74

8.31 8.16 8.26 7.98

0.07 0.04 0.07 1.66

0.013 0.002 0.015 0.26

0.026 0.001 0.018 0.64

0.007 0.009 0.007 0.009

19.37

37.57

9.35

8.11

0.009

0.003

0.001

0.002

It occurs that the most common wind magnitude is 2.5–3.0 m s−1 , showing a seasonal variability in direction. Easterly and southwesterly wind directions are dominant during the winter period having frequencies of 7 and 10%, respectively. Wind directions change to south (4.7%) and south-west (8.3%) during the summer period. Strong winds (3.0–3.5 m s−1 ) in combination with increased tidal range during spring tides are responsible for the low temperatures and turbulent conditions that prevail at these lagoons during the winter. During summer, the dominating south-southwesterly winds moderate the usually warm conditions. The duration of spring and autumn is usually limited. Seawater obtains a residual flow towards the lagoons during autumn (October) and offshore during spring (April) under the influence of the wind shear stress. The seasonal patterns of water temperature, salinity, dissolved oxygen and pH at the lagoons Vassova and Eratino, the adjacent coastal sea and the surrounding drainage canals are shown at Figure 4. Low water temperature is observed during December and January (2.8 ◦ C), resulting in massive fish deaths (10–15% of total fish stock) at the wintering canals of the lagoon. Low winter temperature inhibits


179

Figure 3. Mean daily variability of meteorological and tidal parameters at the broader area of Nestos Lagoons (N.E. Greece).

fish growth, while the influence of strong winds flowing over these shallow lagoons increases turbidity, also causing fish deaths (Barnabe, 1992). Higher values occur during the summer months from June to September (27.8 ◦ C) at the coastal zone of the lagoons, where the smaller annual variability is observed (stand. dev. = 5.82 ◦ C). An important temperature gradient exists at all seasons between the coastal zone and the lagoons, with the sea being warmer during the winter months by 2.7 ◦ C, and the lagoons during the summer period by 3.1 ◦ C. This temperature gradient plays an important role in the physiology, the biological cycle and the movements of lagoonal organisms (Vernberg, 1982). Salinity at Vassova lagoon reaches its maximum value during December (36.9 psu), while reduces to 18.2 psu during August. Similarly, maximum salinity occurs at Eratino during December (38.2 psu) and the minimum during September (23.8 psu). Vassova lagoon retains a high mean annual salinity (S = 32.3 psu) due to diminished fresh water inflows, while Eratino lagoon exhibits higher temporal variability (stand. dev. = 6.8 psu). During the winter, both lagoons are characterized by high water salinity due to both the reduction of irrigation activities in the adjacent fields, and the entrance of high salinity water originated from the Black Sea and flowing, through Dardanellia Straits, along the N. Aegean coastlines. This water

Figure 4. Seasonal patterns of environmental variables at Vassova and Eratino lagoons, a fresh-water canal and coastal sea.

180 G. SYLAIOS AND V. THEOCHARIS


181

type has lower salinity during the summer and higher during the winter, and has been observed at the adjacent coastal area (Yuce, 1995). The two lagoons exhibit different behavior in terms of the processes determining the dissolved oxygen concentration. At Vassova lagoon, the smaller wind fetch leads to the absence of surface waves, which in conjunction with increased biological activity at the central basin, is responsible for the observed low dissolved oxygen values. Reduced dissolved oxygen concentration values at Eratino are mostly associated with the eutrophication phenomena that prevail at the area near the 2 km natural channel of the lagoon (stations E2-E6). Similarly, nutrient concentration (nitrates, nitrites, ammonium and phosphates) at these four stations, representative of fresh, brackish and saline water at the lagoon system of N.E. Greece, are given in Figure 5. Two nutrient minima (in spring and autumn) and an important maximum during the summer period show the strong influence of fresh water entering both lagoons. Nitrate values within the lagoons remain below fish toxicity levels, ranging between 0.03–0.45 µM. Increased nitrate concentration and an important maximum (6.50 µM) were found at the adjacent fresh water canals. Reduced precipitation and fertilizers are the most important factors determining the nitrate concentration. Since this coastal ecosystem is coupled with an adjoining agricultural watershed, the dominant source of nutrients carried to the sea may be fertilizer. Other sources of nutrients in agricultural systems include nitrogen ‘fixed’ by N-fixing crops and animal waste. Nitrites at fresh water samples displayed values ten times higher (0.10 µM) than those observed at the central basin (0.01 µM) of both lagoons. Phosphates ranged between 0.02–0.50 µM at the central lagoon basins and between 0.04–0.58 µM at the drainage canals. The summer phosphate peak follows closely the amount of fresh water discharged into the system. This indicates the long-term influence of fresh water discharged in the lagoon and the settling of phosphates, which reappear in the water column during summer remineralization (Nixon et al., 1980). Furthermore, phosphates tend to determine the fresh water type discharged into the lagoon as a result of intensive agricultural and urban activities of the broader area. Figure 6 relates salinity to nutrient concentration in order to determine the origin of nutrient enrichment at the lagoon basins. All the squared correlation coefficients show relatively low values, something typical for this kind of field data, since nutrient concentrations in the lagoonal environment are not only related to salinity. However, these regression equations are not fitted to produce a law but to show the dependence of nutrient concentration on freshwater influence. Nitrate and nitrite concentrations show a decreasing trend with increasing salinity, proving the important contribution of fresh water drainage canals on the nutrient supply of the lagoon basins. However, the presence of ammonium and phosphorus in the system is not well related to fresh water inflow (lower correlation coefficient r2 ), meaning that there is an important transfer of these nutrients from the adjacent coastal sea, through the lagoon mouth. Increased ammonium and phosphate concentrations at the eastern part of Kavala Gulf are associated with the intensive agricultural and

Figure 5. Seasonal patterns of nutrients at Vassova and Eratino lagoons, a fresh-water canal and coastal sea.

182 G. SYLAIOS AND V. THEOCHARIS

Figure 6. Linear relations of water salinity to nutrient concentrations at the lagoons of N.E. Greece.


183

184


Figure 7. PCA results for the lagoons of N.E Greece (a) when ignoring ecosystem or sampling position, and (b) when considering differences between various parts of each lagoon.

urban activities (e.g., existence of a phosphoric fertilizers factory and a wastewater treatment plant) near the coast of the broader area (Sylaios et al., 2000). Figure 7(a) presents the eigen-values that scale the physical parameters, nutrients and seston biomass as they occur through PCA, irrespectively of the ecosystem or the sampling position. According to these results, Axes 1 and 2 demonstrate that denitrification and its gradient at fresh, brackish and saline water, is the major factor, explaining 44% of the total annual variance. It occurs that nitrites, ammonium and phosphates are closely associated with the first axis, while salinity, suspended sediments, nitrates and the ratio N/P are associated with the second axis. Water and air temperature, pH and dissolved oxygen seem to occupy an intermediate position. Axes 3 and 4 give the degree of differentiation in the effect of fresh water inflow from the adjacent drainage canals into the basins of Vassova and Eratino lagoon. Figure 7(b) refers to the eigen-values of nutrients and seston biomass obtained by the PCA when considering the differences between the various parts of each lagoon (wintering canals, main basin, fish barriers). Axis 2 shows the relative importance of nitrite and ammonium at the wintering canals. It appears that the variability of wet seston biomass, salinity, nitrates and phosphates (axis 1) is significant in the main basin. The variation of N/P-ratio depends mostly on the variability of phosphorus concentration, especially during spring and autumn when the least photosynthetic activity takes place in the lagoon.

Figure 8. Seasonal patterns of environmental variables at Vontas and Kalaga lagoons, Kalamas River and coastal sea.


185

186


4.2. SEASONAL PATTERNS IN THE LAGOONS OF N . W. GREECE The annual variability of the aquatic environmental parameters in the mixed water of the lagoons Vontas and Kalaga (as representative of the system), the fresh water of River Kalamas and the adjacent seawater are shown in Figure 8. The temperate marine climate of the area affects the water temperature variability of the lagoons, which reduces to 10 ◦ C during March and increases to 34 ◦ C at Kalaga lagoon during August. The lagoon system showed a minimum temperature of 5 ◦ C, a maximum of 36 ◦ C and the highest variability of 10.4 ◦ C at the northern station of Loutsa-Papadia lagoon. Water temperature and salinity correlate well with air temperature (r2 = 0.86) and adjacent river discharge (r2 = 0.78), respectively. Water circulation and mixing inside the lagoons is mostly dominated by the intrusion of seawater, depending on the geometry and the orientation of the mouth of each lagoon. Salinity ranged between 12.9 psu at Loutsa-Papadia lagoon during March to 39.8 psu in Alykes lagoon during July. Mean annual salinity has increased values at Alykes (39.7 psu) and Kalaga lagoon (39.8 psu) and moderate to low values at Vontas (31.9 psu), Lousa-Papadia (32.8 psu), Vatatsa (32.4 psu) and Richo (30.5 psu). The system is mainly controlled by evaporation during summer (annual evaporation of 250 mm) and precipitation during winter (annual precipitation > 1000 mm). Dissolved oxygen ranged between 6.6 mg l−1 at Vatatsa lagoon during August and 14.2 mg l−1 at the northern part of Loutsa-Papadia lagoon during November, proving that wind waves are responsible for dissolved oxygen transfer in the water column. However, anoxic conditions have been observed at Loutsa-Papadia lagoon during the summer period, under conditions of increased water temperature and neutral pH values. These conditions result from the continuous consumption of dissolved oxygen by bacteria, which decompose and mineralize dissolved and particular organic material, supplied by land drainage. The oligotrophic nature of the lagoons in N.W. Greece is characterized by the near zero values and the limited temporal variability of nutrients (Figure 9). Nutrient concentrations exhibit considerably increased values at the water of Kalamas River in comparison to those observed at the adjacent coastal lagoons, proving the small effect that fresh water has on the lagoonal ecosystem, mainly due to the general coastal circulation of the area. Hence, the lagoons of N.W. Greece show environmental characteristics according to their distance from fresh water discharge (Kalamas River). The mean annual nitrate concentration at Kalamas River (1.66 µM), Richo and Kalaga lagoons (0.07 µM), Loutsa-Papadia (0.05 µM), Vatatsa (0.04 µM), Alykes and Vontas (0.02 µM), gives an important spatial variability at both sides of fresh water outflow. Nitrites showed mean annual variability ranging between 0.002 µM at Vatatsa lagoon to 0.015 µM at Richo lagoon, with Kalamas River having a mean value of 0.26 µM. Similarly, ammonium illustrated small temporal variability and values ranging between 0.001 µM at Vatatsa lagoon and 0.030 µM at Loutsa-Papadia. Kalamas River, with a minimum value of 0.20 µM

Figure 9. Seasonal patterns of nutrients at Vontas and Kalaga lagoons, Kalamas River and coastal sea.


187

188


during March, a maximum value of 0.95 µM in August and a mean annual value of 0.64 µM, appears to be the main source of ammonium in the area. The mean annual concentration of phosphates in Kalamas River (0.009 µM), Richo, Kalaga and Loutsa-Papadia lagoons (0.007 µM), Vatatsa (0.009 µM), Alykes (0.006 µM) and Vontas (0.002 µM), show the limited contribution of phosphorus on the nutrient enrichment of these lagoons. Nutrients in lagoon water vary in inverse proportion to salinity, showing that nitrogen and phosphorus are being transported to lagoonal basins by fresh water inflow (Figure 10). It occurs that nutrient distribution in these lagoons results primarily from river runoff and biological activities. Figure 11 displays the eigenvalues produced from the Principal Components Analysis, on a correlation matrix of nine abiotic parameters, from the monthly values of the lagoons of N.W. Greece. The cumulative eigen-values of the system, calculated by least squares procedures, show that axes 1 and 2 account for the 54.2% of the total annual variance of the system. It results that nitrites, nitrates and N/P-ratio are positively correlated and closely associated with the first axis, while salinity and phosphates are negatively correlated and associated with the second axis. Water temperature, pH, dissolved oxygen and dry seston biomass seem to occupy intermediate positions. Axes 3 and 4 give the degree of differentiation in the effect of fresh water inflow from the adjacent Kalamas River into the basins of the lagoons of N.W. Greece. As expected, pH and wet seston biomass concentration showed inverse relation to dissolved oxygen, when plotted to the system of 3 and 4 axes. 4.3. LAGOONS WATER BALANCE The net water balance terms for each lagoon basin were analyzed, in order to investigate the order of magnitude of the hydrologic processes responsible for maintaining steady-state conditions. This methodology has also been described by LOICZ Biogeochemical Modeling Guidelines (Gordon et al., 1996) for the application of simple budget models in coastal water bodies. These budget models are generally defined as mass balance calculations of specific variables (water, salt, sediment, carbon, nitrogen, phosphorus, etc.) for a defined geographical area and time period. Thus, the net water balance can be written as: V = QP + QE + QG + QR + QO (1) t where V = lagoon water volume (m3 ), QP = precipitation rate (m3 s−1 ), QE = evaporation rate (m3 s−1 ), QG = groundwater inflow rate (m3 s−1 ), QR = surface runoff discharge (m3 s−1 ) and Qo = net water discharge (m3 s−1 ), defined as (inflow – outflow) through lagoon mouth during mean tidal conditions. These annual water discharges are expressed in m3 s−1 with the sign convention that water gain is positive and water loss is negative. Over the period of study (t = one year), there are no indications that the mean water level changed; thus, the term dV/dt can safely be assumed equal to zero.

Figure 10. Linear relations of water salinity to nutrient concentrations at the lagoons of N.W. Greece.


189

190


Figure 11. PCA results for the lagoons of N.W. Greece. Table III. Water fluxes estimates for the eight lagoons of Northern Greece Lagoon

QP (m3 s−1 )

QE (m3 s−1 )

QR (m3 s−1 )

QO (m3 s−1 )

QT (m3 s−1 )

T50% (days)

Vassova Eratino

0.042 0.054

–0.085 –0.110

0.940 2.110

–0.897 –2.055

12.076 15.655

1.98 2.04

Loutsa – Papadia Kalaga Vatatsa Richo Alykes Vontas

0.070 0.024 0.028 0.035 0.105 0.087

–0.016 –0.006 –0.006 –0.008 –0.024 –0.020

8.220 0.370 2.090 2.170 0.170 3.350

–8.274 –0.389 –2.112 –2.197 –0.251 –3.417

8.945 3.131 3.578 4.472 13.418 11.182

1.78 2.55 2.23 1.78 1.19 2.85

QP = precipitation rate; QE = evaporation rate; QR = fresh water runoff discharge; QO = tidal net water discharge; QT = tidal flushing rate; T50% = calculated flushing half-life.

The term QP represents the total annual precipitation rate on the lagoon surface, calculated as the sum of products of monthly rainfall depths times the lagoon surface area (Table III). QE represents the total annual evaporation rate from the lagoon surface area, computed as the sum of the products of monthly evaporation depth times the lagoon surface area. QG represents the groundwater inflow into the lagoon, which should be at least an order of magnitude lower than the other parameters, thus it can be ignored, since no measured data were available. However, inclusion of real data would probably slightly modify the water balance. The term QR represents the surface runoff discharge from the surrounding drainage basin. QR was calculated using simple tidal prism considerations and salinity data. For a lagoon basin with constant salinity over time, the amount of salt entering


191

Figure 12. Lagoon representation by a single box, where V is the low tide volume and VP is the tidal prism. The outflow and inflow volumes are shown for the ebb and flood tides.

and leaving the basin within a tidal cycle must be zero. If VR is the volume of fresh water entering each lagoon basin from the drainage basin and VP the volume of saline water entering (during flood) or leaving (during ebb) the lagoon basin through its mouth, according to tidal phase, then salt balance requires: VR VR VR VR SR − + V P SL + SR + V P − SO = O (2) 2 2 2 2 where SR , SL and SO the salinity at the river, the lagoon and the coastal sea water, respectively. The first and third terms of equation (2) represent the amount of salt entering the lagoon from freshwater sources during each half (flood or ebb) of the tidal cycle. The second term represents the amount of salt leaving the lagoon during the ebb phase of tide, while the fourth term the amount of salt entering the lagoon from the sea during the flood phase of tide (Figure 12). Since monthly salinity values are available for each lagoon and VP can be estimated from the lagoon area and mean tidal range, Equation (2) can be used to quantify VR , which represents the monthly average volume of fresh water entering each lagoon basin. The magnitude of the monthly QR -term was calculated as QR = VR T−1 , where T the duration of a tidal cycle. These monthly QR -values were used to calculate the mean annual freshwater discharge flowing into the lagoon. The term QO represents the residual water exchange between the inflow of saline coastal water and the outflow of lagoon brackish water on tidal, meteorological and longer time scales. On the assumption that the lagoon volume remains constant within the period of observation, QO was estimated from Equation (1). Table III illustrates the summary of the estimated annual water discharge terms for the lagoon systems of northern Greece. It occurs that fresh water runoff (QR ) and tidal net water discharge (QO ) are the dominant terms in the water balance

192


of these lagoons. Precipitation and evaporation rates appear to play a minor role, being an order of magnitude less than the dominant water rates. Salt balance considerations show that maximum fresh water discharge into lagoon basin occurs at Loutsa-Papadia (8.22 m3 s−1 ) and Vontas (3.35 m3 s−1 ). Kalaga and Alykes lagoons have very small fresh water inflows of 0.37 and 0.17 m3 s−1 , respectively. At N.E. Greece, important fresh water flows occur at Eratino lagoon, through the old deltaic tributary of Nestos River. Precipitation rates have an average value of 0.06 m3 s−1 in N.W. Greece lagoons and 0.05 m3 s−1 in N.E. Greece lagoons. Evaporation rates vary between an average value of 0.1 m3 s−1 at Nestos River lagoons and 0.01 m3 s−1 at Kalamas River lagoons. 4.4. LAGOONS FLUSHING TIME Flushing time is considered as one of the most critical measures in lagoon management, since it approximates the time scale of water exchange, e.g., flushing, turnover or residence time (Zimmerman, 1981; Knoppers et al., 1991). The concept of flushing half-life (T50% in days), or the time needed for the lagoon basin to replace half of its lagoon volume, was adopted as appropriate for the estimation of the steady-state flushing rate. Kjerfve et al. (1996) considered that lagoon flushing half-life is estimated by the ratio: T50% = 0.69 / k

(3)

where k is a rate constant calculated as the average fraction of lagoon water volume replaced each second by the sum of the water fluxes. Hence, k=

[QR + QP + QG + QO + | QT |] V

(4)

The term QT (m3 s−1 ) represents the tidal flushing rate (the oscillating water exchange on a tidal time scale); hence the absolute value sign was used. Since in both lagoon systems predominant tidal constituent is the semi-diurnal tide, QT , was expressed as the prism entering the lagoon system per tidal cycle, although in reality this transport occurs only during half a tidal cycle. QT was calculated by: QT = ±

AL h T

(5)

where h is the mean tidal range (m), AL is the lagoon surface area (km2 ) and T is the period of the semi-diurnal tide (T = 12.42 hr = 44714 s). Kjerfve et al. (1996) calculated the flushing half-life for the lagoon Itaipu (Brazil) having similar geometrical characteristics, fresh water inflow and tidal range, as 1 day. Results show that Eratino, Vontas, Kalaga and Vatatsa lagoons have a flushing half-life time of more than 2 days (Table III). Alykes have the lowest water exchange characteristics of the order of 1 day.


193

Table IV. Calculated nutrient enrichment factors for the eight lagoons of Northern Greece Lagoon

N-NO3

N-NO2

N-NH4

P-PO4

Total Nitrogen

Vassova Eratino

1.633 1.649

0.063 0.070

0.080 0.062

2.113 1.931

0.771 0.772

Loutsa – Papadia Kalaga Vatatsa Richo Alykes Vontas

1.053 1.072 1.042 1.073 1.160 1.010

0.431 0.512 0.417 0.458 0.487 0.417

0.490 0.484 0.450 0.473 0.521 0.455

0.068 0.051 0.060 0.053 0.047 0.029

0.731 0.752 0.709 0.739 0.801 0.697

The above flushing half-life calculations account for both the hydrologic and tidal dispersive processes. The calculated exchange time scales produce a reasonable and easily understandable measure of flushing time, suitable for the ecological and water quality evaluation of coastal lagoons. This model assumes both steady-state and complete mixing conditions of the lagoon, on a time scale that is short, compared to the flushing half-life time. Thus, the calculated flushing halflife should always be considered as the lower limit for water exchange or the ideal flushing time scale. The real advantage of calculating the flushing half-life or the exchange of 50% of the lagoonal water is for comparative analysis of coastal lagoons, thus allowing the better decision-making process in solving lagoonal water quality problems. 4.5. NUTRIENT ENRICHMENT FACTORS In order to estimate the extend of eutrophication in these lagoons, we examined the nutrient content for each basin according to the influence of the various water masses (river, sea, lagoon water), and related these results to background values obtained from a reference station. Table IV presents the relative enrichment factors, illustrating that N.E. Greece lagoons have nitrate content 1.6 times higher than the background, while the N.W. Greece lagoons have equal nitrate content to the background. Nitrites show seven times higher content in Kalamas River lagoons (mean relative factor: 0.453) than that in Nestos River lagoons (mean relative factor: 0.066). Ammonium concentrations were found to be six times higher in the N.W. Greece lagoons (mean relative factor: 0.478) than those observed in Vassova (0.080) and Eratino (0.062). On the contrary, phosphates were found to have concentrations forty times higher in the lagoons of Nestos River (mean relative factor: 2.022), than those observed in the

194


lagoons of Kalamas River (mean relative factor: 0.051). This very high phosphorus content is mostly supplied from the fresh water draining the broader area, where important agricultural activity takes place, and the rich in phosphate coastal seawater, due to the presence of a fertilizer factory and a wastewater treatment plant, located in the eastern part of Kavala Gulf. Generally one can say that the quality of lagoon basins with respect to nutrients depends on the different nutrient sources, the morphology of the basin and the renewal of water with the adjacent coastal sea. The N.E. Greece lagoons have a higher nutrient concentration than N.W. Greece lagoons, due to the more intense human activities (agricultural, industrial and urban) in the broader drainage basin of these lagoons, the transfer of nutrients in the adjacent coastal zone from Nestos River and the influence of the nutrient-rich Black Sea water which outflows through the Dardanelles and moves cyclonically along the Thracian Sea coastline into Kavala Gulf.

5. Conclusions and Management Proposals The comparative hydrology and water quality variability has been studied in this paper, for two of the most significant lagoonal systems in Northern Greece. Low water temperature during the winter and increased phosphate content are two of the most important problems concerning the lagoons of N.E. Greece. Summer denitrification is the most important process in these lagoons, resulting from increased surface temperature, relatively low water flow in the basins, consumption of dissolved oxygen from organic matter, presence of surface algae and bacterial production. The above suggest the need for the continuous control of fresh water quality entering lagoons through the drainage canals and the reduction of phosphoric fertilizers used by agriculture in the surrounded areas. Furthermore, fish production development demands better oxygenation of the water column. One way to resolve these problems is the transfer of fresh warm groundwater to eliminate winter massive fish deaths and to increase the fresh water portion into these lagoons. The flushing half-life of Vassova and Eratino lagoons was calculated to 1.98 and 2.04 days, respectively, when accounting both hydrological and tidal dispersive processes. Simple water budget considerations, based on Equation 1, show that the pumping of fresh groundwater into the lagoon at a rate of 100 m3 hr−1 (0.027 m3 s−1 ) will reduce the flushing half-life of Vassova lagoon to 1.83 days (a 7.5% reduction). Increased salinity (due to evaporation) and temperature, low dissolved oxygen values and increased nitrites and ammonium concentrations during summer suggest the need for bathymetric modifications and the widening of the sea-lagoon channels, allowing better water exchange with the adjacent coastal sea, in the lagoons of N.W. Greece. The flushing half-life of these lagoons was calculated, ranging between 1.19 days (for Alykes) to 2.85 days (for Vontas lagoon). The input of groundwater (100 m3 hr−1 ) will have similar effects, reducing the lagoons


195

flushing half-life from 0.71 days for Richo lagoon (60.2% reduction) to 2.04 days for Vontas lagoon (28.5% reduction).

References Barnabe, G.: 1992, ‘Du potentiel en sommeil au potentiel en danger’, Aqua Revue, 22–25. Colombo, G.: 1977, ‘Lagoons’, in R. S. K. Barnes (ed.), The Coastline, John Wiley and Sons, New York, N.Y., U.S.A., pp. 63–81. Correll, D. L., Jordan, T. E. and Weller, D. E.: 1992, ‘Nutrient fluxes in a landscape: effects of coastal land use and terrestrial community mosaic on nutrient transport in coastal waters’, Estuaries 15, pp. 431–442. Corsi, F. and Ardizzone, G. D.: 1985, ‘Some environmental conditions affecting yellow eels catchability’, Oebalia IX-2, pp. 561–571. Dunteman, G. H.: 1989, Principal Component Analysis, Sage University Paper, 69, Sage, Beverly Hills. Dyer, K. R.: 1973, Estuaries: A Physical Introduction, John Wiley and Sons Ltd., London, 140 p. Friligos, N.: 1988, ‘Nutrient enrichment and circulation of water masses in the Pagasitikos Gulf (Aegean Sea)’, Oceanologica Acta, Suppl., pp. 111–122. Gordon, D. C., Bourdeau, P. R., Mann, K. H., Ong, J-E., Silvert, W. L., Smith, S. V., Wattaykorn, G., Wulff, F. and Yanagi, T.: 1996, ‘LOICZ Biogeochemical Modeling Guidelines’, LOICZ/R&S/955, VI + 96 p., LOICZ, Texel, The Netherlands. Kjerfve, B.: 1986, ‘Comparative oceanography of coastal lagoons’, in D. A. Wolfe (ed.), Estuarine Variability, Academic Press, New York, U.S.A., pp. 63–81. Kjerfve, B. and Magill, K. E.: 1989, ‘Geographic and hydrographic characteristics of shallow coastal lagoons’, Marine Geology 88, pp. 187–199. Kjerfve, B., Schettini, C. A. F., Knoppers, B., Lessa, G. and Ferreira, H. O.: 1996, ‘Hydrology and salt balance in a large, hypersaline coastal lagoon: Lagoon de Araruama, Brazil’, Estuarine, Coastal and Shelf Science 42, pp. 701–725. Knoppers, B., Kjerfve, B. and Carmouze, J. P.: 1991, ‘Trophic state and water turn-over time in six choked coastal lagoons in Brazil’, Biogeochemistry 14, pp. 149–166. Linn, R. L., Centra, J. A. and Tucker, L.: 1975, Multivariate behavioral research 10, pp. 277–288. Luketina, D.: 1998, ‘Simple tidal prism models revisited’, Estuarine, Coastal and Shelf Science 46, pp. 77–84. Mee, L. D.: 1978, ‘Coastal lagoons’, in J. Riley and O. Skirrow (eds.), Chemical Oceanography 7, Academic Press, New York, U.S.A., pp. 441–490. Miller, J. M., Pietrafesa, L. J. and Smith, N. P.: 1990, ‘Principles of hydraulic management of coastal lagoons for aquaculture and fisheries’, in F.A.O. (ed.), FAO Fisheries Technical Paper 314, Rome, 88 p. Nixon, S. W.: 1982, ‘Nutrient dynamics, primary production and fisheries yields of lagoons’, Oceanologica Acta Suppl. 4, pp. 357–372. Nixon, S. W., Kelly, J. R., Furnas, B. N. and Oviatt, C. A.: 1980, ‘Phosporus regeneration and the metabolism of coastal marine bottom communities’, in K. R. Tenore and B. C. Coull (eds), Marine Benthic Dynamics, Univ. of South Carolina Press, Columbia, pp. 219–242. OECD: 1992, ‘Market and government failures in environmental management’, Wetlands and Forests, OECD Publ., Paris, 118 p. Officer, C. B.: 1976, Physical Oceanography of Estuaries (and Associated Waters), Wiley, London, 455 p. Sikora, W. B. and Kjerfve, B.: 1985, ‘Factors influencing the salinity of Lake Pontchartrain, Louisiana, a shallow coastal lagoon: analysis of a long term data set’, Estuaries 8, pp. 170–180.

196


Strickland, J. D. H. and Parsons, T. R.: 1972, A Practical Handbook of Sea Water Analysis, 2nd ed., vol. 167, Fish Res. Board, Ottawa, Canada. Sylaios, G., Ioannidou, D., Stamatis, N. and Kallianiotis, A.: 2000, ‘Water quality estimation of Kavala Gulf’, Proceedings of 6th Panhellenic Congress in Oceanography and Fisheries, Chios, pp. Theocharis, V. S., Sylaios, G. and Stamatis, N.: 2000, ‘Water quality variability at two coastal lagoons in Northern Greece’, Fresenius Environmental Bulletin 9, 30–35. Therianos, A. D.: 1974, ‘The geographical distribution of river water supply in Greece’, Bull. of Geol. Soc. of Greece XI, pp. 25–28. UNESCO: 1981, ‘Tenth technical report of the joint panel on oceanographic tables and standards’, in UNESCO Technical Papers in Marine Science 36, 24 p. Valiela, I., Foreman, K., La Montagne, M., Hersh, D., Costa, J., Peckol, P., DeMeo-Andreson, B., D’Avanzo, C., Babione, M., Sham, C-H., Brawley, J. and Lajtha, K.: 1992, ‘Couplings of watersheds and coastal waters: sources and consequences of nutrient enrichment in Waquoit Bay, Massachusetts’, Estuaries 15, pp. 443–457. Vallejo, S. M. P.: 1982, ‘Development and management of coastal lagoons’, Oceanologica Acta Suppl. 4, pp. 397–401. Vernberg, F. J.: 1982, ‘Environmental adaptation to lagoon systems, Oceanological Acta’, Proc. of Intern. Symp. On Coastal Lagoons, Bordeaux, France, 8–14 September, 1981, pp. 407–415. Xanthopoulos, K.: 1985, ‘River characteristics at Ipiros area, Greece’, Proc. of the Third Panhellenic Hydrology Conference, Patras, pp. 1–17. Yuce, H.: 1995, ‘Northern Aegean water masses’, Estuarine Coastal and Shelf Science 41, pp. 325– 343. Zimmerman, J. T. F.: 1981, ‘The flushing of well mixed tidal lagoons and its seasonal fluctuation’, UNESCO Technical Papers in Marine Science 33, Paris, pp. 15–26.