17 The P Cycle In The Balaton Catchment - A Hungarian Case Study

 

Sándor Herodek1, Vera Istvánovics1,

Géza Jolánkai2, Péter Csathó3, Tamás Németh3 and G. György Várallyay3

1) Balaton Limnological Research Institute of the Hungarian Academy of Sciences, Tihany|
2) Water Resources Research Centre, VITUKI, Kvassay J.út 1, H-1095 Budapest
3) Research Institute for Soil Science and Agricultural Chemistry of the Hungarian Academy of Sciences, Herman Ottó u.15. Budapest II, Hungary

Lake Balaton, the largest lake of Central Europe, is situated in the western part of Hungary in the slightly undulating-hilly "Transdanubia" region. The Lake has a surface area of 593 km2 and a mean depth of 3.2 m. It consists of a sequence of four concatenated basins. The Balaton catchment area covers about 5200 km2 and can be divided into 9 smaller watersheds (Figure 1). The largest is the Zala River watershed, occupying 50.4% of the total area and belonging territorially to four administrative counties: Somogy (34.2%), Vas (9.7%), Veszprém (19.8%) and Zala (36.3%). Most of the other smaller rivers and creeks discharge into the second basin. The only outflow, the Sió Canal, drains the water from the fourth basin into the River Danube (Figure 1). The multi-annual water balance shows 577.106m3y-1 tributary inflow, 373.106m3y-1 precipitation, 545.106m3y-1 evaporation and 405.106m3y-1 outflow.

Northwesterly winds are frequent, and the strong waves swirl up much sediment. The concentration of suspended sediments varies from 2 to 200 g.m3 according to the wind conditions. In long calm periods up to 10% of the light may reach the bottom, whereas during heavy storms less than 1% penetrates down to 2 m depth. The lotic southern shore of the lake is sandy. On the wind protected northern one there are reed belts, but macrophytes cover only 2-3% of the lake surface.

The Lake is covered by ice in January and February, while in July and August the mean water temperature is 24°C. The pH of the water is 8.4, but may increase up to 9.1 in the western part of the Lake during algal blooms. The mean concentrations of the main ions are the following (g.m3): Ca2+ 40; Mg2+ 40; Na+ 20; K+ 6; HCO3- 230; CO32- 15; Cl- 20; SO42- 83.

Balaton was already a favourite summer resort in the first half of this century, but the tourist boom started in the 1960's. Recently the Lake is visited by 2 million tourists every summer which spent ca. 20 million visitor days. The income from tourism is estimated at 40 billion HUF/y (1 USD = 93 HUF) corresponding to about one third of Hungary's total tourism income.

The piped water supply has more or less followed the increase of the summer population, but sewage treatment lags behind. Until the 1980's Lake Balaton remained the recipient of these sewage waters, from which phosphorus was not removed. The storm water runoff from the ever increasing municipalities contributed considerably to the loading of the Lake. Towns and smaller settlements in the catchment area, but beyond the recreational region (e.g. in the Zala River watershed) have also developed rapidly, with multiplied water demands and with significant environmental hazards, e.g. pollution of surface and subsurface water resources. The other two potential sources of P-load to Lake Balaton were:

 

Figure 1. Lake Balaton and the Balaton catchment area with boundaries and numbers of sub-catchment areas.

 

The shallow-lake ecosystems of the Balaton, with its low chemical and biological buffer capacity and high susceptibility to various stresses, is sensitive to these loads. This results in significant environmental consequences, such as unfavourable changes in water quality and in the aquatic ecosystems, eutrophication, deterioration of the reed belt, silting-up, accumulation of potentially harmful chemicals in the sediment, etc.

The crucial problem of the Lake is the ongoing eutrophication. This is mainly due to the external anthropogenic nutrient (mostly phosphorus) loads. In the 80's serious efforts were made to limit the P loads to the Lake: diversion and treatment of about 2/3 of the former sewage discharges; building of several pre-filter reservoirs, re-establishment and reconstruction of the "Kisbalaton" former wetland (peat land with high filtering and storage capacity) at the mouth of the Zala River, the major tributary; control of land use and soil management practices, including fertilizer application in the catchment area; etc. In spite of these efforts no appreciable improvement of water quality has been achieved and there are inevitable increasing trends of eutrophication in almost all parts of the Lake (e.g. algae chlorophyll-a concentration exceeding 200 mg m3 are still being measured in the westernmost Keszthely bay).

Because of the particular economic, environmental and landscape significance of Lake Balaton and its surroundings, numerous integrated national and international research programmes have been initiated and carried out in the region. In the present paper the main concepts, aspects, results and conclusions of this research are briefly and schematically summarized, focusing attention on the P cycle in the Balaton catchment area and on the environmental impact analysis, particularly on the aquatic ecosystems of the Lake.

 

 

SOILS OF THE BALATON CATCHMENT AREA

 

The Balaton catchment area (Figure 1) belongs to three main geographical (agroecological) regions of Hungary (National Atlas, 1989):

 

The simplified soil map of the Balaton catchment area is shown in Figure 2, and the schematic map of soil texture in Figure 3 (Hinrichssen & Enyedy, 1990; Várallyay, 1990; Várallyay et al., 1979-1980).

The major part (about 80-85%) of the area is covered by various types of brown forest soils and only 15-20% by different hydromorphic soils and chernozems. The brown forest soils are increasingly leached from east to west, since the annual precipitation increases from 550-600 mm to 900-1000 mm from the easternmost part of the Balaton catchment area to the western border. Soils along this "leaching sequence" are:

 

The main part (about 2/3) of the Balaton catchment area is covered by lessivated brown forest soils (Stefanovits, 1971). These soils were formed mostly on medium-textured Quaternary loess or loess-like deposits; a smaller part on Tertiary or older deposits with various particle-size distribution but generally with light texture (Inner Somogy); and an even smaller part on Permian sandstone (5. and 6. sub-catchment). The main characteristics of these soils can be generalized as follows: non-calcareous, slightly acidic (pH: 6.2-6.8) soils with medium texture (loam, sandy loam in the A-horizon; loam, clay loam in the B-horizon); 7-10 g Kg-1 organic C content, 250-350 mg Kg-1 total P content; 50-60% base saturation; 70-75% Ca saturation of the CEC, with about 25% exchangeable Mg and negligible amounts of exchangeable K and Na. A few of these soils are shallow (in the Zala catchment area) due to a cemented clay-gravel hardpan near the surface.

 

 

Figure 2. Soil map of the Balaton catchment area (simplified schematic of the original 1:100 000 scale Soil Map of Hungary). Soils occurring in the Balaton catchment, according to the legend of the original Soil Map of Hungary: 1. Lowland chernozems, meadow chernozems (the term "meadow" is related to hydromorphic character), terrace chernozems; 2. rendzinas, erubase soils, "nyirok"; 3. Lessivated brown forest soil; 4. Pseudogelys; 5. Chernozem-brown forest soils; 6. Peat, ameliorated peat; 7. Alluvial soils; 8. Brown earths (Braunerde).

 

Brown earths cover about 15-20% of the Balaton catchment area. They are formed mostly on loamy to sandy loam loess and loess-like deposits in the 7. and 8. sub-catchments and on Tertiary deposits and weathered volcanic materials (basalt) in the 5. and 6. sub-catchments. Their general characteristics are similar to the above mentioned soils, without significant textural differentiation (Bt-horizon) within the soil profile.

Pseudogleys occur in the westernmost part of the Zala catchment area; chernozem-brown forest soils in the 7. and 8. sub-catchments; and chernozems around the NE shore of the Lake (at the border of the Transdanubian Loess Plateau). A considerable part of the 2. (Keszthely Mountain) and 4. (Bakony Mountain) is occupied by rendzinas, formed on limestone. These soils are usually calcareous, Ca-saturated, with high organic matter content in the shallow humus horizon. Because of shallow depth most of these soils are not under agricultural utilization and usually are covered by poor forest.

 

Figure 3. Map of soil texture in the Balaton catchment area (simplified schematic of the original 1:100 000 scale Soil Texture Map of Hungary). Legend of the original map: 1. Sand; 2. Sandy loam; 3. Loam; 4. Clay loam; 5. Clay; 6. Organic soils (peat, partly decomposed peat); 7. Coarse fragments (gravel, non- or partly weathered rocks, etc.)

 

About 15-20% of the Balaton catchment area is covered by various hydromorphic soil types: the - usually narrow - valleys of the Zala River and the smaller creeks by alluvial soils, alluvial meadow soils, meadow soils and peaty meadow soils. All these soils show great spatial (both horizontal and vertical) variability in texture, hydrophysical properties, pH, carbonate and organic matter contents, and consequently in fertility and productivity. The majority of these soils is used as grassland, only a part is arable land.

There are three peat areas in the Balaton region: the Balaton Minor (Kisbalaton); the Nagyberek (in 9. sub-catchment); and the Tapolca Basin (in 3. sub-catchment). The Tapolca Basin is a protected ancient peat relict; a considerable part of Nagyberek was drained and ameliorated in the 50's and used for agricultural production. The Kisbalaton was also partly drained when the Zala river bed was channelled. When it turned out that the project was a failure, re-establishment of the Kisbalaton wetland was implemented, mainly because: (1) the peat area was to be used as a natural filter to prevent (or to reduce) the direct discharge into Keszthely bay (I. sub-basin of the Lake) of the nutrient (particularly P) load which the Zala River collected from the large and rather heavily loaded Zala catchment area; (2) the original wetland ecosystems were to be re-established as a natural protected area.

The soil erosion map of the Balaton catchment area (Figure 4) (Stefanovits, 1964; National Atlas, 1989) shows four grades of soil erosion: (1) non- or insignificantly eroded areas; (2) weakly eroded areas (less than 30% of the original surface layer is eroded); (3) moderately eroded areas (30-70% of the original surface layer is eroded); (4) strongly eroded areas (more than 70% of the original surface layer is eroded). About 70% of the area of the northern catchments (2., 3., 4., 5., 6. and northern part of 1.) are strongly eroded, especially when they are not under forest vegetation, but used as arable land, orchards and vineyards (with up and down cultivation practices in the slope direction). About 15% is moderately, 10% weakly and 5% is not eroded. In the flatter southern sub-catchments (7., 8., 9. and southern part of 1.) the approximate distribution among the strongly-moderately-weakly non-eroded categories is 10, 40, 40 and 10% respectively. In the valleys, local basins and low-lying areas, large portions are affected by sedimentation. In addition to water erosion, the impact of wind erosion has become more significant, particularly on drained and ameliorated peatlands and places with sandy soils, so that both water and wind erosion now play an important role in the nutrient loading of Lake Balaton.

 

 

 

Figure 4. Soil erosion map of the Balaton catchment area (simplified schematic of the original Soil Erosion Map of Hungary). Legend of the original map: 1. Strongly eroded areas. 2. Moderately eroded areas. 3. Weakly eroded areas; 4. Non-eroded areas; 5. Areas with deposits; 6. Wind-eroded areas; 7. Forests.

 

 

LAND USE AND FERTILIZER APPLICATION IN THE BALATON CATCHMENT AREA

 

The land use pattern in the Balaton catchment area can be characterized by the following figures (the figures in brackets show the Hungarian national average): arable land: 38.7 % (50.3%); plantations, orchards, vineyards, gardens: 7.0 % (6.7%); grassland: 13.8% (13.8%); forest: 27.7% (17.6%); uncultivated area: 12.8% (11.6%).

In the period preceding World War II the nutrient status of Hungarian soils was poor due to a negative nutrient balance, resulting from yield exports with minimal recycling of plant residues, organic or green manuring or mineral fertilisation. The average rate of fertilizer application was below 5 kg ha-1. In 1950 and 1955 this value was still only 6 and 10 kg ha-1, respectively (Várallyay et al., 1992). From this time on there has been a sharp increase in the rate of fertilizer application. In 1962 fertilizer application exceeded 50; in 1968, 100; in 1973, 200 kg ha-1 and in 1983 it reached 300 kg ha-1 (equivalent to 118 kg N, 35 kg P, and 85 kg K per hectare). At the same time the average yields of the main agricultural crops increased considerably between the periods 1951-1955 to 1981-1985: (tons/hectare): from 1.5 to 4.6 Mg ha-1 for winter wheat, from 2.1 to 6.1 Mg
ha-1 for Maize, from 19 to 39 Mg ha-1 for Sugar beet, from 1 to 2 Mg ha-1 for Sunflower, and from 9 to 18 Mg ha-1 for Potatoes. In addition to the introduction of new, intensive, high-yielding crop varieties, full mechanisation and integrated pest management, the adequate nutrient supply of crops played an important role in this spectacular and "revolutionary" yield increase.

It was necessary and under the given economic conditions, including high state subsidy on fertilizers, economically rational to apply more fertilizers for three main reasons: (1) to satisfy the higher nutrient requirements of the new, intensive crop varieties; (2) to improve the poor nutrient status of Hungarian soils; and (3) to balance the decreasing rate of organic manure application, which was a consequence of the replacement of more or less evenly distributed farmyard manure from small-scale private farms by the liquid manure from "straw-free" technologies used in over-sized meat, milk or egg production plants. From the mid-70's the proportion of mineral fertilizers reached 75% of the total quantity of plant nutrients applied.

The tendency of fertilizer application in the Balaton catchment area shows characteristics similar to the Hungarian national average. The use of mineral fertilizers and farmyard manure in the 5170 km2 watershed was very low before World War II (Table 1). This was followed by a sharp increase in the late 60's and early 70's growing to 250 kg/ha in 1977, at which level the fertilizer application rate stabilized with an approximate ratio of N:P:K of 24:7:15.

Since the 70's, the quantity of N, P and K added as fertilizers, manures and plant residues exceeded plant uptake and the soils were enriched with plant nutrients. The average P balance of Hungarian soils from 1932 to 1991 (Table 2) shows that between 1965 and 1990 annual P surpluses of fertilizer P were accumulated. About 500 kg ha-1 of applied P was not utilized by plants. Part of this amount remains in the soil in various forms, increasing its plant-available P resources. Another part was lost, mainly by surface runoff from sloping terrains to low-lying areas or to waterways, lakes and reservoirs, with the well-known environmental side-effects such as unfavourable changes in the aquatic ecosystems, eutrophication, etc.

In the late 80's - due to the favourable democratic political changes - the concept of agricultural development changed considerably. Instead of total quantity, quality (exportability), efficiency, economy (based on a real and exact cost-benefit evaluation) and environmental compatibility became more and more important. The restructuring of the planned economy can be characterized by: (1) trends to privatization (mixed ownership: co-existence of private, cooperative and state sectors) and the consequent decrease in size of farming units and fields; (2) market-oriented production with special regard to efficiency, including considerable and rational input reduction; and (3) sustainability, the prevention or at least reduction of unfavourable environmental side-effects. According to the new concept, the high state subsidy was withdrawn from mineral fertilizers. High fertilizer prices and the unpredictable future of state farms and cooperatives resulted in a drastic reduction of fertilizer application to 37 kg ha-1 in 1991; and to 19 kg ha-1 in 1992. Despite a feeling by many new private farmers that systematic fertilizer application is not necessary because they have observed no serious yield reduction when P and K application was stopped on formerly over-supplied fields, this application rate is too low in the long-term and sooner or later will lead to serious yield reductions.

 

Table 1. Use of mineral fertilizers and farmyard manure in the Balaton catchment area (5170 km2) from 1931 to 1992.

 

 

Farmyard manure

 

Mineral fertilizers

Mineral fertilizers*
Years  

N

P

K
  
 

1000 Mg y-1

-------- 1000 Mg y-1 --------

kg ha-1
1931-1940 1023 0.03 0.14 0.02 2
1951-1960 968 1.55 0.66 0.70 15
1961-1965 941 6.64 1.96 2.27 57
1966-1970 1014 12.40 3.49 6.14 109
1971-1975 676 18.88 5.79 15.63 218
1976-1980 653 23.45 6.90 18.00 250
1981-1985 710 23.55 7.10 15.53 282
1986-1990 621 25.00 6.00 14.31 230
1991 370 6.28 0.51 1.24 37
1992   3.75 0.14 0.21 19

* for agricultural area without grasslands

 

P fertilizer application can be both beneficial and detrimental to the environment. In soils with poor natural P supply, P fertilizers and/or organic manures increase plant cover and yields. Surface runoff and soil erosion from the slopes of cultivated areas, however, are the most important causes for nutrient loading of surface waters and their eutrophication. The long-term application of P fertilizers may lead to the accumulation of trace contaminants, such as cadmium, in the soil (Kádár, 1992; Németh & Kádár, 1991; Mortvedt, Ch. 6).

The low solubility of P prevents - in most cases - leaching losses of native and added soil P and the contamination of groundwater resources by soluble P compounds. But flow through preferential pathways (such as cracks, root- and earthworm channels or large inter-aggregate soil pores), surface runoff, water and wind erosion may transport considerable amounts of insoluble P or even fine fertilizer particles to the surface and subsurface waters. Water erosion of soil has particular significance for the P load of Lake Balaton, because the Lake is surrounded by hills with a high percentage of arable land and maize fields, orchards and vineyards which, in many cases are cultivated up and down the slope. The lack of effective soil conservation practices on the highly erodable, often shallow, medium-textured soils with poor aggregate stability and degraded structure leads to serious nutrient losses from eroding slopes with nutrient accumulation in low-lying areas and sediments, and nutrient loading of tributaries and the lake itself.

The Zala River watershed occupies approximately half of the total catchment area of Lake Balaton (Figure 1), with the following distribution among slope categories: flat lands: 2%; < 5% slopes: 57%; 5-10% slopes: 25%; 10-15% slopes: 9%; >15% slopes: 7%.

The P balance of the arable lands in the Zala River catchment area (similarly to the whole country, Table 2) became positive in the mid-60's, continued to increase into the early 80's and stabilized at a level of about 30 kg ha-1 until the late 80's. A yearly 1% increase in the mineral P content of the plough layer from organic manure and mineral fertilizers, contributed between 20-23% to the total plough-layer P.

 

Table 2. Average yearly P balance (kg ha-1) of Hungarian soils, 1932-1991 (Kádár, 1979, 1987; Csathó, 1993).

  1932-1936 1960-1964 1971 1975 1984 1990 1991
               
Crop uptake 7 8 11 13 17 15 18
Applied as              
- Fertilizers - 5 16 28 29 9 2
- Manures 3 3 4 4 7 5 5
- Residues* - - 1 2 1 1 2
Total 3 8 21 33 37 15 8
Net budget -4 0 11 21 20 1 -10
Applied/Uptake % 47 106 200 262 221 106 45

 

* Mainly maize and sunflower stalks; winter wheat straw

 

The P enrichment of soils was shown in regular soil P tests (Egner-Riehm-Domingo AL-method). The AL-soluble P contents were rated in relative P supply categories taking into consideration soil properties which influence the solubility, mobility and plant availability of P compounds and the efficiency of applied P fertilizers. Before the positive P fertilisation balance, the native P status of soils was low due to total P contents of the parent materials and the cereal-farming practice which, during centuries, removed large amounts of P from the fields with grain and straw (Table 3). After 30 years of P fertilizer application the P supply of soils was considerably improved and at the end of the 80's more than half of the arable lands in the Balaton catchment area were well-supplied with plant-available phosphorus (Table 3). A summary of P balances and AL-soluble P contents between 1979 and 1982 for soils on various slope classes of the arable lands of the Balaton catchment area is presented in Table 4.

The P status and balance of arable land of the Zala catchment area was studied by Sisák (1993) (Table 5). Grasslands and forests were excluded because they are not fertilized, and plantations contribute less than 2 percent of the area. The following conclusions can be drawn from the data from about 2000 farm fields studied between 1984 and 1988: (1) the average yearly surplus in the nutrient balance was 55 kg N ha-1, 25 kg P ha-1 and 53 kg K ha-1; (2) the nutrient balance was most positive for sugar beet, maize, sunflower and autumn-seeded rape. Because maize, sunflower and sugar beet do not develop an efficient soil plant cover during the whole year, their highly positive nutrient balance represents a potential source of the nutrient loads of surface waters by surface runoff, erosion and sediment transport.

Table 3. Distribution of the relative P supply categories in the Balaton catchment area (in the percentage of the total investigated area) (Kovács, 1984; Baranyai et al., 1987; Kádár, 1992).

 

Years

 

P supply categories, in % of total investigated area
 

Low

Medium

Good

High
         

1900-1950

40-50

30-40

10-20

0-5

1970-1975

20-30

30-40

30-40

5-10

1978-1981

9

46

41

4

1982-1985

15

32

37

16

 

 

Table 4. P balances and changes in AL-soluble P content of soils between 1979 and 1982 on different slopes in the Zala River watershed
(Debreczeni & Sisák, 1990).

 

Slope

 

P balance

 

Changes in AL-P

%

kg ha-1y-1

mg kg-1

0-5

44

18.2

5-10

40

10.2

10-15

37

9.5

15-20

35

6.5

20-30

31

5.3

 

The role of agricultural land use in the P loading of the Zala River and Lake Balaton has been evaluated by several authors. The estimates show great differences and are not based on systematic and regular measurements and comprehensive nutrient monitoring. Estimates for point-source pollution, such as sewage outfall, are more reliable and mainly based on emission records. The evaluation of non-point sources is much more difficult and requires detailed studies and site-, source- and component-specific evaluation and interpretation. The main causes of the non-point P load of Lake Balaton are the transport of bound, absorbed, biologically immobilized and organic P by surface runoff, water and wind erosion, sediment deposition and redeposition, and suspension flow through preferential pathways. At present we do not have enough reliable data on the relative contribution of these processes in the P transport and on their role in the biogeochemical P cycle.

Various authors estimate fertilizer applications contribute between 1-3% (Debreczeni & Sisák, 1990) and up to 13% of the P load in Lake Balaton. All authors emphasize the particular significance of surface runoff and water erosion (comp. Figure 4.).

 

Table 5. Extension (in % of the total arable land), average yields and nutrient balance of the main crops in the Zala River catchment area, 1984-1988 (Sisák, 1993).

Nutrient balance
 

Extension

Yield

N

P

K
 

%

Mg ha-1

kg ha-1
Winter wheat 29.7 4.1 39 21 32
Winter barley 6.7 3.9 28 22 12
Oat 3.8 3.1 12 18 12
Spring barley 2.1 3.3 30 26 48
Maize 18.8 6.1 86 31 100
Silage maize 9.2 23.5 59 22 27
Sugar beet 2.2 33.7 120 56 147
Winter rape 7.2 1.8 84 29 83
Sunflower 6.0 1.7 46 35 95
Alfalfa 3.7 6.0 80 4 -13
Others 10.6 - - - -

 

MAGNITUDE AND SOURCES OF P LOAD TO LAKE BALATON

 

The crucial problem of the Lake is the continuing eutrophication and the low or even deteriorating water quality, such as temporary bacterial contamination at some of the beaches, or the deterioration of the reed belt. The cause of eutrophication problems are inevitably the external anthropogenic nutrient loads, and among them mainly the P load, P being the growth limiting factor in the lake.

 

BRIEF HISTORY OF ESTIMATING P LOADS

 

With the exception of the Zala River, where daily TP and TN measurements and weekly observation records are available for 15 years, determinations of the magnitude of nutrient loads have been rather uncertain (Jolánkai & Somlyódy, 1981). The reason for the uncertainty is that the monthly two samples taken in the regular monitoring programme do not reveal the actual load, but only fragments of it, since storm runoff events, which cause most of the total load from smaller creeks, will be missed. In addition, the Lake's unmonitored direct catchment area of roughly 800 km2 is suspected to contribute large, but mainly unknown, quantities of nutrients. Since most of the loads at present are from diffuse sources and since large gaps exist in the monitoring, rather rough estimation techniques have to be used in determining total loads.

One of the first published estimates of P loads to the Lake was for the average conditions of the period 1975-1979 (Jolánkai & Somlyódy, 1981) (Table 6). It was based on - at that time - rather infrequent regular monitoring, estimates of inputs by the Zala River (Joó, 1978), storm runoff measurements of Jolánkai (1977), and other estimates (Horváth & Kamarás, 1976; Kaurek & Kovács, 1976). This P load from the direct lake catchment was an underestimate, since detailed storm runoff measurements subsequently indicated that a few large runoff events contribute a large portion of the annual total load (Figures 5 and 6) (Jolánkai, 1993). Several attempts were also made to determine the primary sources of the estimated input loads, as carried by tributary streams or direct runoff. These efforts are based on daring assumptions, concerning the fate of P between the primary source and the final destination, the discharge to the lake.

 

 

Figure 5. Relationship between TP load and discharge, based on regular monitoring and flood sampling (Jolánkai & Somlyódy, 1981).

 

Table 6. Multi-annual average total P (TP) and biologically available P (BAP) loads of Lake Balaton. (Somylodi & van Straten, 1986).

 

Load components
  

 

Total lake

 

Basin I

 

 

Basin II

 

 

Basin III

 

 

Basin IV
 

----------------------------- Mg y-1 -------------------------
Tributaries

TP

BAP

166

92

84

47

56

35

22

9

4

1
Direct sewage TP=BAP 34 1 2 3 28
"Other sewage" TP=BAP 9 - - 9 -
Urban runoff

TP

BAP

58

17

4

1

13

4

14

4

27

8

Direct rural

runoff

TP

BAP

29

9

4

1

8

3

12

4

6

1
Atmospheric TP 18 1 4 6 7
pollution BAP 8 - 2 3 3

Total

external load

TP

BAP

 

314

169

94

50

83

46

66

32

71

41
 

------------------------ mg m-3 y-1 -----------------------

Volumetric

load,

TP

BAP

164

91

1153

624

201

110

113

55

88

51
 

----------------------- mg m-2 y-1 --------------------------
Load per lake surface area

TP

BAP

533

285

2486

1347

577

310

358

172

314

183

It is unfortunate that the relatively intensive field measurement programmes for load determination were discontinued between the early 80's and 1992. The likely reason for this stoppage are the substantial clean-up efforts of the early 80's (diversion and treatment of about 2/3 of the former sewage discharges, and the building of several pre-filter reservoirs, among them the reconstruction of the former Kisbalaton wetland at the mouth of the Zala River) which led the authorities to believe that further exact, systematic monitoring of P loads was not necessary. The Lake's water quality, however, showed little or no sign of improvement, making the determination of the remaining P loads more important than ever. Incidentally, all government documents stated clearly from the beginning that considerable improvement of the Lake's water quality could only be achieved with an at least 75% load reduction. The monitoring task is now more difficult since, with the treatment and diversion of sewage, diffuse sources are the dominant P load contributors (Table 7), and both measuring and modelling of the fate of P in the watershed and the Lake are needed more than ever. In the meantime, indirect estimates of the annual TP load of the Lake are based on water quality monitoring stations (15 inflowing - or pump lifted - streams and canals are monitored bi-weekly to monthly).

 

 

Figure 6. TP concentration vs. flow relationship in Tetves Creek, Lake Balaton.

 

 

THE FATE OF P IN THE CATCHMENT AREA AND IN LAKE BALATON

 

A major research project "Development of Regional Decision Support System for Managing the Water Resources and Water Quality of Large Lakes" was launched in 1991. Its main objective is the development of a computer-aided tool for synthesising past and present research and data, and coordinating field studies in support of the water management and pollution control decisions related to Lake Balaton, in particular to: (1) assess present and past conditions of the Lake, including efforts to determine nutrient (mainly P) loads which are the key factor in the eutrophication process; (2) review past and suggested future control strategies; (3) establish and continuously update a data bank of all available quantitative and qualitative data, including the creation of a Geographical Information System of the drainage basin (far from completion at present); (4) develop lake ecosystem models; (5) adapt a drainage basin model (called SENSMOD) to a GIS of the Lake catchment, with the main purpose of estimating P loads (with special regard to the Lake's own direct drainage basin that is not covered by the regular monitoring network); (6) evaluate statistically hydrological data in order to establish inputs for hydrological and pollutant transport models, including a model of the operation of reservoirs in the drainage basin; (7) intensify field measurements to determine the P load fraction due to storm runoff events (a component that is not reflected by the data of the regular monitoring network).

 

Table 7. Total P load estimates of Lake Balaton in the eighties.

 

Load

component

 

1981

 

 

1982

 

 

1983

 

 

1984

 

 

1985

 

 

1986

 

 

1987

 

 

1988

 

 

1989

 

 

1990
  --------------------------------- Mg y-1 ---------------------------------------

Streams and

canals

 155 160 153 160 171 149 169 92 81 70

Direct

sewage

 17 10 10 8 7.3 7.3 7.3 3.6 3.3 1.6

Other

sewage*

 4.7 5.5 7.7 9.1 5.8 5.1 1.1 0.6 7.1 3.4

Urban

runoff

 58 58 41 60 57 50 64 43 23 47

Other direct

diffuse load

 31 38 23 75 70 28 52 35 36 38

Atmospheric

deposition

 

18

 18 14 20 19 9 21 14 33 34
Total 284 290 248 299 330 248 313 188 184 193

 

In this project, tributary loads will be determined by making use of the stream monitoring data and making corrections in the domain of high flows with the help of concentration vs. flow, or load vs. flow relationships, wherever such relationships can be established on the basis of intensified storm-runoff measurement data. The method of this extrapolation is illustrated in Figure 7. Runoff-induced diffuse loads will be determined by two types of models:

 

LT0-T = C1 QT0-T1 + C2 QT1-T2 + LK(Q)+ C3 QT3-T4 + Lj(Q) + C4 QT5-T

 

Figure 7. Principle of calculating total load on the basis of continuous flow (Q) records and occasional quality sampling (C) data.

 

Some preliminary results (since the GIS is not finalized and the intensified measurement programme is still under way) of such model calculations show the estimate of P loads for the past two years (Table 8). The rather high load contribution of the direct lake catchment might partly be due to the underestimation of the load from the water courses. But it may also be realistic, since the near-shore area is a very densely built-up recreational zone with some two million summer visitors; and orchards, vineyards and arable land are cultivated along the slope down to the shoreline, in most cases without any erosion protection, such as inter-row vegetation. This direct catchment with its short runoff pathways has very high unit area loading rates, as shown by a limited number of in situ measurements.

The above considerations indicate that control measures should concentrate on non-point source pollution, especially in the recreational and agricultural region around the lake. They should be focused on control of erosion and urban storm drainage including treatment in prefilter storage reservoirs; on sewerage and sewage treatment or diversion; and on efficient waste collection and disposal including sewage sludge disposal. Other control measures include the provision of public conveniences in the overcrowded beach resorts, regulation and diversion of motor traffic including motor boats, measures for saving the remnants of the reed belt (a natural means of protection) etc. In-lake control measures might include the dredging of the upper approx. 10 cm P-enriched layer of the bottom deposit in the most problematic parts of the lake basin.

 

Table 8. Recent total phosphorus load estimates for Lake Balaton.

 

 

Basin I

 

Basin II

 

Basin III

 

Basin IV

 

Total
 

--------------------------------- Mg y-1 -------------------------------

1991

Tributaries

 

49.69

 

28.6

 

11.88

 

2.32

 

92.5
Direct sewage

*
 0.463 0.491   0.954
Direct runoff:          
urban 5.48 7.0 9.02 9.31 30.83
arable 5.66 17.65 15.42 9.6 48.33
pasture   2.05 20.5   4.1
forest 0.23 0..57 0.15 0.27 1.23
vineyard 7.71 0.04 8.98 7.19 30.93
Total runoff 19.8 34.31 35.62 26.37 115
Atmospheric 0.813 2.567 2.857 2.715 8.95
Total 70.3 65.9 50.8 31.4 218

 

1992

          
Tributaries 39.97 24.51 9.55 6.33 80.36
Direct sewage

*
 0.501 0.338   0.839
Direct runoff:          
urban 4.84 7.07 9.25 8.88 30.06
arable 5.68 17.87 15.16 8.16 46.88
pasture   2.19 2.05   4.24
forest 0.24 0.61 0.14 0.21 1.21
vineyard 7.76 7.49 8.8 5.57 29.62
Total runoff 18.52 35.23 35.4 22.82 112
Atmospheric 0.642 2.026 2.764 3.39 8.82
Total 59.13 62.27 48.05 32.54 202

(*) missing data

 

 

CHANGES IN THE AQUATIC ECOSYSTEMS OF LAKE BALATON DUE TO NUTRIENT LOAD AND THEIR FUNCTION IN THE P-CYCLE

 

CHANGES IN THE PHYTOPLANKTON

 

Increases in phytoplankton cell numbers and biomass had already been reported for the period 1945-1951, when the Lake was still in an oligo-mesotrophic state. In 1961-1963 the primary production of the phytoplankton measured by 14C was similar in the whole lake, at 70-100 g C m-2y-1. In the following decade the production in the four basins was 830, 301, 274 and 182 g C m-2y-1, respectively (Figure 8), i.e. there was an 8-fold increase in the first basin, a 3-fold one in the central part of the Lake and the production doubled in the eastern basin (Herodek, 1977). Due to the self-shading effect of the phytoplankton in the first basin there was no photosynthesis below 2 m in calm periods. This light deficiency eliminated the rooted submerged macrophytes from the deeper regions.

 

 

Figure 8. The increase of primary production in L. Balaton in space and time.

 

The spring phytoplankton consists of diatoms (Cyclotella bodanica, Cyclotella ocellata, Nitzschia acicularis, Synedra acus etc.). In summer, Ceratium hirundinella used to predominate over the entire Lake, but blooms of blue-green algae appeared from 1965 onwards in the western part of the Lake, and became regular since 1973. Anabaena flos-aquae, Anabaena spiroides and Anabaenopsis raciborskii, i.e. filamentous nitrogen-fixing blue-green algae, form the bulk of the mass.

Chlorophyll-a measurements started in 1973, and showed an increase in the next nine years (Figure 9). The yearly average chlorophyll-a concentration in the four successive basins was 35, 27, 15 and 9 mg m-3 in 1979-1983. Thus, according to the OECD categories, the first basin was hypertrophic, the second was near the limit of hypertrophy, the third was eutrophic, and the fourth was in a transient state from mesotrophy to eutrophy. The summer maxima were three times higher than the yearly average values.

Strong fluctuations of water quality from one year to another can be explained by the dependence of the algal growth on the weather conditions. Blue-green algae prefer long, calm summers. In 1982 the algal bloom (caused by the blue-green Anabaenopsis raciborskii) had already developed by mid-summer in the first basin, and spread successively to the entire Lake area by the end of August. The chlorophyll content reached the hypertrophic level in the first three and rose to the eutrophic level in the fourth basin. Complaints about "dirty" water became wide-spread around the lake.

 

 

 

Figure 9. Long-term changes of chlorophyll-a in the four basins of Lake Balaton.

 

THE LIMITING NUTRIENT PROBLEM AND THE PHOSPHORUS METABOLISM IN THE LAKE

 

Phytoplankton growth is limited by the nutrient in shortest supply relative to its needs. Algae need many elements, but only C, Si, N and most frequently P are reported to be limiting in lakes. C limitations were found only in heavily polluted lakes with very soft water. In view of the high hydrocarbonate content of Lake Balaton water, C can safely be neglected as a limiting factor. In a number of lakes the growth of diatoms is limited by the exhaustion of silicate from the water. In Lake Balaton the silica-containing materials are frequently resuspended from the bottom, and the dissolved silicate level remains high even at the end of the spring diatom bloom, so that Si can also be deleted from the list of potential limiting factors.

The N:P ratio in the streams used to be higher, and in communal waste waters lower than the ratio (7:1 by weight) in which these elements are required by algae. Lakes under temperate climates are, in their natural conditions, most frequently limited by phosphorus, and nitrogen becomes a limiting factor as a consequence of greater P pollution. In Lake Balaton the concentrations of both TP and TN decrease from the first towards the fourth basin (Figure 10). The average TP concentration between 1979 and 1982 was 95 mg m-3 in the first and 30 mg m-3 in the fourth basin, with corresponding TN figures of 1653 and 761 mg m-3. The entire lake was accordingly in the range implying P limitation, though the first basin approached a transitional state. The TN/TP ratio can only be a rough guide, the actual limiting factor depending on a number of circumstances, among which the composition and supply rates of the TP and TN play important roles.

 

 

 

Figure 10. Total P and N concentration gradients along the longitudinal axis of Lake Balaton (averages for 1979-1983). (Research Centre for Water Resources, unpublished).

 

About half of the TN is present in dissolved form. The concentration of ammonia in the water is below 40 mg N m-3 without seasonal changes. In contrast, the concentration of nitrate displays a typical annual cycle. In early spring the concentrations of NO3-N in the first and fourth basin may rise to as much as 900 and 300 mg m-3, respectively, and drop rapidly during the spring diatom bloom. Levels below 40 mg m-3 are typical in summer. Nitrate concentration rises again from late autumn. Accordingly, the TN:TP ratio fluctuates with a strong decrease in summer especially in the western part of the Lake. Here nitrogen may become limiting for a period, which explains the appearance of the N-fixing filamentous, heterocystic blue-green algae. In the water of the western basin, the summer nitrogen fixation by the blue-greens is estimated at 7-13 g N m-2 according to the results of acetylene-reduction analyses. Although this amounts to only about one quarter of the total annual external load of the area, in summer it is still the main source of external N. In the central and eastern parts of the Lake, N fixation is insignificant. Nitrogen is also sorbed in the sediment at an estimated rate of 4 g N m-2 annually, based on measurements performed with intact sediment cores.

The total dissolved P (SP) and the dissolved reactive P (DRP) range from 10 to 30 and from 2 to 6 mg m-3, respectively. Algal tests were used to determine the biological availability of SP. Filtered lake water, enriched with a P-free nutrient solution, and inoculated with a pure alga culture showed no growth of algae, i.e. the amount of algae-available P was below the detectability limit (2 mg P m-3) of the method. This implies that most of the SP is unavailable to the algae.

Phosphate uptake was studied by using 32PO4 (Istvánovics & Herodek, 1985), and followed Michaelis-Menten kinetics

 

v = Vmax S / (K+S)

 

where v = the uptake rate; Vmax = the highest uptake rate attainable at nutrient saturation; K = the half-saturation constant; S = the substrate concentration, i.e. the sum of the concentration, Sn present originally in the water and of the concentration added A, so that S = Sn + A. Using experimental results, the sum of K + Sn , Vmax and the turnover time T of the nutrient dissolved in the water can be estimated. The sum of K + Sn found this way was consistently lower than the DRP determined by chemical analyses, demonstrating that in Lake Balaton water, the orthophosphate concentration is in reality much lower than the already low level measured by the ammonium molybdenate method. The turnover time of the dissolved orthophosphate was only a few hours in winter and a few minutes in summer. This is mainly the consequence of the very low P concentrations, and points to P as the limiting factor. The magnitude of Vmax depends on the quantity of algae and on their nutrient deficiency. The absolute value of Vmax was higher in the first than in the fourth basin, since phytoplankton is much more abundant there. However, Vmax per unit biomass was higher in the fourth basin, indicating a greater P deficiency there. The ratio of photosynthetic rate in the optimally illuminated water layer to the maximum P uptake of the phytoplankton is also used to characterize P deficiency. According to this index, the phytoplankton in the fourth basin is extremely, and in the first basin moderately P deficient.

The impact of nutrient enrichment was also studied in lake enclosures (Istvánovics et al, 1986). Plastic cylinders attached to inflated floating rings at the surface and anchored to a stainless steel hoop in the sediment were used to isolate water columns of 3 m diameter together with the sediment in the fourth, mesotrophic basin of the lake. Phosphate and/or nitrate were added to these daily, at rates similar to the external load in the hypertrophic part of the Lake. In the enclosure to which PO4 and NO3 were added daily at the rates of 5 mg P m-3 and 35 mg N m-3, chlorophyll content and primary production of the phytoplankton quickly attained hypertrophic levels. In the enclosure to which nitrate alone was added, no growth in the phytoplankton mass was observed. In the enclosure enriched with phosphate alone, the phytoplankton increased first moderately, then nitrogen fixing blue-greens appeared and increased the primary production, chlorophyll-content and biomass to levels characteristic of the hypertrophic part of the Lake. Consequently, the mesotrophic water in Lake Balaton could be made hypertrophic by adding phosphate alone.

The studies above suggest that Lake Balaton in its original state was P limited, and algal growth is still limited by the P supply in the greater part of the Lake. In summer the phytoplankton of the most polluted, western area becomes nitrogen limited for a period. In such times however, N-fixing blue-greens appear, to which the atmospheric nitrogen represents an inexhaustible source of supply. Consequently, phytoplankton growth in the hypertrophic part can be reduced also only by P load reduction.

 

SEDIMENT PHOSPHORUS

 

While in deep lakes most of the P may accumulate in the hypolimnion, in the shallow, unstratified Lake Balaton, nutrients are stored in the sediment. The internal loading originating from P release from the sediment may considerably delay the improvement of water quality once the external load is reduced.

About half of the sediment of the lake consists of CaCO3 (mainly calcite). The organic C content of the sediment is low, less than 2 mg g-1 in the eastern and central parts of the Lake, and not higher than 5 mg g-1 in the hypertrophic region. The sediment is frequently resuspended by waves, and the mineralisation is rapid in the warm, well oxygenated water, where the mild alkalinity favours bacterial decomposition.

The siltation rate is ca 0.7 mm y-1 in the largest part of the Lake, but it is one order of magnitude higher in the first basin due to the delivery of suspended solids by the Zala River and to the intensive lime precipitation. This may explain why, despite very different areal loadings, the TP contents of the uppermost 5 cm of sediment in the first and fourth basins are similar at 7000 and 5000 mg kg-1 of dry weight, respectively. In both parts of the lake, sediment P content decreases with depth, but more so in the first than in the fourth basin. The decrease is moderated by the mixing of the uppermost 10-20 cm of sediment due to water motions and bioturbation.

The P of this calcareous sediment was fractionated according to Heiltjes and Lijklema (1980): first the "loosely adsorbed P" was eluted with NH4Cl, the "iron- and aluminum bound P" by NaOH and the "Ca-bound P" by HCl. These fractions corresponded to 5-7, 3-20 and 50-80% of the TP, respectively. In the sediment of the fourth basin there was little iron-bound P, but it contained more Ca-bound P than the sediment of the first basin. The percentage of Ca-bound P increased with depth at all stations, indicating that with aging more and more P becomes bound to Ca.

The part of the P exchangeable under anaerobic conditions with the orthophosphate of the interstitial water has been determined (Herodek & Istvánovics, 1986) using isotope dilution. The reducible iron bound, and the loosely adsorbed fractions showed the highest exchangeabilities. Each fraction had a higher mobility in the hypertrophic region than in the mesotrophic one, and in both areas the mobility decreased in the upper 10 cm layer downwards to about 30% of the surface value (Figure 11). It shows that the older the sediment, the lower is the mobility of its P content. The higher exchangeable P content of the upper layer is due in part to the fact that the "plankton rain" settles first on the sediment surface, but in part also to the upward migration of P from the deeper, reduced layers. The amount of the exchangeable, i.e. mobilisable P in the uppermost 8 cm of the first and fourth basin is 7.8 and 1.3 g P m-2 corresponding to several years' external phosphorus loads.

Maximum PO4-P concentrations in the interstitial water are 200-300 mg m-3, and decrease upwards. The diffusion rate calculated from the concentration gradient is of the same order of magnitude as the external loading. Desorption of phosphate from resuspended sediment results in similarly high internal loading. Phosphorus release is further enhanced by reduction of the sediment surface and rising pH of the water, causing positive feedback during algal blooms (Istvánovics, 1988). Benthic organisms may also enhance P release from the sediment. Many processes are unknown or difficult to quantify, but in the opinion of experts the internal loading may attain or surpass the level of the external one.

 

 

 

Figure 11. Vertical distribution of exchangeable P in the sediments of Basin I and Basin IV (after Herodek & Istvánovics, 1986).

 

Phosphorus transport between the basins was calculated by multiplying the P concentrations measured monthly by the water through-flow of the same month. The retention of the biologically available P by the sediments of the four basins was calculated from a combination of the external loads, the input from the previous and outputs into the next basins (Herodek, 1984). The biologically available P trapped by the sediment of the individual basins ranged from 65 to 73 % of their loads. The outflow of biologically available P at the Sió canal is only 6% of the load on the Lake.

 

 

EUTROPHICATION MODELS

 

Eutrophication control programmes need predictions for the effect of nutrient load reduction on the trophic state. Balaton has the advantage that it consists of four basins, similar in all respects, with the exception of their loading and trophic state. Plotting the primary production or the chlorophyll-a concentration of the four basins vs. the net volume-specific external loadings resulted in a saturation curve (Herodek, 1984). This is suggestive of the fact that in the hypertrophic basin the external BAP loading should be reduced to 1/6 in order to reduce the primary production by half and to prevent the regular blooms of blue-green algae.

According to the OECD model:

 

Chl = 0.37 [Pj / (1+T)]0.79

 

where: Chl = the annual mean chlorophyll-a content of the water (mg m-3),
Pj = the concentration of total phosphorus in the inflow, (mg m-3) and
T = the residence time of water in the lake (years). The chlorophyll contents in the first and second basin of Lake Balaton are higher, in the third and especially in the fourth basin lower than the levels predicted by the model.

 

Figure 12. The Balaton Eutrophication Model (BEM) (after Kutas & Herodek, 1987).

 

While such empirical models are of real practical importance, they cannot allow for the effects of nutrient accumulation, resulting in changes in the internal loading. This may be a task for dynamic models. As part of a IIASA project on modeling and managing shallow lake eutrophication (Somlyódy & van Straten, 1986) several models have been developed, but only one of them has a sediment P compartment (Kutas & Herodek, 1987). In this Balaton Eutrophication Model (BEM) (Figure 12) the Lake is subdivided into four basins, the mass transfer between these being described by a special submodel. The external variables in the individual basin models are light intensity, temperature and the P and N loads. According to the model, the algae incorporate the nutrients. Some of the algae die off, and the resulting organic matter is decomposed by the bacterioplankton, thus liberating the inorganic P and N. Parts of the phyto- and bacterioplankton settle out, and are mineralized in the sediment. One part of the released N is returned to the water, another is denitrified to N2 and escapes the system. The phosphate liberated in the sediment is adsorbed on the sediment particles. Some of it is recycled into the water and some is transformed into unavailable forms and is thus excluded from the subsequent processes. This model and its simplified version (Herodek et al., 1988) were calibrated against a 10-year data set (1975-1985), and were run with different external loading rates for the 1985-2000 period. According to the results, at unchanged external loading the quality of the water will deteriorate further. A reduction of the loading to half of the level of the early 80's would be sufficient to arrest deterioration without initiating significant improvement. Considerable improvement could, however, be expected in each basin by reducing the loading to a quarter. These forecasts are, evidently, of approximate value only. Nevertheless, they emphasize that great efforts are needed to reverse the eutrophication in this Lake.

 

THE STRATEGY FOR PHOSPHORUS LOAD REDUCTION

 

Following the recommendations of the Hungarian Academy of Sciences in 1976 and 1982, the Council of Ministers issued an order in 1983 on a large-scale eutrophication control programme. By 1987 a canal system had been constructed, collecting 2/3 of the effluents of the lake shore sewage treatment plants, and diverting them from the water catchment area of the Lake (Figure 13). In the case of the sewage treatment plants, which are remote from the Sió canal, the Lake remained the recipient, but tertiary treatment was implemented to precipitate P from the effluent. Phosphorus is removed also from the sewage waters of Zalaegerszeg and Tapolca, i.e. the two biggest towns of the tributary catchment outside the recreational area. Important measures have been taken for controlling the loading from liquid manure and the waters leaching from manure dumps.

An interesting part of the programme is the reconstruction of the Kisbalaton. This area upstream of the present mouth of the Zala River formed a bay of Lake Balaton, which was drained when the Sió canal was opened in 1863, dropping the water level of the lake by 3 m. The first, 20 km2 large, reservoir was commissioned by 1985, and its "natural filter system" has had an effective favourable impact on the Lake's P load since 1988. It is only 1 m deep, and its water retention time is one month. The reservoir traps P similar to the other basins of Lake Balaton. The efficiencies of TP and orthophosphate removals are 50 and 90%, respectively (Szilágyi et al., 1990).

By these means, the Lake’s P load halved from the 1982 level. In accordance with the model predictions, this should arrest further eutrophication, but is not enough to improve water quality. Plans call for reducing loading further to 1/5 of the 1982 level. The western part of the Lake is expected to be upgraded from hypertrophic to eutrophic, without regular blooms of blue-greens. It is hoped that the greater central and eastern basins, which attract most of the tourist trade, will become mesotrophic and the algae will not perceptibly colour the water.

To attain this objective it is necessary to complete the second 50 km2 basin of the Kisbalaton, construct similar reservoirs on the other main tributaries, double the sewage treatment capacity, upgrade the P removal from sewage, reduce loads from urban runoff and from agricultural sources, and remove the uppermost 10 cm sediment layer in the first basin. In view of the scale of these efforts, it is important to realise that economic and legal aspects of the control of P transfers need as much attention as the biogeochemical and technical problems.

 

 

 

Figure 13. The sewage water diversion system and the Kisbalaton reservoir.

Phosphorus in the Global Environment.

Edited by H. Tiessen

© 1995 SCOPE. Published in 1995 by John Wiley & Sons Ltd.

Back to table of contents

Last updated: 12.07.2001