Drought-induced changes in nutrient concentrations and retention in two shallow Mediterranean lakes subjected to different degrees of management
- 389 Downloads
While extensive knowledge exists on the relationship between nutrient loading and nutrient concentrations in lakes in the cold temperate region, few studies have been conducted in warm lakes, not least in warm arid lakes. This is unfortunate as a larger proportion of the world’s lakes will be situated in arid climates in the future due to climate change and a larger proportion will suffer from a higher frequency of intensive drought. We conducted a comprehensive 11–13 year mass balance study in two interconnected shallow Mediterranean lakes in Turkey, covering a period with substantial changes in climate conditions. The upstream lake was only affected by natural changes in nutrient loading, while the downstream lake was additionally influenced by sewage diversion and restoration by fish removal. Contrasting to experience from north temperate lakes we found an increase in in-lake concentrations of total phosphorus and inorganic nitrogen (ammonia as well as nitrate) in dry years despite lower external nutrient loading, and submerged macrophytes did not increase the nitrogen retention capacity of the lakes. In contrast, fish removal modulated the nitrogen concentration as in north temperate lakes, but the effect was not long-lasting. Our results suggest that climate warming reduces the nutrient retention capacity of shallow lakes in the Mediterranean and exacerbates eutrophication. Lower thresholds of nutrient loading for shifting turbid shallow lakes to a clear water state are therefore to be expected in arid zones in a future warmer climate, with important management implications.
KeywordsArid regions Nitrogen dynamics Biomanipulation Climate change Eutrophication Water level fluctuation
Riverine phosphorus and nitrogen are retained and nitrate is denitrified during the passage of lakes, the retention and loss percentage being particularly affected by hydraulic retention time (e.g. OECD, 1982; Bachmann, 1984; Lijklema et al., 1989), lake depth (Windolf et al., 1996) and trophic structure (Jeppesen et al., 1998). Several simple empirical mass balance models relating lake concentrations to external loading on an annual (e.g. OECD, 1982; Bachmann, 1984; Saunders & Kalff, 2001) or seasonal (Windolf et al., 1996) scale have been developed and extensively used for lake management purposes. However, these models were constructed mainly for north temperate lakes.
Hydrology is an important element in nutrient mass balances, not least for shallow lakes located in Mediterranean climatic regions. Such lakes are subjected to large variations in water level determined by naturally intra- and interannual variations in rainfall and groundwater discharge or re-charge in alternating drought and wet periods (Coops et al., 2003; Alvarez-Cobelas et al., 2005; Beklioglu et al., 2007). Such changes in hydrology may have major implications for nutrient dynamics and nutrient retention in lakes, not only directly through changes in loading and hydraulic retention time, but also indirectly through alterations in trophic structure, including changes in macrophyte coverage (Beklioglu et al., 2006; Beklioglu & Tan, 2008). Variations in temperature may also affect trophic structure and oxygen concentrations, the latter being a sensitive variable (in addition to temperature) for the phosphate release from the sediment (Jensen & Andersen, 1992; Søndergaard et al., 2003) and for nitrification/denitrification (Windolf et al., 1996; Eriksson & Weisner, 1997). However, only few mass balances exist for arid and semi-arid lakes (e.g. Gophen et al., 1990; Jossette et al., 1999; Romero et al., 2002; Romo et al., 2005) and studies considering the impact of varying hydrological conditions (e.g. drought and wet periods) from dry regions are particularly rare. This is unfortunate as a larger proportion of the world’s lakes are predicted to be subjected to major inter-annual and seasonal variations in hydrology in the future due to global warming. Moreover, lack of data from such regions makes global balances uncertain.
We conducted a comprehensive 11–13 year mass balance study on total phosphorus and inorganic nitrogen and phosphate in two interconnected shallow Mediterranean lakes in Turkey during wet and dry periods entailing substantial changes in climate conditions. Upstream Lake Mogan was only affected by natural changes in nutrient loading determined by variations in hydrology, while downstream Lake Eymir was additionally influenced by sewage effluent diversion conducted in 1995 and restoration by fish removal, which was undertaken twice during the study period (1998–1999 and 2006–2007).
Materials and methods
Lake Mogan is large and shallow (drainage area: 925 km2, surface area: 5.4–6 km2, Zmax: 3.5 m, Zmean: 2.1 m, 39°47′N 32°47′E). The lake is mainly fed by four main inflows, the Sukesen brook in the north, the Gölcük and Yavrucak brooks in the west and the Çölovasi brook in the east. The outflow runs into Lake Eymir through a canal and a wetland in the north.
Lake Eymir is a smaller downstream situated lake (drainage area: 971 km2, surface area: 1.20–1.25 km2, Zmax: 4.3–6 m, Zmean: 2.6–3.2 m; 39°57′N, 32°53′E). The lake receives most of its water from Lake Mogan (Eymir Inflow 1). The other inflow is Kışlakçı brook and the outflow is Eymir Out. The lake received raw sewage effluents for more than 25 years (Altınbilek et al., 1995) until diversion in 1995. Before diversion, Inflow 1 constituted 89% of the total external loading of TP. The fish stock was dominated by tench and common carp in 1997–1998. To reinforce recovery after nutrient loading reduction fish biomanipulation was undertaken during 1998–1999. Fifty percent of the stock of common carp and tench was removed and a ban on pike angling (Esox lucius) was introduced and had a major effect on the lake water quality: a 2-fold and 4-fold decrease in chlorophyll a and suspended solids, respectively, and a 2.5-fold increase in annual Secchi depth occurred (Beklioglu et al., 2003). Seasonal maximum coverage of submerged macrophytes was low (2.5%) before biomanipulation, but expanded after (40–90% coverage) (Fig. 2), being particularly high in the dry year 2001. However, 5 years after the biomanipulation, the fish biomass increased again to the pre-manipulation level, and in 2004 the lake shifted back to a turbid state with scarce submerged vegetation cover (Fig. 2) and higher biomass of both tench and carp, and lower biomass of pike (Özen, 2006). Following a new biomanipulation in 2006–2007 (removing tench and carp) the lake water quality improved (a 2-fold and 1.5-fold decrease in chlorophyll a and suspended solids, respectively, and a 50% increase in annual Secchi depth), though no major change was observed in macrophyte coverage. The phytoplankton community was dominated by chlorophytes during the clearwater period and by cyanobacteria (dominated by Anabaena circinalis, Rabenhorst 1863) during the turbid period. The zooplankton community was largely dominated by A. bacillifer and Daphnia pulex (de Geer 1778) from 1997 until 2003 and by rotifers from 2004 to 2006. In 2007 the zooplankton community was characterised by dominance of D. pulex and D. magna (Strauss 1820).
Water samples were collected from March 1997 to September 2007 in both lakes. Sampling was conducted fortnightly in spring, summer and autumn and monthly in winter. For further details about sampling and the methodology used for analyses of TP, NH4–N, NO2– and NO3–N (see Beklioglu et al. 1999, 2003).
δ V is the change observed in lake volume and B is derived from Eq. 1. We then summed I, P, O, E and G up to annual values.
The lake level, given in meters above sea level (m. a. s. l.), was recorded daily from a fixed gauge positioned at the southwest corner of Lake Mogan, and at the north-east shore of Lake Eymir. Data on water levels and discharge for the inflows and the outflows of the lakes were obtained from the General Directorate of Electrical Power, Resource Survey and Development Administration (EIE, 2007). The volume of the lakes was calculated using the bathymetric map constructed in 2000. Monthly air temperature, rainfall and evaporation data were recorded at a meteorological station located within the catchments of the lakes. The hydraulic residence time was estimated by dividing the lake volume (Vlake) by the volume of water entering the lake (Vin) per unit of time.
As for the water budget, the monthly nutrient budget was constructed first and then summed up to annual values. The DIN concentration for precipitation was 80 μg l−1 (Tuncer et al., 2001). The TP concentration of the precipitation was negligible (Altınbilek et al., 1995) and not included. Groundwater concentrations of TP and DIN were 24 and 3,423 μg l−1, respectively (Altınbilek et al., 1995). The in-lake TP and DIN concentrations were used for water entering the groundwater.
The two lakes were subjected to major interannual variations in water balance (Fig. 2a, b).
The annual mean lake volume ranged from 2.7 to 4.1 km3 (Fig. 3b) and the mean amplitude of the water level fluctuation was 0.9 ± 0.3 m a.s.l. from 1993 to 2007. The contribution of inflows varied from 0.2–16.6 km3 year−1(Fig. 3b). The precipitation contribution became critical in the low water level years (0.2–0.6 km3 year−1) (Fig. 3b). The outflow followed the pattern for Lake Mogan and completely dried out from 2004 and onwards (Fig. 3b). Loss by evaporation was 27–42% of the lake volume annually and the hydraulic residence time was 0.2–13.5 year (Fig. 2b).
The annual TP budget was calculated for a period of 13 years covering 1993–1995 and 1997–2007 (Fig. 3b). Before sewage effluent diversion in 1995, the loading from Inflow I, the major source of TP, was 5.2 g m−2 year−1 and the observed annual mean in-lake TP concentration was 686 μg l−1. Following effluent diversion, the TP loading decreased 16-fold and the in-lake TP concentration 2-fold, followed by an increase during 1999, coinciding with higher hydraulic loading. After fish removal, undertaken during 1998–1999, in-lake TP decreased to172 μg l−1, followed by a major increase during the low water level years (2001, 2004–2007) despite low external TP loading levels. The in-lake TP almost doubled, from 172 μg l−1 in 2000 to 311 μg l−1 in 2001 when the lake volume and water level were at their lowest. A major peak in in-lake TP of 528 μg l−1 coincided with an almost disappearance of submerged macrophytes (Fig. 2h). Following the second biomanipulation, in 2006, when the lake volume and water level exhibited a slight increase, in-lake TP decreased markedly to 243 μg l−1, but in 2007 when the water level was at its lowest and the residence time was at its highest, TP increased again to 337 μg l−1. The predicted annual mean in-lake TP concentration remained above the observed in-lake TP concentration during years with high loading, but was far lower than the level measured during the dry years (Fig. 3b). Net retention of TP in the lake ranged between −1.67 g m−2 year−1 in 1999 to 0.95 g m−2 year−1 in 2002 (Fig. 4b). Also in-lake SRP showed high among-year variability that largely followed the variation in TP (Fig. 5b).
Dissolved inorganic nitrogen budget
Mass balances were conducted for dissolved inorganic nitrogen for the periods 1993–1995 and 1997–2007 (Fig. 3b). The DIN load via inflows and the in-lake DIN concentrations were highest before the sewage effluent diversion. Following the effluent diversion, the DIN load via inflow fell to 0.76 g m−2 year−1 and the in-lake DIN concentration decreased 11-fold. Moreover, following fish removal a further 1.5-fold reduction was found. However, it increased again from 2001 and onwards despite low external loading. The highest total input was 22.7 g m−2 year−1, recorded before the sewage effluent diversion, and the lowest was 0.015 g m−2 year−1 recorded in 2007. The groundwater contribution varied between 0.019 and 0.85 g m−2 year−1. The contribution from precipitation became significant during the drought periods. In-lake nitrate and ammonium declined (and thus DIN) substantially following the drastic external loading reduction in the 90’s, but increased again following biomanipulation and recovery of the macrophytes and remained high during the dry years (Fig. 3b) coinciding with low or negative P retention on an annual basis (Fig. 4b). As for Lake Mogan a high proportion of in-lake DIN consisted of ammonium (Fig. 6b). The proportion of ammonium tended to be higher than in Lake Mogan, coinciding with an overall lower oxygen concentration. DIN loss in the lake was high but more variable relative to external loading than in Lake Mogan, and the variability is here also attributed to both variations in ammonium and to nitrate retention in the lake; negative retention occurring in three of the study years (Fig. 7).
The key finding of our study of these two eutrophic warm temperate lakes are: (i) the concentrations of TP and DIN (ammonium as well as nitrate) increased in dry years despite lower external nutrient loading, which contradicts observations from cold temperate lakes; (ii) unlike in cold temperate lakes, extensive growth of submerged macrophytes did not result in lower inorganic N concentrations, and (iii) as in north temperate lakes fish removal enhanced the N retention, but the effect was not long-lasting. These results suggest that climate warming will enhance the risk of eutrophication in shallow lakes in the Mediterranean climatic region in a future warmer and drier climate. This is supported by a comparative study of 10 shallow Turkish lakes showing an average 11-fold increase in chlorophyll a concentrations from the regular year 2004 to the dry year 2007 (Beklioglu & Özen, 2008).
Our study covered a period with substantial changes in precipitation and runoff that markedly affected the water and nutrient balances of the lakes. While inflows were the major nutrient sources in wet years, precipitation and groundwater inputs were most important in dry years. A 20 and 25% decrease in lake volume and a 5- and 10-fold increase in hydraulic residence occurred during the dry periods due to low inflows and high evaporation. These changes had major implications for the nutrient budgets, most dramatically in Lake Eymir due also to a construction of a new higher sluice gate in Lake Mogan in 2003, implying less water overflow to Lake Eymir.
In both lakes, the in-lake TP concentration was, as expected, relatively high in years with high external loading, that is the wet years, and in Lake Eymir also in years with high sewage input (before 1995). However, in-lake TP was also high in dry years with low external input due to high evaporation and higher internal loading per unit of lake volume, likely exacerbated also by a higher temperature-mediated release as seen in other studies (Jensen et al., 2006). The mass balances clearly show that during the dry periods, in-lake TP became more dependent on internal processes (evaporation and internal loading) than on the external loading. Accordingly, we found predicted in-lake TP concentrations using the Vollenweider equation (Eq. 4) to severely underestimate the annual mean TP concentrations in the dry years in both lakes, while it predicted reasonably well the concentrations in the wet year except after sudden changes in loading.
The TP balance and TP concentrations in Lake Eymir were likely also affected by biomanipulation. Thus, a 2.5-fold reduction in in-lake TP occurred following the 50% reduction in carp and tench biomass. Similar high reductions have been seen in other biomanipulated European lakes (Hansson et al., 1998; Søndergaard et al., 2007, 2008), which has been attributed to reduced fish resuspension and decreased internal loading of phosphorus, the latter mediated by reduced sedimentation of algae (less oxygen demand), less P bound in phytoplankton and increased benthic algae production (Hansson, 1990; Søndergaard et al., 2003). Colonization of submerged plants may have contributed as well, since the vegetation coverage increased from 6% in the pre-biomanipulation period to 50% of the lake surface area afterwards (Beklioglu et al., 2003; Beklioglu & Tan, 2008). However, in the dry years 2004 and 2005, Lake Eymir TP rose to levels comparable to those pre-biomanipulation, despite low external TP loading (fish biomass increased and macrophyte declined), followed by a new decline after the second biomanipulation in 2006.
For nitrogen, we only have information on dissolved inorganic N (DIN, nitrate and ammonium). In Lake Eymir, a sharp decrease in in-lake DIN followed the sewage effluent diversion. A fast response to external nitrogen loading reduction has been seen in many lakes (Jeppesen et al., 2005); reflecting low accumulation of N in the sediment compared to what is typical for phosphorus. However, fish removal likely contributed to the decline in DIN as seen in several whole-lake fish manipulation experiments in the north temperate zone (Jeppesen et al., 1998; Svensson et al., 1999). The role of fish biomanipulation seems supported by the decrease in in-lake DINs in Lake Eymir following the second event of fish removal (2006–2007) after years with higher DIN concentrations. It is evident that high macrophyte coverage was not associated with particularly low inorganic N in this lake. This contradicts results from north temperate freshwater lakes where submerged macrophytes enhance nitrogen removal by offering surfaces for nitrification and denitrification and by enhancing oxygen depletion during night, thereby stimulating denitrification in the sediment and on plant surfaces (Reddy & De Busk, 1985; Eighmy & Bishop, 1989; Eriksson & Weisner, 1997).
Notable is the higher ammonium and nitrate concentrations (and thus DIN concentrations) during the dry, warm years from 2001 and from 2003 and onwards in both lakes despite low external loading and even in years with high macrophyte coverage in Lake Eymir. While higher ammonium accumulation may be attributed in part to a lower oxygen concentration in the dry years it is somewhat surprising that nitrate is higher in dry years. It has often been argued that loss by denitrification is higher in warm lakes (Lewis, 1987), which should have resulted in lower concentrations in the warm dry years in the two lakes. In addition, the loss expectedly should have been amplified by the concurrent reduction in mean depth that at least for north temperate lakes often leads to increase in loss by denitrification (Windolf et al., 1996), likely as a result of enhanced contact between water and sediment. However, a recent lake climate gradient study of shallow lakes from the tropical to the cold regions in South America did not find nitrate concentrations to be particularly low in tropical lakes (Kosten et al., 2009). Actually, in their study high nitrate to ammonia ratios were found in several of the tropical lakes, indicating that denitrification was inefficient, possibly because of lack of high quality organic matter (Kosten et al., 2009). This was hardly the case in our two eutrophic lakes, which is supported by the fact that most of the nitrate entering the lakes was lost on an annual basis, thus nitrate entering during the winter season was lost during summer (Özen et al., unpublished data). Higher nitrate concentrations in the dry years are therefore an effect of evaporation rather than of a reduced capacity of denitrification. Reduced contact to the extensive reed zone around Lake Mogan (102 ha, or approx 15%) may have contributed to higher nitrate concentrations in dry years in this lake as Gökmen (2004) found a high water level to be important for efficient N retention in the reed beds surrounding the lake.
The higher TP and DIN concentrations in the dry years may have destabilised the clearwater state (Moss, 1990; Scheffer et al., 1993) in Lake Eymir, leading to disappearance of the submerged plants, though low water levels in dry years occasionally lead to higher macrophyte coverage (Gafny & Gasith, 1999; Havens et al., 2004; Beklioglu et al., 2006) or to no change in abundance (Özkan et al., 2010). In Lake Mogan, the coverage of submerged plants also decreased.
Our results show that the response of shallow lakes to changes in nutrient loading triggered by variations in hydraulic loading (also affected by climate) and also the effects of biomanipulation on nutrient retention and concentrations in shallow lakes differ from the expectations based on studies of cold temperate lakes (Jeppesen et al., 2009). The result is higher concentrations of both phosphorus and nitrogen in dry years, which enhances the risk of increased eutrophication sympthoms and thus higher turbidity of lakes. We can expect that drought events will occur more frequently in the future and with a higher intensity in semi-arid to arid Mediterranean lakes due to climate warming and changes in irrigation (Alcamo et al., 2007; Jeppesen et al., 2009). Lower critical nutrient loading levels for maintaining a clearwater state in shallow lakes are therefore to be expected in the arid zone in a future warmer and drier climate.
This study and AÖ were supported by a Middle East Technical University, BAP research grant, and the METU-DPT ÖYP programme of Turkey (BAP-08-11-DPT-2002-K120510), the EU projects EUROLIMPACS (www.eurolimpacs.ucl.ac.uk) and WISER (www.wiser.eu), by CLEAR (A Villum Kann Rasmussen Centre of Excellence project), The Research Council for Nature and Universe, Denmark (272-08-0406) and TUBITAK, BIDEB. We thank Anne Mette Poulsen for editorial assistance and Juana Jacobsen for technical assistance.
- Alcamo, J., J. M. Moreno, B. Novaky, M. Bindi, R. Corobov, R. Devoy, C. Giannakopoulos, E. Martin, J. E. Olesen, A. Shvidenko, M. Araujo, A. Dlugolecki, E. Jeppesen, J van Minnen, J de Ronse & J. Sweeny, 2007. IPCC WGII Fourth Assessment Report. Chap. 12: Europe.Google Scholar
- Altınbilek, D., N. Usul, H. Yazıcıoğlu, Y. Kutoğlu, N. Merzi, M. Göğüş, V. Doyuran & A. Günyaktı, 1995. Gölbaşı Mogan-Eymir Gölleri için su kanakları ve çevre yönetim planı projesi. Technical Report No: 93-03-03-04-01. Ankara: Middle East Technical University (in Turkish).Google Scholar
- Alvarez-Cobelas, M., C. Rojo & D. Angeler, 2005. Mediterranean limnology: current status, gaps and future. Journal of Limnology 64: 13–29.Google Scholar
- Bachmann, R. W., 1984. Calculation of phosphorus and nitrogen loadings to natural and artificial lakes. Verhandlungen Internationale Vereinigung für Theoretische und Angewandte Limnologie 22: 239–243.Google Scholar
- Beklioglu, M. & A. Özen, 2008. Ülkemiz sığ gollerinde kuraklık etkisi ve ekolojik tepkiler. In S. Kalaycı & M. E. Aydın (eds), the conference book of International Conference on Global Climate Change and Its environmental Impacts. Damla Press, Konya: 299–306 (in Turkish).Google Scholar
- DSI, 1993. Mogan gölü Limnolojik Etüt Raporu. Ankara: Technical Report. General Directorate of State Hydraulic Works (in Turkish).Google Scholar
- EIE, 2007. Mogan ve Eymir Gölleri havzasının hidrometeorolojik özellikleri Technical Report. Turkish General Directorate of Electrical Power, Resources Survey and Development Administration, Ankara (in Turkish).Google Scholar
- Eighmy, T. T. P. & L. Bishop, 1989. Distribution and role of bacterial nitrifying populations in nitrogen removal in aquatic treatment systems. Water Research 25: 947–955.Google Scholar
- Gökmen M., 2004. Modelling of Nitrogen Removal Efficiency of the Wetlands Surrounding Lake Mogan, MSc thesis, Middle East Technical University, Turkey.Google Scholar
- Gophen, M., S. Serruya & S. Threlkeld, 1990. Long-term patterns in nutrients, phytoplankton and zooplankton of Lake Kinneret and future predictions for ecosystem structure. Archiv für Hydrobiologie 118: 449–460.Google Scholar
- Hansson, L. A., H. Annadotter, E. Bergman, S. F. Hamrin, E. Jeppesen, T. Kairesalo, E. Luokkanen, P. Å. Nilsson, M. Søndergaard & J. Strand, 1998. Biomanipulation as an application of food-chain theory: constraints, synthesis, and recommendations for temperate lakes. Ecosystems 1: 558–574.CrossRefGoogle Scholar
- Jensen, J. P., A. R. Pedersen, E. Jeppesen & M. Søndergaard, 2006. An empirical model describing the seasonal dynamics of phosphorus in 16 shallow eutrophic lakes after external loading reduction. Limnology and Oceanography 51: 791–800.Google Scholar
- Jeppesen, E., J. Windolf, J. P. Jensen, T. L. Lauridsen, M. Søndergaard, K. Sandby & P. H. Møller, 1998. Changes in nitrogen retention in four shallow eutrophic lakes following a decline in density of cyprinids. Archiv für Hydrobiologie 142: 129–152.Google Scholar
- Jeppesen, E., M. Søndergaard, J. P. Jensen, K. E. Havens, O. Anneville, L. Carvalho, M. F. Coveney, R. Deneke, M. T. Dokulil, B. Foy, D. Gerdeaux, S. E. Hampton, S. Hilt, K. Kangur, J. Köhler, E. H. H. R. Lammens, T. L. Lauridsen, M. Manca, M. R. Miracle, B. Moss, P. Nõges, G. Persson, G. Phillips, R. Portielje, S. Romo, C. L. Schelske, D. Straile, T. Tatrai, E. Willén & M. Winder, 2005. Lake responses to reduced nutrient loading – an analysis of contemporary long-term data from 35 case studies. Freshwater Biology 50: 1747–1771.CrossRefGoogle Scholar
- Jeppesen, E., B. Kronvang, M. Meerhoff, M. Søndergaard, K. M. Hansen, H. E. Andersen, T. L. Lauridsen, M. Beklioglu, A. Özen & J. E. Olesen, 2009. Climate change effects on runoff, phosphorus loading and lake ecological state, and potential adaptations. Journal of Environmental Quality 38: 1930–1941.CrossRefPubMedGoogle Scholar
- Jossette, G., B. Leporcq, N. Sanchez & X. Philippon, 1999. Biogeochemical mass-balances (C, N, P, Si) in three large reservoirs of the Seine Basin (France). Biogeochemistry 47: 119–146.Google Scholar
- Lijklema, L., J. H. Jansen & R. M. M. Roıjackers, 1989. Eutrophication in the Netherlands. Water Science and Technology 2112: 1899–1902.Google Scholar
- Manav, E. & S. V. Yerli, 2008. An assessment on the trophic status of Lake Mogan, Turkey. Fresenius Environmental Bulletin 17: 3–8.Google Scholar
- ÖÇKK, 2002. Mogan Gölü Havzası Biyolojik Zenginlikleri ve Ekolojik Yönetim Planı. Ankara: Çevre Bakanlığı Özel Çevre Koruma Kurumu Başkanlığı (in Turkish).Google Scholar
- OECD, 1982. Eutrophication of waters. Monitoring, assessment and control. OECD, Paris.Google Scholar
- Özen, A., 2006. Role of Hydrology, Nutrients and Fish Predation in Determining the Ecology of a System of Shallow Lakes. MSc thesis, Middle East Technical University, Turkey.Google Scholar
- Turkish State Meteorological Service. http://www.meteoroloji.gov.tr/index.aspx.
- Vollenweider, R. A., 1976. Advances in defining critical loading levels for phosphorus in lake eutrophication. Memorie dell’Istituto Italiano Idrobiologia 33: 53–83.Google Scholar