Seasonal Water Column Stratification and Manganese and Iron Distribution in a Water Reservoir: The Case of Pinios Dam (Western Greece)

Abstract

Climate change and extreme events such as droughts, heavy rainfall and flooding can influence the water column stratification in reservoir dams, decrease storage capacity, increase sediment and pollutant loads and, as a result, affect water quality. The seasonal variation in the water column stratification of reservoirs is an important parameter for the study of dam life cycle as well as water management and use. In the present study a detailed bathymetric survey was carried out, and a digital elevation model (DEM) of the reservoir was constructed. Seasonal physicochemical monitoring data such as temperature, dissolved oxygen, pH and conductivity are presented. The seasonal thermal stratification was recorded, resulting in an isolated hypolimnion where anoxic layers formed below 17 m in summer and autumn. Manganese and iron concentrations exhibited values higher than 150 mg/L in the anoxic hypolimnion during summer and autumn, indicating solubilization from the sediment. The observed seasonal and depth-dependent variations in physicochemical parameters underline the reservoir’s susceptibility to eutrophication and metal mobilization, particularly during stratified periods. These findings are critical for designing management strategies to mitigate potential water quality issues.

1. Introduction

Reservoirs are artificial lakes that play a critical role in water resource management worldwide, providing essential services such as water supply for drinking and irrigation, flood control and hydroelectric power generation [1,2]. However, both artificial and natural lakes are susceptible to environmental and anthropogenic influences that can alter their ecological balance and drinking water quality. This balance is mostly controlled by thermal stratification, which is a key physical process that affects the distribution of dissolved oxygen as well as other physicochemical properties which in turn impact water quality and aquatic life [3].

Thermal stratification is characterized by the formation of distinct temperature layers within the water column. This phenomenon arises due to variations in water density with temperature, leading to the development of separate strata that can significantly influence aquatic ecosystems, water quality and biogeochemical cycles.

In monomictic and dimictic lakes, thermal stratification causes the development of distinct layers in the water column that can become isolated from atmospheric oxygen, leading to hypoxic or anoxic conditions in the deeper layers [4,5]. These conditions, often exacerbated by nutrient inputs and organic matter decomposition, can significantly affect water chemistry, particularly the solubilization mobility and exchange of manganese and iron from sediments [6]. Increased concentrations of metals, especially manganese (Mn), are of major concern in water reservoirs as they have been linked with health issues in both humans [7,8] and the environment [9]. Moreover, the isolation of bottom waters can lead to nutrient accumulation, which, upon mixing events, can fuel algal blooms. In the Xiangxi River tributary of the Three Gorges Reservoir, thermal stratification was linked to algal bloom occurrences, influenced by air temperature and water level fluctuations [10].

In natural water bodies such as lakes, thermal stratification typically manifests as a three-layered structure: the epilimnion (warm surface layer), the metalimnion (thermocline with rapid temperature change) and the hypolimnion (cold bottom layer). The stability and duration of these layers are influenced by factors including solar radiation, wind mixing and the lake’s morphology. For instance, deep lakes often exhibit stable stratification during summer months, which can lead to hypoxic conditions in the hypolimnion due to limited oxygen replenishment [11,12,13]. Conversely, shallow lakes may experience frequent mixing events, preventing the establishment of persistent stratification.

Artificial water bodies, such as reservoirs, also undergo thermal stratification that can differ from natural systems. In reservoirs, thermal stratification is primarily driven by solar radiation, which heats the surface water, making it less dense than the cooler, deeper water. This density difference inhibits vertical mixing, leading to the development of stratified layers. Several factors, such as meteorological conditions, reservoir bathymetry and morphology, operation, as well as water inflow and outflow influence the formation and stability of these layers:

Increased air temperatures enhance stratification strength, while wind-induced mixing can disrupt stratification by deepening the thermocline. For instance, a study on the Tianbao Reservoir in Southwest China found that a 10% rise in air temperature increased stratification strength by 18%, and a 3 m/s rise in wind speed deepened the thermocline by approximately 2.09 m [14]. The bathymetry and morphology of a reservoir, along with its operational schemes, significantly affect stratification. Deep reservoirs tend to have more stable stratification. Operational practices, like selective withdrawal, can manipulate stratification patterns. For example, in a monomictic reservoir, deep hypolimnetic withdrawals promoted a thicker epilimnion with lower thermal stability, while shallow withdrawals resulted in a narrower epilimnion with higher thermocline strength [15,16]. The temperature and volume of the water inflow and outflow can influence stratification. Cold inflows can underflow and settle in the hypolimnion, while warm inflows may interflow or overflow, depending on their density relative to the reservoir water. These dynamics affect nutrient distribution and oxygen levels within the reservoir. Stratification also complicates water treatment processes by altering the chemical composition of water withdrawn from different depths. Selective withdrawal strategies are often employed to manage water quality, balancing the needs for temperature, oxygen levels and contaminant concentrations.

Dissolved oxygen in temperate lakes and reservoirs has shown a decreasing trend during the last decades, and an increasing number of cases presenting seasonally hypoxic or anoxic deep-water layers has been reported [17,18,19], raising concerns about the water quality and ecological status of the systems. This is probably associated with climate change as rising global temperatures lead to the earlier onset and prolonged duration of stratification, intensifying hypoxic conditions in bottom waters [5]. A study on the Xin’anjiang Reservoir in China projected that surface water temperatures would increase by up to 3.8 °C by the end of the century, extending stratification periods by up to 36 days [20]. Moreover, changes in precipitation patterns and the increased frequency of extreme weather events can disrupt traditional mixing regimes, leading to unpredictable stratification patterns and associated ecological impacts.

Understanding the mechanisms and implications of thermal stratification in both natural and artificial water bodies is crucial for effective water resource management, ecological conservation and the mitigation of issues such as algal blooms and hypoxia. This brings a necessity to modeling and simulating thermal stratification in water reservoirs, aiding in management decisions. These models incorporate detailed vertical temperature profiles and reservoir geometries to simulate stratification dynamics accurately. For example, a multilayer reservoir stratification model applied to 1400 reservoirs in the United States demonstrated improved predictions of surface and outflow temperatures, enhancing the understanding of anthropogenic impacts on hydrological and ecological cycles [21].

In Greece these models are rare, especially in artificial reservoirs, because modeling thermal stratification variability is a complex process influenced by climatic, morphological and operational factors.

This study aims to investigate the variability of water column stratification in the Pinios Dam reservoir, an artificial lake located in Western Greece, along with the seasonal fluctuations in key physicochemical parameters, including manganese (Mn) and iron (Fe). To achieve this, a detailed bathymetric survey was carried out, and four monitoring stations were established. Seasonal profiles of the water column were recorded, measuring parameters such as pH, temperature, conductivity, dissolved oxygen and concentrations of Mn and Fe. The findings have important implications for water quality, ecosystem dynamics and reservoir management, highlighting the need for a thorough understanding and the development of proactive strategies to address potential adverse effects—particularly in the context of climate change.

2. Materials and Methods

2.1. Study Area

The study area is the Pinios Dam reservoir, which is located at the Ilia Regional Unit Northwest Peloponnese, Greece (Figure 1). The Pinios Dam is a heterogeneous, earth-filled structure with a height of 50 m and a length of 2175 m. It is one of the oldest and longest dams in Greece and has been in operation for 57 years (construction completed in 1968). The dam’s reservoir is an important multipurpose water body as it serves both drinking and irrigation purposes in the region. It covers an area of approximately 20 km² and collects surface water from a 673.41 km² catchment area, with a designed total capacity of 420 × 10⁶ m³ and a maintained capacity of 50 × 10⁶ m³ to support ecological balance [22]. The inflow of the reservoir is mainly located in the Southeastern, Eastern and Northeastern parts of the reservoir (Figure 2).

The geological framework of the study area is primarily composed of Alpine bedrock, including formations from the Ionian, Gavrovo-Tripolis and Pindos geotectonic zones of Greece, consisting of flysch and limestone. Additionally, Neogene and Quaternary sediments are present in the surrounding catchment area [23]. Soil erosion processes are particularly significant in the region, primarily due to the steep slopes and secondarily due to the repeated wildfires that have affected the hilly areas surrounding the artificial lake [24]. More specifically, the complex topography, combined with land use changes—such as those resulting from the extensive 2007 wildfires that burned over 200 km² of forest and agricultural land within the Pinios Basin—intensifies post-fire soil erosion and sedimentation processes, leading to increased sediment transport into the reservoir. [23,25]. Previous research has identified elevated concentrations of manganese (Mn) and iron (Fe) in the reservoir’s water column, particularly during summer stratification periods, raising concerns about water quality and reservoir management. These elevated levels are attributed to the geological formations within the Pinios Dam drainage basin, which contribute sediments and geogenic elements such as Fe and Mn to the reservoir through runoff, weathering and erosion [22].

2.2. Methods

The bathymetric survey of the reservoir was conducted using two methods: (a) an unmanned autonomous surface vessel and (b) a conventional floating vessel (Figure A1). The technical specifications of the unmanned surface vehicle iUSV-170 (Intelligent Machines, Patras, Greece). The HydroLite-Plus Dual Frequency (Seafloor Systems, Inc., El Dorado Hill, CA, USA) single-beam echosounder and Leica GS07 (Leica Geosystems, Heerbrugg, Switzerland) instruments were used for data collection and recording. For the conventional bathymetric survey, a small polyester boat was equipped with a Lowrance 50/200 kHz 600 W transducer and a Lowrance HDS central unit (Lowrance Electronics Co., Seattle, WA, USA) with NMEA-182 output connected to a Panasonic CF-D1 (Panasonic, Tokyo, Japan) field laptop with specialized recording software (Special Devices LTD, Patras, Greece, Version 1.0) (Figure A1).

Water samples were collected from the artificial reservoir of the Pinios Dam at four predetermined stations (ST1, ST2, ST3 and ST4), strategically selected to provide comprehensive coverage of the study area (Figure 2). These stations were located along the reservoir’s main axis, with water depths ranging from 12 m to 36 m (ST1: 36 m, ST2: 28 m, ST3: 20 m, ST4: 12 m) when the reservoir was at its maximum water capacity. The maximum capacity of the reservoir is reached at +93 m (elevation above sea level), and the minimum water level is at +51 m. During the sampling period, the water level in the reservoir varied from +79.8 m to +88.0 m, presenting the lowest level in October 2023. Sampling occurred seasonally, between February 2023 and October 2023. Samples were collected at 4 m intervals from the surface to the bottom using a Ruttner-type water sampler, yielding a total of 74 samples for analysis.

Physicochemical parameters (pH, temperature, dissolved oxygen and conductivity) were measured using an In-Situ Aqua Troll 6000 portable multi-parameter probe (In-Situ Inc., Fort Collins, CO, USA). Vertical profiles were taken every season in each monitoring station at approximately 10 cm intervals.

Water samples were analyzed in the laboratory for manganese (Mn) and iron (Fe) using a DR3800 Hach Lange (Hach Lange GmbH, Berlin, Germany) spectrophotometer according to LCW532 [26] and LCK321 [27] water quality analysis procedures. All laboratory analyses were carried out at the Laboratory of Sedimentology in the University of Patras, which is accredited by the Hellenic Accredited System according to ISO 17025/2017.

3. Results

3.1. Physicochemical Parameters

Seasonal temperature variation indicated that the Pinios Dam reservoir is a typical warm monomictic lake, with winter mixing occurring in February (Figure 3). From February onwards, the surface layer of the lake gradually warmed, and by May, the stratification of the water column into an epilimnion, metalimnion and hypolimnion was clearly observed (Figure 3). The strongest thermal stratification was recorded in August, when the epilimnion extended to a depth of 15 m, and the metalimnion was observed between 15 and 25 m (Figure 3 and Figure 4).

The seasonal variation in surface temperature followed the typical seasonal cycle observed in lakes of temperate geographic regions at low altitudes (Figure 3 and Figure 4). In the surface layer up to a depth of 15 m, the highest temperature was recorded in August (Figure 3 and Figure 4). The maximum surface temperature of the reservoir reached 25.1 °C in August 2023, while the minimum temperature of 12.0 °C was recorded in February 2023. No substantial temperature differences were observed among the sampling stations. A small seasonal temperature variation was also recorded in the deeper layers, where the fluctuation range was approximately 2 °C (Figure 3). Conductivity was lower during winter and higher in summer and autumn (Figure 3).

The dissolved oxygen concentration in the reservoir was influenced by thermal stratification (Figure 4 and Figure 5). During the winter mixing period (February), the water column was well oxygenated from the surface to the bottom, with concentrations exceeding 8.0 mg/L (Figure 3). From May onwards, the epilimnion (up to 17 m depth in October) remained well oxygenated throughout the study period, with average concentrations ranging between 6.5 and 8.5 mg/L (Figure 3). In May, a gradual depletion in the oxygen concentration was observed below 10 m. This decline became more pronounced in August and October, when anoxic conditions (a dissolved oxygen concentration below 0.5 mg/L) were observed at depths greater than 17 m in August and 19 m in October (Figure 5).

The pH in the surface layer of the reservoir (0–10 m) ranged from 8.2 to 8.7 during the study period. The highest values within the water column were recorded in the 5–10 m layer (Figure 3). Below the epilimnion, pH generally decreased with depth. This decline was more pronounced in August and October, when intense hypoxic/anoxic conditions were observed in the hypolimnion (Figure 3). The average pH value in the hypolimnion (below 25 m) ranged from 7.6 to 8.1.

3.2. Manganese and Iron Concentration

Manganese and iron concentrations exhibited obvious depth-dependent trends. Mn concentrations ranged from 0.01 mg/L to 0.198 mg/L, while Fe varied from 0.01 mg/L to 0.204 mg/L (Figure 6,

Table A1. Mn and Fe concentrations in the four monitoring stations in the Pinios reservoir from February to October 2023.

		February 2023		May 2023		August 2023		October 2023
	Depth	Mn	Fe	Mn	Fe	Mn	Fe	Mn	Fe
	(m)	(mg/L)	(mg/L)	(mg/L)	(mg/L)	(mg/L)	(mg/L)	(mg/L)	(mg/L)
ST-1	4	0.059	0.038	0.041	0.011	0.024	0.019	0.061	0.027
	8	0.025	0.033	0.049	0.012	0.025	0.027	0.020	0.033
	12	0.047	0.043	0.029	0.017	0.011	0.033	0.057	0.037
	16	0.053	0.054	0.055	0.020	0.015	0.028	0.059	0.056
	20	0.049	0.095	0.036	0.024	0.102	0.108	0.099	0.076
	24	0.039	0.087	0.045	0.018	0.134	0.198	0.154	0.104
	28	0.043	0.075	0.051	0.046	–	–	0.198	0.174
	32	0.016	0.094	0.018	0.054	–	–	–	–
	36	–	–	0.058	0.048	–	–	–	–
ST-2	4	0.048	0.022	0.032	0.015	0.066	0.020	0.016	0.032
	8	0.023	0.042	0.049	0.022	0.013	0.017	0.054	0.026
	12	0.014	0.035	0.062	0.019	0.056	0.027	0.025	0.031
	16	0.049	0.087	0.033	0.023	0.015	0.176	0.011	0.015
	20	0.022	0.100	0.037	0.032	0.145	0.216	0.024	0.032
	24	0.053	0.097	0.033	0.024	–	–	0.184	0.114
	28	0.031	0.088	0.055	0.034	–	–	0.163	0.204
ST-3	4	0.012	0.018	0.045	0.022	0.012	0.020	0.053	0.015
	8	0.018	0.020	0.022	0.018	0.048	0.034	0.034	0.019
	12	0.016	0.017	0.049	0.039	0.036	0.029	0.054	0.021
	16	0.065	0.022	0.059	0.042	0.048	0.165	0.023	0.028
	20	0.038	0.034	0.016	0.044	0.129	0.203	0.145	0.032
ST-4	4	0.057	0.019	0.054	0.049	0.032	0.014	0.014	0.017
	8	0.028	0.02	0.035	0.046	0.046	0.018	0.023	0.029
	12	0.013	0.017	0.051	0.050	0.026	0.012	–	–
MIN		0.01	0.02	0.02	0.01	0.01	0.01	0.01	0.02
MAX		0.06	0.10	0.06	0.05	0.15	0.22	0.20	0.20
AVERAGE		0.04	0.05	0.04	0.03	0.05	0.07	0.07	0.05

). For manganese, the lowest concentration values were observed during February where the water column was mixed and well oxygenated. A gradual increase in the Mn concentration was observed in the hypolimnion from May until October with the highest values in the anoxic layer (Figure 6). For iron, elevated values were found all year round in the hypolimnion compared to the epilimnion, but the same temporal trend as in the manganese was also observed (Figure 6).

4. Discussion

The Pinios reservoir showed a typical temperate thermal stratification where the water column is strongly stratified for at least six months (May–October) during the year. Thermal stratification isolates the upper part of the water column from the hypolimnion, thus causing a progressive deoxygenation of the deep layers. The duration and intensity of thermal stratification are among the most important factors affecting the dissolved oxygen dynamics in lakes [19], along with eutrophication and climate change [28]. The dissolved oxygen concentration has been considered the most important factor affecting the biological and biogeochemical processes in aquatic ecosystems, such as the distribution of organisms, geochemical processes of nutrients [29,30] and metal kinetics [31,32]. In the Pinios reservoir, the deoxygenation process of the deep layer (below 15 m) seems to begin prior to the establishment of thermal stratification, as the dissolved oxygen concentration in the hypolimnion was reduced by 4 mg/L from February (9 mg/L) until May (5 mg/L), without a substantial increase in temperature, which could affect the solubility of oxygen. The concentration of 5 mg/L in dissolved oxygen is considered as the first critical threshold for the distribution of sensitive organisms [33] and the concentration of 2 mg/L (hypoxia) as the threshold below which the environment becomes unsuitable for most organisms [5], although this threshold has been considered as an underestimation due to the high variance of hypoxic tolerance among different organisms [34]. In this study, the hypolimnion, below 16 m, became hypoxic sometime between May and August, although more frequent profiles are necessary to estimate the oxygen depletion rate. Anoxia (i.e., an oxygen concentration below 0.5 mg/L) was observed below 17 m depth after August, although the exact duration of the anoxic hypolimnion cannot be determined from the seasonal sampling frequency of this study. Anoxia in the hypolimnion of the Pinios reservoir has also been reported in a previous study [22], a fact showing that the problem is relatively constant, but there are very few available data concerning the nutrient concentration and distribution [22] for evaluating the trophic status of the reservoir. Consequently, it is important to establish a more frequent monitoring protocol in the reservoir, as on the one hand the extent and duration of anoxia is directly linked with water quality [35], and on the other, there is strong evidence of widespread hypoxia and anoxia in temperate lakes due to extended thermal stratification [5].

Although the frequency of sampling in the present study was not optimal, the temporal variation in Mn and Fe was clearly associated with the presence of anoxia, presenting the highest values in the anoxic hypolimnion from August to October as was also reported during 2016 [22]. The values measured in the present study during summer and autumn in the hypolimnion are in the same range as the values reported for the same season in 2017 and 2018 [22]. This result is not unexpected as after the establishment of anoxic conditions in the hypolimnion, the most reduced and soluble forms of Mn and Fe diffuse from the sediment and accumulate in the water column [31,36,37,38]. The kinetics of Mn and Fe in aquatic environments depend on complex biochemical processes as well as on the physicochemical properties of the water [37], and as a result, specific studies are required to understand the cycling of these metals in lakes and reservoirs.

In any case, high concentrations of these metals are unwanted, especially in reservoirs where the water is being used for public utilities. In the Pinios reservoir, during August and October, Mn concentrations exceeded 100 μg/L in the samples below 20 m, and the highest value was 198 μg/L. These values are more than two times higher than the accepted level of 50 μg/L which is recommended for drinking water in the European Union [2], and as such, water treatment is imperative. In the case of the Pinios reservoir, the water treatment plant underneath the dam is supplied by the outlet of the dam gate at elevation level of +61 m, a depth which is in the metalimnion during the warm season and presented anoxia along with high concentrations of Mn and Fe. The supply of the treatment plant from the oxygenated epilimnion could solve the problem, but during construction this situation had not been anticipated, and only one outlet was constructed. As a result, water treatment is the only solution in the Pinios reservoir, which, to date, has included chlorination, screening, filtration and ozonation. Yet, there have been cases in the past where very high concentrations of Mn and Fe exceeded the treatment capacity limits of the Pinios treatment station, causing temporary cessation in water supply [22]. The improvement and/or expansion of the water treatment plant has already been stated as a necessity for the Pinios reservoir [22].

Several solutions have been proposed in cases where elevated concentrations of Fe/Mn have been detected, such as applying oxidants in the water treatment plants, direct aeration and/or oxygenation of the water column to lower the concentrations prior to use [8,9,31,35,39,40]. The treatment process for managing Mn is costly and can sometimes be difficult. During the last decades, there has been a growing need for an overall improvement in water quality and the ecological status of these systems, which will lead to the reduction in anoxia and consequently lower the concentration of Fe/Mn and turn the reservoir into the first filter of the water [9].

5. Conclusions

This study provides a detailed seasonal assessment of the physicochemical characteristics and trace metal concentrations in the Pinios Dam reservoir, a critical multipurpose water body in Northwestern Peloponnese, Greece. The results underscore the complex interplay between thermal stratification, dissolved oxygen dynamics and the mobilization of geogenic elements such as manganese (Mn) and iron (Fe), with important implications for water quality and reservoir management. The onset of hypoxia and anoxia in the hypolimnion was closely linked to a substantial rise in Mn and Fe concentrations, particularly from August to October. The highest Mn value recorded exceeded EU drinking water standards by nearly fourfold, while Fe concentrations followed a similar trend, emphasizing the critical need for effective water treatment strategies. The presence of elevated Mn and Fe concentrations poses not only operational challenges for water purification but also potential ecological threats due to their effects on aquatic biota. These elevated concentrations are primarily attributed to the reductive dissolution of Mn and Fe under anoxic conditions and their subsequent release from bottom sediments into the overlying water column. Given the multipurpose use of the Pinios reservoir, for drinking water supply, irrigation and ecological balance, these findings highlight the urgent need for a comprehensive water quality monitoring program and the implementation of mitigation strategies to ensure compliance with public health standards.

The study contributes valuable insights into the temporal dynamics of water quality in a stratified artificial reservoir and underscores the importance of integrated monitoring and management frameworks to safeguard the long-term sustainability of such critical freshwater resources.