Sediment Chemistry and Ecological Risk Assessment in Andean Lakes of Central Ecuador: Influence of Trophic Status on Accumulation Patterns

Abstract

This study evaluated the physicochemical characteristics, organic matter content, and heavy metal accumulation in the sediments of four Andean lakes in central Ecuador, considering their trophic states and protection status. A total of 96 sediment samples were collected and analyzed for electrical conductivity, pH, organic carbon, phosphorus, and heavy metal concentrations (Fe, Mn, Ni, Zn, Cu, Pb). The ecological risk was assessed using the potential ecological risk index (PERI) and pollution load index (PLI), and significant environmental predictors were identified using a classification and regression tree (CART) model. The results showed that protected lakes (Atillo and Magdalena) exhibited higher concentrations of Fe, Mn, and Ni, predominantly from natural sources, while unprotected lakes (Colta and Yambo) had greater Pb and Cu enrichment, associated with anthropogenic inputs. The PERI and PLI indices confirmed a low ecological risk across all lakes, although localized contamination was detected in Yambo and Magdalena. The CART model identified pH and phosphorus as the most significant predictors in protected lakes, whereas heavy metals and phosphate were more influential in unprotected lakes. These findings underscore the role of conservation status in shaping sediment composition and emphasize the need for targeted management strategies to mitigate anthropogenic impacts on Andean aquatic ecosystems.

1. Introduction

The rapid growth of the global population and increasing food demand have intensified the contamination of lacustrine systems, particularly in rural areas. This contamination is driven by both direct and diffuse sources of pollution, leading to excessive nutrients and organic carbon enrichment in water bodies and their sediments [1,2]. Among these sources, intensified agricultural practices have contributed significantly to the degradation of riparian soils, increasing nutrient runoff and sediment accumulation in lakes [3]. As a result, eutrophication accelerates, fostering excessive aquatic vegetation growth and algal blooms, which progressively reduce water surface area and oxygen concentration levels [4,5].

This issue is particularly critical in the Andean region of Latin America, where numerous high-altitude lakes play a crucial role in water regulation and biodiversity conservation [6,7]. These lakes are predominantly oligotrophic due to their cold temperatures, which limit nutrient availability and primary productivity, resulting in high transparency and low chlorophyll content [8,9]. However, changes in land use, climate variability, and anthropogenic activities have disrupted the natural equilibrium of many of these ecosystems, leading to the deterioration of water quality and increased ecological risk [10,11]. Heavy metal accumulation in Andean lakes is closely linked to their trophic state. Oligotrophic lakes, characterized by low organic matter content and high oxygenation levels, limit the solubility and bioavailability of metals, reducing contamination risks [10]. Conversely, eutrophic lakes, enriched with nutrients and organic matter, create favorable conditions for metal retention, as anoxic environments facilitate metal adsorption and precipitation [10,12]. Agricultural and urban runoff are major contributors to metal enrichment, particularly in shallow lakes, increasing potential toxicity risks for aquatic organisms and humans [13,14].

The Yambo, Colta, Magdalena, and Atillo lakes, located in the central Andean region of Ecuador, exhibit distinct environmental challenges resulting from varying levels of anthropogenic intervention and natural processes [15,16,17]. Yambo Lake is undergoing advanced eutrophication due to organic matter and nutrient accumulation from agricultural activities and urban discharges [18]. This has led to algal proliferation, reduced water transparency, and potential alterations in biogeochemical cycles, which may compromise aquatic biodiversity and ecosystem services [15]. Colta Lake is highly influenced by sediment transport and pollutants from agriculture and urban expansion, leading to fluctuations in its trophic state. Human interventions, such as sporadic sediment dredging and waste disposal, have modified its sedimentary dynamics and nutrient cycling, increasing the system’s vulnerability to further degradation [15,16]. Magdalena Lake, located within Sangay National Park, remains oligotrophic and exhibits physicochemical variability influenced by thermal stratification and nutrient dynamics. [17]. Atillo Lake, the largest in the same protected area, also maintains an oligotrophic state due to minimal human impact [17]. However, these ecosystems are not exempt from environmental threats. Natural factors such as temperature fluctuations, precipitation patterns, and sedimentation processes can alter their long-term ecological stability, potentially affecting nutrient availability and carbon sequestration processes [15,17]. Given the varying degrees of anthropogenic pressure and trophic conditions in these lakes, it is essential to evaluate their nutrient and heavy metal accumulation patterns, as well as the potential ecological risks associated with contamination. This study aims to analyze the influence of trophic status on nutrient and heavy metal accumulation in the sediments of Andean lakes in central Ecuador, determining their distribution patterns and assessing the environmental risks associated with sediment contamination.

2. Materials and Methods

2.1. Location and Characteristics of the Study Area

The study was conducted in the central Andean region of Ecuador, encompassing the provinces of Chimborazo and Cotopaxi. In Chimborazo province, the Colta, Magdalena, and Atillo lakes were investigated, while in Cotopaxi province, the research focused on Yambo Lake (Figure 1).

These lakes were selected due to their contrasting characteristics, including their trophic states and their varying levels of anthropogenic influence, which provide a representative gradient of ecological conditions. Additionally, their inclusion allows for a comparative analysis between protected areas (e.g., Atillo and Magdalena within the National Protected Areas System) and unprotected systems (e.g., Colta and Yambo), offering insights into the impacts of conservation efforts and human activities.

The Atillo and Magdalena lakes are located within the Sangay National Park at an altitude of 3485 m above sea level [17]. These lakes, part of the Atillo lacustrine system, are oligotrophic due to their low nutrient levels and high water transparency [19]. The Atillo Lake, the largest of the system, covers an area of 257.2 hectares with an approximate depth of 20 to 25 m, while the smaller Magdalena Lake has an estimated depth of 5 m [17]. These lakes belong to the Chambo River sub-basin within the Pastaza River hydrographic system of the Amazon basin, with periglacial geomorphology characterized by U-shaped valleys formed during the Jurassic, Miocene, and Pliocene periods [20]. The predominant soil types are sandy loam to clay loam, classified as Pachic Melanudands, which have high fertility and water retention capacity, facilitating carbon storage [21]. The climate is cold, with annual temperatures ranging from 5 to 6 °C and an average annual precipitation of 1000 mm [22].

In contrast, Yambo Lake, located at 2600 m above sea level in the Cotopaxi province, exhibits a hypertrophic state due to high anthropogenic influence [18]. With an area of 32 hectares and a maximum depth of approximately 17.1 m, it is situated within the Patate River sub-basin, part of the Pastaza River hydrographic system [15]. Its formation is attributed to a fluvio-lacustrine process, resulting in a depression characteristic of its morphology [23]. The surrounding soil types include mollisols with a loamy texture and a mix of entisols and inceptisols. The region has a temperate climate, with an average annual temperature of 17.1 °C and precipitation ranging between 372 and 500 mm [18]. The land use distribution around the lake shows a predominance of agricultural activities (57.09%), followed by eroded areas (21.00%), mining (14.42%), urbanized zones (5.77%), and natural water bodies (1.72%) [24].

Colta Lake, located in Chimborazo province at 3312 m above sea level, has an area of 280 hectares and a depth ranging from 3.5 to 8 m. It belongs to the Chambo River sub-basin and is considered a wetland with a mesotrophic status [15]. It supports diverse avian species, including ducks, herons, and coots, as well as various aquatic organisms [25]. The lake was formed during the Jurassic and Cretaceous periods, with soils classified as Mollic Endoquents, predominantly sandy loam in texture [16]. The climate is relatively mild, with temperatures between 12 and 15 °C and an annual precipitation of approximately 629 mm [26].

Table 1. Summary of study lakes’ characteristics [15,17,18,19,20,21,24,25].

Lake	Altitude (m a.s.l.)	Area (ha)	Depth (m)	Avg. Temperature (°C)	Annual Precipitation (mm)	Trophic State	Protection Status
Atillo	3485	257.2	20–25	5–6	~1000	Oligotrophic	Protected (SNAP *)
Magdalena	3485	~10	~5	5–6	~1000	Oligotrophic	Protected (SNAP *)
Colta	3312	280	3.5–8	12–15	~717	Mesotrophic	Unprotected
Yambo	2600	32	~17.1	~13	~500	Hypertrophic	Unprotected

* SNAP: Sistema Nacional de Áreas Protegidas (National Protected Areas System).

summarizes the main characteristics of the study lakes, including their altitude, surface area, depth, average temperature, annual precipitation, trophic state, and protection status.

2.2. Sampling

The collection of sediment samples from the study lakes was planned and conducted following the standardized methodologies described by Mudroch & MacKnight (1994) [27]. Sampling took place during the first semester of 2023. Based on the surface area and heterogeneity of each system, between 20 to 25 samples per lake were collected along three strategically located transects, which accounted for variations in sediment deposition patterns. For the analysis of heavy metals, a subset of six sediment samples per lake—referred to as sampling areas (SA1–SA6)—was selected from the total collected samples, ensuring spatial representativity based on depth gradients and proximity to inflows and depositional areas. This selection aimed to capture the variability in metal concentrations influenced by hydrodynamic and depositional processes within each system.

Sediment samples were obtained using a Zodiac-type boat and a Van Veen grab sampler (Eijkelkamp Soil & Water, Giesbeek, Gelderland, The Netherlands) [28], ensuring a uniform sediment retrieval process. The grab was deployed at depths ranging from 10 to 25 cm within the sediment column to obtain undisturbed samples, allowing for accurate assessments of recent sediment accumulation and geochemical processes.

All samples were stored in airtight bags to preserve their integrity and transported to the laboratory in an insulated container, following the guidelines established by the Technical Standard INEN-ISO 10381-1 [29]. Upon arrival, sediment pH and electrical conductivity (EC) were measured. Subsequently, samples were air-dried at controlled temperatures (20–25 °C) to prevent alterations in chemical composition [30]. Dried samples were homogenized and sieved through a 2 mm mesh for further analysis.

2.3. Samples Analysis

Physicochemical parameters such as pH and electrical conductivity (EC) were measured using a portable multiparameter meter HI98194 (HANNA Instruments, Smithfield, RI, USA). Texture was assessed using the Bouyoucos abbreviated method to determine the percentages of sand, silt, and clay [31].

The concentrations of heavy metals, including chromium (Cr), manganese (Mn), nickel (Ni), zinc (Zn), copper (Cu), lead (Pb), and iron (Fe), were quantified to assess potential contamination and ecological risks in the study lakes. These metals were selected due to their environmental relevance, toxicity, and bioaccumulation potential, as well as their documented presence in lacustrine sediments influenced by both natural processes and anthropogenic activities [13,14]. To extract and quantify the heavy metals, a microwave-assisted acid digestion procedure was conducted. A 0.5 g fraction of dried sediment was subjected to digestion using a mixture of 5 mL of concentrated nitric acid (HNO₃) and 2 mL of hydrogen peroxide (H₂O₂), following the methodology described by Gonzalez et al. (2009) [32]. The samples were placed in high-pressure and corrosion-resistant Teflon digestion tubes and left to pre-digest at room temperature for 24 h to facilitate the breakdown of organic matter and the solubilization of metal species [33]. Following the pre-digestion step, the samples were subjected to microwave-assisted digestion using a Milestone ETHOS UP system (Milestone Srl, Curno, Bergamo, Italy) under controlled conditions of temperature (up to 200 °C) and pressure (up to 20 bar), following the manufacturer’s recommended program for environmental solid samples. These conditions ensured complete metal extraction from the sediment matrix. After digestion, the solutions were filtered through 0.47 mm filter paper to remove residual particulates and adjusted to a final volume of 25 mL with deionized water in volumetric flasks to standardize analytical conditions. The quantification of heavy metal concentrations was performed using flame atomic absorption spectroscopy (FAAS) with an iCE 3500 spectrometer (Thermo Fisher Scientific, Waltham, MA, USA), following standard calibration protocols [34]. Metal concentrations were expressed in µg/g of sediment.

Organic matter percentage was determined by the loss on ignition (LOI) method [35]. Total organic carbon (TOC) and total nitrogen (TON) were analyzed via the Flash 2000 organic elemental analyzer (Thermo Fisher Scientific, Milan, Lombardy, Italy), which utilizes high-purity gases such as oxygen (for the combustion chamber) and helium (as the carrier gas). The equipment was calibrated using 2–3 mg of the BBOT standard (6.51% N, 72.53% C, 6.09% H, and 7.44% S) [36]. The Olsen method was employed to determine phosphorus availability [37]. In addition, the concentrations of Ca²⁺, Mg²⁺, Na⁺, and K⁺ were determined using atomic absorption spectrophotometry with the same Thermo Scientific iCE 3500 AA [38]. These cations were extracted from sediment samples using ammonium acetate (CH₃COONH₄), following the standardized methodology described by Azcarate et al. (2017) [39].

2.4. Data Analysis

2.4.1. Identification of Sources of Sedimentary Organic Matter

The carbon-to-nitrogen (C/N) ratio is widely used as an indicator to differentiate the origin of organic matter (OM) in lacustrine sediments [40]. Generally, fresh algae have a C/N ratio between 3 and 8, while terrestrial plants exhibit higher values, typically ranging from 14 to 23 or more [41]. Based on this classification, the C/N ratio was applied to estimate the contribution of autochthonous (endogenous) versus allochthonous (exogenous) OM in the sediments [42]. The calculations were based on the following relationships:

$$TOC=C_1+C_w,with{\displaystyle \frac{C_1}{N_1}}=18,$$

(1)

$$TON=N_1+N_w,with{\displaystyle \frac{C_w}{N_w}}=3,$$

(2)

where $C_1$ represents exogenous organic carbon (%) and $C_w$ endogenous organic carbon (%), while $N_1$ and $N_w$ correspond to exogenous and endogenous nitrogen content, respectively. Total organic nitrogen (TON) was assumed to be approximately 95% of total nitrogen, based on empirical findings [43].

To determine the relative contributions of exogenous and endogenous organic matter in the sediments of Atillo, Magdalena, Yambo, and Colta lakes, the system of Equations (1) and (2) was solved, leading to the application of Equations (3) and (4):

$$C_1={\displaystyle \frac{6TOC-18TON}{5}},$$

(3)

$$C_w={\displaystyle \frac{18TON-TOC}{5}}.$$

(4)

2.4.2. Ecological Risk Assessment Methodology

To assess the environmental risk associated with heavy metal contamination in sediments, two indices were applied: the potential ecological risk index (PERI) and the pollution load index (PLI). These indices provide a comprehensive evaluation of sediment contamination, considering both the individual and combined effects of heavy metals in aquatic ecosystems.

The potential ecological risk index (PERI) was applied to assess the environmental risk associated with heavy metal contamination in sediments. Originally developed by Hakanson (1980) [44] for marine and riverine environments, this method evaluates contamination from a sedimentological perspective, integrating heavy metal concentrations, their toxicological effects, and their potential impact on aquatic ecosystems. The PERI framework is widely recognized as a diagnostic tool for assessing sediment contamination, given the increasing accumulation of heavy metals and their potential remobilization into the water column, where they may pose significant ecological risks [45].

While PERI was initially designed for marine and fluvial ecosystems, recent studies have validated its applicability to high-altitude lacustrine environments [46,47]. The PERI calculation follows a structured approach, as represented in Equation (5):

$$PERI={\sum }_{i=1}^n{E}_i^r={\sum }_{i=1}^n{T}_i^r\times C_i^f,$$

(5)

where $C_i^f$ represents the contamination factor of metal $i$, obtained by dividing its measured concentration ($C_i$) by its background concentration ($C_i^n$). The ecological risk factor $E_i^r$ is calculated by multiplying the contamination factor $C_i^f$ by the toxic response factor $T_i^r$ of each metal. The final PERI value is determined by summing the individual ecological risk factors $E_i^r$ for all metals analyzed.

The toxic response factors for metals are as follows: Mn = 1, As = 10, Cd = 30, Cr = 2, Cu = 5, Ni = 5, Pb = 5, V = 2, Zn = 1 [4]. These values indicate the relative toxicity of each metal in aquatic ecosystems. The ecological risk level criteria for PERI are given as follows: <150 (low ecological risk), 150 ≤ PERI < 300 (moderate), 300 ≤ PERI < 600 (considerable), 600 ≤ PERI < 1200 (danger), >1200 (severe) [48].

The pollution load index (PLI) was also calculated to quantify the overall degree of heavy metal contamination in sediments. Unlike PERI, which focuses on ecological risk, PLI provides a comprehensive measure of cumulative metal pollution by assessing the relative enrichment of multiple contaminants in a given site [49]. The $PLI$ is calculated using Equation (6):

$$PLI={(C_{F1}\times C_{F2}\times C_{F3}\times \dots C_{Fn})}^{1/n},$$

(6)

where $C_F$ represents the contamination factor for each metal, calculated as follows:

$$C_F={\displaystyle \frac{C_i}{C_i^n}}$$

(7)

A PLI value of 1 indicates that metals are at background levels, suggesting no pollution, while values greater than 1 indicate increasing pollution levels [50].

2.4.3. Classification and Regression Tree (CART) Algorithm

The classification and regression tree (CART) algorithm was applied to identify the most influential environmental variables affecting the study lakes. CART is a machine learning technique widely used for decision tree modeling, capable of handling both categorical and numerical data through recursive partitioning [51]. The algorithm systematically partitions the dataset by evaluating potential predictor variables based on impurity reduction criteria, such as Gini impurity for classification tasks and variance minimization for regression problems [52]. This process continues iteratively until an optimal final partitioning is achieved, minimizing selection error while ensuring meaningful differentiation between data groups.

In this study, CART regression was employed to analyze environmental factors influencing the lakes, considering their trophic states and levels of anthropogenic impact. The dataset was structured into two groups: protected lakes (Magdalena and Atillo, within the Sangay National Park) and unprotected lakes (Colta and Yambo, subject to significant human activities). A total of 39 variables were analyzed, including physicochemical parameters, sediment composition, nutrient concentrations, and heavy metal content (pH, electrical conductivity, total cation exchange capacity, organic matter, total nitrogen, total carbon, exogenous carbon, endogenous carbon, carbon-to-nitrogen ratio, phosphorus, phosphate, iron, manganese (Mn), cadmium (Cd), chromium (Cr), copper (Cu), nickel (Ni), lead (Pb), zinc (Zn), contamination factors were calculated for Mn, Cd, Cr, Cu, Ni, Pb, and Zn, while risk evaluations were conducted for Mn, Cd, Cr, Cu, Ni, Pb, and Zn, pollution load index, potential ecological risk index, taxonomic order, suborder, grand order, and sediment age). These variables were derived from both in situ measurements and calculated indices to provide a comprehensive assessment of the factors driving ecological variability.

The CART model was trained independently for each group of lakes, with the sampling sites serving as target outputs. The algorithm identified the main environmental predictors that best distinguished between different lake conditions, facilitating the interpretation of complex ecological interactions. Figure 2 illustrates the CART tree structure, highlighting the sequential variable selection process and the decision thresholds at each node.

2.4.4. Statistical Analysis

The data were processed using IBM SPSS V20 to calculate the central tendency measures of the sediment parameters and variables. The Shapiro–Wilk test was applied to assess the normality of the data distribution. ANOVA analysis was performed to determine significant differences in sediment characteristics among the studied lakes. Statistical significance was considered at p < 0.05. When ANOVA results were significant, a Tukey’s HSD post hoc test was applied to identify pairwise differences between lakes.

3. Results

3.1. Physicochemical Characteristics of Sediments

The electrical conductivity (EC) values of the sediments showed statistically significant differences among the studied lakes (p < 0.05). While Colta, Magdalena, and Atillo lakes exhibited relatively similar EC values of 277.23 ± 2, 237.65 ± 45.1, and 343.31 ± 44.3 µS·cm⁻¹, respectively, Yambo presented a higher EC of 693.67 ± 82 µS·cm⁻¹, indicating a distinct geochemical profile likely influenced by its eutrophic condition.

The pH values of the lakes exhibited significant variability. Yambo, characterized by eutrophic conditions, had the highest mean pH (8.9), followed by Magdalena (7.6), Colta (7.2), and Atillo (7.0).

The predominant sediment texture across three of the four studied lakes is loamy sand, followed by clay loam. However, Atillo exhibits distinct textural characteristics despite its proximity to Magdalena. The sediment in Atillo contains a higher sand fraction, exceeding 78.7%, distinguishing it from the other lakes. This composition classifies Atillo’s sediment within the sandy loam to loamy sand texture range [19].

3.2. Identification of Sedimentary Organic Carbon Content and Its Sources

The statistical analysis of 96 sediment samples revealed significant differences in nutrient parameters among the studied lakes with different trophic states (

Table 2. Nutritional characteristics of sediments in the studied lakes.

Parameter	Colta (n = 26)	Yambo (n = 25)	Magdalena (n = 24)	Atillo (n = 21)	p-Value
SOC (mg/kg)	1.9 ± 0.1 b	4.0 ± 0.3 a	3.7 ± 0.5 a	3.8 ± 0.6 a	0.000
% OC	2.1 ± 0.1 a,c	1.8 ± 0.2 a	2.8 ± 0.3 a,c	3.2 ± 0.5 c	0.009
% N	0.3 ± 0.03 a	0.5 ± 0.05 b	0.3 ± 0.04 a	0.5 ± 0.1 b	0.005
PO43⁻ (mg/kg)	94.9 ± 2.1 a	678.5 ± 162.7 b	126.5 ± 11.6 a	122.2 ± 9.5 a	0.000
% OM	3.3 ± 0.2 a	6.9 ± 0.6 b	6.4 ± 0.8 b	6.5 ± 0.9 b	0.000
CEC (Cmol/kg)	18.9 ± 0.6 b	26.7 ± 3.6 c	1.2 ± 0.03 a	1.4 ± 0.06 a	0.000

Different letters indicate significant differences between lakes (p < 0.05). Values sharing the same letter do not differ significantly. n = number of sediment samples analyzed per lake.

).

The lake with the highest percentage of organic carbon (OC) corresponds to Atillo (3.2 ± 0.5%), which differed significantly from Yambo (1.8 ± 0.2%, p < 0.05). Meanwhile, Magdalena (2.8 ± 0.3%) and Colta (2.1 ± 0.1%) did not differ significantly from either Atillo or Yambo (p > 0.05). These results indicate that organic carbon content is not a direct determinant of the eutrophication level, as Yambo, classified as hypertrophic, exhibits lower OC concentrations than the oligotrophic lakes Atillo and Magdalena. Meanwhile, Colta, which is classified as mesotrophic, presents intermediate OC values.

The C/N ratios show notable differences between the oligotrophic and eutrophic lakes. Magdalena exhibits the highest ratio (9.3), followed by Colta (7.0). In contrast, Atillo and Yambo present lower values of 6.4 and 3.6, respectively. However, these differences were not statistically significant (p > 0.05). These results indicate that eutrophic lakes, particularly Yambo, have lower C/N ratios, which may be attributed to increased microbial activity and organic matter decomposition, leading to greater assimilation of nitrogen relative to carbon [18].

Phosphate (PO₄³⁻) concentrations exhibit significant differences among the studied lakes. Yambo shows the highest concentration, with an average of 678.5 ± 162.7 mg/kg, which was significantly higher than in Colta, Magdalena, and Atillo (p < 0.05), while the other lakes maintain significantly lower and more stable values, ranging from 94.9 ± 2.1 mg/kg to 126.5 ± 11.6 mg/kg (no significant differences among them, p > 0.05).

The cation exchange capacity (CEC), analyzed from the exchangeable ions Ca²⁺, Mg²⁺, Na⁺, and K⁺, presents the highest value in Yambo (26.7 ± 3.6 Cmol/kg), which differed significantly from all the other lakes (p < 0.05), followed by Colta (18.9 ± 0.6 Cmol/kg), and significantly lower values in the oligotrophic lakes Atillo and Magdalena (1.2–1.4 Cmol/kg), which did not differ significantly from each other (p > 0.05). Higher CEC values in eutrophic lakes are associated with increased organic matter content and nutrient availability [53].

Figure 3 illustrates the distribution of exogenous and endogenous carbon contributions to the lacustrine sediments of the studied lakes.

The results indicate that the predominant source of organic carbon is exogenous, which aligns with the erosive processes occurring in the middle and upper terraces of the micro-basin. This pattern is particularly evident in Magdalena, Colta, and Atillo, where exogenous carbon represents the major fraction, with mean values ranging between 70% and 89%.

In contrast, Yambo presents a distinct pattern, showing a significantly higher proportion of endogenous carbon, with values reaching up to 85%. The high variability in endogenous carbon contribution in Yambo (ranging from 15% to 85%) suggests that biological processes, including primary productivity and organic matter decomposition, play a more influential role in its sediment composition compared to the other lakes [18].

The sedimentation of exogenous carbon is closely related to the geomorphological characteristics of the region. Magdalena and Atillo, both of periglacial origin, are characterized by rugged terrains with slopes between 35% and 75% [17], which facilitate the transport of allochthonous material into the lakes. Colta, a colluvial-origin lake, has a hilly terrain with slopes ranging from 20% to 35% [16], supporting moderate but consistent exogenous carbon deposition. Meanwhile, Yambo, with a combination of steep (50–75%) and moderately steep (7–12%) slopes [15], shows a mixed contribution of both exogenous and endogenous sources, reflecting its hypertrophic state and active biological carbon cycling.

These findings are consistent with previous studies on Andean lakes, where steep slopes and erosive forces contribute to high exogenous sedimentation rates, while productive lakes with significant biological activity tend to have a greater proportion of endogenous organic carbon [54,55].

3.3. Sediments Heavy Metals

Figure 4 presents a boxplot summarizing the distribution of heavy metal concentrations including Fe, Mn, Cu, Ni, Zn, and Pb in Andean lagoons of central Ecuador: Colta, Yambo, Magdalena, and Atillo.

The iron (Fe) concentrations in the analyzed sediments of the lakes ranged from 48.3 to 13,184.9 µg/g, with an overall median of 1095.5 µg/g. The highest averages were observed in the oligotrophic lakes: Atillo (7532.0 ± 1138 µg/g) and Magdalena (8197.7 ± 1848 µg/g), both showing high data dispersion. In contrast, the eutrophic Yambo (433.3 ± 131 µg/g) and Colta (843.0 ± 40 µg/g) presented more homogeneous values. The concentrations in Magdalena were up to 19 times higher than those of Colta, highlighting significant differences associated with the trophic state of the lake systems.

Manganese (Mn) concentrations ranged from 31 µg/g to 1082 µg/g, with a median of 103.85 µg/g. The highest average concentrations were identified in Atillo (377.8 ± 156 µg/g) and Magdalena (212.4 ± 62 µg/g), while in Colta (65.9 ± 3 µg/g) and Yambo (102 ± 31 µg/g), the concentrations were lower. Notably, some areas in Atillo showed high values, reaching the maximum concentration observed.

The copper (Cu) concentrations had a mean of 10.0 ± 0.8 µg/g. In eutrophic lakes, values reached up to 13.3 µg/g, while in oligotrophic lakes, they were as high as 17.2 µg/g. The highest concentrations were recorded in Magdalena (11.9 ± 1.4 µg/g) and Atillo (12.4 ± 1.0 µg/g), whereas Yambo (4.05 ± 0.7 µg/g) and Colta (11.2 ± 0.6 µg/g) showed lower values.

Nickel (Ni) concentrations in the studied lakes had a median of 2.81 µg/g, with values ranging from 0.62 to 24.0 µg/g. The median values per lake were as follows: Colta (1.88 ± 0.1 µg/g), Yambo (0.84 ± 0.1 µg/g), Magdalena (10.66 ± 3.0 µg/g), and Atillo (6.77 ± 1.0 µg/g). Variability was lower in oligotrophic lakes.

The median lead (Pb) concentration was 2.63 µg/g, with extreme values observed in Yambo, where concentrations reached 7.21 µg/g. The Colta Lake presented concentrations below 2.5 µg/g. In Yambo and Atillo, Pb values were more consistent, with mean concentrations of 5.6 ± 0.7 µg/g and 2.8 ± 0.1 µg/g, respectively. In Magdalena, Pb concentrations remained low, with a mean value of 2.93 ± 0.1 µg/g. As shown in Figure 4, lead exhibited the highest number of outliers, mainly in Yambo and Atillo.

Zinc (Zn) concentrations in the lakes ranged from 0.91 µg/g to 24.8 µg/g. The mean values were 2.02 ± 0.8 µg/g in Colta, 2.1 ± 0.5 µg/g in Yambo, 14.9 ± 3 µg/g in Magdalena, and 17.7 ± 1.4 µg/g in Atillo. Higher Zn concentrations were observed in lakes with less anthropogenic influence.

The concentrations of chromium (Cr) and cadmium (Cd) in the studied lakes were below 2.5 µg/g and 1.25 µg/g, respectively, as detected by atomic absorption spectroscopy. Due to these low values, Cr and Cd were not included in Figure 4, as their concentrations remained below detectable limits in all the lakes analyzed. These results indicate that these metals are present at minimal levels in the sediments of the studied lakes.

To better interpret these results, the concentrations of the detected trace metals were compared against internationally accepted sediment quality benchmarks [1]. These include two commonly used thresholds: the effects range-low (ERL), below which adverse biological effects are rarely observed, and the effects range-median (ERM), above which such effects are more likely to occur. Concentrations that fall between these two thresholds suggest a possible, though not certain, ecological impact.

Copper, lead, zinc, and nickel concentrations in all lakes were below the sediment quality guideline thresholds. Specifically, copper concentrations were well below the ERL of 34.0 µg/g and the ERM of 270.0 µg/g. Lead levels did not exceed 6 µg/g in any lake, remaining far below the ERL of 46.7 µg/g and ERM of 218.0 µg/g. Zinc, although higher in Atillo (17.7 µg/g) and Magdalena (14.9 µg/g), was still far below the ERL (150.0 µg/g) and ERM (410.0 µg/g). Nickel concentrations, even in Magdalena (10.7 µg/g), did not approach the ERL (20.9 µg/g) or ERM (51.6 µg/g).

Chromium and cadmium concentrations were <2.5 µg/g and <1.25 µg/g, respectively, remaining well below the ERL values (81.0 µg/g for Cr and 1.2 µg/g for Cd) and the ERM thresholds (370.0 µg/g for Cr and 9.6 µg/g for Cd).

Although iron and manganese showed high concentrations in some lakes, they were not compared against ERL/ERM guidelines, as these are essential trace elements that naturally occur at elevated levels in sediments and are not prioritized in toxicity benchmarks for lacustrine systems.

3.4. Ecological Risk Assessment Findings

The pollution load index (PLI) analysis (Figure 5) indicates clear differences among the studied lakes. Colta and Yambo exhibit low and stable contamination levels, ranging between 0.05 and 0.09, whereas Magdalena and Atillo show higher values in certain areas, reaching up to 0.26 in SA1 and 0.24 in SA3. In Colta and Yambo, PLI values remain relatively constant across all sampled areas, suggesting a uniform distribution of sediment contamination. In contrast, Magdalena and Atillo display greater variability, with peaks observed at SA1 and SA3, indicating localized sources of metal enrichment.

The potential ecological risk index (PERI) values in Figure 6 indicate that all the studied lakes fall within the low ecological risk category (<150). Colta Lagoon exhibits PERI values between 1.30 and 1.72, showing minimal variation across sampled areas. Yambo Lagoon presents slightly higher values, ranging from 1.53 to 2.61, with a relative increase at SA5. Magdalena Lagoon shows the highest internal variability, with values ranging from 1.30 to 4.89, although remaining well below the moderate risk threshold (150). Similarly, Atillo Lagoon exhibits values between 2.39 and 3.89, peaking at SA3 but still within the low-risk range.

3.5. CART Predictor Model Findings

The analysis of predictor variables using the CART model, presented in Figure 7, highlights distinct factors influencing the classification of lakes within and outside protected areas.

For lakes within protected areas (Figure 7a), the most influential variables were pH and phosphorus (P), followed by nitrogen (N), organic matter (OM), carbon (C), and electrical conductivity (EC). These variables exhibited the highest importance scores in the classification model.

For lakes outside protected areas (Figure 7b), the primary predictor variables included the contamination factor of copper (Cu-FC), lead (Pb), nickel (Ni), total copper (Cu), phosphate (PO₄³⁻), and pH, all of which showed comparable importance scores.

The CART model demonstrated strong predictive capability, with an R-squared value of 0.71 for lakes within protected areas, explaining 71% of the variability in classification data. Meanwhile, lakes outside protected areas exhibited a slightly lower R-squared value of 0.66, suggesting a moderately lower explanatory power. The differences in predictor variables between protected and unprotected lakes indicate distinct environmental influences affecting these ecosystems.

4. Discussion

4.1. Physicochemical Properties of Lake Sediments

The physicochemical characteristics of lake sediments play a crucial role in determining nutrient dynamics and overall water quality [56]. The comparison of electrical conductivity (EC) and pH levels among Colta, Magdalena, Atillo, and Yambo lakes reveals distinct environmental conditions. Yambo exhibits the highest EC, exceeding 500 µS·cm⁻¹, suggesting an elevated presence of carbonates and bicarbonates, as demonstrated by the analysis of inorganic carbon in sediments [57]. These conditions support its hypertrophic state, promoting phytoplankton proliferation due to the efficient assimilation of CO₂ for primary production [58]. In contrast, the other lakes display lower EC values, indicating differences in sediment composition and geochemical processes.

The pH conditions influence nutrient accumulation and biological activity in the studied lakes [59]. However, the statement that Atillo and Magdalena exhibit high pH values (up to 7.6) contradicts their oligotrophic classification. Oligotrophic systems typically maintain neutral to slightly acidic pH values due to limited biological activity and organic matter decomposition [44]. Thus, further investigation is required to explain the buffering capacity of these lakes and its relation to their trophic status.

pH variability in lakes can be influenced by photosynthetic activity, buffering capacity, and external inputs. In eutrophic systems like Yambo, elevated primary production consumes CO₂, leading to increased pH during daylight hours [60]. In contrast, oligotrophic lakes tend to maintain more stable pH values due to lower biological activity and stronger geochemical buffering. Shifts in pH can also influence phytoplankton composition, as seen in alkaline systems where lower pH reduces cyanobacterial dominance. Moreover, seasonal changes in ecosystem metabolism contribute to natural pH fluctuations and can modulate broader acidification trends [1].

Sediment texture is another determining factor in lake dynamics. The predominant texture is loamy sand, but Atillo stands out with a sandy composition exceeding 78.7%, distinguishing it from the other lakes. This difference is likely due to the geomorphological characteristics of Atillo and Magdalena, where sediments originate from direct runoff from the surrounding hills [15,17]. Their edaphic origin influences sediment stability and nutrient retention, which impacts lake health and productivity [58]. Previous studies on Colta indicate that intensive dredging processes have significantly altered its sediment composition and water quality [15,26]. Escobar-Arrieta et al. (2021) reported that these interventions induce alkaline conditions, which may not align with its mesotrophic classification [26]. This highlights the necessity of long-term studies to assess the impact of sediment removal on lake trophic status and overall ecological health.

4.2. Organic Carbon Content and Trophic Levels

Sediment analysis revealed that Atillo contains the highest organic carbon (OC) percentage (3.2 ± 0.5%), while Colta, Yambo, and Magdalena exhibit lower values, ranging from 1.8 ± 0.2% to 2.8 ± 0.3%. Interestingly, eutrophic lakes like Colta and Yambo display lower OC percentages despite their high phosphate (PO₄³⁻) concentrations, suggesting that OC is not a direct indicator of eutrophication. This pattern is consistent with findings by [61], which reported a weak correlation (0.21) between OC and P.

Significant contrasts in organic matter dynamics exist among the lakes. Magdalena and Atillo exhibit high organic matter accumulation (6.4 ± 0.8% and 6.5 ± 0.9%, respectively), indicating their role as carbon sinks. These conditions suggest low decomposition rates, likely due to climatic constraints such as low temperatures and high humidity, which limit microbial activity [62]. In contrast, Yambo (6.9 ± 0.6%) appears to be a net carbon emitter due to continuous organic matter decomposition, driven by higher temperatures and oxygen availability [62]. These results underscore the influence of local environmental factors in carbon cycling and highlight the need for further research into the specific mechanisms regulating organic matter decomposition in Andean lakes.

4.3. Nitrogen and Phosphorus Dynamics

Nitrogen (N) and phosphorus (P) concentrations are closely linked to sediment origin and algal-driven nitrification processes [18]. In Yambo, elevated N and P levels result from wastewater discharges and agricultural runoff, which lead to nutrient accumulation in sediments, particularly during algal decay phases [58].

Eutrophic lakes such as Yambo and Colta exhibit anaerobic conditions, leading to the production and release of hydrogen sulfide (H₂S) [16,18]. This process disrupts ecological balance, poses environmental risks, and reduces the sediment’s self-purification capacity [63]. Given these conditions, studies on nitrogen dynamics should be conducted to evaluate the potential emission of greenhouse gases such as nitrous oxide (N₂O), a known contributor to climate change.

Cation exchange capacity (CEC) values range from 1.2 ± 0.03 to 26.7 ± 3.6 Cmol/kg, with the highest values observed in eutrophic lakes. Increased CEC indicates greater nutrient retention and bioavailability from organic matter, promoting accelerated aquatic vegetation growth along lake edges [12]. The high retention capacity of these sediments also contributes to heavy metal accumulation, explaining the heterogeneous distribution of contaminants in Yambo [64].

The high phosphate concentrations in eutrophic lakes, particularly in Yambo, indicate elevated P availability in soluble forms, likely due to enhanced mineralization [65]. Meanwhile, the relatively stable PO₄³⁻ levels in Colta may be influenced by sediment dredging, which affects phosphorus cycling. In contrast, lower and more uniform P concentrations in Magdalena and Atillo may result from lower temperatures that reduce microbial activity and phosphorus mineralization rates [66]. Similar patterns have been observed in other Andean lakes, such as Chungará, where physicochemical conditions limit phosphorus availability and primary productivity [67].

4.4. Sedimentary Organic Matter Sources

The sedimentation patterns in the studied lakes are influenced by their geomorphology and basin characteristics [55]. Atillo and Magdalena exhibit a high rate of sedimentation from exogenous sources, consistent with the steep slopes and high-altitude runoff in their catchment areas, as previously documented in the Atillo micro-watershed of Sangay National Park [19]. In contrast, Yambo presents a higher proportion of endogenous organic matter, attributed to the decomposition of its aquatic vegetation and phytoplankton [18,68].

The predominance of exogenous carbon in Atillo, Magdalena, and Colta (70–89%) aligns with the erosion and transport processes occurring in their watersheds [16,17], whereas Yambo, with endogenous contributions reaching 85%, suggests active organic matter cycling and decomposition [69]. These differences influence the biogeochemical stability of each system, affecting nutrient retention and organic matter degradation rates.

The organic matter accumulation in eutrophic lakes like Yambo is strongly linked to anaerobic decomposition processes, which may enhance methane (CH₄) emissions [70]. Studies indicate that aquatic vegetation and phytoplankton contribute significantly to methane production, potentially exceeding emissions from exogenous organic inputs [71]. However, additional research is needed to quantify the specific methane fluxes in these Andean lakes.

In contrast, Atillo and Magdalena may act as carbon sinks, as their high exogenous organic matter content and cold-water conditions limit decomposition rates. Previous studies in Andean lakes suggest that lower temperatures and specific physicochemical conditions can reduce microbial degradation of organic matter, favoring carbon sequestration [62]. This phenomenon could partially explain the differences in organic matter dynamics among the studied lakes.

Phytoplankton and macrophytes likely play an important role in organic carbon cycling in Yambo while in Colta, Atillo, and Magdalena, the organic matter deposition is primarily influenced by terrestrial inputs. Further studies on the microbial degradation of organic matter in these lakes would provide valuable insights into their role in carbon sequestration and greenhouse gas emissions.

4.5. Heavy Metal Concentrations in Lake Sediments

The heavy metal concentrations in the studied Andean lakes of Ecuador provide a reference for understanding sediment contamination levels in high-altitude lacustrine systems. Comparing these results with metal concentrations from other lagoons in the Americas (

Table 3. Results of heavy metal concentrations in American lagoons.

Lagoons/Altitude	Location	Chromium (µg/g)	Cadmium (µg/g)	Iron (µg/g)	Manganese (µg/g)	Nickel (µg/g)	Zinc (µg/g)	Copper (µg/g)	Lead (µ/g)	Reference
Chungará—4520 m a.s.l.	Chile	44.19 ± 4.57	0.13 ± 2.09	NA	NA	NA	40.61 ± 2.39	21.51 ± 2.19	4.10 ± 0.53	[72]
Laja—2362 m a.s.l.	Chile	43.0 ± 3.64	0.12 ± 7.36	NA	NA	NA	36.3 ± 1.06	46.0 ± 3.19	11.1 ± 0.83	[72]
Titicaca—3812 m a.s.l.	Perú	NA	0.04215	NA	NA	NA	8.20 ± NA	6.95 ± NA	0.04 ± NA	[73]
Traful Lake—975 m a.s.l.	Argentina	19.5 ± NA	NA	46.7 ± NA	918	33.6	101.6 ± NA	NA	NA	[74]
Bahía López Lagoon—2075 m a.s.l.	Argentina	57.1 ± NA	NA	45.2 ± NA	931 ± NA	40.2 ± NA	90.7 ± NA	NA	NA	[74]
Bariloche Lagoon—764 m a.s.l.	Argentina	53.83 ± NA	NA	NA	NA	NA	NA	NA	NA	[74]
Bahía Llao-Llao Lagoon—1050 m a.s.l.	Argentina	NA	NA	45.2	931 ± NA	42.4 ± NA	155.8 ± NA	NA	NA	[74]
Punto Panorámico Lagoon—945 m a.s.l.	Argentina	51.6 ± NA	NA	NA	2760 ± NA	27.9 ± NA	172.5 ± NA	NA	NA	[74]
Escondido Lagoon—770 m a.s.l.	Argentina	37.4	NA	NA	1992 ± NA	42.6 ± NA	91.3 ± NA	NA	NA	[74]
Morenito Lagoon—780 m a.s.l.	Argentina	NA	NA	NA	NA	NA	567 ± NA	NA	NA	[74]
Bustillos Lagoon—2000 m a.s.l.	Mexico	25 ± NA	1.51 ± NA	NA	300 ± NA	20 ± NA	90 ± NA	25 ± NA	90 ± NA	[75]
Unare Lagoon—600 m a.s.l.	Venezuela	51.69 ± NA	NA	NA	516.37 ± NA	52.41 ± NA	127.49 ± NA	41.13 ± NA	29 ± NA	[76]
Presa la Purisma—1902 m a.s.l.	Mexico	NA	NA	NA	NA	NA	67.6 ± NA	30.76 ± NA	22.76 ± NA	[38]
Colta Lake—3312 m a.s.l.	Ecuador	<2.5	<1.25	843 ± 40	66 ± 2	1.9 ± 0.1	2.0 ± 0.4	11.2 ± 0.6	<2.5	Actual study
Yambo Lake—2600 m a.s.l.	Ecuador	<2.5	<1.25	433 ± 131	102 ± 31	0.84 ± 0.1	2.1 ± 0.5	4.5 ± 0.7	5.6 ± 0.7	Actual study
Atillo Lake—3485 m a.s.l.	Ecuador	<2.5	<1.25	7532 ± 1138	378 ± 156	6.8 ± 1.1	15 ± 3	12.4 ± 0.9	2.8 ± 0.1	Actual study
Magdalena Lake—3485 m a.s.l.	Ecuador	<2.5	<1.25	8197 ± 1848	212 ± 62	10.7 ± 3.2	18 ± 1.4	11.9 ± 1.4	2.6 ± 0.1	Actual study

NA: Not available.

) offers insights into the regional variability of heavy metal accumulation and potential ecological risks.

Iron (Fe) concentrations ranged from 433 µg/g (Yambo) to 8197 µg/g (Magdalena), with the highest values found in the oligotrophic Atillo and Magdalena lagoons. This pattern aligns with findings from Bahía López and Traful lagoons (Argentina), where Fe concentrations reach up to 46.7 µg/g [74]. The volcanic substrate and mineralogical composition of the Andean region likely contribute to the enrichment of Fe in the sediments of these oligotrophic lagoons [66]. In contrast, Yambo and Colta, with eutrophic conditions, exhibited significantly lower Fe concentrations, suggesting potential Fe depletion due to redox-driven solubilization under anoxic conditions [77].

Manganese (Mn) concentrations in Atillo (378 µg/g) and Magdalena (212 µg/g) were notably higher than those in Colta (66 µg/g) and Yambo (102 µg/g). This is similar to Mn values reported in Unare Lagoon (516 µg/g) and lower than the extreme values recorded in Bustillos (2760 µg/g) and Escondido (1992 µg/g) [74,75,76]. The Mn variability in Andean lakes is likely influenced by redox conditions, as Mn oxides act as adsorption sites for metals like Zn and Ni, affecting their mobility [78]. In eutrophic environments like Yambo, Mn solubilization under anoxic conditions could enhance metal bioavailability, contributing to eutrophication dynamics.

Copper (Cu) concentrations ranged from 4.5 µg/g (Yambo) to 12.4 µg/g (Atillo), with values in Colta (11.2 µg/g) and Magdalena (11.9 µg/g) aligning with regional trends. These values are significantly lower than those reported in Unare Lagoon (127.5 µg/g) and Bahía Llao-Llao (155.8 µg/g) [74,76], suggesting that Cu contamination in Andean lakes is relatively minor. However, in eutrophic systems like Yambo, Cu bioavailability could increase due to its association with organic matter and anoxic conditions [79].

Zinc (Zn) concentrations were highest in Atillo (15 µg/g) and Magdalena (18 µg/g), while lower values were observed in Yambo (2.1 µg/g) and Colta (2.0 µg/g). These values are comparable to those reported in Chungará Lagoon (40.6 µg/g) but significantly lower than those in Punto Panorámico (172.5 µg/g) and Unare Lagoon (127.5 µg/g) [72,74,76]. Zn behavior is pH-dependent, with low bioavailability in alkaline conditions such as Yambo, where Zn²⁺ competes with Ca²⁺ and Mg²⁺ for adsorption sites [80].

Nickel (Ni) concentrations showed a depth-dependent trend, with higher values in Atillo (6.8 µg/g) and Magdalena (10.7 µg/g) compared to Yambo (0.84 µg/g) and Colta (1.9 µg/g). Similar trends have been reported in Traful Lagoon (33.6 µg/g) and Unare Lagoon (52.4 µg/g) [74,76], where Ni accumulation is influenced by organic matter content and alkaline conditions [81]. In Yambo, Ni may be present in carbonate forms, reducing its toxicity [80].

Lead (Pb) concentrations were highest in Yambo (5.6 µg/g), likely due to vehicular emissions from the nearby Panamerican Highway, as observed in Lagos and Ologe Lagoon (Nigeria) [82]. In contrast, Atillo (2.8 µg/g) and Magdalena (2.6 µg/g) had lower Pb values, potentially influenced by natural volcanic sources and historical extractive activities near Sangay National Park [83]. These values are lower than those reported in Unare Lagoon (29 µg/g) and Bustillos (90 µg/g) [75,76], indicating that Pb pollution is relatively low in Andean lakes.

Chromium (Cr) and Cadmium (Cd) concentrations were below detection limits (<2.5 µg/g and <1.25 µg/g, respectively) in all studied lagoons, which contrasts with higher Cr values recorded in Bahía López (57.1 µg/g) and Bariloche (53.8 µg/g) [74]. The absence of Cd and low Cr levels in the Andean lakes suggest minimal anthropogenic inputs of these metals, but further analyses using ICP-MS could help detect trace concentrations and assess potential long-term accumulation risks.

The results indicate that oligotrophic lakes (Atillo and Magdalena) exhibited the highest Fe, Mn, and Ni concentrations, aligning with trends observed in other high-altitude lakes with volcanic substrates and limited anthropogenic influence. Eutrophic lakes (Yambo and Colta) had lower Fe and Mn but higher Pb concentrations, likely influenced by agriculture, vehicular emissions, and redox-driven solubilization of metals. Zn and Cu concentrations were relatively low compared to highly contaminated lakes in Venezuela and Argentina, but Cu dynamics in eutrophic environments require further investigation. Cd and Cr levels were below detection limits, highlighting the need for advanced analytical techniques to assess trace metal accumulation.

4.6. Ecological Risk Assessment

The ecological risk assessment, based on the potential ecological risk index (PERI) and the pollution load index (PLI), reveals that the lakes in central Ecuador present low to very low levels of ecological risk. The PERI values remain below 150 across all lakes, indicating low ecological risk. Similarly, PLI values in all lakes are below 1, suggesting no overall pollution.

The primary source of heavy metal enrichment appears to be geogenic, derived from the geological composition and erosion of the riparian zones [84]. However, there are significant spatial variations in risk, demonstrating the heterogeneity in sediment behavior across different lakes [85].

The pollution load index (PLI) (Figure 5) supports these findings, as Colta and Yambo exhibit stable and low contamination levels (0.05–0.09), indicating a minimal anthropogenic impact. In contrast, Magdalena and Atillo show localized areas of increased contamination (up to 0.26 in SA1 and 0.24 in SA3), suggesting point sources of metal enrichment, likely influenced by natural mineral leaching and, to a lesser extent, human activities [49]. These differences highlight the necessity of spatially explicit monitoring, as contamination is not evenly distributed within lakes.

Regarding the PERI analysis, Yambo Lagoon stands out due to localized areas of increased ecological risk, which are likely linked to anthropogenic sources such as wastewater discharges and cross-contamination from motorboat operations [18], explaining the elevated Pb concentrations (5.6 ± 0.7 µg/g), the highest among the studied lakes. This aligns with similar findings in lacustrine systems where Pb accumulation is associated with vehicular emissions and fuel residues [82].

In contrast, the highest contributions to PERI values (ranging from 1.30 to 4.89) were observed in Atillo and Magdalena, where Fe, Ni, Zn, and Mn concentrations were significantly higher. These metals originate primarily from geogenic sources, particularly volcanic weathering and erosion processes characteristic of high-altitude Andean environments [66]. This finding contradicts studies from European lakes, where high-altitude systems tend to have lower organic matter and heavy metal content [86], suggesting that Andean lakes experience distinct sedimentation and mineral deposition processes.

The role of redox conditions in regulating heavy metal mobility is evident when comparing oligotrophic and eutrophic lakes. In Atillo and Magdalena, oxidative conditions promote the retention of heavy metals via adsorption onto Mn-oxyhydroxides, reducing their bioavailability [78]. Conversely, in eutrophic lakes like Yambo and Colta, high phosphate concentrations (678.5 ± 162.7 mg/kg in Yambo) facilitate heavy metal solubilization, increasing their potential ecological impact [87]. However, in Yambo, despite its hypertrophic conditions, CaCO₃ precipitation appears to reduce heavy metal bioavailability by forming insoluble hydroxides and carbonates that limit their ecological risk [45].

4.7. CART Predictor Model

The results of the classification and regression tree (CART) model reveal that the conservation status of the studied lakes—Atillo and Magdalena (protected) vs. Yambo and Colta (unprotected)—strongly influences the dominant sedimentary processes. The model identified pH and phosphorus (P) as the most significant predictors in protected lakes, accounting for more than 50% of the observed variability, whereas contamination factors of Cu, Ni, Pb, and PO₄³⁻ were more important in unprotected lakes.

The influence of altitude and climate conditions on sediment composition is evident in Atillo and Magdalena, both located at 3485 m a.s.l. with cold temperatures (5–6 °C) and high precipitation (~1000 mm/year). In these oligotrophic lakes, phosphorus is mainly associated with organic matter (OM), suggesting limited bioavailability due to cold temperatures that slow down microbial mineralization processes. This is supported by strong OM–OC correlations (96% in Magdalena and 67% in Atillo), indicating that organic matter is highly recalcitrant and undergoes minimal degradation, stabilizing phosphorus in sediment-bound forms [88]. Moreover, high Fe concentrations (7532–8197 µg/g) in these lakes limit phosphorus release, as Fe-P complexes remain stable under oxic conditions [89].

In contrast, the unprotected lakes Colta and Yambo, located at lower altitudes (3312 m and 2600 m, respectively) with warmer temperatures (12–15 °C and ~13 °C) and lower precipitation (~717 mm and ~500 mm), exhibit a different set of controlling factors. Here, six independent variables—Cu-FC, Pb, Ni, Cu, PO₄³⁻, and pH—each contribute equally (16.5%) to ecosystem classification, suggesting more complex biogeochemical interactions. In these lakes, warmer conditions promote nutrient cycling, and anoxic conditions enhance phosphorus release from Fe-bound sediments [90].

The CART model confirms that Cu, Ni, and Pb accumulation in Colta and Yambo is primarily linked to anthropogenic sources rather than natural geochemical processes. Colta, a mesotrophic lake under restoration, shows a weak correlation between organic matter and organic carbon (35%), indicating that intense microbial decomposition and nutrient release from sediments contribute to system destabilization [77]. Yambo, a hypertrophic lake heavily influenced by agriculture, wastewater discharges, and tourism, exhibits the highest Pb concentrations (5.6 ± 0.7 µg/g), likely due to emissions from boat traffic and nearby highways [82].

Additionally, in eutrophic systems, Cu and Ni accumulation is strongly associated with phosphate enrichment, likely originating from fertilizers, fungicides, and industrial inputs [83]. The high PO₄³⁻ concentrations in Yambo (678.5 ± 162.7 mg/kg) facilitate heavy metal solubilization, increasing mobility and bioavailability [87]. However, the presence of CaCO₃ in Yambo sediments may mitigate metal bioavailability by forming insoluble hydroxides and carbonates, reducing ecological risk [45].

This study focused on the surface sediments of four Andean lakes with different trophic conditions, analyzing nutrient levels and selected trace metals. However, several limitations should be considered. First, sediment–water interactions were not evaluated, which limits the understanding of potential nutrient fluxes and metal remobilization under changing physicochemical conditions. In addition, redox potential (Eh) was not measured in situ, which restricts the interpretation of iron and manganese solubility and mobility across different trophic states, particularly given their redox-sensitive nature. Incorporating redox profiles in future studies would help clarify the biogeochemical behavior of these elements.

Furthermore, the analysis was limited to a subset of metals. Potentially toxic elements such as arsenic (As), mercury (Hg), and others were not included in the scope of this study. Their inclusion in future assessments is recommended to provide a more comprehensive evaluation of sediment quality and ecological risk. Finally, seasonal variation was not addressed; thus, sampling during different hydrological periods could yield a broader understanding of temporal dynamics affecting contaminant distribution and sediment chemistry in high-altitude lakes.

5. Conclusions

This study provides a comprehensive assessment of the physicochemical characteristics, nutrient dynamics, heavy metal accumulation, and ecological risks in the sediments of four Andean lakes in central Ecuador. The findings reveal that the trophic state and conservation status of these lakes strongly influence sediment composition, organic matter dynamics, and metal accumulation patterns.

The physicochemical analysis shows that eutrophic lakes (Yambo and Colta) exhibit higher electrical conductivity, increased phosphate availability, and elevated cation exchange capacity, favoring nutrient retention and promoting primary productivity. In contrast, oligotrophic lakes (Atillo and Magdalena) display lower nutrient levels, higher organic carbon sequestration, and greater mineral enrichment, mainly due to their geological and climatic conditions. These differences highlight the impact of anthropogenic activities on sediment geochemistry, with unprotected lakes showing stronger signs of eutrophication and nutrient mobilization.

The study confirms that sedimentary organic matter in Atillo and Magdalena is predominantly exogenous, transported from the surrounding steep slopes, while Yambo presents a greater contribution of endogenous organic matter derived from phytoplankton and macrophyte decomposition. This suggests that high-altitude lakes function as carbon sinks, whereas eutrophic systems are active in organic matter cycling, potentially contributing to greenhouse gas emissions.

Heavy metal analysis reveals that Fe, Mn, Ni, and Zn concentrations are highest in the oligotrophic lakes due to natural mineral weathering, whereas Pb and Cu are more prevalent in Yambo and Colta, likely linked to anthropogenic sources such as vehicular emissions, agricultural runoff, and wastewater discharges. Compared to other American lakes, the studied Andean systems exhibit lower contamination levels, with Cr and Cd concentrations below detection limits, indicating minimal industrial pollution.

The ecological risk assessment using the potential ecological risk index (PERI) and pollution load index (PLI) confirms that all lakes present a low to very low ecological risk, with PLI values below 1, indicating no significant pollution. However, localized risk hotspots in Yambo and Magdalena, where Pb and Mn concentrations are elevated, highlight the need for targeted monitoring and management strategies.

The CART model successfully differentiates between protected and unprotected lakes, identifying pH and phosphorus as key variables in oligotrophic lakes, while Cu, Pb, Ni, and PO₄³⁻ drive ecosystem classification in eutrophic lakes. The model underscores the importance of land use and anthropogenic pressures in shaping sediment quality, particularly in lakes with intensive human activities.

These findings reinforce the need for adaptive conservation strategies to protect Andean lakes from the ongoing impacts of eutrophication and contamination. Future research should focus on long-term monitoring of sediment dynamics, phosphorus cycling, and metal bioavailability to inform mitigation measures and sustainable watershed management in high-altitude lacustrine ecosystems.