Lake Pomacochas is located in the town of Florida-Pomacochas, in a montane forest zone, on the eastern slope of the Peruvian Andes. It is a lake of tectonic origin at an altitude of 2,233 m. a.s.l., with an approximate surface area of 425.10 ha and an estimated depth of 75.5 m (Wetzel, 2001). It has a humid, warm temperate climate and an average temperature of 15°C. There are two marked seasonal periods with an annual precipitation of 1,104.5 mm (Barboza-Castillo et al., 2014). The wet season from November to April, with the highest precipitation peaks from January to March, and the dry season from May to October with a decrease in precipitation from June to August (Rascón et al., 2021). Precipitation does not present a water deficit throughout the year; on the contrary, there is a surplus of 77.0 and 102.0 mm (Vargas-Rivera). The lake is located in one of the main livestock-raising areas of the Amazon region, where the population settled in the surrounding area is around 7,000 inhabitants (Oliva et al., 2015; INEI, 2018). According to bathymetric studies, precipitation and subway runoff are the main sources of water to the lake (Barboza-Castillo et al., 2014). The main surface effluents are the Fichac and Congona streams, which cross the urban zone, and their effluent (Desaguadero) converges in the Pomacochas River (Figure 1). The main activities in the basin are extensive livestock ranching, vegetable production, aquaculture, and tourism in the western part of the lake (Oliva et al., 2015; ANA, 2016; Chávez et al., 2016).

Samples were collected considering seasonal rainfall. One sampling was executed in March (wet season) and another in August (dry season) of 2021. Considering the lake’s land use and activities nearby, four sampling points were established in the study area. Sampling point one (P1) was chosen for the direction of the lake discharge zone, sampling point two (P2) for its proximity to the area of population settlement and tourist activity, sampling point three (P3) for its proximity to the livestock production zone, and sampling point four (P4) for its proximity to the tributary inflow zone (ANA, 2016). Water and sediment samples were collected in triplicate at each sampling point. Along the water column, three depths were established using the Secchi disk of 20 cm diameter, which combines white and black quadrants alternatively (Utete and Fregene, 2020). The first depth was considered at the surface level from 0 to 50 cm, the second was about the Secchi disk transparency value, and the third was at 2.5 times the Secchi disk transparency value (Potapov et al., 2019). At each depth, temperature (T°), dissolved oxygen (DO), potential hydrogen (pH), and conductivity (EC) were determined with multiparameter equipment, brand WTW and model 3630 IDS. On the other side, turbidity was determined with a portable turbidity meter, brand HACH and model 2100Q. Triplicate samples were collected from each depth in 1-L polyethylene bottles that were thoroughly cleaned and rinsed with sampling water to determine physicochemical parameters (Popek, 2018). However, for the determination of metals, samples were collected in 100-ml polyethylene bottles treated with a 10% 1M nitric acid solution for 30 min and rinsed with distilled or deionized water (EPA, 1992). Sediment samples were collected in triplicate (a 0.5 cm layer from the lake bottom) using an Ekman Dredge (Cross, 1987). All collected samples were immediately transported to the Soil and Water Research Laboratory (LABISAG) of the Universidad Nacional Toribio Rodríguez de Mendoza de Amazonas, located in Chachapoyas. They were stored at −20°C until processing (Hou et al., 2013). In the laboratory, water samples were filtered using a cellulose filter paper, qualitative grade F1002 CHMLab and thickness 190 µm. The filtered product was acidified with nitric acid (1 + 1) to pH < 2 (EPA, 1994). Sediment samples were dried at 50°C before grinding with an agate mortar and sieving on a 200-mm sieve.

Alkalinity was determined by titration with hydrochloric acid (HCL), hardness by titration with EDTA (ethylenediamine tetraacetic acid), and chlorides by titration with silver nitrate (AgNO₃), according to the methodology by APHA, AWWA, and WEF (30). Parameters such as nitrates (NO₃ ⁻) and nitrites (NO₂ ⁻) were determined using the methodology established by HACH (2000). Total dissolved solids (TDS), ammonium (NH₄ ⁺), sulfates (SO₄ ^2-), and chemical oxygen demand (BOD) were determined using the methodology established by APHAAWWA and WEF, 2017).

Pulverized sediment samples were digested with HNO3:H2O2 (EPA, 1996), and elemental concentrations of Cu and Zn were measured (Manoj and Kawsar, 2020). Water samples were determined from the filtrate, acidification, and appropriate digestion in spectroscopy of emission atomic for MP-AES, adapting of methodology for ICP (APHA et al., 2017). The presence and concentration of metals and metalloids in water, such as aluminum (Al), lead (Pb), arsenic (As), boron (B), copper (Cu), cadmium (Cd), zinc (Zn), magnesium (Mg), and calcium (Ca), and the presence and concentration of copper (Cu) and zinc (Zn) in sediment samples were determined.

Microwave plasma atomic emission spectroscopy was performed with an Agilent microwave plasma spectrophotometer, brand Agilent Technologies and model 4100 MP-AES, equipped with a standard torch, Inert OneNeb nebulizer, and dual-pass glass cyclonic spray chamber, brand Agilent Technologies, for all experiments. Nitrogen was obtained from the air using a nitrogen generator, brand Agilent Technologies and model Agilent 4,107. The pump speed was set at 15 rpm. Before reading the samples, 12 s was set for the consumption time, 12 s for the torch stabilization time, and 30 s for the rinsing time. The reading time was 5 s. The spectral intensity was the mean of three repeated readings per sample. The detection wavelength of 193.695, 213.857, 228.802, 249.772, 285.213, 324.754, 393.366, 396.152, 405.781, 588.995, and 766.491 nm was selected for the quantification of As, Zn, Cd, B, Mg, Cu, Ca, Al, Pb, Na, and K, respectively. Before the readings, the equipment was calibrated by using standard solutions of each element in different concentrations, prepared from a 1,000 ppm standard solution (Supplementary Figure S1). The standard solutions used were of Agilent brand, located in the spectrometry area of LABISAG. After each reading, the equipment recovered both concentration and intensity without the need to enrich samples.

All the data obtained were subjected to a normality test applying the Kolmogorov–Smirnov test and homogeneity of a variance test with the application of Bartlett’s test to determine which statistical tests to use. For the evaluation of the behavior of the physicochemical parameters and toxicological elements of the water profile in depth and seasonal period, a principal component analysis (PCA) was applied. PCA is a statistical method used to reduce the dimensions of a large data set (Van Der Maaten et al., 2009). At the same time, selecting the most significant variables is a good technique; discard those that are redundant or have a high correlation (Marin and Robert, 2014; Borcard et al., 2018). PCA recognizes the variance within a set of correlated variables to create a smaller group of uncorrelated variables called principal components (PC), which are weighted linear combinations of the novel variables (Thioulouse et al., 2018). In this study, PCA was determined using a correlation matrix. Eigenvalues were calculated to measure the significance of the components. Once the PCA was calculated, the number of components to be used was determined, using the criterion of considering a sufficient number of components able to explain between 70% and 90% of the total variation of the original variables (Rencher, 2012). Finally, a biplot was used to better interpret the first two principal components (Jolliffe, 2002).

To detect spatial, temporal, and depth variations in both water and sediments a non-parametric multivariate analysis of variance (PERMANOVA) based on permutations was used (Anderson and Walsh, 2013). On the other hand, a U Mann–Whitney test was applied to determine the significant differences between the parameters present in water and sediment.

The mean concentration of metals and metalloids present in the water samples was contrasted with the international standards established by the European Union for environmental quality in the field of water policy of European Union (EQS) (EU, 2008), Canadian standard set by the Ministers of the Environment for the protection of aquatic life (CCME) (CCME, 2007), National Primary Drinking Water Regulations (EPA) (EPA, 2009), and National Environmental Water Quality Standards of Peru (ECAs), for the category of conservation of the aquatic environment (C4) and the category of extraction and cultivation of hydrobiological species in lakes or lagoons (C2) (MINAM, 2017). The toxicological elements present in the sediments were contrasted with the Canadian sediment quality standard for the protection of aquatic life in fresh waters (CEQG) (CCME, 2001), considering the previous conversion to the required concentration units and considering the following evaluation parameters: interim freshwater sediment quality guidelines (ISQG) and probable effect level (PEL).

-Sediment concentration < ISQG = No adverse biological effects.

All statistical analyses were performed at a significance level of p < 0.05, using R software version 4.1.0 (R Development Core Team, 2021).

Table 1 shows the mean values and their standard error for the physicochemical parameters determined in the water. The water’s hydrogen potential (pH) values at the three depths ranged from 6.89 to 8.76, with a mean of 7.68, indicating a near neutral pH environment. The lowest dissolved oxygen (DO) values, which occurred in the dry season (S), ranged from 5.59 mg/L to 5.96 mg/L. Turbidity values are influenced by the wet season (H), with the highest values occurring at the sampling points and depths. TDS generally did not show significant variations, nor did conductivity whose mean value was 236.87 μs/cm². Biochemical oxygen demand (BOD) values ranged from 1.11 mg/L to 13.25 mg/L, with a mean of 4.95 mg/L.

The concentrations of metals and metalloids in the water and sediment profile are shown in Table 2. At the water profile, the highest concentration of aluminum (Al) was found at point 1 (P1) at depth 1 (D1), as well as lead (Pb) and zinc (Zn), and arsenic (As) at depth 2 (D2). Cadmium (Cd) presented the highest concentration value at point 4 (P4), depth 2 (D2), lead (Pb) presented the highest values at points 1 and 2 (P1, P2), and boron (B) presented its highest value at point P2 depth 1 (D1). In the sediments, zinc (Zn) was reported with its highest concentration value at point 4 (P4) in the dry season, as well as copper (Cu).

From the evaluating parameters in the water for both seasonal periods and each depth, seven main components (PC) were selected that explain 42.3% of the total variance. The value of each parameter by component was evaluated, considering a moderate correlation (p ≥ ± 0.50), from which it is reported that T°, DO, pH, EC, turbidity, BOD, Mg, Al, Ca, Zn, and Pb are the parameters with more significant weight for CP1 are shown; for CP2, pH, BOD, hardness, Al, and K; for CP3, chlorides, Na, and Zn; for CP4, chlorides; for PC5, nitrates, and ammonium; for PC6 and As; and for PC7 and Cu (Table 3).

The principal component analysis shows the distribution of the physicochemical parameters and toxicological elements determined in the water by seasonal period. This shows a grouping of those variables that are representative of the wet season, such as BOD, Zn, Mg, turbidity, Ca, DO, T, and pH, whereas for the dry season Pb, EC, and TDS. The analysis of the water profile at the three depths shows an overlap that demonstrates that Pb and TDS behaved similarly at the three depths, whereas other variables differed or were not influenced by depth (Figure 2).

The PERMANOVA analysis of the concentration of physicochemical parameters and toxicological elements present in the water indicates no significant differences by the sampling point (F = 1.694, p = 0.099), whereas at the depth and seasonal level, there are significant differences (F = 6.129, p = 0.001; F = 15.024, p = 0.001). Regarding the concentration of toxicological elements evaluated in the sediment, the PERMANOVA analysis indicates a significant difference in the concentrations by the sampling point (F = 2.701, p = 0.043), whereas at the seasonal period level, there is no significant difference (F = 2.186, p = 0.97).

About the parameters present in water and sediment, only pH, Cu, and Zn were selected as they are the only parameters whose values are available for both matrices. It was found that there are no significant differences between the matrices for pH (W = −1.168, p = 0.245). However, it is observed that there are significant differences between the concentrations of Cu (W = −9.474, p = 0.000) and Zn (W = −7.442, p = 0.000), with values higher in the sediment.

The concentration of the elements presents at each sampling point, at each depth level, and by seasonal period was contrasted with the concentration of the elements established by CCME standards. Overall, collected concentrations from this study are mainly above the established limit concentrations. Concerning the EQS and EPA standards, the concentrations of elements such as Cu, Pb, and As exceed the established limit values. About the national environmental quality standards for water (ECAs) in conservation of the aquatic environment (C4) and extraction and cultivation of hydrobiological species in lakes or lagoons (C2), the concentration of elements such as As and Pb present the highest risk by exceeding the limit concentrations (Figure 3).

After comparing the concentrations of heavy metals found in the sediments with the CEQG standard, it was found that both Cu and Zn values do not exceed the values established by the standard, for both the ISQG limit and the PEL limit (Figure 4).

The biochemical oxygen demand in the water profile of Lake Pomacochas presents lower values of 2 mg/L, more severe at points P1 and P3, at depth 1 in the wet season. These values per sampling point may be due to various external causes, such as land use and agricultural practices that enrich the lake with nutrients (Kemp et al., 2005). For example, P1 is influenced by urban effluent discharges and P3 by tourist activity (Rascón et al., 2021). Small BOD values at depth may be due to nocturnal oxygen depletion by plant respiration (Caraco and Cole, 2002), stratification events (Özkundakci et al., 2010), and wind-generated turbulence, leading to deoxygenation of the water column (Townsend and Edwards, 2003). With all these factors, water quality worsens in the wet season (Jia et al., 2021). Dissolved oxygen values do not present multiple variations possibly because the lake remains in continuous movement. This implies that the water column is frequently oxygenated, and the in situ parameter sampling shows relatively homogeneous values (Wang et al., 2022). Behavior was completely different with biochemical oxygen demand, whose digestion procedure allows for more concise data on the amount of organic matter present (Abd El-Mageed et al., 2022). There are similar studies where biochemical oxygen demand values are near 1 mg/L, as in the case of Lake Veeranam, where it is presumed that these values are the result of organic pollutants, which cause an oxygen demand that causes the death of aquatic life (Ramya et al., 2021). Since 2015, in Lake Pomacochas, values of biochemical oxygen demand between 7.2 and 8.1 mg/L have been reported. Research where microbiological parameters and physicochemical parameters, together with the trophic pollution index (ICOTRO), reported moderate pollution in the lake (Chávez et al., 2016). For 2016 and 2017, monitoring was developed in Lake Pomacochas. In this monitoring, Carlson’s trophic state index, Aizaki’s modified trophic state index, Toledo’s modified trophic state index, and Vollenweider’s trophic state index were calculated. These indices reported that Lake Pomacochas is in a very advanced eutrophic state due to the waste generated by livestock and agriculture carried out in the area. The highest levels of trophism were observed in the wet period, so the trophic state depends on the rainfall regime (Rascón et al., 2021). These results are similar to the data obtained since they show an increase in multiple parameters in the wet season. Special mention should be made of the pH, which increases in the wet season. This increase is mainly due to the geology of the High Andean basins, given that it is composed chiefly of calcium carbonate (limestone and calcite) (Benito et al., 2018; Schmidt et al., 2019). These characteristics suggest that Lake Pomacochas, having these high pH values, has a high buffering capacity in terms of acid contaminations related to agriculture and livestock (Ogato et al., 2015). The sulfate concentrations also showed an increase during the wet season, possibly due to diffuse pollution sources that carry pollutants from urban and agricultural areas. This would also justify the increase in chloride, nitrite, and nitrate values at some points and depths during the wet season (Ahmed et al., 2018; Wei et al., 2019).