Long-Term Variation of Hydrochemical Parameters in a Medium-Sized Tropical River: Effects of a Severe Drought

Abstract

Long-term studies are essential to understand ecological and climatic processes in lotic ecosystems. This study analyzed the effects of changes in river discharge on hydrochemical parameters in the Lower Paraíba do Sul River (PSR) sub-basin over 11 years (2009-2019), in periods before and after a severe drought. The annual ions and particulate flow rates discharge and their relationship with the El Niño Southern Oscillation (ENSO) index were also estimated. Results revealed distinct response of hydrochemical parameters to discharge between the period before and after the severe drought. The occurrence of ENSO phenomena showed an influence on the hydrochemical dynamics at the end of the PSR basin, through changes in the local seasonality and, consequently, in the internal processes of the river channel, intensified by human activities on the soils of the basin. Our results reinforced that the increased frequency and intensity of climatic phenomena will in part affect hydrochemistry and transport patterns along the PSR. However, these effects on hydrochemical parameters prove to be contrasting. This highlights the need for further spatial and temporal scale studies on regulatory factors on hydrochemistry and its impact on water availability in the PSR basin in a scenario of anthropogenic and climate change.

Keywords:
long-term study; hydrochemical; severe drought; climate phenomena; water resources

Introduction

Rivers act as important sites for biogeochemical processing and integrate terrestrial and aquatic ecosystems throughout their basins delivering to the ocean around 0.4 0.5 Pg per year of carbon globally.¹,² The river transport of particulate and dissolved materials is related to the size of the drainage basin, topography, rock geology, weathering, climatic factors, vegetation cover, and human activities.³,⁴ In addition, the hydrochemistry of a river can also be influenced by anthropogenic impacts on water quality such as increased urban and agricultural land use, fertilizer application, soil destabilization, increased population density and pollutants from discharge of untreated domestic and industrial sewage, which are of a significant concern,⁵,⁶ specially where dense urban areas face water deficit problems.⁷

The Paraíba do Sul River (PSR), considered of medium-sized, is one of the most important rivers in the Brazilian Southeast region, transporting an average of 0.08 Tg of dissolved organic carbon (DOC) annually.⁸ However, the quality of its water has been affected by anthropogenic impacts on its basin, aggravated by climate change, with periods of reduced flow due to changes in rainfall patterns.⁹,¹⁰ The Southeast region of Brazil has experienced intense drought events in the last 60 years.¹¹,¹² In the first months of 2014, the region experienced a period of extreme drought caused by atmospheric teleconnections that formed a high-pressure center in the South Atlantic, leading to a precipitation deficit in the region,¹³ drastically affecting the river and reservoirs levels. Therefore, the combination of low rainfall and the inefficiency of the water management system in the region resulted in a severe drought period in 2014 and 2015,¹⁴ with its effects persisting in the river flow values in the following years.

Long-term studies help to identify variations in the ecosystem,¹⁵-¹⁸ being crucial to understanding environmental impacts, providing valuable insights in their analysis for the formulation of public policies in taking adaptation and mitigation measures aimed at water resources in a scenario of intensification of climate events.¹⁴,¹⁵,¹⁷-²¹ In the world, several studies²²-²⁵ have been conducted about river long-term hydrochemistry. In this context, concerns about the effects of human activities on streamflow - stemming from water resource management, power generation, sewage discharge, irrigation, population growth, and the expansion of industrial and agricultural sectors, as well as the presence of heavy metals and other contaminants in the PSR basin - have, in recent years, stimulated a growing number of studies focused on temporal variations.¹⁴,¹⁷,²⁶-³¹ Therefore, the understanding of long-term effects caused primarily by changes in river discharge in consequence of anthoropogenic pressure and climate change in the hydrochemical characteristics of the PSR drainage basin is still scarce.¹⁷,²⁸ In a long-term study conducted by Ovalle et al.²⁸ (from 1997-2007) with samples collected at same station, indicated that the hydrochemistry of the PSR is controlled by a combination of hydrological mechanisms (flow pathways and dilution), biogeochemical processes, and climatic forcings of the El Niño Southern Oscillation (ENSO), in addition to anthropogenic pressures resulting from land use (urbanization, agriculture, industry).

The present study had the following objectives: (i) to perform a long-term hydrochemical analysis (2009-2019) at the PSR outlet and determine its relationship with the discharge variability between a period before and after a severe drought; (ii) to determine the annual export fluxes of hydrochemical parameters and their relationship with the ENSO.

Experimental

Study area

The PSR basin has an area of 61,307 km² and covers the states of São Paulo, Minas Gerais and Rio de Janeiro. The main channel extends for 1,100 km to the mouth, through a highly industrialized and urbanized region, with a hot subtropical climate and an average annual temperature of 18 to 24 °C. With approximately 9.6 million inhabitants, the basin serves 22.6 million people through the transfer of part of its waters to supply the metropolitan region of Rio de Janeiro City. Agriculture stands out for its intensive use of water. Urban areas occupy only 5% of the territory, and the distribution of land use includes fields / pastures and agricultural areas (50%) and forests / forest fragments (43%).³² The basin is divided into sectors: Upper, Middle and Lower. The Lower Paraíba sub-basin is integrated into the Lower sector (14,547 km²) (Figure S1, presented in the Supplementary Information (SI) section), and has a relief with flat areas and steep slopes, with a predominance of fields and pastures, which occupy 79% of the territory. Economic activities in the Lower Paraíba sub-basin are concentrated on livestock, agriculture and the extraction of gravel and ornamental stones.³²

Sampling and analytical methods

Fortnightly water samples were collected at the basin outlet in Campos dos Goytacazes, RJ, Brazil (21°45’06.9”S 41°19’33.0”W), at the Lower Paraíba sub-basin (Figure S1b, SI section), covering dry (April to October) and rainy (November to March) periods, since 1994. This area is the last region of the basin before the mouth, which receives all the pollution produced upstream. It is a fixed monitoring station, allowing fortnightly measurements over decades (from the 1990s to the present).

For this study, we took the data obtained between the years 2009 and 2019. This period was subdivided into: before drough (2009-2013) and drought period (2014 2019). The criteria adopted to define the drought period was based on an anomalous high-pressure event over the Central Atlantic, which led to a precipitation deficit in southeastern Brazil in 2014, resulting in the most severe drought in the Paraíba do Sul River basin in 85 years.¹³,¹⁴,³³ This fact resulted in changes in hydrochemical data also observed through river local discharge. Instantaneous discharge was estimated every sampling day using a flowmeter (General Oceanic model 2030 current meter) and cross-sectional area measurements. A metal sampler attached to an amber glass bottle was used for collection. In the field, water temperature, electrical conductivity (EC; model LF96, WTW) and pH (826 pH Mobile, Metrohm) were measured, while dissolved oxygen (DO) was determined using Winkler method, according to Golterman et al.³⁴ The samples were transferred to plastic bottles and refrigerated before arriving at the laboratory. The suspended particulate material (SPM) was determined by water filtration (0.45-µm acetate pore membrane). The final concentration was determined by the filter weight difference before and after filtering divided by the sample volume filtered.

Subsamples were filtered using GF/F membranes, stored in plastic bottles, and frozen (-20 °C) prior to analysis. Ammonium (NH₄⁺) was determined spectrophotometrically according to Carmouze.³⁵ The concentrations of cations sodium (Na⁺), calcium (Ca²⁺), potassium (K⁺), magnesium (Mg²⁺) were determined by atomic emission spectrometry with inductively coupled plasma with axial viewing geometry (ICP-OES; Varian 720 ES). The plasma operating conditions used for ICP-OES, as well as the analytical wavelengths, are listed in Tables S2 and S3 (SI section). Nitrite (NO₂^-), nitrate (NO₃^-), phosphate (PO₄^3-) and chloride (Cl^-) were determined by ion chromatography with mobile phase in a Metrohm 861 ion chromatograph supported by the IC Net software. Subsamples were filtered through GF/F membranes and re-filtered prior to analysis using cellulose acetate filters (0.2 µm porosity). An eluent for determination was prepared with sodium carbonate (Na₂CO₃; 3.2 μM), sodium bicarbonate (NaHCO₃; 1.0 μM), and sulfuric acid (H₂SO₄; 50 μM). The analyses were performed in duplicates or triplicates, according to the certified method or standard, rejecting variations above 5%. Comparisons with previous analyses were also performed to increase the reliability of the results. Total alkalinity (TA), expressed as carbonate, was determined by automatic titration using a Mettler DL21 titrator. Water samples for chlorophyll a (Chl-a) determination were filtered through Whatman GF/F filters (precombusted at 500 °C for 6 h). The filters were kept at -18 °C prior to analysis. Chl-a was extracted from the filters with 90% acetone and quantified by spectrophotometry (Shimadzu UV-2700) with measurement wavelengths at 630, 647, 664 and 750 nm. After these measurements, the samples were acidified with 0.1 M HCl and reanalyzed at 664 and 750 nm, according to Strickland and Parsons.³⁶ Dissolved organic carbon (DOC) concentrations were determined using a Shimadzu TOC VCPH analyzer through high-temperature catalytic oxidation after acidification and purging with ultrapure air. Annual river export fluxes were calculated according to Ovalle et al.²⁸ multiplying the instantaneous flow rates by the concentrations of dissolved and particulate matter.

Data analysis

The Shapiro-Wilk test was applied to assess the normality of the dataset. As the data did not follow a normal distribution, nonparametric statistical methods were used. Spearman correlation was used to verify the correlation between parameters, Mann-Kendall to analyze the tendency across the long-term discharge series and a Mann-Whitney to analyze the difference between the sampled periods. Linear regression and generalized linear regression model (Gaussian) were carried out to identify potential associations between parameters. The relationship between discharge and surface water contributions can be observed through the analysis of the association between discharge and SPM. Model equations, determination coefficients and p-values were reported in each case. All statistical analyses were performed with software R programming language, version 4.2.2.³⁷

The cross-wavelet power spectrum was used to compare the frequencies of the discharge time series and the ENSO (El Niño and La Niña) index, as well as to analyze their synchronicity over the studied time interval.³⁸,³⁹ Data related to the years of occurrence of the ENSO phenomena and the years with precipitation anomalies were obtained from the Center for Weather Forecasting and Climate Studies (CWFCS)⁴⁰ portal. The Southern Oscillation Index (SOI) data were obtained from the National Oceanic and Atmospheric Administration (NOAA)⁴¹ portal, where they are calculated by standardizing the observed differences in sea level pressure between Tahiti and Darwin (Australia). This index corresponds to changes in the temperature of the Pacific Ocean associated with the occurrence of ENSO phenomena.

Results

Long-term variation of hydrochemical parameters

The instantaneous discharge at the PSR outlet showed seasonal regularity in its volume throughout the years covered by this study (Figure 1), reaching higher values between November and February and lower values between May and October. However, the maximum and minimum discharge varied in their amplitude during the study period, with a tendency to decrease from 2014 onwards (Table 1). Thus, the hydrochemical data obtained were divided into two intervals: before the drought period (2009-2013) and drought period (2014-2019), encompassing a period of 11 years (Figure 1). Before the drought period, the values ranged from 260 to 4546 m³ s^-1, while in the subsequent period they ranged from 31 to 1615 m³ s^-1 (Table 1) (W (Wilcoxon rank-sum statistic) = 14616, p < 0.001). When evaluating discharge variability over the study period (2009-2019), a decreasing trend was observed (tau (Kendall’s rank correlation coefficient) = -0.496, p < 0.0001).

Table 1
Summary of variables analyzed at the outlet of the Paraíba do Sul River (2009-2019) divided into the period before the drought (2009-2013) and the drought period (2014-2019)

Variable	Minimum			Maximum			Median
	Before drought	Drought period		Before drought	Drought period		Before drought	Drought period
Discharge / (m3 s-1)	260	31		4546	1615		735	241
pH	6.2	4.8		7.4	8.2		6.9	6.9
EC / (µS cm-1)	51.7	57.0		140.0	120.0		67.0	77.5
DO / (mg L-1)	3.8	5.5		21.0	10.0		8.4	7.1
TA / (mg L-1)	17.7	9.8		54.3	26.2		24.4	18.3
SPM / (mg L-1)	1.0	0.5		300.0	160.0		34.0	5.3
Chl-a / (µg L-1)	0.5	0.2		36.8	46.0		4.0	1.3
DOC / (mg L-1)	0.3	1.7		6.5	11.5		2.5	3.1
NO2- / (mg L-1)	0.005	0.005		0.161	0.032		0.018	0.014
NO3- / (mg L-1)	0.3	0.2		4.5	9.4		2.7	3.1
NH4+ / (mg L-1)	0.002	0.002		0.126	0.115		0.032	0.018
PO43- / (mg L-1)	0.010	0.010		0.190	0.770		0.048	0.057
Cl- / (mg L-1)	0.7	0.01		9.5	71.7		3.3	6.8
Na+ / (mg L-1)	0.8	2.3		11.8	23.6		6.3	8.1
Ca2+ / (mg L-1)	0.9	2.5		6.5	8.5		3.6	4.4
K+ / (mg L-1)	0.4	0.8		4.0	7.4		2.1	3.1
Mg2+ / (mg L-1)	0.2	0.5		2.1	3.9		1.4	1.7

Figure 1
Historical series of flow in the Lower Paraíba sub-basin. The series were divided into two intervals: before the drought period (2009 2013) and drought period (2014-2019).

The parameters Chl-a, DOC, EC, DO, TA, SPM, NH₄⁺, NO₂^-, NO₃^-, Cl^-, PO₄^3- showed significant differences between pre-drough and drought periods (p < 0.05) (Table S1, SI section). However, the highest maximum values for the analyzed parameters were observed during the period before the drought for the variables: Chl-a, EC, DO, TA, SPM and NH₄⁺ (Table 1). On the other hand, Cl^-, PO₄^3-, Na⁺, Ca²⁺, K⁺ and Mg²⁺ showed their highest values in the period after the drought (Table 1, Figures S3 and S4a, SI section) with a significant increase in PO₄^3- from 2017 onwards (Figure 3c). The availability of the analyzed major ions was as follows: Na⁺ (37%), Cl^- (21%), Ca²⁺ (20%), K⁺ (13%), Mg²⁺ (8%).

Figure 2
Annual variation in discharge (black) and variables of interest (red): correlations with (a) suspended particulate matter (SPM), and (b) total alkalinity (TA).

Figure 3
Annual variation in discharge (black) and variables of interest (red): correlations with (a) electrical conductivity (EC), (b) NO₃^-, and (c) PO₄^3-.

SPM showed an average reduction of 77%, being correlated with the discharge (rho (Spearman’s rank correlation coefficient) = 0.73, p < 0.001) (Table 1 and Figure 2a). Similarly, TA (Figure 3b) showed a positive correlation with the discharge (p < 0.001) and generally decreasing across the long-term studied period. On the other hand, EC and major ions showed a negative relationship with discharge (Table S4, SI section and Figure 3a). The same was observed for NO₃^- and NO₂^- (p < 0.001) with an increase in their concentrations during the drought years, except for NH₄⁺, which showed a reduction in its values during the drought (Figure S2d, SI section). Regarding the distribution of inorganic nitrogen forms (NO₃^-, NO₂^- and NH₄⁺), the most abundant form was NO₃^- (95%) (Figure 3b) followed by NH₄⁺ (4%) and NO₂^- (1%) (Figures S2c and S2d, SI section). However, although changes in discharge did not result in direct impact on the DOC values throughout the temporal analysis (Table 1, Figure S2b), an increase in concentration was observed during the drought period.

Long-term fluxes and export of SPM and dissolved elements in the Paraíba do Sul River

Table 2 presents the annual export flows for discharge, SPM and dissolved elements. When comparing the export results on an annual variation scale, in general, it was possible to notice a reduction in the flow of transported materials, except for PO₄^3- and Cl^- , as there was a reduction in the discharge flow during the dry season.

Table 2
Annual transport flows of dissolved and particulate material at the PSR outlet

Year	Flow rate / (km3 year-1)	TA	SPM	DOC	NO2-	NO3-	NH4+	PO43-	Cl-	Na+	Ca2+	K+	Mg2+
		(kt year-1)
2009	36	2104	6682	173	0.7	51.8	1.2	1.3	179	456	213	107	77
2010	38	1915	5205	238	0.3	71.8	0.9	1.1	321	416	222	124	81
2011	33	1503	8205	205	0.3	60.6	1	0.7	194	340	224	129	83
2012	28	1380	2856	178	0.3	68.1	0.6	1.4	148	320	197	144	83
2013	25	1122	1844	171	0.2	62.3	1	2.4	137	309	198	133	79
2014	13	468	261	63	1.4	67.3	0.7	1	-	202	102	63	44
2015	10	327	155	101	1.8	79.4	0.9	0.8	-	148	81	57	31
2016	12	396	558	116	0.3	85.2	0.9	1	-	150	103	71	34
2017	8	268	189	69	0.3	40.5	3.7	6.3	75	121	65	52	27
2018	9	310	243	58	0.2	3.8	0.1	0.6	119	132	79	56	32
2019	6	197	822	61	0.4	66.5	9.4	6.9	251	112	45	37	19

Climatic factors controlling long-term flow variation at the PSR basin outlet

Figure 4a shows the relationship between discharge and the SOI. Thus, prolonged periods with negative SOI values (in red) are associated with the occurrence of El Niño phenomena and prolonged periods of positve SOI (in blue) are associated with the occurrence of La Niña phenomena. According to CWFCS, during the study period, the years with occurrence of El Niño (in red) were 2009-2010 and 2015-2016, and the years with occurrence of La Niña (in blue) were 2010-2011 and 2017-2018.

Figure 4
Climatic variation and generalized linear regression models relating flow to proxy variables. (a) Median flow and the Southern Oscillation Index (SOI) over the study period (2009-2019), with negative SOI values indicating El Niño (red) and positive SOI values indicating La Niña (blue). (b) Model explaining the relationship between flow and SPM (p < 0.0001).

The relationship between flow and surface water contributions can be observed through the analysis of the association between discharge and SPM. Given the nonparametric nature of the data, we used a generalized linear model with Gaussian distribution to evaluate the relationship between flow and the variables mentioned. The model explained 62% of the variation found between SPM and discharge (Figure 4b).

Figures 5a and 5b revealed common and significant (p < 0.05) power areas during most years between SOI and the discharge and SPM variables, respectively. The series were in phase (arrows to the right) during most of the period, indicating that they reached maximum values simultaneously at 180 and 90°, corresponding to the positive and significant correlations between the two series (rho = 0.31, p < 0.001 and rho = 0.19, p < 0.05, respectively). However, there were periods of anti-phase (arrows to the left) in 2009-2010, again in 2011, and briefly in 2015-2016, associated with high power levels, indicating strong correlation over a year. Figure 5c suggests that the SOI and Na⁺ time series share common power areas throughout most of the series, being in phase during the 8-year period and in anti-phase in some parts of the 4-year period, presenting an inverse and significant behavior (p < 0.001), as in 2017, with this series showing more periods in anti-phase compared to the discharge and SPM series.

Figure 5
Analysis of the cross-wavelet spectrum between the medians of the time series of discharge, suspended particulate matter (SPM) and Na⁺. The contour lines show the 5% statistical significance level against red noise. The cone indicates where edge effects can distort the image, where lighter spots represent greater significance (p < 0.05) in the power spectrum, which indicates a greater variance for each period (years). (a) Cross-wavelet spectrum between the historical series discharge and Southern Oscillation Index (SOI); (b) cross-wavelet spectrum between the historical series SPM and SOI index; (c) cross-wavelet spectrum between the historical series Na⁺ and SOI index.

Discussion

Long-term variation of hydrochemical parameters

In general, the hydrochemical parameters analyzed in this study varied throughout the historical series in response to changes in discharge (Figures 2 and 3). However, responses exhibited distinct patterns considering the periods before and after the extreme drought event. In general, TA and SPM presented a positive pattern in relation to the discharge (Figure 2) across the whole temporal series (Table S4), with TA presenting a deacreasing tendency. This finding was distinct from that observed by other studies⁸,²⁸,⁴² in the same area, where the authors found a negative relationship with river discharge, with lower values during the wet period associated to higher dilution capacity, source contributions (weathering and soil) and local seasonality.

In the present study, the positive relationship between alkalinity and river discharge across the long-term temporal serie may be partly explained by land use and cover. The Lower PSR basin is characterized by the removal of natural vegetation, caused by unplanned agricultural activities, tending to promote erosive processes. The sugar-alcohol industry is the main economic activity in this area, which compromises the natural vegetation cover of the region, predominantly composed of fields and pastures (79%).³² In this process, carbonate minerals are added to correct soil acidity, which, through weathering of these minerals produces HCO₃^-, a processes known as liming widely used in the region. Consequently, the weakened soil becomes much more susceptible to surface runoff, increasing the transport of sediments and minerals to the river.⁴³ Therefore, during the rainy season, this flux tends to intensify, which explains part of the correlation found between alkalinity and discharge in the present study.

During the drought period, the decreasing trend in alkalinity showed that the hydrological dynamics modulates the alkalinity through changes in river connectivity. In this sense, urbanization, inputs of agricultural activities and/or untreated domestic sewage have been pointed out as the a possible alkalinity sources.²²,²³,⁴²,⁴⁴-⁴⁶ This relationship was demonstrated throught fluctuations in anions and cations as indicative of contributions from anthropogenic activities.²²,⁴⁷ In the present study, it was possible to observe a trend of increasing dissolved cations and anions (NO₃^-, Cl^-, Na⁺, Ca²⁺ and K⁺, Figures 3b and S3) during the drought period. However, the alkalinity variability was negative correlated with major ions and NO₃^- (Table S4). Contributions through these pathways depend on factors such as precipitation and flow rate, and are intensified in areas with soils susceptible to erosive processes.²⁸,⁴³ However, during the drought period, this effect weakened, reducing contributions from surface runoff and enhancing internal biogeochemical processes within the channel. Another possible explanation for alkalinity variations, mainly during the drought period, is the contribution of metabolism. The decreasing trend across the temporal long-term variation was followed by a higher amplitude and oscillation of pH values, with ranged from 6.23-7.43 before the severe drought to 4.83 8.15 during drought period. Rivers are known to be CO₂ supersaturated because of the net heterotrophic metabolism sustained by inputs of organic carbon from the catchment combined with autotrophic sources.⁴⁸,⁴⁹ This explanation can be reinforced by the by the increasing trend in DOC and the reduction in O₂ levels across the drought period in the present study (Figures S2b and S4c).

The major ions followed the decreasing order of Na⁺ > Cl^- > Ca²⁺ > K⁺ > Mg²⁺ throughout the sampling period. Geological characteristics are known to play a key role in hydrochemistry characterization. In this sense, Ovalle et al.²⁸ observed that the waters of Lower sector are of the sodium-bicarbonate type (Na-HCO₃^-), showing chemical similarity to other tropical rivers such as the Niger, Congo, Amazon, and Orinoco, indicating that the weathering of rocks in the region exerts a strong influence on the local hydrochemistry. Meneguelli Souza et al.,⁵⁰ when characterizing groundwater in the Lower sector, found similar results to this study, associating ion variation with regional geology and the intensive practice of sugarcane monoculture. In the region, compounds based on Na⁺ and Ca²⁺ are commonly used for fertilization and soil acidity correction. Besides, potassium-based fertilizers such as potassium chloride (KCl) and polyhalite (K₂Ca₂Mg(SO₄)₄.2H₂O) are also commonly used in sugarcane cultivation.⁵¹,⁵² The latter is a slow-release fertilizer with high water solubility, also providing other ions such as Ca²⁺ and Mg²⁺.⁵³ These fertilizers can alter the chemical composition of surface and groundwater through processes such as leaching and surface runoff.

In addition to the influence of agricultural practices, it is important to highlight the intense industrialization and urban densification as factors that contribute to changes in water quality.⁵⁴ Domestic effluents and solid waste represent the main sources of pollution in the PSR basin.⁵⁵ During the period analyzed, in the Lower sector, 63% of the population is served by the sewage collection network, but only 17% receive treatment.³² The discharge of untreated effluents can impact ion concentrations in river water, mainly increasing Na⁺ and Cl-^- levels, which are strongly associated with human diet.⁵⁴,⁵⁶

The inverse relationship with river discharge was also observed for NO₂^- and NO₃^-, with NO₃^- being the major inorganic nitrogen form found during the drought period (Figure 3b). This finding have been associated with the intensification of the nitrification process related to the discharge of untreated wastewater and agricultural practices.⁵⁷,⁵⁸ This may also explain the increase in PO₄^3-concentrations during the drought period (Figure 3c). Otherwise, it would be not possible to observe an increase in chlorophyll a during the drought period associated with the reduction in water volume and higher retention time, which would favor phytoplankton activity.⁵⁹ In addition, this period also showed the lowest levels of oxygenation, which may be related to the reduction of primary productivity rates and the increase in organic matter mineralization processes, fueled by the increase in DOC. This lower authocthonous carbon contribution has been also pointed out to stimulated microbial heterotrophy rather than autotrophy in turbidity rivers.⁶⁰ Moreover, the increase in DOC levels during the drought period may also be explained by the reduction in dilution processes associated with the increased contributions from internal production, groundwater and lateral input from anthropogenic sources, such as sewage discharge.⁸,²⁸

Long-term discharge and climatic factors driving the fluxes and export of SPM and dissolved elements in the Lower sector of PSR

The discharge and hydrochemistry at the PSR are influenced by the occurrence of climatic events. The increase in the frequency and intensity of these events affects the hydrochemical dynamics of the region, as well as the transport of nutrients to the estuary.⁶¹-⁶³ Studies¹⁴,²⁸,⁶⁴,⁶⁵ have pointed to a relationship between changes in climatic phenomena and precipitation anomalies in the PSR basin. The ENSO phenomena are associated with positive and negative precipitation anomalies in the basin. The occurrence of La Niña is associated with a predominance of negative precipitation, while El Niño is associated with a prevalence of positive precipitation and variation in long-term dicharge was found be related with ENSO.²⁸

In the present study, the analysis of export flows showed that most of the hydrochemical variables analyzed presented a reduction in annual discharge, being associated with the drought in 2014 and persisting until 2019 (Table 2), although the relationship of most parameters with the discharge in terms of concentration was negative.This fact can be explained because the total discharge in a given location takes into account the area of the drainage basin, considering the annual discharge. Thus, when the discharge is reduced, there is consequently a reduction in the transport of materials, as they are transported by the water flow.²⁸,⁶⁶ In other words, the reduction in the annual discharge was responsible for the reduction in the transport of ions (Na⁺, K⁺, Ca²⁺, Mg²⁺, NO₂^-, NO₃^- and NH₄⁺), TA and SPM during the drought period (Table 2). On the other hand, ions such as PO₄^3- and Cl^- showed an increase in discharge after the drought period (Table 2) following the tendency discussed in the sub-section “Long-term fluxes and export of SPM and dissolved elements in the Paraíba do Sul River”. In general, the relationship between precipitation and extreme discharge values has been observed by other authors,⁶⁷,⁶⁸ who also observed that extreme discharges directly affect the loads of transported materials, affecting the river renewal capacity, causing changes in the dilution potential and favoring the accumulation of materials present in higher concentrations.

SPM was associated with surface inflows²⁸,⁶⁶ according to the model performed in the present study (Figure 4b). Furthermore, the results of the cross-wavelet spectrum in the present study confirmed the connection between SOI and the variables discharge and SPM, with greater intensity in years with El Niño occurrence (2015-2016, Figures 5a and 5b), as well as demonstrating the connection between SOI and Na⁺ with greater intensity in La Niña years (2017, Figure 5c). Thus, those parameters can function as a proxy between the drainage basin and ENSO events.²⁸

However, the observations made in this study indicate that, even after the normalization in the rainfall volume in the region in years with El Niño occurrence, it was not enough to return to the discharge values and hydrochemical variables prior to the drought, except for PO₄^3- and Cl^-. Thus, the changes in the precipitation volume possibly caused by El Niño do not fully explain the variation between precipitation and discharge. In addition to precipitation, these variations are being influenced by changes in land use and climatological changes that have been affecting the Southeast region of Brazil.¹²,¹³,⁶⁴,⁶⁹

Despite of not considered in the present study, reservoirs and diversions stand out as one of the main factors associated with reductions in PSR water discharge across the whole drainage basin.³⁰,⁷⁰,⁷¹ Reservoirs significantly alter river discharge by storing water and releasing it in a controlled manner, which typically results in a more regulated and attenuated flow downstream. The combination of reduced precipitation and anthropogenic processes was responsible for intensifying the water deficit in 2014.¹⁴,⁶⁵ These processes reduces peak flows during floods but can also cause lower average flows, especially during certain seasons, and can shift water from wet to dry periods. These changes directly affected the PSR flow, resulting in a significant reduction.⁷² The water usage in the region includes the generation of hydroelectric energy; consumption by activities such as industry, agriculture, mining, and fishing; and urban supply, responsible for the water used by 14 million people.³² Nowadays, 83 different reservoirs and hydroelectric dams of varying sizes are found throughout the basin. Several dams are installed along its length, with the Usina Elevatória de Santa Cecília and Ilha dos Pombos (Pombo Island) dams located the closest to the river mouth, at 357 km and 184 km, respectively.⁷³

Over the study period, the population of Campos dos Goytacazes grew significantly, rising from about 460,000 to 507,000 inhabitants. The official data obtained from National Sanitation Information System⁷⁴ indicate that only in recent years has the municipality achieved sewage treatment coverage of around 70% of the collected volume, which corresponds to approximately 86% of the total population. In addition, in 2013, a new hydroelectric plant was installed in the basin.⁷⁵ Therefore, factors such as the spread of point and diffuse sources of pollution, the reduction in water flow associated with the presence of hydroelectric reservoirs, together with the reduction in precipitation in the Southeast region of Brazil, significantly affected the hydrochemical dynamics of the PSR.

Conclusions

The hydrochemical parameters analyzed in this study varied throughout the historical series in response to changes in discharge. However, these responses presented distinct patterns considering the period before and after the extreme drought in this study. We noted a decrease in annual discharge for most variables, except to PO₄^3- and Cl^-, which may be directly associated with the main land use class of the basin (fields and pasture) and the reduction in the potential for water dilution during the drought. Thus, we believe that these factors may have influenced the biochemical processes inherent to these macronutrients. However, a study of the local sources of these elements is necessary to identify possible changes in their biochemical cycles.

The SOI proved to be a possible proxy between the variation in discharge volume and ENSO phenomena, showing that the occurrence of these events can affect transport and local hydrochemical dynamics, which was evidenced through correlations with SPM and Na⁺. However, we observed that other factors can influence this dynamic, such as changes in land use that affect surface runoff and climate changes that interferes with rainfall formation. Thus, in the present study, we identified that the main factors controlling material transport and hydrochemical dynamics in the Lower Paraíba sub-basin are the discharge and internal processes of the channel, which are directly influenced by local seasonality and anthropic actions on the basin soils. The increasing frequency of ENSO phenomena suggests the intensification of this scenario, highlighting the need for local and global studies to understand the impact of these climatic events on water resources.

Our results reinforced that the increased frequency and intensity of climatic phenomena will, in part, affect hydrochemistry and transport patterns along the PSR. However, these effects on hydrochemical parameters prove to be contrasting. Moreover, anthropogenic pressures such as agriculture and sewage must also be considered. This highlights the need for further spatial and temporal scale studies on regulatory factors on hydrochemistry and its impacts on water availability in the Paraíba do Sul drainage basin in a scenario of anthropogenic and climate change.

Supplementary Information

Supplementary material 1

Supplementary material 1

Supplemental data (sampling location data, annual variation and flux of pH, dissolved organic carbon (DOC), NO₂^-, NH₄⁺, Na⁺, Ca²⁺, K⁺, Cl^-, Mg²⁺, chlorophyll a (Chl-a), dissolved oxygen (DO), Mann Whitney test, and Spearman correlation matrix) are available free of charge at http://jbcs.sbq.org.br as PDF file.

Acknowledgments

The authors are grateful for the support from the Laboratory of Environmental Sciences and to the Graduate Program in Ecology and Natural Resources of the State University of Norte Fluminense Darcy Ribeiro. We also thank Water Resources Fund CT Hidro, CNPq (409378/2022-4), Instituto Nacional de Ciência e Tecnologia de Transferência de Materiais Continente-Oceano (INCT) TMCOcean (405.765/2022-3) and CAPES (grant No. 88887.956201/2024-00).

Data Availability Statement

All data included in the text, as well as any additional information, may be requested from the corresponding author.

References

• ¹ Battin, T. J.; Lauerwald, R.; Bernhardt, E. S.; Bertuzzo, E.; Gener, L. G.; Hall, R. O.; Hotchkiss, E. R.; Maavara, T.; Pavelsky, T. M.; Ran, L.; Raymond, P.; Rosentreter, J. A.; Regnier, P.; Nature 2023, 613, 449. [Crossref]
  » Crossref
• ² Drake, T. W.; Raymond, P. A.; Spencer, R. G. M.; Limnol. Oceanogr. Lett. 2018, 3, 132. [Crossref]
  » Crossref
• ³ Dodson, S. I.; Introduction to Limnology, 1^st ed.; McGraw-Hill Education: New York, USA, 2005.
• ⁴ Calow, P.; Petts, G. E.; The Rivers Handbook: Hydrological and Ecological Principles, vol. 2, 1^st ed.; Blackwell Scientific: Cambridge, USA, 1994.
• ⁵ Chen, S.; Zhong, J.; Ran, L. S.; Yi, Y. B.; Wang, W. F.; Yan, Z. L.; Li, S. L.; Mostofa, K. M. G.; Biogeosciences 2023, 20, 4949. [Crossref]
  » Crossref
• ⁶ Zhou, Y.; Yao, X.; Zhou, L.; Zhao, Z.; Wang, X.; Jang, K. S.; Tian, W.; Zhang, Y.; Podgorski, D. C.; Spencer, R. G. M.; Kothawala, D. N.; Jeppesen, E.; Wu, F.; Limnol. Oceanogr. 2021, 66, 1730. [Crossref]
  » Crossref
• ⁷ Agência Nacional de Águas e Saneamento Básico (ANA); Conjuntura de Recursos Hídricos Brasil 2020, ANA: Brasília, 2020. [Link] accessed in February 2026
  » Link
• ⁸ Figueiredo, R. O.; Ovalle, A. R. C.; de Rezende, C. E.; Martinelli, L. A.; Rev. Ambiente Agua 2011, 6, 7. [Crossref]
  » Crossref
• ⁹ Cotovicz, L. C.; Vidal, L. O.; de Rezende, C. E.; Bernardes, M. C.; Knoppers, B. A.; Sobrinho, R. L.; Cardoso, R. P.; Muniz, M.; dos Anjos, R. M.; Biehler, A.; Abril, G.; Mar. Chem. 2020, 226, 103869. [Crossref]
  » Crossref
• ¹⁰ Vidal, L. O.; Lambert, T.; Cotovicz, L. C.; Bernardes, M. C.; Sobrinho, R.; Thompson, F.; Garcia, G. D.; Knoppers, B. A.; Gatts, P. V.; Régis, C. R.; Abril, G.; Rezende, C. E.; Sci. Total Environ. 2023, 857, 1. [Crossref]
  » Crossref
• ¹¹ Cunha, A. P. M. A.; Zeri, M.; Leal, K. D.; Costa, L.; Cuartas, L. A.; Marengo, J. A.; Tomasella, J.; Vieira, R. M.; Barbosa, A. A.; Cunningham, C.; Garcia, J. V. C.; Broedel, E.; Alvalá, R.; Ribeiro-Neto, G.; Atmosphere 2019, 10, 642. [Crossref]
  » Crossref
• ¹² Neves, A. O.; Vilanova, M. R. N.; Eng. Sanit. Ambiental 2021, 26, 339. [Crossref]
  » Crossref
• ¹³ Coelho, C. A. S.; de Oliveira, C. P.; Ambrizzi, T.; Reboita, M. S.; Carpenedo, C. B.; Campos, J. L. P. S.; Tomaziello, A. C. N.; Pampuch, L. A.; Custódio, M. S.; Dutra, L. M. M.; da Rocha, R. P.; Rehbein, A.; Clim. Dyn. 2016, 46, 3737. [Crossref]
  » Crossref
• ¹⁴ Marengo, J. A.; Nobre, C. A.; Seluchi, M. E.; Curstas, A.; Alves, L. M.; Mendiondo, E. M.; Obregón, G.; Sampaio, G.; Rev. USP 2015, 38, 30. [Crossref]
  » Crossref
• ¹⁵ Lohner, T. W.; Dixon, D. A.; Environ. Monit. Assess. 2013, 185, 9385. [Crossref]
  » Crossref
• ¹⁶ Elliott J. M.; Freshwater Biol. 1990, 23, 1. [Crossref]
  » Crossref
• ¹⁷ Marengo, J. A.; Alves, L. M.; Rev. Bras. Meteorol. 2005, 20, 215. [Link] accessed in February 2026
  » Link
• ¹⁸ Wilby, R. L.; Hydrol. Res. 2019, 50, 1464. [Crossref]
  » Crossref
• ¹⁹ Kellner, E.; Wiley Interdiscip. Rev.: Water 2021, 8, e1514. [Link] accessed in February 2026
  » Link
• ²⁰ Riveros-Iregui, D. A.; Covino, T. P.; González-Pinzón, R.; Hydrol. Processes 2018, 32, 2441. [Crossref]
  » Crossref
• ²¹ Wheater, H. S.; Gober, P.; Water Resour. Res. 2015, 51, 5406. [Crossref]
  » Crossref
• ²² Zhang, S. R.; Lu, X. X.; Higgitt, D. L.; Chen, C. T. A.; Sun, H. G.; Han, J. T.; J. Geophys. Res.: Earth Surf. 2007, 112, F01011. [Crossref]
  » Crossref
• ²³ Stets, E. G.; Kelly, V. J.; Crawford, C. G.; Sci. Total Environ. 2014, 488-489, 280. [Crossref]
  » Crossref
• ²⁴ Wu, J.; Cheng, S. P.; He, L. Y.; Wang, Y. C.; Yue, Y.; Zeng, H.; Xu, N.; Water Res. 2023, 244, 120492. [Crossref]
  » Crossref
• ²⁵ Cooper, R. J.; Hiscock, K. M.; Lovett, A. A.; Dugdale, S. J.; Sünnenberg, G.; Vrain, E.; Sci. Total Environ. 2020, 724, 138253. [Crossref]
  » Crossref
• ²⁶ Silva, R. C.; Fish, G.; Geociências 2019, 38, 587. [Crossref]
  » Crossref
• ²⁷ Azevedo, L. S.; Pestana, I. A.; Rocha, A. R. M.; Meneguelli Souza, A. C.; Lima, C. A. I.; Almeida, M. G.; Bastos, W. R.; Souza, C. M. M.; Chemosphere 2018, 202, 483. [Crossref]
  » Crossref
• ²⁸ Ovalle, A. R. C.; Silva, C. F.; Rezende, C. E.; Gatts, C. E. N.; Suzuki, M. S.; Figueiredo, R. O.; J. Hydrol. 2013, 481, 191. [Crossref]
  » Crossref
• ²⁹ Silva, M. A. L.; Calasans, C. F.; Ovalle, A. R. C.; Rezende, C. E.; Braz. Arch. Biol. Technol. 2001, 44, 365. [Crossref]
  » Crossref
• ³⁰ Gomes, P. R.; Pestana, I. A.; de Almeida, M. G.; de Rezende, C. E.; J. Hazard. Mater. 2023, 460, 132442. [Crossref]
  » Crossref
• ³¹ Oliveira, E. C.; Barboza, R. D.; Silva, B. G. G.; Diaz Filho, M. C.; An. Acad. Bras. Cienc. 2023, 95, e20220576. [Crossref]
  » Crossref
• ³² Associação Pró-Gestão das Águas da Bacia Hidrográfica do Rio Paraíba do Sul (AGEVAP); Relatório de Situação Bacia do Rio Paraíba do Sul, AGEVAP: Resende, Rio de Janeiro, 2020. [Link] accessed in February 2026
  » Link
• ³³ Britto, A. L.; Formiga-Johnsson, R. M.; Carneiro, P. R. F.; Ambiente Soc. 2016, 19, 183. [Crossref]
  » Crossref
• ³⁴ Golterman, H. L.; Clymo, R. S.; Ohnstad, M. A. M.; Methods for Physical and Chemical Analysis of Freshwaters, vol. 8, 2^nd ed.; Blackwell Scientific Publications: Oxford, UK, 1978.
• ³⁵ Carmouze, J. P.; O Metabolismo dos Ecossistemas Aquáticos: Fundamentos Teóricos, Métodos de Estudo e Análises Químicas; Edgard Blucher: São Paulo, Brazil, 1994.
• ³⁶ Strickland, J. D. H.; Parsons, T. R.; A Practical Handbook of Sea-Water Analysis, 2^nd ed.; Fisheries Research Board of Canada: Ottawa, Canada, 1972.
• ³⁷ R Core Team; R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing, Vienna, Austria, 2022.
• ³⁸ Addison, P. S.; The Illustrated Wavelet Transform Handbook Introductory Theory and Applications in Science, Engineering, Medicine and Finance, 2^nd ed.; CRC Press: Boca Raton, USA, 2017.
• ³⁹ Rösch, A.; Schmidbauer, H.; WaveletComp 1.1: A Guided Tour Through the R Package, 2018. [Link] accessed in February 2026
  » Link
• ⁴⁰ Center for Weather Forecasting and Climate Studies (CWFCS). [Link] accessed in February 2026
  » Link
• ⁴¹ National Oceanic and Atmospheric Administration (NOAA). [Link] accessed in February 2026
  » Link
• ⁴² Pezzini Jr., A.; Ovalle, A. R. C.; Geochim. Bras. 2009, 23, 255. [Link] accessed in February 2026
  » Link
• ⁴³ Borges, S. A.; Cunha, A. H. N.; Silva, S. M. C.; Vieira, J. A.; Nascimento, A. R.; Multi Sci. J. 2015, 1, 74. [Crossref]
  » Crossref
• ⁴⁴ Mortatti, J.; Oliveira, H.; Bibian, J. P.; Lopes, R. A.; Bonassi, J. A.; Probst, J. L.; Geochim. Bras. 2006, 20, 267. [Link] accessed in February 2026
  » Link
• ⁴⁵ Raymond, P. A.; Hamilton, S. K.; Limnol. Oceanogr. Lett. 2018, 3, 143. [Crossref]
  » Crossref
• ⁴⁶ Tappin, A. D.; Navarro-Rodriguez, A.; Comber, S. D. W.; Worsfold, P. J.; Environ. Sci.: Processes Impacts 2018, 20, 1361. [Crossref]
  » Crossref
• ⁴⁷ Zali, M. A.; Juahir, H.; Ali, M. M.; Retnam, A.; Idris, A. N.; Sefie, A.; Tawnie, I.; Malays. J. Chem. 2023, 25, 110. [Crossref]
  » Crossref
• ⁴⁸ Abril, G.; Bouillon, S.; Darchambeau, F.; Teodoru, C. R.; Marwick, T. R.; Tamooh, F.; Ochieng Omengo, F.; Geeraert, N.; Deirmendjian, L.; Polsenaere, P.; Borges, A. V.; Biogeosciences 2015, 12, 67. [Crossref]
  » Crossref
• ⁴⁹ Cole, J. J.; Caraco, N. F.; Kling, G. W.; Kratz, T. K.; Science 1994, 265, 1568. [Crossref]
  » Crossref
• ⁵⁰ Meneguelli-Souza, A. C.; Pestana, I. A.; Azevedo, L. S.; de Almeida, M. G.; de Souza, C. M. M.; Environ. Monit. Assess. 2021, 193, 57. [Crossref]
  » Crossref
• ⁵¹ Bhatt, R.; Oliveira, M. W.; Santos, D. F.; Rev. Gestão Secretariado 2024, 15, e3781. [Crossref]
  » Crossref
• ⁵² Otto, R.; Vitti, G. C.; Luz, P. H. C.; Rev. Bras. Cienc. Solo 2010, 34, 1137. [Crossref]
  » Crossref
• ⁵³ Chen, X.; Chen, X.; Jiao, J.; Zhang, F.; Chen, X.; Li, G.; Song, Z.; Sokolowski, E.; Imas, P.; Magen, H.; Bustan, A.; He, Y.; Xie, D.; Zhang, B.; Sustainability 2022, 14, 5646. [Crossref]
  » Crossref
• ⁵⁴ Ren, J.; Han, G.; Liu, J.; Gao, X.; Urban Water J. 2024, 21, 393. [Crossref]
  » Crossref
• ⁵⁵ Kury, K. A.; Bol. Obs. Ambient. Alberto Ribeiro Lamego 2008, 2, 117. [Crossref]
  » Crossref
• ⁵⁶ Sousa, D. N. R.; Mozeto, A. A.; Carneiro, R. L.; Fadini, P. S.; Sci. Total Environ. 2014, 484, 19. [Crossref]
  » Crossref
• ⁵⁷ Li, S. L.; Liu, C. Q.; Li, J.; Liu, X.; Chetelat, B.; Wang, B.; Wang, F.; Environ. Sci. Technol. 2010, 44, 1573. [Crossref]
  » Crossref
• ⁵⁸ Zheng, B.; Zhao, Y.; Qin, Y.; Ma, Y.; Han, C.; Environ. Earth Sci. 2016, 75, 1219. [Crossref]
  » Crossref
• ⁵⁹ Zhou, Y.; Evans, C. D.; Chen, Y.; Chang, K. Y. W.; Martin, P.; J. Geophys. Res.: Oceans 2021, 126, e2021JC017292. [Crossref]
  » Crossref
• ⁶⁰ Cotner, J. B.; Biddanda, B. A.; Ecosystems 2002, 5, 105. [Crossref]
  » Crossref
• ⁶¹ Hao, S.; Li, X.; Jiang, Y.; Zhao, H.; Yang, L.; Environ. Sci. Pollut. Res. 2016, 23, 17953. [Crossref]
  » Crossref
• ⁶² Hao, Y.; Lu, J.; Environ. Sci. Pollut. Res. 2021, 28, 41807. [Crossref]
  » Crossref
• ⁶³ Prasad, M. B. K.; Sapiano, M. R. P.; Anderson, C. R.; Long, W.; Murtugudde, R.; Estuaries Coasts 2010, 33, 1128. [Crossref]
  » Crossref
• ⁶⁴ Oliveira, D. M.; Carvalho, V. S. B.; da Silva, B. C.; Reboita, M. S.; de Campos, B.; Climate 2023, 11, 138. [Crossref]
  » Crossref
• ⁶⁵ Nobre, C. A.; Marengo, J. A.; Seluchi, M. E.; Cuartas, L. A.; Alves, L. M.; J. Water Resour. Prot. 2016, 8, 252. [Crossref]
  » Crossref
• ⁶⁶ Tardy, Y.; Bustillo, V.; Boeglin, J. L.; Appl. Geochem. 2004, 19, 469. [Crossref]
  » Crossref
• ⁶⁷ Carpenter, S. R.; Booth, E. G.; Kucharik, C. J.; Limnol. Oceanogr. 2018, 63, 1221. [Crossref]
  » Crossref
• ⁶⁸ Grabb, K. C.; Ding, S.; Ning, X.; Liu, S. M.; Qian, B.; Environ. Res. 2021, 195, 110759. [Crossref]
  » Crossref
• ⁶⁹ Paiva, A. C. E.; Martins, M.; Canamary, E. A.; Rodriguez, D. A.; Tomasella, J.; J. South Am. Earth Sci. 2024, 133, 104707. [Crossref]
  » Crossref
• ⁷⁰ Pacheco, F. S.; Miranda, M.; Pezzi, L. P.; Assireu, A.; Marinho, M. M.; Malafaia, M.; Reis, A.; Sales, M.; Correia, G.; Domingos, P.; Iwama, A.; Rudorff, C.; Oliva, P.; Ometto, J. P.; Limnol. Oceanogr. 2017, 62, S131. [Crossref]
  » Crossref
• ⁷¹ Paiva, A. C. E.; Nascimento, N.; Rodriguez, D. A.; Tomasella, J.; Carriello, F.; Rezende, F. S.; Sci. Total Environ. 2020, 720, 137509. [Crossref]
  » Crossref
• ⁷² Comitê de Bacia Hidrográfica do Baixo Paraíba do Sul (CBH BPSI); Nota à Imprensa - Fechamento da Foz do Rio Paraíba do Sul, 2019. [Link] accessed in February 2026
  » Link
• ⁷³ Agência Nacional de Energia Elétrica (ANEEL). [Link] accessed in February 2026
  » Link
• ⁷⁴ South-South Cooperation (SSC); National Sanitation Information System (SNIS), 2024. [Link] accessed in February 2026
  » Link
• ⁷⁵ Eletrobras; Usina de Simplício / Anta, 2023. [Link] accessed in February 2026
  » Link

Edited by

• Editor handled this article:
  Josué Carinhanha Caldas Santos (Associate)

Publication Dates

• Publication in this collection
  16 Mar 2026
• Date of issue
  2026

History

• Received
  19 Dec 2025
• Accepted
  20 Feb 2026