1 Introduction

The water flow connectivity plays a crucial role in controlling the hydrogeological and hydrochemical processes in the landscape, influencing ecosystem diversity, nutrient mobilization and water availability, which causes great social impact [1]. The presence of isolated waterlogged topographic depressions, known as geographically isolated wetlands (GIW) in a topographic depression within a hydrographic basin, particularly in the interfluves, often does not clearly indicate their hydrological connectivity with other water bodies of water within the integrated system [2]. According to the authors, the multiple interactions in the space and time between GIW and river channel can occur through subsurface water movement in the same direction as the surface water, as well as through lateral flows exchanged with associated floodplain and riparian areas. Understanding connectivity is of particular interest in the study of catchment hydrology, and this subject is relevant to protect hydric resources and aquatic ecosystem [3]. When isolation, or absence of a surface connection, is recognized, the variation in water level (hydroperiod) in isolated wetlands is associated with various possible flow pathways that must be evaluated within a conceptual framework [4]. Hydrochemistry and potentiometry flow studies can facilitate the understanding of the hydrological context and connectivity within the landscape [5].

In the Paulista Peripheral Depression, a sedimentary geological-geomorphological compartment between the crystalline Atlantic Plateau and the basaltic cuesta, small wetlands apparently disconnected with river channels are widespread in the smoothed hills. In the field, the topographic depression covered by grassy vegetation are the main characteristics to distinguish small wetlands from the surround agricultural soil which originally was covered with a mosaic of vegetation comprising Savanna and Atlantic Forest. In the municipality of Cordeirópolis, located in São Paulo State, Brazil, isolated wetlands have developed over diabase rock from Serra Geral Formation and are surrounded by intensive sugar cane agriculture. These isolated topographic depressions fill with rainwater, which, due to the local climate and rainfall pattern, accumulates during the months of October to March (encompassed in the spring and summer seasons), and lose water during the months of April to September (encompassed in the autumn and winter seasons), thereby characterizing them as seasonal isolated wetlands [6]. Because there is no visible connection between wetland, river and groundwater, eventual subsurface flow can be a major factor affecting the hydrochemical composition of wetland soil–water, which also varies in quality due to changing dry and wet seasons [7].

In this region, the intensive land use conversion has exacerbated environmental degradation. Soil erosion, deforestation of native vegetation, and consequently, the vulnerability of springs and rivers have reduced the water storage and availability potential. Similarly, small, isolated wetlands are strongly affected, both directly and indirectly, by changes in land use and land cover [8]. The integration of small wetlands in the anthropogenic landscape mischaracterizes these ecosystems, impacting water quality and storage. Under these conditions, wetlands are vulnerable to the input of sediments, nutrients, inorganic and organic toxic elements, as well as the lowering of the water table, which affects aquifer recharge and river flow [9].

In Brazil, where 20% of the territory is covered by various types of wetlands, the lack of effective measures or legislation regulating their protection and management is a precondition for their destruction [10]. The environmental condition is more critical for small wetlands as they are more susceptible to be conversion than large ones [11]. Conservation is a challenge because the information on status, inventory, and accuracy of wetland delineations is still limited [12]. In addition to anthropogenic pressures, uncertainty regarding the hydrological dynamics of natural wetlands and their connections among surface soil water and aquifers contributes to disrupting the water cycle, thereby affecting biodiversity in small wetlands [8, 13].

The objectives of this study were (1) investigate the subsurface connection between a GIW with the adjacent water body to understand water connectivity in the slope and (2) evaluate the hydrochemical composition in the GIW to assess the contribution of surface water and groundwater to the wetland supply.

The hydrochemical and potentiometric results of this study will enhance the understanding of sustainable management of water resources. Additionally, it will support future research on the conservation and ecological management of GIW in the Paulista Peripheral Depression, which have been degraded since the mid-1970s due to intensified land use for urban development, mining, and intensive agriculture.

2 Materials and methods

2.1 Study area

The geographically isolated wetland (GIW) studied (Fig. 1) has approximately 56.5 m2 located in the southwest Cordeirópolis municipality (22º 29′ 22″ S, 47º 29′ 14″ W), adjacent to the Santa Gertrudes municipality, in the São Paulo State.

Fig. 1
figure 1

Source: elaborated by authors based on base maps

Hydrographic map and location of studied area. [14,15,16,17]

Full size image

Both municipalities have economic significance due to ceramic production and agroindustry, with a particular focus on sugarcane for ethanol production. Sugarcane accounts for 83.8% of the GDP in the agricultural sector of Cordeirópolis [18]. Geologically, the studied area is in the context of the Paraná Sedimentary Basin [19], locally outcropping the Serra Geral Formation [6] which is represented by diabase dykes and sills [20]. The presence of faults and fractures, resulting from magmatic intrusions in underlying units and their subsequent crystallization, is common. These structures play a significant role in the hydrodynamics of a topographic depression because fractures and faults tend to accelerate chemical weathering by increasing the contact between water and rock or soil interactions [21].

The diabase of the Serra Geral Formation is composed, on average, of 40% of plagioclase (labradorite to oligoclase—(Na, Ca)Al(Si, Al)Si2O8)); 23% of pyroxene (augite—Ca(Mg, Fe, Al)(Al, Si)2O6); 4% of pyroxene (pigeonite—(Mg, Fe2 + , Ca)(Mg, Fe2 +)Si2O6); 7% of opaques (magnetite (Fe3O4), ilmenite (FeTiO3) and sulfides); 0.5% of olivine ((Mg, Fe)2SiO4); and 25.5% of quartz-feldspathic mesostasis and vitreous material [22].

Geomorphologically, the study area is in the Western Geomorphological Province, called Paulista Peripheral Depression [19]. This compartment is situated between outstanding topographical structures, from the Atlantic Plateau and the crests of the basaltic Cuestas. On a local scale, Casagrande et al. (2021) presented the Digital Elevation Model (DEM) in which the authors characterized three surface compartments in the landscape which the same nomenclatures will be adopted in this work: (i) Uplands, (ii) Wetland (flooding area), and (iii) potential expansion of the flooded area (Fig. 2).

Fig. 2
figure 2

Photographs of the studied area (looking south) highlighting its topography ((a) corresponds to the image photographed on 08/2021 and (b) on 09/2021). The dashed lines and white arrows correspond to the delimitation of the flooding area; the yellow dotted lines correspond to the potential expansion of the flooded area and the red dashed line delimits the uplands

Full size image

In Fig. 2a (August 2021), taken after the wetland was burned, the flooding area (dashed line and white arrows) shows burnt vegetation, the zone of potential expansion of the flooded area (marked with yellow dotted lines) exhibits dry vegetation and uplands zone is characterized by sugarcane cultivation (outlined by the red dashed line). In the Fig. 2b (September 2021) the flooding area (dashed line and white arrows) displays healthy green vegetation, with soil still darkened by the burning; the potential expansion of the flooded area zone features dry vegetation (yellow dotted line); and uplands zone continues to show fully grown sugarcane cultivation (red dashed line).

The region's climate is humid subtropical with hot summers (Cwa), according to the Koppen classification (1931) for Brazil [23], with two well-defined seasons: a dry season during the coldest months (April to September) and a rainy in the hottest months (October to March). The highest rainfall record in the municipality of Cordeirópolis was in January 2022 (344.17 mm) and the lowest in July of the same year, with no precipitation [24].

The GIW is located near the source of the Santa Gertrudes Stream, a tributary of Corumbataí River, responsible for the water supply, along with three groundwater wells (Santa Gertrudes City Council [25]). As a result, the study area has a more significant influence on Santa Gertrudes due to its proximity and possible subsurface connection with the homonymous stream than on the municipality in which it is geographically located.

Water balance data, released by the River Basin Commitee [26], indicated that a sub-basin neighboring the studied area has been classified as a region of very high criticality. This classification occurs when the water balance between demand and water availability is zero or negative. The water balance represents the equilibrium between water availability, consumptive demands, and return flows.

2.2 Field procedures: well installation and water monitoring

Four monitoring wells (MW) were installed in the month July 21 (during the dry season), within the GIW’s inner perimeter. They were strategically positioned to represent the northeast, southeast, southwest, and northwest quadrants (Fig. 1). Drilling was carried out using a manual auger until reaching the impenetrable material, which is the diabase of the Serra Geral Formation. The depths achieved and the characteristics of the wells are detailed in Table 1.

Table 1 Data from wells installed in the study area
Full size table

The wells were lined with PVC tube (50 mm diameter), containing a filter at the lower end corresponding to holes of about 2 mm in diameter. To avoid clogging, the tubes were covered with fine yarn synthetic fabric and sealed with bentonite solution. At the end of the installation, the wells were drained with a bailer for cleaning and developed to ensure the representativeness and samples quality.

2.3 Sampling and hydrochemical analysis

Water Level (WL) and the following physical and chemical parameters were measured in situ: temperature (T), dissolved oxygen (DO), electrical conductivity (EC), total dissolved solids (TDS) and hydrogen potential (pH).

A measuring tape was used to measure the water level, and a Pro2030 multiparameter meter (YSI) was employed for measuring physical and chemical parameters. YSI was previously calibrated in the laboratory for EC and TDS with a KCl standard solution (1.0 mmol/L), EC (147 μS/cm at 25 °C) and DO with sodium sulfite solution (Na2SO3) < 0.1 mg L−1 of O2. The detection limits are as follows: -5 to 55ºC ± 0.3ºC for T; 0 to 20 mg L−1 ± 0.2 mg L−1 for DO; 0 to 500 µS/cm ± 1 µS/cm for EC; 0 to 100 g/L ± 0.1 g/L for TDS. For pH measurement, a Kasvi K39-0014PA portable pH meter was used, calibrated in the laboratory for pH 4.00 (4.01 ± 0.01 at 25 ± 0.2 ºC) and 7.00 (7.01 ± 0.01 at 25 ± 0.2 ºC) from 0 to 14 pH units ± 0.1.

The campaigns were conducted three times per semester, spanning one year, resulting in a total of six monitoring campaigns and five sampling campaigns, due to the low water level in September (dry period), insufficient for sampling. Thus, samples where representative of different hydrological periods. One day before each sampling campaign, the wells were drained to ensure the homogeneity and representativeness samples, so that they accurately reflected the aquifer composition.

For water sampling, a 96 cm long, 40 mm diameter bailer sampler was used. It was lifted using a 0.3 mm nylon thread. The collected samples were stored in opaque white polyethylene bottles, which were previously cleaned in the laboratory using a deionized water and hydrochloric acid (0.1 M) solution. The cleaning process was repeated three times with deionized water before the bottles were left to dry (upside down) for approximately four days. Subsequently, the bottles were placed in a thermal box maintained at a temperature of 4 °C, located at the Aquatic Ecosystems and Soils Laboratory (LEAS, Brazil).

The substances investigated in laboratory analysis were sodium (Na+), potassium (K+), calcium (Ca2+), magnesium (Mg2+), chloride (Cl−), sulfate (SO42−), bicarbonate (HCO3−), phosphate (PO43−), nitrate (NO3−) and carbonate (CO32−), following the NBR9898 standard guidelines [27]. In the laboratory, the samples were filtered using a glass fiber microfilter, model GF-1, 47 mm (pore 0.7 µm) in diameter.

Bicarbonate (HCO3−) concentrations were obtained by titrimetric method, titration with sulfuric acid 0.02 N between 1.0 and 500 ± 1.7% [28]; Carbonate (CO32−) was calculated by H+ in HCO3− subtraction [29]; Chloride (Cl−) was assessed using either argentometric Mohr's method with titration in a standard 0.0141 M silver nitrate solution for the concentration range between 0.0 and 100 ± 1.7% [28]; The Sodium (Na+) and Potassium (K+) were determined using the Flame Photometer MOD. 910 MS (20 and 100 ppm Na+ and K+ standard) 1.0 ppm ± 0.1 ppm resolution [30].

Using a Hach Company model DR 890 spectrophotometer [31] were quantifeid; Calcium (Ca2+) and magnesium (Mg2+) by colorimetric calagite method (0.0 to 4.00 mg L−1, Ca2+  ± 0.09 mg L−1 and Mg2+  ± 0.08 mg L−1 with CaCO3); Nitrate (NO3−) by cadmium reduction method 0 to 5.0 ± 0.2 mg L−1; Phosphate (PO43−) using ascorbic acid method from 0 to 2.5 ± 0.05 mg L−1; and Sulfate (SO42−) by sulfaver4 method 0 to 70 mg L−1 ± 0.5 mg L−1. For results above the concentration range, sample dilution techniques were performed.

Hardness is a measurement based on the calcium and magnesium content of the water. The hardness relative to CaCO3 (mg L−1) is obtained according to Eq. 1 by [32]:

$${\text{Th}}=2.5\cdot {\text{Ca}}+4.1\cdot {\text{Mg}}$$
(1)

where: Th is the total hardness (mg L−1); Ca is calcium content (mg L−1); and Mg the magnesium content (mg L−1).

To classify and compare various water groups based on dominant ion compositions, Piper’s diagram was generated through QualiGraf software [33].

2.4 Statistical analysis

Descriptive statistics consisted of calculating the mean, median, maximum, minimum, amplitude, interquartile range, standard deviation, variance, asymmetry and kurtosis according to [34].

To correlate the variables, the calculation of Pearson’s correlation coefficient (r) was applied, a measure of linear association between two variables defined by Eqs. 2 and 3:

$$r={\sum }_{i=1}^{n}\left({x}_{i}-\overline{x }\right)\left({y}_{i}-\overline{y }\right)$$
(2)
$$\sqrt{{\sum }_{i=1}^{n}{\left({x}_{i}-\overline{x }\right)}^{2}{\sum }_{i=1}^{n}{\left({y}_{i}-\overline{y }\right)}^{2}}$$
(3)

where: r is Pearson's correlation coefficient; xi and yi are the values of the observations on the two variables; and x̄ and ȳ are the means of the observations on the two variables.

The correlation coefficient (r) varies between − 1 and + 1, where zero means that there is no correlation, where 1 means a complete correlation and the sign of the coefficient shows the direction of the correlation, with negative r meaning that the variables are inversely related and although there is no specific rule for the qualitative interpretation of the coefficient [35], in this work the following meanings for the strength of the correlation will be adopted (Table 2):

Table 2 Qualitative interpretation adopted for Pearson's correlation coefficient values
Full size table

2.5 Potentiometry and subsurface water flow

For the potentiometric map, it was necessary to calculate the hydraulic loads by subtracting the water level (measured in meters) from the topographic elevation (also in meters) of each well. This calculation provided the topographic elevation relative to sea level datum, following the method described by [36].

The water level (WL) used was the well static WL before sampling, and the elevation was obtained through the photogrammetric acquisition conducted in February 2020 by Casagrande et al. (2021) using a drone model Phantom 4 Pro DJI (dji.com) coupled with dual GPS/GLONASS receiver and 20-megapixel RGB DJU 1'' CMOS sensor. Precise coordinates for the wells were acquired using GPS after their installation.

The hydraulic loads were interpolated by the mathematical method Inverse Distance Weighted (IDW), from the hydraulic head data and UTM coordinates, using the ArcGis Pro 3.1 software [37], to estimate the hydraulic head in points not sampled [38].

The interpolation map is the basis for the contour map of equipotential lines generated at intervals of 0.5 m. The flow direction was determined according to Darcy's Law, knowing that the components of the flow vector are tangential to the equipotential line (Qs) and normal to the equipotential (Qn) in relation to the coordinate system (s,n), as seen in Eq. 4 [39]:

$${{\varvec{Q}}}_{{\varvec{S}}}=-\frac{\partial{\varvec{\Phi}}}{\partial {\varvec{s}}}\boldsymbol{ }\boldsymbol{ }{\varvec{e}}\boldsymbol{ }\boldsymbol{ }{{\varvec{Q}}}_{{\varvec{n}}}=-\frac{\partial{\varvec{\Phi}}}{\partial {\varvec{n}}}$$
(4)

where: \({{\varvec{Q}}}_{{\varvec{S}}}\) is the tangential component of flow in the direction of flow (or parallel to the equipotential line); \({{\varvec{Q}}}_{{\varvec{n}}}\) is the normal component of the flow perpendicular to the flow direction (or perpendicular to the equipotential line); \(\frac{\partial{\varvec{\Phi}}}{\partial {\varvec{s}}}\) is the partial derivative of hydraulic potential (\({\varvec{\Phi}}\)) in relation to the coordinate \({\varvec{s}}\); and \(\frac{\partial{\varvec{\Phi}}}{\partial {\varvec{n}}}\) is the partial derivative of hydraulic potential (\({\varvec{\Phi}}\)) in relation to the coordinate \({\varvec{n}}\).

3 Results

3.1 Water physical and chemical parameters

The water level fluctuates by 1.68 cm between the wet season (October to April) and the dry season (May to September). Notably, the water level recorded in November, at the start of the wet season, exhibits characteristics similar to those of the dry season. This observation suggests a gap between rainwater percolation and the water level response, which can influence the availability of water volume in the area and in the wells (water levels) (Table 3).

Table 3 Water physical and chemical parameters for the studied area
Full size table

In general, the water level ranged from 0.11 m (lowest of all campaigns) to 0.93 m (wet period maximum) below the land surface in the wet periods and from 1.25 m (dry period minimum) to 1.79 m (highest of all campaigns) in the dry periods (Table 4).

Table 4 Descriptive statistical analysis for the values of the physical and chemical parameters for the studied area
Full size table

Temperature showed a 4 °C total and slight seasonal variation, higher during wet and warmer months and lower in the dry periods. The values of dissolved oxygen (DO) are within the expected for groundwater [40], ranging from 0.1 (lowest of all campaigns) to 5.6 (highest of all campaigns) and showed variation according to temperature oscillation.

The electrical conductivity (EC) showed seasonal variation with higher values in wet periods (520 μS cm−1) and lower values in dry periods (min: 24.8 μS cm−1). Second campaign had exceptionally high values due to specific wells MW-01 and MW-02, this may indicate that the upstream portion, where these wells are located, is receiving and retaining salts from the surrounding crops due to surface runoff. Amplitude in each campaign exceeded 200 μS cm−1, influenced by well discrepancies. Total dissolved solids (TDS) values closely followed electrical conductivity oscillations indicated high salinity and dissolved solids presence. Wet period had a maximum value of 327.4 mg L−1, while the dry season had a minimum of 12.3 mg L−1. Wells MW-02 and MW-01 influenced the TDS amplitudes significantly.

The pH values varied between 5.1 and 6.8 during the same campaign. Slight variation was also determined in the means values of each campaign, with the exception of the last campaign, which showed higher values, less acidic, than the previous campaigns. Slight seasonal variation was observed, with the most acidic pH values during the wet periods.

3.2 Hydrochemical assessment

Considering the anions and cations, the most abundant, above 5.0 mg L−1, were bicarbonate (HCO3−) and carbonate (CO32−) followed by sodium (Na+), sulfate (SO42−) and chloride (Cl−). The least abundant constituents were potassium (K+), calcium (Ca2+), magnesium (Mg2+), phosphate (PO43−) and nitrate (NO3−) (Tables 5 and 6).

Table 5 Analytical results for the substances of interest
Full size table
Table 6 Descriptive statistical analysis for the analytical results. All campaigns represents the statistics for each parameter for the entire investigation period (all 5 sampling campaigns)
Full size table

The maximum average concentrations of bicarbonate and carbonate (188.16 mg/L, 4th campaign) were observed during the rainy season, correlating with the rise in water levels within the topographic depression. These values were approximately 2.7 times higher than the concentrations observed at the end of the dry season (70.38 mg/L, 2nd campaign), likely due to the influx of water into the system. Conversely, the average concentrations of sodium (11.90 mg/L, 2nd campaign) and chloride (8.17 mg/L, 5th campaign) reached their peaks at the end and during the dry seasons, respectively, indicating that there is a greater concentration of salts retained as the water level decreases.

Samples collected from well MW-02 consistently exhibited carbonate, bicarbonate, and sodium concentrations approximately one order of magnitude higher than those in other wells. Carbonate, bicarbonate, and sulfate were also detected in wells MW-01 and MW-04 during the early and dry seasons, specifically in the 03/23 and 05/27/2022 campaigns. Slight concentrations of chloride and sodium were observed in well MW-03 during the first campaign in September 2021, still within the dry season. These results align with the electrical conductivity values, which similarly indicated higher concentrations in MW-02 and MW-01.

Analyzing the Piper diagram (Fig. 3), all samples were categorized as Sodium Bicarbonate, except for the MW-03 sample in the first campaign, which exhibited a classification of Mixed Sodic. This classification was due to the slight presence of chloride and sodium in that particular sample.

Fig. 3
figure 3

Source: prepared with software QualiGraf [33]

Piper’s diagram for the analyzed samples.

Full size image

3.3 Correlations with variable

Correlations were conducted using all the samples, exploring relationships between physical and chemical parameters and ion concentrations (Fig. 4), as well as correlations among the ions themselves (Fig. 5). These procedures allowed us to assess interdependencies and gain a better understanding of the mutual influences between the analyzed parameters.

Fig. 4
figure 4

Pearson's correlation coefficient between the values of physical and chemical parameters and ion concentrations (considering all campaigns)

Full size image
Fig. 5
figure 5

Pearson's correlation coefficient between ion concentrations (considering all campaigns)

Full size image

3.4 Potentiometry and water flow direction

The potentiometry results, when combined with the drainage map and topography, aimed to establish a relationship between water flow and rock fractures or surface drainage axes. Given that a topographic depression situated in an interfluvial position can be generated and oriented along fault lines, this positioning can determine the preferred paths for the flow of stored water.

The hydraulic interpolation results indicated a predominant direction to the East (Fig. 6), with slight oscillations towards the northeast in the 3rd campaign (January—wet) (Fig. 6b) and towards the southeast in the 6th campaign (September—dry) (Fig. 6f). Considering the layout of the monitoring wells in the area, the flow direction begins from wells MW-01 and MW-02 (west) towards MW-03 and MW-04 (east). It is also important to consider the movement of groundwater by gravity, since in wet areas the groundwater table is usually a moderate reflection of climate and topography [41].

Fig. 6
figure 6

Potentiometry and flow pathways for each monitoring campaign, being them: a 1st campaign (09/23/2021); b 2nd campaign (11/15/2021); c 3rd campaign (01/26/2022); d 4th campaign (03/23/2022); e 5th campaign (05/27/2022); f 6th campaign (09/16/2022)

Full size image

4 Discussion

4.1 Influence of landscape dynamics on physical and chemical parameters

Water level values demonstrate a direct rainwater contribution to the maintenance of the wetland, resulting in expansion during rainy periods and contraction in the dry season. This recurring dynamic defines the hydroperiod, which is influenced by hydrological and climatic processes.

Temperature values suggest that the collected water is influenced by atmospheric conditions rather than confined aquifers, where variations would be more closely linked to sampling depth (geothermal gradient) than ambient temperature [42].

The hydroperiod, influenced by rainwater, exhibits a monomodal flooding pulse, as described by Junk et al. [10], which corresponds to the wet months. While further monitoring is needed, groundwater fluctuation values have not exceeded the land surface, possibly due to climate change or intensive agriculture, which can impact soil water holding capacity [43, 44]. This wetland provides numerous environmental, social, economic, and water ecosystem services, including water storage and recharge, flood regulation, and mitigation of the effects of climate change and land use [45].

According to CETESB [40], electrical conductivity values exceeding 100 μS/cm generally indicate a high presence of salts. This suggests that these wells, located upstream in the watershed, are receiving, and retaining salts from the surrounding crops, a typical characteristic of wet areas. Additionally, the increased salinity of wetland waters is often attributed to the enrichment of ions such as Na+, Mg2+, and SO42− through weathering and groundwater flow [46]. Therefore, considering the heterogeneity, it is plausible that the introduction of sulfates through surface runoff (leaching) plays a substantial role in the rise of salt concentration in certain wells.

4.2 Hydrogeological context and the hydrochemical signatures

Seasonal influences are readily apparent in the water composition Bicarbonate in water can originate from two primary sources: (1) the dissociation of atmospheric CO2, involving the gaseous diffusion of CO2, and (2) by the dissolution of carbonates [47]. In the studied area, the first source is probably predominant. Bicarbonate stands out as the most abundant anion in water bodies, while chloride is reported as relatively prominent in lakes, stemming from natural sources due to mineral weathering or anthropogenic sources linked to agricultural fertilizer use [48, 49].

In a hydrogeological context similar to the one studied, an unconfined aquifer within the Serra Geral Formation, Gastmans et al. [50] demonstrated that groundwater flow is primarily influenced by topography and local drainage of discharge. Water composition in such aquifers typically features low solute content, with bicarbonate being the dominant anion and calcium and magnesium as the primary cations. There may be increases in sodium related to total dissolved solids, a characteristic also observed in the waters of the studied GIW. In the region, rainwater composition is predominantly composed of HCO3− (bicarbonate) and other ions of both natural and anthropogenic origin [51], a trend similar to that found in this research.

According to the piper diagram, the fractured aquifer Serra Geral waters are predominantly calcium or magnesian bicarbonates and, secondarily, sodium bicarbonates and the recharge occurs through rainfall on the basaltic soils, which reaches the alteration zones and fissures of the parent rock [52], the latter being consistent with the classification obtained in this study. Although there is no evidence of an intense vertical water flow at the level of rock substrate to characterize the studied area aquifer as fractured, the water percolates in soil of diabase alteration (rock of the Serra Geral aquifer), and the main source of water input is also rainfall.

4.3 Correlations between geological framework, seasonal hydrodynamics, and water parameters

The existence and dynamics of the GIW are dependent on meteoric water input. Precipitation serves as the primary water source for the topographic depression, justifying the seasonal variations in water levels and the extent of flooding during both dry and wet seasons. Due to this origin, it is expected that the water composition collected in the wells will be directly influenced by the composition of rainwater and by the chemical and mineralogical characteristics of the soil resulting from diabase alteration. Additionally, it may also be influenced by land use, which has been dominated by intensive sugarcane agriculture and conventional management practices for decades.

The chemical concentration pattern of water in a GIW is exceedingly complex due to various conditions and processes, including: (i) an increase in chemical concentration as a result of runoff from slopes directed toward the wetland during the rainy season, (ii) an increase in chemical concentration in response to evapotranspiration during the dry season, (iii) a decrease in chemical concentration during the wet period due to dilution, and (iv) variation in concentration as a response to nutrient absorption by the rhizosphere [53, 54]. Furthermore, correlating these concentrations with physical and chemical parameters is essential for assessing the condition, productivity, and sustainability of the environment, as well as contributing to its management [55].

The diabase, due to its mineralogical composition and commonly associated faults, is intensely susceptible to chemical weathering since its mineralogy is mostly dominated by less stable minerals (augite and plagioclase) according to the Goldich [56] series. In the region, according to Machado et al. [22], diabase is composed by plagioclase (labradorite and oligoclase), pyroxene (Augite and Pigeonite), opaque minerals, olivine and quartz-feldspathic mesostasis. These minerals, when undergoing weather changes in a humid environment, as studied by Clemente e Marconi [57], tend to release soluble ions such as Ca2+, Mg2+, K+ and Na+ into the water and aluminum and iron remain immobile, forming the soil constituent alteration minerals such as kaolinite and oxides, respectively. The diabase chemical weathering can be triggered by several chemical reactions (dissolution, hydrolysis, acidolysis and oxidation), and hydrolysis is one of the main responsible for changing the aqueous composition medium through the interaction of H+ or hydroxyl ions (OH−) present in the water with the unchanged rock minerals.

The silicate minerals chemical alteration, such as feldspars, is processed through the incongruent reaction of plagioclase, where the alkaline ions (Na+ and Ca2+) present in the feldspar structure are exchanged for the H+ solution, increasing the water pH, giving rise to an amorphous phase that will be transformed into kaolinite (Eqs. 5 and 6) [58]:

$$\begin{array}{c}CaA{l}_{2}S{i}_{2}{O}_{8}\\ Anorthite\end{array}\begin{array}{c}+2{\text{H}}^{+}\rightleftharpoons {\text{H}}_{2}{\text{A}}{\text{l}}_{2}{\text{S}}{\text{i}}_{2}{{\text{O}}}_{8}+{\text{C}}{\text{a}}^{2+}\\ \end{array}$$
(5)
$$\begin{array}{c}NaAlS{i}_{3}{O}_{8}\\ Albite\end{array}\begin{array}{c}+{\text{H}}^{+}\rightleftharpoons {\text{HAlS}}{\text{i}}_{3}{{\text{O}}}_{8}+{\text{Na}}^{+}\\ \end{array}$$
(6)

It is worth mentioning that the Serra Geral aquifer waters composition, according to Gastmans et al. [50], may reflect water–rock/soil interactions due to the incongruous dissolution of diabase minerals such as plagioclase (labradorite, with 30 to 50% albite) [59] and pyroxene (augite) that in the presence of atmospheric CO2, lead to the secondary formation of kaolinite, quartz minerals and clay, releasing calcium, sodium, magnesium, silica and bicarbonate (soil constituents), according to the reactions (Eqs. 7, 8 and 9):

$$\begin{array}{c}CaA{l}_{2}S{i}_{2}{O}_{8}\\ Anorthite\end{array}\begin{array}{c}+2{\text{C}}{\text{O}}_{2}+3{\text{H}}_{2}{\text{O}}\to \\ \end{array}\begin{array}{c}A{l}_{2}S{i}_{2}{O}_{5}{\left(OH\right)}_{4}\\ Kaolinite\end{array}\begin{array}{c}+{\text{Ca}}^{2+}+{2{\text{HCO}}}_{3}^{-}\\ \end{array}$$
(7)
$$\begin{array}{c}2{\text{NaAlS}}{\text{i}}_{3}{O}_{8}\\ Albite\end{array}\begin{array}{c}+2{\text{CO}}_{2}+3{\text{H}}_{2}{\text{O}}\to \\ \end{array}\begin{array}{c}A{l}_{2}{Si}_{2}{O}_{5}{\left(OH\right)}_{4}\\ Kaolinite\end{array}\begin{array}{c}+2{\text{Na}}^{+}+{2{\text{HCO}}}_{3}^{-}+4{\text{Si}}{\left({\text{OH}}\right)}_{4}\\ \end{array}$$
(8)
$$\begin{array}{c}{\left(Ca,Mg,Fe\right)}_{2}S{i}_{2}{O}_{6}\\ Augite\end{array}\begin{array}{c}+2C{O}_{2}+5{H}_{2}O\to 2\left(C{a}^{2+},M{g}^{2+},F{e}^{2+}\right)+2HC{O}_{3}^{-}+2Si{\left(OH\right)}_{4}+\frac{1}{2}{O}_{2}\\ \end{array}$$
(9)

In this sense, the sodium concentration, in non-confined conditions as it is considered, can be interpreted as a result of the exchange reactions consistent with the limit between Na+ and Ca2+ (forming secondary minerals with the elements mentioned), since it is more difficult to reach the limits of calcite saturation.

The strong positive correlation among Na + (plagioclase), CO32−+HCO3-, and SO42− (sulfate) ions, as well as their positive correlations with electrical conductivity, suggests that these ions are released into the water concurrently, likely due to mineral alteration processes occurring in the diabase and soil. Additionally, the moderate negative correlation between magnesium and temperature indicates that this ion is released in higher quantities during periods of drought, when there is reduced inflow and exchange with rainwater.

The transport or accumulation of solutes in a humid area is dependent on evapotranspiration [60]. Solute concentration increases until saturation during dry weather conditions and decreases during wet weather conditions [61]. In wetlands where water volume primarily comes from precipitation and runoff from slopes, water loss occurs through evapotranspiration. For example, in the Okavango Delta in Senegal, approximately 98% of the wetland's water volume originates from rain, while only an average of 2% represents the inflow from distal parts, and some is absorbed by plants [60]. According to the authors, plants play a significant role in solute assimilation and nutrient retention, which also affects the quantity and distribution pattern of solutes [60].

In the studied GIW, the low levels of Ca2+ and Mg2+ cations suggest that saprolite or rock is not the dominant source of ions dissolved in water. However, the increase in Mg2+ concentrations (associated with pyroxenes) during the dry period may indicate this association due to the greater interaction of stored water with rock (augite) and/or soil (clay minerals).

The soil's mineralogical composition can also influence the chemical composition of water through the alteration of secondary minerals, as they come into contact with water for an extended period. This is a significant aspect of GIW hydrology, where a mixture of meteoric, interstitial, and groundwater occurs. Considering the mineralogical constitution of the soil within the studied area, which originates from diabase alteration, its composition comprises minerals, such as kaolinite/halloysite, quartz, gibbsite, goethite, hematite, cristobalite, anatase, chlorite-serpentine, ilmenite, montmorillonite-chlorite, and nontronite. Both the rock substrate and soil play pivotal roles in influencing water composition. Chemical weathering reactions of these minerals may culminate in the liberation of cations into the aqueous medium. The clay minerals formed during the diabase alteration in the soil can influence the composition of groundwater through processes such as adsorption, ion exchange, and chemical reactions. These processes can lead to changes in water concentration as well as the release or retention of chemical elements. Adsorption occurs when ions in the water bind to the surface of clay minerals, while ion exchange involves the replacement of adsorbed ions with other ions present in the water.

The smectites group (in this case, montmorillonite and nontronite) can adsorb and exchange ions (Na+, K+, Ca2+, Mg2+ and H+) with water, which can affect its composition. In addition, with the advancement of weathering these minerals also release cations when changed to kaolinite [62]. Chlorite and serpentine can also interact with groundwater and release ions (Fe2+, Fe3+, Mg2+, Al3+ and H+) during the alteration process. Kaolinite is a clay mineral that has a low ion exchange capacity and usually represents the final product of weathering.

These processes are influenced by water availability, as measured at the monitoring well level, and water quality, which is determined by physical and chemical parameters [63]. Therefore, the duration of interaction between water and mineral surfaces, the composition of the mineral assembly, and the degree of weathering all directly influence hydrochemistry. This hydrochemistry can vary significantly from the composition of rainwater [64].

It is noteworthy that despite being a wet area, the strong seasonality and the mineral composition of the soil suggest an oxidizing environment that allows for significant chemical changes. Oxide-reduction features resulting from the remobilization of iron in soil horizons are common characteristics across all profiles. However, it's important to mention that the presence of low chroma colors (< 2), which typically indicate prolonged saturation periods [65], were not described.

Regarding the influence of NPK (nitrogen, phosphorus and potassium) fertilizers on water quality, the moderate negative correlation between PO43− and Ca2+ and PO43− and Mg2+ suggests that the wetland is fulfilling the function of a biogeochemical reactor, where the PO43− concentration tends to be controlled by its calcium phosphate (Ca3(PO4)2 adsorption and precipitation [66] and Mg3(PO4)2 and MgHPO4 [67], which may also corroborate the low Ca2+ and Mg2+ water concentration.

4.4 Potentiometry and flow pathways

The shallow water level and its fluctuations observed during different hydrological periods in the GIW indicate that the studied aquifer is unconfined and characterized by significant interactions and storage of meteoric water [68].

The GIWs, flooded or not, are the evolution landscapes result formed by natural or anthropic processes [69]. In the case of natural evolution, such as in the studied area, with no visible connection to rivers, the hydroperiod may be directly related to the behavior (flow and direction) of the unconfined aquifer, which is fed by precipitation.

Wetlands can receive flow inputs, recharge aquifers or both depending on the fluctuation of the water table that can advance and retreat, discharging and recharging, and the same wetland can be a discharge zone in a time of high flow and a recharge zone in periods of drought [70]. Considering these factors, the studied feature can be compared with a flow lake model, where it receives discharge from upstream areas and acts as a recharge zone downstream through subsurface and lateral flow.

The proximity with the Santa Gertrudes Stream source to the east, suggests a subsurface connection between the water bodies, with the possible availability of water at the spring being dependent on the flow from the GIW, which would act as a sinkhole. Springs are natural outlets of groundwater, typically occurring when the water table is close to the surface due to impermeable layers intersecting it. They are often found in lowlands, geological faults, or karst formations. Contact springs are common, forming on slopes, hills, or depressions, with defined points (water sources) or diffuse areas forming wetlands and ponds from water accumulation [71].

The potentiometric maps provide an initial flow simulation in the shallow aquifer and suggest a connection, through subsurface flow, between the GIW and the Santa Gertrudes Stream. This discovery is significant since the stream, which may be partially fed by the GIW, contributes to ensuring water security in the Cordeirópolis and Santa Gertrudes municipalities, both of which face recurring water scarcity during seasonal drought periods. However, to establish the spatio-temporal relationship between the flow of the wetland and the stream, it is necessary to quantify the residence time of soil water and the shallow aquifer [64]. Additionally, given that the aquifer is shallow in the wetland, a portion of the interstitial water is likely absorbed by plant roots, which is of great importance for ecological characterization and structural function, and it can also contribute to flow regulation [72].

5 Limitations and future aspects

One of the limitations of this study is the spatial and temporal restriction of the data sampled, which focuses on a limited number of wells and specific monitoring periods within a single GIW. The hydrochemical and potentiometric analyses were conducted over a single hydrological cycle, which may not fully capture seasonal and interannual variations. Additionally, the lack of a broader monitoring network constrains the understanding of subsurface connectivity on a larger scale.

Future studies could benefit from expanding the monitoring network to include more GIWs and integrating continuous water level and quality sensors. Furthermore, applying additional methods, such as isotope tracers, could enhance the accuracy of inferences regarding hydrogeological processes and connectivity between water bodies. This would provide a more comprehensive understanding of the dynamics at play and support more effective water resource management.

6 Conclusion

The results obtained through the measurement of physical and chemical parameters of water, assessments of water composition, and potentiometry have significantly enhanced the understanding of the ecological context of the study area. These findings have also aided in the development of hypotheses regarding its hydrogeological characteristics and potential socio-environmental significance. Notably, the results support the hypothesis of a possible subsurface connection between the study area and the Santa Gertrudes stream, a crucial water source for the city of the same name, indicating the relevance of this research in understanding hydrological dynamics in such environments. Additionally, this insight contributes significantly to the development of strategies for the conservation and management of water resources in the region, which is extremely important for promoting environmental and social sustainability.

The monitoring campaigns allowed identifying that water level showed variations between the rainy and dry periods, suggesting that the wetland receives a direct contribution from rainwater, and that the electrical conductivity values (greater than 100 µS/cm) may indicate that wells MW-01 and MW-02 (upstream) are receiving and retaining salts from the surrounding cultivation due to surface runoff.

With the hydrochemical results it was possible to notice seasonal influence on ion concentrations and the Piper’s diagram showed that the water composition can be classified as Sodium Bicarbonate. Additionally, Low levels of calcium and magnesium suggests the non-dominance of saprolite or rock as primary sources of ions dissolved in water, although the increase in magnesium (from pyroxene) concentrations of the dry period samples may lead to this association.

The correlations allowed the interpretation that the strong positive correlation between sodium (from plagioclase), bicarbonate and carbonate, and sulfate (from sulfides) with the electrical conductivity (Ce) and among themselves suggests that these are being released into the water simultaneously due to the alteration of the minerals present in the diabase and soil. Furthermore, the moderate negative correlation between phosphate and calcium, and phosphate and magnesium suggest that phosphate concentration may be controlled by its adsorption and precipitation of calcium phosphate and magnesium phosphate. This highlights the function of the wetlad as a biogeochemical reactor, as the land use can be the source of some ions and compounds due to the use of agrochemicals such as fertilizers (NPK, nitrogen-phosphorus-potassium), nutrients, pesticides, salts, metals, among others.

The potentiometry results showed predominant flow direction to the East, from wells MW-01 and MW-02 to MW-03 and MW-04 and the characterization of the flow suggests that the studied feature can be compared with the flow lake model, which supports the hypothesis of subsurface connection with the Santa Gertrudes stream. Furthermore, the shallow water level and the fluctuations in the hydrological wet and dry periods suggest that the studied aquifer is a free, phreatic aquifer, with strong conditioning and storage of rainwater.

7 Recommendations

It is important to highlight that there are several tools and modeling capable of expanding knowledge regarding water dynamics in wetlands that can be a field of study for further research, highlighting the quantification of the residence time of groundwater and the study of isotopes to investigate hydrological and geochemical processes, aiming to provide more detailed information about the origin, flow, and evolution of water in these areas.