Groundwater Quality and Health Risk Assessment in Trenggalek Karst Springs and Underground Rivers as a Drinking Water Source

Abstract

The karst landscape of Trenggalek Regency, located in several sub-districts including Dongko, Kampak, and Watulimo, is shaped by the Wonosari Formation and is characterized by springs and underground rivers. Due to water scarcity in the region, local communities heavily depend on these natural water sources. This study assesses the groundwater quality of 16 springs and 20 underground rivers to evaluate their suitability for consumption and associated health risks. Using the groundwater quality index (GWQI), human health risk assessment (HHRA), and statistical methods, various physicochemical parameters were analyzed, including pH, total dissolved solids (TDS), electrical conductivity (EC), and concentrations of iron (Fe²⁺), manganese (Mn²⁺), calcium carbonate (CaCO₃), and sulfate (SO₄). Water generally meets the World Health Organization standards for safe drinking. However, correlation analysis reveals notable mineral dissolution and possible anthropogenic influence. TDS strongly correlates with EC (r = 0.97), while Fe²⁺ shows significant relationships with Mn and TDS. Conversely, CaCO₃ shows a negative correlation with EC and TDS, suggesting alternative sources beyond rock weathering. The HHRA indicates higher non-carcinogenic health risks from Fe²⁺ contamination in underground rivers compared to springs. The study’s novelty comes in its integrated assessment of groundwater quality and health hazards in Trenggalek’s karst region, which uses GWQI, HHRA, and statistical analysis to show geochemical interactions and highlight iron-related health issues in underground rivers.

1. Summary

The unique topography in Trenggalek Regency, which includes geological and hydrogeological characteristics, found springs and underground rivers which are typical of karst landscapes formed by karstification. This area faces urgent challenges in the form of water scarcity, which drives local communities to rely on springs as a vital source of daily water supply. This study aims to evaluate the groundwater quality of springs and underground rivers in the Trenggalek karst area. The main objective is to assess the health risks of water use in this area. This study uses several methodologies, including the groundwater quality index (GWQI), human health risk assessment (HHRA), and statistical analysis. To meet these objectives, this study involved the collection and analysis of groundwater samples from 16 karst springs and 20 underground rivers. The novelty of this study is the correlation analysis of physicochemical parameters and ion compound content in water in the karst area. Based on the World Health Organization (WHO) standards, groundwater in the Trenggalek karst area meets the criteria for safe drinking water in terms of its physicochemical properties. However, the correlation analysis shows an increase in rock mineral dissolution in groundwater. The results of the health risk assessment revealed that the non-carcinogenic health risks from consuming iron-contaminated water in underground rivers were significantly higher than those associated with springs. The correlation matrix showed that most of the water quality parameters had statistically significant relationships. Meanwhile, the negative correlation of calcite rocks implies that its presence is not only caused by rock weathering but can also be influenced by anthropogenic sources. This suggests that ongoing communication with local residents regarding the risks posed by inorganic contaminants to human health in the area is needed.

2. Introduction

The rapid growth of the global population will lead to significant challenges, particularly in meeting the demand for groundwater resources [1]. The demand for water resources, particularly surface and groundwater, is closely linked to the global population growth rate [2]. Groundwater is essential for daily life, serving as a crucial resource for drinking water, household use, agriculture, and industrial activities [3]. Groundwater quality, particularly in tropical regions, is influenced by geological and climatic conditions, as well as human activities [4]. Groundwater, which is commonly utilized for daily needs and agricultural purposes, serves as a readily accessible water source within our living environment [5]. Important water sources throughout the world can be obtained in karst areas, which can store large amounts of water [6]. Pollution in karst aquifers exhibits more complex behavior compared to other types of aquifers due to the rapid infiltration of surface water into the groundwater system, which facilitates the easy transport of contaminants. This process can lead to significant changes in water quality [7].

The hydrological system in karst morphology, which develops due to the dissolution of soluble rock, consists of a combination of conduits formed by dissolution processes and water movement through the spaces between particles, known as diffuse flow. The availability and quality of water resources in such environments are significantly affected by human activities and other life processes occurring in the surrounding area [8]. Karst areas exhibit a higher susceptibility to pollution compared to other natural landscapes due to the limited ability of the rock formations to filter contaminants from groundwater. The dissolution of limestone or dolomite leads to an elevated concentration of calcite in the water, distinguishing karst hydrogeochemical conditions from those found in other geological settings [9]. Total dissolved solids and ion content in aquatic conditions can be measured by TDS parameter values and electrical conductivity (EC) values [10].

The geographical characteristics of Trenggalek Regency indicate its vulnerability to drought, particularly during prolonged dry seasons. This phenomenon is attributed to the region’s geological composition, which predominantly consists of limestone lithology. Consequently, groundwater extraction through dug wells is not a viable option for meeting domestic water needs, unlike in areas with more suitable aquifer systems. The dug wells in this region function primarily as rainfed reservoirs, accumulating water during periods of high rainfall intensity and drying up when precipitation decreases. A significant portion of the population relies on the local water utility for their daily water requirements. However, during the dry season, the limited capacity of the water supply infrastructure prevents equitable distribution across all areas, necessitating a rotational allocation system. In light of these challenges, the utilization of springs emerges as a crucial strategy to enhance the availability of clean water for residents.

Despite significant efforts over the past decade to improve access to safe drinking water in Trenggalek, East Java, numerous small rural communities continue to depend on individual wells, natural springs, or informal local water distribution systems with uncertain or insufficiently monitored water quality. As a result, these communities face potential health risks due to the consumption of substandard drinking water. This study aims to assess the number of springs and underground river sources, as well as evaluate groundwater quality and associated health risks in the research area. Specifically, this study examines the physicochemical properties of water from 16 karst springs and 20 underground rivers in the Trenggalek karst region, alongside an assessment of the potential health risks for communities that rely on these water sources for drinking purposes. The research employs the groundwater quality index, health risk assessment, and statistical analysis as its primary methodologies. A key aspect of this study is its originality, as no comprehensive examination of the water chemistry in these springs and underground rivers has been conducted before.

Geology and Hydrogeological Setting of the Study Area

The study area in Trenggalek is the regency of East Java Province. Trenggalek Regency has an area of 1261.40 km², with geographical conditions consisting of mountains and plains. The population is 680,929 people, with a population density of 540 people per square kilometer, (km²) and a population growth rate of 0.25%. Trenggalek has several types of soil: among them are Mediterranean MK, alluvial, and podzolic humid. The average rainfall is 1910 mm/year (Trenggalek Regency Statistics Agency, Trenggalek, Indonesia, 2022) (Figure 1). Springs and underground rivers are water sources used to meet the drinking water needs of the Trenggalek people.

The geological conditions of the study area are described based on findings by Samodra and Gafoer, [11] as well as Hartono, Baharuddin, and Brata [12]. During the Tertiary period, various types of rocks were formed, including intrusion rocks (Tom); ancient volcanic rocks, including the Mandalika Formation (Tomm) and the Wuni Formation (Tmw); clastic sedimentary rocks such as the Arjosari Formation (Toma), Jaten Formation (Tmj), Nampol Formation (Tmn), and (Tmo); and limestone formations, including the Campurdarat Formation (Tmcl) and the Wonosari Formation (Tmwl) (see Figure 2). Additionally, young volcanic rocks (Qav) and alluvial deposits (Qa) were formed during the Quaternary (Cenozoic) era.

The old volcanic rocks (Tomm) exhibit the broadest lateral distribution, covering approximately 45% of the area. They are primarily composed of volcanic breccias, tuffs, and limestone intercalations. Clastic sedimentary rocks (Toma) follow with the subsequent widest distribution, encompassing around 20% of the area. These rocks consist of breccias of various materials, siltstone, conglomerate, and limestone intercalations. This study focuses on limestone formations, specifically the Campurdarat Formation (Tmcl), which comprises crystalline limestone with claystone intercalations containing carbonate (about 15%), and the Wonosari Formation (Tmwl), which is dominated by reef limestone, bedded limestone, and sandy limestone (approximately 3%). Alluvial deposits (Qa) constitute around 15% of the area and are primarily composed of gravel, sand, silt, clay, and silt. Other rock units have a relatively small distribution area, collectively accounting for about 2% of the total area of the Trenggalek Regency (see Figure 2).

The tertiary group of rock units—comprising intrusive rocks (Tom), old volcanic rocks (Tomm, Tmw), and clastic sedimentary rocks (Toma, Tmj, Tmn, Tmo)—exhibits low hydraulic permeability. Consequently, due to the reasons mentioned above, this region falls into the category of areas with non-exploitable groundwater, meaning that the potential for economically viable groundwater extraction is extremely limited. Groundwater is only found in limited quantities within the weathering zone in valleys and can be located within fissure and fracture aquifer systems. This geological condition covers approximately 65% of the total area of Trenggalek Regency.

In unconsolidated alluvial deposits, intergranular aquifer systems are predominantly composed of fine-grained materials such as clay and silt, with occasional sandy layers. These aquifers are generally characterized by moderate to low permeability, limited thickness, and low transmissivity. As a result, the construction of dug and driven wells in such areas typically produces a water yield of less than 5.0 L per second.

In limestone aquifer systems (Tmcl, Tmwl), groundwater flow is primarily controlled by fissures, fractured zones, and solution channels. The permeability of these aquifers ranges from moderate to high, depending on the degree of karstification. The Tmcl aquifer is characterized by moderate permeability, whereas the Tmwl aquifer exhibits high permeability. Aquifer productivity in the Tmcl region is considered moderate, with well and spring discharges varying across a wide range. Figure 3 presents the spatial distribution of groundwater sampling locations in the study area.

In contrast, the Tmwl area boasts high aquifer productivity, with good discharges exceeding 10 L per second. According to the Indonesian Groundwater Basin Map by Minister of Energy and Mineral Resources Decree (No. 2, 2017), the study area lies outside a potential groundwater basin or does not qualify as a groundwater basin. Consequently, groundwater management in this study area requires separate regulation through a ministerial decree by the authority responsible for groundwater resources.

Land use in Trenggalek Regency encompasses a variety of categories, including coppice, forest, secondary forest, secondary mangrove, agriculture, mining areas, housing, mixed agriculture, paddy fields, and fish ponds. Underground rivers and springs are primarily in agricultural, mixed agricultural, forested, and coppice areas. It is worth noting that activities related to land cover can have both positive and negative impacts on water resources [13,14]. On the other hand, changes in land cover can lead to significant variations in hydrochemistry and groundwater quality, a phenomenon well-documented in studies such as Das and Pal [15].

3. Materials and Methods

The groundwater object point was gathered during the dry season, with 36 water samples taken from 16 karst springs and 20 subsurface rivers in the Trenggalek karst area (Figure 1). The study has some limitations, one of which is that the survey was conducted solely during the dry season, which occurred between August and September of 2022. The collection of data pertinent to groundwater samples and the evaluation of the physicochemical properties of groundwater in the selected area comprised readings of pH, electrical conductivity (EC), and temperature. The Horiba U-10 was used to measure the water quality at both dug and drilled wells, as well as springs. A 500 mL plastic bottle, which had been washed with water before collection, was used to store the water samples.The local community relies on these water sources for drinking and daily household needs. All the springs and groundwater-fed rivers studied flow continuously throughout the year.

The chemical characteristics of water samples include CaCO₃, SO₄²⁻, Fe²⁺, and Mn²⁺. Atomic absorption spectrometry (AAS) has been used with Shimadzu AA-7000 Atomic Absorption Spectrophotometer (Shimadzu Corporation, Kyoto, Japan)to estimate the quantities of Fe²⁺ and Mn²⁺, employing SNI 6989.84.2019 techniques. CaCO₃ was estimated using wet analysis and titration according to SNI 06-6989.12:2004 procedures. Sulfate was measured using a Shimadzu UV-VIS 1700 spectrophotometer (Shimadzu Corporation, Kyoto, Japan) and ASTM D 4327 procedures. TDS values were quantified using gravimetric methods as per SNI 6898.27.2019. Total hardness (TH) was computed as equivalent to CaCO₃. The measurements were conducted at the. Hydrogeology analyses were conducted in the Hydrogeology Laboratory at the Centre for Groundwater and Environmental Geology.

3.1. Groundwater Quality Index

The groundwater quality index (GWQI) serves as a comprehensive method for assessing overall groundwater quality in the study area by analyzing the concentrations of major cations, anions, and metal elements. The use of GWQI, which integrates an extensive dataset encompassing multiple sampling points and seasonal variations, provides a reliable framework for evaluating annual trends, spatial and temporal fluctuations, and broader patterns in groundwater quality dynamics [16,17].

The GWQI was determined through a systematic approach involving the following steps: (i) assigning weights (wi) to each water quality parameter according to its significance; (ii) computing the weighted values (Wi) and assessing the water quality rating (qi) using Equation (1); (iii) calculating the sub-index (SLi) for each parameter using Equations (2) and (3); (iv) deriving the final GWQI through Equation (4). The GWQI values were calculated using the following algorithmic procedures to ensure accurate and reliable assessment of groundwater quality [18,19,20]:

$$Wi={\displaystyle \frac{wi}{{\sum }_{i=1}^n.wi}}$$

(1)

$$qi=Ci/Si\times 100$$

(2)

$${SI}_i=Wi\times qi$$

(3)

$$GWQI=\sum {SI}_i$$

(4)

where

• W_i: weight of each indicator;
• W_i: relative weight;
• Q_i: ranking for each parameter;
• W_i: weight for each indicator was determined according to the value of water quality;
• C_i: measured concentration;
• S_i: guideline value according to the drinking water quality guidelines [21].

The water quality parameter’s weight value (wi) was determined based on its importance to the overall quality of water acceptable for human consumption. The weighting for each parameter was determined using references based on an analysis of each parameter’s influence on human health (

Table 1. Chemical parameters calculation of the water quality index.

Parameter	Unit	Weight (wi)	Relative Weights (Wi)	Si *
pH		5	0.23	8.5
EC	μS/cm	4	0.18	500
TDS	mg/L	4	0.18	500
Fe2+	mg/L	3	0.14	0.3
Mn2+	mg/L	3	0.14	0.1
CaCO3	mg/L	2	0.09	300
SO42−	mg/L	1	0.05	250

* Si guideline value according to the World Health Organization Guidelines for Drinking Water Quality [22]. Weight of each parameters according to [23].

).

The values of the GWQI are put into categories and graded based on five different levels. To assist in deciding when groundwater is suitable for various applications, such as drinking and agricultural, the ground water quality index (GWQI) score is classified into five categories based on the quality of the ground water [24]. These different groups are determined by the quality of the ground water. This particular method was selected from among the several that are available for determining the current state of water quality (

Table 2. Physicochemical properties of analyzed groundwater samples.

No	Code	Type	EC (µS/cm)	TDS (mg L−1)	pH	Fe2+ (mg L−1)	Mn2+ (mg L−1)	CaCO3 (mg L−1)	SO42− (mg L−1)
1	S 1	Spring	915.5	443.5	6.88	0.25	0.23	322	234.5
2	S 2	Spring	732	356.5	7.43	0.09	0.07	344.5	652
3	S 3	Spring	1137.5	739.375	6.44	0.19	0.11	223	445
4	S 4	Spring	886.5	576.225	6.74	0.22	0.24	765	356.5
5	S 5	Spring	1451	912.5	7.22	0.08	0.08	112.3	145
6	S 6	Spring	883	573.95	7.11	0.16	0.21	321	243
7	S 7	Spring	326.5	234	6.35	0.19	0.05	89.4	164
8	S 8	Spring	1154	50.1	7.23	0.12	0.04	87.3	157
9	S 9	Spring	987	661.5	7.51	0.09	0.07	69.2	321
10	S 10	Spring	1276	779	6.44	0.18	0.09	443	228
11	S 11	Spring	665	432.25	7.05	0.08	0.05	590	398.5
12	S 12	Spring	734	421	7.11	0.07	0.07	78.4	149
13	S 13	Spring	1115.5	725.075	6.63	0.25	0.14	146.2	187.5
14	S 14	Spring	998	634	6.44	0.18	0.09	115	342
15	S 15	Spring	1187	771.55	7.23	0.21	0.18	89.2	221
16	S 16	Spring	1099.5	714.67	6.14	0.19	0.22	117.3	167.5
17	UGR 1	Underground river	1766.5	1148.2	6.23	0.44	0.25	187	226.5
18	UGR 2	Underground river	987	769.5	6.67	0.29	0.17	298.5	199
19	UGR 3	Underground river	945	614.25	7.18	0.34	0.36	177.5	334
20	UGR 4	Underground river	2117	1366	7.23	0.38	0.39	87.43	187
21	UGR 5	Underground river	1284	784.5	7.18	0.45	0.07	254	186
22	UGR 6	Underground river	1469.5	835.5	6.21	0.14	0.03	74.22	211
23	UGR 7	Underground river	1287	836.55	7.06	0.33	0.29	338	332
24	UGR 8	Underground river	1349	876.85	7.25	0.48	0.57	218	167
25	UGR 9	Underground river	2216	1548	6.44	0.55	0.43	123	204
26	UGR 10	Underground river	966	627.9	6.5	0.33	0.22	498.5	312
27	UGR 11	Underground river	1087	886.5	6.88	0.42	0.37	338	446.5
28	UGR 12	Underground river	1289	899	6.66	0.08	0.06	366	347.5
29	UGR 13	Underground river	1187	774	6.82	0.06	0.08	388.5	555
30	UGR 14	Underground river	1011	657.15	6.22	0.26	0.16	442.5	346.5
31	UGR 15	Underground river	1341	871.65	7.11	0.15	0.09	366	334.5
32	UGR 16	Underground river	1094	711.1	7.05	0.20	0.10	288.5	333
33	UGR 17	Underground river	663.5	431.28	6.22	0.08	0.05	522	240
34	UGR 18	Underground river	554	360.1	6.75	0.19	0.14	345.5	332
35	UGR 19	Underground river	889	547.5	6.88	0.08	0.05	544	239
36	UGR 20	Underground river	1284	887.5	6.55	0.09	0.07	332	237.5

) [18]. For the purpose of calculating the GWQI, variables that were pertinent to both natural geological processes and probable anthropogenic impacts were utilized. A value of 5 was assigned to the pH scale. According to the pH scale, water can be classified as either acidic or alkaline, which has an effect on the solubility and toxicity of contaminants. The parameters EC, TDS, Fe, and Mn were given the most weight because of the enormous impact they have on water quality, particularly when it comes to drinking water. For example, the existence of pH, EC, TDS, Fe, and Mn accumulations in water can be an indication of either natural or man-made sources, which could possibly pose a threat to human health if they reach dangerous levels. Fe and Mn are the most crucial factors to take into account because both of these parameters are frequently discovered in karst regions, and an excessive amount of either one of them can have a substantial impact on the flavor, color, and health of the food [25]. Electrical conductivity (EC) is a measure of the capacity of water to carry electrical current, which is directly proportional to the amount of ions that are dissolved in the water. In groundwater, total dissolved solids (TDS) refers to the total amount of all inorganic and organic components. The purpose of these two metrics is to ensure that the carbonate and osmotic balances are preserved. According to [26]), SO₂ is a way to determine the buffering capacity of water as well as the possible sources of contamination.

3.2. Human Health Risk Assessment

HHRA is a computational model designed to evaluate the extent to which contact with polluted water endangers human health [27]. The detrimental effects of groundwater contaminants on human health through oral consumption and skin contact have been examined in several earlier studies [28]. Non-carcinogenic risks from ingestion and skin contact were assessed for two age categories: children (0–21 years) and adults (21–72 years) [8]. The hazard quotient (HQ) for both oral and dermal exposure was determined using Equation (5), which relies on the average daily dose (ADD) calculated through Equations (6) and (7)). These are as follows:

$$HQ={\displaystyle \frac{ADD}{Rfd}}$$

(5)

$$ADD oral={\displaystyle \frac{CW\times IR\times EF\times ED}{BW x AT}}$$

(6)

$$ADD dermal={\displaystyle \frac{CW\times SA\times Kp\times ET\times EF\times ED {10}^{-4}}{BW x AT}}$$

(7)

where CW is the pollutant concentration in water [mg L⁻¹], IR is consumption rate [L day⁻¹], EF and ED are exposure frequency [days year⁻¹] and duration [years], BW is body weight [kg], AT is the meantime [days], SA is the exposed skin area [cm²], Kp is the coefficient of skin permeability in water [cm h⁻¹], and ET is the exposure time [h days⁻¹].

The potential for non-carcinogenic effects is calculated by the number of HQ for each pollutant and is expressed as a hazard index (HI) (Equation (8)). The total hazard index (THI) is calculated according to Equation (9) [23]. HQ, HI, or THI values above 1.0 indicate a non-carcinogenic health risk [29]. The equations are as follows:

$$HI=\sum HQ$$

(8)

$$THI = HI oral \times HIdermal$$

(9)

The distribution of major ions in the groundwater was examined using the hydrogeochemical method. The multivariate statistical analysis examined factors affecting the spatial distribution of CaCO₃, SO₄²⁻, Fe²⁺, and Mn²⁺. All statistics among EC, TDS, pH, CaCO₃, SO₄²⁻, Fe²⁺, and Mn²⁺ were conducted using Past 4.03 software. The methods used in this study are shown in Figure 4.

4. Results

The physicochemical parameters of the research area were compared with the guideline values recommended for drinking water and human consumption set by the World Health Organization. The water was neutral, with a pH between 6.14 and 7.51. pH influences the solubility of toxic metals that harm aquatic living organisms and human health [30]. The presence of dissolved minerals and organic substances in water can be seen from high levels of EC and TDS [31]. The WHO set the acceptable limit of TDS in drinking water at 1000 mg L⁻¹. The EC ranged between 326.5 and 2216 µS/cm. The total dissolved solids (TDS) ranged between 234 and 1548 mgL⁻¹, indicating the concentration of inorganic salts dissolved in water (see Table 2).

Total hardness (TH) is equivalent to calcium carbonate (CaCO₃) expressed as the concentration of calcium and magnesium ions in water, indicating the natural process of metal ion dissolution (8). Ca and Mg are essential elements for human health, and there is no evidence of adverse health effects caused by water hardness. The CaCO₃ is a crucial factor in the dissolving process of limestone [31]. Dissolved CaCO₃ values are affected by carbonate rocks and spring discharge (10). Total hardness (TH) ranged between 96.2 and 765 mg L⁻¹ (Figure 2). Carbon will be diluted at a large flow volume. In contrast, the concentration of carbon increases at low flow volumes, which enhances its aggressiveness when dissolved in water and accelerates the dissolution process.

Based on TH, water is classified as soft (TH < 17), slightly hard (17–60), moderately hard (60–120), hard (120–180), or very hard (>180). Based on the average TH values, the research area was classified as moderately hard, indicated by a blue to greenish color in the north from the hills to the central and eastern valleys; hard waters, represented by a green color, mostly in the central valley area; and very hard waters, marked in red, found in the southern part.The high concentration of CaCO₃ in the southern part can be influenced by the Campurdarat and Wonosari Formations (Figure 2), which are dominated by limestone. The dissolution rate in each catchment area determines the difference in the amount of CaCO₃. The concentration of sulfate (SO₄²⁻) ranged between 145 and 652 mg L⁻¹ (Figure 3).

Sulfate (SO₄²⁻) is likely derived from human activities, such as the use of agricultural fertilizers (e.g., urea, ammonium sulfate), livestock waste, and the release of sewage. Sulfate may also have resulted from dissolving gypsum, anhydrite, and sulfide minerals. This can be seen on the (SO₄²⁻) distribution map, which tends to be present in the western region of the study location (Figure 5). At this location, land use is dominated by agriculture and mixed agriculture (Figure 3).

The spatial distribution of sulfate (SO₄²⁻) concentrations across the study area is illustrated in Figure 6, showing higher levels in agricultural zones likely influenced by fertilizer application.

The variation of dissolved iron (Fe²⁺) in groundwater is presented in Figure 7, indicating elevated concentrations in the southeastern part of the study area.

Iron (Fe²⁺) is a chemical element found in almost all places on Earth, including rock layers and bodies of water [32]. Iron in water is generally dissolved as Fe²⁺ and suspended as colloidal grains—high concentrations of Fe²⁺ cause rusting of metals, smelly water, and high hardness [33]. Iron (Fe²⁺) ranged between 0.06 and 0.55 mg L⁻¹ (Figure 4). From a health standpoint, Fe²⁺ levels that exceed the threshold in the long term in drinking water will cause intestinal disorders and eye and skin irritation.

Iron (Fe²⁺) and manganese (Mn²⁺) are heavy metals generally found in groundwater [3]. Manganese (Mn²⁺) ranged between 0.03 and 0.57 mg L⁻¹ (Figure 5). The presence of Fe²⁺ and Mn in groundwater can be linked to geological origins, derived from minerals such as pyroxene, amphibole, biotite, iron oxide (e.g., hematite), and hydroxide (e.g., goethite) found in sediments [34]. The distribution of Fe²⁺ and Mn²⁺ can be affected by various physicochemical factors. The distribution of Fe²⁺ and Mn²⁺ ions in water samples also has a similar distribution pattern, which tends to increase in the southeastern region of the study location. When viewed from the geological map (Figure 2), this location contains a Mandalika Formation which is a volcanic rock, and the volcanic rock itself may carry heavy metals from pyroclastic deposits [35]. In addition, at this location there is alluvium that can enrich the Fe²⁺ and Mn²⁺ content from sedimentation results.

Springs with low discharge are more easily contaminated by human impacts than springs with high discharge because low discharge can cause pollutants to stay longer in the karst network. In contrast, high discharges shorten the contact time between the water and the host rock and reduce the possibility of the leaching of pollutants into the water [36]. Based on the GWQI, the springs have a good water status, which is better than that of the underground rivers.

Detailed physicochemical parameters of all groundwater samples are presented in Appendix A, while the complete GWQI values and their classifications are provided in Appendix B to support these results.

5. Discussion

The Pearson correlations in Trenggalek karst area indicated the relationships among parameters within a specific season.

Table 3. Pearson’s correlation in the research area.

	EC	TDS	pH	Fe	Mn	CaCO3	SO4
EC	1
TDS	0.974 **	1
pH	−0.022	−0.050	1
Fe	0.510 **	0.551 **	−0.063	1
Mn	0.441 **	0.486 **	0.103	0.808 **	1
CaCO3	−0.335 *	−0.307	−0.125	−0.161	−0.095	1
SO4	−0.244	−0.225	0.134	−0.211	−0.115	0.428 **	1

** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).

shows correlation matrices and all data had positive and negative correlations, representing a direct relationship with all hydrogeochemical data.

The Pearson correlation analysis indicates robust positive correlations among electrical conductivity (EC), total dissolved solids (TDS), and metal contents (Fe and Mn), implying a common geochemical source likely associated with mineralization. Electrical conductivity (EC) and total dissolved solids (TDS) have a near-perfect correlation (r = 0.974), signifying that dissolved ions predominantly influence conductivity. Iron (Fe) and manganese (Mn) exhibit a robust mutual connection (r = 0.808) and are significantly linked to electrical conductivity (EC) and total dissolved solids (TDS), indicating their contribution to the ionic load of the water. Conversely, pH demonstrates minor associations with all other parameters, indicating its independent variation from the concentrations of key ions and metals. Calcium carbonate (CaCO₃) and sulfate (SO₄) have a notable positive correlation (r = 0.428), perhaps indicative of carbonate–sulfate lithologies.; however, CaCO₃ demonstrates negative correlations with electrical conductivity (EC) and total dissolved solids (TDS), implying that carbonate-rich waters may possess reduced ionic strength. The matrix delineates two primary geochemical clusters, one characterized by mineralization (EC, TDS, Fe, Mn) and the other by carbonate–sulfate chemistry (CaCO₃, SO₄), with pH functioning as an independent variable.

The Pearson correlation studies elucidated the interdependence of many physicochemical factors. The specific characteristics and compositions of the parameters present in the aqueous solution determined the correlation between TDS-EC and Fe²⁺–Mn²⁺. These findings indicate that the dissolving of rock minerals influenced the geochemistry of the groundwater. The correlation between parameters shows the interactions of groundwater with rock, dissolution, and mineral decomposition.

The study examined correlation matrices of different water quality measures and observed that a majority of the metrics exhibit statistically significant correlations, indicating a strong link among themselves (

Table 4. Principal component analysis in the research area.

Parameter	1	2	3
EC	0.88	−0.04	−0.02
TDS	0.89	0.01	−0.06
pH	−0.02	0.12	0.96
Fe2+	0.80	0.34	−0.09
Mn2+	0.74	0.46	0.09
CaCO3	−0.44	0.68	−0.33
SO42−	−0.41	0.68	0.15
% of Variance	44.45	18.29	15.46
Cumulative %	44.45	62.73	78.19

). TDS exhibits a highly significant positive correlation with EC (r = 0.97), but its links with Fe²⁺, Mn²⁺, CaCO₃, and SO₄²⁻ ions are quite weak. Fe²⁺ exhibits a significant link with Mn²⁺ and TDS, with correlation coefficients of 0.81 and 0.55, respectively. The negative correlations between CaCO₃ and EC and TDS (r = −0.34 and −0.31) suggests that CaCO₃ is not solely derived from weatheringbut that other anthropogenic sources may potentially be contributing to the presence of CaCO₃ in groundwater. The alkalinity of a substance, usually CaCO₃, is determined by the presence of several ionic species such as hydroxide, carbonates, bicarbonates, and organic acids. These elements are inherent sources of water and numerous natural processes occur at each given location.

Hierarchical cluster analysis (HCA) was applied, utilizing squared Euclidean distance for linkage to categorize the chemical parameters in the underground river and spring waters according to their similarities. Figure 8 identified the influence of two distinct sources on the spring water. In the figure, “Ma” refers to spring (code S), while “Sb” refers to underground river (code UGR).

The parameters of Cluster A are from both spring and underground river waters, primarily resulting from water–rock interactions The primary source of constituents in the spring waters and underground rivers within this cluster is the dissolution of carbonate minerals. Furthermore, the origin of this cluster can be attributed to rainfall, as well as the dissolution of chloride-bearing minerals and sedimentary rocks.

Cluster B represents an underground river characterized by poor water quality. The elements in this cluster stem from a combination of geogenic factors, including weathering, the dissolution of minerals, and anthropogenic sources. The anthropogenic contributors encompass domestic wastes, sewage systems, irrigation return flow, and the application of chemical fertilizers. Magnesium originates naturally from dolomite and magmatic minerals such as biotite, hornblende, and olivine, whereas its anthropogenic sources are primarily linked to fertilizers.

Principal Component Analysis using Varimax rotation was applied. The rotated loading factors and communalities influence the groundwater and its chemical parameters (Table 4). PC1 depicts mineralization and metal content, with significant positive loadings for Fe (0.80), Mn (0.74), TDS (0.89), and EC (0.88). The interpretation of PC1 signifies general mineralization and metal concentration in the water. Elevated electrical conductivity (EC) and total dissolved solids (TDS) indicate increased levels of dissolved solids, whereas iron (Fe) and manganese (Mn) are prevalent geogenic metals. The environmental implications suggest that this component may represent regions affected by geological formations rich in Fe and Mn, potentially ultramafic or mafic rocks. PC2 reflects carbonate and sulfate, with significant positive loadings for CaCO₃ (0.68) and SO₄ (0.68). The interpretation: PC2 encompasses carbonate and sulfate chemistry, potentially associated with sedimentary or evaporitic rock origins. Environmental implication: This axis may differentiate waters affected by limestone or gypsum-bearing deposits. PC3 indicates acidity or alkalinity. Strong positive correlation: pH (0.96). The interpretation of PC3 is mostly influenced by pH, signifying fluctuations in acidity or alkalinity irrespective of other chemical factors. The environmental implications of this component may indicate localized pH alterations resulting from biological activity, acid drainage, or buffering capacity.

The dendrogram of groundwater samples based on hierarchical cluster analysis is shown in Figure 9, which classifies the samples into two main hydrochemical clusters.

The dendrogram analysis classified the groundwater samples into two main clusters, namely Cluster A and Cluster B. Cluster A is further divided into two sub-clusters (A1 and A2), which indicate closer similarities in hydrochemical characteristics among those sample groups. In contrast, Cluster B represents a distinct group with different water quality parameters. This clustering pattern highlights the hydrochemical heterogeneity within the study area.

The principal component analysis reveals that geogenic factors have a significant impact on the levels of EC, TDS, Fe²⁺, and Mn²⁺ in both groundwater and river water. The dissolution of carbonate rocks significantly impacts the quality of groundwater. Conversely, the dissolving of compounds that originate from agricultural and household residues also affects the total dissolved solids (TDS), iron (Fe²⁺), and manganese (Mn²⁺) levels.

The Trenggalek Regency is primarily characterized by its diversified agricultural practices. The application of fertilizers in large quantities is necessary for the preservation of open agricultural land and the enhancement of the quality of crop yield. The hydrochemical conditions of groundwater will inevitably be influenced by the utilization of fertilizers and the dissolved chemical leftovers that result after fertilization. There will be an impact on the hydrochemical conditions of groundwater and river water as a result of the growing relationship between the removal of agricultural land and plantations.

The principal component analysis shows that geogenic and anthropogenic influences have a considerable impact on both spring and underground river water. In contrast, the dissolution of chemicals originating from agricultural and household residues has a significant impact. The breakdown of carbonate rocks has a profound impact on groundwater quality. The Trenggalek Regency is chiefly distinguished by its diverse agricultural methods. Fertilizers must be used in significant quantities to preserve open agricultural land and improve crop yield. The use of fertilizers and the dissolved chemical leftovers that follow from fertilization will undoubtedly have an impact on the hydrochemical conditions of groundwater. As the relationship between agricultural land removal and plantations grows, the hydrochemical conditions of groundwater and river water will be impacted.

5.1. Groundwater Quality Index

The groundwater quality index (GWQI) in Trenggalek Regency exhibits a clear relationship with land cover conditions. Areas characterized by poor WQI scores often feature mixed agricultural land cover. Factors contributing to this degradation include the residues from organic fertilizers and erosion on agricultural land and plantations, which collectively affect the quality of groundwater sources. It is noteworthy that, on the whole, springs tend to have better WQI values compared to underground rivers. This difference can be attributed to the superior pollutant filtering capabilities of springs emerging from rock fractures. Conversely, wild streams and subterranean tributaries can serve as pathways for pollutants from the surface to enter underground rivers. Therefore, continuous monitoring and appropriate treatment measures are essential to mitigate contamination risks and ensure safe groundwater usage.

The groundwater in the area under study is utilized for a variety of reasons, including domestic use. The quality of groundwater is affected by a wide variety of causes, including geological conditions and conditions caused by human influences. It is necessary to treat it to reduce the amount of pollution, particularly the Fe and Mn contents. The significance level of the correlation coefficients and the values of the correlation coefficients will be of assistance in selecting the appropriate experimental procedures that will be employed for the treatment of water to reduce the contamination of underground rivers in the study region. To accomplish the objective of raising awareness among the general public about the need to preserve groundwater quality, the current study may prove to be helpful.This work has evaluated and mapped the geographic variability of groundwater quality.

The groundwater quality index (GWQI) in Trenggalek Regency exhibits a clear relationship with land cover conditions. Areas characterized by poor WQI scores often feature mixed agricultural land cover. Factors contributing to this degradation include the residues from organic fertilizers and erosion on agricultural land and plantations, which collectively affect the quality of groundwater sources. It is noteworthy that, on the whole, springs tend to have better WQI values compared to underground rivers. This difference can be attributed to the superior pollutant filtering capabilities of springs emerging from rock fractures. Conversely, wild streams and subterranean tributaries can serve as pathways for pollutants from the surface to enter underground rivers.

5.2. Human Health Risk Assessment

The human health risk assessment (HHRA) evaluates the potential adverse effects of iron (Fe²⁺) and manganese (Mn²⁺) exposure through groundwater consumption in the study area. The results indicate that iron concentrations in underground rivers pose a significant health risk, particularly for adults, with a risk characterization value exceeding the safe threshold. Long-term exposure to excessive iron levels may lead to health complications such as oxidative stress, organ damage, and increased susceptibility to chronic diseases. Conversely, manganese levels in groundwater sources generally remain within safe limits, with no substantial health concerns identified. However, continuous monitoring is essential to prevent any potential accumulation of contaminants over time. Given these findings, local authorities should prioritize water quality management strategies, including regular monitoring, public awareness campaigns, and the implementation of appropriate water treatment solutions to safeguard public health.

In this study, the health risk assessment considered the oral ingestion of iron and manganese through spring water and underground rivers, specifically for both adults and children. The health risk assessment (HQ) results for iron indicate an average value of 1.27, while the HQ value for children averages at 0.47. The HQ value for iron from underground rivers is notably higher for adults, suggesting a potential adverse health impact associated with unsafe consumption. Excessive iron intake is a concern due to its potential links to chronic diseases such as heart disease and diabetes.

For manganese the HQ values for adults exhibit a wide range (ranging from 0.019 to 2.459), with a median value of 0.348 and an interquartile range between 0.186 and 0.764. In contrast, the HQ value for manganese in children is 0.04. Notably, the HQ values for manganese from underground rivers indicate no adverse health effects, reflecting a safe status.

It is recommended that the results of the health risk assessment (HQ) value for iron be evaluated with greater caution than the results of the health risk assessment value for manganese in humans. Geographical considerations and the possibility of severe iron contamination in the vicinity of water sources need the local administration to pay serious attention to both of these factors.

The findings from the health risk assessment underscore that the non-carcinogenic health risks related to the consumption of iron-contaminated water are considerably higher in underground rivers compared to springs. Recognizing the potential threat that inorganic contaminants pose to human health, engaging with the local population consuming these elements in the research area is imperative. To address this issue, several steps are recommended, including conducting periodic tube well water quality tests, raising awareness about providing safe and affordable water sources, and developing low-cost household water treatment systems. These measures aim to mitigate health risks and ensure access to clean and safe drinking water for the affected communities.

The calculated groundwater quality index (GWQI) values for springs and underground rivers are summarized in

Table 5. Groundwater quality index (QWQI) of the studied karst water.

No	Code	Type	GWQI	Water Quality
1	S 1	Spring	97.43	Good
2	S 2	Spring	72.94	Good
3	S 3	Spring	88.39	Good
4	S 4	Spring	109.33	Poor
5	S 5	Spring	81.93	Good
6	S 6	Spring	92.97	Good
7	S 7	Spring	48.03	Good
8	S 8	Spring	69.75	Good
9	S 9	Spring	71.66	Good
10	S 10	Spring	87.95	Good
11	S 11	Spring	68.15	Good
12	S 12	Spring	57.49	Good
13	S 13	Spring	88.72	Good
14	S 14	Spring	76.71	Good
15	S 15	Spring	95.84	Good
16	S 16	Spring	94.40	Good
17	UGR 1	Underground river	132.75	Poor
18	UGR 2	Underground river	96.35	Good
19	UGR 3	Underground river	123.64	Poor
20	UGR 4	Underground river	159.90	Poor
21	UGR 5	Underground river	95.75	Good
22	UGR 6	Underground river	74.58	Good
23	UGR 7	Underground river	126.05	Poor
24	UGR 8	Underground river	169.15	Poor
25	UGR 9	Underground river	177.34	Poor
26	UGR 10	Underground river	107.69	Poor
27	UGR 11	Underground river	140.55	Poor
28	UGR 12	Underground river	83.08	Good
29	UGR 13	Underground river	85.85	Good
30	UGR 14	Underground river	96.29	Poor
31	UGR 15	Underground river	91.94	Poor
32	UGR 16	Underground river	86.75	Good
33	UGR 17	Underground river	61.47	Good
34	UGR 18	Underground river	76.04	Good
35	UGR 19	Underground river	69.78	Good
36	UGR 20	Underground river	81.54	Good

, highlighting variations from good to poor categories.

The results of the health quality index assessment for Fe²⁺ and Mn²⁺ exposure are presented in

Table 6. Health quality index.

No.		Fe2+ = 0.3		Mn2+ = 0.4
1	Average Concentration (mg/L)	0.22	0.22	0.16	0.16
2	Old	Children	Adults	Children	Adults
3	R (liter)	1	2	1	2
4	Fe2+ (day/year)	350	350	350	350
5	Dt (year)	6	30	6	30
6	Wb (kg)	15	55	15	55
7	t avg (day)	10,950	10,950	10,950	10,950
8	RFd (mg/kg.day)	0.006	0.006	0.14	0.14
appe9	CDI (mg/kg.day)	0.00	0.01	0.00	0.01
10	Risk Characterization	0.47	1.27	0.01	0.04
11	Status	Unsafe	Unsafe	Safe	Safe
12	Safe Concentration (mg/L)	0.47	0.17	0.01	0.02

, which shows that iron concentrations in underground rivers pose potential health risks.

6. Conclusions

This study confirms that geological and anthropogenic variables significantly impact the quality of spring water and underground river water, while the use of fertilizers and agricultural and domestic waste also influence the hydrochemical conditions of groundwater. Pearson’s correlation analysis reveals interdependence among various physicochemical factors, where the composition of parameters in the aqueous solution determines the correlation between TDS-EC and Fe²⁺–Mn²⁺. These findings indicate that the dissolution of rock minerals affects groundwater geochemistry, with interactions between groundwater and rock reflecting dissolution and mineral decomposition processes. Correlation matrices show that most water quality parameters exhibit statistically significant relationships, with TDS having a strong correlation with EC (r = 0.97) but a weaker correlation with Fe²⁺, Mn²⁺, CaCO₃, and SO₄²⁻. Fe²⁺ shows significant correlations with Mn²⁺ and TDS (r = 0.81 and 0.55), while the negative correlations between CaCO₃ and EC and TDS (r = −0.34 and −0.31) suggests that CaCO₃ is not solely derived from rock weathering but may also originate from anthropogenic sources. Alkalinity, typically represented by CaCO₃, is determined by the presence of hydroxide, carbonate, bicarbonate, and organic acid ions, which come from various natural processes.

The health risk assessment results reveal the presence of non-carcinogenic health risks due to high iron levels, particularly in underground rivers, with greater health threats compared to springs. Therefore, effective communication with communities is crucial, with measures such as regular tube well water quality testing, raising awareness about affordable and safe water provision, and developing cost-effective household water treatment systems. While sulfate levels were also analyzed their health impacts were found to be secondary compared to heavy metals (Fe and Mn), which constitute the main contributors to non-carcinogenic health risks in the study area. The novelty of this study resides in the fact that it uses GWQI, HHRA, and statistical analysis to conduct an integrated assessment of groundwater quality and health risks in the karst region of Trenggalek. This assessment reveals geochemical interactions and brings attention to health problems associated with iron in underground rivers.

Further research is needed to gain a deeper understanding of health risks associated with iron and manganese in drinking water sources to develop more effective water quality improvement and risk mitigation strategies. By implementing the recommended measures and continuing research efforts health risks can be minimized, ensuring the safety of water resources for communities in the study area.