Abstract
Karst aquifers are strategically important as they supply domestic water and are a resource for irrigation and industry in northern China. The heterogeneity of the karst aquifer medium makes it vulnerable to external influences. Here, samples of surface water and groundwater in a typical karst groundwater system of Jinan (China) were collected in December 2021, and the sample data were analyzed to further elucidate the hydrogeochemical processes. Results showed that the predominant water chemical type is HCO3-Ca, with a lesser proportion of type HCO3·SO4-Ca and SO4-Ca. The dominant water–rock interactions comprise the dissolution of carbonate minerals, gypsum, and halite and ion exchange. Dissolution/precipitation of calcite, dolomite, and gypsum determines the concentrations of Ca2+, Mg2+, HCO3-, and SO42-. In terms of spatial distribution, the indirect and direct recharge areas are dominated by calcite dissolution, followed by dolomite dissolution, and are prone to ion exchange. The hydrogeochemical formation mechanism of the discharge area is more complicated by other hydrochemical processes and anthropogenic activities. These results provide guidance for global karst groundwater resource management and pollution prevention.
HIGHLIGHTS
Study the scientific issues to be solved in the research of hydrogeological characteristics in the process of karst groundwater circulation in Jinan.
Enrich the technical connotation of karst groundwater resource management and pollution prevention.
INTRODUCTION
Karst groundwater is an important component of regional water sources and it has become the primary water resource for meeting human demand in many areas (Ford & Williams 1989; C. S. Li et al. 2018; Yang et al. 2019). Approximately 20%–25% of the world's population depends on karst groundwater for potable water (Ma et al. 2011; Ford & Williams 2007; Lin et al. 2019). It is therefore important to study the hydrogeochemical characteristics and spatial evolution of karst groundwater before appropriate environmental protection measures can be taken (Frank et al. 2018; Wu et al. 2020).
The particular hydrogeological characteristics of a karst system, together with the high heterogeneity, openness, and high vulnerability of the aquifer (Bakalowicz 2005; Ghezelayagh et al. 2021), make karst groundwater particularly sensitive to external environmental influences (Eftimi et al. 2017; Hao et al. 2021). Generally, the hydrogeochemical characteristics of karst groundwater are controlled by many processes such as climate change (Berner 1992), land use (Sullivan et al. 2019), aquifer minerals (Williams 1993), and anthropogenic activities (such as agricultural activities and domestic sewage etc.) (Moral et al. 2008; Nguyen et al. 2014). Therefore, the interaction between surface water and karst groundwater and the influence of hydrochemical formation in the process of groundwater flow need to be fully elucidated (Eftimi et al. 2017; Basack et al. 2022; Raji & Packialakshmi 2022). Recently, hydrochemical methods have been widely employed as an important means to investigate hydrogeochemical characteristics (Nguyen et al. 2014). In Guizhou Province, China, Hao et al. (2021) utilized hydrochemical methods to determine the seasonal variation of chemical properties and define the water quality. Martos-Rosillo & Moral (2015) revealed the hydrochemical changes due to the intensive use of karst groundwater in Seville, southern Spain, in order to achieve its sustainable use. However, further research is needed to better understand how the hydrochemical characteristics of groundwater near rivers are formed.
Jinan City, located in a typical karst area in the semiarid region of northern China, has karst groundwater as its essential source of water supply (Chen et al. 2021). Numerous previous studies on the karst groundwater resource of Jinan City have focused on the groundwater flow (Qian et al. 2006; Wang et al. 2016), hydrologic processes (Kang et al. 2011; Zhang et al. 2018), spring formation (Zhang et al. 2018; Luo et al. 2020), and hydrochemical characteristics (Wu et al. 2020). Over past decades, with the development of industry and agriculture, the continuous increase in the utilization of the karst groundwater resource has led to a decline in the groundwater level and deterioration of groundwater quality (Schöpke et al. 2017; C. S. Li et al. 2018; Luo et al. 2020; Zhang & Wang 2021).
The components of the karst groundwater system of Jinan comprise the indirect recharge area (IRA), direct recharge area (DRA), and discharge area (DA) (Chen et al. 2021). The hydrogeochemical characteristics of karst groundwater are controlled by the recharge, runoff, and discharge conditions (Wu et al. 2009). Earlier research on the Jinan spring area has focused mainly on the DA, which is always preferentially selected for groundwater development. However, along the general path of groundwater flow, the ion concentrations change due to water–rock interactions between the groundwater and the aquifer materials. It is important to identify differences in the hydrogeochemical characterization of karst groundwater between the recharge area and the DA; however, previous research has largely focused on the environmental quality of karst groundwater, while ignoring the spatial variation of karst groundwater (Wang et al. 2006).
For the hydrogeochemical processes, it is understandable that the spatial hydrogeochemical characteristics and evolution processes of the Jinan karst groundwater system could be very different from the spatial distribution. This motivates us to collect samples from IRA, DRA, and DA and conduct this investigation, and several methods including hydrochemical analysis, Piper diagram, Gibbs diagram, the ion ratio method, and statistical analysis were applied to answer the following three questions: (1) What are the hydrogeochemical characteristics of karst groundwater and how do they differ from those of surface water? (2) What water–rock interactions control the formation and evolution of karst groundwater? (3) What are the spatial differences in karst groundwater within the study area and how are they manifested? The objectives of this study were to identify the hydrogeochemical characteristics of karst groundwater and to provide a reference for both the management of karst groundwater and the exploration of effective methods for karst groundwater development and utilization.
MATERIALS AND METHODS
Study area
The study area is located in the Jinan karst aquifer system, Shandong Province, China (Figure S1, Supplementary Information). This region has a tropical semiarid continental monsoon climate, with an average annual precipitation of 698 mm, and an average annual evaporation of 1,476 mm, and an average annual temperature of 14.3 °C.
The study area is located on the northern wing of the western Luxi anticline, which is a gentle monoclinal structure. The terrain is high in the southeast and low in the northwest, changing from low hills, to remnant hills, and to piedmont plain landforms. The strata are inclined to the north and northwest, with an inclination of 5°–12°. From top to bottom, the main strata in this area are Quaternary, Tertiary, Carboniferous, Permian, Ordovician, Paleozoic Cambrian, and the Archaean Taishan Group of ancient metamorphic rock. In the northern part, there is an igneous rock mass intrusion of the Yanshanian period.
Karst groundwater generally flows from the south toward the north and is mainly controlled by topography and bounded by the watershed. The main aquifers of the sampling wells used in this study are the middle Cambrian Zhangxia Group, upper Cambrian Fengshan Group, and Ordovician aquifers. The aquifers consist of mainly carbonate rock fissure water, and the lithology is mainly limestone, dolomitic limestone, and argillaceous limestone. The karst groundwater system of Jinan is divided into IRA, DRA, and DA with different recharge sources and flow characteristics.
Sample collection and analysis
Piper diagram of all water samples.
Prior to sampling, all bottles were washed and rinsed thoroughly with water. The SW samples were collected directly. Most KG samples were obtained from wells after purging for at least 5 min; however, the KG-3 (artesian discharge well) and KG-14 (spring) samples were collected directly. All samples were filtered through 0.45 μm membrane filters immediately after collection and stored at 4 °C until laboratory analysis, which was conducted within a week of collection.
Field parameters, such as alkalinity, pH, total dissolved solids (TDS), electrical conductivity (EC), and dissolved oxygen (DO), were measured on-site using a multi-index portable meter (DZS-708; INSEA, China), which was calibrated using standard solutions prior to use. Major cations (i.e., 
, K+, 
, and 
) and anions (i.e., 
, 
, 
, and 
) were analyzed using ion chromatography (IC883; Metrohm, Switzerland), and 
was determined by titration. Total organic carbon (TOC) was determined using a Shimadzu TOC-VCPH analyzer. The water samples used for cationic analysis were acidified to pH < 2 with ultrapure nitric acid and stored in polyethylene bottles. All chemical analyses were performed at Shandong Provincial Key Laboratory of Water Resources and Environment.
Data processing and statistical analysis
The correlation analysis was conducted to identify the relationships among the various hydrogeochemical compositions. Multivariate principal component analysis (PCA) was performed using ORIGIN 2022 to determine the primary factors that influence the hydrogeochemical characteristics.
RESULTS AND DISCUSSION
Characteristics of hydrochemical parameters
The on-site indicators and hydrochemical parameters of the SW and KG samples were analyzed and the results are listed in Tables S1 and S2. The Piper diagram of all samples divided into six zones is shown in Figure 1 and most samples fall in Zone 2 and Zone 5. The hydrochemical type of all SW samples and some KG samples is 
, while the type of other KG samples is 
. The hydrochemical type of sample KG-3 is 
. With respect to cations, all samples fall in Zone B (Ca type), reflecting the prominence of carbonate weathering. With respect to anions, most samples fall in Zone F (HCO3) and Zone H (mixed type), although one sample falls in Zone E (SO4 type), representing the dominance of carbonate weathering and dissolution of gypsum. Generally, the hydrochemical type of KG shows consistency with that of SW, and this hydrochemical similarity suggests that they are related and probably have a close transformation relationship and hydraulic connection. According to previous research (Wu et al. 2020), the hydrochemical type of karst groundwater evolves from 
to 
. The main reason for the difference in the current study is that the sampling sites are mostly located along the Yufu River, which is affected substantially by surface water recharge.
Evidence of major hydrochemical formation
and 
as a function of TDS, which can be used to assess the natural sources of dissolved chemical constituents and investigate how chemical constituents form, including precipitation dominance, rock weathering dominance, and the evaporation–crystallization process (Gibbs 1971; Talib et al. 2019). Figure 2 shows that rock weathering contributes to the main process that controls the chemical composition of the groundwater. The 
and 
values are <0.5, indicating that the anions and ions are mainly composed of 
and 
(Huang et al. 2017). The main karst aquifers within the study area are Ordovician and Cambrian carbonates and therefore weathering dissolution of tuffs and dolomites occurs (Wang et al. 2016). There is no substantial change in TDS with an increase in 
, showing that ion exchange also affects the chemical compositions by increasing 
and decreasing 
(Su et al. 2023). 
Gibbs diagram: (a) 
(mg/L) versus log TDS (mg/L) and (b) 
(mg/L) versus log TDS (mg/L).
Gibbs diagram: (a) (mg/L) versus log TDS (mg/L) and (b) (mg/L) versus log TDS (mg/L).
Evolution process between hydrochemistry and lithology of karst rocks
Ion exchange
and 
(Han et al. 2013; Lin et al. 2019). As can be seen in Figure 3(a), the relationship between (
) and [
] is linear with a slope value close to −1, which suggests 
and 
exchange with 
adsorbed on the rock (X. X. Li et al. 2018). Almost all samples are located above the −1:1 line, indicating that 
and 
in groundwater might have been affected by the dissolution of albite and anorthite, in addition to ion exchange. The samples in the DA deviate far from the −1 line, suggesting that anthropogenic input or reverse ion exchange might have increased the contents of 
and 
. 
Plots of (a) 
versus 
, and (b) CAI-1 versus CAI-2 indices of water in the study area.
Plots of (a) versus , and (b) CAI-1 versus CAI-2 indices of water in the study area.
and decrease in 
(Equation (4)). However, the CAI of some KG samples in the DA are >0, signifying that reverse ion exchange is an important source of ions (Equation (5)). Ultimately, 
and 
in the aquifer are exchanged with 
and K+ in the groundwater (Yang et al. 2016; Wang et al. 2017).
Rock weathering
Ion ratios characterizing the hydrogeochemical reactions in different water.
The 
value of all SW and KG samples in the DRA and DA is close to 1 (Figure 4(c)), indicating that dissolved calcite is dominant in producing 
and 
. With increase in concentration, some KG samples in the IRA and DA fall below the 1:1 line, indicating that 
is released by reactions other than carbonate dissolution, e.g., gypsum. Meanwhile, the 
value of most samples falls between the 1:2 line and the 1:3 line (Figure 4(d)), indicating that dolomite minerals are the important source of 
and 
in DRA and DA, which aligns with previous studies (Su et al. 2023). The IRA-KG samples fall above the 1:3 line, indicating other sources of 
. The reasonable correlation between 
(Figure 4(e)) suggests that the dissolution of both carbonate and sulfate plays an important role (Wu et al. 2009). Samples in the IRA and DA are slightly above the 1:1 line, indicating that dissolution of carbonate is the main process of ion exchange (Dehnavi et al. 2011; Senthilkumar & Elango 2013), whereas samples in the DRA are slightly below the 1:1 line, which might be related to reverse ion exchange and sulfate reduction that causes the decrease in 
and 
(Zhang et al. 2020). In Figure 4(f), most samples are above the 1:1 line, indicating that the formation of the hydrochemical components of karst groundwater is attributable mainly to the dissolution of carbonate rocks, and partly in association with evaporate dissolution in some instances.
The effects of ‘non-gypsum source calcium’ and ‘non-carbonate source calcium’ on the hydrochemistry of karst water are shown in Figure 4(g) and 4(h) (Wang et al. 2006). In Figure 4(g), the KG samples are distributed mainly between the 1:1 line and the 1:2 line, whereas the SW samples are located mostly near the 1:2 line. The results show that the 
and 
of the KG samples are mainly from dissolved calcite, followed by dolomite, while the SW samples are mainly from dissolved calcite. The KG samples are on the right of the plot, indicating that dolomite dissolution is enhanced. As evident in Figure 4(h), most samples are scattered above the 1:1 line, indicating that 
and 
originate primarily from the dissolution of gypsum (Wu et al. 2014). With an increase in 
, the DA KG samples deviate further from the 1:1 line, indicating the presence of 
from non-hydropetrogenic sources in the karst water, which might be related to precipitation and anthropogenic activities (Wang et al. 2006).
and 
, especially for 
(Yang et al. 2016). The excess of 
over 
in the DA KG samples indicates additional sources of chloride. On the one hand, reverse ion exchange might have occurred to reduce 
, which would explain the slight excess of 
and 
over 
and 
(Bakalowicz 1994). On the other hand, the characteristics of groundwater recharge and discharge represent other important influences on the content of 
(Xanke et al. 2015). This further explains the deviation of 
in Figure 4(e) of the DA KG samples. Most samples in the IRA and DRA are above the 1:1 line, indicating that 
has sources other than halite mineral dissolution. Guo et al. (2020) found the same conclusion in the closed karst groundwater basin of Jinan. This result is consistent with the content of Figure 3, which indicates that ion exchange might have occurred in the IRA and DRA. 
Water sample saturation indices (SI) for (a) calcite, (b) dolomite, (c) gypsum, (d) halite, (e) SIdolomite versus SIcalcite, and (f) SI values in the IRA, DRA, and DA.
Water sample saturation indices (SI) for (a) calcite, (b) dolomite, (c) gypsum, (d) halite, (e) SIdolomite versus SIcalcite, and (f) SI values in the IRA, DRA, and DA.
Saturation indices
It can be seen from Figure 5(a) that most SIcalcite values are >0, and that the SIcalcite of the KG samples is greater than that of the SW samples, indicating that calcite is in a saturated state. The SIdolomite of the SW samples is <0, whereas the SIdolomite of most KG samples is >0, indicating that precipitation of dolomite occurs in groundwater (Figure 5(b)). Wang et al. (2006) reported that the dissolution rate of calcite is higher than that of dolomite, and that dolomite might still dissolve when calcite is oversaturated in KG, resulting in calcite precipitation. However, this is not consistent with the increase in 
(Figure 4(c)), which is due to gypsum in the aquifer, whereby the dissolution of gypsum leads to an increase in 
. When 
reaches a specified level, de-dolomitization will occur. The precipitation of calcite leads to a decrease in 
, such that the dolomite is unsaturated and 
increases (Appelo & Postma 2005).
SIgypsum is in the range of −1.97 to −0.03 (Figure 5(c)), indicating an unsaturated state of gypsum, whereby more gypsum dissolves in the water with groundwater flow. This is in marked contrast to the findings of other research in karst areas in China (Lin et al. 2019) and is related to dolomite dissolution and preferential exchange of 
by 
. SIhalite is also unsaturated, indicating the presence of halite dissolution (Figure 5(d)). SIgypsum and SIhalite increase with an increase in TDS, indicating that the 
, 
, 
, and 
are from the dissolution of gypsum and halite. There is a notable relationship between SIdolomite and SIcalcite (Figure 5(e)) that indicates that the composition of dolomite and decomposition of calcite are synchronous, which is in agreement with the results of analysis of the main ion sources (Wang et al. 2017).
From the IRA, DRA, and DA, there is no substantial change in SIcalcite. Dolomite is further dissolved and transitioned from the unsaturated state to the saturated state, and gypsum and halite are unsaturated. SIhalite in the DA is greater than that in the IRA, and SIgypsum shows a trend of decrease (Deng et al. 2023). This indicates that groundwater has a strong ability to dissolve gypsum in the DA (Han et al. 2015).
Source similarity analysis of hydrogeochemical formation
, 
, and 
all contribute substantially, together accounting for 42.9% of the total. It shows that the dilution effect of ions related to limestone and gypsum dissolution is caused by karst groundwater evolution, evidenced by the close negative correlation of PC1 with pH, DO, TOC, and DO. Closely related to PC2 are alkalinity, 
, 
, 
, and 
, suggesting that the dissolution of halite and anthropogenic effects have a substantial impact on the hydrogeochemical characteristics (Cao et al. 2016). 
PCA results showing relationships between water samples and hydrogeochemical parameters in terms of spatial variation.
PCA results showing relationships between water samples and hydrogeochemical parameters in terms of spatial variation.
Pearson correlation analysis represents the major contributions of ions to the groundwater totals and reveals the importance of certain effects (Figure S4). Alkalinity is correlated negatively with pH (r > 0.5) and correlated positively with 
(r > 0.9). There are strong correlations between TDS and 
, 
, 
, and 
, with the most significant correlations between TDS and 
and 
. Water–rock interactions control the composition of TDS, which is consistent with the results of the Gibbs analysis. There are moderate correlations between 
and 
and 
, because the main source of ions is related to the dissolution of dolomite, limestone, and gypsum. The results show that the weathering and dissolution of carbonate rock and sulfate minerals play important roles in controlling the chemical composition (Zhou et al. 2015). The aquifer in the Jinan spring area is mainly composed of limestone and dolomite, although some rock layers are mixed with gypsum. This is consistent with the ion ratio analysis results (Figure 5(c)). A strong correlation between 
and 
indicates that 
, derived mainly from pollutants, increases the ionic strength of the karst groundwater, leading to strong dissolution of halite (Wang et al. 2023). There is also a positive correlation between EC and 
, indicating that anthropogenic activities cause an increase in 
in groundwater. Moreover, 
has a significant correlation with 
(r = 0.78), which shows that 
pollution promotes the accumulation of 
.
CONCLUSIONS
Hydrochemistry provided important information about the hydrogeochemical characteristics of a karst groundwater system in Jinan and revealed the evolution processes in IRA, DRA, and DA in Jinan, Shandong Province. The major conclusions drawn are as follows:
- (1)
The hydrochemical type of SW and KG is predominantly
, but the type of some samples is 
and 
. The ion content is higher in the DRA and DA than in the IRA. This is mainly influenced by rock weathering and ion exchange in the aquifer, and especially by human activities such as agriculture in the DRA. - (2)
The process and extent of rock weathering cause differences in the IRA, DRA, and DA, mainly due to different flow paths and residence times, which indicate the extent of groundwater circulation.
and 
as the monitoring tracer provide additional insight into the evolution process. Groundwater in the DRA is readily saturated with calcite, but unsaturated with gypsum, and the dissolution of gypsum minerals leads to de-dolomitization. Groundwater in the IRA and DA is mainly affected by the decomposition of carbonate rock, especially in the IRA. - (3)
The groundwater in the DRA is recharged by the IRA and has experienced a longer path. The increase in salinity is mainly controlled by the dissolution of gypsum and halite minerals. The groundwater in the IRA and DRA is more vulnerable to pollution discharges. In addition to ion exchange and rock weathering, there are other hydrochemical processes that supply notable amounts of
and 
. The mechanisms affecting groundwater chemistry in the study area are still largely controlled by geogenic processes rather than by anthropogenic activities.
In addition, we emphasize the need for future studies to assess the differences in hydrogeochemical characteristics between different regions, and for targeted work to be undertaken to develop and protect groundwater resources.
AUTHORS CONTRIBUTIONS
D. L. and X. C. are responsible for the formulation and methodology of overarching research aims. C. T. analyzed the major hydrochemical formation. W. Z. and Z. W. analyzed the major ion sources and hydrogeochemical evolution. X. Z. and D. X. performed the principal component analysis. Y. C. performed the Pearson correlation analysis. All authors read and approved the final manuscript.
FUNDING
This research was funded by the Natural Science Foundation of Shandong Province (ZR2021QD031; ZR2021MD086), Water Conservancy Technology Demonstration Project (SF-202210), Optional Subjects of the Water Resources Research Institute of Shandong Province (SDSKYZX202121-1) and the Key Technology and Application Project of Flood Control and Waterlogging Control in Plain Waterlogging Depression (PCTXPQ-KY202001-2).
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.
%20cropped.png?versionId=7153)