The Characteristics and Application of Deuterium and Oxygen Isotopes to Karst Groundwater, Southwest China

Abstract

Stable isotopes in natural water bodies serve as hydrologic tracers, with a history of extensive application in karst groundwater research. The present paper is a systematic review of previous research results, with the objective of sorting out and summarizing the stable isotopic characteristics of karst groundwater in southwest China, which is the most typical karst region worldwide. Comprehensive analysis reveals that the deuterium and oxygen isotope values of most karst groundwater are characterized temporally as ‘enriched in the rainy season and depleted in the dry season’, which is the opposite of the pattern of precipitation. While the spatial distribution feature is basically consistent with the spatial variation pattern of precipitation, which is characterized as ‘depleted from the coast to the interior’. Additionally, the main applications of stable isotopes to karst groundwater are discussed, including karst groundwater recharge source and recharge elevation identification, research on karst groundwater in the hydrological cycle, the hydrological process of karst groundwater, and karst groundwater contamination tracking. Finally, we looked forward to future research on karst groundwater based on deuterium and oxygen isotopes. It is our hope that this review may provide insight into the study of karst groundwater.

1. Introduction

The development and application of stable hydrogen and oxygen isotopes have made significant contributions to the field of hydrology. The superiority of the application of stable isotopes in hydrological studies enhances accuracy [1], enables detailed pathway analysis [2,3,4], and facilitates non-destructive sampling [5], making it widely popular among scholars. Naturally, water molecules contain two stable isotopes of hydrogen, ¹H (protium) and ²H (deuterium, also marked as D), and three stable isotopes of oxygen, ¹⁶O, ¹⁷O, and ¹⁸O [6,7]. The process of evaporation and condensation of water is accompanied by isotopic fractionation [8]. And fractionation that occurs at all stages of the hydrological cycle determines the isotopic composition of water in atmospheric water vapor [9]. Furthermore, the evaporation of water exhibits variations in regions of different latitudes and altitudes [10]. Consequently, there are noticeable disparities in the hydrogen and oxygen isotope compositions of rainfall, surface water, and groundwater in different geographical areas [8,11]. These compositional differences are often metaphorically referred to as the ‘fingerprint’ of water. Scholars have conducted numerous empirical studies by harnessing these stable isotopes, including tracing the natural hydrological cycle [12,13,14], determining the features of stable isotopes of oxygen and hydrogen in precipitation and its water vapor source [15,16], and identifying the origin and dynamics of surface water (e.g., rivers, lakes) [17,18]. Furthermore, the analysis of the stable isotopes of groundwater has been applied to understand recharge sources [19,20,21], residence time [22,23], and the interactions between surface water [24,25]. Overall, stable hydrogen and oxygen isotope analysis has emerged as an essential tool for studying hydrological systems and advancing our knowledge in this important field. Scientists have gained valuable insights into the movement, origin, and transformation of water in different environments based on stable isotopes.

To date, the literature has summarized the isotopic theoretical basis and characteristics of water in the hydrological cycle [8]. Researchers have also studied the isotopes of the water cycle at spatial scales, from specific regions to the globe, focusing on the large-scale isotopic data to understand the water cycle in the atmosphere and the land-atmosphere systems, in karst critical zones and karst catchments systems and in human-managed water distribution systems [26]. In addition, the isotopic composition and characteristics of precipitation [27,28,29], to identify the origin and trajectory of air masses in the regional area, have been studied [10,30,31,32]. The isotopic research on water vapor, including the isotopic composition measurements, simulations, and interpretations, with the hydrological cycle, has also been reviewed [9]. Another review has provided an overview of the spatial–temporal characteristics of stable isotope variations in water bodies and summarized the environmental and eco-hydrological implications for the cold regions of western China [33]. What is more, the temporal and spatial distribution of isotopes in river water systems has also been investigated [34]. Last but not least, from the perspective of global isotope hydrogeology, a review has summarized and synthesized the isotopic techniques for the hydrogeological cycle of groundwater recharge, storage, and discharge. Specifically, isotope-based approaches to water management and hydrological sciences, the method of isotopic data interpretation, and the applications, limitations, and opportunities of isotopic techniques for groundwater hydrology are considered [35]. Despite the extensive literature, the summary and synthesis of stable isotopes in karst groundwater research is still deficient.

Karst groundwater is a significant freshwater resource in a karst area and has attracted much attention due to its occurrence in unique karst landforms. The hydrological system in a karst area consists of a surface system and a subsurface system [36], forming a spatial distribution pattern of ‘soil on the ground and water underground’ [37], which is different from non-karst areas. Groundwater in karst areas is mainly stored in the fissures, conduits, and caverns of carbonate rocks, and its occurrence state depends on the degree of karst development [38]. The spatial variation of aquifers is very large, which poses great difficulties for the investigation and utilization of karst water [39]. Furthermore, it is difficult to describe and quantify the flow characteristic with traditional methods due to the hydrological processes in karst areas are rather complex [40]. Therefore, the stable isotopes have been well developed and applied in the karst groundwater research.

China is one of the countries with the largest karst distribution over the globe, with a total area of 3.44 million km², accounting for one-third of the total area of China [41,42]. The southwest of China, with a total outcropping karst area of 540,000 km² including six provinces (Guizhou, Yunnan, Sichuan, Hubei, Hunan, Guangdong), one autonomous region (Guangxi) and one municipality (Chongqing), is the most representative karst region in China and even in the world [43,44,45]. The total area of the three major watersheds in southwest China (including the Yangtze River Basin, the Pearl River Basin, and the Red River Basin flowing into Southeast Asia) is 808,579.18 km², of which the allowable yield of karst groundwater is 61.5 billion m³ per year [46]. There is no doubt that karst groundwater has great value and potential to be exploited and utilized. Up to now, many achievements of hydrogen and oxygen isotope research of karst groundwater in southwest China have been investigated, with limited attention given to regional systematical summary. Thus, the motivation of this review paper is to collect the state of the art of the available stable isotope literature on karst groundwater in southwest China and to synthesize the specific application content of hydrogen and oxygen stable isotopes. The regions of southwest China referred to in this paper include Guizhou, Guangxi, Yunnan, Sichuan, and Chongqing, where karst development is concentrated and typical (Figure 1). In this paper, we provide a systematic review, including (i) the characteristics of stable isotopes of karst groundwater in southwest China, mainly focusing on the temporal and spatial characteristics; (ii) the stable isotopes’ application to karst groundwater in southwest China, which includes karst groundwater recharge source and recharge elevation identification, karst groundwater in the hydrological cycle, the hydrological process of karst groundwater, and karst groundwater contamination tracking; (iii) finally, the development prospects of the application of stable isotopes to karst groundwater are discussed in detail. It is expected that this review could provide a comprehensive and systematic summary of the karst groundwater resources in southwest China, serve as a foundation for the scientific development, usage, and protection of these resources in the area, as well as provide a reference for future hydrological research in the area.

2. The Characteristics of Stable Isotopes of Karst Groundwater

2.1. Seasonal Characteristics of Stable Isotopes of Karst Groundwater in Southwest China

The stable isotope values in karst groundwater exhibit a pronounced interannual variability in the southwestern regions, and the δD value fluctuates between −114.80‰ and −29.99‰, while the δ¹⁸O value fluctuates between −15.79‰ and −5.48‰ (Table S1, Figure 2). Yunnan has the biggest differences in the interannual fluctuation of stable isotope concentrations in karst groundwater; the δD value ranges from −114.80‰ to −68.33‰, and the δ¹⁸O value ranges from −15.79‰ to −10.14‰, according to data from the published paper (Table S1). Large variations in stable isotope concentrations in karst groundwater indicate that they are influenced by complex factors. In fact, atmospheric precipitation is the main recharge source of karst groundwater in the southwest region, where the interannual variation range of stable isotope concentrations in precipitation is also significant [47,48,49]. For instance, the δD and δ¹⁸O values of precipitation in Guiyang (the provincial capital of Guizhou) range from −153.6‰ to 37.9‰ and −19.4‰ to 2.9‰, respectively [50]. The stable isotope concentrations (δD and δ¹⁸O) of precipitation in Chongqing vary from −122‰ to 36‰ and −16.8‰ to 0.6‰, respectively [51]. In addition, the stable isotope compositions of precipitation in Chengdu (the provincial capital of Sichuan) [52], Guilin (typical karst region of Guangxi) [53], and Kunming (the provincial capital of Yunnan) [54] ranged from −119.38‰ to +22.10‰, −147.3‰ to +25.4‰, and −110.9 to −22.1 ‰ (δD) and from −15.84‰ to +2.68‰, −15.4‰ to +3.3, and −14.3 to −3.3 ‰ (δ¹⁸O), respectively. It can be seen that the range of fluctuations of isotope values in karst groundwater is basically within the range of fluctuations of isotope values in rainwater, and the interannual variability of karst groundwater isotopes is relatively small when compared to those of rainwater.

Furthermore, the regularity of stable isotope concentrations in karst groundwater in the southwest of China is evident. As shown in Figure 3, the groundwater samples were collected along the Lijiang River Basin [55] and from the Yaji karst experiment site [56], the spring of Guancun karst catchment [57], the Shuanghe Cave Basin [58], the Chenqi karst spring catchment [59], the Shanwan karst test site [60], the Huanglong spring catchment [61], as well as the Chongqing karst groundwater [62]; the δD and δ¹⁸O values showed significant temporal differences. During the rainy season, the isotopic composition (δD and δ¹⁸O) of the karst groundwater was more enriched, while the isotopic value was depleted during the dry season in comparison. Interestingly, this seasonal feature is opposite to the pattern of stable isotopes in precipitation in the southwestern karst region. The seasonal characteristics of stable isotopes in precipitation are characterized by depletion in the rainy season and enrichment in the dry season [53,63,64]. It is generally accepted that shifts in atmospheric circulation and differences in moisture vapor sources are the main causes of this seasonal variation in precipitation isotopes in the southwest karst region [65,66,67]. Specifically, the study area is located in the typical monsoon climate zone. Summer (the rainy season) is dominated by the Indian and western North Pacific monsoons, which are marked by high air humidity [68]. As water vapor moves inland from the oceans, there are multiple precipitation events. This results in a ‘rain-out effect’, where heavy isotopes fall preferentially in the early phases, while the remaining air mass contains relatively high concentrations of light isotopes [69]. In winter (dry season), the continental air mass predominates, with lower air humidity and less precipitation. The accumulation of rainfall over water vapor transport pathways is the dominant factor influencing the isotopic composition of water vapor [70]. Thus, compared to oceanic water vapor from tropical oceans, continental water vapor has a greater isotope value [71]. For karst groundwater, due to the large amount of rainfall during the rainy season, atmospheric precipitation can quickly recharge groundwater through karst morphology, the water-rock interaction time is short, and the effects of evaporation and fractionation are small [62,72,73,74]. Thus, the δ D and δ¹⁸O values are more easily enriched in the rainy season than in the dry season. Secondly, due to the ‘piston effect’ generated by the infiltration and replenishment process of precipitation, the ‘old water’ stored in karst cracks and pipelines is an additional source of supply during the rainy season [61,75]. Binet et al. [76] noted that the groundwater evolves from recently recharged water to old water coming from the local recharge area during post-flood drainage. Previous studies have also indicated that δ¹⁸O in groundwater exhibits a close positive correlation with residence time, leading to the conclusion that ‘old water’ is enriched in heavy isotopes [77,78]. However, the seasonal variation of deuterium and oxygen isotope values of karst groundwater in areas such as the Huanjiang karst spring [79], the Banzhai karst catchments, the Dengzhanhe karst spring catchments [59], and Black Dragon Pool Park (including Xiaoshuitan, Qingshuitan and Hunshuitan) [80] is not significant (Figure 3). This is attributed to the buffering and mixing capacity of aquifers in karst catchments [81,82]. For example, a study conducted by Perrin et al. [83] suggested that the soil and epikarst sub-systems in the study area acted as the important groundwater storage elements of the karst system and had a strong buffering on the rainfall isotope signals. In addition, Hu et al. [84] found that epikarst and soils controlled the rainfall recharge dynamics and had a small response to rainfall events. The epikarst had a strong water-holding capacity, which resulted in the seasonal variation in the deuterium content of the epikarst spring being less evident than in rainfall in a small karst catchment. Similar results were observed at the other karst catchment [85,86]. Additionally, it is worth noting that the δD and δ¹⁸O values of the karst groundwater in the Caohai Wetland catchment [87] are heavier during the dry season (low-flow season, April) (Figure 3). This phenomenon is also found in the Huixian karst wetland [88], which is the opposite of the seasonal characteristics of most karst regions and is attributed to the fact that the evaporation was more intense during the dry season in the karst wetland compared to the rainy season, resulting in more heavy isotope enrichment in the surface water [89]. The surface water is the main source for recharging karst groundwater during the dry season in wetland catchment.

In summary, a comprehensive analysis of previous research reveals that there is significant heterogeneity in the temporal distribution of stable isotopes. Apparent seasonal variations in the isotopic composition of karst groundwater are characteristic of most karst regions and are usually higher in the rainy season than in the dry season. The primary source of recharging the karst groundwater, atmospheric precipitation, has a major influence on the stable isotopic composition of karst groundwater. It is also influenced by a combination of factors such as the karst landscape itself and evaporation from the water body.

2.2. Spatial Characteristics of Stable Isotopes of Karst Groundwater in Southwest China

Stable isotope values in karst groundwater in the southwestern regions generally decrease from east to west (Figure 4). This spatial distribution pattern is basically consistent with the spatial variation pattern of precipitation. Studies have shown that the isotopic composition of precipitation in the southeastern region of China is spatially depleted from the coast to the interior as a result of the movement path of the air mass. Apart from the meteorological factors, the degree of heavy isotope depletion of precipitation is related to geographical parameters (such as altitude) [30]. Previous studies have shown that topographic factors such as altitude are thought to be closely related to the isotopic composition of precipitation [90,91]; i.e., isotopes become increasingly depleted with increasing altitude [92,93]. The phenomenon is attributed to the gradual decrease in temperature and condensation precipitation that occurs as altitude increases during the transport of water vapor. This leads to a depletion of isotope values in the remaining water vapor [94,95]. Additionally, Jasechko [35], who sorted out the correlations between meteoric water δ¹⁸O and sampling elevation, found precipitation δ¹⁸O values to decrease with elevation in most places. As seen in Figure 4, this is evident in Yunnan Province, where the isotopic composition of karst groundwater is more depleted than in Guangxi to the east. Yunnan is located on the Yunnan-Guizhou Plateau, which borders the Tibetan Plateau region to the northwest, and its altitude is the highest of the other four provinces. This suggests that the relationship between karst groundwater and elevation is consistent with the relationship between precipitation and elevation. Karst groundwater is recharged by precipitation from high elevations. Its isotopic composition may be lower. Overall, this suggests that the effect of topography on stable isotopes in precipitation also indirectly affects the stable isotope composition in karst groundwater. In general, precipitation is still the dominant factor influencing the spatial distribution of deuterium and oxygen isotopes in karst groundwater in the study area.

3. Application of Stable Isotopes to Karst Groundwater

3.1. Karst Groundwater Recharge Source and Recharge Elevation Identification

3.1.1. Recharge Source of Karst Groundwater

Stable isotope tracers have long been applied to identify the recharge sources of karst groundwater. They are mainly determined by interpreting the hydrogen and oxygen isotope values of groundwater and its possible recharge sources, such as atmospheric rainfall, ice or snowmelt, and surface water [96,97,98]. It is generally recognized that rainwater undergoes isotopic fractionation as it falls to the ground and infiltrates into the subsurface, often changing the concentration ratio of water isotopes. Differences between the δD or δ¹⁸O values of groundwater and rainfall can determine the relationship between them and help to obtain information on the sources of groundwater recharge. Specifically, the isotopic values (δD and δ¹⁸O) of local rainwater samples can be plotted on a line, which is usually named the Local Meteoric Water Line (LMWL) [99,100,101]. It is assumed that the isotopic composition (δD and δ¹⁸O) of karst groundwater essentially scatters along the LMWL, which confirms that the collected groundwater samples mainly originate from precipitation infiltration [102]. If the composition of hydrogen and oxygen isotopes in karst groundwater deviates to some extent from the LMWL, it indicates that evaporation has occurred during the process of atmospheric precipitation supplying karst groundwater. The degree of evaporation is usually represented by the deuterium excess values (d-excess), which are defined as d = δD − 8 δ¹⁸O [29]. The evaporation degree is inversely proportional to the d value, with higher evaporation degrees corresponding to lower d values. Conversely, after a raindrop is condensed, the stronger the re-evaporative fractionation under the cloud, the smaller the d value [103].

In fact, precipitation is the main source of recharge for karst groundwater in the southwest area. Some works have identified groundwater δD and δ¹⁸O values that are similar to those that characterize local rainfall. In Bijie City, Guizhou Province, for instance, Yuan et al. [104] found that the δD and δ¹⁸O values of 20 groups of karst groundwater samples were close to the LMWL expression, δD = 8.59 δ¹⁸O + 17.7 [105], indicating that the groundwater in the study area is mainly recharged by rainfall. Zhang [106] collected 108 groups of groundwater samples in eastern Yunnan and western Guizhou in the summer, and the relationship of the δD and δ¹⁸O values of the karst groundwater was defined by the equation δD = 8.732 δ¹⁸O + 16.85, which shows a slight difference to the LMWL: δD = 7.848 δ¹⁸O + 11. In a study on Chongqing karst groundwater, Pu [62] analyzed the δD and δ¹⁸O values of all underground river samples and found they are distributed along the meteoric water line, which indicates that precipitation is the main recharge source for karst groundwater. Furthermore, Pu [62] found that the slope and intercept of the groundwater linear equation formed by δD and δ¹⁸O values in the rainy season are close to those of the GMWL (Global Meteoric Water Line [107]) and CLMWL (Chongqing Local Meteoric Water Line); compared to the dry season, the d-excess value in the dry season is significantly lower than that in the rainy season. The main reason for this difference is related to the residence time of karst groundwater in the aquifer. The longer the residence time, the more the d-excess value decreases. It is also related to the hydrogeological characteristics of karst; that is, karst aquifers have a strong hydraulic connection, and karst fissures and pores are developed; the precipitation in the rainy season directly feeds underground rivers through sinkholes, shafts, dolines, etc. and basically has no obvious evaporation or only slight evaporation [62]. A study of karst groundwater in Xide country, Sichuan Province, conducted by Yuan et al. [108] indicated that almost all groundwater was distributed around the LMWL (δD = 7.54 δ¹⁸O + 4.84). There was no obvious isotopic shift in the isotopic values for karst groundwater, indicating that all water samples were of meteoric origin. Furthermore, Wang et al. [79] obtained the same conclusion in a study of a typical cockpit karst catchment in Huanjiang County, Guangxi Province. The results showed that the isotopic composition of the spring was distributed along the meteoric water line LMWL (δD = 8.1 × δ¹⁸O + 13.9, R² = 0.94), which revealed that rainwater was the only source of karst groundwater. In addition, snowfall during the winter and spring seasons in the high-altitude southwest karst areas also serves as an indirect source of groundwater recharge after melting in summer [109,110]. It is not difficult to distinguish the sources of recharge because the isotope values of ice and snow meltwater are less deleted in comparison to rainwater [111]. In conclusion, there are plenty of documents indicating that the recharge sources of the karst groundwater in the southwest area are related to precipitation, which was influenced by the environmental background of the regional climate and local geographic factors [32,112].

3.1.2. Recharge Elevation of Karst Groundwater

Stable isotope values in precipitation not only have the ‘seasonal effect’ mentioned above but also show an obvious ‘altitude effect’ [113,114], which concisely describes the shift in the isotopic composition of the precipitation with the change in the elevation of the terrain; with the increasing elevation, the δD and δ¹⁸O values are gradually depleted [115,116]. Knowing the groundwater stable isotopic values (δD and δ¹⁸O) and the precipitation stable isotope elevation gradient, it is possible to figure out the elevation of the precipitation that recharged the aquifer [117,118,119]. Several researchers previously identified recharge elevations of karst groundwater in southwest areas based on this pattern. For instance, Liu et al. [120] investigated the recharge sources of karst springs in the triangle area between the north bank of the Beipanjiang River and the west bank of the Dabang River in Guanling County, Guizhou Province. Based on the δD value of spring water samples and the correlations between isotopic values (δD and δ¹⁸O) in rainwater with altitude, the average elevation of the supply area is calculated. Additionally, they analyzed the results of hydrochemical and isotopic values of karst spring water, and the recharge strata were preliminarily obtained. Moreover, studies have investigated the height variation trend of isotopic values in karst groundwater [121,122,123]. Given that the isotope values of karst groundwater at the same altitude in different regions may be influenced by a variety of sources of recharge, it is possible to eliminate the influence of the local environment on groundwater isotope values by taking the weighted average of the isotope values of groundwater samples at each certain altitude range [74,124]. In the karst area of the East Yunnan-West Guizhou region, for example, a linear regression between the δD and δ¹⁸O values and the elevation values of the sampling sites revealed a high correlation, as δ¹⁸O (‰) = −0.00259 H − 5.657 (r = 0.749), and δD (‰) = −0.0236 H − 31.08 (r = 0.765). Specifically, the ¹⁸O value decreased by −0.259‰ and the δD value decreased by −2.36‰ for every 100 m increase in elevation [106]. In addition, a binary regression model was constructed to analyze the relationship between δ¹⁸O values and the elevation of underground rivers in the Chongqing area [62]. The results of the regression analysis indicated that the variation rate of δ¹⁸O in Chongqing groundwater with elevation was −0.34‰ per 100 m and −0.31‰ per 100 m in the dry and rainy seasons, respectively. This is due to the water bodies in both seasons being replenished by atmospheric precipitation, which results in a negligible difference in the height gradient between the dry and wet seasons [62]. However, the isotopic values (δD and δ¹⁸O) of karst groundwater can infer only the altitude of the precipitation entering the aquifer and not necessarily the site of recharge [125]. Therefore, additional geological, meteorological, and hydrologic data are required to identify the location of recharge [126].

3.2. Karst Groundwater in Hydrological Cycle

On a global scale, the components of the water cycle include water vapor, precipitation, surface water, groundwater, tap water, soil water, plant water, and ocean water [26]. The basin or regional water cycle focuses on precipitation, evapotranspiration, runoff, and the mutual transformation between precipitation, surface water, and groundwater [127,128]. Compared with other areas, the high permeability of the structure of karst watersheds makes subsurface runoff more abundant than surface runoff [129]. Most of the rainfall in the karst area is transformed into groundwater, some returns to the atmosphere in the form of evaporation and transpiration through the soil and vegetation, and a small proportion forms surface runoff, so karst areas often lack surface water systems [130]. Meanwhile, due to the presence of special geological structures in karst areas, karst groundwater, and surface water are hydraulically connected through numerous karst features (such as sinkholes, conduits, and shafts) that facilitate the exchange of water between the surface and the subsurface [131,132,133].

With the continuous movement of water in the hydrologic cycle, the water molecule exhibits a unique signature identified by the ratio of light to heavy stable isotopes of oxygen (¹⁸O/¹⁶O) and hydrogen (²H/¹H) [11]. Consequently, different water bodies will have different deuterium and oxygen isotope values. The compositional differences of isotope values are useful for understanding the processes of the water cycle [134]. The application of stable isotope ratios of hydrogen and oxygen to karst water cycle research is extremely popular. As an important component of the water cycle in karst areas, the interaction between karst groundwater and surface water has received a lot of attention [135,136,137]. Generally speaking, there is a difference in the isotopic composition of surface water and groundwater, and the recharge relationship between groundwater and surface water can be determined based on this difference [11]; i.e., when the δD and δ¹⁸O values show that groundwater is greater than surface water, it indicates that surface water is predominantly recharged groundwater, and conversely, groundwater is mainly recharged surface water [58]. For instance, in the case of the Lijiang River Basin in Guilin, groundwater and surface water were found to have similar trends with respect to δD and δ¹⁸O. This is due to the spatial and temporal connection between surface water and groundwater, and the frequent exchange of water bodies reduces the difference between them, with surface water mainly recharging groundwater during the wet period and groundwater recharging surface water during the dry period [55]. Similarly, Pu et al. [109] investigated the characteristics of surface water and karst groundwater in the Lijiang Basin and Yulong Mountain region and found that surface water and groundwater samples in the region have similar characteristics in terms of δ¹⁸O values, indicating that surface water and karst groundwater have the same recharge source and a close hydraulic connection. Surface water generated by precipitation and ice and snow meltwater from Yulong Mountain rapidly recharges groundwater aquifers, and the residence time of river water is short. The small difference between the δ¹⁸O values of surface water and groundwater indicates the frequent transformation between them. In general, the conversion relationship between surface water and groundwater can be determined by the difference in stable isotopes. The degree of karst development is also an important factor. In karst areas, karst conduits and fissures are developed; surface water could infiltrate into the ground quickly, and the hydraulic residence time is short, so the difference in stable isotopes between the two is minor, but on the contrary, the difference is obvious. This approach has been applied in many studies because it allows the acquisition of basic information on the interconversion of surface water and groundwater [138,139,140,141,142]. However, the aforementioned studies have only conducted simple qualitative analyses of the relationship between karst groundwater and surface water based on the characteristics of deuterium and oxygen isotopes. Furthermore, on the basis of the mass conservation principle of isotopes, a mass balance model can be established using stable isotopes of hydrogen and oxygen to quantify the specific transformations between surface water and groundwater [76,143,144]. In an important karst catchment in southeastern Jiuzhaigou, Sichuan Province, Liao et al. [145] investigated groundwater–surface water interactions. Using a coupled end-member mixing model, they found that the average contribution of groundwater to the surface runoff in the catchments was 36%. In addition, Sun et al. [146] used a model based on mixing water from a solute mass balance and evaluated the water fluxes contributions of three end-members in the main channel mixing water of Guizhou Province. The results showed that the percentage contributions of the AMD-impacted, non-AMD-impacted, and spring waters were 34.6%, 10.6%, and 54.8%, respectively. Similarly, Liu et al. [147] adopted the d-excess and TDS value as tracers for the end-member mixing analysis to determine the three sources mixed in the cistern recharge process: slope runoff (SR, SSR), rainwater (RW), and epikarst runoff (ER). The mixing ratios of the three end-members in the rainy season show that the SR + SSR was the predominant recharge type, accounting for 64%, while the ER accounted for the smallest proportion, but some cisterns recharged by epikarst springs can accumulate spring water in the absence of rainfall.

Moreover, the water cycle in the epikarst zone has attracted considerable attention. Specifically, the epikarst zone is characterized by strong dissolution, which results in the development of karst fissures and pores. The groundwater stored in these water-bearing media, including fissures, pores, channels, and weathering joints in the strongly weathered zone, is a significant factor in the water cycle [148,149,150]. This plays a pivotal role in water circulation and possesses a powerful storage capacity [84,151,152]. The degree of development of the epikarst zone is also closely related to the coverage of surface soil and vegetation [153,154]. Essentially, the karst area is characterized by a high degree of erosion, with a limited number of insoluble components and a significant proportion of soluble components being transported by water flow during the erosion process [155], so the soil layer in the karst area is thin and low in fertility, especially in the exposed karst area, where there is virtually no soil cover, which has a great impact on the vegetation growth in the karst area [156,157,158]. Consequently, in addition to soil water, epikarst-zone water represents a stable water source for karst plants. A considerable number of studies have employed stable isotopes to trace the ‘water use strategy’ of karst plants in the southwest karst area. For example, research conducted at Libo Karst Forest in Guizhou revealed that epikarst-zone water serves as a stable water source for karst plants, both during the rainy and dry seasons. This was determined by analyzing the δD and δ¹⁸O values of plant stem water and water extracted from soil layers, as well as the subcutaneous zone [159]. Additionally, the researchers observed that different species employ different water-use strategies. Furthermore, by measuring the seasonal change in isotope ratios (δD and δ¹⁸O) in twig sap, soil water, rainfall, and spring water, Deng et al. [160] conducted an investigation into the water sources utilized by karst plants at the Nongla karst dynamic monitoring station. Their findings indicated that adult trees exhibited a tendency to utilize deeper water stored within the epikarst zone, while young trees were observed to extract soil water when precipitation decreased. Additionally, the study revealed that all sampled trees primarily depended on water stored within the epikarst zone during the dry season. The paper also demonstrated that trees growing on exposed carbonates were mainly reliant on water in the epikarst zone. In a study conducted in the Yaji karst dynamics experimental station, Deng et al. [161] employed the dual stable isotopes of δD and δ¹⁸O to quantify the plant water uptake patterns and plant responses to different groundwater depths at four distinct sites. This was completed to estimate the soil and rock water contributions to the plant xylem water. It was observed that two species mainly absorb soil water during the rainy season and rock water during the dry season to maintain their growth. The seasonal fluctuations in water uptake observed in these habitats in the karst area are a consequence of the limited availability of surface water. Similar instances of such seasonal water uptake patterns have been reported in other karst regions [162,163,164]. In short, previous research indicated that deuterium and oxygen isotope tracing state clearly that epikarst zone water is an essential component of groundwater in karst areas, and it also plays an important role in the growth of karst plants. Unfortunately, the unreasonable use of epikarst-zone water may result in more severe environmental problems, such as rocky desertification in karst areas [165]. When rocky desertification occurs in karst areas, it becomes challenging for plants to use epikarst water, which in turn intensifies the process of rocky desertification, creating a vicious circle. Consequently, the utilization of deuterium and oxygen stable isotopes to elucidate the intrinsic interconnectivity between groundwater, soil water, and plant water in karst areas has a positive impact on the investigation of ecological aspects of karst areas.

3.3. Hydrological Process of Karst Groundwater

As a hydrological system, the karst aquifer structure of karst catchments exhibits a high degree of complexity and heterogeneity, which is characterized by the multi-media nature of karst aquifers, including fissures, pores, and pipelines. This feature results in marked variability of hydrological processes in both spatial and temporal domains. Generally, the hydrological behavior of a karst system can be described as temporally and spatially highly variable, with regard to processes involving recharging (diffuse and concentrated), storage (in epikarst, vadose, and phreatic zones), and flow type (diffuse and along preferential conduits) [22]. Hydrometric observations and tracer investigations are traditional and effective approaches to inferring hydrological processes within karst systems [166]. The application of stable isotopes can provide powerful tools for the study of the dynamic hydrological processes of karst groundwater [167,168]. Some research has employed high-resolution isotope hydrological monitoring of karst caves to obtain details of the runoff characteristics of epikarst flows and to gain deep insight into the recharge mechanisms in the epikarst zone [169]. Other works have identified water δD and δ¹⁸O values to track mixing processes and mixing characteristics of ‘new’ event water from rainwater and ‘old’ pre-event water from low-permeability aquifers [82]. Additionally, the contributions of deep warm groundwater to mixing in a carbonate-hosted ore deposit have been investigated [170]. Moreover, the hydrogen and oxygen isotope parameters have also been employed to assist in the construction of hydrological models, which depict complex hydrological phenomena and processes of groundwater in different landscape units within the karst catchment. For example, Zeng et al. [97] selected the Jade Dragon Snow Mountain (JDSM) as the subject of their study, employing hydrochemical and stable isotopic measurements to calculate the proportion of glacier meltwater that penetrates the aquifer of the study area. Subsequently, a conceptual hydrogeological model of the JDSM alpine karst aquifer of Baishui Spring was proposed, which represents the glacier-influenced and non-influenced components of the aquifer supplying Baishui Spring. Zhang et al. [171] developed a tracer-aided model that can estimate storage and water age dynamics, and it has been successfully applied in the Chenqi catchment in Guizhou Province. In general, the conceptual model can reflect the overall input–output relationship of hydrology in the karst area. However, it is not sufficient to portray the actual physical processes and spatial differences of the hydrological system. Another study was conducted in the same study area based on three years of daily measured stable isotopes (δD and δ18O) in precipitation, catchment outlet, and hillslope spring flows to calibrate the model parameters. The objective was to describe the flow path and temporal variabilities in the travel time distribution of the complex karst systems [172]. Later, Li et al. [173] constructed a coupled flow-tracer mode to reflect the hydrological connection between hillslopes and depressions in the Chenqi watershed. This model is capable of more accurately describing the hydrological structure and flow paths of hillslopes and depressions in karst watersheds, thereby improving the accuracy of the simulation of rainfall–runoff processes and model prediction accuracy in the watersheds. In brief, a hydrological model coupled with deuterium and oxygen isotope tracing has been developed rapidly, which provides a new path for the study of hydrological models in karst areas. It is also noteworthy that the process-based hydrological models are mainly classified into fully distributed, semi-distributed, and lumped hydrological models [174]. However, the literature on coupled distributed models and stable isotopes remains relatively scarce. Currently, the application of hydrological models to karst groundwater is still in its infancy, and the highly heterogeneous spatial structure of karst areas presents significant challenges in establishing accurate physical models in karst areas. Therefore, the application of a coupling deuterium and oxygen stable isotope approach to construct a hydrological model for a complex hydrological system in a karst area is still in need of further investigation.

3.4. Karst Groundwater Contamination Tracking

The dual isotope method (δ¹⁵N-NO₃⁻ and δ¹⁸O-NO₃⁻ or δ³⁴S-SO₄²⁻ and δ¹⁸O-SO₄²⁻) has been extensively applied for the investigation of the sources, migration, and transformation process of NO₃⁻and SO₄²⁻ pollution [175].

On the one hand, NO₃⁻ is a common environmental pollutant that not only exists in nature but is also influenced by human activities. The primary sources of nitrate are typically associated with the excessive utilization of chemical fertilizers, the discharge of domestic sewage, the accumulation and stacking of feces, and the release of organic nitrogen from soil [176]. In karst areas, where agricultural activities are concentrated, studies of NO₃⁻ pollution source identification are conducted by analyzing the end-member isotopic compositions and clarifying the distribution of NO₃⁻ contaminated characteristics in karst groundwater [177,178]. The common feature in these studies is that the sources of NO₃⁻ pollution are mainly related to agricultural fertilizers [179,180,181]. In karst scenic areas, for example, septic tank effluent and sewage contribute NO₃⁻ to the karst groundwater, which results in the enrichment of the δ¹⁵N-NO₃⁻ and δ¹⁸O-NO₃⁻ isotopes in the karst groundwater, which also poses a threat to the water quality of karst groundwater systems [182]. The mass balance calculation [183] and the statistical model (e.g., the Bayesian isotope mixing model [184,185]) are also useful for specifying the proportional contribution of each end-member. This is based on the truth that NO₃⁻ from different sources of pollution have different isotopic values. For example, NO₃⁻ derived from fertilizers exhibits a range of δ¹⁵N values between −3‰ and 3‰ and δ¹⁸O values between 17‰ and 25‰ [186]. δ¹⁵N values of NO₃⁻ originating from manure and sewage are between +5‰ and +25‰ and between +4‰ and +19‰ [175], respectively, and δ¹⁵O values of NO₃⁻ originating from manure and sewage are between 0‰ and +15‰ [187]. Nitrate in natural soil organic matter has values of δ¹⁵N between −3‰ and +5‰ [188] and δ¹⁸O values between −15‰ and 15‰ [186]. The isotopic compositions of atmospheric nitrogen deposition are situated between −13‰ and +13‰ for δ¹⁵N-NO₃⁻ [175] and range from +25‰ to +70‰ for δ¹⁸O-NO₃⁻ [187]. In addition, the influence of hydrodynamic changes [88,179,189] and microbial processes [181,190] has a significant impact on the migration and transformation process of NO₃⁻ pollutants.

On the other hand, in addition to agricultural activity, another important factor influencing karst groundwater quality in the southwest is mining activity. Acid mine drainage (AMD) water is formed by the oxidation of sulfide minerals in waste-rock storage facilities, tailings storage facilities, and abandoned or active mine structures [191]. AMD water is typically distinguished by a low pH and a high SO₄²⁻ concentration and high concentrations of hazardous trace metals [192,193,194]. The contamination of karst aquifers by AMD waters is becoming increasingly prevalent. Sun et al. [195] demonstrated that a coalfield in Guizhou Province has a significant impact on the karst spring water. The compositions of δ³⁴S-SO₄²⁻, δ¹⁸O -SO₄²⁻, and δ¹⁸O-H₂O, in conjunction with a three-end-member mixing model, indicated that the spring water was primarily influenced by AMD water, river water, and carbonate rock dissolution. Although the AMD water was characterized by the lowest mixing percentages of spring water during the wet (14.1%) and dry (26.9%) seasons, it was a critical forcing factor for spring water quality. In addition, the Babu karst aquifer, which is situated downstream of a tailings dump, has been contaminated by AMD. The contribution of AMD to the dissolved sulfates at the mouth of the BSS has increased significantly over time, from 39% in November 2014 to 93% in November 2018. Ren et al. [196] have observed that the δ³⁴S-SO₄²⁻ value has decreased (from 3.4‰ in 2014 to −13.2‰ in 2018) as a result of the severe impact that AMD waters have had on the underground water. A study by Song et al. [197] revealed that between 34% and 70% of mine water infiltrated directly through karst fissures and karst pipes and could not be collected at the mine entrance. This was determined through an analysis of δ¹⁸O-H₂O and δ³⁴S-SO₄²⁻ concentrations and a water balance assessment and implies that AMD infiltrating directly through runoff can easily be overlooked due to the concealed migration pathway.

The availability of groundwater is a vital resource for human survival. Nevertheless, the extraction of groundwater for human consumption and the discharge of wastewater from agricultural, domestic activities, and various industrial processes can have a deleterious impact on the water environment. The karst region is particularly susceptible to environmental contamination due to its highly interconnected hydrological structure, which renders it more vulnerable to contamination than other regions [198,199]. Pollutants in karst groundwater exhibit mobility and diffusivity, undergoing chemical, physical, and biological interactions with the surrounding environment. Such pollutants are often discharged from subterranean river outlets or in the form of springs. Therefore, it is of great importance to identify the sources, sinks, and pollution migration pathways of karst groundwater pollution, which can provide key information for source control and pollution pathway-blocking measures.

4. Conclusions and Outlook

The analysis of stable isotopes in karst groundwater in southwest China revealed that the interannual variability of stable isotope values fluctuates considerably. The majority of karst areas exhibit seasonal variations, which are characterized as ‘enriched in the rainy season and depleted in the dry season’. This is in contrast with precipitation isotopes in some way. Spatially, isotopic variation generally decreases from east to west. In essence, atmospheric precipitation, as the main recharge source of karst groundwater, strongly influences the isotopic composition of karst groundwater, along with the geographical location, hydrogeological conditions, and altitude. What is more, the information revealed by stable isotopic values allows for the acquisition of additional details regarding karst groundwater. These include the identification of the recharge source and elevation of karst groundwater, the delineation of karst groundwater within the hydrological cycle, and the tracking of the hydrological process of karst groundwater. Furthermore, the application of stable isotopic values allows for the identification of contamination within karst groundwater. In summary, the application of deuterium and oxygen isotopes in karst groundwater is highly extensive and has acquired significant achievements.

However, there are still some issues that need to be explored continuously, and the following aspects can be studied in the future:

(1) The further use of deuterium and oxygen isotopes should be used to study the water cycle and water balance processes in karst groundwater. A considerable number of studies have recently concentrated on small karst watersheds or specific segments of the karst water cycle. It would be highly beneficial to expand the scope of the research to encompass a larger watershed. These studies have the potential both to contribute to our understanding of the important effects of the interdependence of karst groundwater (especially epikarst water), soil water, and plant water on the ecology of karst areas and to provide scientific advice on the sustainable use and management of valuable water resources by understanding and quantifying the processes of exchange between karst groundwater and other water bodies.

(2) The further exploration of deuterium and oxygen isotopes should be used for the study of hydrogeological processes between karst groundwater and rocks. The process of the karst groundwater–rock interaction is accompanied by energy and material exchanges, which have a significant impact on groundwater quality and flow characteristics. Heavy water has different chemical properties and geochemical behavior in comparison to light water. Therefore, by using high-resolution deuterium and oxygen isotope monitoring data, it is possible to study the relative content and distribution between karst groundwater and rock, thereby revealing the process of interaction between groundwater and subsurface media.

(3) The integration of modeling techniques and data should be explored. Hydrological models are of great importance for understanding the prevention and control of droughts and floods, the management and exploitation of water resources, and other related fields. However, the high heterogeneity of karst areas makes the hydrological model difficult to use widely, mostly because catchment models cannot be replicated and applied to another karst watershed. To more effectively manage and utilize the water resources in karst areas, the construction of a more accurate and reliable karst groundwater model through the integrated use of deuterium and oxygen isotope measurements, numerical simulation techniques, and field monitoring data should be a developing trend in future research.

(4) The multi-parameter joint study of deuterium and oxygen isotopes with other environmental indicators should be explored. One way to uncover biogeochemical processes in karst groundwater is to combine variations in the ratio of deuterium and oxygen isotopes with other isotopes (such as hydrogen, carbon, nitrogen, etc.). In addition, by integrating deuterium and oxygen isotopes with environmental monitoring data (e.g., organic matter, heavy metals), a thorough analysis of the origin, movement, and elimination of contaminants in karst groundwater can be conducted. Furthermore, combining these isotopes with meteorological data (e.g., temperature, precipitation) helps uncover climate change impacts on karst groundwater systems. Overall, the utilization of deuterium and oxygen isotopes and other multi-parameter environmental indicators allows a comprehensive understanding of the energy and material exchange process in karst groundwater. This will also strengthen the scientific guidance and support for groundwater resource management and environmental protection.