Insights into the Landslide Processes by Hydrogeochemical and Isotopic Characterization: The Case Study of the Slano Blato Landslide (SW Slovenia)

Abstract

This study evaluates the role of groundwater in the dynamics of the Slano blato landslide using hydrogeochemical and stable isotope data. Results show that deep groundwater inflow significantly affected the landslide behavior, as demonstrated by pronounced hydrogeochemical and isotopic differences among springs. Springs within the landslide differ markedly from those in similar geological settings of the Vipava Valley, indicating a distinct local groundwater system. Groundwater is present within the landslide body even during dry periods. Waters originate mainly from a higher karstic recharge area and flow through deep flysch strata, particularly fractured sandstones, where they become enriched in dissolved ions, especially K⁺ and SO₄²⁻, and show increased mineralization in the lower parts of the landslide. Saturation indices indicate slight oversaturation with calcite and dolomite and equilibrium with quartz for most samples, reflecting interaction with carbonates and flysch sandstones. Elevated sulphate concentrations and near-equilibrium conditions for mirabilite and thenardite suggest salt-related deterioration of landslide material, enhanced by evaporation. Stable isotope data (δ¹³C_DIC, δ¹⁸O, δ²H) indicate dominant carbonate recharge, meteoric origin, evaporation effects, and long-term water–rock interaction. This study highlights the need for additional isotope tracers, groundwater age indicators, seasonal monitoring, and on-site meteorological measurements to improve interpretation.

1. Introduction

Aim of the Study

Hydrogeochemistry can provide important insights on the subsurface lithology and structure; however, interpreting hydrogeochemical processes in landslides is often challenging [1,2,3]. Although most landslides are triggered by precipitation, rainfall is not the only source of water responsible for mass movements or for mineralogical and geomechanical changes in sediments and rocks. Groundwater plays an equally important role. Groundwater circulation alters the mechanical behavior of soil and rock, thereby influencing mass movements as well as mineralogical and geomechanical transformations. From a hydrogeochemical perspective, landslides are particularly difficult to study because they are geologically and geomorphologically complex phenomena and highly heterogeneous in both space and time [4]. Fluctuations in groundwater levels can have a severe impact on slope stability due to increased pore pressure in soils [5]. While these fluctuations are most commonly attributed to precipitation events, groundwater inflow can also fluctuate with a time delay in response to precipitation. This response of groundwater to precipitation depends on hydrological characteristics, antecedent soil saturation, and the prevailing groundwater conditions. Water levels may respond rapidly to intense rainfall or exhibit gradual rises and declines during wet and dry periods of the year [6]. A deeper understanding of the processes governing slope mass movements is essential for managing the increasing risks associated with these natural hazards. Although landslide activity is often directly linked to precipitation patterns, the spatial distribution of pore pressures, the origin and age of groundwater, and groundwater flow paths within the landslide body are less frequently investigated. These aspects are particularly important for understanding creep processes and for designing effective mitigation measures [2,3].

We present a hydrogeochemical study of waters associated with one of the major landslides in Slovenia—the Slano blato landslide—which ranks among the largest landslides in the country and for which special legislation was enacted to address the damage it caused (Measures to Repair the Damage Caused by Certain Large-Scale Landslides in 2000 and 2001 Act, Official Gazette of RS, no. 3/06; [7]). The study provides a comprehensive hydrogeochemical and isotopic characterization of landslide waters using δ¹³C, δ¹⁸O and δ²H analyses. A distinctive feature of the Slano blato landslide, compared with other landslides in the wider region, is the occurrence of sodium sulfate salts (thenardite and mirabilite) on the landslide surface [8]. These salts give the landslide its name Slano blato, meaning “salty mud.” Owing to its karstic hinterland [9,10] and its complex hydrogeological and hydrogeochemical setting, the Slano blato landslide represents a particularly challenging system to characterize and to interpret the geochemical processes that may also influence landslide dynamics and movement.

2. Materials and Methods

2.1. Description of the Slano Blato Landslide and Its Geological Setting

The earliest documented reports of landslide activity date back to 1885, when landslides inundated villages and the main road in the valley [11]. Initial mitigation measures were implemented in the late nineteenth century and included slope regrading to reduce inclination and widening of the stream channel [9]. Over time, however, these structural measures were not maintained, and the landslide was reactivated during a period of intense rainfall between 17 and 19 November 2000 [11].

The landslide extends for approximately 1.3 km in length (Figure 1) and reaches its maximum width in the upper sector, where it ranges from about 60 to 200 m [9,11]. The main scarp is located near a tectonic contact where Mesozoic carbonate rocks are thrust over Eocene flysch. Upstream of the landslide, the carbonate limestone massif forms a detached tectonic block and a structural depression [10], which allows groundwater to accumulate above the less permeable flysch. The following geological composition is presented from the review paper [9] of the Slano blato landslide in 2005, when the first phase of mitigation was finished. The Eocene flysch consists of layers of marl and sandstone few centimeters up to few meters thick, and the flysch beds are strongly folded and fractured. The contact between limestone and flysch has the NW–SE (Dinaric) orientation with a dip of 10° in NE direction. Several faults with Dinaric orientation (330–345°/55–75°) were observed throughout the whole landslide area. In the region of the landslide, flysch can be divided into three regions: layers of predominant sandstone in the thickness of at least one or several meters, region with alternation of marl and 10 cm thick layers of sandstone and into region with predominant layers of marl. Limestones are of Triassic age, bedded and highly fractured. Generally, it is known that limestones have much higher permeabilities than the flysch, and the saturated hydraulic conductivities (K) obtained from 8 slug tests and 1 pumping test were the following: 10⁻⁷–10⁻⁸ m/s, for the marl and massive sandstone (flysch layers) and dense gravely clay as a part of the landslide material, 10⁻⁶–10⁻⁵ m/s for tectonized marl and fissured sandstone, and 10⁻⁴–10⁻³ m/s for clayey gravel with pieces of limestone and siltstone as a part of the landslide mass. Flysch has a very low intergranular porosity, mostly in the sandstone layers, which is accompanied by secondary fracture porosity. Contrarily, limestones exhibit mostly fracture porosity with no matrix porosity (this is more typical for dolomites [12]), and some karstic porosity (channels) can appear, but was not observed in the Slano blato region. For additional geochemical properties of flysch, one should refer to the reference of Logar et al. [9]. Groundwater accumulation in the landslide contributes significantly to the saturation and instability of the upper part of the landslide.

Since the last major movements in 2000, the Slano blato landslide has been extensively studied and monitored by several researchers [1,8,9,10,13,14,15,16]. The landslide initially developed as a rotational slide at an elevation of approximately 570 m a.s.l. Within a few days, it expanded to a length of 500 m, a width of 60–250 m, and a depth of up to 10 m. The upper section consisted of clay-rich gravel, while the lower section was composed of weathered flysch [16]. Continuous rainfall, combined with the uneven landslide surface, led to the formation of temporary lakes. The water-saturated material transformed into a mudflow, reaching a maximum velocity of 100 m/day and coming to rest at an elevation of 460 m, at the site now known as “Mud Lake.”

Between 2001 and 2004, the landslide material continued to move as a series of earthflows, occasionally evolving into very rapid or extremely rapid mudflows. To protect the nearby village of Lokavec, a small rockfill dam with a retention capacity of 5000 m³ was constructed in 2002, and approximately 200,000 m³ of landslide material was removed. From 2004 to 2007, several reinforced concrete shafts were installed in the upper part of the landslide to prevent progressive widening. Each shaft is 8 m in diameter, 24 m deep, and functions both as a dewatering measure and a retaining structure [17].

In addition to the shafts, a 2 m-high concrete dam was built to reduce slope inclination below the shafts, the surface was reshaped, and a surface drainage system was installed. Another 13 m-high concrete dam is currently under construction at the “Slap” (waterfall) site to mitigate erosion and stabilize the lower part of the landslide. As a result of these measures, landslide activity has been significantly reduced and is now mostly confined to areas north and west of the main scarp. In recent years, additional surface drainage channels have been implemented, and excavation of six additional reinforced concrete shafts is planned in the still-unstable upper part of the landslide.

2.2. Sampling Locations

We conducted hydrogeological and isotopic investigations of springs and groundwater throughout the landslide area and its surroundings. Our goal was to sample all occurrences of groundwater on the landslide body and springs in its immediate vicinity, based on available literature and several previous field observations. All spatial data were stored and analyzed within a Geographic Information System (GIS) using ESRI ArcGIS Pro 3.3.

At the Slano blato landslide, monthly measurements of spring discharge and water physical–chemical parameters were performed from November 2021 to December 2022. Measurement locations are shown in Figure 2 and

Table 1. Coordinates (LON/LAT) and altitude of the sampling sites on the Slano blato landslide.

ID	Site	Latitude	Longitude	Altitude
SB1-S	Channel below retention wall	45.907247°	13.871403°	259 m
SB2-G	Spring below retention wall	45.907666°	13.872121°	274 m
SB3-G	Red spring	45.910166°	13.868302°	356 m
SB4-G	New drainage pipe	45.910152°	13.868022°	360 m
SB5-G	New spring	45.910206°	13.867748°	367 m
SB6-G	Drainage pipe upper part	45.917118°	13.862400°	620 m
SB7-G	Spring in the scarp	45.917118°	13.861349°	624 m
SB8-G	Draining well	45.915901°	13.863229°	567 m
SB9-G	Jovšček spring	45.909662°	13.864278°	421 m

. At each site, water temperature (°C), pH, specific electrical conductivity (EC, μS/cm), total dissolved solids (TDS, mg/L), oxidation–reduction potential (ORP, mV), and dissolved oxygen (O₂, % and mg/L) were recorded using a WTW Multi 3420 multimeter (WTW GmbH, Weilheim, Germany).

For hydrogeochemical and isotopic analyses of surface and groundwater, eight water samples were collected from the Slano blato landslide on 11 October 2022. Weather conditions were sunny and warm, with an outdoor temperature of 23 °C. Sampling locations were selected to allow withdrawal of sufficient water volumes, and the sites were well distributed across the landslide area (Figure 2 and Figure 3).

Sample SB1-S was collected from a channel and represents surface water (S), corresponding to the main flow of drained water from the landslide. Sample SB2-G is a spring (G) emerging from slope gravel on the left flank of the landslide. SB3-G represents drainage from the hinterland (drainage channel SB3 was constructed in April 2022), while SB4-G was taken from a drainage pipe crossing the landslide and conveying subsurface water. Water was sampled in a natural drainage pipe installed transversely to the landslide at 1 m below the surface, which drains shallow groundwater. SB5-G is a small spring that flows only during high-water periods; consequently, many measurements from this site are missing. During our hydrogeochemical and isotope sampling, this spring was dry, so 8 of total 9 springs were sampled. SB6-G represents drainage from a horizontal well in the carbonate hinterland just above the landslide scarp, and SB7-G is a spring emerging from the scarp itself. At SB7-G, three distinct flows occur along the steep edge, and only the largest spring was sampled. SB8-G was collected from a central well (shaft) approximately 24 m deep, where groundwater is continuously drained; the sample was taken at a depth of about 10 m. The primary lining of the shaft consists of reinforced cast-in-place concrete rings, up to 30 cm thick and 1 m high, with an internal diameter of 5 m. These rings were connected vertically and separated from the sliding mass by an outer layer of drainage geotextile. As the well was gradually deepened, the rings were perforated as needed to allow water infiltration and, if necessary, reinforced with steel wire mesh or ribs and up to 10 cm of shotcrete to withstand the loads from the sliding mass during construction. In the stable base, additional high-density polyethylene (PEHD) drainage tubes (Φ125 mm) were installed at vertical intervals of 3 m, discharging into the interior of the well [17]. Finally, SB9-G is the Jovšček spring, located just outside the active landslide area. This spring comprises several partially covered sources, and its water characteristics can be correlated with groundwater on the western side of the landslide.

2.3. Hydrogeochemical Measurements

For major cation analysis, water samples were filtered through a 0.45 μm membrane filter, collected in HDPE bottles with internal stoppers, and acidified with a few drops of HNO₃ to lower the pH below 2. For major anion analysis, 50 mL of water was stored in HDPE bottles with internal stoppers. Total alkalinity was measured from 60 mL of water, filtered through a 0.45 μm membrane filter and stored in HDPE bottles. For stable isotope analyses, 30 mL of water was collected in HDPE bottles for δ¹⁸O and δ²H measurements, while δ¹³C_DIC was analyzed from two 12 mL glass ampoules filled with water filtered through a 0.45 μm membrane. All bottles and ampoules, except those used for cation analysis, were rinsed twice with the respective water sample before being filled to the top.

Analyses of 60 major and trace elements were conducted at ActLabs in Canada (Activation Laboratories Ltd., Ancaster, ON, Canada) using inductively coupled plasma mass spectrometry (ICP-MS; Code 6–Hydrogeochemistry-ICP/MS, details available at http://www.actlabs.com/ accessed on 19 December 2025). Anions (F⁻, Cl⁻, Br⁻, NO₂⁻, NO₃⁻, PO₄³⁻, SO₄²⁻) were analyzed by ion chromatography (IC; Code 6B–Ion Chromatography).

The saturation states of calcite, dolomite, gypsum, quartz, thenardite, and mirabilite in groundwater were estimated using PHREEQC for Windows 2.18, based on measured pH, alkalinity, and temperature [18,19]. Values below the detection limit (LOD) were replaced with LOD/√2 for modeling purposes, following best practices reported in the literature [20,21,22,23]. The saturation index (SI) indicates the degree of equilibrium between water and a given mineral: positive values signify supersaturation and potential mineral precipitation, negative values indicate undersaturation and no precipitation, while water is considered in equilibrium with a mineral at SI ≈ 0.1–0.2. The calculation error for individual sites is considered acceptable up to 5%.

2.4. Total Alkalinity Measurements

Total alkalinity of the samples was determined using Gran’s titration method [24], a mathematical approach for identifying the second endpoint of the carbonate equilibrium. This method allows quantification of carbonate alkalinity, expressed as the concentrations of HCO₃⁻ and CO₃²⁻. The procedure involves monitoring pH changes as a solution of unknown alkalinity is titrated with an acid of known concentration. In this study, 0.05 N HCl was used for titration, and measurements were conducted with a CAT-brand titrator (Germany). Approximately 7–10 g of each sample was weighed into a dedicated plastic container using a Mettler Toledo PL202-S2 balance. The pH electrode was calibrated prior to use with certified Mettler Toledo buffers at pH 7.00 and 4.01 ± 0.02. Detailed descriptions of the method can be found in Zuliani et al. [25].

2.5. Isotopic Measurements

The isotopic composition of carbon in dissolved inorganic carbon (DIC) (δ¹³C_DIC) was determined following Spötl’s method [26]. Ampoules containing 100–200 μL of saturated phosphoric acid were flushed with helium, after which 3–5 mL of water sample (depending on total alkalinity) was added. The resulting CO₂ was analyzed using a Europa Scientific 20-20 isotope-ratio mass spectrometer (Sercon Limited, Crewe, UK) equipped with a TG preparation module, with a measurement time of 500 s per sample.

Measurement quality was controlled using a Na₂CO₃ standard solution (Carlo Erba; 8 mg dissolved in 12 mL distilled water) with a known δ¹³C_DIC value of −10.8‰ ± 0.1 and tap water from Reaktorski Center Podgorica (δ¹³C_DIC = −12.5‰ ± 0.1). Both the standard and tap water were introduced into helium-flushed ampoules containing phosphoric acid using a needle and plastic syringe. Results were normalized to the Carlo Erba standard and are reported in δ notation (‰), representing the relative difference between the sample and the reference standard.

The isotopic composition of hydrogen (δ²H) and oxygen (δ¹⁸O) in water was determined using the H₂–H₂O equilibration [27] and CO₂–H₂O methods [28,29] on a Finnigan MAT DELTA plus mass spectrometer with a dual introduction system (Finnigan MAT GmbH, Bremen, Germany) and an automatic H₂–H₂O/CO₂–H₂O equilibrator (HDOeq48) maintained at 18 °C at the JSI laboratories [30,31]. Measurements were performed with laboratory reference materials calibrated periodically against primary IAEA standards (IAEA, 2017) on the VSMOW/SLAP scale [32]. Results were normalized to VSMOW/SLAP using two-point normalization in the LIMS program and are reported in conventional δ notation (‰). Overall analytical uncertainties were below 1‰ for δ²H and 0.05‰ for δ¹⁸O.

3. Results and Discussion

3.1. Physico-Chemical Parameters

Of the nine measurement points, only two groundwater sources remained active throughout the entire observation period: SB2-G and SB8-G. The remaining springs, as well as the central channel where surface water was monitored, dried up during the summer or exhibited such low discharge (dripping or seeping through the soil) that insufficient water was available for sampling. Consequently, the physico-chemical data (Table S1) for some springs are incomplete due to the temporary absence of water.

Rainfall data were obtained from the nearest rain gauge, located in the Otlica settlement, approximately 3.5 km northeast of the landslide. Cumulative rainfall (in mm) represents the total precipitation recorded between two consecutive measurements at the site.

3.1.1. Temperature

The spring temperature curves exhibit clear seasonal fluctuations for most springs (Figure 4A). Spring SB9 shows the most stable and lowest temperatures, reflecting its complete karstic recharge area and a large hinterland supplying deeply sourced, geochemically stable groundwater. Spring SB2, which drains from carbonate scree and also has a karstic recharge from higher elevations, shows slightly larger temperature variations. The well at SB8 exhibits moderate fluctuations, with the highest temperatures during summer and autumn and the lowest in late winter to early spring (March–April). These relatively small variations suggest a deeper water source, sampled from a significant depth in the well, with minimal influence from precipitation, except during an exceptionally rainy September 2022.

The largest temperature fluctuations are observed at spring SB3, primarily due to drainage works carried out in March 2022. Initially, measurements were taken in the soil where water accumulated and flowed slowly from the landslide. After construction of the drainage channel and installation of a pipe, larger water volumes could be sampled due to the increased flow. The SB3 temperature curve closely resembles that of SB2, with a maximum in the summer months. During the second half of the observation period, temperatures reached 19.3 °C in August and dropped to 13 °C in April. The lack of correlation with precipitation indicates that the water originates from a deeper source, which is likely warmed during its passage through the pipe.

3.1.2. pH

The pH values (Figure 4B) were generally stable, ranging from 7.5 to 8.5. An exception is spring SB4. Unusual fluctuations in pH, as well as in other physico-chemical parameters, were also observed at SB4 and spring SB3.

3.1.3. Specific Electrical Conductivity (EC)

Figure 4C shows the electrical conductivity (EC) fluctuation curves. The lowest and most stable EC was observed at karstic spring SB9, consistent with the explanation for its temperature stability. Spring SB2 also exhibited stable EC, but values were approximately twice as high. The highest EC values were recorded at SB3 and SB8, both representing deeper groundwater with high mineralization due to prolonged residence times in the flysch. This extended interaction with the host rocks promotes greater dissolution of minerals, resulting in elevated ion concentrations. At these locations, EC fluctuated in response to precipitation, with higher values during rainfall events. Despite being deeper-sourced, the groundwater at SB3 and SB8 responds relatively quickly to precipitation while remaining highly mineralized.

The relationship between EC and discharge at SB8 (well) is shown in Figure 5. Here, higher EC coincides with greater water discharge, which may be attributed to the “piston effect” typical of karst systems. In this process, incoming groundwater pushes out older, highly mineralized water that has interacted with surrounding rocks over a longer period [33,34,35]. In this case, slow groundwater flow through fractured sandstones of the flysch formation allows extended water–rock interaction; during heavy rainfall, increased pressure likely displaces this mineralized water, leading to simultaneous increases in EC and discharge. However, as measurements were taken only monthly, this hypothesis cannot be fully confirmed.

In contrast, springs SB2 and SB9, which originate in carbonate rocks and carbonate scree deposits, display constant EC due to the higher permeability of these materials, resulting in more rapid water flow and limited mineralization.

3.1.4. Oxidation–Reduction Potential (ORP)

The oxidation–reduction potential (ORP) values (Figure 6A) were generally stable, ranging from −80 to −40 mV. Similarly to the pH results, measurements at sites SB3 and SB4 were distinctive. At SB3, the spring initially exhibited more oxidizing conditions, as measurements were taken in small soil depressions (Figure 7A). The surrounding red-colored soil indicated iron oxidation resulting from reactions with pyrite. Following the construction of the drainage channel in April 2022, ORP values at SB3 stabilized around −70 mV (Figure 7B). At SB4, ORP initially indicated strongly reducing conditions (~−180 mV), but subsequently increased and stabilized between −80 and −60 mV.

3.1.5. Dissolved Oxygen (DO)

Dissolved oxygen (DO) values (Figure 6B) were generally uniform across all springs, ranging between 95 and 105%, with the exception of SB3 and SB5. At these two sites, measurements were taken in small basins, where the presence of silty fractions in the water may have affected the readings. At SB3, lower DO values were recorded before the construction of the drainage, when measurements were taken directly in the soil; after the drainage was completed, DO stabilized around 100%. A similar pattern of fluctuation was observed at SB5.

3.1.6. Spring Discharges

The landslide remains saturated even during periods of severe drought. At a depth of 10 m in the well at the upper part of the landslide (SB8), sufficient groundwater was available throughout the entire measurement period, despite 2022 being extremely dry until September, while most other groundwater locations dried up. This suggests that substantial water is retained within the flysch layers of the landslide for extended periods.

Discharge fluctuations (Figure 6C) in relation to monthly precipitation were similar across sampling locations, differing primarily in absolute discharge (L/s) and the likelihood of the spring drying out. Most springs, except SB2 and SB8, completely dried up during the summer. The largest fluctuations were observed at SB4. The curve for SB4 shows extremely large amplitude fluctuations, although data are incomplete due to the pipe’s installation in March 2022. Similarly, SB6 displayed large amplitude fluctuations. In contrast, SB7 exhibited smaller amplitude fluctuations. Flow measurements generally showed delayed responses to precipitation events, indicating longer retention times and slower seepage through the flysch.

The response of spring SB9 was particularly notable. It was dry on 1 August and 2 September 2022, but more than 600 mm of precipitation fell between 2 September and 11 October. This resulted in a rapid, turbulent flow that later diminished and stabilized by 11 November, demonstrating a typical karstic response with high discharge fluctuations due to recharge from the carbonate area. Conversely, SB5, which generally exhibits very low discharge as water seeps from a crack in the slope, remained completely dry between 1 August and 11 October. Even by the final measurement on 8 December, the discharge was insufficient for parameter measurements.

3.2. Hydrogeochemical Analyses

Table 2. Concentration of major cations, anions, total alkalinity and isotopic composition of carbon in dissolved inorganic carbon (δ¹³C_DIC), oxygen (δ¹⁸O) and hydrogen (δ²H) for different sites (from sampling 11 October 2022).

Parameter	Unit	SB1-S	SB2-G	SB3-G	SB4-G	SB6-G	SB7-G	SB8-G	SB9-G
EC (20 °C)	µS/cm	746	523	1358	955	810	583	1543	241
DO	mg/L	9.10	10.36	9.22	9.34	8.63	9.11	9.52	10.69
DO	%	94.1	99.6	98.1	100	104.3	100.7	98.0	98.3
ORP	mV	−67.1	−54.4	−43.9	−66.6	−63	−55.3	−65.7	−55.1
pe (from Eh)	/	−1.19	−0.96	−0.78	−1.18	−1.11	−0.98	−1.16	−0.97
pH	/	8.065	7.854	7.655	8.059	7.971	7.866	8.038	7.871
T	°C	16.2	12.4	16.7	17	21.3	17.2	14.3	9.6
Alkalinity	mmol/L	4.90	4.40	8.4	3.8	6.30	5.4	7.1	2.0
HCO3	mgHCO3−/L	299	268	513	232	384	330	433	122
Ca	mg/L	96.8	83.8	143	83.7	145	109	164	38.9
K	mg/L	4.26	1.94	5.88	5.73	2.41	1.88	9.4	0.19
Cl	mg/L	3.3	2.9	4.4	2.8	2.8	2.9	5.4	1.8
Mg	mg/L	27.6	16.1	59.6	45.7	29.9	19.7	81.5	8.38
Mn	mg/L	0.0013	0.0003	0.0446	0.0005	0.0026	0.0019	0.0438	0.0002
TDS	mg/L	749	521	1366	953	812	583	1532	241
Na	mg/L	24.4	5.6	92.1	78.2	8.7	6.46	92.9	1.41
NO3	mgNO32−/L	0.94	1.58	0.43	0.25	0.14	0.85	1.12	3.18
SO4	mgSO42−/L	187.0	39.1	383.0	326.0	145.0	55.8	572	4.25
Total Fe	mg/L	<0.01	<0.01	0.01	0.01	0.01	0.02	0.01	<0.01
As	mg/L	0.00026	0.00022	0.00038	0.00026	0.00018	0.00013	0.00023	0.00017
Ba	mg/L	0.0446	0.0472	0.0382	0.0301	0.0409	0.0958	0.0272	0.004
Br	mg/L	<0.1	<0.09	<0.3	<0.1	<0.1	<0.09	<0.3	<0.06
F	mg/L	<0.04	<0.03	<0.1	<0.05	<0.04	<0.03	<0.1	<0.02
Total Cr	mg/L	0.0021	0.0007	0.0015	0.0061	<0.0005	<0.0005	<0.0005	<0.0005
Li	mg/L	0.02	0.005	0.049	0.039	0.022	0.017	0.057	<0.001
Se	mg/L	0.0006	0.0007	0.0002	0.0007	0.0004	0.0011	0.0021	<0.0002
Sr	mg/L	0.816	0.28	2.07	1.61	0.943	0.731	3.1	0.0354
Pb	mg/L	0.00024	0.00011	0.00009	0.00009	0.0001	0.0005	0.00007	0.00022
Al	mg/L	0.011	0.004	0.009	0.009	0.006	0.1	0.007	0.006
Sb	mg/L	0.00024	0.00005	0.00009	0.00022	0.0001	0.00008	0.00011	0.00003
Cu	mg/L	0.0012	0.0005	0.0006	0.0009	0.0004	0.0007	0.0007	0.0002
Be	mg/L	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
Cs	mg/L	0.000015	0.000001	0.00002	0.000011	0.000004	0.000006	0.000064	<0.000001
Zn	mg/L	0.0039	0.0071	0.0026	0.0041	0.0039	0.0068	0.0336	0.0072
Cd	mg/L	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
Co	mg/L	0.000057	0.000022	0.000086	0.000054	0.000022	0.000033	0.000169	0.000012
Sn	mg/L	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
Mo	mg/L	0.0009	0.0002	0.0007	0.0011	0.0003	0.0002	0.0004	<0.0001
Ni	mg/L	0.0011	0.0004	0.0023	0.0013	0.0031	0.001	0.0032	<0.0003
Rb	mg/L	0.00389	0.000796	0.00377	0.00317	0.00172	0.00178	0.00872	0.000141
Ag	mg/L	<0.002	<0.002	<0.002	<0.002	<0.002	<0.002	<0.002	<0.002
U	mg/L	0.000584	0.000359	0.000581	0.000692	0.00048	0.000338	0.000777	0.000217
V	mg/L	0.0009	0.0004	0.0006	0.0005	<0.0001	0.0002	0.0002	0.0005
Hg	mg/L	<0.0002	<0.0002	<0.0002	<0.0002	<0.0002	<0.0002	<0.0002	<0.0002
Total P	mg/L	<0.08	<0.06	<0.2	<0.1	<0.08	<0.06	<0.2	<0.04
Tl	mg/L	0.000008	0.000004	0.000011	0.000007	0.000011	0.000005	0.000008	0.000004
Ti	mg/L	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
Te	mg/L	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
Si	mg/L	4.2	3.2	6	4.6	5.5	4.4	5.1	0.9
δ13CDIC	(‰)	−10.5	−12.2	−14.1	−9.3	−9.8	−11.0	−7.3	−6.5
δ18O	(‰)	−6.58	−7.01	−6.77	−6.45	−7.22	−7.36	−7.24	−7.67
δ2H	(‰)	−40.5	−44.9	−42.4	−40.4	−44.9	−45.4	−45.3	−47.4

presents the results of the hydrogeochemical analyses. Spring SB5 was dry at the time of sampling and is, therefore, not included in the following results. Sampling was conducted on 11 October 2022, under an outdoor temperature of approximately 22 °C.

The most abundant ions associated with interaction with carbonate rocks are Ca²⁺, Mg²⁺, and HCO₃⁻. Limestone (CaCO₃) and dolomite (CaMg(CO₃)₂) in the karstic hinterland are the primary sources of these ions. Elevated concentrations of these ions indicate a carbonate origin, where groundwater infiltrates into the carbonates and subsequently flows underground through the flysch, which contains permeable fractured sandstone layers. Highest Ca²⁺ concentrations were observed at SB8, whereas SB9 exhibited the lowest values. The highest Mg²⁺ concentrations were measured at SB3 and SB8. Alkalinity, expressed as the content of HCO₃⁻ and CO₃²⁻ ions, is comparable for SB3, SB4, SB6, and SB8.

Ions originating from flysch rocks include Na⁺, K⁺, and SO₄²⁻. Sodium primarily derives from the weathering of thicker layers of quartz-feldspar lithic sandstone, where plagioclase and lithic grains of pre-existing rocks provide a local source of Na⁺ [36]. Potassium is likely released from K-clays. The highest concentrations of Na⁺ and K⁺ are observed at spring SB8. Springs SB3 and SB8 also exhibit high concentrations of other ions, reflecting the overall elevated mineralization of these waters.

The sulfate ion (SO₄²⁻) originates from the oxidation of pyrite during the weathering of flysch, with pyrite commonly occurring as a coating on sandstone grains (Figure 8) [36]. Extremely high SO₄²⁻ concentrations are observed in SB3, SB4, and SB8. Comparing the relative proportion of sulfate in the landslide area (Figure 9) shows that the springs within the landslide contain a much higher proportion of SO₄²⁻, whereas springs SB2 and SB9, have significantly lower sulfate content.

According to TDS measurements, springs SB3 and SB9 exhibit the highest concentrations of total dissolved solids (>1000 mg/L), indicating a deeper groundwater source with slower flow through the flysch layers. In contrast, springs SB2 and SB8 have the lowest TDS values (approximately 500 and 240 mg/L, respectively), reflecting shorter water retention times and faster groundwater flow. The slightly elevated TDS at SB2 is likely due to a longer flow path and greater interaction with the rock, as this spring is located at a lower altitude than SB9.

Groundwater entering the landslide differs significantly from the springs SB2 and SB9, which are located in the immediate vicinity of the landslide. Both SB2 and SB9 (and SB7) exhibit a Ca-Mg-HCO₃ facies, with dominant Ca²⁺ and Mg²⁺, indicating strong interaction with carbonate rocks. SB4 and SB8 have more mixed facies of Ca-Mg-Na-SO₄-HCO₃, SB3 and SB6 have a similar facies of Ca-Mg-Na-HCO₃-SO₄ and SB1 has a facies of Ca-Mg-HCO₃-SO₄. The relatively stable temperature of these springs suggests a deep aquifer, largely unaffected by precipitation or seasonal temperature variations. SB2 also displays elevated sulfate (SO₄²⁻) concentrations, reflecting limited interaction with flysch rocks. In contrast, the waters flowing through the landslide obviously exhibit more mixed facies, dominated by SO₄²⁻ and also Na⁺, which reflects prolonged water retention and enhanced mineral dissolution in the flysch.

The lowest ion concentrations and overall mineralization are observed at spring SB9, which has the shortest retention time and minimal interaction with the surrounding rocks. Interestingly, SB9 also shows the highest nitrate concentration (3.18 mg/L NO₃⁻), likely due to water infiltration through soil and the oxidation of nitrogen-rich organic matter [37].

The geochemical composition of the water samples is illustrated in the Piper diagram (Figure 10). The diagram highlights differences in water composition and origin. Springs SB3, SB4, and SB8, representing groundwater within the landslide, are influenced by flysch weathering, which contributes to elevated Na⁺ and K⁺ concentrations compared to the Ca²⁺ and Mg²⁺ derived from carbonate dissolution. Among these, SB3 and SB8 are most similar, reflecting deep groundwater sourced primarily from flysch.

Site SB1 (blue circle) occupies an intermediate position in the diagram, indicating a mixture of waters from different origins—both carbonate-fed and flysch-fed springs. In contrast, springs SB2, SB6, SB7, and SB9 contain higher calcium concentrations. SB2 and SB9 are springs in carbonate rocks and slope debris, while SB6 and SB7 emerge near or on the scarp at the top of the landslide. At these locations, interaction with flysch is limited, whereas interaction with carbonates in the hinterland dominates. Further down the landslide, mineralization increases markedly due to prolonged contact with flysch rocks, with Na⁺, K⁺, and SO₄²⁻ becoming the dominant ions (Figure 10).

3.3. Saturation Indices

Saturation indices for all samples, except SB9, indicate slight oversaturation with calcite and dolomite, while remaining in equilibrium with quartz. This reflects interaction with both carbonate rocks and flysch sandstones. All samples show strong oversaturation with dolomite; however, dolomite is generally not known to precipitate under typical atmospheric conditions [38]. In contrast, Ca-carbonate precipitation usually occurs in waters that are 5–10 times saturated with respect to calcite [39]. Therefore, the observed oversaturation primarily reflects prolonged water–rock interaction, with carbonate dissolution being the dominant geochemical process driving this saturation state. Among the samples, SB9 shows the lowest dolomite saturation, indicating only limited interaction with dolomite (

Table 3. Results of the Saturation Index (SI) modeling in PHREEQC. ERR = analysis error in %, CAL = calcite, DOL = dolomite, GYP = gypsum, QTZ = quartz, THE = thenardite, MIR = mirabilite.

Name	ERR (%)	CAL	DOL	GYP	QTZ	THE	MIR
SB1-S	−3.94	0.86	2.51	−1.33	0.05	−8.70	−7.48
SB2-G	4.54	0.55	1.72	−1.97	0.02	−10.60	−9.20
SB3-G	−0.89	0.78	2.51	−1.00	0.20	−7.35	−6.16
SB4-G	4.43	0.65	2.37	−1.21	0.07	−7.49	−6.31
SB6-G	4.35	1.11	2.91	−1.31	0.06	−9.74	−8.77
SB7-G	5.76	0.81	2.24	−1.76	0.05	−10.36	−9.19
SB8-G	0.22	1.06	3.14	−0.83	0.18	−7.21	−5.90
SB9-G	10.59	−0.07	0.50	−3.13	−0.46	−12.66	−11.13

).

All samples were undersaturated with gypsum, thenardite, and mirabilite (Table 3). Gypsum (CaSO₄·2H₂O) and thenardite (Na₂SO₄) are common surface salt deposits [36], which were also observed during fieldwork (Figure 11). Thenardite forms primarily under warmer conditions and dissolves rapidly during precipitation events. It is the dehydrated form of mirabilite (Na₂SO₄·10H₂O), which converts to thenardite under low humidity and at room temperature [36]. Both minerals are known to cause significant weathering and damage to building stones [36,40,41,42], and similarly contribute to the degradation of rocks and sediments in landslides. Rodriguez et al. [42] have stated that rapid evaporation and the high degree of solution supersaturation reached in micropores before thenardite precipitation result in high crystallization pressure and greater damage to porous materials than mirabilite, which crystallizes at lower supersaturation ratios and generally as efflorescence. The presence of these two minerals can therefore accelerate the breakdown of flysch, producing larger quantities of fine sediments, which may influence the reactivation and movement of the slope on a broader scale.

We modeled the potential precipitation of thenardite and mirabilite under extreme evaporation conditions (Figure 12) using saturation indices. The results indicate that precipitation is not possible under the sampled conditions, as both minerals exhibited negative saturation indices, likely because the water samples were collected below the surface. However, under an extreme evaporation scenario, where only a small percent of the water remains, the saturation approaches equilibrium. Our field sampling was conducted in October, whereas evaporation predominantly occurs during late spring and summer (May to September). Higher temperatures during these periods could therefore promote the surface precipitation of both minerals. The observed occurrence of thenardite and mirabilite on the landslide surface may result from more complex processes not captured in this simplified geochemical modeling. A comparable situation has been reported at Slani potok creek in Croatia, where efflorescence of thenardite on flysch was confirmed by XRD and SEM analyses [43].

Whether the formation of thenardite and mirabilite in the landslide is driven by evaporation can also be assessed using the Sr²⁺/Ca²⁺ molar ratio, with evaporative waters typically exhibiting values >1‰. According to Meybeck [44], elevated strontium concentrations originate from the dissolution of celestite, which commonly occurs in association with gypsum. Using this approach, de Montety et al. [2] determined the origin of groundwater in the Super-Sauze landslide in France. The Sr²⁺/Ca²⁺ ratios for individual springs at the Slano blato landslide are presented in

Table 4. Sr²⁺/Ca²⁺ molar ratio (in permille) of springs on the Slano blato landslide.

Location	SB-1	SB-2	SB-3	SB-4	SB-6	SB-7	SB-8	SB-9
Sr2+/Ca2+	3.86	1.53	6.62	8.80	2.97	3.07	8.65	0.42

. The results indicate that evaporation influences all springs except the carbonate spring SB9, supporting the conclusion that extreme evaporation contributes to the precipitation of both minerals. The highest ratios were observed at SB3, SB4, and SB8, all of which represent deeper, highly mineralized groundwater sources.

3.4. Isotopic Analyses

3.4.1. Isotopic Composition of Carbon in Dissolved Inorganic Carbon (δ¹³C_DIC)

The results of the δ¹³C_DIC analyses are presented in Figure 13. The accompanying table shows the calculated contributions of the organic and carbonate fractions for each water sample. The organic fraction originates from the decomposition of plant material and soil microorganisms, while the carbonate fraction reflects the dissolution of carbonate minerals. The highest proportion of the carbonate fraction was observed at SB9, reaching 75%, whereas the lowest value, 45.8%, was measured at SB3.

The δ¹³C_DIC results indicate a pronounced influence of carbonate dissolution in most springs (Figure 14). The contribution of soil-derived organic carbon is significant only at SB3 (Figure 2), where inorganic carbon from organic matter exceeds 50%, resulting in the lowest proportion of the carbonate fraction. This is consistent with the dense vegetation cover in the SB3 hinterland. The low δ¹³C_DIC value of −14.1‰, combined with a high total alkalinity of 8.4 mM, suggests a biological origin of CO₂, likely from microbial activity in the weathered material, reflecting enhanced alkalinity and prolonged infiltration through the soil profile.

A high carbonate fraction was also observed at SB8 (Figure 13), the well in the upper landslide. Although the groundwater at SB3 and SB8 has a similar chemical composition, the influence of carbonate dissolution in the hinterland is more pronounced at SB8, while the effect of surface water infiltration is significantly lower than at SB3. In contrast, the SB2 spring, located in the carbonate slope deposits, exhibits a relatively low carbonate content.

The relation between total alkalinity and δ¹³C_DIC (Figure 14) indicates that most springs, except SB3 (Figure 2), behave as open systems in which DIC equilibrates with soil CO₂ and carbonate dissolution occurs. In the case of SB3, both biogeochemical processes influence DIC (falling between fractionation lines 1 and 2 in Figure 14): open-system equilibration with soil CO₂ derived from organic matter degradation (δ¹³C_CO2 = −26.0‰) and carbonate dissolution (δ¹³C_CaCO3 = 0‰) [43]. Consequently, SB3 water exhibits a slightly different isotopic signature than other springs, characterized by higher total alkalinity and a greater contribution from soil-derived CO₂, rather than mineral dissolution (feldspars and carbonates). This reflects its primarily subterranean flow through flysch layers with minimal contact with carbonates.

In contrast, spring SB9 (Figure 2) shows minimal influence from soil CO₂, with the highest δ¹³C_DIC value (−6.5‰) and the lowest alkalinity (2.0 mM; Figure 14, Table 2). Its DIC originates almost entirely from carbonate dissolution, reflecting its fully karstic source. Thus, the carbonate recharge area of Mala Gora (Figure 1), located in the landslide’s hinterland, exerts a major influence on the water composition of the landslide and its springs, consistent with the overall geochemical findings (Figure 10).

3.4.2. Isotopic Composition of Hydrogen (δ²H) and Oxygen (δ¹⁸O)

The isotopic values of δ²H range from −47.4 to −40.4‰ (range 7‰) and δ¹⁸O from −7.67 to −6.45‰ (range 1.22‰) (Table 2). No information about local isotopic composition of precipitation, surface and groundwater for Slano blato is publicly available. Therefore, the δ²H versus δ¹⁸O relationship shown in Figure 15 is presented alongside meteoric water lines (MWLs) from nearby SLONIP precipitation monitoring stations [45]: Portorož, located approximately 55 km southwest, and Kredarica, approximately 60 km north. Those MWLs can serve as potential precipitation input end-members. All data obtained in this study plot between these MWLs, indicating that the main water source is a mixture of local modern precipitation typical of the northeastern Adriatic coastal region and the Alpine region in western Slovenia [45].

Isotope data from Slano blato align along the line δ²H = 5.93 × δ¹⁸O–2.15, which has a lower slope and intercept than the LMWLs. This suggests that meteoric water has undergone evaporation and/or other geochemical processes that modified its stable isotope composition. Among these processes, changes in redox potential—particularly pronounced on the landslide surface during summer—and water–rock interactions at elevated temperatures may be significant as indicated also with other geochemical results. A more precise assessment of redox effects as well as water residence time estimations would require higher-frequency sampling of precipitation, surface water, and groundwater on the landslide.

A distinct difference in δ²H and δ¹⁸O values is observed between lower-altitude sampling locations (258–359 m) and higher-altitude sites (420–624 m), with an estimated vertical δ¹⁸O gradient of approximately 0.27‰ per 100 m. This gradient is consistent with previous altitude effect estimates for precipitation in the northeastern Adriatic [46] and for groundwater in the Alps and coastal regions [47] of Slovenia.

4. Conclusions

• The chemical composition of springs within the Slano blato landslide differs markedly from springs in similar geological settings in the broader Vipava Valley.
• Waters primarily originate from the higher karstic recharge area and flow through deep flysch layers (fractured sandstone beds), where they acquire higher concentrations of dissolved ions—especially K⁺ and SO₄²⁻—and increased mineralization in the lowland section of the landslide.
• Groundwater persists within the landslide body even during dry periods.
• Saturation indices for all samples, except SB9, indicate slight oversaturation with calcite and dolomite and equilibrium with quartz, reflecting interactions with both carbonates and flysch sandstones.
• Salts such as mirabilite and thenardite can contribute to the degradation of landslide material. Although simple geochemical modeling could not directly confirm their crystallization, it suggests that near-equilibrium—and potential precipitation—could occur under increased evaporation. Landslide processes expose, mix, hydrate, and oxidize materials, promoting secondary mineralization (e.g., gypsum, mirabilite) and bringing primary minerals (e.g., pyrite, andradite) to the surface. Higher sulfate concentrations are observed in the central part of the landslide, with the exception of one spring located outside the landslide.
• Waters from carbonate scree locations respond more rapidly hydrologically than deeper groundwater sources.
• δ¹³C_DIC values indicate a dominant influence of carbonate dissolution with minimal contribution from organic matter in most springs.
• δ¹⁸O and δ²H values indicate a meteoric origin of the groundwater, modified by evaporation and prolonged water–rock interaction.
• Our hydrogeochemical and isotope investigations confirm the idea of the structural model of the Slano blato landslide proposed by Placer et al. [10], in which the limestone massif above the landslide scarp forms a detached tectonic block and a structural depression. This configuration allows groundwater to accumulate in the carbonates above the less permeable flysch and enables delayed, deeper water flow through the flysch.
• We have shown that the water chemistry changes significantly during this underground flow, and both hydrogeochemical composition and isotope data support the idea of deeper underground water flow through flysch. For further confirmation, a hydrogeological model with quantitative flow modelling should be conducted; this was beyond the scope of our study and is suggested for future investigations. The same conclusion applies to the observation of the relationship between discharge and specific electrical conductivity values at the SB8 site, which suggests a delayed groundwater response to precipitation within the flysch. In summary, this study represents an investigation based on a single sampling campaign using stable isotope tracers (C, H, O) to assess landslide behavior. Future research should include additional stable isotope tracers (e.g., S: δ³⁴S_SO4, δ¹⁸O_SO4), water-age tracers (³H, CFCs), seasonal hydrogeochemical analyses for more robust interpretations and hydrogeological modeling. Installing a meteorological station directly on the landslide would allow precise monitoring of local precipitation events and improve correlations between rainfall and hydrogeochemical responses.