Temporal Variation of Phosphate Contamination In Tropical Epikarst Springs, Gunungsewu Karst Area, Indonesia

Epikarst springs in less developed karst areas typically have predominant diffuse flows where groundwater chiefly moves through spaces between rock grains. This might lead to different flow characteristics between flow dominated by diffuse and dominated by large conduits resulting from dissolution processes. This research aimed to analyze the multitemporal characteristics of phosphate contamination of epikarst spring water in the Gunungsewu Karst Area, Java Island, Indonesia. Phosphate levels were determined from the spring water samples collected every two weeks for nine months to encompass two seasons characteristic of tropical environments: dry and rainy. Results show that the phosphate concentration of the Gedaren Spring exceeded the water quality standard at different times, was likely high in the rainy season and low in dry season and showed positive correlation with rainfall and flow discharge. Further, it was identified that the phosphate was probably generated by agricultural activities or different sources from sulfate and nitrate, as indicated by their weak positive correlations between phosphate-sulfate or phosphate-nitrate.


Introduction
Karst is a landscape with distinctive hydrological systems that play a vital role in clean water provision [1,2].Another distinguishing feature is the dissolution process that creates conduits (a type of secondary porosity), even though water still oftentimes moves through intergranular pores via diffusion that are developing in soluble rocks [3,4].Under these conditions, water falling on the ground moves down through subsurface fissures and channels and subsequently accumulates in karst aquifers or underground rivers [5].
Karst drainage comprised two zones: subcutaneous (epikarst) and subterranean (endokarst) [6].Both have different roles but are comparably significant for this system.Epikarst is located in the uppermorst zone and has high permeability and porosity, which is essential for creating a suitable water storage medium [7,8].Epikarst is also important for feeding of underground rivers even during very long periods of drought, and regulates water infiltration, storage, and evapotranspiration in the region [9][10][11][12].For these reasons, Epikarst zone has an essential part in karst hydrological systems.
With all these characteristics, karst areas are particularly prone to pollution.Chemical weathering enlarges pores, fissures, and conduits, creating rapid flow and exchange of water and thin soil layers into which pollutants on the surface quickly seep [13].These regions are also characterized by ponors and basins that potentially act as input points for contaminants.In addition, a high degree of connectivity between surface runoff and underground flow, limited surface water, and thin soil cover can trigger this rapid infiltration [14], transporting contaminants to a very great distance [15].Some parts of a karst area may be less developed with dominant diffuse flows.At these locations, epikarst springs are often found with small to moderate discharge.Also, areas with developed soils have high clay contents [14].Under these conditions, the groundwater in the region has different chemical properties from karst at later development stages [16,17,18], which consequently shape the characteristics of its contamination.
This research was intended to determine the multitemporal characteristics of phosphate contamination in epikarst springs.Excess phosphate has raised concerns about water quality decline, contamination, and threats to dependent organisms [19] because surface runoff can transfer the phosphate to nearby water sources, such as groundwater or larger bodies of water.In addition, this pollutant can increase algae growth and block sunlight underwater plants need [20].Generally, the phosphate found in water comes from mammal feces, rinse water containing soap and detergent, and waste from pulp and paper industries [19].Phosphate contamination can be linked to improper domestic and industrial management practices with by-products that affect water quality levels, posing challenges that developing countries must overcome [21].

Method Research Location
The research location is Gedaren Spring in the Kapanewon (Subdistrict) of Ponjong, Gunungkidul Regency, Province of D.I. Yogyakarta (Figure 1).Gedaren is one of the springs the local government relies on to provide raw water for a local drinking water company.However, there are intensive anthropogenic activities in its catchment area.The land use consists largely of dryland/non-irrigated fields, amounting to 45.99% of the catchment.The rest is covered by shrubs (40.68%) and settlements (13.36%).These anthropogenic activities generate pollutants, including phosphate, and increase the probability of phosphate contamination because this element is easily dissolved and transferred by water into the spring.

Data Acquisition Method
The data used in this research comprised phosphate, nitrate, and sulfate concentrations, rainfall intensity, discharge, and cropping calendar.Phosphate, nitrate, and sulfate levels were determined in the laboratory from the water samples collected every two weeks for nine months, including both dry and rainy seasons.This interval is considered efficient and representatively describes the hydrogeochemical condition of karst area [22,23].Sampling began in August 2022 and ended in April 2023.River discharge was also measured in the field biweekly at the same time as the water sampling.Rainfall intensity (secondary data) was obtained from the Ponjong Agricultural Extension Center, and information on the cropping calendar was gathered during in-depth interviews with informants in the Gedaren Spring's catchment.It sets the maximum allowable presence of phosphate at 0,2 mg/l.Temporal analysis was performed statistically using bivariate regression to determine the phosphate-discharge and phosphate-rainfall correlations.In addition, bivariate analyses of phosphate, nitrate, and sulfate concentrations were performed to ascertain whether these pollutants come from the same source(s).Further, in-depth interviews with informants were conducted to identify community behavior that might contribute to phosphate fluctuations in the Gedaren Spring.

Results and Discussion
The phosphate analysis showed concentrations above the regulated standard throughout the research period in seven out of the 17 collected samples (see bold values in Table 1).This finding suggests that despite predominant diffuse flow, phosphate was still found in high amounts in the spring.In addition to fissures and thin soils in the epikarst zone, fractures and pores enlarged by the growth of plant roots might create input points for phosphate left on the surface or the top soil layer to seep into the subterranean drainage, thus contaminating the spring water.Source: Data Analysis (2023) Table 1 shows that samples 2, 9, 11, 12, 13, 15, and 17 contained more phosphate than the standard, with the highest concentration at 3.365 mg/l on February 4 and the lowest at 0.059 mg/l on April 2, 2023.During the research period, the spring water had an average of 0.370 mg/l of phosphate or above its permissible maximum level, 0.2 mg/l.Many pollutants can reach groundwater because fissure development allows pollutant-carrying water to enter karstic aquifer systems easily [24]; however, this process is still controlled by natural infiltration and percolation through intergranular voids.This answers why most of the spring water samples had lower phosphate levels than the standard, even though the surrounding land was primarily used for agriculture.
The multitemporal analysis in this research took seasons into account, given that rainfall is the main factor influencing variations in the spring's hydrological conditions.In the Gunungsewu Karst Area, the rainy season starts in October and ends in March, whereas the dry season lasts from April to September [25].Figure 2 shows four charts illustrating the correlations between phosphate, flow discharge, and rainfall in different seasons.The first two charts (Figures 2a-b) suggest phosphate fluctuated more widely as rains intensified.It increased from 0.08 to 0.168 mg/l at the beginning of the rainy season in October but decreased from 0.211 to 0.059 mg/l entering the dry season.This trend indicated a positive correlation between rainfall and phosphate dynamics.Other than rainfall, phosphate is also positively correlated with flow discharge.In particular, higher rainfall intensity flow discharge relust in a higher phosphate concentration in the spring and vice versa.This condition is called phosphate mobilization [26,27].As seen in Figures 2c and 2d, the discharge-phosphate and rainfall-phosphate correlations are positively correlated as shown by the Rvalues of 0.45 and 0.40, respectively.From these values, it can be inferred that flow discharge is strongly correlated with phosphate concentration.This might be caused by the direct influence of the discharge on phosphate fluctuations.
Further bivariate analyses assessed the correlations between phosphate, nitrate, and sulfate levels to determine if they came from the same source(s).Nitrate is mainly generated by farming activities [28], and sulfate is from the dissolution of gypsum, naturally occurring sulfate in the soil, atmospheric deposition, and the by-products of anthropogenic activities, e.g., industrial waste, wastewater, and remnants or residues from applications of organic and synthetic fertilizers [29].
Figures 3a-d show R-values of 0.33 and 0.18 for the phosphate-nitrate and phosphate-sulfate correlations, respectively, suggesting that phosphate correlates more strongly with nitrate than sulfate.Nevertheless, both R-values indicated weak correlations,suggesting that the three pollutants did not originate from the aforementioned sources.According to previous studies by Cahyadi et al. [7] at the Selonjono Spring and Adji et al. [30] in the Seropan Underground River in the Gunungsewu Karst Area, phosphate comes from agricultural fields, sulfate from domestic waste (gray water or detergent), and nitrate possibly from fecal matter.Agricultural practices apply fertilizers that leave residues of phosphate in the soil, and Loganathan et al. [31] confirmed that the groundwater's phosphate content is likely to increase with the added fertilizers.

Conclusion
The phosphate concentration at the Gedaren Spring, a tropical epikarst spring, varies or fluctuates temporally according to season.It is positively correlated with flow discharge and rainfall, meaning that when either of the increases, the phosphate concentration also tends to become higher.These positive correlations, however, differ in strength: phosphate is more strongly regulated by discharge that exerts a direct influence rather than rainfall.In addition, there are weak correlations between phosphate, sulfate, and nitrate concentrations, indicating that the three elements come from different sources.Phosphate is generated by agricultural practices, while sulfate and nitrate can be found in gray water and fecal matter, respectively.

Figure 1 .
Figure 1.Map showing the location of the Gedaren Spring and the land use of its catchment (Sources: DEMNAS Geospatial Information of the Republic of Indonesia, Topographic Map 1:25,000, and field survey)

Figure 2 .
Figure 2. a. Phosphate-rainfall correlation chart, b.Phosphate-discharge correlation chart, c.Bivariate analysis of flow discharge vs. phosphate, and d.Bivariate analysis of rainfall vs. phosphate.

Figure 3 .
Figure 3. a. Bivariate analysis of nitrate vs. phosphate, b.Bivariate analysis of phosphate vs. sulfate, and c.Bivariate analysis of nitrate vs. sulfate.

Table 1 .
Phosphate, Nitrate and Sulfate concentrations, discharge, and rainfall at the Gedaren Spring from August 2022 to April 2023