Analytical Science and Technology (분석과학)

The Korean Society of Analytical Sciences (한국분석과학회)
권호 표지
DOI QR코드

DOI QR Code

Identification of pollutant sources using water quality and stable isotope ratios of inflow tributaries in the lower reaches of the Han-River

  • Hong, Jung-Ki (Han-River Environment Research Center, National Institute of Environmental Research) ;
  • Lee, Bo-Mi (Han-River Environment Research Center, National Institute of Environmental Research) ;
  • Son, Ju Yeon (Han-River Environment Research Center, National Institute of Environmental Research) ;
  • Park, Jin-Rak (Han-River Environment Research Center, National Institute of Environmental Research) ;
  • Lee, Sung Hye (Han-River Environment Research Center, National Institute of Environmental Research) ;
  • Kim, Kap-Soon (Han-River Environment Research Center, National Institute of Environmental Research) ;
  • Yu, Soon-Ju (Han-River Environment Research Center, National Institute of Environmental Research) ;
  • Noh, Hye-ran (Han-River Environment Research Center, National Institute of Environmental Research)
  • Received : 2018.10.19
  • Accepted : 2019.04.01
  • Published : 2019.04.25
Download PDF
⟨ Previous Next ⟩

Abstract

Despite the expansion of sewage treatment facilities to reduce pollutants in the tributaries of the Han River, water pollution accidents such as fish deaths continue to frequently occur. The purpose of this study was to identify the pollutant sources using water quality and stable isotope ratio (${\delta}^{15}N$, ${\delta}^{13}C$, ${\delta}^{15}N-NH_4$, ${\delta}^{15}N-NO_3$) analysis results in the three inflow tributaries (Gulpocheon (GP), Anyangcheon (AY) and Sincheon (SC)) of the Han River. Water quality was analyzed in June and October from 2013 to 2017, and the results showed that the concentrations of nutrients, such as T-N, $NO_3-N$, and T-P, were increased at GP4, AY3, SC3, and SC4, which lie downstream of sewage treatment facilities. The results of ${\delta}^{15}N$ for June 2017 indicated that the source of nitrogen was sewage or livestock excreta at GP4 and SC4, and organic fertilizers at AY3 and SC3. ${\delta}^{15}N-NO_3$ results suggested that the source of nitrogen was related to organic sewage, livestock or manure at GP4, AY3 and SC4. Therefore, GP4 and SC4 were more influenced by effluent from sewage treatment facilities than by their tributaries, AY3 and SC3 were considered to be influenced more by their tributary than effluent from sewage treatment facilities. With the results of this study, the source of contamination (sewage treatment facility effluent) of river inflow downstream of Han River could be confirmed using water quality and stable isotope ratio.

Keywords

1. Introduction

Urbanization, and the associated concentrated population, has increased pollutant emissions and water pollution in the inflow streams to the downstream reaches of the Han River. Despite the expansion of sewage treatment facilities, river water pollution level is high in the dry season when the flow rate isinsufficient and there are stagnant sections. Urbanrivers exposed to non-point source pollution showdrastic changes in water quality even in the samestream, depending on the inflow of pollutants, changes in flow velocity, and shape. Factors such as industrial wastewater, livestock manure, fertilizer, and pesticides are aggravating water pollution in rivers.1 Furthermore, water quality deteriorates due to an increased inflow of point contamination sources, such as sewage and wastewater, and decreased discharge.

As of 2018, there are 1,936 water quality monitoringnetworks to assess the quality of water and aquatice cosystems in public water bodies, including rivers and lakes. The collected data are the basis for evaluating water quality nationwide and establishing relevant policies. 2 In addition, the water quality and flow rates derived from the monitoring networks have been used in various studies.3 The pollutant discharge loadin the Anyangcheon River (AY4) was 25.2 % of the maximum BOD of 65.90 kg/d/km2, higher than that of the Gulpocheon (GP4) at 20.5 % and Sincheon (SC5) at 11.1 %. The GP4 showed 44.5 % of the maximum T-P of 9.409 kg/d/km2, higher than that of AY4 (7.3 %) and SC5 (2.4 %), due to the many metal processing companies along its reaches. Along the SC5, there are many textile and leather manufacturers that generate wastewater with high concentrations oforganic matter, and the pollutant discharge loadreached the maximum BOD of 7.17 kg/d/km2.3

Statistical analysis of water quality data has been widely used to evaluate water quality of varioussurface and fresh waterbodies. In particular, correlation analysis is useful for understanding the linear correlation between two variables.3,4,5 In addition, carbon stable isotope ratios (δ13C) and nitrogen stable isotope ratios (δ15N) are used to estimate the origin of organic matter in the water.6,7,9 Internally, organic matter is produced by algae, phytoplankton, and aquatic plants, while external sources include ground vegetation.6,8Analyzed inorganic carbon and nitrogen can becompared to previously measured compositions of phytoplankton and terrestrial plants; δ13C is widely used to identify the origin of organic matter.5-7,9,12,13C3 plant species such as wheat and rice among terrestrial plants show a δ13C range from -35 to-20 ‰,6,10 and δ13C of freshwater phytoplankton using dissolved inorganic carbon is a range from -40 to-20 ‰.6,11 Plants show a wide range between -5 and +2 ‰ of δ15N, soil show a wide range between +2 and +5 ‰, and nitrate from livestock manure show a wide range between +10 and +20 ‰.14,15 The range of δ15N-NO3 is a range from +5 to +25 ‰ in manureand a range from +4 to +19 ‰ in sewage. In terms of general water pollutants, the range of δ15N-NO3 is from -3 to +5 ‰ for inorganic matter such as synthetic fertilizers, Showing a wide range between +5 and +10 ‰ for mixed inorganic and organic matter, and a range from +10 to +20 ‰ for organic matter suchas livestock manure and excreta.16,17 The range of δ15N-NH4 and δ15N-NO3 in manure and sewage is arange from +10 to +35 ‰.18 Although several studies have investigated the origin of organic matter using stable isotope ratios in the Yeongsan River,9 therehas been insufficient research on contaminationsources of the Han River.19

The objective of this study is to investigate the physico-chemical characteristics of water quality and determine the contamination sources, i.e., effluence from sewage treatment facilities, to the lower reaches of the Han River. Stable isotope ratios (δ13C and δ15N of particulate matter, dissolved δ15N-NH4 and δ15N-NO3) are measured in selected upstream and downstream sites of effluence from sewage treatment facilities in the Gulpocheon (GP), Anyangcheon (AY), and Sincheon (SC) Rivers. In addition, correlation analysis is used to identify the relationship between water quality analysis items and stable isotope ratios.

 

2. Study Location and Method

 

2.1. Study sites

The Han River is 494 km long, from Geum daesan(1,418 m) in Taebaek-si, Gangwon-do to the estuary and consists of 920 tributaries, including 19 national rivers, 15 first-grade local rivers, and 886 second-grade local rivers.2

The study area site includes tributaries that have been identified as vulnerable to water pollution based on the water quality monitoring network and locations of dewatering outlets from large-scalese wage treatment facilities. Three major urban rivers were selected: the GP, AY, and SC. Four sites werechosen along the GP, which is characterized by a high density of metal-working process companies and intensive land use. Five sites were chosen along the AY, which has distributed livestock contaminationsources. Finally, five sites were chosen along the SC, which has many textile and leather manufacturing facilities that generate wastewater containing significantorganic matter (Fig. 1).

 

BGHHBN_2019_v32n2_65_f0001.png 이미지

Fig. 1. The location of study sites in the Han-River.

 

2.2. Methods

 

2.2.1. Data collection and water quality analysis

National monitoring data obtained from the Water Environment Information System of the Ministry of Environment was used in this study. Samples werecollected and measured 12 times/year (monthly) for the general water quality monitoring network and 36 times or more/year (average 8 days) for the total water quality monitoring network (Table 1). Waterquality data during June and October, before and after the rainy season, from 2013 to 2017 were compiled. Data included daily water temperature (WT), dissolved oxygen (DO), electric conductivity (EC), total organic carbon (TOC), biological oxygen demand (BOD5), chemical oxygen demand (COD), total nitrogen (T-N), total phosphorus (T-P), dissolved total nitrogen (DTN), ammonia nitrogen (NH3-N), nitrate nitrogen(NO3-N), phosphate-phosphorus (PO4-P), Chlorophyll-a (Chl-a), and suspended solids (SS). The sampling and analysis of the water samples collected in 2017 were performed in accordance with the water pollution process test standards in the same laboratory as the stable isotope analyses. WT, DO, EC, TOC, BOD5,COD, T-N, T-P, and SS; NH3-N, NO3-N, and PO4-P, which are eutrophication-causing substances in therivers, and Chl-a, produced as a result of eutrophication, were measured.

 

Table 1. Target areas for the characteristic analysis of water quality

BGHHBN_2019_v32n2_65_t0003.png 이미지

 

2.2.2. Measurements of stable isotope ratios

Samples for the analysis of stable isotope ratiosand Water quality metrics were collected from 6 L of surface water in June and October of 2017. Glassfilber fiter papers were heated for 5 h at 450 °C toremove organic matter of them. They were thenfiltered through glass fiber filter paper (GF/F) with adiameter of 25 mm and thickness of 0.7 μm. The filter paper sample for δ13C measurement was steamed with 12 N hydrochloric acid (HCl) for 24 h toremove inorganic carbon. The filter paper sample for δ15N was not pretreated with HCl. The prepared samples were freeze-dried at -80°C to remove water. The samples for δ15N-NH4 and δ15N-NO3 measurementwere prepared by freezing approximately 800 mL offiltrate passed through the GF/F and analyzed using the Kjeldahl-Dumas pretreatment method. All stable isotope ratios were measured with a stable isotoperatio mass spectrometer (continuous flow type, Vision-EA, Isoprime, UK). The stable isotope ratio was defined as the δ value of the isotope ratio difference between the standard and analytical samples and expressed as ‰:

\(\begin{aligned} &\delta(\%)=\left[\left(\mathrm{R}_{\text {sample }} / \mathrm{R}_{\text {standral }}\right)-1\right] \times 1000\\ &\text { where } \mathrm{R}=^{13} \mathrm{C}^{12} \mathrm{C} \text { or }^{15} \mathrm{N} /^{14} \mathrm{N} \end{aligned}\)

 

2.2.3. Correlation analysis

Correlation analysis has been widely used to evaluate the water quality of various surface and freshwaterbodies. 3-5 In this study, correlation analysis of each river sample’s quality metric was performed using the SPSS (ver. 14.0) program. Correlation analysisis useful for identifying linear correlations between two variables and understanding their relationship.4

 

3. Results and Discussion

 

3.1. Water quality metrics

Fig. 2 is a box-type graph showing water quality measurement data (minimum: 5 and maximum: 21) from each of the 14 study sites in June and October, before and after the rainy season, from 2013 to 2017. In addition, the samples collected for stable isotoperatio measurements were analyzed to obtain the standard water quality metrics and the data included in the 2013-2017 database. T-P concentrations ranged from 0.024 to 2.590 mg/L in all study sites and from 0.044 to 2.590 mg/L in GP samples, higher than that of other rivers. The mean DO concentration of GPsample was 6.8 mg/L, which was lower than the mean value of all sites (8.1 mg/L), and the T-P concentration was 0.442 mg/L, 1.8 times higher thanthe mean of all sites (0.248 mg/L). The low DO may have resulted from stagnated flow, while changes in the T-N concentrations of GP4 samples were likely influenced by effluents from the upstream se wagetreatment facility. The ammonium nitrogen (NH3-N) of AY sample was 2.748 mg/L, which was higher than the mean of all study sites (2.050 mg/L), and the mean NH3-N concentrations of GP and SC samples were 1.682 mg/L and 1.645 mg/L, respectively. Themean BOD of SC samples was 7.2 mg/L, higher thanthat of all study sites (5.6 mg/L), GP samples (6.4 mg/L), and AY samples (3.4 mg/L). These results are inagreement with the results of Choi et al.3, whoinvestigated the water pollution characteristics of the GP, where wastewater from metal-working companiesis introduced, and SC, where wastewater from textileand leather manufacturers is introduced. Differences between the 2013-2017 monitoring data and additional 2017 data from the samples collected for stable isotopeanalyses may have been due to different sampling times.

 

BGHHBN_2019_v32n2_65_f0002.png 이미지

BGHHBN_2019_v32n2_65_f0003.png 이미지

Fig. 2. Water quality variations of each sites in June and October, 2013~2017. (a) WT, (b) DO, (c) EC, (d) TOC, (e) BOD, (f) COD, (g) T-N, (h) NH3-N, (i) T-P, (j) PO4-P, (k) Chl-a, (l) SS, (m) DTN, (n) NO3-N

 

3.2. Stable isotope ratios

The range of δ13C of GP samples decreased from a wide range between -27.25 and -21.83 ‰ in June to a wide range between -30.47 and -26.31 ‰ in October, while that of δ15N decreased from a widerange between +0.94 and +10.24 ‰ in June to a wide range between -4.11 and +1.30 ‰ in October. The range in δ13C of AY samples increased from a wide range between -27.76 and -25.73 ‰ in June to a wide range between -26.46 and -25.28 ‰ in October, while the range of δ15N decreased from a wide range between +5.31 and +13.37 ‰ in June to a wide range between +2.58 and +5.73 ‰ in October. The range of δ13C of SC increased from a wide range between-30.86 and -28.78 ‰ in June to a wide range between-28.57 and -24.48 ‰ in October, and the range of δ15N decreased from a wide range between +4.77 and +14.74 ‰ in June to a wide range between +2.42 and +6.40 ‰ in October. Except for somesites, δ13C increased while δ15N decreased in October compared to that in June (Fig. 3(a)).

 

BGHHBN_2019_v32n2_65_f0004.png 이미지

Fig. 3. Comparison of δ13C and δ15N values at sampling sites in June and October, 2017

Notable differences are observed between June and October when comparing δ15N of upstream and downstream samples along the GP (GP3, GP4) based on effluence from domestic and sanitary se wagetreatment facilities. Upstream, soils were the dominantsources, while downstream domestic sewage orlivestock excreta were dominant sources in June. However, in October, while plants were the dominantsource upstream and soils were dominant downstream (Fig. 3(b)). That is, sewage treatment facility effluents was more important in June than October. A similarchange is observed in δ15N from samples upstreamand downstream (AY2, AY3) along the AY River in June and October. In June, domestic sewage orlivestock excreta was dominant upstream, while organic fertilizer was dominant downstream. In October, bothupstream and downstream sources were predominantlysoil (Fig. 3(c)). In June, the influence of tributaries was greater than that of sewage treatment effluents, and the influence of tributaries and effluents from sewage treatment facilities was small in October. In contrast, comparing δ15N of upstream and downstreamsamples (SC2, SC3) along the SC, organic fertilizer was the primary source in June with no significantspatial change, and soil dominated in October with no significant spatial change (Fig. 3(d)). Comparisons of δ15N-NO3 from the downstream sample (SC3) along the SC revealed that the inorganic matter was primarily from synthetic fertilizer and fertile soil (from -3 to +5 ‰),20 with a value of -1.13 ‰ (Fig. 4). The influence of tributaries appeared greater thanthat of effluent from sewage treatment facilities, which is in agreement with the δ15N results.

In the AY, there was a significant δ15N-NH4 decreaseand a δ15N-NO3 increase from AY1 to AY2 (Fig. 4), which we hypothesize was due to nitrification in which ammonia nitrogen was oxidized to nitrous acid and nitrate nitrogen by microbial action. However, because δ15N can be influenced by mixing of various NO3− sources and various processes, i.e., nitrification, denitrification, and assimilation, the entire N cycleshould be considered when identifying NO3− sources. 20,21

 

BGHHBN_2019_v32n2_65_f0005.png 이미지

Fig. 4. δ15N Comparison of NO3 and NH4 values at sampling sites in June and October, 2017.

Differences in δ15N values (Δδ15N) were comparedupstream and downstream of effluent from se wagetreatment facilities at each site and in both study months. The Δδ15N of AY2 and AY3 was 7.22 ‰ in June and 1.74 ‰ in October, while the Δδ15N of SC3 and SC4 was -8.70 ‰ in June and -0.57 ‰ in October. These results may have been influenced by otherinflux sources, such as tributaries, between sites. The absolute values of Δδ15N and Δδ15N-NH4 of GP3 and GP4, Δδ15N-NO3 of AY2 and AY3, Δδ13C of SC2 and SC3, and Δδ15N-NH4 of SC3 and SC4 weresimilar in June and October (Table 2). The similarity of the absolute values of the stable isotopes in June and October suggests that there was a constant influx source in each river; however, to identify the source of pollution more precisely, surveys of each pollutionsource should be conducted.

 

Table 2. Difference stable isotope ratios (upstream-downstream) of each sites in June and October, 2017

BGHHBN_2019_v32n2_65_t0002.png 이미지

 

3.3. Changes in water quality and stable isotope ratios at each site

Effluent from sewage treatment facilities is a keynitrogen source to river waters and is characterized by high nitrogen isotope ratios.16 In the periods of heavy rainfall, ammonia nitrogen can enter the riversystem. In comparison, phosphate sources includesoil, sewage, and agricultural fertilizers.9

  

Fig. 5 shows changes in T-N and T-P; NH3-N, NO3-N, and PO4-P, which cause river eutrophication, and stable isotope ratios in June and October at each site upstream and downstream of sewage treatment facilities. The T-N and NO3-N concentrations and δ15N increased, but T-P and NH3-N concentrations and δ13C decreased in June and October upstream (GP3) and downstream (GP4) of the sewage treatment facility (facility capacity: 900,000 m3/day) along the GP4. There are two sewage treatment facilities (facility capacity: 300,000 m3/day, 150,000 m3/day) along the AY. T-N, NH3-N, NO3-N, T-P, and PO4-P concentrations and δ13C increased in June and October in bothupstream (AY2) and downstream (AY3) samples. Along the SC, upstream (SC2) and downstream (SC3) of these wage treatment facility (facility capacity: 700,000m3/day) had higher T-N, NH3-N, NO3-N and δ15N-NH4 concentrations and δ13C and lower T-P and PO4-P concentrations, a likely result of an influx of contamination sources containing nitrogen and phosphorus. Further down the SC, upstream (SC3) and downstream (SC4) of the sewage treatment facility (facility capacity: 86,000 m3/day) the T-Nand NH3-N concentrations and δ13C decreased but NO3-N concentrations and δ15N, δ15N-NH4, and δ15N-NO3 increased in June. The difference betweencarbon and nitrogen stable isotope ratios indicates differences between the survey period, sites, and sources of influx.9 GP4 and SC4 samples seemed to have been influenced by the effluent from se wagetreatment facilities, while the influence of agricultural fertilizer seemed to be small. Based on the stable isotope analysis results, AY3 and SC3 were influenced by an influx of diverse contamination sources containingnitrogen and phosphorus.

 

BGHHBN_2019_v32n2_65_f0006.png 이미지

Fig. 5. Water quality and stable isotope variations of each sites in June and October, 2017.

 

3.4. Correlation between water quality variables and stable isotope ratios

Table 3 shows the results of the correlation analysis between water quality variables and stable isotoperatios from GP, AY, and SC samples from the lowerreaches of the Han River in June and October of 2017. Correlation coefficients between COD and BOD and COD and TOC, which are indirect indices of organic matter, were 0.934 and 0.957 (n=28, p< 0.01), respectively, showing a strong positive correlation. This strong correlation is in agreement with the correlation analysis of Gwak et al.4,22. The correlation coefficient between BOD and Chl-a is 0.913 (n=28, p< 0.01), showing a strong positive correlation, indicating that organic matter may have increased Chl-a, a productof eutrophication. The correlation coefficient between NH3-N and δ15N-NO3 is -0.595 (n=28, p<0.01), showing a negative correlation overall, which may have beeninfluenced by nitrification.20,21 Table 4 shows the results of correlation analysis of differences in water quality and stable isotope ratios (upstream-downstream); the correlation coefficient between ΔCOD and ΔChl-a is 0.722 (n=22, p<0.01). These results are similar to those in Table 3 and previous studies,4,22 indicating that the amount of change in upstream and downstreamsamples for each variable was linked. The ΔT-P and δ Chl-a are positively correlated, with a correlationvalue of 0.603 (n=22, p<0.01), revealing that T-P from nutrient salts and Chl-a, were related. The correlation between Δδ13C and ΔTOC is 0.613 (n=22, p< 0.01), showing a positive correlation. The correlation coefficients between ΔNH3-N and Δδ15N and between δ NH3-N and Δδ15N-NO3 are -0.617 and -0.731 (n=22, p< 0.01), respectively, showing a negative correlation, which may have been related to the nitrogen cycle.

The results in Table 3 indicate no correlation between phosphates and nitrogenous nutrients, suggesting that their origins were different.5 The correlation analysis of the upstream-downstream difference shows a correlation between ΔT-P and ΔChl-a and between ΔTOC and δδ13C, suggesting that their sources were the same, which resulted in the similar changes. Therefore, additional research should be conducted that surveysthe contamination sources.

 

Table 3. Pearson correlation coefficient among the water quality and stable isotope ratio parameters, 2017.

BGHHBN_2019_v32n2_65_t0004.png 이미지

 

Table 4. Pearson correlation coefficient among the difference(upstream-downstream) water quality and stable isotope ratio parameters, 2017

BGHHBN_2019_v32n2_65_t0005.png 이미지

 

4. Conclusions

This study confirmed the influence of effluents from the sewage treatment facilities using water quality metrics and stable isotope ratios of rivers flowinginto the lower reaches of the Han River. Waterquality data downstream from 2013 to 2017 indicated that the T-N and NO3-N in GP4 samples, T-N, T-P, NH3-N, NO3-N and PO4-P in AY3 samples, T-N, NH3-N and NO3-N in SC3 samples, and NO3-N in SC4 samples downstream of the sewage treatment facility were higher than those upstream in June and October. δ15N in the upstream and downstreameffluents from sewage treatment facilities in 2017 indicated that GP (GP3 and GP4) and SC (SC3 and SC4) samples were influenced by effluent from se wagetreatment facilities. In contrast, AY (AY2 and AY3) and SC (SC2 and SC3) samples were more affected by tributaries than by effluent from sewage treatment facilities. Samples collected in October indicated that the GP (GP4) was influenced by soil sources, and the effects of sewage treatment facility effluent and tributaries were small in the AY (AY3) and SC (SC3 and SC4). Correlation analyses between water quality and stable isotope ratios in 2017, revealed coefficients between COD and BOD and COD and TOC (whichare indirect indices of organic matter) as 0.934 and 0.957 (n=28, p<0.01), respectively. In addition, the correlation coefficient between BOD and Chl-a was 0.913 (n=28, p<0.01), showing a strong positive correlation. This suggests that organic matter affected the increase in Chl-a. Correlation analysis of differences due to location (upstream-downstream) showed apositive correlation between Δδ13C and ΔTOC with the value of 0.613 (n=22, p<0.01). This shows that changes in organic matter and carbon stable isotoperatios were related. These results indicate that periodicsurveys should be performed in conjunction with surveys of precipitation, flow rate and pollutant source to more clearly determine contaminant sources torivers flowing into the lower reaches of the Han River.

 

Acknowledgments

This study was supported by the National Institute of Environmental Research with financial support from the Ministry of Environment (NIER-2018-03-03-002).

References

  1. Ministry of Environment (MOE), Han River Water Management Committee, 'A study of the investigation and improvement on the water quality deteriorating area in urban stream of Han River', 2017.
  2. Ministry of Environment (MOE), National Institute of Environment Research (NIER), http://water.nier.go.kr/publicMain/mainContent.do, accessed sep. 2018.
  3. O. Y. Choi, K. H. Kim, and I. S. Han, J. Korean Soc. Water Environ., 31, 700-714 (2015). https://doi.org/10.15681/KSWE.2015.31.6.700
  4. B. R. Gwak and I. K. Kim, J. Korean Soc. Environ. Eng., 38, 71-78 (2016). https://doi.org/10.4491/KSEE.2016.38.2.71
  5. J. K. Gal, M. S. Kim, Y. J. Lee, J. W. Seo, and K. H. Shin, Korean J. Limnol., 45, 242-251 (2012).
  6. M. S. Kim, J. M. Kim, J. Y. Hwang, B. K. Kim, H. S. Cho, S. J. Youn, S. Y. Hong, O. S. Kwon, and W. S. Lee, KJEE, 47(2), 127-134 (2014). https://doi.org/10.11614/KSL.2014.47.2.127
  7. M. S. Kim, Y. J. Lee, K. G. An, B. H. Kim, S. J. Hwang, and K. H. Shin, KJEE, 47, 53-61 (2014). https://doi.org/10.11614/KSL.2014.47.1.053
  8. B. Fry and E. B. Sherr, '${\delta}^{13}C$ measurements as indicators of carbon flow in marine and freshwater ecosystems', Contrib. Mar. Sci., 27, 196-229 (1984).
  9. Y. J, Lee, B. K. Jeong, Y. S. Shin, S. H. Kim, and K. H. Shin, KJEE, 46, 175-184 (2013). https://doi.org/10.11614/KSL.2013.46.2.175
  10. J. R. Ehleringer, 'Carbon isotope ratios and physiololgical processes in Aridland plants', p14-54, Springer-Ver-lag, New York, 1989.
  11. R. I. France, Marine Ecology Progress Series, 124, 307-312 (1995). https://doi.org/10.3354/meps124307
  12. P. A. Meyers, Organic Geochemistry, 27, 213-250 (1997). https://doi.org/10.1016/S0146-6380(97)00049-1
  13. J. E. Tesdal, E. D. Gallbraith, and M. Kienast, Biogeosciences, 10, 101-118 (2013). https://doi.org/10.5194/bg-10-101-2013
  14. B. Fry, Ecology, 72, 2293-2297 (1991). https://doi.org/10.2307/1941580
  15. C. Kendall, 'In Isotope Tracers in Catchment Hydrology', pp. 519-576, Elsevier: New York, 1998.
  16. NIER, 'Studies on the origin of nitrogen sources from watersheds (I)', pp. 17-34, 2015.
  17. D. Xue, J. Botte, B. D. Baets, F. Accoe, A. Nestler, P. Taylor, O. V. Cheemput, M. Berglund, and P. Boeckx., Water Res., 43, 1159-1170 (2009). https://doi.org/10.1016/j.watres.2008.12.048
  18. NIER, Environmental Measurement & Analysis Center, 'A review : stable isotope analysis of environmental pollutants', 2017.
  19. J. K. Kim, B. C. Kim, S. M. Jung, C. G. Jang, M. S. Shin, and Y. Y. Lee, Contrib. Mar. Sci., 40, 93-102 (2007).
  20. J. Vrzel, B. Vukovic-Gacic, S. Kolarevic, Z. Gacic, M. Kracun-Kolarevic, J. Kostic, M. Aborgiba, A. Farnleitner, G. Reischer, R. Linke, M. Paunovic, and N. Ogric, Sci. Total. Environ., 573, 1460-1471 (2016). https://doi.org/10.1016/j.scitotenv.2016.07.213
  21. W. J. Choi, J. H. Kwak, S. S. Lim, H. J. Park, S. X. Chang, S. M. Lee, M. A. Arshad, S. I. Yun, and H. Y. Kim, Agric. Ecosyst. Environ., 237, 1-15 (2017). https://doi.org/10.1016/j.agee.2016.12.020
  22. E. S. Song, S. M. Jeon, E. J. Lee, D. J. Park, and Y. S. Shin, Korean J. Limnol., 45, 302-313 (2012).