Title Page
Contents
ABSTRACT 16
Chapter 1. Introduction 18
1.1. Background and research necessity 18
1.2. Objectives and scope 21
1.3. Study procedure 23
Chapter 2. Literature Review 25
2.1. DPSIR Framework 25
2.2. Anthropogenic human activities 28
2.3. Species distribution model 31
2.4. Stream aquatic ecosystem health 33
Chapter 3. Materials and Methods 37
3.1. DPSIR Framework for stream aquatic ecosystem health 37
3.2. N, P emission from anthropogenic human activities 39
3.3. SWAT (Soil and Water Assessment Tool) 47
3.4. SDM (Species Distribution Model) 51
3.5. T-AEHI (Total Aquatic Ecosystem Health Index) 61
3.6. Response action for stream aquatic ecosystem health 62
Chapter 4. Results and Discussion 65
4.1. Analysis of anthropogenic human activities 65
4.1.1. Anthropogenic nitrogen, phosphorus emission 65
4.1.2. Analysis of characteristic for anthropogenic nitrogen and phosphorus 71
4.2. Results of SWAT 78
4.2.1. SWAT calibration and validation 78
4.2.2. Analysis of SWAT hydrological and water quality results 93
4.3. Results of SDMs 97
4.3.1. Results of SDM for representative species of fish 97
4.3.2. Results of SDM for representative species of trophic diatom 104
4.3.3. Results of SDM for representative species of benthic macro-invertebrate 113
4.4. Evaluation of T-AEHI 121
4.4.1. Select species to develop T-AEHI 121
4.4.2. Comparison previously used index (FAI, TDI, and BMI) 125
4.4.3. Analysis of spatial distribution of T-AEHI 128
4.5. Response to restore aquatic ecosystem health 134
4.5.1. Setting management target 134
4.5.2. Evaluation T-AEHI change based on response actions 138
Chapter 5. Conclusion 145
5.1. Summary and conclusions 145
5.2. Recommendations for future work 147
References 149
Appendices 162
[Appendix A] Statistic Data for human activities 162
[Appendix B] SDMs results for occurrence probability by year 169
Abstract (in Korean) 192
Table 2-1. Review of categorized variables in DPSIR framework 26
Table 2-2. The AEH (Aquatic Ecosystem Health) Indices assessment criteria based on score (0~100) 34
Table 3-1. Example data of Fertilizer use in 16 cities and provinces (2019) 40
Table 3-2. Example data of raised livestock head in 16 cities and provinces 42
Table 3-3. Livestock manure generation load unit 42
Table 3-4. Characteristics of selected representative fish species 54
Table 3-5. Characteristics of selected representative trophic diatom species 55
Table 3-6. Characteristics of selected representative benthic macro-invertebrate species 55
Table 3-7. Environmental variables used for SDMs in this study 57
Table 3-8. Representative metrics calculated from the confusion matrix 60
Table 3-9. Definition of anthropogenic human activity regulation scenarios. 62
Table 3-10. Definition of BMPs scenarios and SWAT implementatioin criteria. 64
Table 4-1. Summary of total nitrogen, total phosphorus emission per unit area from anthropogenic human activities in the major tributary watersheds in ND River 72
Table 4-2. summary of landuse change in major tributary watershed and ND watershed 74
Table 4-3. Calibrated parameters for the SWAT 80
Table 4-4. Statistical summary of precipitation, streamflow, and runoff ratio for calibration and validation periods 86
Table 4-5. Statistical results of calibration and validation for streamflow and inflow of hydraulic structures 87
Table 4-6. Statistical results of calibration and validation for water quality (SS, TN, TP loads) at 8 water quality monitoring stations 92
Table 4-7. Summary of water quality load per unit area in the major tributary watershed and ND watershed 96
Table 4-8. Performance evaluation of SDMs for 6 fish species 97
Table 4-9. Summary of occurrence probability of 6 fish species by stream order 101
Table 4-10. Performance evaluation of SDMs for 9 trophic diatom species 104
Table 4-11. Summary of occurrence probability of 9 trophic diatom species by stream order 112
Table 4-12. Performance evaluation of SDMs for 8 benthic macro-invertebrate species 113
Table 4-13. Summary of occurrence probability of 8 benthic macro-invertebrate species by stream order 118
Table 4-14. Statistical results of T-AEHI and each AEH by stream order 126
Table 4-15. Annual average T-AEHI by stream order (2008~2019) 128
Table 4-16. Analysis of change on average T-AEHI in major tributary watershed and ND river watershed (2008~2019) 130
Table 4-17. Summary of T-AEHI in Watershed with good and poor aqua-ecological status by year 135
Table 4-18. Comparison of water quality characteristic in tributary watersheds with good and poor ecological status 137
Table 4-19. Analysis of water quality change under human activities regulation scenarios 138
Table 4-20. Analysis of water quality change under BMPs scenarios 142
Table 4-21. Analysis of T-AEHI change under BMPs scenarios by stream order 143
Table A-1. Amount of generated nitrogen wastewater from livestock manure during 1999~2019 163
Figure 1-1. Description of study area. 22
Figure 1-2. Flowchart of the study. 24
Figure 2-1. The DPSIR framework. 25
Figure 2-2. Conceptual diagram for calculating anthropogenic nitrogen and phosphorus from human activities. 30
Figure 2-3. Process of species distribution models. 32
Figure 2-4.. Representative species for grade A to D of FAI, TDI and BMI. 35
Figure 2-5. Correlation of the three indices (FAI, BMI, and TDI) by year in (a) spring, and (b) fall. 36
Figure 3-1. DPSIR Framework in this study. 38
Figure 3-2. schematic illustration for the impact of anthropogenic human activities on watershed. 39
Figure 3-3. Livestock manure treatment process. 43
Figure 3-4. Location of sewage treatment plants according to capacity. 44
Figure 3-5. Location of acid deposition monitoring stations in South Korea. 46
Figure 3-6. GIS data for SWAT (a) DEM, (b) Soil type, (c) Land use and (d) location of monitoring stations. 48
Figure 3-7. Example of converting monitoring location data to species distribution data classified presence, transition, and absence. 51
Figure 3-8. Population of observed fish species in study area during 2008-2020. 52
Figure 3-9. Population density of observed trophic diatom species in study area during 2008-2020. 53
Figure 3-10. Population density of observed benthic macro-invertebrate species in study area during 2008-2020. 54
Figure 3-11. Random Forest regression algorithm 59
Figure 3-12. General confusion matrix of classification. 59
Figure 4-1. Distribution map for anthropogenic nitrogen emission from fertilizer use during (a) early 2000s, (b) late 2000s, (c) early 2010s, and (d) late 2010s. 66
Figure 4-2. Distribution map for anthropogenic nitrogen emission from livestock manure during (a) early 2000s, (b) late 2000s, (c) early 2010s, and (d) late 2010s. 67
Figure 4-3. Distribution map for anthropogenic nitrogen emission from point source pollution during (a) early 2000s, (b) late 2000s, (c) early 2010s, and (d) late 2010s. 67
Figure 4-4. Distribution map for anthropogenic nitrogen emission from atmospheric nitrogen deposition during (a) early 2000s, (b) late 2000s, (c) early 2010s,... 67
Figure 4-5. Analysis of anthropogenic total nitrogen emission change trend. 68
Figure 4-6. Distribution map for anthropogenic phosphorus emission from fertilizer use during (a) early 2000s, (b) late 2000s, (c) early 2010s, and (d) late 2010s. 69
Figure 4-7. Distribution map for anthropogenic phosphorus emission from livestock manure during (a) early 2000s, (b) late 2000s, (c) early 2010s, and (d) late 2010s. 69
Figure 4-8. Distribution map for anthropogenic phosphorus emission from point source pollution during (a) early 2000s, (b) late 2000s, (c) early 2010s, and (d) late 2010s. 69
Figure 4-9. Analysis of anthropogenic total nitrogen emission change trend. 70
Figure 4-10. Riverine export relative to (a) anthropogenic TN and (b) TP 75
Figure 4-11. Riverine N export relative to anthropogenic TN in (a) NSC, (b) WC, (c) GC, (d) GH, (e) MY, (f) HG, and (g) NG watershed 76
Figure 4-12. Riverine P export relative to anthropogenic TP in (a) NSC, (b) WC, (c) GC, (d) GH, (e) MY, (f) HG, and (g) NG watershed 77
Figure 4-13. Comparison of the observed daily streamflow with SWAT-simulated streamflow during the calibration (2010-2014) and validation (2015-2020) periods at... 83
Figure 4-14. Comparison of the observed daily dam inflow with SWAT-simulated dam inflow during the calibration (2010-2014) and validation (2015-2020) periods... 84
Figure 4-15. Comparison of the observed daily weir inflow with SWAT-simulated weir inflow during the calibration (2012-2015) and validation (2016-2020) periods... 85
Figure 4-16. Comparison of the observed monthly SS loads with SWAT-simulated SS loads during the calibration (2010-2014) and validation (2015-2020) periods... 89
Figure 4-17. Comparison of the observed monthly TN loads with SWAT-simulated TN loads during the calibration (2010-2014) and validation (2015-2020) periods... 90
Figure 4-18. Comparison of the observed monthly TP loads with SWAT-simulated TP loads during the calibration (2010-2014) and validation (2015-2020) periods... 91
Figure 4-19. Analysis of monthly average total runoff of major tributary watersheds (NSC, WC, GC, GH, MY, HG, and NG) in Nak-dong river basin during 2010~2020. 93
Figure 4-20. Comparison of streamflow, low flow duration, and extreme flow events at (a) NSC, (b) WC, (c) GC, (d) GH, (e) MY, (f) HG, and (g) NG. 95
Figure 4-21. Analysis of SDM for Rhinogobius brunneus (a) variable importance and response curve of (b) 1st important variable (LU_200) and (c) 2nd important... 98
Figure 4-22. Analysis of SDM for Coreoleuciscus splendidus (a) variable importance and response curve of (b) 1st important variable (Drainage Area) and (c) 2nd... 99
Figure 4-23. Analysis of SDM for Zacco koreanus (a) variable importance and response curve of (b) 1st important variable (LU_100) and (c) 2nd important... 99
Figure 4-24. Analysis of SDM for Rhynchocypris oxycephalus (a) variable importance and response curve of (b) 1st important variable (Flow) and (c) 2nd... 99
Figure 4-25. Analysis of SDM for Zacco platypus (a) variable importance and response curve of (b) 1st important variable (Drainage Area) and (c) 2nd important... 100
Figure 4-26. Analysis of SDM for Carassius auratus (a) variable importance and response curve of (b) 1st important variable (LU_100) and (c) 2nd important... 100
Figure 4-27. Occurrence probability distribution map of (a) Rhinogobius brunneus, (b) Coreoleuciscus splendidus, (c) Zacco koreanus, (d) Rhynchocypris oxycephalus,... 102
Figure 4-28. Analysis of SDM for Achnanthes minutissima (a) variable importance and response curve of (b) 1st important variable (LU_200) and (c) 2nd important... 106
Figure 4-29. Analysis of SDM for Achnanthes convergens (a) variable importance and response curve of (b) 1st important variable (LU_100) and (c) 2nd important... 106
Figure 4-30. Analysis of SDM for Nitzschia inconspicua (a) variable importance and response curve of (b) 1st important variable (LU_100) and (c) 2nd important... 106
Figure 4-31. Analysis of SDM for Nitzschia fonticola (a) variable importance and response curve of (b) 1st important variable (TOP) and (c) 2nd important variable (LowD). 107
Figure 4-32. Analysis of SDM for Gomphonema clevei (a) variable importance and response curve of (b) 1st important variable (LU_100) and (c) 2nd important... 107
Figure 4-33. Analysis of SDM for Cocconeis placentula var. lineata (a) variable importance and response curve of (b) 1st important variable (FLOW) and (c) 2nd...[이미지참조] 107
Figure 4-34. Analysis of SDM for Fragilaria construens f. binodis (a) variable importance and response curve of (b) 1st important variable (HighE) and (c) 2nd... 108
Figure 4-35. Analysis of SDM for Fragilaria elliptica (a) variable importance and response curve of (b) 1st important variable (LU_200) and (c) 2nd important... 108
Figure 4-36. Analysis of SDM for Cymbella affinis (a) variable importance and response curve of (b) 1st important variable (NH₄) and (c) 2nd important variable (SS).[이미지참조] 108
Figure 4-37. Occurrence probability distribution map of (a) Achnanthes minutissima, (b) Achnanthes convergens, (c) Nitzschia inconspicua, (d) Nitzschia fonticola, (e)... 110
Figure 4-38. Analysis of SDM for Ecdyonurus kibunensis (a) variable importance and response curve of (b) 1st important variable (LU_100) and (c) 2nd important variable (LU_200).[이미지참조] 114
Figure 4-39. Analysis of SDM for Serratella setigera (a) variable importance and response curve of (b) 1st important variable (LU_100) and (c) 2nd important variable (DrainA).[이미지참조] 115
Figure 4-40. Analysis of SDM for Ecdyonurus levis (a) variable importance and response curve of (b) 1st important variable (LU_200) and (c) 2nd important variable (LU_100).[이미지참조] 115
Figure 4-41. Analysis of SDM for Hydropsyche kozhantschikovi (a) variable importance and response curve of (b) 1st important variable (DrainA) and (c) 2nd...[이미지참조] 115
Figure 4-42. Analysis of SDM for Cheumatopsyche brevilineata (a) variable importance and response curve of (b) 1st important variable (NH₄) and (c) 2nd important...[이미지참조] 116
Figure 4-43. Analysis of SDM for Asellus sp. (a) variable importance and response curve of (b) 1st important variable (LU_200) and (c) 2nd important variable (LU_100).[이미지참조] 116
Figure 4-44. Analysis of SDM for Baetis fuscatus (a) variable importance and response curve of (b) 1st important variable (LU_100) and (c) 2nd important... 116
Figure 4-45. Analysis of SDM for Chironomidae spp. (red type) (a) variable importance and response curve of (b) 1st important variable (LowD) and (c) 2nd... 117
Figure 4-46. Occurrence probability distribution map of (a) Ecdyonurus kibunensis, (b) Serratella setigera, (c) Ecdyonurus levis, (d) Hydropsyche kozhantschikovi, (e)... 119
Figure 4-47. Comparison for occurrence probability of (a) Rhinogobius brunneus, (b) Coreoleuciscus splendidus, (c) Zacco koreanus, (d) Rhynchocypris oxycephalus, (e)... 122
Figure 4-48. Comparison for occurrence probability (a) Achnanthes minutissima, (b) Achnanthes convergens, (c) Nitzschia inconspicua, (d) Nitzschia fonticola, (e)... 123
Figure 4-49. Comparison for occurrence probability (a) Ecdyonurus kibunensis, (b) Serratella setigera, (c) Ecdyonurus levis, (d) Hydropsyche kozhantschikovi,... 124
Figure 4-50. Comparison T-AEHI with (a)FAI, (b)BMI and (c)TDI. 126
Figure 4-51. Comparison T-AEHI with Euclidean distance of AEH using 3 dimensions scatter plot. 127
Figure 4-52. Boxplot of T-AEHI in major tributary watershed and ND river 129
Figure 4-53. Distribution map of T-AEHI during 2008~2019. 133
Figure 4-54. Comparison of occurrence probability in tributary watersheds with good and poor ecological status. 136
Figure 4-55. Change on species occurrence probability of (a) fish, (b) trophic diatom, and (c) benthic macro-invertebrate under human activities regulation scenarios. 140
Figure 4-56. Distribution map of T-AEHI difference under human activities regulation scenarios. 141
Figure 4-57. Comparison of T-AEHI in major tributary watersheds under human activities regulation scenarios. 144
Figure 4-58. Distribution map of T-AEHI difference under BMPs scenarios 144