1. October 5, 2005, The 4th IAHR Symposium on River, Coastal and Estuarine Morphodynamics Field Observation and WEPP Application for Sediment Yield in an Agricultural Watershed Kazutoshi Osawa & Syunsuke Ikeda Tokyo Institute of Technology, Tokyo, Japan. Satoshi Yamaguchi Ministry of Land, Infrastructure and Transport, Japan

2. Location of the Okinawa region Okinawa region Location of the Okinawa region in Japan Tokyo 1000km Ishigaki Island 1000km INTRODUCTION

3. Recently many farm land reclamation projects are carried out. Land use is changed dramatically from forest to upland fields. Red-soil runoff problems in Okinawa region Soil erosion is accelerated heavily  Subtropical climate (high intensity of rainfall)  Red soil (high erodibility)  Steep slope ( ~14%) Upland fields and bare lands become the main source of sediment Introduction

4. Red soil runoff problems in Okinawa region Heavy soil erosion occurs Marine pollution  Disruption of the oceanic ecosystems (coral, fishes, etc.) Flow into channels and the rivers  Very short river length (5-10km)  Discharged into the sea directly Dead corals in near of the mouth of the river Corals barely live in Nagura Bay, Ishigaki Island Corals in the unpolluted water Corals have been damaged by oversupply of sediment and nutrients from the river basin including agricultural zone.

5. Objectives Soil erosion plot test was carried out to compare the amounts of sediment yield depending on the difference of agricultural management Water Erosion Prediction Project (WEPP) model was applied to these test fields in predicting sediment yield at farmland and simulated sediment runoff in watershed Introduction

6. Field plot test Outlines St-1: Non-cultivating (control) St-2: Spring sugarcane by conventional tillage St-3: Spring sugarcane with grass strip St-4: Perennial sugarcane by zero-tillage farming Canopy cover effect Grass strip effect Zero-tillage effect

7. Temporal variations of measured parameters only put the data during each rainfall event

8. Canopy cover effect 59 % Zero-tillage effect (vs. non-cultivating) 94 % Zero-tillage effect (vs. spring sugarcane) 85 % Grass strip effect 8 % Sediment runoff reduction ratio Amount of sediment yields 59% 94% 85% 8%

9. Water Erosion Prediction Project (WEPP) Constructed by Nearing et al. in 1989 as hillslope erosion model In 1995, the model was expanded to the watershed scale Physically-based model Erosion, climate, hydrology, daily water balance, plant growth, residue decomposition, etc. Different from the USLE, the WEPP model was constructed for the purpose of estimating soil loss at every rainfall event Constructed as a post-USLE, however, little study has been done to apply the WEPP model to Japan

10. Plant growth Infiltration Soil condition conductivity conductivity erodibility erodibility Percolation Overland flow Climate 1. Climate 2. Overland flow 3. Water balance 4. Plant growth 5. Soil condition 6. Managements 7. Erosion Managements Tillage, Plant seedlings, Harvest, etc. Evapo- transpiration Hillslope components of WEPP model Rill erosion Interrill erosion

11. Sediment continuity equation G: sediment load, D i : interrill erosion rate, D f : rill erosion rate K i : interill erodibility, I e : effective rainfall intensity, σ ir : interrill runoff rate, SDR rr : sediment delivery ratio F nozzle : adjustment factor for sprinkler irrigation, R s : rill spacing, W: rill width Interill erosion T c : transport capacity of flow in the rill τ f : flow shear stress acting on soil particles τ c : critical shear stress of the soil K r : rill erodibility β : raindrop-induced turbulence coefficient V f : effective fall velocity q ir : flow discharge. Rill erosion and deposition Erosion process of WEPP model erosion deposition

12. Watershed scale Hillslope Channel Impoundment deposition transportation erosion Interrill erosion Rill erosion Plant Management Soil Hillslope scale Watershed components of WEPP model transportation deposition erosion

13. WEPP model verification Most of calculated results were agree with observed ones. Disagreements of discharge at St-4 would be attributed to the overestimate of the hydraulic conductivity. Differences of sediment runoff at St-1 can be attributed to the growth of weeds and the loss of fine and easily erodible sediment at the actual plot. If these conditions were expressed properly with the model, these gaps will be improved.

14. Simulated results by WEPP model spring-perennial sugarcane cycle was more effective cropping method in view of sediment yield reduction than summer sugarcane. The measures of residue mulch and no-tillage planting at summer sugarcane field reduced sediment yield more effective than that at spring-perennial sugarcane field.

15. Outlines and land use of Kandabara basin Farmlands occupied most of the basin. Sugarcane: 49% (Summer: 38%, Spring: 4%, Perennial: 6%) Pineapple: 2%, Paddy: 24%, Grassland: 14%

16. Calculated sediment yield and discharge The feasible combinations of sediment yield reduction methods: (1)shifting land use of summer sugarcane into spring- perennial sugarcane cycle (2)mulching by residue of sugarcane (3)installing the grass strip. Sediment yield tended to be large at summer sugarcane fields or pineapple fields. As slope length or slope angle enlarged, sediment yield tended to be increased.

17. Calculated sediment yield reduction ratios The reduction ratio was high at sugarcane fields with measure (1) and pine-apple fields with measure (2) and (3). Sediment discharge at outlet of the basin was decreased by 56% in comparison with present situation.

18. Conclusion We have carried out multi-points observations at four test plots in sugarcane fields to compare the amounts of sediment yield depending on the difference of agricultural management at each plot. Zero-tillage perennial farming is found to reduce sediment effectively. The WEPP model is effective to estimate the sediment yield at farmlands affected by various agricultural management conditions. In the present application to the watershed, the authors carried out some case studies to choose proper combinations of sediment yield reduction methods. The calculation has shown that the sediment runoff can be reduced by 56% at the watershed if the combination is adequately chosen.

19. Field plot test Outlines Amount of discharged sediment is calculated as the product of water discharge and sediment concentration.

20. g ・ L -1 mg ・ L -1 SS concentration N P June 8, 2004 Nutrients yield