Coupled Control of Land Use and Topography on Suspended Sediment Dynamics in an Agriculture-Forest Dominated Watershed.

Suspended sediment (SS) and associated nutrients transport from land to watercourse is an immense problem that has threatened soil and water conservation in the world. Understanding SS transport mechanism is important for soil erosion estimation and making the best management practices.

This study was conducted in an agriculture-forest dominated Shibetsu River watershed (SRW) and its two sub-watersheds with different land uses and topography located in Hokkaido, Japan, to understand effects of land use, topography and hydrological process on SS dynamics. The agriculture-dominated sub-watershed (AW) had flat topography, the forest-dominated sub-watershed (FW) was characterized by steep slopes. During the snowmelt seasons and rainfall events, stream water level (H) was measured and stream discharge (Q) was calculated from H-Q equations. Water samples were collected by auto-sampler and filtered through 0.7 μm glass microfiber filters for analyzing suspended sediment concentration (SSC). Measured hydrograph of six rainfall events and wavelet analysis for daily streamflow from 2003 to 2008 were analyzed to understand the time-lag between streamflow and rainfall. Relationships between Q and SSC were analyzed by hysteresis loops for two snowmelt events and six flood events.

Measured hydrograph and wavelet analysis results showed that response of streamflow to rainfall was faster in AW than FW during the flood events. During snowmelt events and flood events, results showed that mean and maximum SSC were much higher in AW than FW. However, stream water yield was much higher in FW than AW and this resulted in higher SS yield in FW than AW in two flood events and a snowmelt event in May 2006, even though the mean SSC was lower in FW than AW. For most of the flood events, SS yield was much higher in AW than in FW, indicating that soil erosion was more serious in AW due to the plant cover (most area was covered by pasture) and management practices, such as grazing, even though slopes of AW was more flatter than FW. Results of hysteresis loops showed that most of the flood events were characterized with anti-clockwise hysteresis (peak of SSC happened after peak of Q) in AW, indicating that the SS source maybe mainly from distant areas with relatively steeper slopes from the streams. However, most of the flood events were characterized with clockwise hysteresis (peak of SSC happened before peak of Q) in FW, indicating that the SS sources maybe mainly from the near stream areas with steep slopes. At the whole watershed outlet SRW, most flood events were characterized with clockwise hysteresis, it is concluded that existence of earlier SS supply or increased portion of streamflow during the recession limb of discharge. The earlier SS sources might be mainly from agriculture land, such as AW, due to its higher soil erosion, faster response of streamflow to rainfall, and located nearer to the main outlet SRW compared with FW. The increased portion of streamflow during the recession limb of discharge at SRW might be from the forest land, which had more stream water yield and slower response of streamflow to rainfall during flood events, and also located further to the main outlet compared with agriculture land.

In conclusion, streamflow response to rainfall was faster, mean and maximum SSC were higher in AW than FW. Stream water yield was higher in FW than AW during snowmelt and flood events. Regarding the SS transport dynamics during all flood events, AW was mostly characterized with anticlockwise hysteretic loop, indicating SS source from distant areas, while FW and SRW were mostly characterized with clockwise hysteretic loops, indicating SS source from near stream areas. Agriculture land was identified with critical soil erosion, thus, management practices should be carefully conducted in these areas, especially in distant areas with steep slopes, such as avoiding tillage, harvest and dairy cattle grazing practices during snowmelt and rainfall events and reducing grazing frequency.