Field Investigations of Suspended Sediment Loading from Ward Creek Watershed At Lake Tahoe Andy Stubblefield, Charles Goldman and John Reuter Tahoe Research Group, University of California, Davis      This project focuses on gathering data for the support of adaptive watershed management in the Lake Tahoe Basin and the Sierra Nevada. By themselves, water quality standards and thresholds are not adequate to assess lake and watershed management alternatives. At Lake Tahoe we need to know (1) what are the sources and relative contribution of sediment and nutrients to the lake, (2) how much of a reduction in loading is necessary to achieve the desired water quality, and (3) how will this reduction be achieved? The answers to these and other questions must be evaluated within a dynamic framework based on solid scientific knowledge.      During water year 2000, research continued the synoptic sampling of Ward Creek. A second year of turbidity data was collected to support the data collected in water year 1999. The goals of the project were to (1) quantify sediment loading coming from small tributaries, channel bank erosion, and a steep denuded region referred to on USFS maps as "the Ward Badlands", (2) collect field data for the watershed sediment delivery model of Levant Kavvas' group, (3) evaluate the use of turbidity and turbidometers as proxies for phosphorus and suspended sediment, and (4) conduct an analysis of long term monitoring databases to support research findings.      Weekly sampling trips during the snowmelt season and during storm events were augmented by continuous turbidity readings. Four infrared backscatter turbidometers were installed in the Ward Creek drainage. The in-stream turbidometers provided continuous readings to Campbell Scientific dataloggers. The turbidometer readings complemented the sampling trips in the following ways. The continuous record provided a background and context for the weekly sampling. Ecosystems with snowmelt-driven hydrology have huge seasonal, weekly and daily variation in water discharge and sediment and nutrient loading. Without a continuous record, it is not possible to be assured that the events sampled on a weekly basis are representative of the actual conditions prevailing during the rest of the week. The additional function of the turbidometer is to characterize temporal variation. The synoptic sampling is supposed to represent a snapshot of conditions at seven locations on Ward Creek. However, in snowy conditions it can take several hours to sample all of the locations. The turbidometer serves as assurance that the differences between locations are due to differences in sediment loading between the locations, not due to differences in the time of day that the samples were gathered.      During each sampling trip the turbidometers were cleaned and the data downloaded into a laptop computer. At each of the four turbidometer locations and three additional ones stream water was collected total suspended sediment analysis. The stream discharge was determined with a flow meter at each location, except for the three locations that have continuous USGS stream gages.      Regressions were calculated to determine the relationship between the measured parameters and to calibrate the in-stream turbidity readings. Instantaneous loads were calculated by multiplying concentration times discharge. Yearly loads were calculated with the help of the turbidometer record. The yearly loading for significant stream reaches were compared in sediment budgets. Excellent correlations were found between turbidity and total suspended sediment (r2= 0.95). Supporting data was gathered from the analysis of long term monitoring data from the Tahoe Research Group. Biweekly sampling of three locations on Ward Creek for water quality parameters has been undertaken since 1992. Continuous discharge information is available from the USGS for these sites. Using the rating curve method, sediment budgets were created for the period 1992-2000.      The sediment budget for Ward Creek for water year 1999 found 22% of the loading coming from the South Fork, and 6% from the North Fork. The Upper Main Stem contributed 64% of the sediment load and the Lower Main Stem contributed 8% of the sediment load. The sediment budget for water year 2000 found 41% of the loading coming from the South Fork, and 7% from the North Fork. The Upper Main Stem contributed 34% of the sediment load and the Lower Main Stem contributed 18% of the sediment load.      Sediment budgets calculated from the long-term data set support these findings. Analyzing the period 1992-1996 and 1998-2000 together we find that 42% of the loading came from the Upper Main Stem, 28% from the Lower Main Stem and 30% from the headwaters region that contains both the North and South Forks. In 1997 a flood with a recurrence interval of 100 years occurred in Ward Creek. A landslide deposited large amounts of sediment into the channel. The sediment budget for the period 1992-2000 including the flood year gives 22% for the upper main stem, 64% for the lower main stem and 14% for the headwater region.      Given the stable condition of the South Fork channel banks it appears that the majority of the sediment coming from the South Fork is coming from the Ward Badlands. This region of deeply dissected topography, rills and gullies, is the result of past land use practices in the upper watershed, grazing and timber harvest. The large amount of sediment coming from this source relative to other parts of the basin makes it a high priority for restoration efforts. The North Fork has much more limited areas of gullying. The metasedimentary deposits forming the parent material are much more resistant than the fine volcanic mudflow material composing the parent material of much of the South Fork region.      In the main stem the upper reaches appear to have higher sediment loading for most years. Two sediment sources are predominant in this reach: 1) sediment from bank erosion during high flows, 2) release of sediment washed into the Upper Main Stem from the Ward Badlands during fall thunderstorms. Light development in the watershed of the upper main stem concentrates surface flow into culverts and gullies. Researchers in other regions (Booth and Jackson 1997, Trimble 1997) have found elevated rates of bank erosion and channel degradation from increased frequency and magnitude of storm flows after urbanization. This altered hydrology is a possible mechanism for the observed high sediment loading rates from the upper main stem.      The largest sediment source in this watershed was the landslide occurring in 1997. In a single event over 16,000 tons of sediment was moved into Lake Tahoe. This event highlights the importance of landslides to sediment budgets. Restoration efforts and future land-use decisions should take into account factors leading to landslide initiation including alteration of drainage patterns, road construction and timber harvest.      Hand sampling of tributaries draining the Scott Peak subwatershed and the Paige Meadows subwatershed, both on the northern slope of the Upper Main Stem, indicate a very minor role for these subwatersheds in the sediment budget of Ward Creek. For seven sampling dates between 4/23/00 and 6/4/00 the Scott Peak subwatershed was less than 1% of the sediment loading at the mouth of Ward Creek. Paige Meadows was less than 1/10 of 1%. The Paige Meadows tributary is a smaller watershed with a very low gradient and significant wetland areas. It is likely that the reduced sediment loading is from sediment trapping characteristics of the wetland areas. The Scott Peak subwatershed is forested with no roads or disturbance. Erosion rates off undisturbed forest landscapes are very low.      Analysis of the ratio of suspended sediment concentration to discharge for the long term data set and the turbidity record reveals a flushing process occurring in Ward Creek on a yearly cycle. Fall storms have the highest turbidities. Early spring rain on snow events produce moderate turbidity. By the late spring the turbidity is quite low for comparable flows. Figure 2 illustrates this phenomenon. The flushing effect shows that sediment exhaustion is occurring on a seasonal basis. This finding has strong implications for the restoration of Sierran watersheds. With restoration of degraded upland regions and unstable channel banks, less fine sediment will enter the river. This strong flushing effect suggests that the water quality of the river will respond quickly to the reduction of sediment inputs.