Monitoring Sediment Size Distributions in a Regulated Gravel-Bed Coastal Stream

Abstract:

Lagunitas Creek drains 282 km² in coastal Marin County, California. The watershed contains water supply reservoirs, urban areas, parks and habitat for threatened species (e.g. coho salmon). Water quality is impaired by excess fine sediment, and a plan to improve water quality (i.e. TMDL) was adopted by State authorities in 2014. The TMDL asserts changes in sediment delivery, transport, and storage contributed to the decline of coho. A sediment source analysis found a 2x increase in sediment supply. Concentrations of sand and fine gravel in the channel are elevated and, during high flows, more mobile. The Federal Coho Salmon Recovery Plan (2012) describes sediment conditions affecting coho habitat as “fair”. Reservoir managers were directed by the State in 1995 to reduce sedimentation and improve riparian vegetation and woody debris to improve fish habitat. Prior sediment monitoring found variability related primarily to intense winter runoff without identifying clear trends. A new sediment monitoring program was implemented in 2012 for ongoing quantification of sediment conditions.

The goal of monitoring is to determine with specified statistical certainty changes in sediment conditions over time and variation among reaches throughout the watershed. Conditions were compared in 3 reaches of Lagunitas Cr. and 2 tributaries. In each of the 5 channel reaches, 4 shorter reaches were sampled in a systematic grid comprised of 30 cross-channel transects spaced at intervals of 1/2 bankfull width and 10 sample points per transect; n=1200 in 5 channel reaches. Sediment diameter class (one clast), sediment facies (a patch descriptor), and habitat type were observed at each point. Fine sediment depth was measured by probing the thickness of the deposit, providing a means to estimate total volume of fine sediment and a measure of rearing habitat occupied by fine sediment (e.g. V*). Sub-surface sediment samples were collected and analyzed for size distribution at two scales: a larger sample of a spawning site in each sample reach and 20 smaller sub-samples of fine sediment facies. These data provide a robust description of streambed sediment conditions (e.g. % < 1 mm) expected to vary systematically across the watershed (e.g. fining downstream) and over time in response to management of watershed resources.