Mendocino Redwood Company Sediment Sampling Methods

MRC studied channel geometry and in-stream sediment at five stream segments in the Big River watershed in 2000. Channel geometry was measured through the use of thalweg profiles in all of the stream segments and further defined by surveying multiple cross-sections along each of the thalweg profiles. The thalweg profiles were run for a length of 20-30 bankfull channel widths upstream from a known reference point, which defined the distance of the stream segments. Evidently, channel geometry measurements were also conducted by MRC in 1998 at or near the stream segments studied in 2000. Based on antidotal descriptions of this fieldwork, it appears as though this data was collected without permanent benchmarks, making repeatability difficult. In any case, the data from this survey was not available for this assessment. In each of the stream segments, pebble counts, McNeil bulk samples, and permeability measurements comprised the in-stream sediment measurements. The five stream segments studied were Daugherty Creek above the South Fork Big River, Lower Ramon Creek, South Fork Big River above the Big River, Big River below the North Fork Big River, and the Lower East Branch of the North Fork Big River.

For the thalweg profiles and cross section surveys, reference points that mark the upstream and downstream ends of the monitoring segments were permanently monumented using nails driven into trees that appeared stable. This allows for a more permanent "benchmark" that can be used in future surveys. Distances and azimuths from these benchmarks to the start of a thalweg or a cross-section survey were recorded. By doing this, it is possible to begin and end surveys in the exact same location each year. These also provide a place of "known" elevation that should not change over time. This will presumably increase accuracy and confidence in comparability of data between years.

Working upstream, the thalweg depth (elevation) and distance along the stream was surveyed. The thalweg is the deepest point of the flowing channel, excluding any detached or "dead end" scours and/or side channels. These areas were excluded in the thalweg profile. As the team measuring the thalweg moved upstream, the distance from the previously measured point and the elevation of the current location was taken at every visually apparent change in location or depth of the flowing channel. Distance was measured by stretching a 300-foot tape along the channel and then reading a station distance during the survey. In the absence of visually apparent changes, thalweg measurements were taken every 15-20 feet up the center tape.

As specific landmarks were encountered along the reach, (e.g. tributary channels, particularly large pieces of woody debris, permanent survey stakes, armored bend, or other features of interest) the recorder made note of their location and size. Where a channel split into two components, the surveyor decided which is the main channel and then continued moving upstream (making measurements) along that channel.

Approximately every 5 to 8 bankfull channel widths along the thalweg profile, the location for a cross section survey was monumented and recorded in the thalweg profile survey notes. The cross sections were located across riffles on relatively straight reaches of channel. Cross sections were surveyed from above the bankfull channel margins on both banks. Typically 2 to 4 cross sections were measured along each stream segment.

Cross section rebar pins were established at both ends of the cross section well above the flood-prone channel margin to monument the cross section location. The elevation and the distance from the left bank pin was measured at least every five feet or at any visually apparent topographic change along the cross section. At each cross section a pebble count was also conducted to determine the median particle size of the stream bed (D₅₀) by measuring 100 randomly selected pebbles along a single transect. The pebble counts collected in this manner may be comparable to similar data collected by DFG and can provide a snapshot of the median size of surface material at the cross-section. When the median pebble counts at multiple cross-sections are reviewed as a whole, it may be an indicator of trends in the amount of fine sediment moving through the stream segment. However, it may not be comparable to pebble counts that are taken along multiple closely spaced transects, such as the pebble counts conducted by Knopp (1993). Further study is necessary to determine the comparability of these two different pebble count measurement techniques.

Substrate samples (McNeil cores) were taken from four randomly selected pool tail-outs in each segment from all of the pool tail-outs that are suitable for spawning (i.e., not dominated by bedrock or covered in substrate too large for a fish to make a redd). Bulk gravel samples were taken at the same location as the deepest permeability site in a given tail-out. Bulk gravel samples were collected using a modified McNeil sampler (a 12" cylinder placed on the streambed and worked downward as the sample is manually removed). The original McNeil sampler allowed for suspended material dislodged during the removal process to be included in the sample. This is done by using a stopper in opening to the McNeil sampler to retain the water, and therefore the suspended sediment, in the sample. Because the MRC methods do not use a stopper to retain the free water, it may slightly under-represent the finer fraction of the bed substrate (Hames et al. 1996).

To sort the streambed substrate samples, MRC used a gravimetric (dry sieve) method. This consisted of drying their gravel samples, weighing the total sample, and then passing the sample through seven progressively finer screens (50, 25, 12.5, 6.3, 4.75, 2.36, 0.85 mm). The material retained on each screen was then weighed to calculate the percent finer for each size class. None of the size classes were truncated and the entire sample was used.

MRC measured gravel permeability in the Big River following the methods of Barnard and McBain (1994). A perforated standpipe was driven into the streambed to a depth of 25 centimeters, which was chosen as an intermediate depth for a coho redd. The perforations in the bottom of the standpipe allow the interstitial water in the streambed to flow into the standpipe up to the height of the stream surface. An electric pump was then used to draw a suction on the standpipe and evacuate 2.54 centimeters of water in the standpipe into a sealed graduated cylinder. The lowered hydraulic head in the standpipe induces interstitial water in the streambed to flow into the standpipe at a rate equal to the water flowing into the graduated cylinder. The flow into the graduated cylinder was measured and timed, resulting in values for flow rate. Through a calibration curve, these flow rate values can be related to gravel permeability in units of centimeters per hour (cm/hr), which describe the interstitial flow rate in the streambed. At each measurement location, repetitive measurements were taken until the permeability readings ceased to increase.

From a power analysis, MRC determined that a total of 26 permeability measurement locations were needed for each stream segment to predict the survival of emerging fry within 20 percent accuracy. The 26 permeability measurement locations were distributed equally among each of the pool tail-outs in each stream segment, with any extra measurements taken in tail-outs behind the deepest pool(s). The measurement location in each pool tail-out was randomly selected from an evenly spaced 12-point grid. For example, if there are six pool tail-outs in a given stream segment, there might be four permeability measurements taken at each tail-out with the remaining two permeability measurements taken in the tail-outs behind the two deepest pools. In this example, the permeability measurements at each pool tail-out would consist of four randomly selected points in a 12 point grid with an additional two random grid points in the tail-out below the deepest pool. In all cases, the permeability measurements were adjusted for the viscosity of the water, which is a function of the water temperature. This required recording the temperature of the water at each location.

The median value of all of the permeability measurements were then used to determine the overall permeability for the given stream segment. The median values were used instead of the averages to avoid the data becoming skewed by permeability measurements that were unusually high or low. The overall median permeability value was then used for direct comparisons against the other sites monitored by MRC, and also to relate the permeability values with chinook and coho survival to emergence. The relationship between permeability and survival to emergence was derived by Tagart (1976) and McCuddin (1977) (as cited in McBain and Trush 2000), using a best fit regression curve (r²=0.8521):

In a few cases, the survival index was a negative number. In these cases, the index was reported as zero. The survival relationship is an index of spawning gravel quality and interpretations based on this can be only considered preliminary. However, this is currently one of the few approaches that quantitatively links a biological relationship to permeability data.

References

Hames, S.H., R. Conrad, A. Pleus and D. Smith. 1996. Field comparisons of the McNeil sampler with three shovel based methods used to sample spawning gravel substrate composition in small streams. Northwest Indian Fisheries Commission. Part of Timber Fish and Wildlife Ambient Monitoring Program Series. TFW-AM-9-96-005.

Barnard, K. and S. McBain. 1994. Standpipe to Determine Permeability, Dissolved Oxygen, and Vertical Particle Size Distribution in Salmonid Spawning Gravels. As FHR Currents # 15. US Forest Service, Region 5. Eureka, CA. 12 pp.

McBain and Trush. 2000. Spawning gravel composition and permeability within the Garcia River watershed, CA. Final Report. Prepared for Mendocino County Resource Conservation District. 32 pp. without appendices.