Watershed and stream channel monitoring

Monitoring entails the initial collection of accurate and careful data at sites within a watershed, and the subsequent remeasurement of the same variables or conditions at a later date. Monitoring data is used to define changes and trends in watershed conditions, and may involve physical data from hillslopes or stream channels, or biological data concerning species presence and abundance.

Early monitoring of channel conditions in the South Fork Trinity River and its tributaries has not been well organized or conducted with an overall strategy. Historic data on channel conditions dates back to the installation of USGS stream gaging stations at several locations on the South Fork Trinity River (Hawley and Jones, 1969). At each gage site, channel cross sections have been re-surveyed at intervals and give an isolated picture of changing bed morphology. CDWR (1992) has plotted channel cross section data for three channel sites over a number of years to identify changing channel conditions: Forest Glen (1961, 1964, 1980), Hyampom (1973, 1979, 1980, 1990) and Salyer (1955, 1956, 1964, 1965, 1967, 1967, 1980, 1990).

Spot samples of bed material have been collected and analyzed by a number of agencies, including CDFG, CDWR and the USFS, but data has not been collected or presented in a systematic, repetitive manner. Most bed material sampling was performed for isolated sediment transport studies (CDWR, 1979; 1992), or for evaluation of the suitability of spawning habitat (CDWR, 1982; Bill Jong, personal communication).

More recent monitoring efforts have been undertaken to determine the effects of land use and wildfire on stream systems in the basin (Ranken, et al., 1989; USFS, 1990; Veevaert, et al., 1991). These recent efforts have been organized and undertaken to document the present condition and health of selected streams, and to provide a base-line of data that can be monitored for changes over time.

The initiation of this most recent monitoring work was spurred by the occurrence of upper basin wildfires in 1987 and 1988, and was planned to continue for five years. The $200,000 program was formulated to monitor watershed conditions, fire recovery measures and fire-related BMP implementation (Ranken, et. al., 1989; Veevaert, et. al., 1991). Much of the effort was directed at stream channel monitoring, since streams were believed to serve as an important indicator of overall watershed health and could be easily monitored for change. As a part of the same program, best management practices were monitored and evaluated on salvage sites throughout the burn areas.

Stream channel monitoring employed both physical and biological monitoring techniques. The following variables were measured at 35 stream stations, selected to diagnose the health of burned watersheds:

1. fish habitat typing and channel condition survey

2. channel cross sections and profiles

3. fish population survey

4. macro invertebrate index

5. spawning gravel composition

6. water temperature

The plan initially called for five years of monitoring. Timber management in the burned watersheds was to be postponed until monitoring revealed that conditions were stable (George Cruz, personal communication). Unfortunately, funding for the monitoring program has been cut substantially, and discontinuities in professional staffing at the District level has resulted in gaps in selected data categories (Darrel Ranken, personal communication).

As might be expected, data for more than preliminary assessment of changing stream and watershed conditions is not yet available. No large storm has occurred since the monitoring stations were established in 1989, so physical conditions have changed little. After the first two years, it was thought that several years of additional data would be needed for an adequate evaluation of conditions (USFS, 1990). With low flow years characterizing the first three years of the monitoring effort, and "considering the scope of the plan's objectives, it is becoming increasingly evident that further monitoring, beyond the initial five years, will be necessary" (Veevaert, et. al., 1991). A scaled-down monitoring effort has been conducted during low-flow years through 1993.The winter of 1992-93 was the first better-than-average rainfall year since the 1987 wildfires. While winter base flows were relatively high compared to recent years, no streams within the South Fork Trinity basin experienced peak flows of greater than a two year return interval (John Veevaert, personal communication). Consequently, measurable changes in channel morphology have still not occurred.

Since monitoring stations were initially established and measured, there have been several changes to the plan to accommodate field conditions (USFS, 1990). For example, once temperature data for selected sites was well established, recorders were moved to other locations. Measurements of cross section and profile data, as well as habitat typing, has not been repeated in low-flow years when channel changes could not be visually detected (Darrel Ranken, personal communication; John Veevaert, personal communication). In addition, macro invertebrate studies may be discontinued because of inconclusive results (Darrel Ranken, personal communication).

Past Efforts To Assess Fish Habitat and Fish Populations

A great deal of valuable scientific information has been collected on South Fork Trinity River salmon and steelhead stocks in recent years, partially as a result of the Trinity River Fish and Wildlife Restoration Program, but also because of California Department of Fish and Game concerns over the potential loss of spring chinook. Some of these monitoring efforts are described below, but extensive descriptions of past inventories, habitat surveys and fish population assessments have been previously reviewed (see Chapters II and III).

Selected Monitoring and Research Tools Needed to Guide Future Restoration

A recurring theme in USFS analyses and descriptions of watershed areas is the thought that watershed conditions continue to improve, and that the effects of the 1964 flood are lessening. Monitoring is needed to develop a better understanding of: 1) the natural biologic and physical system, 2) how and at what rate the basin is recovering, and 3) the nature and extend of past and future effects of human management and restoration activities.

Monitoring of watershed conditions must be continued (or intensified) to determine effects of management and restoration, and to chart the changing characteristics of watershed conditions, fish habitat and fish populations (Haskins and Irizarry, 1988). Those involved in restoration will need to continually check fish populations and other biological parameters, as well as water quality and physical stream channel (habitat) conditions.

The options available for monitoring hillslope, stream channel and biologic conditions in the watershed are numerous (MacDonald, et. al., 1991). Some study and monitoring methods described below represent a continuation of current, accepted sampling regimes. Other unused techniques have not been employed to the degree that is needed to understand and evaluate watershed and fish population changes. Still other, newer techniques are listed and described as possible candidates for future inclusion in the restoration program.

Assessing and Monitoring Fish Population Trends

Any fisheries restoration program must continually check population trends to measure success. However, the need is even more critical in the South Fork Trinity River basin where preventing stock loss mandates even closer monitoring. Carcass surveys, redd counts, and direct dive observations are not complex scientifically, but do require significant agency commitment of funds and personnel. Downstream migrant trapping of juvenile salmon has also been conducted, but is not a reliable population estimation tool.

Fall chinook salmon, spring chinook salmon and winter steelhead populations have all been estimated using a combination of tagging at weirs and spawning ground surveys. Because salmon die after spawning, carcass counts can be conducted for salmon, but steelhead populations can not be estimated using this technique. Spring chinook and summer steelhead are also counted using direct dive observation methods during summer low flows. Winter steelhead populations are difficult to estimate because they require maintenance of a weir to trap adult fish throughout the winter and because steelhead spawn during higher flows and are elusive when spawning.

Fall Chinook: Spawning escapement estimates were made by CDFG in 1964 (LaFaunce, 1964), and for six consecutive years beginning in 1985 (Jong and Mills, in press). During the more recent efforts, a weir was set up so that a portion of the fall chinook run could be tagged. Expansion of tag recoveries following spawner surveys provided a fairly accurate estimate of the population. The primary problems are posed by heavy fall rains, which sometimes cause the weir to be washed out, and elevate turbidity, making redd counts and carcass surveys impossible

.Since CDFG discontinued counts in 1991, no estimate is available for fall chinook salmon populations in the South Fork Trinity River in 1991 and 1992. CDFG (1993a) did operate the Sandy Bar weir to mark steelhead in 1992 and fall chinook counts at the weir were higher than in 1990 or 1991. Because most recent estimates show a population that is low and unstable, it is critical that population estimates and monitoring be resumed. A fall chinook salmon spawning estimate is being made for 1993 but was unavailable as this report went to press (Barry Collins, personal communication). A continuing annual population estimate of fall chinook salmon is clearly needed. If CDFG lacks necessary resources to accomplish this task, it is possible that they could enlist the cooperation of Six Rivers National Forest fisheries staff to continue spawner counts for this species.

Spring Chinook Salmon: Spawner counts by Healy (1963) provided the first estimate of the spring chinook population of the South Fork Trinity River. LaFaunce (1964) followed with an even more accurate assessment, using the Peterson mark-and-recapture method. Several hundred adult spring chinook were tagged in holding pools then tags were recovered during spawner surveys. If these worthwhile studies had not been conducted, no baseline information would have been available to judge the effects of past land use and the 1964 flood, or to help shape ultimate goals for restoration of Spring Chinook salmon runs. CDFG conducted index counts on various reaches of the South Fork Trinity River through the 1970's. Shasta Trinity National Forest began dive counts for spring chinook in the late 1980's and continued them for several years. CDFG has recently initiated an intensive study of spring chinook.

Since 1990, CDFG has counted and tagged spring chinook at Gates weir on the lower South Fork Trinity River. The tagged fish have then been counted through direct dive observation and during spawner surveys and redd counts (Dean, in press). The ratio of tagged to untagged fish is used to generate a population estimate using the Peterson mark recapture interval. The current CDFG spring chinook study is expected to continue for only five years, but this vital information needs to be gathered in the future. While it is optimal to have both summer holding estimates and spawner surveys, at least one comprehensive survey is needed annually. Good spawner counts require several passes and using this technique for a population estimate would necessitate continued tagging at a weir. If sufficient funding is not available, a direct dive observation sweep is still a very good index. However, summer dive counts do not give any indication of pre-spawn mortality levels.

Spring chinook (and summer steelhead) dive counts on the South Fork Trinity River could involve volunteer efforts similar to those pioneered by Klamath National Forest on the Salmon River. A dozen or more volunteers have supplemented Klamath National Forest staff during dive sweeps in recent years where over 40 miles of the river are covered in a span of two days. Volunteers are screened for swimming ability and stamina, provided wetsuits, and paired with more experienced divers. Because of the rough terrain in many reaches of the South Fork Trinity River and few access points, logistics for use of volunteers may be problematic (Mike Dean, personal communication).

Summer Steelhead: Summer steelhead have been counted during dive counts and at Gates weir during activities targeting spring chinook. No population estimate of this species has been carried out, however. This species is also at critical levels in the South Fork Trinity River basin. At the least, annual spring chinook studies should continue to make note of summer steelhead numbers. Additional effort to better understand the life history of this species and track population trends is warranted.

Coho Salmon: Because of the extremely low population level of coho salmon, and their late run timing, dive counts may be difficult or impossible to conduct. Lower river tributaries with some natural reproduction, such as Madden Creek, should continue to be monitored for coho. If further study of South Fork Trinity River coho salmon takes place, it should be part of a comprehensive lower Trinity River coho restoration effort.

Winter Steelhead: Adult population estimates of winter steelhead have relied on marking of returning adults at weirs, creel census, live fish counts on spawning beds, and downstream migrant trapping of those steelhead that survive spawning. Despite extensive effort, confidence intervals on estimates remain wide because of the few tagged fish recovered or sighted.

Redd counts for winter steelhead were conducted in selected reaches of Hayfork Creek in the early 1960's (LaFaunce, 1965), and during the early 1970's (Rogers, 1972). Since 1990, extensive redd counts have taken place in many tributaries of the South Fork Trinity River and Hayfork Creek (Wilson and Mills, 1991; Wilson and Collins, in press). Winter steelhead should respond most rapidly to restoration efforts, particularly to cooler water temperatures and increased summer flow in Hayfork Creek. Tracking improvements in these water quality parameters may provide us with the most immediate measure of restoration success in the Hayfork Creek basin but redd counts should also continue. The possibility of augmenting efforts by paid fisheries workers with trained volunteers should also be considered if funding for future redd counts is limited.

Juvenile Salmonid Population Estimation and Downstream Migrant Trapping: Downstream migrant traps have been operated periodically in the South Fork Trinity basin to determine the run timing and species composition of salmonids. Because the trapping efficiency is unknown and flows may vary considerably, data collected cannot be used for the purpose of population estimation. If artificial culture is continued or expanded in the basin, downstream migrant traps will be needed to monitor success or side effects from these efforts.

The standing crop of winter steelhead juveniles on Big Creek, Hayfork Creek, Salt Creek, Rusch Creek and Rattlesnake Creek has been monitored by the USFS since 1986, using electroshocking. Because of the high variability of anadromous fish populations, only long term data sets can provide information on trends associated with improved habitat. It is critical that USFS studies continue using the same sample reaches and methodology. Chapter 13 continued

Table of Contents