Large Tributaries to the Lower South Fork Trinity River
Madden Creek
The habitat typing survey for this lower South Fork Trinity River tributary was conducted from July 24 to August 14, 1989. The survey report included a watershed history and recommendations for possible fish habitat improvement projects (Dale, 1992a). Stream temperatures ranged" in the 50's and low 60's" during the survey and a complete canopy of alder covered the stream. The report suggests a long, steep boulder cascade located about one mile up Madden Creek is the upstream extent of available anadromous fish habitat.
Juvenile steelhead, chinook salmon and coho salmon were all found in low densities throughout the lower reach. The stream channel was comprised of 40% low gradient riffles and 20% high gradient riffles, by length. Pools made up only 17% of the channel and yet salmon, steelhead, and resident rainbow trout strongly favored pool habitats. Plunge pools and main channel pools, those features with greatest depth, had by far the highest concentration of all species. Low and high gradient riffles and runs also were inhabited by all species, but in very low numbers. Larger steelhead juveniles that had spent one or two years in freshwater (1+ and 2+) were found in step runs and cascades.
Only 10% of this watershed is held privately and only 17% of the total area has been logged since 1948. Despite this low level of timber harvest, debris flows occurred in the Madden Creek channel during the 1964 Flood (Grunert, 1974). Dale (1992a) concluded that "past management activities (in the watershed)...may have adversely affected both the availability of rearing habitat and suitable gravels for spawning." Madden Creek is in late stages of recovery from the flood at this time.
Sand and fine sediment comprised only 20% of sediments in pools. This was substantially less than the 34% found in pool habitats in the nearby Grouse Creek watershed (Fuller, 1990), which is known to have continuing problems with incursions of sediment. In Madden Creek, alders have created a riparian corridor, the stream channel has narrowed, but pools are infrequent and lack depth. Lack of rearing space and spawning gravels seems to be limiting productivity of salmon and steelhead.
Grouse Creek
Habitat typing has been completed by the USFS on 25 miles of the main stem of this fifth order tributary, which flows off South Fork Mountain (Dale, 1992b). The survey was conducted during July-August, 1988. Grouse Creek is one of the six largest South Fork tributaries, but a large landslide (Devastation Slide), located 1.6 miles from the mouth, blocks upstream access for anadromous fish. The watershed is 41% privately owned. Concerns about the cumulative effects of land management on private and Forest Service land have lead the Forest Service to suspend all timber harvest in the basin (Fuller, 1990). Numerous studies have been conducted to estimate sediment budgets, major erosion sources, changes in the stream channel in response to land management, and impacts of sediment on biota in the stream (Raines and Kelsey, 1991, McHugh, 1989, Mangum, 1990, Fuller, 1990). Dale (1992b) incorporated information from these reports in his interpretations of the habitat typing data.
Grouse Creek was broken into three reaches, one above and one below Mosquito Creek, and the lowest five miles of Mosquito Creek itself. All channel types were moderately confined, falling into "B" types in the Rosgen classification system. In the reach below Mosquito Creek, the most abundant habitat types (by length) were high gradient riffles (20%), low gradient riffles (18%), main channel pools (13%) and runs (12%). Pools (all types) comprised 27% of all habitat units, by length.
Above Mosquito Creek, low gradient riffles comprised approximately 42% of all units, by length. Other prevalent habitat types included step runs (16%), runs (11%), high gradient riffles (10%), and lateral scour boulder pools (7%). In the lowest five miles of Mosquito Creek, relative abundance of habitat types were very similar to those in upper Grouse Creek: low gradient riffles (45%), step runs (15%), runs (12%), and high gradient riffles (11%). Grouse Creek is a B channel type, so it is only moderately susceptible to channel filling. The predominance of low gradient riffle habitats and lack of pool habitats are probably in part a result of aggradation.
Resident rainbow trout inhabited upper Grouse Creek, including Mosquito Creek, while chinook salmon and steelhead trout were found only below Devastation Slide. Only five young of the year chinook were counted; too few to determine patterns of habitat utilization. No counts of 0+ steelhead were made. Pools were preferred habitat types for both resident rainbow and 1+ and 2+ steelhead juveniles. Plunge pools and lateral scour bedrock pools had the highest densities of steelhead with step runs and main channel pools also showing considerable use. Resident rainbow trout were most abundant in main channel pools, boulder formed lateral scour pools, and bedrock lateral scour pools, respectively. Runs and step runs were the next most preferred trout habitat. Low and high gradient riffles and glides also held some fish.
Fuller (1990) used selected habitat typing information from Grouse Creek and compared it with other streams on Six Rivers National Forest to demonstrate problems resulting from cumulative effects. Grouse Creek had, by far, the highest amount of sand and fines of any stream measured (Figure 3-1). The depth of the stream relative to its width was less than any other stream as well. Fuller (1990) concluded that Grouse Creek was suffering from chronic high sediment levels, which resulted in decreased rearing habitat for fish, reduced egg survival, and decreased invertebrate production. Macroinvertebrate comparisons in Grouse Creek, Willow Creek, and Horse Linto Creek showed that biotic conditions in Grouse Creekwere impaired relative to the less impacted watersheds in the Trinity River basin. Fuller (1990) also noted that stream widening had decreased shade on the stream, leading to increases in summer water temperatures. Dale (1992b) stated that stream temperatures in lower Grouse Creek exceed 70°F during summer.
Figure 3-1. Comparison of fine sediment and sand in pool habitats, as determined by habitat typing surveys, in various streams throughout Six Rivers National Forest. NOT AVAILABLE IN ELECTRONIC FORMAT
Eltapom Creek
This third order stream joins the South Fork Trinity River from the east just below the town of Hyampom. Its watershed includes area both within the Hayfork and Big Bar Ranger Districts of Shasta-Trinity National Forest. Between August 1-5, 1991, habitat typing was conducted in the lowest mile of Eltapom Creek, below a 50 foot waterfall which blocks anadromous fish migration (Mayo, 1992a). The most frequently occurring habitat types were step runs (33%), lateral scour bedrock pools (19%), main channel pools (16%), and low gradient riffles (14%).
Densities of juvenile steelhead in Eltapom Creek were much higher than in any other stream in the South Fork Trinity River basin and among the highest of all Trinity basin streams. Young of the year steelhead trout (0+) favored step runs, low gradient riffles and runs. They were also present, along with 1+ and 2+ steelhead, in most pool types except lateral scour bedrock pools. Only 2% of the gravels in Eltapom Creek were considered suitable for spawning. Fine sediment levels in those gravels were estimated to range between 17 and 20 percent.
The Eltapom Creek watershed is comprised of relatively stable geologic formations. Despite past problems with logging (Coots, 1952) and recent fires, no major sediment problems have been apparent. Coho salmon were once present in this stream (Coots, 1952), but past flood effects may have played a part in their disappearance. Because the lower reach has a high gradient, confined channel, any sediment entering the creek would be flushed during subsequent high stream flows. The steep canyon also provides shade, even in areas where trees on hillslopes have been burned recently, so water temperatures remain cool in summer. High fish densities may reflect, in part, a movement of fish up from the main South Fork to escape high summer water temperatures.
Small Tributaries to the Lower South Fork Trinity River
Ammon Creek, Mingo Creek, Hell's Half Acre Creek and Underwood Creek were surveyed from June 20 through August 20, 1990 (Dale 1990). The report noted that the temperature of the main South Fork Trinity River ranged between 68-77°F during the survey. The primary reason for the survey was to determine how important the lower reaches of these tributaries are for providing cold water refuges for salmon and steelhead juveniles during summer. All tributaries, except Hell's Half Acre Creek, were found to be contributing cooler water and providing suitable habitat for juvenile salmonids in their lower reaches, and below stream mouths in the South Fork Trinity River.
Ammon Creek
This stream is approximately 2.6 miles in length and flows off South Fork Mountain. Its steep gradient forms a channel with abundant high gradient riffles (32%), low gradient riffles (24%), step pools (18%) and cascades (10%). No 2+ juvenile steelhead were found in the stream, but steelhead young of the year (0+) and 1+ fish were found throughout the survey reach. Highest concentrations of young steelhead were found in the lowest mile of stream, where pools were favored habitats. Step pools had the highest concentrations of steelhead juveniles (22%); 19% were observed in lateral scour pools; main channel pools had 10%, and 10% were found in plunge pools. Some use of riffle habitats by 0+ steelhead was also observed.
Mingo Creek
This stream also drains South Fork Mountain. A 4.8 mile reach was habitat typed. High gradient riffles were again most numerous (42%). Step pools (20%) and low gradient riffles (16%) were the other frequently occurring habitat units. The report does not mention any resident rainbow trout population in the stream's upper reaches, but the lowest 1.2 miles held juvenile steelhead. A large waterfall 2 miles upstream would definitely prevent anadromous fish passage. Divers found 88% of the steelhead in the stream were young of the year (0+) with the remainder being yearlings (1+). No 2+ steelhead juveniles were counted. The majority of fish resided in lateral scour pools (28%), main channel pools (19%), step pools (19%), and low gradient riffles (20%). The highest concentrations of fish were found just above the confluence with the South Fork.
Hell's Half Acre Creek
This creek spills into the South Fork Trinity River from the east, over a 10 foot high bedrock falls. No resident or anadromous fish were found in Hell's Half Acre Creek. Most frequently occurring habitat types were low gradient riffles (35%), high gradient riffles (13%), cascades (13%), and step pools (14%). Average stream temperatures were 65°F, with highs in the high 70s. Hell's Half Acre Creek was noted to be deficient in shade.
Underwood Creek
This second order tributary is three miles in length and also drains into the South Fork from the east. A short, low gradient reach extends approximately 500 feet up from the mouth and the rest of the stream channel is very steep and confined. Only the lowest 1000 feet of the stream had juvenile steelhead, but all age classes were represented. Over 70% of the steelhead observed were in lateral scour pools. The most abundant habitat types were low gradient riffles (30%), step pools (28%), cascades (17%), and high gradient riffles (14%).
South Fork Trinity River (Big Slide Creek to East Fork of South Fork)
Shasta Trinity National Forest habitat typed the South Fork Trinity River from Slide Creek, below Hyampom, to the East Fork of the South Fork during the summer of 1989 (USFS, 1990b). The survey found that the predominant habitat types, by length, were glides (20%), low gradient riffles (18%), runs (18%), high gradient riffles (8%) and step runs (8%). The mild stream gradient and relatively unconfined channel present in this portion of the South Fork have made the stream much more susceptible to aggradation than the gorge area downstream from Slide Creek. An increase in glide and low gradient riffle habitats relative to the number of pools may be, in part, a result of this channel filling.
Pool habitat types comprised 24% of all habitat types by length, but only 30% by volume. Only four pool types had average depths greater than 1 meter (m) and only plunge pools exceeded 1.5 m in average depth. The shallowness of pools indicates that most of the South Fork is only in early stages of recovery from the 1964 flood. The maximum stream temperature measured was 76°F and the average temperature for the reach was 67°F, which is higher than optimal for salmonids (Reiser and Bjornn, 1979). Lack of pool depth exacerbates temperature problems in this reach because it impedes thermal stratification.
Substrate composition and embeddedness also suggest that excessive quantities of fine sediment persist in the South Fork. Pools had a total of 39% sand and fines, which is high relative to other Klamath and Trinity basin streams (Figure 3-1). Fuller (1990) usedthis criterion as an index of cumulative effects. Embeddedness is measured during habitat typing to see to what degree spaces aroundcobbles, gravels or boulders are filled in by sand and fine sediment. Pool tail crests in this reach of the South Fork had 42% embeddedness. Interstitial spaces in stream gravels can produce a tremendous abundance and diversity of invertebrates, often to a depth of two feet or more (Hynes, 1970). Therefore, high embeddedness also indicates a substantial decrease in food available for juvenile salmonids. Studies also show that embeddedness in excess of 25% in spawning areas can lead to poor survival of salmon and steelhead eggs (Reiser and Bjornn, 1979).
The habitat typing report (USFS, 1990b) noted that spawning areas had a "large percentage of fine sediment" and that "embeddedness of spawning gravels was especially high below streamside slides, which were numerous." Dean (personal communication) has seen spring chinook salmon redds covered with fine sediment stemming from landslides just upstream. DWR (1982) suggested that the entire gravel matrix of the mainstem South Fork was unstable. The instability has lead to limited success for salmon spawning due to frequent shifts in the streambed during even low to moderate winter storm flows. There is evidence that problems with gravel quality and stability may still be a major limiting factor for spring chinook salmon in the mainstem of the South Fork Trinity River.
The highest densities of juvenile steelhead were found in dammed pools, plunge pools, and lateral scour log formed pools, all of which combined to form only 1.4% of all habitat units, by length. The most frequently occurring habitat types, such as glides and low gradient riffles, had some of the lowest fish densities, possibly because high stream temperatures made them uninhabitable. Both 1+ and 2+ juvenile steelhead showed a strong preference for cascades (2% of all units). Turbulence in cascades and high gradient riffles increases dissolved oxygen, which may explain increased fish densities in these habitats.
Fish densities in this reach of the South Fork Trinity River were the lowest of any stream for which data was available (Table 3-1). When compared with other South Fork sub-basins and Trinity River tributaries, fish densities were lower by an order of magnitude. and two orders of magnitude lower than fish densities in Fish Creek, in Oregon (Everest et al., 1986). Only two juvenile chinook salmon were counted during the entire survey. Older age steelhead juveniles (1+ and 2+) were more numerous than young of the year (0+).
The river was divided into 15 reaches of varying length for the purpose of this survey. Tributaries flowing off South Fork Mountain from Sulphur Glade Creek downstream are thought to be the major source of sediment still entering the South Fork. A great deal of sediment also remains stored in river fill terraces along low gradient reaches from Sulphur Glade Creek downstream (Haskins and Irizarry, 1988). One would expect that fish habitat recovery would be further advanced in reaches upstream of the sediment sources, and that fish densities would increase progressively upstream. No persistent trend toward increasing juvenile steelhead abundance in upper reaches was evident (Figure 3-2).
Table 3-1. Steelhead juvenile densities (in fish/meter2), by age, for various streams (information taken from Shasta-Trinity National Forest reports). Stream Name 0+ Steelhead 1+ Steelhead 2+ Steelhead Big French Creek .291 .075 .054 Grider Creek .232 .070 N/A E.F. of N.F. Trinity .364 .080 .018 N.F. Trinity River .141 .057 .009 Canyon Creek .170 .055 .011 Eltapom Creek .807 .048 .065 Plummer Creek .136 .010 .015 Butter Creek .256 .015 .010 Rusch Creek .104 .067 .054 Upper Salt Creek .050 .059 .020 Lower Salt Creek .180 .001 .001 Big Creek (Hayfork) .046 .014 est .005 Rattlesnake Creek .020 .034 .013 Potato Creek .014 .040 .033 Hayfork Creek R#11 .050 .030 .001 Hayfork Creek R#22 N/A N/A N/A Hayfork Creek R#33 .060 .001 .0001 Hayfork Creek R#44 .049 .003 .003 E.F. of S.F. Trinity .030 .011 .006 Dubakella Creek .008 .003 .011 Lower S.F. Trinity5 .002 .003 .024 Middle S.F. Trinity6 .002 .004 .004 Fish Creek (Oregon)7 .552 .124 N/A Idaho Streams8 .230 .069 .058 1 Hayfork Creek from the town of Hayfork to the East Fork 2 Hayfork Creek in Hayfork Valley 3 Hayfork Creek from Nine Mile Bridge to lower Hayfork Valley 4 Hayfork Creek above the East Fork up to Dubakella Creek 5 Main stem South Fork from mouth to Coon Creek 6 Main stem South Fork from Slide Creek to East Fork of South Fork 7 Fish densities from Everest et al. (1986) 8 Fish densities from Hanson et al. (1977) as cited in USFS (1990c)
Figure 3-2. Densities of juvenile salmonids by reach from Slide Creek to the East Fork of the South Fork Trinity River. NOT AVAILABLE IN ELECTRONIC FORMAT
Large Tributaries to the Middle South Fork Trinity River
Butter Creek
The lowest 1.6 miles of Butter Creek was habitat typed by Shasta-Trinity National Forest staff from August 12-29, 1989 (Gilroy et al., 1993b). The main stem of Butter Creek above the reach surveyed and Indian Valley Creek, a major tributary, are inaccessible to anadromous fish due to large, impassable waterfalls. The average gradient of the habitat typed section was 2.5%. Most of the reach surveyed was moderately entrenched and confined, and was classified as a B-2 Rosgen channel type. A short section near the upstream end of the survey reach was steeper (7.5% gradient) and very confined (A1-a). Stream temperatures averaged 61°F during the survey and ranged from 57-65°F.
The most abundant habitat types were step runs (19%), runs (15%), main channel pools (12%), glides (11%), and low gradient (9%) and high gradient (9%) riffles. Boulders and cobbles were the predominant substrate in Butter Creek, which is not surprising given the high gradient of the stream. Sand and fine sediment comprised 18% of pool substrates. Only 2% of the stream had gravels suitable for spawning and fine sediment within these spawning beds was judged to be 16%. Embeddedness at pool tail crests was visually estimated at 41%.
No chinook salmon juveniles were counted, but all age classes of juvenile steelhead were found. Young of the year (0+) comprised 93% of steelhead present, with 1+ and 2+ making up just 5% and 2%, respectively. While 0+ were found in all habitat types, they showed greatest preference for runs, root wad formed lateral scour pools, and step pools. Runs were most preferred by 1+ steelhead also, with plunge pools and step pools harboring the next highest densities. Plunge pools, lateral scour boulder pools and step pools held the highest number of 2+ steelhead.
Densities of 0+ steelhead were some of the highest of any South Fork Trinity River tributary and compared favorably with the East Fork of the North Fork Trinity River (Table 3-1). The density of 1+ and 2+ were comparable to other South Fork basin tributaries, such as Rattlesnake Creek, Big Creek, Plummer Creek and the East Fork of the South Fork. Cool water temperatures allowed fish to use all habitat types. Two large, streamside landslides were noted, and potential sediment contributions from lands disturbed by the 1987 fires are still possible. Large fill terraces near the mouth of the stream are evidence of the 1964 flood, but Butter Creek seems to be in advanced stages of recovery at this time. The high energy associated with the confined, steep stream channel has helped to flush excess sediments. Coots (1952) noted that coho salmon juveniles were present in the stream and that chinook salmon used the stream for spawning. Changes in Butter Creek's channel, such as diminished pool frequency and depth, stream armoring with larger cobbles and small boulders, or flushing of smaller gravels suitable for spawning may have reduced the ability of the stream to support salmon species. Chapter 3 continued