LIFE HISTORY, DISTRIBUTION, RUN SIZE, AND HARVEST OF SPRING CHINOOK SALMON IN THE SOUTH FORK TRINITY RIVER BASIN (continued)
The study area included the lower 124 km of the SFTR, the lower 7 km of the East Fork of the SFTR, and the lower 16 km of Hayfork Creek, totaling 147 km of river. Lafaunce (1967) and USFS surveys (Appendix 1) broke this area into 16 roughly equal sections. We attempted to use these same sections for comparison, but for logistical reasons deviated slightly from their delineations (Figures 1 & 2). We also snorkel surveyed the lower 4 km of Grouse Creek.
This study is comprised of several distinct elements, each intended to generate an escapement estimate or provide information on in-stream life history or distribution.
To meet Job Objective 1, we used the Petersen mark and recapture method, with some variation. We operated a weir at which fish were trapped, tagged, and released. We recovered fish or observed tags in three ways: 1) a recapture weir in the mainstem SFTR, and one in Hayfork Creek; 2) snorkel surveys of the entire study area; and 3) carcass recoveries during the spawning season. Data from each recovery technique were intended to be used in making separate Petersen estimates. We used several recovery techniques to insure that at least one would yield statistically- valid results, and to make comparisons between the different methods. Petersen estimates represent point-in-time run-size estimates upstream of the tagging weir. Snorkel surveys were also used to determine in-river distribution, and to continue documenting run timing once the tagging weir was removed. The number and distribution of redds were determined by foot and kayak surveys (redd surveys).
To meet Job Objective 2, we utilized non-reward tag returns and a limited creel survey. Historically, poaching has been a problem in the SFTR. Non-reward tags were chosen so the potential of poaching, primarily for the reward, was not increased.
To meet Job Objective 3, we analyzed scales collected during the adult trapping operation and carcass recovery surveys, and performed emigrant juvenile trapping.
The primary trapping and tagging weir (Gates Weir) was located at river kilometer (RKM) 31.7, 16 km downstream from the township of Hyampom (Figure 1). The weir functioned as a fence across the river, guiding fish into a trap. The weir was constructed of 1.5-m-wide by 1.2-m-high panels, which reached completely across the river. Each panel was constructed of 1.9-cm-diameter galvanized conduit welded horizontally on 5.7-cm centers to 2.5-cm by 2.5-cm steel angle-iron uprights. Panels were wired together with steel tie-wire, and supported with conventional steel fence posts driven into the river bottom. Netting was placed atop the panels to prevent fish from jumping over the weir.
The trap was 2.4 m long by 2.4 m wide by 1.2 m high (vertical dimension) and was constructed with the weir panels described above. Two 1.2-m2 panels were placed inside the open end of the trap forming a fyke, guiding fish inside and deterring their escape. The conduit of the upstream and side panels was sleeved with clear vinyl tubing to minimize potential abrasion to trapped fish. To make fish more "at ease" in the trap and less likely to try to jump out, a piece of dark blue nylon fabric was floated on the water surface inside the trap. It was attached only at the upstream end of the trap, so if a fish were to jump and land atop the fabric, it would sink, allowing the fish to settle back into the water. This device also provided cover and made fish difficult to see from outside the trap. Great care was taken to insure that there were no sharp projections, wire, etc. inside the trap which might injure fish. Foam pipe insulation was used in areas where unavoidable abrasion might otherwise occur. The trap was provided with a lockable plywood lid and solid plywood bottom.
Fish were netted from the trap with a knotless-nylon-mesh net and placed in a tagging cradle. The tagging cradle consisted of a frame, constructed from 1.9-cm-diameter copper pipe, measuring 100 by 50 cm, and was fitted with a nylon cradle and a metric ruler for measuring fork lengths (FL). The cradle assembly was designed to slide into a channel in the front of the trap. A sliding door made from perforated aluminum plate (0.32-cm holes) formed the upstream end of the cradle. Once marked and measured, fish were released by raising the sliding door.
During tagging, fish were examined for marks, scars, and general condition, their FL measured to the nearest cm, and a scale sample was taken. A small knife was used to collect scales from the left side of the fish just below the dorsal fin. Spring chinook from the 1992 cohort, which appeared healthy, were marked in one of two ways: either a one-half left ventral (2LV) fin clip and a numbered Floy anchor tag, or a one-half right ventral (2RV) fin clip and a Lotek1/ implantable radio transmitter. Anchor tags were placed on the left side, just below the dorsal fin, and just posterior to the midline. Radio transmitters were inserted into the stomach of adult spring chinook through the esophagus with the aid of a small length of 0.95-cm-diameter plastic pipe. The radio tagging operation was done in cooperation with a project led by Dr. Roger Barnhart of the U.S. Fish and Wildlife Service, California Cooperative Fishery Unit, Humboldt State University. Thirty-nine spring chinook were tagged with Floy tags and nine received radio transmitters. (Note: Spring chinook were marked as described above during the last reporting period [1991-1992] and are discussed in the RESULTS section of this report. Spring chinook of the 1993 cohort, were marked at the end of this reporting period with a 2RV fin clip and colored anchor-tags, and will be discussed in the 1993-1994 Annual Report.) Spring-run steelhead were marked with a 2LV fin clip.
Tagged fish were sprayed with a 10-20% aqueous solution of Propolyaqua (artificial slime) to help prevent infection caused by the removal of mucus during handling. Spraying was focused on areas such as the caudal peduncle, scale-sample site, and the tag location. Care was taken to insure that the head, operculum, and gills were not sprayed with the solution.
After processing, fish which appeared fresh and strong were immediately released from the cradle to the river, upstream of the weir, without further handling. During periods of warm water temperature (> 15.5 oC) or when they appeared stressed, fish were allowed to swim from the cradle into a recovery tube and held there for at least 60 minutes. The recovery tubes were made from plastic pipe measuring 3.5 m long by 25 cm in diameter. The upstream and downstream ends were fitted with sliding plexiglass doors, each with numerous 2-cm holes allowing ample water to flow through the tube. The tubes were oriented with their long axis parallel to the current and held on the river bottom with large rocks or steel fence posts. After recovery, the upstream door was opened and fish were allowed to leave of their own volition.
Instead of a weir operation, we conducted snorkel surveys and pool follow-up observations to determine the size and distribution of the late-entering segment of the spring chinook run. We felt that the operation of a weir during August and early September, when minimum water temperatures regularly exceed 21 oC, would result in unacceptable fish mortality.
Another significant problem encountered in operating a weir at this time of year, was defining spring-run vs. fall-run chinook salmon (fall chinook), since both are often present at this time. Late-entering spring chinook were identified as those fish which were dark, brassy, and may have had other physical marks indicating they had over-summered lower in the Klamath-Trinity system. Fall chinook were identified as those fish which appeared fresh, bright, nickel-colored, and usually lacked old marks and scars.
Two Alaskan-style weirs were operated in the basin as recovery stations. These weirs were located in Hayfork Creek at Bar 717 Ranch, 8 km upstream from its confluence with the SFTR, and in the mainstem SFTR at Forest Glen Campground (RKM 89.5) (Figure 1). The Alaskan weir also utilized 1.9-cm galvanized conduit panels as the "fence", but the support and orientation of the pipe was markedly different than the Gates Weir. The conduits slid through holes in 7.6-cm-wide by 3.3-m-long aluminum channel, and contacted the natural river bottom. The aluminum channel was supported on tripods constructed of 8.9-cm x 14-cm, and 3.8-cm x 14-cm Douglas fir beams (standard mill-run). The aluminum channel was oriented horizontally and the conduit was oriented vertically. The center-to-center spacing between conduit elements was 5.7 cm, leaving a 2.5-cm gap.
The trap construction was the same as that of the Gates Weir, except that vinyl tubing was not used to sleeve the conduit elements of the Hayfork Creek trap. Fish captured in these traps were netted, examined for marks, scars, and general condition, then immediately released. Artificial slime was also applied to each fish just prior to release.
All weirs were operated 7 days-per-week, 24 hours-per-day. Each was serviced every morning and often staffed 24 hours-per-day during busy holiday weekends.
Digital recording thermographs were used to continually monitor water temperatures at the weir sites. Thermographs were protected inside a steel casing and chained to each weir. Hand-held thermometers were used to check water temperature each morning during routine weir service.
During the summer of 1992, snorkel surveys were conducted during late July and late August, and covered the entire study area (Figures 1 & 2). Our primary goal was to observe and record the numbers of marked and unmarked spring chinook for making population estimates. We also documented the number and location of over-summer holding pools utilized by three or more spring chinook. We also recorded the numbers of marked and unmarked adult spring-run steelhead seen.
We used teams of two to three individuals, equipped with mask, snorkel, wetsuit, anti-slip footwear or fins, notepads, and appropriate safety gear (e.g., rescue rope and first aid kit). We typically entered the river at approximately 9:30 AM and covered 7.0 to 10.5 km of river per-day, depending on the length and difficulty of each river section. Each team floated or swam downstream, recording the number of adult salmonids and the relative abundance of juvenile salmonids. We also noted habitat types and conditions, water temperatures, presence of tributaries and their respective temperatures, and the presence or absence of summer holding habitat. The most difficult task was finding adult fish. We spent a great deal of effort searching beneath undercut rocks, ledges, vegetation, overhangs, etc., where fish often hid to avoid divers. Some sections required a good deal of walking and investigation of pools, step-runs, pocket-water, and other habitat types which afforded good cover.
We surveyed two contiguous river sections per-day, four days-per-week. This year we surveyed the lowest sections first and progressed upstream. We were careful to minimize disturbance to fish so that fish movement from one river section to another, and possible double counting, was negligible.
Once we determined which pools were being utilized by spring chinook, we made follow-up observations of fish at these sites. We used binoculars from a vantage point which afforded a good view, without the fish being aware of us. Almost every pool had an adjacent steep bluff which was ideal for this purpose. Our goals were to determine if fish were moving into or out of the pools, assess summer mortality, make counts and look for tagged and marked fish, and to observe pre-spawning behavior in order to begin our spawning/redd surveys at the appropriate time.
Redd and carcass surveys began in late September and continued through mid-November. We made aerial surveys by helicopter every seven to fourteen days covering the entire river to ensure we were performing ground surveys frequently enough, and to observe overall trends. Each river section was covered more thoroughly by two-person crews, on-foot or in kayaks. When redds were located, their location was documented (by RKM and local landmarks) and each was assigned a specific identification number. We measured overall redd size and position in the stream, water depth, current velocity, and estimated gravel size. We also estimated the percent fines in surrounding gravels and noted various aspects of fish behavior (e.g., female present or absent, evidence of false redd activity, estimated time spent on redd). We repeated the surveys until two consecutive trips noted no new redds or live fish.
The carcass recovery effort was conducted in the same manner as redd surveys and focused on those areas where redds and spawning fish were seen during previous surveys. Carcasses were examined for tags and tag scars, fin clips, spawning success, and signs of predation, and a scale sample was taken. Species, sex, FL, and general condition were also noted. We attempted to correlate each carcass with a known redd. We hoped to determine if redds might actually contain eggs, based on the spawning success of the correlated carcass. We also hoped to determine a tag shedding rate from recovered carcasses.
The angler harvest estimate was based entirely on tag returns and creel surveys. Creel survey days were chosen at random (using a table of random numbers) from two strata, weekends and weekdays. The creel survey covered the river between its mouth and RKM 48, where over-summering spring chinook were most likely to be caught by anglers. Fishing was prohibited upstream of RKM 48, and no creel survey was performed upstream of this point. Angler access sites in the creel survey area were identified prior to the survey period.
Creel survey clerks followed a set route from the predetermined schedule, and examined each access site for anglers. Anglers observed fishing during the survey periods were interviewed for hours fished that day, target species, success, angling method, and county or state of residence. Sport-caught chinook were measured (FL, cm), and examined for fin clips and external tags. The number of any tag observed was recorded, the fish's sex determined, and its spawning condition noted. Scale samples were taken from creeled fish in the same manner as for fish at the Gates Weir.
In-stream life-history patterns were determined from analysis of adult scales, emigrant juvenile trapping, and data from the weirs and snorkel surveys.
Scales obtained from immigrant chinook and from carcass recoveries were cleaned and mounted between two glass microscope slides and examined with a microfiche reader. The number of annuli and patterns on the scale, indicating ocean- or stream-type life history, were noted. An ocean-type life history was indicated by the presence of the first annulus outside the point of ocean entry. A stream-type life history was indicated by the presence of the first annulus inside the point of ocean entry (Snyder 1931; Mills 1986; Sullivan 1989). The point of ocean entry was defined by the first obvious, pronounced increase in the distance between circuli, as measured outward from the scale nucleus. The number of circuli were counted, and the radial distance (mm) measured from the scale focus to the mark indicating ocean entry and to the first annulus, between each annulus, and from the last annulus to the scale margin. Each scale set was examined by two readers and their results compared. If the readers were in agreement, the interpretation was assumed correct. If readers were not in agreement, both readers re-examined the scale set together to determine the correct interpretation.
We monitored juvenile emigration patterns by trapping in the SFTR at Forest Glen, 400 m downstream from the Highway 36 river crossing. We chose this location for three reasons: 1) neither in our field work nor in the literature did we find evidence of fall chinook spawning this far upstream, so we were reasonably sure that any juvenile chinook salmon captured would be spring-run fish; 2) many of the spring chinook redds we documented this season were less than 16 km upstream of this point; and 3) this site afforded easy access and was less subject to high storm-flows than areas farther downstream.
Juveniles were captured using fyke nets attached to trap boxes. The nets were constructed of 1.3-cm nylon mesh, had a 1.8-m x 2.4-m upstream opening, and extended 10.1 m to a trap attachment frame at the terminal end. Trap boxes were constructed of plywood and hardware cloth, and measured 0.8 m wide by 1.2 m long and 0.5 m in depth (vertical dimension). The fyke-net traps were fished overnight, usually about 24 hours, and examined the following morning. To minimize the chances of current-induced fish mortality, we placed two trap boxes in tandem so that the current velocity in the last box was less than 0.3 m-per-second. We also formed an enclosure inside the last trap box using hardware cloth with 1.3-cm holes, which allowed chinook salmon fry refuge from larger Age 1+ and 2+ juvenile steelhead.
Captured fish were identified to species and enumerated. Individual chinook salmon and steelhead were measured for FL (mm). A displacement volume was measured for all chinook salmon caught each day. Scale samples were taken from yearling chinook salmon and some steelhead captured. Flows through the net were measured with a Marsh-McBirney flow meter to estimate the total volume of water sampled. Water temperatures were monitored using hand-held thermometers or digital recording thermographs. When flow conditions permitted, we trapped two nights-per-week beginning 2 February, and increased to three nights-per-week near the end of March. We trapped on this schedule until no juvenile chinook salmon were caught for two successive trapping weeks, and emigration appeared to be complete. Results are reported by trap-night, defined as one juvenile trap, fished for one night.
We determined the number of effectively marked fish by subtracting the number of tagging or marking mortalities recovered at or near the Gates Weir from the number of marked fish. Mortality was considered to be a result of the tagging operation if the fish was discovered dead within 30 days of processing. We did not subtract mortalities discovered during the snorkel surveys from the effectively marked population since some over-summer mortality is normal.
To determine the run-size above the Gates Weir, we used Chapman's version of the Petersen Single Census Method (Ricker 1975):
N = (M+1) (C+1) / (R+1), where
N = estimated run-size; M = number of effectively tagged fish; C = the total number of spring chinook observed during snorkel or carcass recovery surveys, or at recovery weirs; and R = number of weir-tagged and -marked fish which were seen during the snorkel or carcass recovery surveys, or at recovery weirs.
In using this method, we assumed that fish trapped and marked were a random and representative sample of the population; marked and unmarked fish were equally likely to be observed in snorkel and carcass surveys, and captured at recovery weirs; tagged and marked fish were randomly distributed throughout the population; marked and unmarked fish did not suffer differential mortality; all tagged and marked salmon were recognized upon recovery at weirs or during the carcass recovery survey; and that only tagged fish would be recognized during snorkel surveys.
Some data collected are presented in Julian week (JW) format. Each JW is defined as one of a consecutive set of 52 weekly periods, beginning 1 January, regardless of the day of the week on which 1 January falls. The extra day during leap years is added to the ninth week, and the last day of the year is included in the 52nd week (Appendix 2). This procedure allows inter-annual comparisons of identical weekly periods.
