Staff of the California Department of Fish and Game's, Trinity Fisheries Investigations Project conducted a mark-and-recovery, salmon spawner survey of a portion of the mid-Trinity River basin from 18 September 1989 through 26 January 1990. The main-stem Trinity River and its major tributaries were surveyed from the upstream limit of anadromous migration at Lewiston Dam to a point 64.4 km downstream at the confluence of the North Fork Trinity River. We examined 8,785 chinook salmon (Oncorhynchus tshawytscha) and 1,369 coho salmon (O. kisutch) carcasses during the survey.
Chinook salmon spawning was unevenly distributed in the main-stem Trinity River. We estimate 42% of the chinook salmon spawned in the uppermost 3.2 km just below the dam, followed by 25.7% in the next 7.9 km, 10.5% in the next 9.7 km, 8.0% in the next 10.8 km, and 13.8% in the remaining 31.9 km.
Coho salmon spawning distribution in the main-stem Trinity River was similar to that of chinook salmon. We estimate that 43.9% of the coho salmon spawned in the uppermost 3.2 km just below the dam, followed by 23.5% in the next 7.9 km, 7.7% in the next 9.7 km, 3.6% in the next 10.8 km, and 21.3% in the remaining 31.9 km.
Little chinook salmon spawning occurred in the tributaries we surveyed. The North Fork Trinity River and Canyon Creek had the most spawners, with fewer than 150 fish in each. Only 15 coho were observed in the seven tributaries surveyed. Of these, six were found in Weaver Creek.
The percentage of female chinook salmon which died prior to spawning averaged 62.8% for spring-run and 32.1% for fall-run chinook. The overall female chinook salmon prespawning mortality rate during the survey was 31.3%. This excessively high mortality rate is probably related to stress induced by the limited holding and spawning habitat found in the upper main-stem Trinity River.
Approximately 6.2% of the female coho salmon observed in the main-stem Trinity River died prior to spawning.
Both spring-run and fall-run chinook salmon were recovered in the survey. Spring-run fish dominated recoveries until late October, thereafter fall-run fish were more abundant. Coho salmon were first noted in the main-stem Trinity in late October. Their numbers peaked in early December, and they were essentially gone by late January.
Fork lengths of adult spring-run and fall-run chinook salmon from the main-stem Trinity River averaged 74.5 cm (range: 38-100 cm) and 69.8 cm (range: 36-98 cm), respectively. Adult chinook salmon comprised 98.0% of the spring run and 95.0% of the fall run, with grilse comprising the remainder. Chinook salmon recovered in the tributaries were significantly smaller than main-stem Trinity River fish, and had a lower overall adult percentage of 79.6%. Coho salmon were not measured during the survey.
1. To determine, through a system of spawning ground surveys, the distribution of naturally spawning chinook and coho salmon in the main-stem Trinity River and its tributaries upstream of, and including the North Fork Trinity River.
2. To determine the incidence of pre-spawning mortality among naturally spawning salmon in this area.
3. To determine the size, sex composition and incidence of marked and tagged individuals among the naturally spawning populations in this area.
4. To determine the relative distributions of spawners in different areas of the basin up-stream of, and including the North Fork Trinity River.
This year the California Department of Fish and Game's (CDFG) Trinity Fisheries Investigations Project (TFIP) completed the twenty-second salmon spawner survey conducted in the main-stem Trinity River since 1942. The first three surveys (Moffett and Smith 1950, Gibbs 1956, and Weber 1965) were fishery evaluations prior to the construction of Lewiston Dam. The remaining eighteen (La Faunce 1965, Rogers 1970, 1973, 1982; Miller 1972, 1973, 1974, 1976, 1978, 1979, 1980, 1981, 1982, 1984, 1985; Smith 1975, Stempel 1988, and Zuspan 1991a) were designed to evaluate the effects of the existing dam on the salmon resource.
In 1984, The Trinity River Basin Fish and Wildlife Management Program was enacted by Congress (Public Law 98-541). This law appropriated approximately $57 million to be spent for fishery and wildlife restoration, and monitoring within the Trinity River basin.
This survey, and those scheduled for following years by CDFG's TFIP, will help to evaluate the effectiveness of increasing spawning and holding habitat within the basin through habitat improvement efforts that are part of the restoration program.
Our study area included the main-stem Trinity River from its upstream limit to anadromous fish migration at Lewiston Dam (River km 180.1) to the confluence of North Fork Trinity River, 64.4 km downstream (Figure 1). Previous studies have divided the river into either a four-or seven-zone system. The seven-zone system was used in 1987 by the United States Fish and Wildlife Service (USFWS) (Stempel 1988) and again in 1988 (Zuspan 1991a) (Table 1) . Prior to this, with the exception of Moffett and Smith 1950, all surveys were based on a system utilizing four zones in the river reach below Lewiston Dam (Gibbs 1956; La Faunce 1965; Rogers 1970, 1973, 1982; Miller 1972, 1973, 1974, 1976, 1978, 1979, 1980, 1981, 1982, 1984, 1985; Smith 1975; Weber 1965). Our 1989-1990 data were collected based on both zone systems. We will summarized data in this report based only on the seven-zone system as it allows comparisons of different river sections in finer detail. By recording data also using the four-zone system, we will be able to compare historic and current trends in subsequent reports.
(Figure 1) (Table 1)
TFIP staff conducted the survey using 12-ft Avon inflatable rafts equipped with rowing frames. Raft crews consisted of a rower, and one or two personnel to recover carcasses. To increase coverage of the highly productive upper two sections, two rafts were used simultaneously, with one covering each side of the river. Carcasses were recovered on foot along the shore or, in deep water, from the rafts with long handled gigs. We surveyed the entire main-stem Trinity River study section once a week throughout the salmon spawning season.
We determined spawning condition in female salmon by direct observation of the ovaries. Fish were classified as either spawned or unspawned based on egg retention. Females which retained over 50% of their eggs were classified as unspawned. Male spawning condition was not assessed, as its determination was considered to be too subjective.
All carcasses we observed were identified by species and examined for an adipose fin clip (Ad-clip) indicating the presence of a coded-wire tag (CWT) in their snout. Fish were further examined for the presence of an external tag (spaghetti tag) and an operculum punch, applied as part of an ongoing study by other elements of the CDFG's Klamath-Trinity Program. Spaghetti tags and operculum punches (Program marks) are placed on returning adult fish by CDFG staff at three trapping and tagging stations downstream of the spawner survey area, to monitor escapement and harvest of returning adult salmonids. The spaghetti-tagged salmon also receive an identifying operculum punch in order to estimate tag shedding rates at the three sites. The first site is located at the mouth of the Klamath River where returning fall-run chinook salmon, coho salmon, and steelhead trout are captured in a seine and tagged. The second site upstream is Willow Creek Weir, located at river km 32.2 on the main-stem Trinity River. The last site is Junction City Weir at river km 136.5 on the main-stem Trinity River. Spring-run and fall-run chinook salmon, coho salmon, and steelhead are trapped and tagged at both Willow Creek and Junction City weirs.
We classified all chinook salmon carcasses as either condition one or two, based on the extent of body deterioration. Condition-one fish were the freshest, having at least one clear eye and a relatively firm body. Condition-one fish were assumed to have died within one week prior to recovery. Condition-two fish were in various advanced stages of decomposition and assumed to have died more than one week prior to recovery. We did not count partially intact fish skeletons, because they could have represented Program-marked or condition-two fish which had already been counted and chopped in half during a previous week's survey.
All chinook salmon we recovered were further classified into four categories: 1) Ad-clipped fish; 2) Program-marked fish; 3) condition-one, unmarked fish; 4) condition-two, unmarked fish. The category assigned determined what data we collected from each fish.
We sexed and measured Ad-clipped fish to the nearest cm fork length (FL), and determined their condition and spawning success. Heads of Ad-clipped fish were removed and retained for later CWT recovery and decoding.
Program-marked fish were sexed, measured (cm, FL), and assessed for spawning condition. We removed any spaghetti tags and then cut the fish in half with a machete to prevent recounting in future weeks. Spaghetti tags have a unique number which allowed determination of date and location of tagging.
Condition-one fish which were neither Ad-clipped nor Program-marked were flagged and returned to moving water for subsequent recovery, and a systematically collected subsample of them were measured for FL (cm). Flags consisted of plastic survey tape wrapped tightly around a colored hog ring and affixed to the left mandible of the carcass. The survey tape was wrapped so tightly around the hog ring, that it amounted to no more than a colored coating, with less than 2.5 cm of tape extending from the hog ring at any time. Flag colors were changed weekly so that, on recovery, the week of flagging could be determined. The hog rings used to attach the flagging were color coded to indicate in which zone they were affixed, so that we could determine the incidence of carcasses drifting into another recovery zone. Chinook < 56 cm were preliminarily classified as grilse during the carcass surveys. Actual grilse to adult ratios for the whole population of chinook in this year's run were determined from post-season evaluations of CWT data. Adult and grilse salmon analysis in this report is based on the post-season size determinations.
Condition-two fish which were neither Ad-clipped or Program-marked were checked for the presence of a flag and, if possible, the spawning condition was assessed. If a flag was present, the color of the flagging tape and the underlying ring were recorded, and all fish were then cut in half to prevent later recovery and re-counting of the same fish.
All coho salmon collected were checked for the presence of Ad-clips or Program-marks. When possible, sex and spawning condition were determined and then all coho salmon were cut in half to prevent future re-counting. Coho carcasses were not used in the flagging experiment, since they would have required a separate series of flag colors to segregate them from flagged chinook salmon.
Tributaries to the main-stem Trinity River, specifically Rush Creek, Grass Valley Creek, Indian Creek, Reading Creek, Browns Creek, Weaver Creek, Canyon Creek, and the North Fork Trinity River, were surveyed on foot once a week throughout the chinook salmon spawning season. Sections surveyed for each tributary ranged in length from 2.4 to 4.8 km, and were chosen based on accessibility and their historic use by chinook salmon spawners (Figure 1). The survey began with the onset of chinook salmon spawning in each tributary and continued until spawning ended (Table 2.)
We classified all identifiable chinook salmon recovered into the four categories used in the main-stem spawner survey and handled them accordingly (see above). However, sex and prespawning condition was assessed only for fish collected from the main-stem Trinity River, and not its tributaries, because too few fish were observed in the tributaries to compose an adequate sample and most of those observed were condition-one fish which we needed to flag for spawning escapement estimates. Coho salmon were counted and cut in half upon recovery. Chinook salmon redds, when observed for the first time, were counted and recorded.
Aerial flights and ground-truthing surveys were made of each tributary to determine the percentage of the total available spawning area within each tributary represented by each of our ongoing spawner survey zones. Flights were made during the peak of spawning activity to observe redds and locate the upstream limit of spawning. Follow-up ground-truthing surveys were made, when necessary, to make total redd counts for both the whole tributary and its spawner survey zone. The percentage of the total redds occurring in a survey zone during the aforementioned count was assumed to represent the percentage of the total spawning in each tributary that took place within the survey zone.
We observed 8,785 chinook salmon during the spawner survey which included 195 Ad-clipped fish (six also program-marked), 218 Program-marked fish (six also Ad-clipped), 4,886 unmarked condition-one fish which we flagged, and 3,495 unmarked condition-two fish (Appendix 1). We recaptured and re-examined 2,270 flagged chinook salmon, but did not see any whole skeletons.
We recovered 1,360 coho salmon in the spawner survey, including three Ad-clipped fish and 75 Program-marked fish (Appendix 2), and did not see any whole skeletons.
We observed 127 chinook salmon in the eight tributaries surveyed this season. Included in the total are 5 Ad-clipped fish, 9 Program-marked fish, 115 unmarked condition-one fish which we flagged, and 12 skeletons (Appendix 3). We recaptured and re-examined 50 flagged fish.
We recovered nine coho salmon in the tributaries this season (Appendix 3), but observed no whole skeletons.
Only chinook salmon recovered in the main-stem Trinity River were used to determine spring and fall spawning interval. Both spring and fall race chinook salmon were observed in the main-stem survey. A date separating the two races was determined from CWT and Program-marked chinook salmon. Spring-run chinook salmon dominated our recoveries through the fourth week of the survey ending 9 October 1989. Some overlap of spring-run and fall-run chinook salmon occurred during the fifth week ending 16 October. Fall-run chinook salmon became dominant by the sixth week of the survey, which began 23 October. For the purposes of this report, all chinook recovered prior to 23 October 1989 are considered spring race, while those recovered afterwards are considered fall race (Figure 2).
Chinook salmon spawning interval determined from weekly carcass recoveries of coded-wire-tagged and Program-marked fish in the 1989-90 Trinity River spawner survey. The arrow indicates the date separating the spring from the fall run.
We measured the size (cm, FL) of 607 spring-run chinook salmon during the survey. Adults (fish >48 cm FL as determined from post-season analysis of length frequency and coded-wire tag recovery) (Bill Heubach, Assoc. Fishery Biologist, CDFG, pers. commun.) comprised 98.0% (595/607) of the spring-run chinook salmon observed in the spawner survey, while grilse (fish <48 cm FL) comprised the remaining 2.0% (12/607). (Table 3, Figure 3). The percentage of spring-run chinook salmon grilse in the survey closely matched that observed at Junction City Weir but varied from that observed at Willow Creek Weir and Trinity River Hatchery (Table 3). The reason for the difference in grilse percentages at the different sites is unknown.
(Table 3)(Figure 3)
Based on the date at which we first observed spawning activity and the lack of spring-run codes among the CWT recoveries, we assume that no spring-run chinook salmon were observed in the tributaries.
We measured the FL's of 634 fall-run chinook salmon. Based on a minimum FL of 52 cm for adults (Bill Heubach, Assoc. Fishery Biologist, CDFG, pers. commun.), 99.3% of the fall-run chinook salmon measured were adults and 4.7% were grilse. (Table 4, Figure 4). For comparison, the percentage of fall-run chinook salmon grilse at the different sampling sites ranged from 2.1% to 20.2% (Table 4). As with spring-run chinook salmon grilse, the reason for the difference in rates between the sample sites is unknown.
We measured 114 chinook salmon from the tributaries. Adults comprised 79.8% of the chinook observed while grilse comprised the remaining 20.2% (Table 4, Figure 5). The percentage of grilse observed in the tributaries was significantly different (X2=35.0, df=1, P=.0001) than observed in the main-stem Trinity River spawner survey. The higher percentage of grilse in the tributaries may have been due to competition with larger fish for prime spawning locations in the main-stem Trinity River, or larger fish may have found it harder to enter tributaries during the low flow conditions encountered this year.
(Figure 4) (Table 4) (Figure 5)
Sex was determined only for fish recovered from the main-stem Trinity River that were either condition-two unmarked fish, Program-marked fish, or flagged fish recaptured in the carcass survey.
We determined the sex of 1,877 adult chinook salmon during the survey (498 spring-run and 1,379 fall-run). Both spring and fall adult chinook salmon runs had more females than males (69.9% and 58.2% females, respectively). The percentage of females in the survey was highest during the early and late weeks of the survey and lowest during the middle weeks (Figure 6). The preponderance of females in the adult run has been noted in all but two of the previous surveys and has ranged from 73.6% to 25.8% (Appendix 4). The predominance of females among adult fish results when males return as grilse, thereby decreasing the number of males left to return as adults.
Sex was determined for 1,282 coho salmon, 57.2% (733) of which were females. No attempt was made to differentiate adults from grilse for coho salmon. For comparison, only 42.4% of the coho examined last year were females (Zuspan 1991a). In a pattern similar to that observed for chinook salmon, female coho salmon were most prevalent early and late in the survey (Figure 7).
Prespawning mortality was determined only for fish recovered in the main-stem Trinity River that were either condition-two unmarked fish, Program-marked fish, or flagged fish recaptured in the carcass survey.
We checked the spawning condition of 2,531 adult female chinook salmon this season (521 spring-run and 2,010 fall-run fish). Prespawning mortality was 62.8% and 23.1% for spring-run and fall-run chinook salmon females, respectively. The rate of prespawning mortality decreased through time, starting at 92.2% and gradually decreasing to 11.1% by the end of spawning (Figure 8). The higher prespawning mortality rate for spring-run chinook salmon females is probably related to the added stress imposed by the extended time they spend in the river.
The overall prespawning mortality rate of both races of female chinook salmon was 31.3%. Overall (spring-run and fall-run) prespawning mortality for female chinook salmon has ranged from 1.5% to 44.9%, averaging 11.7% during previous surveys (Appendix 5).
Spawning condition was determined for 689 adult female coho salmon during the survey. The overall prespawning mortality rate was 6.2% (43/689). In 1988 prespawning mortality for female coho was 25.6% (46/180) (Zuspan 1991a). Coho prespawning rates have not been reported in surveys prior to 1988. The weekly rate of prespawning mortality varied between 20% and 4% during the survey (Figure 9).