Comprehensive Monitoring, Assessment, and Research Program

CALFED=s

Comprehensive Monitoring, Assessment, and Research Program

for Chinook Salmon and Steelhead

in the Central Valley Rivers.

This plan describes the comprehensive monitoring and research needs to determine whether the habitat for chinook salmon (Oncorhynchus tshawytscha) and steelhead (Oncorhynchus mykiss) in Central Valley rivers will be restored by the CALFED=s Ecosystem Restoration Program, other CALFED programs, and other restoration programs, such as the CVPIA Anadromous Fish Restoration Program. The rivers included in this plan are the San Joaquin River tributaries, including the Stanislaus, Tuolumne, and Merced rivers; the Eastside Rivers, including the Mokelumne, Cosumnes, and Calaveras rivers; the American, Feather, Yuba and Bear rivers; the Dry Creek basin; and the upper Sacramento River and its tributaries. All fishery biologists, including agency employees, irrigation district consultants, and watershed advisory members, that have experience with the San Joaquin, Eastside, American, Feather, Yuba, and Bear river basins were asked to contribute to this plan. Study issues for the upper Sacramento River and its tributaries were provided by agency biologists working on these rivers and from descriptions of stressors described in the Review Draft of the Ecosystem Restoration Program Plan (CALFED 1997) and Volume 3 of Working Paper on Restoration Needs (U.S. Fish and Wildlife Service 1995).

When considering studies on salmon and steelhead, other plans in this appendix should be reviewed. This plan identifies studies on the relationships between fish, their habitat, and ecological processes that are more fully described in other plans in this appendix such as AFluvial Geomorphology and Riparian Issues@, AFishes in Shallow Water Habitats@, ABenthic Macroinvertebrate Communities@, and AWater Quality in the San Joaquin Basin@. There is a separate plan for steelhead in this appendix that more fully describes the level of existing knowledge and the need for future monitoring and research. The monitoring and research needs for salmon and steelhead rearing in and migrating through the Delta are described in another document, AJuvenile Salmon in the Delta@, in the appendix. The ARiver Resident Fishes Species@ plan includes study methodologies that are similar to those described here for salmon and steelhead, and these studies should be coordinated.

I. MONITORING OBJECTIVES

1

CMARP - Fall-Run Chinook Salmon and Steelhead January 8, 1999

The Ecosystem Restoration Program has the goal to improve and increase aquatic and terrestrial habitat and improve ecological functions in the Bay-Delta to support sustainable populations of diverse and valuable plant and animal species. Within the Ecosystem Restoration Program, there are four categories of implementation objectives that include (1) Ecological Processes; (2) Habitat; (3) Species and Species Groups; and (4) Stressors. Within the third category, Species and Species Groups, the implementation objective for the chinook salmon sub-program is to ensure the recovery of the Sacramento winter-run chinook salmon, a species listed as endangered under the federal and California Endangered Species Acts in order to ensure overall species richness and diversity and reduce conflict between the need for its protection and other beneficial uses of water in the Bay-Delta. The objective is also to ensure the restoration of Sacramento fall-run, spring-run, late-fall-run, and San Joaquin fall-run chinook salmon is adequate to support sport and viable commercial fisheries.

The objectives of the monitoring proposed by this salmonid team are twofold:

1)To document whether implementation of the CALFED common programs, primarily the Ecosystem Restoration Program, results in an increase in the abundance of juvenile and adult chinook salmon in the San Joaquin tributaries and eastside rivers; and

2)To monitor indicators consisting of Akey@ features of salmon health, habitat quality, and ecosystem processes to provide a basis for adaptive management.

II. CONCEPTUAL MODEL: HYPOTHESES AND ASSUMPTIONS OF THE SYSTEM

The following conceptual model was developed to help define hypotheses and assumptions regarding the Ecological Restoration Program. The model is based on the ecological interactions between habitat features, stressors, and specific life stages of salmonids. The model begins with the upstream migration of adult salmon and steelhead, followed by spawning, egg incubation and emergence, rearing, juvenile migration, and ocean residence. Figure 1 shows an example of the ecological interactions for salmonids.

A. Upstream Migration of Adults

The upstream migration of adult fall-run chinook salmon through the Delta to their natal streams is thought to begin earlier for Sacramento basin populations (Gerstung 1971) than for Eastside and San Joaquin basin populations. Adult San Joaquin fall-run chinook salmon begin to enter the western Delta near Jersey Point in September and they migrate upstream slowly, typically entering the San Joaquin tributaries in late October or early November and continuing to migrate into the rivers through December (Hallock et al. 1970; Department of Fish and Game annual reports; Carl Mesick Consultants 1998a). Video and trapping data at Woodbridge Dam in 1990 and 1991 indicate that most adults migrate into the Mokelumne River from late October through December (BioSystems Analysis, Inc. 1992), more recent studies during higher flows indicate that adults begin to enter the river in mid August (EBMUD 1998 unpublished data). Adult salmon have been trapped in the northern Delta in August and September and most of these fish are probably migrating toward the Sacramento River basin. Sport fishing for adult chinook salmon in the American River is frequently intensive in September and October. Adult migration is also reported to begin in September when attraction flows are adequate in the Yuba River (Department of Fish and Game 1991). In the Feather River, adult salmon first appear in August

1

CMARP - Fall-Run Chinook Salmon and Steelhead January 8, 1999

Figure 1. Diagram of potential stressors for each life stage of chinook salmon and steelhead.

1

and migration peaks in September through November and continues into December (Ted Sommer, Department of Water Resources, personal communication).

Steelhead are reported to begin their upstream migration into the American, Feather, Yuba, and Mokelumne rivers in August through October depending on water temperature, weather conditions, and flow, and peak migration occurs in November and December (Department of Fish and Game 1991; Busby et al.1996; Joe Merz, EBMUD personal communication; Ted Sommer, Department of Water Resources, personal communication). However, the data on migration timing are inconclusive because surveys from December to April are frequently hampered by high flows and turbid water, cessation of hatchery operations that monitor fish movement, and the difficulty in locating steelhead carcasses because many adults leave the spawning reaches before they die (Dennis McEwan, Department of Fish and Game, personal communication; Joe Merz, EBMUD, personal communication). Steelhead surveys are not conducted in the Stanislaus River and most of the information on run timing is provided by anglers. In May 1997, a four- and a six-pound steelhead with brightly colored red stripes and opercula were caught by anglers just downstream of Knights Ferry (Jason Anderson, U.S. Army Corps of Engineers, personal communication) and so the migration period could be quite long in the Stanislaus River. The stressors that affect fall-run chinook salmon probably also affect steelhead except that the critical period may be year-round for adult steelhead.

Important considerations include (1) straying of adult salmon and steelhead from the San Joaquin tributaries and Eastside Rivers to the Sacramento basin due to inadequate flow for attraction through the Delta; (2) stressors, such as high temperatures, that reduce gamete viability; (3) delays in adult migration that delay juvenile outmigration until periods when stressors are worsening, (4) migration barriers in the eastside tributaries and in the South Delta; pre-spawning adult mortality caused by poaching, disease, and thermal stress; and (5) migration barriers, particularly large dams without ladders, that prevent the adult salmon and steelhead from reaching upstream historical spawning areas.

Adult Straying: A preliminary analysis of coded-wire-tagged (CWT) Merced River hatchery chinook salmon that were released in the San Joaquin basin as juveniles and recovered as adults from 1983 to 1996 relative to Vernalis flows and Delta exports was conducted by Carl Mesick Consultants (1998a). The concern was that Delta export rates were increased in fall 1996 and in subsequent years to Amake-up@ for reduced pumping rates during the spring outmigration period. A high number of Merced River CWT fish were recovered in the Sacramento basin as adults in 1987, 1988, and 1989 when San Joaquin River flows were low and Delta export rates were high. By extrapolating the tag recoveries to the entire populations for all the rivers surveyed in the Central Valley, straying rates were estimated to be less than 5% when Delta export rates were less than about 300% of San Joaquin River flow at Vernalis in mid-October. However during mid October in 1987 through 1989, export rates exceeded 300% of Vernalis flows and straying rates increased to a maximum of 18%. Pulse flow releases from the San Joaquin tributaries for 8 to 10 days in mid-October since 1993 appear to have kept straying rates below 5%. The accuracy of the straying estimates is in question because accurate records of the number of adult fish examined for CWTs are not available, particularly for the San Joaquin tributaries.

1

CMARP - Fall-Run Chinook Salmon and Steelhead January 8, 1999

A similar analysis with CWT fish reared at the Mokelumne Fish Hatchery and released as juveniles in the Mokelumne River has not been conducted. It is possible that high export rates that result in Areverse flows@ in the San Joaquin near Antioch may cause adult chinook salmon from the eastside rivers to stray as well.

Hypothesis A-1: High Delta export rates relative to flow in the San Joaquin River (at Vernalis) and in the Delta (Qwest) in mid-October increase the percentage of adult chinook salmon and steelhead from the San Joaquin and eastside tributaries that stray into the Sacramento basin.

A recent study by Habicth et al. (1998) suggests that half-length CWT inserted near the olfactory organs and nerves of fry (0.2 g) may have caused straying in adult pink salmon returning to Prince William Sound, Alaska. Currently only large juveniles are tagged in the Central Valley and straying rates have been low when attraction flows are adequate and so there does not appear to be a problem. However, tagging programs for hatchery reared fish are expanding and it is possible that fry may be tagged in the future. If so, olfactory damage caused by tagging will be a concern.

Hypothesis A-2: CWT placement can affect the homing ability of chinook salmon and steelhead.

Delayed Migration: Hallock et al. (1970) showed that radio-tagged adult chinook salmon delayed their migration at Stockton whenever D.O. levels were less than 5 mg/l in October during the 1960s. Hallock also reported that D.O. levels usually increased to suitable levels by November. D.O. levels near Stockton in October and November had been greater than 5 mg/l from 1983, when DWR began monitoring, to 1990. Then in 1991 and 1992, D.O. levels were substantially lower than 5 mg/l for most of October. The Head of the Old River Barrier was installed in fall 1992, but it did not correct the problem. In 1993, D.O. levels were low until about 10 October and it is likely that pulse flows that raised Vernalis flows to about 4,000 cfs on 7 October were responsible for increasing D.O. levels at Stockton. Similarly in 1994, D.O. levels were low until 15 October when pulse flows raised Vernalis flows to about 2,000 cfs. In 1995, D.O. levels were near 5 mg/l in mid to late September until Vernalis flows increased from about 3,000 cfs to 6,000 cfs through mid October. Low D.O. levels also occurred in 1996 until 12 October when pulse flow releases increased Vernalis flows from 2,000 to about 3,000 cfs.

Previous studies (McCarty 1969) and a recent evaluation performed by Jones & Stokes Associates indicate that the low D.O. levels in the ship channel during summer and early fall months are partly (if not primarily) a result of the decomposition of algal biomass that is produced in the comparatively shallow, nutrient-rich water upstream of Mossdale and subsequently transported into the much deeper waters of the ship channel. The algae, mostly diatoms, are not adapted to deep water conditions and quickly settle out and decompose on the streambed. Simulations performed using the City of Stockton=s DO model (Schanz and Chen 1993) indicate that increasing flow at Vernalis with the head of Old River barrier closed generally improves DO conditions at Stockton during most months. But in October, warm temperatures and the DO demand exerted by ammonia from the Stockton wastewater plant, the rotting algal biomass and other organic matter, usually keep DO levels well below the 6 mg/L standard.

1

CMARP - Fall-Run Chinook Salmon and Steelhead January 8, 1999

The chlorophyll levels at Vernalis are literally among the highest ever recorded for streams worldwide and much of this production may be fueled by a few easily controlled feedlot operations in the catchment. On the other hand, it is possible that the nutrient loading responsible for the high algal production stems from much more diffuse processes, such as tile drainage from row crops or orchards. An EPA-style TMDL (ATotal Maximum Daily Load@) analysis for nutrients (especially phosphorus) for the San Joaquin catchment would be the first step toward resolving these issues.

The Stockton DO model does not yet explicitly include algae, however, so its predictions about the effects of increased flow should be viewed with healthy skepticism. It is conceivable that under some circumstances sending more Vernalis water to the ship channel could make matters worse by increasing its organic matter loading rate. Ideally, the continuous monitoring stations upstream of the ship channel would be equipped with fluorometers calibrated to measure chlorophyll concentration (an indirect measure of algal biomass). Such a system would alert

managers when algal biomass levels at Vernalis or further upstream are extremely high and give them time to take appropriate action.

Under most circumstances, the loading of algal biomass produced naturally in the San Joaquin river upstream is probably a much more serious problem for DO in the ship channel than organic matter loading from the Stockton wastewater treatment ponds. The loading of dissociated ammonium from the wastewater facility, however, may pose a potential toxicity problem. When algae are abundant and DO upstream becomes supersaturated (due to photosynthesis), pH levels also increase. High pH and high ammonium concentration lead to higher levels of undissociated ammonia, which is toxic to fish and aquatic invertebrates. It is possible that the salmon are responding to this toxicity rather than to low DO.

Hypothesis A-3: Low dissolved oxygen levels (<5 ppm) in the mainstem San Joaquin River near Stockton in mid-October delay upstream migrating adult chinook salmon on their way to spawn in the San Joaquin tributaries.

Hypothesis A-4: Low dissolved oxygen levels (<5 ppm) near Stockton result from decomposition of algal biomass produced from upstream of Mossdale. This high upstream productivity is supported by excessive nutrient-loading from diffuse agricultural sources and feedlot effluent.

Hypothesis A-5: Low D.O. levels near Stockton result from localized nutrient loading, such as effluent from the City of Stockton=s sewage treatment plant at Rough and Ready Island, effluent from canneries near Stockton, fertilizers flushed through the storm drains at the Port of Stockton, and dairy farms near Mossdale.

Hypothesis A-6: High levels of undissociated ammonia in the mainstem San Joaquin River near Stockton in mid-October delay upstream migration of adult chinook salmon on their way to spawn in the San Joaquin tributaries.

1

CMARP - Fall-Run Chinook Salmon and Steelhead January 8, 1999

There is concern that pulse flow releases in mid October to attract adult salmon may cause the fish to enter the rivers earlier than normal which may expose them to high water temperatures when the pulse flows cease. Hallock et al. (1970) reported that the number of adults trapped at the Orange Blossom Bridge (rivermile 47) in 1965 and 1967, when there were suitable D.O. levels near Stockton during migration, gradually increased from mid October to a high in mid November, when suitable temperatures typically occur only in the upstream reaches. In recent years, the timing of the runs appears to have been delayed by about two weeks, particularly in the Stanislaus River, where typically few fish enter the river by late October. Although the effect of reduced flows may be responsible for the delay, Thomas R. Payne & Associates (1997) reported that even when D.O. concentrations were about 7 ppm and Vernalis flows ranged between 6,000 and 7,000 cfs during the first half of October 1995, few salmon were observed in the Stanislaus River above Riverbank (rivermile 34) before early November. Instead of flow, the appearance of salmon on the spawning riffles in Stanislaus River has coincided with the first major storms (i.e., declining barometric pressure and air temperatures without flow changes), particularly in 1997 when many fish arrived in mid October. Although the adults also tend to arrive at the spawning riffles in substantial numbers in late October or early November in the Tuolumne and Merced rivers, salmon were observed migrating in the San Joaquin River at the Hills Ferry Barrier near Newman in late September and early October 1995 (Department of Fish and Game 1997). These observations along with those of Hallock et al. (1970) suggest that adult salmon migrate through the Delta to the spawning riffles at a very slow rate, taking approximately two months to migrate from Antioch to the spawning riffles. If true, then migratory cues such as pulse flows or storm related influences (e.g., barometric pressure) could influence the salmon anywhere along their migration route and their response may not be observed in the tributaries for several weeks.