|
|
|
|
|
|
|
The Stanislaus River Restoration Plan Information Site
Home About CRRF Accomplishments Current Projects Membership Info Conservation Partners Contact Us
4.1 INTRODUCTION:
The following conceptual models are summaries of the available evidence and the "best guesses" of the members of the Stanislaus Fish Group as to how habitat degradation and other limiting factors have caused the decline of Chinook salmon and steelhead in the Stanislaus River. The conceptual model is the centerpiece of this restoration plan because it identifies important data gaps in our understanding of the ecological processes that led to the fishes' decline in the lower Stanislaus River. It also highlights the obvious problems that if corrected should result in a measureable increase in the abundance of the target species.
The model helps identify data gaps in two ways. First, it provides a summary of all the evidence regarding the causes that led to the fishes' decline. Second, it includes all the alternative theories presented by various stakeholders of what caused the decline of the fish populations. By viewing the model as a set of interlinked hypotheses, it provides the guidance needed to design the monitoring, focused research, and pilot restoration actions to resolve uncertainties to the satisfaction of all stakeholders and to determine the most effective way to restore and manage the river. Important data gaps identified within the conceptual model are categorized as high priority.
The model is also useful for identifying actions for which there is sufficient evidence that they should provide measureable increases in fish abundance. Even when there is sufficient evidence that a problem should be corrected, most projects have been designed with an adaptive management approach to help refine our restoration methods. These projects are also recommended as a high priority for funding and implementation.
Conceptual models were developed for the life stages for each of the three target species in the lower Stanislaus River: spring-run Chinook salmon (Section 4.2), fall-run Chinook salmon (Section 4.3), and steelhead (Section 4.4).
4.2 SPRING-RUN CHINOOK SALMON:
Although spring-run salmon probably declined in response to the hydraulic mining in the Stanislaus River watershed and the dredging of the Delta during the late 1880s, it is highly likely that their populations have not rebounded because the upstream dams block the migration of adult fish into the upper watershed. Migration into the upper watershed during high spring flows may have been critical for spring-run because it separated them from the fall-run salmon that were confined to the lower watershed by low fall flows. Although the habitat downstream of Goodwin Dam should provide suitable habitat for spring-run due to the year-round hypolimnetic releases of relatively cold water from New Melones Reservoir, competition for spawning habitat and genetic contamination with fall-run are believed to prevent the recovery of spring-run confined to the lower Stanislaus River.
Other factors that depress spring-run Chinook salmon populations are probably similar to those that impact fall-run Chinook salmon and are discussed in Section 4.3.
4.2.1 Adult Upstream Migration:
It is likely that flows and water temperatures are adequate for adult spring-run to complete their migration into the Stanislaus River between mid-February and mid-May. Hallock and others (1970) reported that tagged fall-run adults migrating in the San Joaquin River usually moved past Stockton when water temperatures dropped to less than 66 oF. Water temperatures in the San Joaquin River near Vernalis were usually less than 65 oF prior to mid-April when Vernalis flows exceeded about 1,500 cfs between 1962 and 1994 (Mesick 2001a). Water temperatures should also be less than 65 degrees between late-April and mid-May (extended to June 1 in wet years) due to thirty-day pulse flow releases of about 1,500 cfs from each San Joaquin River tributary that began in 1993 to enhance the survival of juvenile fall-run salmon migrating through the Delta.
However after the pulse flows cease in mid-May, the migration of late-arriving adults may be delayed or blocked due to unsuitably warm water temperatures in the lower San Joaquin River. Water temperatures at Vernalis frequently exceeded 65 oF in late May regardless of the streamflow (up to about 30,000 cfs) at Vernalis and presumably this would block the passage of adult salmon in the Delta near Stockton where D.O. and water quality are poor (Mesick 2001a). On the other hand, migrations may be possible intermittently when water temperatures are below 65 oF for several days at a time presumably due to Delta breezes and reduced air temperatures (Mesick 2001a). If the San Joaquin River basin populations of spring-run salmon had similar migration patterns to those in the Sacramento River basin, then up to one-third of the spring-run population in the Stanislaus River would have migrated through the Delta after mid-May. The extent of this impact depends on (1) whether adult spring-run are able to complete their migration into the Stanislaus River during the brief periods when water temperatures are below 65 degrees in the San Joaquin River after mid-May; and (2) whether delays of migrating fish in the Delta due to warm temperatures would harm the unfertilized gametes in the adult fish. These questions cannot be answered at this time.
Another concern is that excessive exports in the Delta at the Central Valley Project (CVP) and State Water Project (SWP) may increase adult straying rates. The percentage of coded-wire-tagged San Joaquin fall-run Chinook salmon that strayed to the Sacramento River basin or one of the eastside streams increased up to about 20% whenever more than 400% of the San Joaquin River flow at Vernalis was exported from 1979 to 1996 (Mesick 2001b). A 10-day pulse flow released from the San Joaquin River tributaries that resulted in export rates less than 3% in mid-October, which was the peak migration period in the Delta, was sufficient to minimize straying. It is reasonable to assume that if exports exceed 400% of San Joaquin River flows during the upstream migration period, then migrating adults might stray to the Sacramento basin or they could be delayed until the spring pulse flow. Straying of adult spring-run to the Sacramento basin jeopardizes the San Joaquin basin populations by reducing the number of San Joaquin basin juveniles produced.
4.2.2 Adult Holding Habitat:
CDFG (1998) describes spring-run holding habitat as consisting of (1) deep pools, (2) adequate cover, such as bubble curtains created by flowing water, (3) proximity to spawning gravel, and (4) adequate water temperatures and dissolved oxygen concentrations. The high recreational use of the Stanislaus River by rafters probably increases the importance of cover or particularly deep pools for holding habitat. Although the long deep pools in Goodwin Canyon do not contain spawning gravel, a few spring-run were observed there in summer 2000 (Fisheries Foundation 2002). Holding habitat currently exists at numerous sites in Goodwin Canyon, six of the Knights Ferry Gravel Replenishment sites near Lovers Leap (riffles R13, R14, R14A, R15, R19, and R19A), and four riffles adjacent to deep mine pits near Frymire Ranch (RM 53.4), "Willms Pond" (RM 51.8), and Button Bush Park (RM 48).
It is likely that summer water temperatures were suitable for holding adult salmon in recent years, except perhaps 1991. Mean daily flows at the Orange Blossom Bridge (DWR gage) were usually greater than 200 cfs from June through September between 1991 and 2001; however, flows were near 110 cfs for about 45 days during summer 1991. The US Bureau of Reclamation Stanislaus River Basin Water Temperature Model (Rowell 1993) predicts that at a base flow of 200 cfs, weekly mean water temperatures at the Orange Blossom Bridge would range between 59.5 and 66.8 oF from June through September when the temperature of releases from Goodwin dam were moderate. Rowell (1993) did not model temperatures for base flows less than 200 cfs. USGS measurements indicate that hourly water temperatures ranged between 55 and 65 oF at the Orange Blossom Bridge during summer 2001 when flows ranged between 287 and 840 cfs. The Biological Opinion for the USBOR operations for New Melones Reservoir (NMFS 2002) directs that sufficient flows are to be released to maintain water temperatures below 65 degrees (at the Orange Blossom Bridge or below Goodwin Dam?). Overall, it is highly unlikely that either the quality or quantity of adult holding habitat currently limits the spring-run population in the Stanislaus River.
4.2.3 Spawning and Incubation Habitat:
It is likely that fall-run salmon destroy many of the spring-run redds because spring-run spawn first which leaves their redds vulnerable to superimposition by fall-run salmon. Spring-run typically spawn in September and October, whereas fall-run salmon spawn between late-October and December. Redd superimposition by fall-run salmon would probably kill many of the spring-run eggs or entomb spring-run alevins by burying the spring-run redd with silt and sand (CMC 2002b). When fall-run escapement was approximately 8,500 fish in fall 2000, redd superimposition from fall-run salmon completely disturbed 33% of egg pockets and buried another 21% of egg pockets of artificial redds constructed in September and October between Knights Ferry and Willms Pond (CMC 2002b). Redd superimposition rates may also be high when escapement is relatively low as salmon prefer to spawn in loose gravel, as occurs in newly constructed redds in the highly compacted, natural spawning sites in the Stanislaus River. In fall 1996, when escapement was estimated at 168 fish, redd superimposition completely disturbed 24% of the egg pockets of salmon redds marked with pipe piezometers over an eight-day period in mid November (Mesick 2001b). Although redd superimposition by fall-run salmon may destroy 40 to 50% of the spring-run eggs each year, it seems unlikely that this mortality factor alone could prevent the recovery of the spring-run Chinook population in the Stanislaus River.
Another problem for spring-run using the lower river is that fall-run salmon probably interbreed with spring-run fish and interbreeding probably results in genetic contamination that would alter their migratory behavior. There are no behavioral differences that would prevent spring-run and fall-run from interbreeding as fish from each run would be present together in late October when both would be spawning. This may be a substantial problem for the recovery of the spring-run since there would be several fall-run adults that could interbreed with each spring-run adult in late October.
Other factors such as high water temperatures or turbid storm runoff probably do not have a significant impact on incubating spring-run eggs. Surface and intragravel water temperatures were almost always less than 56 degrees during late-October and early-November in 1996 (CMC 1997), 1999, 2000 (CMC 2002a, 2002b) between Goodwin Dam and the Orange Blossom Bridge, whereas egg mortality begins to occur at 58 degrees (CDFG 1998). Moreover, the eggs would have hatched and most would have emerged by mid-January, when intensive turbid storm runoff usually begins in the Stanislaus River.
4.2.4 Juvenile Rearing Habitat:
Although spring-run juveniles are subjected to the same impacts that affect fall-run juveniles, such as in-channel gravel mining, exotic predators, and reductions in flow and floodplain inundation, spring-run juveniles emerge before the fall-run juveniles and so they would have a competitive edge for the best rearing habitat. Therefore, it is unlikely that degraded juvenile rearing habitat has prevented the recovery of the spring-run population.
4.2.5 Juvenile Migration:
Spring-run juveniles would also begin their migrations earlier than the fall-run juveniles when mortality rates are probably at their lowest. Although it is possible that the application of dormant sprays in the lowermost reaches of the Stanislaus River, San Joaquin River, and Delta during the winter could cause substantial mortality of spring-run juveniles, this would primarily affect fry that migrate from the upstream rearing areas during the early winter. However, since spring-run juveniles would be the first to emerge and establish territories in suitable rearing habitat, it is likely that most remain in the upstream areas that are unaffected by the dormant sprays.
4.3 FALL-RUN CHINOOK SALMON:
The fall-run Chinook salmon population in the Stanislaus River is probably impacted by many factors, although mortality rates have not been quantified for any of the factors. One of the most prominent limiting factors appears to be the high rates of mortality for juveniles migrating through dredged channels in the Stanislaus River and Delta, particularly the Stockton deep-water ship channel. The survival of juvenile fish in the deep-water ship channel is highest during flood flows or when a barrier is placed at the head of the Old River that more than doubles the flow in the ship channel. Escapement is also directly correlated with springtime flows when each brood migrates downstream as smolts. However, the cause of the mortality in the ship channel has not been studied. It is possible that mortality results from the combined effects of high water temperatures, low dissolved oxygen concentrations, ammonia toxicity, and predation.
Dredging for gravel and gold in much of the Stanislaus River's streambed primarily during the 1940s has substantially limited the availability of spawning and rearing habitat. It is also possible that the dredged areas store fine sediments that are transported to spawning and rearing habitats during high flows. The dredged areas also contain an abundance of large predatory fish although there is uncertainty as to whether predation is a substantial source of mortality of juvenile salmon.
Spawning and rearing habitat in the lower Stanislaus River are further degraded by highly regulated flows and the diking of floodplains for agriculture. The Stanislaus River's streamflow is highly regulated by over forty dams, with 85% of the total storage contained in New Melones Reservoir, which began to fill in 1981 (Kondolf and others 2001). The reservoirs have greatly reduced the amplitude and frequency of flood flows and together with the dikes, very little of the historical floodplains are inundated. Without inundation, the floodplains cannot provide terrestrial food for juvenile salmon or organic matter that helps produce more food within the river. Moreover, the lack of peak flood flows allows encroachment of riparian vegetation, which along with the dikes tend to confine flood flows to the river channel. This in turn accelerates the rate that gravel is scoured from spawning and rearing habitat. With high rates of scour, spawning and rearing habitat tends to erode away and the river tends to widen because the upstream reservoirs block gravel recruitment from the upper watershed (Kondolf and others 2001).
Another problem that resulted from the construction of New Melones Reservoir is that housing developments began to rapidly encroach on the floodplain. Housing construction near the river has led to high rates of sedimentation in the upstream areas near the Orange Blossom Bridge. The sedimentation may result in mortality of incubating Chinook salmon eggs and it probably reduces the production of aquatic invertebrates that provide food for juvenile salmon.
Water diversions for urban and agricultural use in all three San Joaquin River tributaries reduce flows and potentially result in unsuitably high water temperatures during the spring and fall for juveniles and adults rearing and migrating in the lower San Joaquin River and Delta. Currently, flow releases from Friant Dam on the mainstem San Joaquin River do not reach the confluence with the Merced River except during wet years. Moreover, base flows on the Tuolumne and Merced rivers during dry and normal years are low compared to those on the Stanislaus River. In addition, agricultural return flows are particularly high in the mainstem San Joaquin River, primarily from Mud and Salt sloughs, and the concentration of contaminants and warm temperatures are probably detrimental to rearing juveniles.
The federal and state pumping facilities in the Delta have impacted juvenile and adult salmon since pumping began at the federally operated Central Valley Project in 1951. Pumping up to a combined current maximum of 11,000 cfs for the federal and state pumping facilities during the spring diverts water into the Old River from the mainstem San Joaquin River. Studies suggest that high rates of pumping reduce juvenile survival in both the deep-water ship channel and in the Old River, except during flooding (Mesick 2001a). Recently, smolts migrating in the Delta have been protected between mid-April and mid-May with a barrier at the head of the Old River, a reduction in pumping rates, and pulse flow releases from all three tributaries. As a result of these protections, juvenile survival in the Delta has more than doubled to between 30% and 60% based on studies of tagged smolts that were reared in the Merced River Fish Facility and released at Mossdale and Dos Reis (Mesick 2001a, SJRGA 2002). However, fry that migrate into the mainstem San Joaquin River are usually subjected to the maximum pumping rates without the protection of the barrier at the head of the Old River. Moreover, adult salmon migrating upstream to spawn in the San Joaquin tributaries can stray into the Sacramento basin if pumping rates exceed 400% of San Joaquin River flows in mid-October (Mesick 2001c). Pulse flows have been released from all three tributaries since 1994 to minimize straying; however, DWR's Interim South Delta Program may increase fall pumping rates in the near future and pulse flows were relatively low in fall 2002.
Overall, these impacts affect every aspect of the salmon's life history:
a) Dikes and
encroached vegetation that isolate floodplain habitat from the
river on spawning habitat suitability, food availability for juvenile
salmonids, juvenile refuge habitat during flooding;
b) Predation, particularly in mined channels;
c) Disease;
d) Unscreened diversions downstream of Oakdale, the San Joaquin
River, and Delta;
e) Contaminants in agricultural and urban runoff;
f) High water temperatures during late spring, summer, and early
fall; and
g) Reduced dissolved oxygen concentrations in dredged channels.
4.3.1 Adult Upstream Migration:
Poor water quality in the deep-water ship channel near Stockton and excessive exports at the State Water Project and Central Valley Project at Tracy can either delay the upstream migration of adults or cause them to stray to the Sacramento River basin. Since 1993, 10-day pulse flows from the three San Joaquin River tributaries in mid-October have probably minimized these problems.
Delayed Adult Migration:
Hallock and others (1970) showed that radio-tagged adult chinook salmon delayed their migration at Stockton whenever dissolved oxygen (D.O.) concentrations were less than 5 mg/l in October during the 1960s (and possibly water temperatures that exceeded about 65 oF) and there are concerns that if adult salmon are delayed in the Delta that egg and milt viability may be reduced and the harvest of adults may be increased before spawning occurs. They reported that D.O. levels usually increased to suitable levels by November when upstream migrations resumed.
D.O. concentrations near Stockton in October and November were greater than 5 mg/l from 1983, when DWR began monitoring, to 1990, but were substantially lower than 5 mg/l for most of October in 1991 and 1992. 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.
Adult Straying:
Delta export rates at the State Water Project and Central Valley Project were increased to near maximum (about 9,600 cfs) in fall 1996 and in subsequent years to "make-up" for reduced pumping rates during the spring outmigration period to improve salmon smolt survival. The adult fall-run salmon are migrating upstream through the Delta primarily in October typically when San Joaquin River flows measured at Vernalis are low (Mesick 2001c). It is likely that when exports are high relative to San Joaquin River flows, little if any San Joaquin River water reaches the San Francisco Bay where it may be needed to help guide the salmon back to their natal stream. An analysis by Mesick (2001c) of the recovered adult salmon with coded-wire-tags (CWT) that had been reared at the Merced River Fish Facility and released in one of the San Joaquin tributaries suggests straying occurred when the ratio of exports to flows was high. The analysis, which included an adjustment to the estimated number of fish examined for CWTs in some rivers, indicates that during mid October from 1987 through 1989 when export rates exceeded 400% of Vernalis flows, straying rates ranged between 11% and 17%. In contrast, straying rates were estimated to be less than 3% when Delta export rates were less than about 300% of San Joaquin River flow at Vernalis during mid-October. Since 1993, pulse flow releases from the San Joaquin tributaries for 8 to 10 days in mid-October appear to have kept straying rates below 2%.
Delta Migration Barriers:
In the South Delta, rock barriers are installed to maintain hydraulic head for small pump diversions and these barriers are impassible for adult salmonids. Hallock and others (1970) found that when the head of Old River barrier was installed in fall 1964, adult salmon migrated through the mainstem San Joaquin River. However, when the barrier was not installed in fall 1965 through 1967, some of the salmon migrated through the South Delta. Their study suggests the rock barriers block the flow of water that attracted migrating adults and so few, if any, adults would be expected to migrate into channels where rock barriers have been installed.
4.3.2 Spawning and Incubation Habitat:
Although the historical spawning reach for fall-run Chinook salmon has been described as a ten-mile reach that extended from the marshlands above Oakdale (possibly Kerr Park at rivermile 43.5) to Knights Ferry by Clark (1929), spawning typically occurs at about 120 natural riffles between Goodwin Dam and Jacob Meyers Park in Riverbank with only a slight decline in redd density in a downstream direction (CMC 2001; Mesick 2001b). However, from fall 1998 to 2000, redd densities declined substantially in a downstream direction with few fish spawning near Oakdale (CMC 2001, 2002a, 2002b).
Surveys in 1994 and 1995 indicated that approximately 73% of the redds occurred in the uppermost 30-foot sections of the riffles although the habitat in the downstream sections of the riffles appeared to have similar water depths, velocities, and gravel sizes compared to the highly used upstream sections (Mesick 2001b). Detailed measurements at 12 riffles indicated that most redds were constructed where the streambed was rising in a downstream direction (e.g., tail of a pool); intragravel dissolved oxygen concentrations were usually high at these areas although downwelling could not be detected with a manometer (Mesick 2001b). This suggests that adult Chinook salmon require cover as provided by the pools immediately upstream of their spawning habitat. Apparently, gravel more than about 30 feet from pool habitat or other forms of cover are not used by spawning adults.
Although several studies have indicated that there are relatively high concentrations of sand and silt in the spawning riffles in the Stanislaus River (DWR 1994; Mesick 2001b, CMC 2001), Chinook salmon are able to effectively remove most of the fines during the construction of their redds (CMC 2002b). However, the fines pose a problem during redd superimposition when fish construct their redd immediately upstream of a pre-existing redd. During the construction the new redd, the superimposed redd becomes covered with fines which can entomb the alevins (CMC 2002b) and possibly reduce downwelling of oxygen-rich surface flows into the egg pocket.
Redd superimposition may be a substantial problem for salmon in the Stanislaus River. High rates of redd superimposition have been observed in the Stanislaus River since 1994 regardless of the number of spawners, partially because they tend to crowd themselves into a relatively few sites in the upstream areas where much of the spawning habitat has been excavated. Approximately 57% of the spawning habitat has been mined for gravel and 33% of the riffles between Goodwin Dam and the Orange Blossom Bridge have become somewhat armored due to the lack of gravel recruitment (Mesick 2002). Although gravel has been added to several sites in Goodwin Canyon since 1997 by the Department of Fish and Game and to 18 sites between Two-Mile Bar and Oakdale in 1998 (CMC 2002a), the amount of gravel added to the river is less than 1% of the estimated 1,031,800 cubic-yards of gravel that were excavated from the active channel between 1939 and 1999 (Kondolf and others 2001).
Most of the riffles that remain in the unmined areas are either somewhat armored, such that large rocks, which are difficult for the adults to move, cover the surface layer of the streambed, or they are cemented because the loose gravel has been gradually washed away by high flows. It is likely that adult salmon tend to superimpose their redds on top of existing redds in the Stanislaus River simply because it is easy to dig in a new redd compared to nearby areas that are either covered with large rocks or consist of cemented gravel. Redd superimposition would have the greatest impact on redds constructed early in the season (e.g., late October and early November) because they would be the most vulnerable to late-arriving fish. Based on fall 2000 studies with artificial redds (CMC 2002b), approximately 33% of the early redds would be completely destroyed during redd superimposition and another 21% of the early redds would be buried by the fines cleaned from the superimposing redd that could entomb many of the alevins and reduce the downwelling of oxygen-rich surface water into the egg pocket. Mortality rates due to entombment in superimposed redds have not been quantified.
Another likely problem is that intensive rainstorms, which typically occur in January and February, result in turbid storm runoff. When runoff is high from intensive rainstorms, flows become turbid at Knights Ferry and turbidity increases in a downstream direction, particularly starting near the housing developments just upstream of the Orange Blossom Bridge. After an intensive rainstorm on 25 January 2000, there was a thick blanket of clay-sized silt covering the riffles at Knights Ferry and the downstream areas, particularly those below the Orange Blossom Bridge (CMC 2002a). This amount of silt had not been observed following intensive storms in January 1996 and so the new housing developments are the most likely cause. It is possible that the recent heavy siltation from the housing developments caused a shift in the distribution of spawners to the upstream areas beginning in fall 1998 (CMC 2001, 2002a, 2002b). However, spring 1999 pulse flows flushed most of the silt from the spawning beds (CMC 2002a) and so there is no obvious mechanism that would have caused the fish to shift their spawning distribution. Although agriculture probably contributes much less turbidity than the new housing developments, turbidity increases in response to erosion of unpaved roads and the trenching of pastures to drain standing water that accumulates during intensive rainstorms. Bank erosion occurs at isolated locations and does not appear to be a substantial problem between Goodwin Dam and Oakdale.
Turbid storm runoff in January and February probably impacts redds constructed after mid-November that would still contain incubating eggs during the storms. Even though salmon can clean the gravel and create highly permeable intragravel conditions during redd construction (CMC 2002b), suspended clay-sized particles in turbid storm runoff can penetrate the egg pockets and coat the surface of the eggs. Coating the surface of the eggs with silt probably reduces the ability of the egg to absorb oxygen (CMC 2002a). Eggs incubating in clean laboratory conditions can survive at dissolved oxygen concentrations as low as about 2 mg/l whereas those incubated in silty, natural environments require up to 8 mg/l dissolved oxygen concentrations for survival. Eggs incubated under low oxygen concentrations also tend to produce stunted alevins that may be too weak to emerge through the sand layers covering the egg pocket or compete with healthy alevins for food after emergence.
Another problem
with turbid storm runoff is that fine sediment intrusion can seal
the sand layer covering the egg pocket thereby reducing the downwelling
of oxygen-rich surface water into the egg pocket and simultaneously
increasing the flow of oxygen-poor groundwater into the egg pocket.
In February 1996, intragravel dissolved oxygen concentrations
declined by 25% to 75% in several artificial redds after four
intensive rainstorms in January increased mean daily flows by
an average of 200 to 500 cfs for several days after each storm
(Mesick 2001b). Again, this problem would affect primarily incubating
eggs in redds constructed after mid-November, whereas alevins,
which have gills, would probably survive low dissolved oxygen
concentrations.
Although studies have not been conducted to quantify the effects
of redd superimposition and sedimentation on Chinook salmon egg
survival in the Stanislaus River, the Stanislaus River can support
the progeny of only about 1,250 adult female salmon based on a
stock-recruitment analysis for the Stanislaus River salmon population
from 1946 to 1998 (Mesick 2001a).
Gravel Mining Impacts:
The coarsening of substrate size in spawning beds in the lower Stanislaus River has resulted from the combined impacts of gravel mining, upstream dams that block recruitment of spawning-sized gravel and increased rates of scour due to the isolation of floodplain lands from the river channel. Instream gravel and gold mining, which peaked during the early 1940s and ceased prior to 1980, completely excavated approximately 44% of the spawning habitat from the 10-mile reach between Goodwin Dam and the Orange Blossom Bridge (CMC 2002c), where most Chinook salmon and steelhead spawn in the lower Stanislaus River (CMC 2001, Mesick 2001b). The upstream dams that blocked the coarse sediment supply and the isolation of floodplain habitat from the active river with dikes worsened the problem as some of the riffles in the unmined reaches have became armored and smaller as the gravel gradually eroded away (CMC 2002c). Furthermore, there is evidence that the blockage of gravel recruitment and the isolation of floodplain habitat is causing channel widening during periods of high flows (Kondolf and others 2001). However, there are a few sections of the river between Goodwin Dam and the Orange Blossom Bridge that were not mined where gravel is mobilized from riffle to riffle and most of the habitat is still highly suitable for spawning (CMC 2002c). The presence of these highly functional sections of river suggests that lack of spawning-sized gravel in the lower Stanislaus River is primarily due to past gravel mining operations.
Loss of Floodplain Habitat:
During flooding when floodplain habitats are overtopped, fine sediments are deposited on the floodplain's surface, which helps remove fines from in-river spawning habitat (Kondolf and others 2001). Diking of floodplain habitat from the active channel prevents the separation of fine sediments from spawning-size gravel during periods of bed mobilization. Gravel deposited at spawning beds in the lower Stanislaus River near the Horseshoe Road Park and Oakdale during the spring 1997 high flows appeared to have high concentrations of fine sediments (Mesick 2001b). Intragravel dissolved oxygen concentrations were suitable at the Horseshoe Road Park site (Riffle R27) but low for incubating salmonid eggs at the Oakdale site (Riffle R78) during fall 1997 (Mesick 2001b).
4.3.3 Juvenile Rearing Habitat:
Soon after emergence, concentrations of fry were observed in slow-water, margin habitats of eight study sites from about a mile below Goodwin Dam (rivermile 57.5) downstream to Oakdale (rivermile 40) in 2000 and 2001 (Fisheries Foundation 2002). Fry densities were highest between Knights Ferry and the Honolulu Bar sites. As they grew through the spring, juvenile salmon were evenly distributed between Two-Mile Bar and Oakdale, where they were relatively abundant in fast-water, riffle habitat, particularly at the Knights Ferry Gravel Replenishment Project (KFGRP) restoration sites. Although the juvenile salmon utilized fast-water areas without bankside vegetation, densities were highest where bankside vegetation provided cover adjacent to fast-water habitat. During the summer through mid September, large juveniles were observed downstream to the Orange Blossom Bridge site. From September 27 to November 19, 2001, few juveniles were observed in the river although water temperatures never exceeded 65 oF at the Orange Blossom Bridge during summer 2001 (CDEC web site). The Fishery Foundation surveys did not address the suitability of the habitat downstream of Oakdale for rearing juveniles.
It is likely that excavation of riffle habitat for gravel mining has substantially degraded juvenile rearing habitat. The snorkel surveys conducted by the Fisheries Foundation (2002) indicate that juvenile densities were highest at the KFGRP sites where gravel berms were constructed in the shape of the tails of pools. This highly porous gravel probably provided shallow habitat that was relatively free of predators and contained an abundance of aquatic invertebrates for food. In contrast, the mined channels have abnormally high densities of large Sacramento pikeminnow (Ptychocheilus grandis) that feed on newly emerged fry (Walser, personal communication, see "Notes"). However, no studies have quantified the effect of mining on mortality rates or food production and the attraction of juvenile salmon to the KFGRP gravel berms is not an indication of improved survival.
It is also possible that the lack of rearing habitat due to in-river gravel mining between Goodwin Dam and Oakdale causes fry to pre-maturely migrate from the Stanislaus River to the Delta. Survival of fry and larger juveniles in the lower Stanislaus River and San Joaquin Delta is relatively low during dry and normal years (Brandes and McLain 2001), and it is widely believed that most die from a combination of factors including high water temperatures, poor water quality in the deep-water ship channel, predation, contaminants from agricultural runoff, and entrainment in unscreened diversions (Herren and Kawasaki 2001).
Flows and Water Temperatures:
Streamflow releases from Goodwin Dam are probably adequate to provide suitable water depths, velocities (USFWS 1993), and water temperature (USBOR 1993) for juvenile salmon under the 1987 agreement between CDFG and USBOR, except perhaps during extremely dry years when the fishery flow allocation is reduced 69,000 acre-feet under a recent variant to the SWRCB Decision 1422. This is partly true because fishery flows are augmented under the CVPIA (b)(2) program depending on the availability and need for increased flows. Moreover, the 2002 NMFS biological opinion on interim operations of the Central Valley Project (CVP) and State Water Project (SWP) on federally listed threatened Central Valley spring-run chinook salmon and threatened Central Valley steelhead in accordance with section 7 of the Endangered Species Act of 1973 (ESA) includes a non-discretionary measure for the USBOR to maintain water temperatures with flow releases to no more than a mean daily average of 65 oF (between Goodwin Dam and the Orange Blossom Bridge or just below Goodwin Dam?) from June 1 through September 30 to protect over-summering steelhead from thermal stress and from warm water predator species. During a hot spell in early August 2002 when maximum daily air temperatures in Modesto ranged between 99 and 103 oF, flow releases of 200 and 225 cfs kept the mean daily water temperatures at Orange Blossom Bridge at or below 64 oF. The maximum daily water temperature was 66.3 oF during this period.
Suitable water temperatures for juvenile fall-run Chinook salmon in the Central Valley are somewhat in dispute because laboratory tests suggest that Central Valley Chinook salmon are relatively sensitive to high water temperatures compared to Chinook salmon from the Pacific Northwest. For example, juvenile Chinook salmon from the American River suffered chronic sublethal stress at water temperatures above 60 oF (15.6 oC; Rich 1987) and mortality occurred at 77 oF (25 oC) for fish from the Mokelumne River Fish Hatchery after about 100 minutes of exposure (Hanson 1991). In contrast, juvenile spring-run Chinook salmon from British Columbia grew at maximum rates at water temperatures between 66 and 68 oF (19 to 20 oC) when fed to satiation, whereas mortality also occurred at 77 oF (Brett and others 1982).
Dissolved Oxygen:
The excavation of riffle habitat reduces flow turbulence, which probably reduces dissolved oxygen concentrations. This potential problem would be particularly associated with large gravel mine pits, which collect decaying organic matter and provide virtually no flow turbulence such that dissolved oxygen concentrations would be reduced within and downstream of the pit. Although reduced dissolved oxygen concentrations are probably not lethal to juvenile salmon, mortality may result from the combined effects of reduced dissolved oxygen concentrations with other stressors including high water temperatures, disease, contaminants, and predation. For example, many dead O. mykiss, Sacramento pikeminnow, Sacramento sucker, and riffle sculpins were observed in the Tuolumne River near the La Grange Bridge in early June 2002 (Walser, personal communication, see "Notes") when water temperatures were between 59 and 61 oF. Mortalities should not have occurred at these temperatures unless dissolved oxygen concentrations were unusually low in the dredged channel near the bridge.
Floodplain Inundation:
The highly regulated flows and the construction of dikes to convert floodplain to agricultural uses on the Stanislaus River have substantially reduced that amount of floodplain that is inundated annually. Although the extent or benefits of floodplain inundation have not been quantified in the Stanislaus River, floodplain inundation probably provides a rich source of terrestrial food for juvenile salmon and organic matter that helps generate more food in the river. Growth, survival, feeding success and prey availability for juvenile Chinook salmon were higher in the Yolo Bypass, the primary floodplain of the lower Sacramento River, than in the adjacent mainstem channel in 1998 and 1999 (Sommer and others 2001). Most of the remaining functional floodplain in the Stanislaus River consists of approximately half-mile long, narrow strips that are inundated at flows of 1,500 cfs near Knights Ferry (RM 54), Frymire Ranch (RM 53), and Honolulu Bar (RM 47).
Predation:
Although predation of juvenile salmon by striped bass, largemouth bass (Micropterus salmoides), and smallmouth bass (M. dolomieu) is thought to be a major source of mortality in the San Joaquin River tributaries, the evidence does not fully justify the theory. EA Engineering, Science, and Technology (EA) conducted riverwide electrofishing surveys in the Tuolumne River in spring 1989 and 1990 (EA 1992) and concluded that largemouth and smallmouth bass are the primary predators of juvenile salmon. However, very few bass contained juvenile salmon in their stomachs except during May 1990 when 93,653 hatchery reared salmon smolts were released at Old La Grange Bridge for survival studies. The following table summarizes the results of EA's studies.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Although EA estimated that almost 70% of the tagged hatchery fish died during their three-day migration from the La Grange Bridge to the confluence with the San Joaquin River, presumably from predation, their study cannot be extrapolated to the population of naturally produced fish. First, naturally produced juveniles usually migrate at night when predation rates are lowest, whereas hatchery fish typically migrate during the day and they are thought to be naïve at avoiding predators. Second, relatively few bass were feeding on naturally produced fish particularly in early 1990 when salmon densities would have been highest and water temperatures were low. Finally, the predation study was conducted under flows that are much lower than those typically released in the Stanislaus River and it is likely that higher flows would reduce water temperatures and therefore, predation rates as well, throughout the juvenile rearing period.
Studies conducted by the California Department of Fish and Game and S.P. Cramer and Associates suggest that predation occurs in the Stanislaus River as well, but they do not prove that it is a substantial source of mortality for naturally produced juveniles. CDFG (undated memorandum) conducted a single-pass electrofishing survey at a large instream mine pit adjacent to the Oakdale Recreational Park in the late 1980s and stomach content sampling suggests that several species of centrarchids were preying on juvenile salmonids. However, the percentage of the juvenile population eaten by predators was not quantified.
A radio tracking study conducted by S.P. Cramer & Associates (1998) in the Stanislaus River during May through July 1998 suggests that 30 of 43 (70%) tagged fish of the tagged fish were presumably eaten by predators. Seven of the tagged fish stopped their migration at the in-river gravel pit at Oakdale Recreational Park. A subsequent electrofishing survey at this pit indicated that several of the tagged fish had been consumed by largemouth bass. However, the results of the radio tagging study are questionable because implanting radio tags in the fishes' stomachs may have affected the susceptibility of the tagged fish to predation, particularly from the 12-inch antennae trailing behind the fish.
S.P. Cramer & Associates have also conducted screw trapping studies in the Stanislaus River that suggest juvenile salmon survival migrating between Oakdale and Caswell State Park have been poor in some years in spite of apparently suitable flows and water temperatures (Demko and others 1999, S.P. Cramer and Associates 2001). Although these studies suggest that is an unknown but substantial source of mortality in the lower river, the survival estimates were derived by comparing the relatively low number of juveniles estimated to be migrating past a downstream screw trap at Caswell Park with high estimates of juvenile migrants at the upstream screw trap at Oakdale and the confidence intervals for these estimates are too great to detect whether the differences are real.
In contrast, observations by professional fishing guides suggest that adult Sacramento pikeminnow, which form large schools in the 3-8 foot deep mined channels in the Stanislaus River, prey on salmon fry, whereas smallmouth bass, largemouth bass, and striped bass don't begin feeding until most of the juvenile salmon have emigrated from the river (Walser, personal communication, see "Notes"). The stomachs of large adult Sacramento pikeminnow caught in the Stanislaus River frequently contained newly emerged fry in January and February. On the other hand, largemouth bass, smallmouth bass, and striped bass typically do not begin to strike at lures, and presumably feed, until after the 1,500 cfs pulse flows cease in late May, when water temperatures increase and most juvenile salmon have emigrated to the Delta. Moreover, striped bass are unlikely predators of juvenile salmon because they primarily strike at large lures resembling O. mykiss smolts, yearling salmon, yearling Sacramento pikeminnow, and large crayfish. Fishing for largemouth and smallmouth bass in the Stanislaus River was particularly good during the 1987 to 1992 drought but became quite poor after the 1995, 1997, and 1998 floods, which further suggests that bass are displaced by high flows and predation rates are low when water temperatures are low.
Other potential predators include the riffle sculpin (Cottus gulosus), which is relatively abundant in the Stanislaus River, and adult American shad (Alosa sapidissima). Another species of sculpin, the torrent sculpin (C. rhotheus), is known to prey on coho salmon (O. kisutch) fry. Schools of 20 or more American shad were observed in June and July near Lovers Leap (Fisheries Foundation 2002). In June and July, shad might prey on a few of the juveniles that remain in the river during the summer and then emigrate as yearlings.
The belief that striped bass and Sacramento pikeminnow are substantial predators of juvenile salmon is primarily based on studies conducted in the Delta. Pickard and others (1982) reported that juvenile salmon predation was high for both Sacramento pikeminnow and striped bass in the Sacramento River Delta between 1976 and 1978. They used gill nets set in Horseshoe Bend and near Hood to collect predators between February 1976 and February 1978. The results suggest that 150-1,050 mm fork length striped bass and 300-700 mm fork length Sacramento pikeminnow primarily fed on fry and relatively few smolts as feeding rates were highest rates in winter (Dec-Feb), when 77.7% had fish in their stomachs, and low during the spring (Mar-May), when only 23.3% had fish in their stomachs. Unfortunately, the species of fish contained in the predators' stomachs were not reported. Relatively few steelhead, white catfish (Ictalurus catus), channel catfish (I. punctatus), and black crappie (Pomoxis nigromaculatus) were caught in the gill nets at Horseshoe Bend.
Studies conducted outside of the San Joaquin basin suggest that predation of juvenile salmonids primarily occurs with hatchery-reared salmonids or naturally produced juveniles in reservoirs or below low-head dams where flows are altered and predators are abnormally abundant. Brown and Moyle (1981), who reviewed the literature on pikeminnow predation on juvenile salmonids, reported that under natural riverine conditions salmonids are not major prey items of pikeminnow, but that pikeminnow preyed heavily on newly released hatchery salmon after passing over the irrigation diversion dam at Red Bluff, California. Ward and others (1995) similarly reported that northern pikeminnow consumption of juvenile salmonids was highest in tailraces downstream from dams. However, other studies suggest that habitat conditions can affect whether predators feed on juvenile salmonids. Zimmerman (1999) reported that juvenile salmonid predation was much greater for northern pikeminnow than for smallmouth bass and walleyes through the lower unimpounded Columbia River, but not in the John Day Reservoir. Curet (1993) estimated that smallmouth bass predation on wild, subyearling Chinook salmon from April through June exceeded that of northern pikeminnow in Lower Granite Reservoir on the Snake River. Tabor and others (1993) reported that juvenile salmonids were the primary prey item for smallmouth bass >200 mm FL whereas crayfish was the dominant prey item for pikeminnow in a 6-km stretch of river located at the upstream end of McNary Reservoir on the Columbia River.
Overall the studies and observations described above suggest that in the Stanislaus River, predation may be limited to the large schools of adult pikeminnow and riffle sculpin feeding on newly emerged fry, whereas smallmouth bass, largemouth bass, and possibly American shad probably feed on relatively few parr that remain in the river during late spring and summer when water temperatures are high. However, it is possible that predation is high for juveniles rearing in the deep-water ship channel in the Delta as observed by Pickard and others (1982). One problem for predation studies is that many species of predators may feed at abnormally high rates on hatchery-reared juveniles that migrate during daylight and tagged juveniles that behave abnormally.
Disease:
The disease Ceratomyxa is present in the Central Valley and studies indicate that it causes a high mortality rate of chinook smolts migrating through the lower Willamette River, Oregon. This disease relies on tubifix worms for an intermediate host and the worms flourish in organic sediments. It is likely that the worms multiply and the disease spreads in years when organic sediments are not flushed by high flows. There are indications that mortality of smolts due to this disease increases in drought years and decreases in wet years. This disease is a particular concern for the Stanislaus River because there is a tubifix worm farm near the Orange Blossom Bridge. It is also possible that organic sediments accumulate and produce tubifix worms in captured mine pits.
Contaminants:
Mud and Salt sloughs and many small agricultural return channels contribute a variety of contaminants to the mainstem San Joaquin River and the tributaries, particularly during the winter when dormant sprays are applied to crops and rain storms flush the contaminants into the rivers in a pulse. However, experimental studies have indicated that there were no detrimental effects of agricultural return flow from the west side of the San Joaquin on the growth and survival of chinook salmon reared at the Merced Fish Facilities when the return flows were diluted by 50% or more with San Joaquin River water (Saiki and others 1992). Bioassays with fathead minnows with water samples from the San Joaquin, Merced, Tuolumne, and Stanislaus rivers showed little evidence of toxicity (Brown 1996). Low or no detectable concentrations of organochlorine pesticides and polychlorinated biphenyls were detected in fish collected from Don Pedro Reservoir on the Tuolumne River, San Joaquin River at Fremont Ford and Mossdale (Goodbred and others 1997). Although contaminants in the San Joaquin basin may not have had direct effects, a study conducted in Puget Sound, Washington, (Arkoosh and others 1998) indicates that emigrating juvenile chinook salmon exposed to contaminants, polycyclic aromatic hydrocarbons and polychlorinated biphenyls, suffered increased susceptibility to common marine pathogens (Vibrio anguillarium).
Contaminants may reduce the abundance of food organisms in the lower reaches of the Stanislaus River and in the Delta (draft CALFED Salmonid White Paper 2002). Although it is likely that reduced food availability affects the growth and survival of fry and parr rearing in the lower river and Delta, this impact has not been quantified.
Unscreened Diversions:
There are 44 small, unscreened screened diversions primarily in the lower Stanislaus River (Herren and Kawasaki 2001) although the entrainment rates at these sites have not been studied. The radio tagging study in the Stanislaus River, which tagged 49 fish, did not detect any entrainment of tagged fish at several moderately sized unscreened pumps in the lower river (SP Cramer & Associates 1998). Studies in the Delta suggest entrainment rates increase exponentially with increases in diversion rate. If true, a majority of entrainment would occur at the Banta-Carbona diversion, which is downstream of the mouth of the Stanislaus River and has a maximum capacity of 240 cfs. Screens were installed in the late 1970s but were later abandoned due to maintenance problems.
4.3.4 Juvenile Migration:
Juvenile salmon migrate downstream to the San Joaquin River and Delta in response to storm runoff and smoltification, during which the fish undergo physiological changes for adaptation to seawater. Smoltification usually begins in April when the juveniles reach a fork length between 70 and 100 mm. Environmental factors, such as streamflow, water temperature, photoperiod, lunar phasing, and pollution can affect the onset of smoltification (Rich and Loudermilk 1991).
Juvenile Rearing in the Delta:
High flows trigger large numbers of juvenile salmon, primarily fry (< 40 mm fork length) and some parr to migrate from the rivers into the San Joaquin River and Delta where survival rates are low compared to survival in the tributaries. Studies by Erkkila and others (1950) indicate that many more Chinook salmon fry (about 40 mm long) were collected by trawling in the San Joaquin Delta in March 1949, a relatively dry year, compared to the number of parr and smolts collected from April through June of the same year. Recent screw trapping studies indicate that 87% of the outmigrants captured at Caswell State Park emigrated as fry during 1999, when trapping occurred between January 18 and June 30 (Demko and others 2000). Fry migrations from the Stanislaus River coincided with the onset of peak flows in 1996, 1998, and 1999 (Demko and others 1999, 2000, S.P. Cramer & Associates, Inc. 1997). Some of the large juveniles (45 to 80 mm in fork length), which are called parr and show no external signs of smoltification, generally migrate downstream between mid-March and late-April.
Many fry and parr are caught with seines throughout the Delta and the San Pablo and San Francisco bays between January and March, particularly during wet years (Brandes and McLain 2001). However, relatively few juveniles are caught with seines in the Delta between April and July and mark and recapture experiments with fry released in the upper Sacramento River, Delta and San Francisco Bay between 1980 and 1987 suggest that survival is low in the Delta and Bay (Brandes and McLain 2001). Ocean recovery rates of the fry obtained from the Coleman National Fish Hatchery that were tagged with coded wire half tags indicate that survival was higher for fry released in the upper Sacramento River below Red Bluff Diversion Dam than for fry released in the North Delta in wet years when the floodplains of the Yolo and Sutter bypasses are inundated. Those released in the Bay had the lowest recovery rates in all years. Other USFWS studies indicate that fry survival was lower in the Central Delta near the mouth of the Mokelumne River than in the North Delta near Courtland, Ryde, or Isleton during dry years, although the difference was not statistically significant (Brandes and McLain 2001). However during flooding in 1982 and 1983, tagged fry survived at similar rates in the Central Delta and South Delta in the Old River compared to the North Delta. These results suggest that survival of juveniles rearing in the San Joaquin Delta is low compared to those rearing in the Stanislaus River, except perhaps during flooding. The poor survival of juveniles rearing in the Delta in dry and normal water years may be caused by predation, entrainment at numerous small, unscreened diversions, unsuitable water quality, and direct mortality at the state and federal pumping facilities in the Delta. Entrainment at the Delta pumping facilities may be nil during very wet years as tagged fry were collected at the pumping facilities only during the dry years whereas none were collected in wet years (Brandes and McLain 2001).
Currently, there are no flow or water temperature standards to maintain suitable habitat for juvenile salmon in the mainstem San Joaquin River below the mouth of the Stanislaus River. The relationship between streamflow at Vernalis and the daily range in water temperature at Vernalis for periods in April, May, and early June in 1962, 1963, 1970, and 1973 to 1994 suggest a flow of about 3,500 cfs from mid April to mid May is adequate to maintain maximum daily water temperatures below 65 F at Vernalis (Mesick 2001a). Usually adequate water temperatures occurred in the San Joaquin River except during drought years (1977 and 1987 to 1992), and when high flows entered the San Joaquin River from the James Bypass upstream of Newman during spring 1986. By the end of May, water temperatures exceeded 65 F even when flows exceeded 30,000 cfs; whereas flows greater than about 3,000 cfs were sufficient to keep water temperatures between 65 F and 70 F.
Smolt Migration Through The Delta:
The downstream migration of smolts generally begins in early April, peaks between late April and mid May in dry years (or late May in normal and wet years) and then rapidly declines in June as determined by trawling at Mossdale (CDFG 1991 to 1998). Screw trapping at Caswell State Park in the Stanislaus River between 1996 and 1999 indicates that smolts migrate downstream between early April and mid July (Demko and others 2000).
A correlation analysis with escapement-based estimates of population recruitment, which is the number of adult salmon produced from a single brood or cohort, suggests that smolt survival is strongly correlated with flow in the Delta. Recruitment to the Stanislaus River between 1946 and 1995 is correlated (r = 0.769, P = 0.000) with the mean flow in the San Joaquin River at Vernalis from April 15 to June 15 when each brood outmigrated as smolts (Mesick 2001a). In contrast, recruitment was less well correlated with flows in the Stanislaus River and the percentage of San Joaquin River flow exported at the SWP and CVP facilities and poorly correlated with mean maximum daily water temperatures at Vernalis, ocean harvest, ocean climate and upwelling, and ocean prey abundance (Mesick 2001a).
Smolt survival has been studied by the U.S. Fish and Wildlife Service (USFWS) and CDF&G since 1982 by releasing groups of about 25,000 to 100,000 hatchery reared juveniles with coded-wire-tags (CWT) at various locations in the tributaries and Delta in April and May and recapturing them with a trawl at Mossdale and Chipps Island to investigate the effects of flows and exports (Mesick 2001a). These studies provide three important results: (1) smolt survival may be lowest in the deep-water ship channel from the Port of Stockton to the mouth of the Mokelumne River; (2) survival in this reach may be correlated with water temperature and (3) installing a barrier at the head of Old River (HORB) appears to ameliorate the impacts of exports by preventing entrainment of smolts in the Old River and toward the pumping facilities and by increasing flow and reducing water temperatures in the deep-water ship channel.
The absolute survival estimates for juveniles migrating through the Delta in the mainstem San Joaquin River in spring 1991 were somewhat lower in the deep-water ship channel between Stockton and Jersey Point than in the upstream reach between Dos Reis and Stockton. Absolute survival for Feather River hatchery smolts in mid April was 72.0% between Dos Reis and Stockton (Buckley Cove) but only 17.4% between Stockton and Jersey Point based on ocean recoveries. This computes to a mortality rate of 2.55% per mile for the 11-mile reach between Dos Reis and Stockton and a mortality rate of 3.05% per mile for the 27-mile reach between Stockton and Jersey Point. During a second test in early May 1991, no releases were made at Dos Reis but absolute survival of the Feather River smolts migrating between Stockton and Jersey Point was similarly low at 17.9%; which equates to a mortality rate of 3.04% per mile. During the April 1991 test, Vernalis flows averaged 1,150 cfs, total Delta exports averaged 4,283 cfs, dissolved oxygen averaged 6.3 ppm at Rough and Ready Island near Stockton, water temperature was about 60 oF near Stockton, and flows in the Delta Cross Channel and Georgiana Slough averaged about 4,000 cfs. During the May 1991 test, Vernalis flows averaged 959 cfs, total Delta exports averaged 2,613 cfs, dissolved oxygen averaged 5.4 ppm at Rough and Ready Island near Stockton, water temperature was about 65 oF near Stockton, and flows in the Delta Cross Channel and Georgiana Slough averaged about 3,500 cfs.
The smolt survival tests in 1991 did not provide consistent results regarding the effects of increased flows downstream of the lower Mokelumne River provided by diversions of Sacramento River water through the Delta Cross Channel and Georgiana Slough. In April 1991, absolute smolt mortality rate in the 20-mile reach between Stockton and the lower Mokelumne River was 3.0% per mile whereas it was 8.14% in the 7-mile reach with increased flows between the lower Mokelumne River and Jersey Point. In contrast during May 1991, the absolute smolt mortality rate between Stockton and the lower Mokelumne River was 3.93% per mile whereas it was 2.3% per mile in the reach with increased flows between the lower Mokelumne River and Jersey Point. The difference between these tests is caused by differences in the number of fish recovered from the groups of test fish released at the lower Mokelumne River. Recoveries of the Stockton groups and the Jersey Point groups declined slightly for the May studies compared to the April studies as would be expected if mortalities increased due to increased water temperatures. However, the number of test fish recovered for the lower Mokelumne River group during May was about 60% greater than the number recovered in April even though survival should have better for the April test. One possible explanation is that the fish released at the lower Mokelumne River site in April were in relatively poor condition (e.g., unusually small or diseased) compared to the other groups.
The table below presents the number of tagged fish released at the various study sites, the number recovered in the ocean fisheries, and the survival index used to estimate absolute survival and mortality rates for the 1991 studies.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
More recent studies with Merced River hatchery fish migrating from Mossdale with the head of the Old River installed or Dos Reis without the barrier to Jersey Point suggest that survival may be correlated with a water temperature threshold of about 66 oF in the deep-water ship channel near Stockton (SJRGA 2002). The following table presents the absolute estimates of survival for juvenile fish from the Merced River Fish Facility migrating between Mossdale with the barrier installed or Dos Reis without the barrier. The estimates are based on the ocean catch for studies between 1996 and 1999 and the Chipps Island trawls estimates for 2000 and 2001, estimated flow near Stockton, and mean daily water temperature near Stockton (Station 4) during a 10-day period after the study fish were released.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
These data suggest that Merced River smolt survival for fish migrating
through the San Joaquin Delta ranged between 30% and 60% when
water temperatures near Stockton were less than 66 oF, except
for the April 28, 2000 test, and ranged between 14% and 19% when
water temperatures exceeded 68 oF. The survival estimate for the
April 28, 2000 test may change when computed with the more accurate
ocean recovery data.
Net pen studies at the release points indicated that most (>95%)
of the test fish survived for 48 hours during the 2000 and 2001
studies although water temperatures ranged between 68 and 72 oF
during the second release tests (May 7 and 8) at the Durham Ferry,
Mossdale, and Jersey Point test sites (SJRGA 2001, 2002). These
tests suggest that either mortality occurred more than 48 hours
after release and/or other factors, such as low dissolved oxygen
in the deep-water ship channel and predation, contributed to the
mortalities.
Densities of black
bass and striped bass are about three times higher in the central
Delta downstream from Rough and Ready Island near Stockton and
in the Mokelumne River (eastern Delta) than in the northern or
southern areas of the Delta based on the CDF&G resident fish
study conducted from 1980 to 1983 (see following table, CDF&G,
unpublished data). CDF&G introduced Florida largemouth bass
into the Delta in the early 1980s and again in 1989 and catch
rates of black bass have increased since 1993 (Lee 2000). Although
predation of juvenile salmon in the Delta has not been well studied,
it would account for the low survival rates of juvenile salmon
migrating between Dos Reis and Jersey Point and for Sacramento
River juveniles migrating into the Mokelumne River through the
Delta Cross Channel.
Number and mean fork length of largemouth bass, smallmouth bass,
and striped bass per kilometer that were collected during CDF&G
electrofishing surveys in the Sacramento-San Joaquin Delta, 1980
to 1983. The sampling sites in each region of the Delta are shown
in Figure 1 of Schaffter (2000).
|
|
|
|
|
|
Central Delta |
|
|
|
|
Eastern Delta |
|
|
|
|
Southern Delta |
|
|
|
|
Northern Delta |
|
|
|
|
Western Delta |
|
|
|
Delta Exports:
Although the survival of Merced River hatchery smolts from 1996 to 2001 was not correlated with Delta export rates between 1,529 cfs and 3,163 cfs (SJRGA 2002), mortality occurs as a result of entrainment at the pumping facilities and predation losses at Clifton Court Forebay (Brandes and McLain 2001). Many of the marked fish were observed at the pumping facilities when they were released into the upper Old River (average 19%) compared to those released at Dos Reis (average 3%; Brandes and McLain 2001). In addition, early smolt survival studies with Feather River Hatchery Fish indicate that smolt survival was low when Delta exports were high, but water temperatures should have been suitable. In mid April 1991, the survival of smolts from the Feather River Hatchery was only 9.4%, when total Delta exports were 7,880 cfs, Vernalis flows were 809 cfs, and the mean water temperature near Stockton was 61.2 oF for a 10-day period following the fishes' release (Mesick 2001a).
Barrier At The Head Of The Old River:
Installing a barrier at the Head of the Old River would reduce entrainment into the Old River and reduce water temperatures in the mainstem near Stockton and thereby increase smolt survival up to 400%. By reducing entrainment into the Old River and direct mortality at the CVP and SWP pumping facilities, the barrier would be expected to increase survival by about 100% when exports are about 3,100 cfs or greater based on releases of tagged Merced River hatchery smolts at Mossdale and Dos Reis in 1999. The survival for those released upstream of the Old River at Mossdale was 30