| Abstract | Introduction | Procedure | Discussion | Literature cited |

Results and Discussion

Tributaries investigated

Figure 1 shows tributaries of the Sacramento River between Keswick Dam and Chico. Table 1 provides a summary description of the tributaries targeted in this study. Twenty-four streams were visited at least once. Lack of time, difficulty of access, or inability to obtain permission, prevented visits being made to all potential tributaries. Stream kilometer (str km), the stream distance from the tributary's confluence with the Sacramento River to the sample site, is used throughout this report to define sample sites. Table 2 provides a physical reference for each site to make them easier to find on a map or in the field. (A set of GIS overlays showing the data are in preparation).

Physical conditions

Table 3 summarizes descriptive data for the sample sites. PFC ratings and channel width to depth ratios will be discussed below under "stream problems". Flow velocity at sites where juvenile chinook were collected ranged from zero to about 3 ft/sec. but were usually less than 1 ft/sec.. Turbidity was generally less than 5 NTU except immediately after a storm event (events were infrequent this year). Tributary temperature varied diurnally, seasonally, and between streams, but remained within suitable range for rearing chinook throughout the season. Figure 2 shows water temperature for a shaded reach of Kusal Slough through the 1997 season. A month's worth of data from an unshaded reach of Elder Creek is shown for comparison. Temperatures in both creeks follow the same pattern, but Elder displays a substantially greater daily excursion. The usable rearing season ended about a month earlier in Elder than Kusal, in part because of Elder's more extreme temperature fluctuation As long as maxima remain below critical limits, fluctuating temperatures, such as observed in tributaries, seem to be more favorable than constant temperatures to rearing salmonids (Brett, 1971; Behnke, 1992; Moore, 1997).

Sites sampled and juvenile chinook captured

Table 4 summarizes data from the sample sites investigated, listing values for:

Presence and distance moved upstream

Of the streams visited, all with a near-mouth gradient of less than 1 % (see Table 1) supported non-natal chinook rearing. Juveniles were not found in Paynes Creek (gradient 1.39%) or Spring Creek (gradient 1.66%). Juveniles were found only in the first 0.1 km of Sulphur Creek (gradient 0.89%).

One might logically expect to find a reciprocal relationship between gradient and distance the rearing chinook move upstream, but the data show this relationship to be very weak (r2=0.14; Figure 3). An obvious complicating factor would be presence of barriers such as beaver dams, debris dams, and high velocity culverts. It would have been too time-consuming to walk each stream looking for barriers, but in general, they should be more common in smaller streams. However, the observed relationship between stream size and distance moved upstream is also weak (r2=0.21; Figure 4). An unexpected variable, the river mile at which the tributary enters, shows the strongest correlation with the distance juveniles move upstream in the tributary (r2=0.31; Figure 5). One might logically think that there would be more juveniles entering downstream tributaries, perhaps as a result of excess fry dumping by Coleman Hatchery, which usually releases excess fry in the vicinity of Red Bluff (RM 244), and that crowding forces the juveniles further up the tributaries. However correlation between river mile and juvenile density in the mouth of tributaries is very weak (r2=0.10; Figure 6) and there is no correlation between juvenile density in the mouth of tributaries and distance they move upstream (Figure 7). There is a correlation between river mile and the gradient of tributaries studied (r2=0.27; Figure 8), however, and this plus random chance may explain the correlation between river mile and the distance juveniles move upstream. It will be interesting to see if this relationship persists in other years.

One surprise was the short distance the juveniles moved upstream in Stillwater and Churn Creeks. Both are low-gradient streams with what appears to be suitable habitat and both had good populations near the mouth, yet the apparent upstream movement was only about 10% of what was expected from observations in similar creeks.

Comparison of tributaries above and below Red Bluff

Area Juvenile Salmon Captured
Trips Sites Sampled Total per site Fall Spring Winter Late Fall
Down River from Red Bluff 24 68 2243 33 1366 783 96 0
Up River from Red Bluff 14 49 667 14 584 47 16 20

Catch per unit effort was about twice as great below Red Bluff. This seems to relate more to the fact that juvenile chinook are going further upstream than to the density at the mouths of the tributaries (See Figure 5 and Figure 6). If releases by Coleman Hatchery were the cause, one would expect greater density at downstream tributary mouths, as well as capture of some coded wire tagged fish (none were captured this year). The phenomenon of more juveniles, particularly in the Spring chinook size category, in tributaries downstream of Red Bluff was also observed by (Maslin, et al. 1996a, 1996b), who suggested a relationship with the Spring chinook spawning streams (Mill, Deer, and Big Chico). Genetic analysis to reveal the origin of tributary rearing juveniles may shed more light on this phenomenon.

Condition Factor

Calculated condition factors varied with fish size for juvenile chinook smaller than 50 mm fork length. Above 50 mm condition factor should be independent of size (See Moore, 1997). Average condition of fish greater than 50 mm fork length are summarized in Figure 9. Regardless of year or area, the condition factor for juvenile chinook in tributaries averaged 1.0 or better, indicating that the fish were finding adequate food. All plots show the same general curve, with lower conditions corresponding to the colder months. One possible explanation is that temperatures In that period are below optimum for juvenile chinook and probably also for many of their prey items. Figure 10 shows changes in mean condition in each tributary (all chinook) at different dates. Directly below is shown minimum and maximum fork lengths observed. Condition shows a distinct increase with time (r2 = 0.345). Close observation shows the same dip as seen in average condition factors (Figure 9), but suggests a different explanation. Most of the fish before mid-February are relatively large (probably winter-run) fish which have been in the tributaries for some time and are in good condition. They emigrate at about the same time that small (probably fall-run) fish in relatively poor condition are entering the tributaries. These, in turn, improve in condition the longer they remain in the tributaries. Correlation of condition with date for values between February 22 and April 20 gives an r2 value of 0.61, which is identical to the correlation between mean fork length and mean condition factor. In other words, the juveniles are remaining in the tributaries, growing longer and getting fatter at about the same rate (see Figure 10 and Figure 12).

Density calculations and population estimates

Estimates were made of the number of juvenile chinook per linear meter of stream for most sample sites. The number of juveniles in the streams varied through-out the season, with the greatest number being present from mid-February through mid-April. Some examples are shown in Figure 11. Densities ranged up to nearly 6 juveniles per linear meter, but were usually much lower. Many sites show a wide range of density values as would be expected since the number of juveniles should range from zero early in the season through a maximum about mid season back to zero by the end of the season. Projecting the observed mid-season densities to estimate the juveniles rearing in a stream is questionable, since density estimates are available for only a few sites on most streams and values between sample sites might vary considerably with the quality of the habitat. The only streams for which we would be comfortable with population projections are shown below:

Estimated Juvenile Chinook Present in March, 1997
Creek Number
Mud 32,000
Elder 4,000
Toomes 10,000

On Mud, Elder, and Toomes Creeks, we have been able to explore the length of the habitat used and consequently have a good idea of how representative the sample sites are. Calculations from density and distance are reduced by 30% for Mud Creek to account for about 30% of the creek being lower quality habitat than the average sample site. For Elder and Toomes Creeks, we believe the sample sites are representative of available habitat.

These population estimates only include the main cohort of juveniles in the stream. A different (much smaller) cohort may have used the tributary during the early part of the season and have already emigrated by mid season. The probable growth of this cohort is shown by line W on Figure 12, whereas the growth of the major cohort is shown by line F. Potentially, a third cohort may enter the tributary past the mid season peak. That was not observed during the (relatively dry) 1997 season but may be clearly seen in Mud and Blue Tent Creeks in 1996 (Figures 13B and Figure 16). Indication of the extended time period of juveniles entering the tributaries are also seen in the small sizes present throughout much of the season in some tributaries during the 1995 and 1996 seasons (Figures 13 through 16). The single point at the bottom right of Figure 12 (arrow) does represent the appearance of a new cohort in Olney Creek, but this cohort was apparently spawned in the creek and thus does not represent non-natal rearing.

Juvenile density compared to habitat quality

Maximum juvenile density observed is compared to our ranked value for quality of cover in Figure 17. Although it seems logical that quality of cover would be an important parameter for juvenile salmonids, no correlation can be seen. Too many other factors (time of year, distance from stream mouth, partial barriers downstream) influence the density of individuals at a site.

Comparison with other years

Unfortunately, only two density estimates are available for these tributaries in former years. In February/March, 1996 , we estimated 5.9 juveniles/linear meter in Mud Creek at stream km 4.0 and 0.74 per linear meter in Blue Tent Creek at stream km 0.5 (Maslin, et al. 1996). Both values are within the range of densities measured this year.

One difference between the 1997 sample year and former years is suggested from careful observation of juvenile size ranges at different times (Figures 13-16). In 1997, essentially all the juveniles entered the creeks at one time in conjunction with the only mild winter storm. The cohort of fish entering then can be followed through time as they grew in size all the way into late April. In fact, because of lack of spates to move them out, the juveniles remained longer and grew larger than usual. Except for winter chinook which appear at large sizes in February or March in several years (Kusal, stream km 4 in 1994 [Figure 14]; Mud Creek in 1994, 1995, and 1996 [Figure 13]), only 1994 and 1997, both low rainfall years, show any quantity of juveniles exceeding 100 mm fork length. Observation in wetter years suggested that most juveniles left the tributaries at fork lengths between 80 and 90 mm (Maslin, et al. 1996a).

A single entry time is characteristic of dry years, and shows up in the data for most streams in 1994 and 1997. Multiple invasions are characteristic of wetter years such as 1995 and 1996 and can be seen in the graphs for Toomes, 1995; Figure 15) and Mud Creek at stream km 4 in 1995 and especially 1996 (Figure 13-B).

Data gathered in such an atypical water year cannot easily be extended to generalizations applicable to most years. Probably the number of juvenile chinook rearing in tributaries was substantially lower than usual. Eggs and fry may have been swept out of the system by the early-season high flows. The number of winter-run juveniles observed is probably depressed even more than the total because of our inability to sample effectively during much of the early part of the season when winter-run would be expected to be most abundant. The lack of subsequent storm events has several probable impacts on the number and distribution of juveniles in tributaries: First, it eliminated a major stimulus which causes juveniles to move around looking for suitable habitat, thereby reducing the number entering the tributaries. Second, it probably resulted in juveniles not moving as far up the tributaries for the same reason. Finally it reduced size diversity since most juveniles entered the tributaries on the same high water event.

Chinook Races:

As discussed in previous reports (Maslin, 1996a, 1996b; Moore, 1997) a wide variety of sizes of juvenile chinook are found in seasonal Sacramento River tributaries. The number of each "race" (based on fish size compared to the daily length table of Johnson, et al., 1992) captured in tributaries in this study is shown below:

"Race" Fall Late Fall Spring Winter
Number Captured 2951 20 984 111

Because of the faster growth in the tributaries and because of the difficulty in determining how long a fish has been in the tributary growing at that faster rate, tributary juveniles cannot be classified by size into races with any confidence. Figures 18-23 show the size distribution of juvenile chinook captured at each site against a background of date-specific size ranges for each race from the Sacramento River Daily Length Table. Very small juveniles (40 - 50 mm fork length), having recently emerged from redds and found their way into the tributaries, will be close in size to juveniles rearing in the mainstem. Although the cut-off between races may be at a slightly larger size, the distribution of very small fish should cluster on the proper side of the dividing line. Toomes Creek on 3/8/97 (Figure 21 A and B) provides clear examples. The smaller-size cohort in Kusal Slough on 2/5/97 (Figure 19B) provides a much more ambiguous case, since the distribution clusters exactly on the dividing line. However, one can apply the rule of a slightly larger cut-off size for tributaries and assign them to the fall race. The group of much larger fish on that date are probably not fall-run, but there is no certainty that they are winter-run. On 3/22/97, the smaller-size cohort falls clearly into the spring-run size, but these are probably the same fish assigned to the fall-run on 2/5/97. The 4/26/97 distribution adds more confusion. If the fish continued growing faster than the river rate, the mode should have moved closer to the right edge of the spring distribution, but instead, it has become very diffuse and, if anything, closer to the left edge of the spring distribution. The probable explanation is that most of the larger members of the cohort smolted and left. An experienced researcher can generally make a guess about the race based on relative size and adjustments for faster tributary growth, but it will be, at best, an educated guess.

Because of the lack of spates to move fish around this year, it is probable that most samples at the same site on different dates represent subsamples from the same population. In several instances summarized below it is apparent that the population at a site would be classified as different races on different dates.

Figure Creek Site Date 1-race 1 Date 2-race 2
18-B Mud km 3.9 3/5 - Fall 4/16 - Spring
18-D Mud km 8.4 3/5 - Fall 4/16 - Spring
19-A Kusal km 4.0 2/19 - Fall 3/22 - Spring
20-A Pine km 6.4 2/26 - Fall/Spring 4/26 - Spring/Winter
21-B Toomes km 1.5 3/8 - Fall 4/30- Fall/Spring

In all these instances the apparent change in race from date to date could have resulted from the same cohort growing at a rate similar to those reported for tributaries by Maslin, et al., 1996b.

Despite the difficulty of classifying juveniles to race, it is likely that several races utilize tributaries for rearing. Figure 12 shows the pattern of growth of two distinct cohorts of fish, most probably winter and fall races. In Figures 13-16, distinct groups of juveniles can be followed, as they enter the tributaries at different times, grow, and eventually emigrate. Distinct size groups can also be seen in Figures 18A&B, 19A&B, 22 and 23. Such diverse size groups were more conspicuous in wetter years (Maslin, et al., 1996b). Also, in previous years, we captured coded-wire-tagged chinook which were verified to be of winter, fall, and late-fall hatchery-released fish (Maslin, et al., 1996a). (No hatchery spring chinook were being tagged in the system.) The variety of hatchery races found, coupled with the variety of dates juveniles enter the tributaries and the variety of sizes of juveniles present at any one date, suggests that some members of all races enter tributaries for rearing. Probably the only way to verify this is with genetic testing.

Egress

The lack of significant rainfall between February and May this year resulted in a twofold problem for tributary-rearing chinook. In smaller streams such as McClure, Blue Tent and Dibble, stream flow became interrupted before the juveniles reached smolting size, so that most were trapped within pools and lost to avian predators. In the larger (and more permanent) tributaries, without the stimulus of rising water juveniles tended to remain past typical smolting size. This can be seen by comparing the maximum size juveniles observed this year with other years (Figures 13-16). While fish in most tributaries still seemed to leave before the stream became impassable, some were observed to become trapped in Pine Creek.

Importance of intermittent tributaries to other species

Table 5 shows the fish species encountered in each tributary while Table 6 summarizes information about each species. Most fish encountered were natives. While 31% of the fish species we examined were exotic, they comprised only 3.2% of the individuals. The seasonal stream habitat has been part of the Sacramento River ecosystem since the Pleistocene Era and bas been well incorporated into the life cycle of native species. However, it seems to be unique enough to exclude or inhibit exotic species. Many native species have adapted behaviorally or physiologically to take advantage of seasonal streams. Sacramento suckers, Sacramento squawfish, and hardhead, which migrate into seasonal streams for spawning and rearing, were observed in almost all the tributaries investigated. Hitch, while less frequently encountered, seem to have a similar reproductive strategy, but use only the lower ends of low gradient tributaries. We have encountered three sexually ripe Sacramento splittail in seasonal streams; perhaps this species (classified as being of special concern) has a similar reproductive strategy. Female tule perch migrate into the lower ends of the seasonal streams to give birth, then depart, leaving the young to rear for a while in the relatively predator-free tributary before following their mothers back to the river. The California roach, which is non-migratory but tolerates the extremes of high temperature and low oxygen found in pools of interrupted streams, was the most abundant fish species encountered in many streams and is probably found in the foothill reaches of most Central Valley seasonal streams.

Species other than fish also use seasonal streams extensively. Baetid mayflies and Capniid stoneflies, specialized for life in a temporary habitat, are common. Both are important food sources for rearing chinook (Moore, 1997). Pacific treefrogs, western toads and spadefoot toads use drying pools of seasonal streams as spawning areas. Cliff swallows and bats nest or roost under bridges and feed on emerging aquatic insects. Mallards and wood ducks nest along the intermittent pools.

Stream problems and restoration potential

Proper Functioning Condition (PFC)

Very few of the stream reaches visited could be assessed as being in proper functioning condition as defined by the Bureau of Land Management, 1995. Some are functioning at risk because of the destabilizing influence of human structures or activities in adjacent reaches. Unfortunately, a high proportion must be classified as non-functioning. While we did not have a multidisciplinary team to assess the reaches by the standard protocol, we made educated guesses based on our observations as compared with PFC protocol. Our approximation of PFC classification is included on Table 3. It should be emphasized that non-functioning refers to the physical stability of the stream reach and not to its ability to serve as fish habitat. Many non-functioning reaches still serve as rearing habitat for chinook. However, that habitat is probably substantially poorer than it could be and may be vulnerable to further degradation.

The poor showing of these tributaries relative to proper functioning condition results primarily from the long-standing attitude of managing agencies and riparian landowners that these small tributaries serve primarily as drainage ditches that should be straightened and cleared to transport storm water efficiently. Unfortunately this approach is exactly opposite to the concept of proper functioning condition, a primary goal of which is to keep water on the land as long as possible. Such an ingrained lack of understanding of stream function can only be countered by extensive education.

Anthropogenic barriers

A few streams have clearly observable anthropogenic barriers that impede upstream movement of juvenile chinook. These fall into two categories:

  1. Poorly designed road crossings which result in high velocity flow through culverts or in waterfalls, or both. Examples are found at:

  2. High gradient rapids created by placement of large boulders around bridge foundations. Examples are found at:

It should be noted that such obstructions serve as fish filters rather than absolute barriers. A few juvenile chinook have been observed upstream of most of them.

Riparian Vegetation

As suggested in Figure 2, riparian shade can be critical in preventing diurnal thermal maxima from reaching dangerous levels, thereby extending the usable season for these small streams. Riparian vegetation serves another function. Aphids, tree hoppers, etc., falling from overhanging vegetation, are an important source of food for rearing chinook early in the season before the invertebrate fauna has had time to develop in intermittent streams, and a supplemental food throughout the season.

Table 7 lists stream reaches shown on the GIS Riparian Plot (Geographical Information Center, 1997) to be deficient in riparian vegetation. In most of these reaches the devegetation resulted from deliberate management decisions by landowners or controlling agencies. Restoration would therefore involve an educational component probably far more difficult than the replanting component. In fact, in most sites the riparian vegetation would regenerate on its own if not continually suppressed.

Destabilization from Upland Activities

In other creeks the stream degradation results primarily from upland destabilizing activities such as mining, construction, logging, or improper grazing, perhaps in a geologically fragile soil or substrate. The result is mass movement of rock debris. This tends to deposit in lower gradient reaches of the stream, producing high gravel bars which are difficult for plants to colonize and which force lateral scouring widening the channel and disrupting lateral riparian vegetation. The resultant channel has an extremely high width to depth ratio (See Table 3) and very poor habitat quality. In many cases this problem has been compounded by attempts to mechanically shape the agraded channel. Examples are found at:

Elder Creek is a special case. While it has the mass movement of small rock debris typical of west side streams, it has been artificially confined between levees which prevent lateral scour. Consequently it has a very unstable bed, and an almost uniform flow rather than the stair-step of riffles and pools characteristic of most streams. Riparian vegetation is slowly recolonizing the artificial banks, but it is still small, providing little fish cover and no recruitment of large woody debris to force the scouring of pools. The combined effect is to reduce available cover for fish to a minimum. A possible long-term solution might be to plant its banks with large-growing trees such as cottonwoods and western sycamores. Of course, stabilizing the headwaters would facilitate restoration.

Sulphur Creek probably never recovered from extensive historical gold dredging, but it appears to have also been recently destabilized, probably as a result of residential development. This year many salmon and steelhead adults (probably several hundred of each) migrated into Sulphur Creek to spawn. Rapid dewatering of the cobble-filled bed led to interrupted flow which stranded adult steelhead and caused eggs of both species to be lost. (We also observed evidence of spawning in Middle, Olney, and Churn Creeks.) Whether the extensive spawning activity is a regular phenomenon in tributaries just downstream of Keswick Dam or just a function of the unusual water year should be investigated.

Diversion

Most of the streams probably experience either opportunistic or regular diversion for irrigation or stock watering. Unfortunately, our sampling protocol did not readily identify upstream diversions. Residents of Bear Creek complained that both base flow of the creek and runs of adult chinook and steelhead were much reduced from former years and attributed this to upstream diversion. (Bear was one of the creeks in which we identified spawning success.)

Some anthropogenic changes may even enhance the suitability of small streams for salmonid rearing. Leakage of water from the ACID Canal into Olney Creek keeps the volume up and temperature down late in the season. Anderson Creek may benefit from the same phenomenon. Judicious diversion of small amounts of Sacramento River water from existing canals into tributaries could potentially enhance the survival of non-natal rearing chinook. One stream-side resident even suggested that it would require relatively simple engineering to divert water from Shasta Reservoir down Churn Creek to make it more suitable for both spawning and rearing. (Several adult carcasses were observed in Churn Creek, but flow became reduced or interrupted before successful hatching and no fry were observed.)

To: Discussion