SALMONID LIFE-CYCLE MONITORING DESCRIPTION Trap adult and juvenile salmonid migrants (all species) in selected index streams. RESPONSIBLE PARTY The Salmonid Life-Cycle Monitoring Project of the Western Oregon Fish Research and Monitoring Program is responsible for this work. This program is administered by the Northwest Region, and is supervised by Tom Nickelson [(541) 757-4263 Ext. 223; email: nickelsont@fsl.orst.edu]. Project staff are: Project Leader: Mario Solazzi [Corvallis, (541) 757-4263 Ext. 242 e-mail: solazzim@ucs.orst.edu] Assistant Project Leader: Jeff Rodgers [Corvallis, (541) 757-4263 Ext. 231 e-mail: rodgersj@fsl.orst.edu] Assistant Project Leader: Steve Johnson [Newport, (541) 867-0300 Ext. 238; e-mail: johnsost@ccmail.orst.edu] Assistant Project Leader: Bruce Miller [Charleston, (541) 888-5515; e-mail: ] Assistant Project Leader: Tim Dalton [Tillamook, (541) 842-2741; e-mail: ] QUESTIONS ADDRESSED Are there trends in abundance of adult or downstream migrant anadromous salmonids in selected index streams? Are trends in abundance of adult coho salmon in selected index streams primarily due to changes in freshwater survival or to changes in marine survival? Are there geographic differences in the patterns of freshwater and marine survival of coho salmon? Are trends in freshwater and marine survival of coho salmon in western Oregon correlated? Are geographic patterns of freshwater survival of coho salmon associated with differences in habitat quality? (Addressed in conjunction with the Aquatic Inventory Project) What are the influences of climate and land-use activities on coho salmon survival rates? How do survival rates of wild and hatchery coho salmon compare? Addressed in conjunction with the Stock Assessment Project) What are the life history characteristics (time, size, and age at juvenile and adult migration) of the anadromous salmonids in the index streams? How accurate are methods of estimating spawning abundance of different anadromous salmonid species? (Addressed in conjunction with the Coastal Salmonid Inventory Project) CONSIDERATIONS 1. Because of life history differences among anadromous salmonid species, the questions addressed by life-cycle monitoring vary by species. The following is a species-specific description of the type of information that the program will provide. Coho Salmon Juvenile coho salmon, (at least the large majority of those life history types that still exist), spend their entire freshwater residence in or near their small natal streams. If trapping sites are located on large enough streams such that juvenile rearing occurs primarily above the trap sites, adult and juvenile migrant trapping will provide information on the freshwater and marine survival of coho salmon. Marine survival, as we use it here, encompasses the survival of fish from the time the smolts1 migrate out of the study stream until the adults return to the stream. Thus, this survival includes migration through mainstems and estuaries as both smolts and adults. With the survival information, the trapping program for coho salmon can address the following question that is critical to evaluating the effectiveness of plan measures: Chinook Salmon In Oregon coastal streams, most fall chinook juveniles migrate out of their natal stream by the early summer, and continue rearing in the mainstem rivers and estuaries before migrating to the ocean in late summer and fall. Because of this life history pattern, the trapping program will not be able to estimate marine survival rates for chinook salmon. Trapping will provide estimates of the number presmolt chinook leaving the streams each year. In addition, information on size of migrants and the timing of the migration will be collected. Chum Salmon Chum salmon may occur in some of the northern index streams. They have the shortest freshwater life history of any of the anadromous salmonids in Oregon, with juveniles migrating to the estuary shortly after emergence from the gravel. Thus, it will be possible to make estimates of marine survival. Steelhead Steelhead juveniles may move and rear considerable distances from their natal streams before they make their seaward migration. Therefore, unless trapping operations are located near the ocean, no estimate of the total number of ocean migrating juvenile steelhead produced from a known number of adult spawners can be obtained. Consequently, in most cases trapping will not provide information on the marine and freshwater survival of steelhead. Those sites located in the lower portions of river basins will 1 Coho Smolts are defined as age 1+ spring migrants and in some cases, age 0 coho that are silvery in appearance and have fork length >80mm. in late spring. provide information on smolt abundance each year. The sampling will also provide information on the migration timing, and the size and age of the migrants. Searun Cutthroat Trout The freshwater life history of searun cutthroat trout, which is similar to that of steelhead, presents similar obstacles to using trapping information to estimate their freshwater and marine survival. In addition, the small size of returning searun cutthroat trout adults makes them difficult to trap. Most returning searun cutthroat are small enough to swim through the upper picket fence in the adult trap. In most cases the spacing of the bars in the picket fence cannot be reduced to insure the capture of all searun cutthroat because it would result in the adult trap clogging with debris during high stream flows. Therefore, in most cases, trapping will only enable monitoring of trends in the number of downstream juvenile migrant cutthroat trout. Experiments are currently being conducted with an infrared fish counter that may enable us to count returning adult searun cutthroat trout without actually capturing them in a trap. 2. The selection of streams for salmonid life-cycle monitoring is of critical importance to the success of the program. To provide the most useful and accurate information, the trapping program must select sites in the most unbiased way possible. Unfortunately, there are a number of obstacles preventing the implementation of a totally unbiased trap site selection process. One obstacle is the reality that not all streams have sites that are conducive to successful trap operation. For example, streams may either be too large for traps to accurately estimate migrant numbers, or too small so that the degree of error associated with the abundance estimates masks any trends in abundance or survival. Other factors such as bank and substrate stability, stream gradient, site access, and landowner cooperation also impairs the unbiased site selection process. Funding realities and the logistics of maximizing the amount of information obtained in relation to the dollars spent are another obstacle. For example, on a daily basis, one person can run two downstream migrant traps successfully during most streamflow events only if the sites are located within a 30-minute drive or less from each other. This reality means that in some cases, some sites may be selected or rejected based on their proximity to another site, rather than on their overall merit as being the most “representative” of other streams in the area. All of these obstacles mean that only a limited number of streams will effectively be a candidate for trapping. Because the number of sites is limited in relation to the diverse nature of coastal streams, it is impossible to pick a suite of study streams that “represent” all of the other streams present in each of the five Gene Conservation Group Areas (GCG Areas). LOCATIONS Based upon a preliminary review of potential candidates, many potential streams have been identified for salmonid life-cycle monitoring). As a first cut, sites for monitoring both smolts and adults are being established at the following locations (See attached map): North Fork Nehalem River Little Nestucca River Mill Creek (Siletz River) Bales Creek (Yaquina River) Cascade Creek (Alsea River) West Fork Smith River North Fork Coquille River These streams will provide a relatively good latitudinal representation of streams in four of the five Coho GCG Areas along the Oregon Coast. In addition, smolt trapping is beginning in two Tillamook Bay tributaries (South Fork Kilchis River and Little North Fork Wilson River) with the hope that funding can be secured for adult monitoring. We are also considering other possibilities near the Smith River and Coquille River sites. Trapping Site Selection Criteria 1. Good geographic spread of sites coast-wide. Currently, ODFW has partial funding for field crews to be based in Tillamook, Newport, and Charleston. Without additional funding, it will be difficult for ODFW to operate traps that are long distances from these three areas (e.g. South Coast streams). 2. One person can run two traps. Paired sites should not be more than a 30minute drive apart so that trap watcher can cycle between traps during high streamflows. This is particularly important during smolt trapping. Trapping sites do not necessarily need to be within 30 minutes of the field crews office if travel trailers, or some other means of housing can be arranged. 3. Candidate streams should have spawning populations of coho, steelhead, and cutthroat, and where possible, chinook. 4. To maximize the number of fish sampled, streams should be as large as trapping technology allows. In practice, this generally means fourth to fifth order streams that are no wider than approximately 30 meters active channel width. 5. Existing fish ladders should be used where possible as adult trap sites. This will reduce construction costs and enable more adult traps to be operated, improving the geographic range of the monitoring effort. 6. Sites must be of sufficient depth (> 2.5 feet) and of sufficient velocity (> ft/sec) at low spring stream flows to allow operation of rotating screw smolt trap (or the site must be amenable to modification to meet these criteria). The site should also be neither too constrained or high gradient so that the smolt trap will be damaged due to excessive water turbulence, or be too unconstrained so that the stream becomes too wide and slow for efficient screw trap operation during high stream flows. 7. Land owner willingness to allow access to site for long term (> 10 years) monitoring. 8. Candidate streams without existing fish ladders need to have sites with the following characteristics to enable the construction of an adult weir: a) Uniform (preferably bedrock) bottom and stable streambanks. b) 1-2 percent gradient. c) Road access (close enough for delivery of materials needed to construct weir). PROTOCOLS Estimating Abundance of Downstream Migrants 1. Trap Design Downstream-migrating juveniles salmonids are trapped each spring using either a rotary screw trap or an inclined plane trap. In small streams, a revolving screen trap identical to that described by McLemore et al. (1989) is used. These traps consist of a revolving screen suspended between two pontoons. A 12-volt DC motor turns the screen. Downstream migrants are swept onto the screen by the stream current and transported to the back end of the trap with the aid of perforated L-shaped cups that are attached to the screen at two-foot intervals. Fish reaching the back of the trap are dropped into a live box where they can later be enumerated. The revolving screen traps currently in use have either 2 ft or 3 ft wide screens. In larger streams a floating trap that relies on an archemedes screw built into a screen covered cone that is suspended between two pontoons is used. The large end of the cone, (either 5 ft or 8 ft in diameter, depending on the size of the stream) is placed upstream into the stream current. With the lower half of the cone in the water, the water pressure forces the cone to turn on a central shaft, much like a turbine. Downstream migrating fish that enter the cone are trapped by the rotating screw and forced into a holding box at the end of the trap. Neither of these two trap designs is appropriate for all streams or flow conditions. The type and size of trap used is both a function of the size and flow characteristics of the stream being sampled, and the size and species of the fish that are targeted for trapping. In general, the screw trap is more effective in larger streams because it requires adequate water depth to accommodate the fixed fishing position of the screw as well as adequate water velocity to turn the screw. Because of their smaller size and adjustable screen depth, the revolving screen traps are best suited to smaller streams. The passive capture design of the screw trap generally means that it is more effective in catching larger steelhead and cutthroat trout than the revolving trap because these larger fish are sometimes strong enough to swim off the revolving screen before being deposited in the trap live box. Because these traps must operate during high streamflows, there is a risk that the traps will become jammed with debris and thus not fish the entire time between trap checks. For the revolving screen traps, a 12-volt clock is connected to a fuse inline with the trap motor. If the trap is jammed with debris, the fuse is blown, stopping the clock. By recording the starting and ending time on the clock, the length of time the trap was fishing before becoming jammed can be determined. The fuse has the additional benefit of preventing motor burn out by cutting off power to the motor if the trap becomes jammed. For the screw traps, a hubodometer, identical to those used by the trucking industry is used. The hubodometer is placed on the front spindle of the screw trap. The hubodometer records the distance that the screw turns between trap checks. By monitoring the number of revolutions per minute, and knowing the distance that the screw turned, the length of time the trap fished before jamming can be estimated. Predation by larger fish on smaller fish can be a problem in the trap live box. To minimize this, fir boughs or other hiding cover are often placed in the trap live box to give smaller fish a place to hide. Fish mortality can also occur in the screw trap live box during high streamflows. The resulting high water velocities can push smaller fish to the back of the live box, eventually exhausting them and leading to mortality. To reduce water velocities in the screw trap live box, a v-shaped plywood deflector that creates a pocket of calm water for small fish is sometimes placed in the live box. 2. Site Selection Regardless of the type of trap used, the selection of the site to place the trap is critical for successful trapping. The traps should generally be fished near the head of a pool, just below a section of fast flowing water. Streamflow should be moving in a straight line as it enters the trap. Pools with sharp changes in direction that result in large back eddy currents should generally be avoided. In smaller streams, small boulders, sandbags, or screened weir panels can be used to improve the hydrodynamics of the site. Both trap designs require streamflows that with a minimum water velocity to allow traps to function properly. For revolving screen traps, McLemore et al. (1989) suggests a minimum water velocity of 0.9 meters/second, with most efficient operation at 1.5-2.0 meters/second. We have found an optimum water velocity of 0.8-2.0 meters/second for the screw traps operated in Oregon coastal streams. Other considerations important in the site selection process are the availability of large trees or other suitable anchor sites, and anticipation of how well the trap will function at different streamflows and how easy it will be to monitor at high streamflows. Special attention should be given to insuring that the trap has a safe refuge during extreme high streamflow events. 3. Sampling Duration Sampling begins by the first week in March and continues until the catch decreases to low levels, usually by the first of June. Traps generally are operated 24 hours per day seven days per week and are monitored daily. 4. Trap Efficiency Estimates Because neither trap samples 100 percent of the water column, only a portion of the total number of downstream migrants are captured. A variety of factors, including changing streamflow, changing fish size, behavior, and species composition can influence the proportion of the total migrant population that is captured by the trap (trap efficiency). To accurately estimate the number of downstream migrating juvenile fish, trap efficiency must be measured on a regular basis. To accomplish this, up to 25 fish of each age class/species (see data sheet, appendix A) are marked and released each day. To insure that the fish used for trap efficiency estimates are migrating fish, they are selected from those fish captured in the trap the previous night. Because frequently not enough fish are caught in certain species/size classes to enable an accurate trap efficiency estimate on a daily basis, a weekly estimate is calculated instead using the following formula: Ni = (ni) / (mrecapi/mreli) Where Ni = total number of migrants passing trapping location in week I, ni = number of unmarked fish caught in trap in week i (Monday-Sunday trap catches), mrecapi = number of marked fish recaptured in trap on week i (Monday-Sunday trap catches) mreli = number of marked fish released above the trap in week i (Sunday- Saturday) The total number of fish migrating past the trap site for the season is then estimate by: Ntot = Ni The importance of expanding trap counts by appropriate measures of trap efficiency is illustrated in tables 1 and 2. In the examples shown, misinterpretation of the data would have resulted if trap efficiency estimates based on species and size class characteristics had not be used to adjust the estimate of downstream migrants. To obtain the most accurate estimate of trap efficiency, marked fish are released at dusk. This is done for two reasons: 1) most downstream migrating juvenile salmonids on the Oregon coast migrate at or a few hours after dusk, and; 2) daytime releases of marked fish at the same spot every day can lead to increased predation by resident cutthroat trout that establish feeding stations at the release point. Dusk releases appear to reduce this predation by reducing the exposure time of marked fish to predators. Fish marked for efficiency estimates are given adequate time to recover from handling prior to release. Because the smolt traps are typically checked in the morning, a timer-activated, self-releasing live box is used that enables marked fish to recover all day before being released at dusk. This device consists of three dark-colored five-gallon buckets that are suspended between two small floating pontoons. A spring wound timer is connected to a 12-volt automobile door lock actuator. At the appropriate time, the timer energizes the door lock actuator, which pulls a pin releasing the buckets. The buckets pivot on a pipe inserted through holes in their base, turn upside down, and release the fish. Each bucket has wire mesh panels along their sides to allow transport of oxygenated water into them. To avoid predation problems in the holding buckets, fish are separated by size. Typically fry are placed in one bucket, coho smolt sized fish in another, and larger trout in a third bucket. Periodically the fish are examined just prior to release to make sure that there is no mortality and that the buckets dump at the appropriate time. The release location for marked fish for trap efficiency estimates is located far enough upstream so the fish can evenly mix with unmarked fish moving downstream, yet not be so far upstream as to cause an extracted period of migration of marked fish over multiple days. Marked fish are typically released at least two pool/riffle units, but no more than 300 meters, above the trap. This is based on our experience and is consistent with the recommendations of Roper (1995). 5. Fish Handling Generally all fish handled for marking or length measurements are anesthetized with MS222. When anesthetizing fish, it is important to remember that water temperature, anesthetic concentration, and fish density and size can all increase the stress load on the fish. Care is taken that no more fish are anesthetized at one time than can be safely processed. During warm water periods or when large numbers of fish are being processed, the anesthetic water is regularly changed to keep it cool and well oxygenated. Several methods of marking fish for trap efficiency estimates are used. Usually, a small scissors or a razor blade is used to remove a portion of a fin. When using fin clips, typically the tip of the upper or lower caudal fin is removed. This technique is the only reliably safe way to mark small fry. For larger fish, a Panjet needleless injector is typically used to inject a small amount of colored acrylic paint under the skin. All newly marked fish are placed in the timer actuated release device, and a count is made of unmarked and recaptured marked fish. Each week, up to 25 length measurements are made for each species/size class (see data sheet instructions). Once all recaptured marked fish and unmarked fish are processed, the unmarked fish are released far enough downstream to minimize the potential of them swimming back upstream and reentering the trap. 6. Data Analysis During the early or late portion of the migration period, only a few fish may be caught during the week, and no marks may be recovered, making it impossible to calculate a weekly trap efficiency (note that this is the case for the first two week and the last week of trapping in Table 3.). To expand trap catches for these weeks a trap efficiency estimate based on the entire season is used. While this may not be extremely accurate, the total catches for these weeks are generally small, and their expanded estimates of migrants represent a small portion to the total population. For some streams, the number of migrants of a particular species that are caught in the trap are not sufficient to obtain a weekly trap efficiency estimate. This may result from a low number of migrants, a low trap efficiency for this particular species, or a combination of both. For these fish, a trap efficiency for the entire season is calculated based on the total number of marks released and recaptured while the trap was in operation (Table 4). This seasonal trap efficiency estimate is used to expand the number of fish caught in the trap during the season to obtain an estimate of total migrants. This approach may lead to a less accurate estimate, because most marked fish may be released under different flows than when the majority of marks are recovered. More importantly, if the total number of marks recovered is low (say less than 5), the expanded estimate of migrants may change significantly if the number of marks recovered is changed by even one fish. In some streams, trap efficiencies and total number of fish caught can be improved by adding weir panels to force more of the stream flow in front of the trap. ADD SECTION ON BOOTSTRAP? Trapping Adults Protocols under development. Estimating Adult Abundance Protocols under development. REFERENCES McLemore, C.E., F.H. Everest, W. R. Humphreys, and M.S. Solazzi. 1989. A Floating Trap for sampling downstream migrant fishes. Research note PNWRN-490, Pacific Northwest Research Station, United States Forest Service, Corvallis, Oregon. Roper, B.B. 1995. Ecology of anadromous salmonids within the Upper South Umpqua Basin. Doctoral Dissertation. University of Idaho, Moscow, Idaho.