U.S. Dept Commerce/NOAA/NMFS/NWFSC/Publications

The Northwest Fisheries Science Center of the National Marine Fisheries Service, NOAA, uses the NOAA Technical Memorandum NMFS series to issue informal scientific and technical publications when complete formal review and editorial processing are not appropriate or feasible due to time constraints. Documents published in this series may be referenced in the scientific and technical literature.

The NMFS-NWFSC Technical Memorandum series of the Northwest Fisheries Science Center continues the NMFS-F/NWC series established in 1970 by the Northwest & Alaska Fisheries Science Center, which has since been split into the Northwest Fisheries Science Center and the Alaska Fisheries Science Center. The NMFS-AFSC Technical Memorandum series is now being used by the Alaska Fisheries Science Center.

Reference throughout this document to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA.

Ledgerwood, R.D., E.M. Dawley, P.J. Bentley, L.G. Gilbreath, T.P. Poe, and H.L. Hansen. 1993. Effectiveness of predator removal for protecting juvenile fall chinook salmon released from Bonneville Hatchery, 1991. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-NWFSC-9, 63 p.

National Technical Information Service
U.S. Department of Commerce
5285 Port Royal Road
Springfield, VA 22161
1-800-553-8647
[email protected]

Abstract

List of Figures

List of Tables

Methods

Results

Discussion

Conclusions

Citations

Appendices

Appendix A—Marking Information, Tag Loss Estimates, Release Information, and River Conditions

Appendix B—Northern Squawfish Electrofishing Information

Appendix C—Estuarine Recovery Information

Appendix D—Statistical Analyses of Juvenile Recovery Data

Despite a belief that removal of northern squawfish (Ptychocheilus oregonensis) would increase survival of juvenile salmonids (Oncorhynchus spp.) in the Columbia River Basin, there has been no direct demonstration of the benefit of predator removal.

In 1991, we assessed the survival increases for juvenile salmon before and after the removal of northern squawfish in the vicinity of the hatchery release site, while the U.S. Fish and Wildlife Service tested the effectiveness of electrofishing to remove northern squawfish. Short-term survival differences among release groups of juvenile salmon were assessed from comparisons of coded-wire tagged (CWT) fish recovered near the upper boundary of the Columbia River estuary at Jones Beach (River Kilometer 75). Captured northern squawfish were examined to determine the effects of predator size and density on the rate at which juvenile salmonids are consumed.

A total of 2,012 northern squawfish were removed from nine transect areas near the hatchery in about 20 hours of electrofishing between the two release dates. With few exceptions, the daily catch, catch rate, mean fork length, and mean weight of northern squawfish and the number of CWTs recovered in the digestive tracts of northern squawfish (representing ingested juvenile salmon) declined over time. Analysis of CWT-fish recoveries at Jones Beach indicated that the recovery percentages for fish released into the midstream Columbia River were significantly higher than for fish released into Tanner Creek before predator removal (0.37% versus 0.30%; P = 0.01) and after predator removal (0.39% versus 0.33%; P = 0.02). After the removal of northern squawfish, the difference in recovery percentages between the two release sites was reduced from 23.3 to 18.2% (insignificant; P = 0.92).

In 1990 and 1989, the difference in recovery percentages between the two release sites (no predator removal) were considerably greater and may have been related to the lower river flows. We speculate that higher flow volumes in 1991 dispersed test fish more rapidly, reduced their exposure time to predation, and resulted in higher survival rates for Tanner Creek releases. In addition, the Columbia River flow increased about 25% between the first and second release dates and the higher flow may have increased survival of the Tanner Creek-released fish regardless of predator removal efforts. The 18.2% difference in recovery between midstream and Tanner Creek release following northern squawfish removal suggests that the resident population of northern squawfish was large and removal of 2,012 predators was insufficient to significantly improve survival of juvenile fish emigrating from Tanner Creek.

Despite the almost universal belief that removal of northern squawfish (Ptychocheilus oregonensis) will increase survival of juvenile salmonids (Oncorhynchus spp.) in the Columbia River Basin (Fig. 1), there has yet to be a direct demonstration of the benefit of predator removal. Heretofore, research has largely focused on estimating abundance of northern squawfish in selected locations (e.g., tailraces and forebays of dams, and reservoir reaches) and assessing northern squawfish predation on smolts near hydroelectric projects (Thompson 1959, Uremovich et al. 1980, Nigro 1990, Poe et al. 1991, Vigg et al. 1991). In 1989 and 1990, the National Marine Fisheries Service (NMFS) and Oregon Department of Fish and Wildlife (ODFW) cooperated in a release-site study at Bonneville Hatchery (Harold L. Hansen, unpubl. data, ODFW, Clackamas, Oregon). Subyearling fall chinook salmon (O. tshawytscha) were marked and simultaneously released into Tanner Creek (the normal hatchery release site) and into the midstream Columbia River, lateral to the confluence of Tanner Creek (Fig. 2). Seine recoveries of juveniles in the estuary indicated that survival following the 157-km migration was dramatically better for midstream Columbia River release groups than for Tanner Creek release groups. In 1989 and 1990, the differences were about 65 and 40%, respectively. These differences were thought to be related to greater predation by northern squawfish on fish released into Tanner Creek than on fish released into the deep-water, high-current area of the midstream Columbia River. Northern squawfish are known to inhabit protected shoreline areas; a large population of northern squawfish exists in the tailrace area of Bonneville Dam, adjacent to Tanner Creek (Vigg et al. 1990, Petersen et al. 1990).

This report summarizes a 1991 cooperative study by NMFS, ODFW, and U.S. Fish and Wildlife Service to demonstrate the effectiveness of removing northern squawfish from the migration route of juvenile salmon released at Bonneville Hatchery. The study had three objectives: 1) assess survival increases for juvenile salmon after the removal of northern squawfish from Tanner Creek and adjacent shoreline areas of the Columbia River; 2) assess effectiveness of electrofishing to remove northern squawfish from the migration route of juvenile salmon in the vicinity of the hatchery release site; and 3) assess prey consumption by northern squawfish before and after large-scale predator removal efforts to determine the effects of predator size and density on the rate at which juvenile salmonids are consumed.

Prior to northern squawfish removal efforts, one uniquely marked group of 100,000 juvenile salmon was released into Tanner Creek and another into the midstream Columbia River, lateral to the confluence of Tanner Creek. During the following four nights, extensive electrofishing efforts were made to remove northern squawfish from the immediate area in and around Tanner Creek and from the adjacent shoreline areas of the Columbia River extending 5 km downstream. Catch per unit effort (CPUE), size of fish removed, numbers of salmon ingested, and overall food consumption of northern squawfish were assessed to evaluate changes in the local population and their impacts on released salmon. Following northern squawfish removal efforts, a second pair of uniquely marked 100,000-fish groups was released at the two study sites. Purse and beach seining were conducted near the upper boundary of the Columbia River estuary at Jones Beach, River kilometer (RKm) 75, to recover marked fish. Recovery percentages were used for evaluating short-term survival differences between fish groups released at the two study sites before and after northern squawfish removal efforts. Similar comparisons of the relative contribution of marked fish returning to ocean and river fisheries and to the hatchery will provide a long-term evaluation for all release groups.

Test fish were the progeny of fall chinook salmon (upriver bright stock) collected by ODFW personnel at Bonneville Hatchery. About 400,000 of these fish were reared at the hatchery for this study. At release, the mean size of these subyearling-age fish was 7.4 g (61 fish/lb), somewhat larger than the fish used in the 1989 (7.0 g) and 1990 (6.3 g) studies.

Test fish were marked from 3 to 17 June, Monday through Friday, by a 14-person crew marking fish 8 hours per day; about 35,000 fish were marked each day. Each marked group had unique coded-wire tags (CWT) (Bergman et al. 1968). Cold brands (Mighell 1969) were applied to allow visual identification of fish from different treatment groups in samples seined from the estuary.  Logistics for marking fish were similar to those described by Ledgerwood et al. (1990). Two measures were taken to ensure that marked groups did not differ in fish size, fish condition, rearing history, or mark quality: 1) the four groups were marked simultaneously; and 2) differences in mark quality among groups were minimized by rotating fish markers and mark codes among fish marking stations every 2 hours so that each marker and each station contributed equivalent numbers of marked fish to each treatment group. To assess and maintain quality control in the tagging process, samples of about 100 fish from each treatment were collected about every 2 hours from outfall pipes at the marking trailer and checked for CWTs (Appendix Table A1). Similarly, samples of about five fish from each treatment were diverted into net-pens at 1-hour intervals throughout the marking day and held for a minimum of 30 days to determine tag loss. Samples from each treatment were held in separate net-pens. Estimates of tag loss ranged from 4.0 to 5.4% (mean = 4.4, n = 2,076; Appendix Table A2). Release numbers for each CWT group (treatment) were adjusted for estimated tag loss based on tag loss for the marked fish held a minimum of 30 days.

Groups of marked fish were released into Tanner Creek (the normal hatchery release site) and into the midstream Columbia River, lateral to the confluence of Tanner Creek (Fig. 2). The specific release locations and procedures were as follows:

1)

Tanner Creek: Test fish were released using the normal hatchery procedure of drawing down the water in the rearing pond and crowding fish into an underground flume. The flume carried fish about 650 m to Tanner Creek, where they were free to migrate to its confluence with the Columbia River, about 400 m downstream. At the confluence, fish were lateral to and about 150 m from the midstream Columbia River release site. Tanner Creek releases began at 2030 h, about an hour prior to midstream releases, to provide extra time for fish traveling to the Columbia River.

2)

Midstream Columbia River: Test fish were pumped through a hose with a diameter of 15 cm into 4,000-L tanker trucks; three trucks were used on each release night. Each truck was loaded with about 34,000 fish to maintain transport densities of about 60 g fish/L water (0.5 lb/gal). The trucks were loaded aboard a barge at the boat launch on Hamilton Island with one truck per barge trip. At midstream, the fish were released into the river through a 3-m-long hose with a diameter of 15 cm. Releases occurred between about 2130 and 2300 h at about RKm 232.

Two 5.5-m electrofishing boats (Smith-Root brand, model SR-18E) were used to capture northern squawfish. The bow platform of each boat was equipped with a pair of adjustable booms fitted with umbrella anode arrays. These arrays consisted of six stainless-steel cables, which were lowered into the water when fishing. All electrofishing was with pulsed direct current using 60 pulses/sec, 400-500 volts, and 4-5 amperes.

Electrofishing activities began at 0300 h on 25 June, about 6 hours following the first pair of releases (Appendix Table B1). On subsequent nights through 28 June, electrofishing was conducted from 2100 h to 0900 h.

Electrofishing was delayed the first night to allow test fish to disperse following release. Nine transect areas were electrofished: one in lower Tanner Creek, and eight others in nearshore areas in the Columbia River (Fig. 3). Each area was electrofished at least twice for about 30 minutes during each electrofishing period. Though transects on both the Oregon and Washington side of the Columbia River were electrofished, removal efforts were more concentrated in transect areas closest to the release locations.

Northern squawfish, stunned from electrofishing, generally came to the water surface and were collected with a dipnet; some stunned fish were lost in the swift currents. Netted fish were placed in a lethal solution of tricaine methane sulfonate (MS-222) and within about 40 minutes of capture, taken to a processing station on shore where weight (g), fork length (mm), sex, and state of sexual maturity were recorded for each fish. The digestive tract (esophagus to anus) was removed from each fish, placed in a plastic bag, and frozen for later analysis.

In the laboratory, frozen digestive tracts were thawed and prepared for analysis using a digestive enzyme solution (pancreatin) to dissolve flesh and leave diagnostic bones and CWTs from ingested fish intact (Petersen et al. 1990). The 2% (by weight) pancreatin solution, prepared using lukewarm tapwater, also contained 1% sodium sulfide. This solution was added to the plastic bags containing the digestive tracts and the bags were placed in a 40°C desiccating oven for 24 hours. The stainless-steel CWTs, having a higher density than bone, sank to the bottom after agitation of the digested sample, and were removed. In addition, these samples were checked for missed CWTs using an electronic tag detector. The CWTs were decoded using a compound microscope (Appendix Table B2). The solid contents of the bags were then rinsed through a 425-m m sieve using tap water. A compound microscope and forceps were used to remove diagnostic bones (primarily cleithra, dentaries, and opercles) from the samples (Hansel et al. 1988). Diagnostic bones were identified and paired to enumerate salmonids and other prey consumed.

Short-term survival differences among release groups were assessed from comparisons of tagged fish recovered near the upper boundary of the Columbia River estuary at Jones Beach (RKm 75). In addition to determining recovery differences, captured fish were observed for differences in descaling, injuries, size, and migration behavior. Dawley et al. (1985, 1988) described the sampling site and the fishing gear.

Sampling was conducted by two or three crews working 7 days per week for 8 to 12 hours per day, beginning at sunrise (Appendix Table C1). Both purse seines (midriver) and beach seines (Oregon shore) were used to determine whether study fish were more abundant in midriver or near shore (Fig. 4) and to maximize effort using the gear type that captured the greatest numbers of study fish.

All captured fish were processed aboard the purse seine vessels. The catch from each set was anesthetized using a 50 mg/L solution of ethyl-p-aminobenzoate (benzocaine) and enumerated by species. Numbers of dead, injured, or descaled salmonids were recorded. Subyearling chinook salmon were examined for excised adipose fins and brands (possible study fish) and separated for mark processing. Nonstudy fish were returned to the river immediately after counting, evaluation, and recovery from anesthesia. Descaling was judged rapidly while counting and separating study fish from nonstudy fish. Fish were classified as descaled when 25% or more of their scales were missing on one side. Descaling of fish captured at Jones Beach was generally related to waves from wind or passing ships, which rolled fish in the nets. Great care was taken to minimize descaling.

Brands were used to identify study fish for collecting CWTs, to mark biological samples, and to compare fish size among treatment groups. Daily sampling effort was adjusted to attain the desired minimum sample size of 0.5% of the number of fish released. Brand information and biological and associated sampling data (i.e., date, vessel code, gear code, set number, time of examination, fork length, and descaling) were immediately entered into a computer database and printed. Fork lengths of marked fish were recorded to the nearest mm. All branded fish (including those with illegible brands) were sacrificed to obtain CWTs, which identified treatment group and day of release.

The heads of branded fish were processed individually by recovery day, site, and time of capture. A 40% aqueous solution of potassium hydroxide was used to dissolve the heads and obtain CWTs. All CWTs were decoded and later verified; additional details of tag processing are presented in Appendix D of Ledgerwood et al. (1990).

Purse seine data, obtained from 28 June to 16 July, were standardized to a 10-set-per-day effort and beach seine catch data from 28 June to 13 July were standardized to a 5-set-per-day effort. The following formula was used for standardizing each marked group:

where:

On the day when there was no sampling effort for a particular gear type (beach seine, 4 July), the standardized catch was derived by averaging standardized catches for one day prior to and one day after the missed day. Few fish were captured after the data standardization periods and effort was reduced during the final week of sampling; thus those data were not included in the standardized data set. Dates of median fish recovery for each marked group were determined using the combined standardized data from purse and beach seine catches. Movement rates for each CWT group were calculated as the distance from the midstream Columbia River release site (RKm 232) to Jones Beach (RKm 75) divided by the travel time (in days) from release date to the date of the median fish recovery.

The hypothesis that recovery ratios at Jones Beach were equal for fish released into Tanner Creek or the midstream Columbia River was tested using a paired difference z-test. The hypothesis that different marked groups, released the same day, had equal probability of capture through time was tested using chi-square goodness of fit (Zar 1974).

We marked 400,615 fish with freeze brands, CWTs, and excision of the adipose fin before release (Table 1). Between the two release dates, 2,012 northern squawfish were captured and removed from the study area. We recovered 1,326 study fish in the estuary (about 0.3% of fish released); 71% of these were captured with purse seines in midriver (Appendix Table C2). Handling mortality for all subyearling chinook salmon captured at Jones Beach was less than 0.5% and descaling averaged 1.0%. However, no study fish were descaled.

We captured and removed 2,012 northern squawfish from the nine transect areas in about 20 hours (70,833 seconds) of electrofishing (Table 2). Forty-one percent (817) of these removals were caught in Tanner Creek or its adjacent transect areas (O1 and O2) (Table 3). The daily catch, catch rate, mean fork length, and mean weight of northern squawfish declined over time (with few exceptions, which mostly occurred during the initial abbreviated removal period on 25 June). In addition, the number of CWTs recovered in the digestive tracts of northern squawfish (representing ingested juvenile salmon), also diminished over time. Of the CWTs recovered, 86% (147) were from the digestive tracts of northern squawfish captured in Tanner Creek or its adjacent transect areas, and all were from study fish released into Tanner Creek (Appendix Table B2; Fig. 5). The CPUE was highest in transect area W1, along the Washington side of the river, but no CWTs from study fish were recovered from northern squawfish in this transect area. Only three CWTs from study fish released in the midstream Columbia River were found in northern squawfish. These were caught in transect areas O3 and O4 along the Oregon shore, which are the farthest transects from the release sites.

No significant differences were observed in migrational timing of study fish groups between either pair of groups released on the same day (a  = 0.05; Appendix D). Temporal catch distributions of each release group are presented in Figure 6. Movement rates of study fish to Jones Beach ranged from 15.7 to 22.4 km/day; faster than in 1989 or 1990 (Table 4).

Movement rates of fish from the second pair of release groups were about 26% higher than those of the first pair, probably due in part to increased river flow at the time of the second release (Fig. 7).

Generally, fish from all release groups showed increasing mean lengths during the recovery period, but no differences were apparent among treatment groups (Fig. 8). The randomized marking procedures produced groups with similar size distributions, which enabled comparisons of length-frequency distribution for marked groups after their migration to the estuary (Fig. 9). There was no indication that smaller fish were missing from the Tanner Creek release groups. Evidence of missing size-groups may have been apparent if size-selective predation by northern squawfish had occurred.

Analysis of CWT-fish recoveries at Jones Beach (Appendix D) indicated that the recovery percentages for fish released from the midstream Columbia River were significantly higher than for fish released from Tanner Creek for both the first (0.37% versus 0.30%; P = 0.01) and the second pair of release groups (0.39% versus 0.33%; P = 0.02). After the removal of northern squawfish, the difference in recovery percentages between the two release sites was reduced from 23.3% to 18.2% (Table 5; Fig. 10); this 22% reduction in recovery percentage differences ((23.3 - 18.2) ¸ 23.3 * 100) was insignificant (P = 0.92). Although the recovery percentages of the second release pair were higher than those for the first release pair, they are not directly comparable because releases made on different dates were subject to different river conditions and sampling effort.

To further assess data consistency, we analyzed purse seine, beach seine, and total recoveries, and standardized these recovery data to a constant daily effort (Appendix D). Conclusions regarding differences among recovery ratios derived from the standardized data were similar to those reached with the actual catch data. Recoveries of study fish released from the midstream Columbia River were higher than those for fish released into Tanner Creek; no significant change in the difference between recovery percentages occurred following removal of northern squawfish.

In 1991, recovery of subyearling chinook salmon released from the midstream Columbia River was significantly higher (a  = 0.05), averaging about 21% greater, than for fish released from Bonneville Hatchery into Tanner Creek. The difference in recovery percentages for midstream Columbia River releases in 1991 was considerably less than the differences of 40% in 1990 and 65% in 1989. One factor in the reduced difference between midstream and Tanner Creek releases may have been the increased river flow during the majority of the 1991 outmigration compared to previous years, especially compared to the drought year 1989 (Fig. 7). We speculate that higher flow volumes dispersed test fish more rapidly, reduced their exposure time to predation, and resulted in higher survival rates for Tanner Creek releases. The percent survival benefit for midstream releases was inversely correlated with the movement rate of Tanner Creek-released fish (Fig. 11). Movement rate may be a function of both river flow and state of smoltification (Zaugg and Mahnken 1991). Smoltification was not assessed in this study; however, release dates were similar each year (between 24 and 30 June).

In 1991, the Columbia River flow increased about 25% between the first and second release dates (Table 4). The higher flow resulted in faster movement to Jones Beach for the groups released on 28 June and may have increased survival of the Tanner Creek-released fish regardless of predator removal efforts. Yet the difference between midstream and Tanner Creek release declined only slightly (22%) following northern squawfish removal. This may suggest that the resident population of northern squawfish was large, and removal of 2,012 predators was insufficient to significantly improve survival of juvenile fish emigrating from Tanner Creek.

It was difficult to determine if the higher numbers and catch rates of predators in the transect areas nearest Tanner Creek occurred because of northern squawfish congregation near the hatchery release site or because high densities of northern squawfish were prevalent throughout the study area. The high catches of northern squawfish in transect area W1 support the latter explanation. The observations that CWT recoveries were concentrated in transect areas closest to the Tanner Creek release site, and that nearly all the CWTs recovered were from the Tanner Creek release groups, are evidence that juvenile salmonids released from the hatchery were more vulnerable to predation by northern squawfish than juveniles released in midstream. The decline in catch and size of northern squawfish captured may indicate a depletion of the local population, especially of the larger fish, during the removal period. Other explanations for the decline in catch may be emigration of predators from the study area (along with the released salmon), or a change in avoidance reaction to electrofishing gear. In total, over 60,000 northern squawfish were removed from the tailrace area of Bonneville Dam during 1991 (with most removals done after this study). The sharp drop in numbers of CWTs in the digestive tracts of northern squawfish by the final day of electrofishing indicates emigration of the released salmon.

1)

Releases of subyearling chinook salmon from Bonneville Hatchery into the midstream Columbia River exhibited significantly higher short-term survival rates than fish released into Tanner Creek. The difference in survival is thought to be related to predation by northern squawfish on fish released at the hatchery.

2)

It was difficult to determine if the higher numbers and catch rates of predators in the transect areas nearest Tanner Creek occurred because of northern squawfish congregation near the hatchery release site or because high densities of northern squawfish were prevalent throughout the study area.

3)

The predominance of CWTs from Tanner Creek-released juvenile salmon in the digestive tracts of northern squawfish indicated that juvenile salmon released from the hatchery were more vulnerable to predation by northern squawfish than juveniles released in midstream.

4)

The percent survival benefit between midstream Columbia River and Tanner Creek release groups appears to be inversely related to the movement rate of Tanner Creek release groups. Higher movement rates for fish were associated with higher river flows and may also have been influenced by smoltification differences between years.

5)

Electrofishing efforts to remove northern squawfish from the migration route of juvenile salmon emanating from Bonneville Hatchery did not significantly reduce the survival difference between midstream Columbia River and Tanner Creek release groups.

Bergman, P. K., K. B. Jefferts, H. F. Fiscus, and R. C. Hager. 1968. A preliminary evaluation of an implanted coded wire fish tag. Wash. Dep. Fish. Fish. Res. Pap. 3(1):63-84.

Dawley, E. M., L. G. Gilbreath, and R. D. Ledgerwood. 1988. Evaluation of juvenile salmonid survival through the Second Powerhouse turbines and downstream migrant bypass system at Bonneville Dam, 1987. Report to U.S. Army Corps of Engineers, Contract DACW57-87-F-0323, 36 p. + Appendix. (Available from Northwest Fisheries Science Center, 2725 Montlake Blvd. E., Seattle, WA 98112-2097.)

Dawley, E. M., L. G. Gilbreath, R. D. Ledgerwood, P. J. Bentley, B. P. Sandford, and M. H. Schiewe. 1989. Survival of subyearling chinook salmon which have passed through the turbines, bypass system, and tailrace basin of Bonneville Dam Second Powerhouse, 1988. Report to U.S. Army Corps of Engineers, Contract DACW57-87-F-0323, 78 p. (Available from Northwest Fisheries Science Center, 2725 Montlake Blvd. E., Seattle, WA 98112-2097.)

Dawley, E. M., R. D. Ledgerwood, and A. L. Jensen. 1985. Beach and purse seine sampling of juvenile salmonids in the Columbia River estuary and ocean plume, 1977-1983. Volume I: Procedures, sampling effort, and catch data. U.S. Dep. Commer., NOAA Tech. Memo. NMFS, F/NWC-74, 260 p.

Hansel, H. C., S. D. Duke, P. T. Lofy, and G. A. Gray. 1988. Use of diagnostic bones to identify and estimate original lengths of ingested prey fishes. Trans. Am. Fish. Soc. 117:55-62.

Ledgerwood, R. D., E. M. Dawley, L. G. Gilbreath, P. J. Bentley, B. P. Sandford, and M. H. Schiewe. 1990. Relative survival of subyearling chinook salmon which have passed Bonneville Dam via the spillway or the Second Powerhouse turbines or bypass system in 1989, with comparisons to 1987 and 1988. Report to U.S. Army Corps of Engineers, Contract E85890024/E86890097, 136 p. (Available from Northwest Fisheries Science Center, 2725 Montlake Blvd. E., Seattle, WA 98112-2097.)

Mighell, J. H. 1969. Rapid cold-branding of salmon and trout with liquid nitrogen.  J. Fish. Res. Board Can. 26:2765-2769.

Nigro, A. A. 1990. Developing a predation index and evaluating ways to reduce salmonid losses to predation in the Columbia River basin. Oregon Dep. Fish Wildl. Report to Bonneville Power Administration, Contract DE-A179-88BP92122, 338 p. (Available from Oregon Department of Fish and Wildlife, P.O. Box 59, Portland, OR 97207.)

Petersen, J. H., M. G. Mesa, J. Hall-Griswold, W. C. Schrader, G. W. Short, and T. P. Poe. 1990. Magnitude and dynamics of predation on juvenile salmonids in Columbia and Snake River reservoirs. Report of Research, 1989-1990 to Bonneville Power Administration, Contract DE-AI79-88BP91964, 82 p. (Available from Oregon Department of Fish and Wildlife, P.O. Box 59, Portland, OR 97207.)

Poe, T. P., H. C. Hansel, S. Vigg, D. E. Palmer, and L. A. Prendergast. 1991. Feeding of predaceous fishes on out-migrating juvenile salmonids in John Day Reservoir, Columbia River. Trans. Am. Fish. Soc. 120:405-420.

Sokal, R. R., and F. J. Rohlf. 1981. Biometry, 2nd Edition. W. H. Freeman, San  Francisco, 776 p.

Thompson, R. B. 1959. Food of the squawfish (Ptychocheilus oregonensis Richardson) of the lower Columbia River. U.S. Fish Wildl. Serv. Fish. Bull. 158:43-58.

Uremovich, B. L., S. P. Cramer, C. F. Willis, and C. O. Junge. 1980. Passage of juvenile salmonids through the ice-trash sluiceway and squawfish predation at Bonneville Dam, 1980. Report to U. S. Army Corps of Engineers, 46 p. (Available from Oregon Department of Fish and Wildlife, P.O. Box 59, Portland, OR 97207.)

Vigg, S., C. C. Burley, D. L. Ward, C. Mallette, S. Smith, and M. Zimmerman. 1990.  Development of a system-wide predator control program: stepwise implementation of a predation index, predator control fisheries, and evaluation plan in the Columbia River Basin. Report to Bonneville Power Administration, Contract DE-BI79-90BP07084, 111 p. (Available from Oregon Department of Fish and Wildlife, P.O. Box 59, Portland, OR 97207.)

Vigg, S., T. P. Poe, L. A. Prendergast, and H. C. Hansel. 1991. Rates of consumption of juvenile salmonids and alternative prey fish by northern squawfish, walleyes, smallmouth bass, and channel catfish in John Day Reservoir, Columbia River. Trans. Am. Fish. Soc. 120:421-438.

Zar, J. H. 1974. Biostatistical Analysis. Prentice-Hall. Englewood Cliffs, NJ. 620 p.

Zaugg, W. S., and C. V. W. Mahnken. 1991. The importance of smolt development to successful marine ranching of Pacific salmon. In R. S. Svrjcek (editor), Marine ranching. Proceedings of the Seventeenth U.S.-Japan Meeting on Aquaculture, Ise, Mie Prefecture, Japan, October 16, 17, and 18, 1988, p. 89-97. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 102.

Return to top of Table of Contents