BPA Fish and Wildlife FY 1997 Proposal

Section 1. Administrative
Section 2. Narrative
Section 3. Budget

see CBFWA and BPA funding recommendations

Section 1. Administrative

Title of project
Developing Techniques to Improve the Quality and Survival of Hatchery Reared Salmonids

BPA project number   5509300

Business name of agency, institution or organization requesting funding
National Biological Service

Sponsor type   WA-Federal Agency

Proposal contact person or principal investigator

BPA technical contact   , EWN

Biological opinion ID   None

NWPPC Program number   6.2B

Short description
Using existing hatcheries in the Columbia River Basin, techniques to rear a more wild like salmon smolt will be explored. Studies will center around four components: (1) riparian-like covers; (2) predator avoidance - naive vs trained; (3) raceway substrate and bottom oriented flows , and; (4) in-water feeding. Control fish will be those produced under “normal” hatchery operations.

Project start year   1997    End year   2003

Start of operation and/or maintenance   1997

Project development phase   Implementation

Section 2. Narrative

Related projects
The NBS has completed a two-year pilot study, without coded wire tags, to look at the effects of riparian-like covers on the physiology and behavior of hatchery reared coho salmon and are nearing the end of the second year (three years total) of a cover study and predator avoidance study using coded-wire tagged hatchery reared spring chinook salmon. All studies are being conducted at U.S. Fish and Wildlife Service production hatcheries on the Columbia River. NMFS NATURE Studies and some state hatchery projects are looking at other aspects of changing the way hatchery fish are produced, but none are as extensive as these we are proposing in that we are using production fish in their respective hatchery environments to conduct the studies.

Project history
The project was actually begun using brood year 1991 coho salmon progeny at the Willard National Fish Hatchery. The funding to date has been primarily NBS funding and no additional staff have been hired to assist with the study. Material costs have been minimal thus far, but to complete phases 2, 3, 4 and 5 (see Testable Hypothesis below), funding will be required to bring on one other biologist to assist, and to buy materials to be able to achieve the aforementioned in-water manipulations to existing hatchery raceways.

Biological results achieved
Although not directly funded by BPA or other outside agencies, two years of work have been completed using riparian-like covers to test their effects on the physiology, behavior and morphology of hatchery reared coho salmon. The data analysis is nearly complete and a peer-reviewed journal article will be prepared during the spring and summer of 1996. Some work has also begun, in cooperation with BPA funded predation studies (82-003), to develop training techniques that will enable salmon smolts to avoid predators once released from the hatchery.

Annual reports and technical papers

Management implications

Specific measureable objectives
The primary objective is to develop techniques that change the behavior of hatchery reared salmonids to more resemble their wild counterparts when they are released into the rivers and streams for migration to the estuaries and ocean. This should lead to increased survival of the hatchery stocks, and therefore allow hatcheries to reduce production (i.e., potentially less pressure placed on wild outmigrating stocks) needed to meet escapement goals for tribal, sport and commercial fisheries. Our results are very measurable because we use “normal” hatchery production fish as our controls to test against our treatments, both for “in-hatchery” and escapement comparisons.

Testable hypothesis
1. Cover Study: Null hypothesis - Raceway covers have no effect on the physiology (stress, growth), morphology (coloration), escapement, or behavior (spread out vs clumping) of hatchery reared salmonids. Overall survival is not effected and therefore release densities remain high to meet escapement goals.

2. Predator Avoidance: Null hypothesis - Exposure of hatchery reared salmonids to predators, causes no observable differences in escapement between treatment and control groups.

3. Substrate & Flows: Null hypothesis - The addition of substrate to the bottom of existing production raceways and redirecting the flows to form resting pockets and quiet areas in conjunction with the substrate, causes no significant differences between the treatment and control fish stress, behavior, survival and escapement over the three year study period.

4. In-Water Feeding: Null hypothesis - Hatchery fish fed by underwater methods are more fit than those fed on the surface. By training hatchery reared salmonids to feed underwater, versus having to be attracted to the surface for food, there are no significant differences in the behavior of the fish once released from the hatchery environment and hence no increased survival to the estuary and ocean.

5. Combined Effects: Null hypothesis - There are no significant differences in overall survivability and/or escapement between treatments and controls when all, or portions, of the above studies are combined.

Underlying assumptions or critical constraints
The underlying assumption is simply that hatchery salmonids be reared in such a way that their behavior can be changed to enable them to be better prepared for optimum survival once they leave the hatchery and migrate downstream to the ocean. The only foreseeable constraints to a study such as this would be a lack of brood stock to produce enough progeny for the study, i.e., if too few adults returned to a study hatchery there might possibly be too few progeny to complete a statistically valid comparison between treatments and controls. However, if adequate funding is not secured for 1997, we will not be able to complete phases 3, 4 and 5 of the overall study, which are critical to combining results and hence providing an environment in which hatchery fish are reared to survive in the wild.

Methods
Background: Each phase of the study will have different methods used in that the treatments are different. However, the underlying assumptions are the same: Change the behavior of hatchery reared salmonids, in this case coho for the initial two year study (field work completed) and spring chinook for the remaining 5-7 years, in order to increase overall survival from hatchery to ocean. Statistical methodology will be the same for each treatment. Physiology and growth data during hatchery rearing will be compared between control and treatment groups using the students “t” test where P < 0.05. Using data compiled from Elrod and Frank (1990) in their studies with Lake Trout, the number of recaptured fish needed to detect a departure of 20% from the null hypothesis (a = 0.05, 1 - b = 0.9) using a chi-square design based is 25,000 coded wire tags (CWT)/raceway, replicated by the number of raceways in each phase of the study, i.e., naive vs trained, covers, substrate, feeding, etc. Laboratory experiments are also being conducted in conjunction with the naive vs trained portion using spring chinook salmon smolts as the prey and northern squawfish as predators. Eight naive vs trained laboratory trials have been conducted to date with at least 20 more planned over the next two years to gather enough data for a rigorous statistical analysis combining simple comparative tests and power statistics.

Coho Study
Background: In 1992, coho salmon in four production raceways at the Willard NFH were selected for the study; two control and two treatment. The selection was based on: 1) close proximity to the NBS-Columbia River Research Laboratory, and; 2) similarity of raceways to each other, e.g., centrally located in a bank of 20 so that end raceways would not be included. Arrangements with the National Marine Fisheries Service to purchase the netting were completed in late June. Although the project could not begin until 24 July 1992 after completion of the initial set of covers (the fish had been ponded in March), it was decided to proceed because the fish still had approximately 9 months left prior to release. Some "sunburn" was occurring but it appeared equally distributed among the study raceways. Regardless of the fact that the study fish would not be covered for their entire outside lives at the hatchery, the experience gained working with the covers and perfecting data collection, sampling, and observation techniques, was quite valuable. A literature search was conducted in March 1992 through Oregon State University's Compact Cambridge system.

The first set of covers were constructed by building 14, 10 ft long X 8 ft wide frames from S-40, 1 inch PVC pipe. Once the frames were put together, 8 ft X 10 ft pieces of surplus military camouflage netting were cut and fastened to the frames with eight inch long nylon electrical ties placed approximately every two feet. Fifty-six frame supports (4/frame), made from 2 ft long X 1 inch wide X 1/4 inch thick stainless steel flat-bar, were heated and bent to conform to the top and edges of the concrete raceways. Once in place, this design enabled hatchery employees and researchers to move about freely on the concrete walkways to carry out feeding, cleaning and sampling.

Plasma cortisol, gill Na+K+-ATPase, lysozyme, skin reflectance, length, and weight were sampled, and general external characteristics were noted approximately monthly through March 1993. Sixty fish were sampled per sampling period (15 fish/raceway). Seasonal (summer, fall, winter, spring) light levels, in lumens, were recorded in the covered and uncovered raceways in 1992 using an International Light Digital Photometer.
Feeding, cleaning and mortality collection were carried out by hatchery personnel so that all factors, except the addition of covers over the treatment raceways, remained the same. Temperatures were recorded daily on a hatchery thermograph.

In conjunction with the first year of study, a seawater challenge was performed on a group of coho smolts coinciding with the release from hatchery on March 15, 1993. The aforementioned physiological samples were taken along with lysozyme samples which would indicate whether changes were taking place within the fish’s immune system during the seawater transition. The second years replicate with coho was identical to the first accept we didn’t conduct the seawater challenge.

Spring Chinook Study:
Background: In early April 1993, in conjunction with the beginning of the second year of coho work at the Willard NFH, covers were placed over spring chinook salmon juveniles in two of four bridge ponds (raceways) at Little White Salmon NFH; the two controls remaining uncovered. The Lower Columbia River Fisheries Program Office agreed to implant 100,000 coded wire tags (CWTs) at 25,000 tags per raceway, into Little White Salmon NFH juvenile spring chinook beginning with the 1993 brood, not in time for the 1993 study (92 brood year fish). The first 25,000 fish were tagged from each of the four raceways in December 1994. Sampling regimes and comparative analysis are the same as those previously described for coho. Adult returns will be monitored in 1997, 1998, 1999 and 2000. The study will be replicated over three years using the 1993, 1994 and 1995 brood year spring chinook at Little White Salmon NFH.

A 25,000 CWT per raceway figure was determined to be adequate for Little White Salmon NFH spring chinook salmon based on return data compiled by Mr. Steve Pastor, Lower Columbia River Program Office (personal communication), for the years 1984, 1985, 1986, 1987 (spring chinook salmon). See statistical methodology above.

Naive vs Trained:
Background: In March of 1995, 15 northern squawfish were captured alive using electroshocking techniques, transported to the Little White Salmon NFH and acclimated for four days in the head end of an unused raceway. At the end of the acclimation period the squawfish were netted and placed, 5 each, into three treatment raceways containing normal production spring chinook. The timing was such that the juvenile spring chinook would be exposed to the predators for approximately two weeks prior to release. We had hoped to expose the smolts to the predators for one month but were unable to capture any squawfish that early in the year. Based on previous CRRL laboratory studies, northern squawfish have been shown to be opportunistic feeders. Therefore, we felt that they would most likely eat only the sick and lame in the raceway, of which there were sufficient numbers due to a minor BKD infection. However, the water temperature was not optimum to induce northern squawfish feeding activity, i.e., approximately 5-6o C versus a desired 8-10o C (T. Poe, personal communication). Along with the previously mentioned 100,000 CWT’s implanted into the smolts for the cover study, an additional 100,000 were also implanted into 25,000 spring chinook smolts each from four more raceways for the study; giving us three additional treatment and one additional control raceway. We were able to utilize the two cover study control raceways for this study because controls for both studies were normal production fish. The fish were all released on 13 April 1995 giving the smolts a two week exposure. The study will be conducted for two more years to provide enough escapement data to properly evaluate whether or not training leads to higher escapement.

FUTURE CONSIDERATIONS
A number of researchers have conducted valuable work in laboratories that has demonstrated variations in how fish rear under different conditions. Conversely, working with fish in the wild is often logistically and financially very difficult.

Over the last few years, the NBS-Columbia River Research Laboratory has developed considerable expertise in using radio telemetry technology to observe movement and feeding patterns of northern squawfish and salmon smolts. For 1997 we propose the use of radio tags in both hatchery reared salmonids, and the predators that we use for training, to look at smolt recovery after release as well as learning more about predator behavior when confronted with an overwhelming number of hatchery salmonids prior to and during a typical hatchery release. Columbia River Research Laboratory has developed considerable expertise in using radio telemetry.

Substrate and in-water feeding protocols will be developed later in 1996. Some work has been done in both areas by other researchers, but never in a production role. We will begin by developing techniques in raceways without fish (the first year) so that substrate type, flow and current patterns can be optimized prior to actual testing with fish in 1997. We have been assigned two standard raceways at the Willard NFH where we will begin work this spring and summer.

Finally, we would eventually need to build a database that reflects the many, real or perceived, differences between hatchery and wild stocks within the Basin. However, we fully understand the difficulty in obtaining enough information on wild populations, especially in the Snake River Basin given the endangered status of many of the stocks.

Brief schedule of activities
1. Complete final two years of cover study with spring chinook at Little White Salmon NFH - end of FY98

2. Complete final two years of naive vs trained study in-hatchery - end of FY98

3. Begin developing substrate, flow and in-water feeding techniques - mid FY98

4. Implement substrate, flow, and in-water feeding techniques in hatchery raceways with control and treatment groups - FY00

5. Begin evaluating adult returns from cover and naive vs trained studies - FY01

6. Write up and publish/report in-hatchery portions of studies as they are completed - FY01

7. Combine portions of all four phases to determine optimum hatchery rearing scenario to increase smolt survival from hatchery to ocean - FY01

8. Publish/report results of escapement between controls and treatment. Develop wild vs hatchery database - FY01(+)

9. Program Administration - FY01(+)

Biological need
Given the need today to optimize returns and protect ESA listed Columbia and Snake River basin salmonids, it is imperative that we do everything we can to protect the genetic integrity of the remaining wild populations. In systems like the Columbia and Snake rivers where hydroelectric dams and extensive irrigation projects have altered the system to the point where anything resembling the old “natural” system will be impractical to achieve, we will have to mitigate the remaining wild populations with hatchery stocks to meet the escapement goals of the many user groups. Therefore, we must strive to develop a hatchery fish whose behavior mimics their wild counterparts as close as possible in order to insure maximum survival to the estuary and ocean. By doing that, we will ultimately reduce competition between hatchery and wild stocks because we will be able to reduce hatchery production, even below today levels, and still meet escapement goals. Lower densities will also render disease outbreaks less likely since much of today outbreaks are very likely due to overcrowding and the accompanying stress.

Critical uncertainties
Whether or not hatchery reared fish can be produced that have qualities resembling their wild/naturally reared counterparts. Can behavior be changed/corrected enough to make a discernable difference in escapement?

Summary of expected outcome
As previously mentioned, the expected outcome will be fewer hatchery reared fish reared to meet escapement goals while putting less pressure on wild outmigrants.

Dependencies/opportunities for cooperation
There are presently no actions or events that might affect this proposals timing or budget. However, if funding for Basin hatcheries continues to decline, as it has in FY-96, some hatcheries may be forced to close or consolidate operations, and in some cases that may not be all bad. The U.S. Fish and Wildlife Service has been cooperating closely with us over the past three years and there is no indication that the good working relationship will end. I have been in contact with state (OR, WA, and ID) and federal (NMFS) researchers and I am seriously considering putting together a small workshop of individuals who are actively working in this field. A workshop of that type will only increase cooperation amongst researchers interested in improving the overall quality and survivability of hatchery reared salmonids.

Risks
There are always some risks associated with any research project. If the project results indicate that we have made no improvement in the overall escapement, or perhaps even reduced it, then we can simply revert to today methods of rearing fish. The biggest risk facing a study such as this will be, as previously mentioned, a drastic curtailment of funding that may jeopardize the operations of lower river hatcheries which are actually supporting fairly large adult returns.

Monitoring activity
As previously described, the project’s overall outcome will simply be measured by statistically comparing the treatments vs controls in adult returns. We will also evaluate the physiological, behavioral and morphological variations in the young fish while rearing in the hatchery. If stress can be relieved in the hatchery by manipulating the rearing environment then we are on the right track to producing a fish that will probably be better able to cope with it’s environment once it leaves the hatchery and begins it’s journey to the sea.

Section 3. Budget

* For most projects, Authorized is the amount recommended by CBFWA and the Council. Planned is amount currently allocated. Contracted is the amount obligated to date of printout.

Funding recommendations

CBFWA funding review group   System Policy

Recommendation    Tier 2 - fund when funds available

Recommended funding level   $125,200