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
Effects of Flow Pulses on the Migrations of Juvenile and Adult Salmonids

BPA project number   5500200

Business name of agency, institution or organization requesting funding
S.P. Cramer & Associates

Sponsor type   OR-Consultant

Proposal contact person or principal investigator

BPA technical contact   Deborah Docherty, EWN 503/230-4458

Biological opinion ID   13.g

NWPPC Program number   5.0F.4

Short description
Synthesis and analysis of existing data to determine the effects of pulses in flow on migration and survival of anadromous salmonids in the Columbia River.

Project start year   1997    End year   1998

Start of operation and/or maintenance   

Project development phase   implementation

Section 2. Narrative

Related projects
Project Relationship to this project (No. 5500100)
8332300 Tagging and detection of PIT tags that contribute to database analyzed in this project
8401400 PIT tag detections at Bonneville and John Day will be included in this analysis
9102800 Tagging and detection of PIT tags that contribute to database analyzed in this project
9302900 Tagging and detection of PIT tags that contribute to database analyzed in this project
9008000 Ptagis Database will be information source for this project
8910800 Migration rate and survival estimates from this project will be used in CRiSP model calibrations.
9203200 Findings from this project can be used for calibration or validation of SLCM model assumptions
9204100 Radio tracking data will be sorted for this project and applied specifically to analysis of sharp changes in flow

Project history

Biological results achieved

Annual reports and technical papers

Management implications
Information from this study will assist fishery managers in managing hydropower operations to improve survival of anadromous salmonids.
More efficient use of water to aid survival of migrants may be identified.
New methods for increasing survival of migrants with flow manipulations may be identified.
Impacts to be avoided from flow manipulations may be identified.

Specific measureable objectives
Objective 1.1: Determine the effects pulsed flow events have on the timing of juvenile salmon and steelhead outmigration.

The key questions that need to be addressed concerning migration timing are:

Q1. How high should pulse flows be to stimulate migration?
Q2. How long should pulse flows last to stimulate migration?
Q3. Are there limiting factors before or after the pulse that determine its benefit?
Q4. Do subyearlings and yearlings respond similarly to pulses in flow?
Q5. Does each salmonid species respond similarly to pulses in flow?

Objective 1.2: Determine the effects pulsed flow events have on the rate of juvenile salmon and steelhead migration

The key questions that need to be addressed concerning migration rate are:

Q1. How do pulses in flow affect migration rate?
Q2. How long does the stimulus to migration rate last after the pulse has passed?
Q3. Do subyearlings and yearlings respond similarly to pulses in flow?
Q4. Does each salmonid species respond similarly to pulses in flow?


Objective 1.3: Determine the effects pulsed flow events have on survival of juvenile salmon and steelhead during outmigration

The key questions that need to be addressed concerning chinook survival are:
Q1. What proportion of smolts die during migration out of the river, and is loss related to migration rate?
Q2. Is loss rate during outmigration related to flow or other environmental factors?
Q3. Are survival of subyearlings and yearlings affected similarly by pulses in flow?
Q4. Is survival of each salmonid species affected similarly by pulses in flow?


Objective 2.1: Determine the effects pulsed flow events have on the rate of adult salmon and steelhead migration

Q1. How does flow affect migration rate other than during passage at dams?
Q2. Do pulses in flow act as a temporary stimulus or retardant to migration?
Q3. Does each salmonid species respond similarly to pulses in flow?


Objective 2.2: Determine the effects of pulsed flow events on passage of adult salmon and steelhead at dams.

Q1. Do increases in flow cause a delay in fish passage at each dam? How long?
Q2. Do increases in flow lead to a higher rate of fallback at each dam?
Q3. Is prespawning survival affected by passage time at each dam or by fallback?


Objective 2.3: Determine the effects pulsed flow events have on survival of adult salmon and steelhead

Q1. Is prespawning mortality related to upstream migration rate or time of arrival in holding areas?
Q2. Is loss rate during upstream migration related to flow or other environmental factors?
Q3. Is prespawning survival of each salmonid species affected similarly by pulses in flow?

Testable hypothesis
Hypotheses for Smolt Migration

l H.1: Sharp changes in flow do not stimulate migration rate of smolts any differently than the absolute volume of flow.

l H.2: The elapsed time interval between flow pulses does not alter the increase in migration rate of smolts that is stimulated by subsequent pulses in flow.

l H.3: The size of an individual migrant is not related to his response to pulses in flow

l H.4: Water quality of a pulse in flow, such as turbidity or temperature, does not affect a fishes response to the pulse

l H.5: The magnitude of increase in migration rate is proportional to the increase in flow.
l H.6: The duration of the flow pulse does not affect the change in migration rate it stimulates.

Hypotheses for Adult Migration

l H.1: Sharp changes in flow do not affect migration rate of adults any differently than the absolute volume of flow.

l H.2: Pulses in flow do not stimulate fall back of adults

l H.3: Any change in migration rate in response to a pulse in flow lasts no longer than the flow pulse.

l H.4 Water quality of a pulse in flow, such as turbidity or temperature, does not affect a fishes response to the pulse

l H.5: The magnitude of change in migration rate is proportional to the increase in flow.

l H.6: Subsequent survival of adults is not affected by their response to a pulse in flow.

Underlying assumptions or critical constraints
Assumption There are many observations of PIT-tagged salmon and steelhead juveniles that were migrating during a sharp change in flow.

Assumption: The response of PIT-tagged juveniles to pulses in flow is similar to that for untagged fish.

Assumption: There are many observations of radio-tagged adult salmon and steelhead that were migrating during a sharp change in flow.
Assumption: The response of radio-tagged adults to pulses in flow is similar to that for untagged adults.

Methods
Task 1.1 Locate and Assemble Data on Juvenile Salmonid Outmigrations from Studies on the West Coast Performed with Continuous Sampling of Outmigrants.

The purpose of this task is to locate potentially useful data sets for analyzing migration parameters associated with pulse flow events. This is the first step in a sorting process in which we first identify all possible members of the set, and then in subsequent steps eliminate those members without the desired qualities. We would like to locate between 5 to 10 data sets that meet our needs.

Activity 1.1.1 Interview Biologists. Because most data sets from sampling of juvenile outmigrants have not been specifically designed to evaluate pulsed flows, we must first locate the data sets representing continuous monitoring of outmigrants, and then determine which of those data sets includes freshets or artificial pulse-flow events. We expect that few of these data sets have been published. We will begin our search by interviewing fisheries researchers from each state agency, federal agency, and university on the west coast. We will also interview long-established fisheries consultants to learn of the existence of candidate data sets.

For each data set we identify that has contiguously sampled juvenile outmigrations, we will determine the following:

! Did sampling efficiency vary and was it monitored?
! Were marked fish released?
! Was there some sort of downstream sampling that enabled consistent periodic recovery of marked fish such that migration rate or survival could be estimated?
! Was streamflow monitored?
! Did flows change distinctly at any time when significant numbers of migrants were present?

Activity 1.1.2 Literature Search. We will also complete a thorough literature search for studies of juvenile salmonid outmigrations. We will obtain and review all published studies that may discuss flow effects on outmigration timing, rate, or survival. We will include studies for areas other than the West Coast.

Task 1.2 Analyze Data from Qualifying Data Sets to Compare Migration and Survival Before, During and After Pulse Flow Events.

We will subject data sets to various levels of analysis, depending on the amount of information they contain. We anticipate that the number of qualifying data sets will be greatest for migration timing and least for migration survival.

Activity 1.2.1 Enter Data into Spreadsheet. Some data sets will be available in electronic format, but many will be on data sheets. We will translate and format electronic data sets for our use in Quattro Pro and Statgraphics. We will enter data from hard copy into Quattro Pro. We will obtain appropriate flow data and enter it or translate it into Quattro Pro.

Activity 1.2.2 Compare Frequency Distributions of Migration Parameters Before During and After Flow Pulses. Data we have analyzed in the past suggest that a change in flow acts as a temporary stimulus to migration, but we are not sure how long the stimulus lasts. Therefore we will compare frequency distributions of catches, migration rates, and survivals for varying periods of time before, during and after sharp changes in flow. We will examine both parametric and nonparametric statistics to test for changes in migration timing, rate, and survival. Additionally, because we believe that the degree of smolt development affects a fish's response to a flow stimulus, we will also seek statistical methods of segregating our analyses between fish that are likely to be smolted and those that are unlikely to be smolted. Possible parameters useful for segregation include length, migration rate, appearance, and biochemical measures.

Activity 1.2.3 Estimate Parameter Values for Pulse Flow Effects on Migration. After we have examined frequency distributions and decided upon appropriate statistical techniques, we will apply those techniques to estimate parameters and their confidence intervals for flow effects on migration timing, rate, and survival. Our analyses may be as simple as analysis of variance, but are likely to require more complex techniques. Our analyses will seek to address each of the questions listed under Objectives 1-3. In order to answer these questions, we will also have to account for the influence of factors other than flow, such as turbidity, river temperature, size, degree of smolting, age, time of year, and hatchery v. wild rearing history. Thus, data on all of these factors, where available, will have to be included in the analyses.


Task 1.3 Analyze migration and survival rates of PIT-tagged juveniles in the Columbia River Before, During and After Pulse Flow Events.

Migration and survival rates of PIT-tagged juvenile salmonids in the Columbia Basin are already being analyzed under BPA contract. However, those analyses have generally employed a multiple regression approach. Such an approach cannot be effective if flow operates as a temporary stimulus that only affects a portion of the population that is predisposed to migrate. Therefore, analytical procedures employed for Task 1.2 will also be applied to the PIT tag data set for the Columbia River. We need to search the data set, and analyze information for juveniles marked immediately before, during, or after pulsed flow events. We will work collaboratively with those already analyzing the PIT tag data set.

Task 1.4 Prepare Paper for Publication.

Methods and findings from this synthesis of information will be written in a paper suitable for publication and submitted to a scientific journal acceptable to BPA.


Task 2.1 Locate and Assemble Data Sets on Upstream Migration of Adult Salmonids on the West Coast in Streams with Monitoring at Least at Two Locations.

The purpose of this task is to locate potentially useful data sets for analyzing migration parameters associated with pulse flow events. This is the first step in a sorting process in which we first identify all possible members of the set, and then in subsequent steps eliminate those members without the desired qualities. We would like to locate between 5 to 10 data sets that meet our needs.


Activity 2.1.1 Interview Biologists. The process would be similar to that described in Activity 1.1.1 for juveniles, but would likely lead us to a different set of researchers with data. Again, we expect that most pertinent studies have not been published in the scientific literature. Our key target will be to locate data sets were sampling at up and downstream locations enables estimation of migration rates and survival. We expect that radio tracking studies will be most useful.

Activity 2.1.2 Literature Search. We will also complete a thorough literature search for studies of adult salmonid migrations. We will obtain and review all published studies that may discuss flow effects on migration rate, passage at dams, or survival. We will include studies for areas other than the West Coast.

Task 2.2 Analyze Data from Qualifying Data Sets to Compare Migration and Survival Before, During and After Pulse Flow Events.

We will subject data sets to various levels of analysis, depending on the amount of information they contain. We will only analyze data sets which include a pulsed flow event.

Activity 2.2.1 Enter Data into Spreadsheet. Some data sets will be available in electronic format, but many will be on data sheets. This activity is similar to that described for Activity 1.2.1.

Activity 2.2.2 Compare Frequency Distributions of Migration Parameters Before During and After Flow Pulses. Our experience indicates that a change in flow can act as either a temporary stimulus or as a retardant to migration, but we are not sure how long these effects lasts. Therefore we will compare frequency distributions of counts, migration rates, passage times, fallback rates, and survivals for varying periods of time before, during and after sharp changes in flow. We will examine possible uses of both parametric and nonparametric statistics.

Activity 2.2.3 Estimate Parameter Values for Pulse Flow Effects on Migration. After we have examined frequency distributions and decided upon appropriate statistical techniques, we will apply those techniques to estimate parameters and their confidence intervals for flow effects on migration rate, fallback rate, passage delay, and prespawning survival. Our analyses may be as simple as analysis of variance, but are likely to require more complex techniques. Our analyses will seek to address each of the questions listed under Objectives 1-3. In order to answer these questions, we will also have to account for the influence of factors other than flow, such as turbidity, river temperature, size, time of year, and distance to spawning areas. Thus, data on all of these factors, where available, will have to be included in the analyses.

Task 2.3 Analyze Migration and Survival Rates of Radio-Tagged Adults in the Columbia River Before, During and after Pulse Flow Events.

Numerous radio-tagging studies of salmon and steelhead migrating upstream have been carried out in the Columbia River and it tributaries. However, analyses of the data have generally focused on the flow effects over a continuous scale, rather than discrete or temporary effects of flow pulses. Prolonged flow pulses have been recommended in hopes of benefiting juvenile salmonids, and effects on adult salmonids have received little consideration. However, it is probable the radio-tagged adults were being tracked at different locations in the Columbia Basin during times that flows changed substantially. Multiple regression techniques of data analysis cannot be effective if flow operates as a temporary stimulus or retardant to migration. Therefore, analytical procedures employed for Task 2.2 will also be applied to radio tracking data in the Columbia Basin. We will work collaboratively with those already analyzing radio tracking data, in order to complete this task.

Task 2.4 Prepare Paper for Publication.

Findings from this synthesis of information on adult migration will be written in a paper suitable for publication and submitted to a scientific journal acceptable to BPA.

Brief schedule of activities
Fall: Examine flow records and sort the PIT tag database into data sets useful for testing pulse flow effects
Winter-spring: Interview west coast biologists and review literature to locate useful data sets of juvenile outmigrant monitoring and of adult radio tracking.
Spring: Obtain, sort and input portions of data sets that reflect migration behavior before, during, and after pulses in flow.
Summer: Perform statistical analysis of data sets
Oct Prepare rough draft and distribute for peer review
Nov Annual Report
Dec. Prepare one paper on juvenile migration and one paper on adult migration, and submit them to a peer-reviewed journal, such as the Journal of the American Fisheries Society.

Biological need
There is evidence from many streams that rapid changes in flow stimulate downstream migrations of juvenile salmon and steelhead, and the purpose of this study is to assemble and analyze that evidence and determine its implications for use of the Water Budget on the Columbia River. The Columbia River Basin Fish and Wildlife Program (NWPC 1994) identified an urgent need to fully understand how flow affects survival of outmigrating juvenile salmonids. Measure 3.7F of the Program states,

"The Council joins with the National Marine Fisheries Service and other regional interests in insisting that this relationship (between spring and summer flow, velocity and fish survival) immediately receive the highest priority in the region's research efforts."

Measure 3.7F then goes on to direct BPA to:

"As soon as possible, fund additional, independent, third-party scientific evaluations to determine the relationship of flow and water velocity to the travel time and survival of juvenile spring, summer and fall chinook and sockeye salmon."

Although third-party evaluations have been conducted, the best models developed to predict migration behavior and survival fail to account for a substantial portion of the observed variation within and between years. Little analysis has focused on the evidence that changes in flow (distinct from the absolute volume of flow), act as a stimulus to migration. In this proposal, we refer to rapid increases in flow as "pulses" in flow.

The Biological Opinion issued by NMFS in March 1995 for the operation of the Columbia River federal power system specifically identifies the need to understand the effects of flow pulses on survival of juvenile outmigrants. Measure 13.g. under Reasonable and Prudent Alternatives states,

"Potential, overall survival benefits to juvenile migrants from pulsing flows is unknown. The BPA, COE and BOR should cooperate in carrying out such studies."

Because juvenile and adult salmonids are migrating through the Columbia River at the same time, any pulses in flow designed to stimulate migration of juveniles will also expose adults to those pulses in flow. Therefore, we need to understand how both the juvenile and adult life stages will be affected. In the section on adult salmon migrations, the Basin Fish and Wildlife Program states (p. 4-1),

"flow and spill conditions intended to assist juvenile migrants at some dams tend to discourage upstream fish migration, mask the flows that attract fish to the fishway, or induce fallback so the fish must relocate and reascend the ladder."

Measure 4.1 (13.c.) directs BPA, Idaho Power and others to,

"Conduct additional (adult) salmon and steelhead migration studies, and coordinate with ongoing fish migration and behavior such as timing, movement, fallback, straying and other characteristics."

Studies of adult migration and survival in the Columbia River have generally demonstrated that increases in flow create two problems for adult migrants: delay at dams and fallback (Dauble and Mueller 1993). As noted in the Columbia Basin Fish and Wildlife Program, these effects conflict with the benefits that accrue to juvenile outmigrants from high flows. Short term pulses in flow may provide a means of minimizing the negative effects in adult migrants while maintaining most of the benefits for juvenile migrants. In order to determine how well these positive and negative effects can be balanced, we must quantify the effects of short term pulses in flow on adult migrants. Information needs for this section are divided into three objectives. We have listed the key questions which need resolution under each objective

Critical uncertainties
Do the databases of PIT tags and radio tracking include enough data before, during and after pulse flow events to enable quantification of fish behaviors and subsequent survival?

Will researchers of juvenile outmigrations and adult upstream migrations be willing to release their data in the detail that we need to complete our analyses?

Summary of expected outcome
We expect to find numerous repetitions over time and space of how juvenile and adult salmonids respond to flow. We are already aware of several data sets on migration and survival of juvenile chinook, and we present a summary of some findings that demonstrate the utility of this approach.

Studies on streams elsewhere on the west coast have demonstrated that outmigration of juvenile chinook is stimulated temporarily by changes in flow, not by high constant flow. We cite here examples from the Yakima River in Washington, the Rogue River in Oregon, the Sacramento and Klamath rivers in California, and the Snake River. The outmigration of juvenile chinook was studied extensively in the Yakima River for nine years, 1982 through 1990, and in the final report of that study (Fast et al. 1992), the following conclusions were offered:

"Flow induced stimulation of passage is especially pronounced when it occurs on the heels of a number of days of declining flows. Interestingly, the peak of the migratory response to increased flows usually occurs before the discharge peak."

"Inspection of daily passage and flow data has revealed that consecutive days of declining, or even stable, flows are usually associated with declining outmigration rates. It should be noted that descending flows stall passage, even when absolute discharge during the decline remains relatively high. During such periods, smolts accumulate somewhere between Sunnyside and Prosser dams, and are subject to longer periods of vulnerability to predators."

"Stalled migrations are stimulated by rapid increases in flow. The increase need not be especially large, but should be abrupt; gradual increases do not evoke a sharp response in passage. An analysis of natural flow pulses gauged below Sunnyside Dam indicates the "minimal stimulated pulse" should be about 20% of the pre pulse "base flow," and that the pulse should occur over no more that two days."

Similarly, studies in the Rogue River, where the peak outmigration of juvenile chinook is typically during mid summer, showed that a sharp increase in flow during the period of juvenile outmigration, stimulated a sharp, but short lived, increase in the number of outmigrant chinook (Cramer et al. 1985). Cramer et al. (1985) found that a unique event during the 10 year study occurred in 1976 when a record setting freshet caused a sharp increase in flow during early August. Immediately following the increase in flows, the number of outmigrants passing Savage Rapids Dam (RM 173) increased dramatically. However, the peak in outmigration lasted less than one week (outmigration for the season was only about 50% complete), while the river flows remained at double the summer base flow for more than three weeks. Cramer et al. (1985) did not observe similar spikes in outmigration (or flow during the summer) during any other year of the study.

The outmigration of juvenile chinook from the upper Sacramento River has been indexed by catches in rotary screw traps at several locations in recent years. Catches in these traps show effects of increases in flow at various times of year. In 1993, Cramer et al. (1993) fished a rotary screw trap in the Sacramento River near Wilkins Slough (Rm 118) from late April through mid July, and observed substantial peaks in outmigration during late April and again in the first week in June. The peak outmigration in late April was composed predominantly of hatchery fish that had just been released from Coleman Hatchery, but the peak on June 1 and 2, which was atypical for the Sacramento River, coincided with an unseasonal flow event. The sharp increase in flow in June clearly stimulated some fish to migrate for about two days, although flows remained elevated for over two weeks.
Craig (1994) examined the effects of pulsed flows in the Klamath River on outmigration of juvenile chinook. Craig (1994) found that pulsed flow had little effect in early May, but strong effect in mid June. The flow pulses tested were only 300 cfs increases above a base flow of about 2500 cfs. Four pulses were tested during May and early July. Craig (1994) reported, "The initial migration rates for IGH (Iron Gate Hatchery) released fingerling chinook during 1989 (20 rkm/day), 1991 (9 rkm/day), and 1992 (30 rkm/day) were substantially lower than observed in 1994. The increased rate of migration observed in 1994 indicates that the pulse flow of June 16 benefitted hatchery chinook by decreasing travel time." "It is further suggested that until fish reach appropriate physiological development or "readiness" to migrate, increased or pulsed flow events may do little other than displace fish downstream."

For adult migrating upstream, we expect to find that pulses in flow have variable and temporary effects migration behavior. We should be able to determine how the magnitude and duration of pulses modulate these effects on behavior. We expect to find sufficient evidence to determine whehter these changes in behavior are likely to affect prespawning survival.


Figure 1. Timing of subyearling chinook passage at Lower Granite Dam on the Snake River during 1994, compared to river flow. Data from Fish Passage Center, Columbia Basin Agencies and Tribes, Portland, Oregon.

REFERENCES CITED

Craig, J.L. 1994. Klamath River juvenile salmonid emigration monitoring pulsed flow evaluation and Klamath Basin drought impacts 1994 salmon production. Preliminary Report. US Fish and Wildlife Service, Coastal California Fishery Resource Office, Arcata.

Cramer, S. P., and J.T. Martin. 1978. Rogue basin fisheries evaluation program, juvenile salmonid studies. Oregon Department of Fish and Wildlife, Fish Research Project DACW57-77-C-0027, Annual Progress Report, Portland.

Cramer, S. P., T. Satterwaite, R. Boyce, and B. McPherson. 1985. Impacts of Lost Creek Dam on the biology of anadromous salmonids in the Rogue River. Submitted to U.S. Army Corps of Engineers.

Cramer, S.P., E. S. Van Dyke and D. B. Demko. 1993. Evaluation of sound and electrical fish guidance systems at the Wilkins Slough diversion operated by Reclamation District 108. Annual Report submitted to Reclamation District 108, Grimes, California, 110 pp.

Dauble D.D., and R.P. Mueller. 1993. Factors affecting the survival of upstream migrant adult salmonids in the Columbia River basin. Technical Report, Recovery Issues for Threatened and Endangered Snake River Salmon. Prepared for BPA, project #93-026.

Fast, D., J. Huble, M. Kohn, and B. Watson. 1991. Yakima river spring chinook enhancement study. Yakima Indian Nation Fisheries Resource Management, Final Report, May 1991. Prepared for U.S. Department of Energy, Bonneville Power Administration. Project No. 82 16.

Northwest Power Planning Council (NWPC). 1994. 1994 Columbia River Basin fish and wildlife program. Northwest Power Planning Council, Portland, Oregon.

Dependencies/opportunities for cooperation
We are dependent on a number of other researchers to release portions of their data for our analysis. We fully expect that some data sets will not be released to us. On the other hand, there are a large number of data sets existing on juvenile outmigrations and radio tracking of adult upmigrations that we should easily obtain a sufficient variety of data to accomplish our objectives. The process of obtaining some of these data sets is likely to be time consuming.

Risks
Other researchers from whom we obtain data may perceive risk that their data sets will be misused. To help overcome this perception, we will involve them in ongoing review of our analyses.

Monitoring activity
We will aggressively seek peer review by:
' Participating in regional coordination and planning meetings regarding juvenile and adult passage survival
' Scheduling meetings in early summer to review our findings with other groups working on smolt survival, including NBS, NMFS, FPC, IDFG, WDFW and ODFW.
' Distributing progress reports for peer review
' Hiring statistical consultants for critical review

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   Mainstem

Recommendation    Tier 3 - do not fund