Friends of the Cowlitz

WDFW has submitted a genetics research proposal, so instead of doing our own study we will be
partnering with them.

We will still have our own Fisheries Scientist and volunteers participating in the study.

This is a win as it will cut our costs considerably while still giving us input and oversite.

WDFW RESEARCH PROPOSAL

Authors: Anne R. Marshall, Wolf Dammers, and Julie Henning
Washington Department of Fish and Wildlife
Fish Program - Science Division, Conservation Biology Unit & Fish Management Division, Region 5
January 2008

TITLE

Genetic relationships among naturally spawning Steelhead (Oncorhynchus mykiss) in Lower Cowlitz River tributaries and hatchery Steelhead stocks released in the Cowlitz Basin: implications for recovery planning

INTRODUCTION

Naturally spawning steelhead in lower Cowlitz River tributaries are included in the Lower Columbia River Evolutionarily Significant Unit (ESU), which was ESA-listed as threatened in 1998. Within this ESU, there is a high degree of uncertainty regarding the existence of native steelhead populations or gene pools, and the impact of hatchery-origin steelhead on natural populations in terms of genetic diversity, population structuring, productivity, and viability or persistence. Busby et al. (1996) concluded that the major threat to genetic integrity of Lower Columbia ESU steelhead occurred from past and present hatchery practices.

The late-winter steelhead stock produced at Cowlitz Trout Hatchery is derived from native Cowlitz Basin steelhead and may include a mixture of formerly distinct populations from drainages upstream of Mayfield Dam. Genetic characteristics of the late-winter hatchery stock were distinctive compared to the non-native Cowlitz Hatchery early-winter and summer-run stocks, and to other Lower Columbia population samples (Phelps et al. 1997). No genetic data exist for natural-origin winter steelhead in Lower Cowlitz tributaries or mainstem. It is likely that native Lower Cowlitz fish have a shared ancestry with the late-winter Cowlitz Hatchery stock. Due to past transplantation of non-native hatchery-origin fish and high levels of on-station hatchery production it is reasonable to expect that Lower Cowlitz populations have been affected by genetic introgression and/or ecological impacts from hatchery stocks. However, differences in spawn-timing and potential lower reproductive success of hatchery stocks could limit their impacts on native wild-spawning populations.

In order to protect remaining genetic resources and plan appropriate recovery actions, it is essential to genetically characterize natural-origin steelhead in Lower Cowlitz tributaries. We propose a minimum two-year sampling of natural-origin adults returning to at least three major spawning areas and a comprehensive genetic comparative analysis to determine within-basin population structuring, and to estimate hatchery stock introgression. We plan to sample adults from the three Cowlitz Trout Hatchery stocks for comparative analyses. Genetic data from existing wild and hatchery steelhead samples from the Toutle sub-basin, Coweeman River, Lewis and Kalama basins will also be utilized in comparative analyses.

Restoring wild steelhead abundance and productivity is a goal for ESA de-listing. We believe data from this study will allow the development of recovery strategies that have the best chance of succeeding. These data may also assist in prioritizing habitat restoration actions and should be informative for hatchery management.

STEELHEAD STOCK STATUS

The 2002 update of the Salmon and Steelhead Stock Inventory (SaSI), a product of WDFW and Western Washington Treaty Indian Tribes, described Cowlitz winter steelhead status as "Unknown" because no adequate abundance trend data were available, but could be "Depressed" because access to 80% of historic habitat has been lost due to dam construction. Only a single winter steelhead stock was identified in the Cowlitz River, based simply on spawning distribution (SaSI 2002). Four other winter steelhead stocks are identified in the Cowlitz Basin- Coweeman, Mainstem/North Fork Toutle, South Fork Toutle, and Green River (SaSI 2002). Summer-run steelhead were not historically present in the Cowlitz Basin (Myers et al. 2002), and no current populations have been identified (SaSI 2002).

COWLITZ WINTER STEELHEAD DISTRIBUTION

Most spawning downstream of Mayfield Dam takes place in the lower mainstem
Cowlitz River and in Ostrander and Salmon creeks, and also occurs in Olequa, Stillwater, Whittle, Arkansas, and Delameter creeks (SaSI 2002). Tipping et al. (1985) conducted an extensive spawning ground survey in Cowlitz Basin areas below Mayfield Dam (except Toutle and Coweeman rivers) and counted 10 or more redds per mile in the following locations (creeks): South Fork Ostrander, Leckler, mainstem Arkansas, Monahan (Arkansas drainage), mainstem Olequa, mainstem Stillwater (Olequa drainage), North Fork Stillwater, Campbell (Olequa/Stillwater drainage), Curtis (Olequa drainage), Otter, and Mill. Non-ad-clipped (presumed natural-origin) late-winter steelhead are collected at the barrier dam and trucked to the Tilton River and to Lake Scanewa, the uppermost reservoir. Late-winter hatchery stock are also transported to these areas. Spawning occurs in the Tilton River, the Cispus River drainage, and the upper Cowlitz drainage.

EXISTING GENETIC DATA

Steelhead smolts from the three Cowlitz hatchery stocks were sampled for genetic material in 1996, along with wild smolts from the South Fork Toutle and Green (North Fork Toutle) Rivers (Phelps et al. 1997). Based on observed variation of allozymes, Phelps et al. (1997) reported that the late-winter hatchery stock was distinctive compared to the early-winter and summer hatchery stocks in the Cowlitz and to other Lower Columbia stocks. Allozyme-based analyses reported in a memo by Steve Phelps (WDFW, 1997) also indicated that the Cowlitz late-winter hatchery stock was genetically distinct from other hatchery stocks in the Cowlitz (summer steelhead and Chambers Creek winter stock), as well as hatchery steelhead stocks in nearby river systems. Phelps (1997) reported: "The allele frequencies of the late Cowlitz winter-run steelhead are significantly different from all other steelhead collections assayed by this lab to date…. The dendrogram using Cavalli-Sforza and Edwards chord distance with UPGMA clustering combined the early-run steelhead collection (26Cowlitz EWR96) with several lower [Columbia] river streams and then with a group of winter-steelhead hatcheries. The late winter collection (27CowlitzLWR96) did not cluster with any other collection." No genetic sampling has been conducted on natural origin winter-run steelhead in the Lower Cowlitz and its tributaries, other than in the Toutle sub-basin and Coweeman River.

RESEARCH GOAL AND OBJECTIVES

Our primary goal is to characterize the genetic structure and diversity of populations of steelhead originating throughout the Cowlitz River Basin. We will test for genetic differentiation among non-native hatchery steelhead (early winter-run Puget Sound-origin stock and summer-run Skamania Hatchery origin stock), late winter-run Cowlitz-origin hatchery steelhead, and naturally spawning winter-run steelhead populations in Lower Cowlitz tributaries. We will determine whether any natural populations sampled can be classified as genetically distinct native gene pools. Genetic evaluations of hatchery stock introgression will be possible if hatchery stocks are genetically distinct from naturally spawning populations. Our objectives are to non-lethally sample adult, un-marked steelhead as close to their home stream as possible in several tributaries, compare their genetic characteristics to the three Cowlitz Hatchery stocks, to each other, and to natural populations in Coweeman and Toutle rivers. We will also compare Cowlitz Basin steelhead samples with those available from other Lower Columbia tributaries.

Hypothesis of this study include:

1. Un-marked (presumably natural-origin) adult steelhead sampled from different tributaries of the Lower Cowlitz River Basin are genetically distinct from hatchery-origin steelhead.
2. The genetic profiles of un-marked adult steelhead are distinct among different tributaries.
3. Lower Cowlitz River Basin steelhead populations are genetically distinct from steelhead in other Lower Columbia tributaries.

SCOPE OF WORK

Methods and Study Design Approach

To adequately address the genetic impacts of hatchery steelhead stocks on wild-spawning winter steelhead we propose a 2-part phased sampling design. The first phase of the study will collect DNA samples from adult steelhead and the second phase will collect DNA samples from juvenile steelhead that could be progeny of previously sampled naturally spawning adults.

Phase 1 - Adult sampling
Year 1 of the study - Collect DNA samples using hook and line sampling. Collection timeframe will be from late February through June. This method has been successfully utilized in the past. Following training in low-impact fishing and sampling methods, volunteer and agency personnel using angling gear will be deployed in key steelhead spawning streams in the lower Cowlitz River. WDFW will supervise and coordinate the collection of DNA samples. This will include soliciting volunteer interest from angler organizations to help collect the samples.

If this sampling method is successful, we will continue to implement this method in year 2 and if needed in year 3. However, if this sampling method cannot obtain enough samples (less than 25 NORs per site in year 1), then other sampling methods will be deployed. These include installing fishway traps in suitable areas and/or using tanglenets in high-density spawning areas.

Adult, non-adipose fin-clipped, presumed natural-origin steelhead will be sampled at the rate of 25 to 50 fish per year from a few of the following locations: Ostrander Creek, Arkansas mainstem, Monahan Creek, Olequa mainstem, Stillwater Creek, Salmon Creek. We will chose the sites based on population and stream conditions in the year sampling is initiated and the previous two or three years. We will also sample 25-50 natural-origin adult steelhead per year transported to Upper Cowlitz/Cispus drainages because of the possibility that some could originate from below the barrier dam.

Adult steelhead will be sampled at the rate of 50 fish per year for 2 years from the three Cowlitz Hatchery populations: late winter broodstock, early winter broodstock, summer broodstock

A small fin clip will be taken non-lethally from each fish, and pertinent sampling and biological data will be recorded for the individual. Depending on resolving issues of handling stress, we will also sample scales to determine age, freshwater residency period, and check for absence of hatchery yearling release pattern to verify natural origin of unmarked fish. Fish handling is intended to be as minimal and non-invasive as possible, and all fish will be released immediately post-sampling. Each individual sampled will be given a unique number, which will be associated with all genetic, biological and sampling data collected.

The minimum sample size goal for any putative population or broodstock is 50 fish, which could be accumulated over time. Thus, for natural origin populations, and natural origin component of a spawning population, sample sizes between 25 to 50 adult fish will be required per year. In the lower Cowlitz tributaries it is important to sample both unmarked, presumably natural-origin, and marked hatchery-origin steelhead, if they are present. When possible, genetic relationships between the two types will be compared within the tributary to understand past influence and natural production of particular hatchery stocks.

Phase 2 - Juvenile sampling
If it is found that hatchery fish are entering the spawning grounds but little or no introgression is found in natural origin adults from the previous brood years, then juvenile sampling should begin. In Year 3 of the study we will collect DNA samples from juvenile steelhead if needed. Collection timeframe will occur over the entire period of outmigration and for all stream rearing life stages. This phase is important for determining whether (1) hatchery fish are producing offspring in the tributary where they were captured, (2) introgression is occurring and can be detected in juveniles, and (3) the level of juvenile production among hatchery-origin fish may pose an ecological risk to natural population.

Genetic Methods

Laboratory data collection

We will analyze nuclear, microsatellite DNA markers in this study. Microsatellite DNA loci are non-coding, selectively neutral DNA sequences that typically show high levels of variation. We will determine genotypes at 16 microsatellite loci for each fish sampled. The 16 loci are a group that has been standardized for genetic data collection among several laboratories working on Columbia Basin steelhead research. We will extract template DNA from fin tissue and microsatellite alleles will be amplified using the polymerase chain reaction (PCR) with fluorescently labeled primers, then separated by size on a DNA sequencer. Genotypes will be assessed using the appropriate computer software.

Genetic data analyses - microsatellite DNA

Genotypic data will be examined first with a suite of tests to gather basic population information. For each sample we will test microsatellite loci for Hardy Weinberg equilibrium (genotypic proportions expected in a randomly mating population) and linkage disequilibrium. Besides being informative about the nature of population samples, these tests help identify sampling error, data collection error, mixed-origin samples, and whether loci are independent.

To examine whether significant genetic variation exists between natural- and hatchery-origin steelhead, we will conduct a variety of statistical tests. Genotypic differentiation among samples, and between or among brood years, will be evaluated. We will conduct pair-wise tests to determine whether samples differ significantly in allele frequencies. We will evaluate differentiation among samples through cluster analyses, and in factorial correspondence analyses. In cluster analysis, genetic distances between all possible sample pairs are calculated from allele frequencies, and samples are then plotted on a phenogram with placement corresponding to genetic distances. In correspondence analysis, samples are plotted in multi dimensional space based upon allele frequencies, illustrating distances in three dimensions.

We will also evaluate individual fish in a correspondence analysis in which individuals are plotted in multi dimensional space based upon their genotype. Outlier genotypes are visually identified and subtle differences among samples may become apparent. We will also assess data sets for genetically distinguishable samples using the computer program STRUCTURE (Pritchard et al. 2000). This program implements a Bayesian algorithm to identify the most likely number of genetically distinct groups in all sampled individuals and provides probabilities of assignment to putative groups for each individual. Assuming the three hatchery stocks are genetically distinct from each other, any hatchery-origin fish sampled on tributary spawning grounds will be tested for their most likely hatchery of origin.

Estimating introgression from hatchery stocks

Interbreeding between Lower Cowlitz native winter-run steelhead and the two non-native hatchery stocks poses the highest risk to the natural spawning populations. Given the history of hatchery production at the Cowlitz facilities (e.g. Chamber Creek stock steelhead were first released in 1968) interbreeding impacts could have been occurring continuously, assuming strays were always present. The extent of introgression in original populations would depend of the reproductive success of returning offspring from any hatchery by wild crosses that occurred in streams. Returning interbred fish would have the opportunity to spawn with others like themselves, with native fish, and hatchery strays. Introgression of genes from a hatchery stock could increase over time if crossbred offspring are successful at interbreeding with native-origin fish, and hatchery strays are present annually.

Measuring introgression is possible where the interacting populations are genetically distinct and baseline population data prior to any interbreeding are available. Without baseline data for original lower Cowlitz steelhead populations, we can attempt to understand introgression by making assumptions about initial conditions for the populations. Thus, if we assume that lower Cowlitz steelhead would have been most similar genetically to the late winter-run hatchery stock, then introgression from the two non-native hatchery stocks may be the most easily detectable. However, using the current late winter-run hatchery stock as a baseline profile may be problematic because this stock may have diverged from original founders. Genetic data for natural late winter-run populations in the Coweeman and Toutle basins, and their comparison with Cowlitz hatchery stocks may help us gauge introgression levels in lower Cowlitz steelhead. Also, a comparison among natural late winter-run populations in the Coweeman and Toutle basins to natural and hatchery late winter-run Cowlitz populations could help us gauge whether it is appropriate to use Coweeman and Toutle samples as baseline profiles of native populations. We will conduct data analyses exploring potential scenarios to assist in estimating introgression.

MANAGEMENT IMPLICATIONS

WDFW believes it is important to understand the genetic composition of various hatchery origin and natural spawning populations in the Cowlitz River System in order to better manage the hatchery production and fishery harvest. We believe that it is important to look at several different genetic metrics and perspectives that elucidate the degree of distinction (and conversely relatedness) of existing steelhead stocks found in the lower Cowlitz River system - hatchery production, natural production, and out-of-basin strays.

Results from the genetic study are expected to provide a basis for any management actions needed to assist recovery of natural Lower Cowlitz steelhead populations. A variety of possible outcomes from the genetic analysis can be anticipated. Three possible results scenarios and projected management responses are outlined below. These are only given as examples and we do not assume that they model the most likely outcomes.

Potential Outcome 1. All natural populations sampled are significantly genetically distinct from hatchery broodstocks and show little evidence of hatchery stock introgression.

Management response:
Assume any straying hatchery steelhead do not interbreed successfully (i.e. do not produce adult offspring) and that natural populations are capable of maintaining at least a minimum level of sustainability. Through field studies investigate actual straying levels and reproductive behavior of strays. Sample subyearling offspring to determine whether strays produce juveniles. Determine whether any juvenile production poses ecological risks to steelhead or other listed species recovery. If significant risks are present, modify appropriate hatchery production program(s) to substantially reduce these risks. Modifications could include a variety of actions, such as altering release protocols, reducing hatchery stock production, increasing hatchery fish harvest, targeted in-stream strays removal.

Potential Outcome 2. All natural populations sampled are moderately genetically distinct from hatchery broodstocks and show intermediate levels of hatchery stock introgression.

Management response:
Assume straying hatchery steelhead do interbreed successfully with native populations and have an annual production contribution dependent on abundance. Assume that native gene pools are present but at diminished natural production capacity. Through field studies investigate actual straying levels and reproductive behavior of strays. Sample subyearling offspring to determine the magnitude of interbreeding. Modify appropriate hatchery production program(s) to substantially reduce the straying behavior or proportion of strays. Modifications could include a variety of actions, such as altering release protocols, reducing hatchery stock production, increasing hatchery fish harvest, targeted in-stream strays removal. Monitor natural populations to evaluate genetic change in response to reduction of hatchery fish interbreeding.

Potential Outcome 3. All natural populations sampled are not genetically distinct from hatchery broodstocks and are likely offspring from hatchery broodstocks.

Management response:
Assume native populations are extinct in the sampled drainages. Through field studies investigate annual straying levels of first generation hatchery fish and reproductive behavior of strays. Determine whether existing naturally spawning populations are self-sustaining. Assess whether a restoration strategy using a local, native population is possible and feasible. A restoration strategy may require modifications of hatchery production activities.

The potential outcomes and management responses from the study that are listed above are examples of actions the Department may take. WDFW considers the results of this study as one important tool in their toolbox that will be used along with other studies and recommendations to manage the Cowlitz steelhead population and the steelhead hatchery program.

Deliverables from this project include annual progress reports and a final report at the end of the study. The final report will include genetic impact information with steelhead management recommendations.

PROPOSED BUDGET FOR YEAR 1

Salaries and Benefits

Phase 1 - Year 1 Field Sampling Adults -

Scientific Technician 2 (4 mo. @ 3,003 + benefits) $16,660
Sampling Supplies and general equipment $ 1,000
Vehicle mileage $ 3,000

Subtotal $20,660

Laboratory analyses-
Current estimated costs for microsatellite DNA marker analyses are
$43/fish; thus N fish sampled per year X $43 = Total Annual Lab Costs

Potential 1st year maximum: Analyze 300 samples $12,900
Research Scientist 1- design, interpretation, report writing (1 mo) $7,091

Subtotal for field and laboratory work $40,651
Indirect (29%) $11,788
TOTAL - YEAR 1 $52,439

Years 2 and 3 of the study
The study design and budget for years following year 1 will depend on the success in year 1 of collecting DNA from fish in the small tributaries. If year 1 is successful, the sampling methods and budget for year 2 is anticipated to be similar to or less than year 1. However, if few samples are collected than other sampling methods will need to be implemented to obtain samples. These methods will be more costly and may include using fishways and tanglenets. In addition, juvenile sampling may need to be initiated in year 3 depending on adult sampling results. It is anticipated that this study will need to be funded for a minimum of two years but could be funded longer if adult samples are difficult to obtain or juvenile sampling is initiated.