Lake Roosevelt is one of 11 lakes and reservoirs within the Columbia River drainage in Washington State that supports a Burbot population. Standardized stock assessment data is needed to facilitate management of the Lake Roosevelt Burbot population. In the absence of standardized stock assessment data, Burbot populations in Washington State, including Lake Roosevelt, have been managed conservatively. A suspected increase in abundance and high survival suggests that the population in Lake Roosevelt could potentially support greater harvest, although this needs confirmation.

For the 2011 Resident Fish Categorical Review, the Colville Confederated Tribes (CCT) submitted a proposal to develop a standardized stock assessment program for Burbot in Lake Roosevelt. The objective was to facilitate management of the Burbot population to provide for sustainable harvest through the development of a status and trends monitoring program. Because burbot are commonly observed in the bycatch of annual Lake Roosevelt Fall Walleye Index Netting (FWIN) surveys, the Independent Scientific Review Panel (ISRP) and Northwest Power and Conservation Council (NPCC) recommended that the CCT first evaluate the potential for using FWIN as the basis for Burbot stock assessment in Lake Roosevelt. Accordingly, in 2013 the CCT hired a subcontractor to analyze ten years of FWIN data (2003-2012) and assess its adequacy for Burbot stock assessment. The overarching conclusion of the analysis was that the FWIN program is adequate for monitoring year-to-year differences in Lake Roosevelt Burbot population indices.

Having determined that FWIN is generally adequate for Burbot stock assessment in Lake Roosevelt, we working toward our goal of a achieving a Burbot population that has stable abundance and is managed to provide some harvest. The primary objective of this project is to provide technical advice to the Lake Roosevelt Co-Managers regarding management alternatives so that they can develop realistic fishery targets and appropriate implementation strategies. Our approach will be to estimate the harvest potential of Lake Roosevelt Burbot under current management and population characteristics using Fishery Analysis and Model Simulation (FAMS) software. Model inputs are recruitment, mortality (natural and fishing), and growth, which will be derived from the FWIN data with some supplemental field sampling to evaluate biases. Thus, we (with a subcontractor) will analyze the entire FWIN data set with the 2014 data included.

Despite the demonstrated utility of FWIN as a tool for Lake Roosevelt Burbot stock assessment, there is a need to evaluate the size selectivity of the gill net gear. Gill net size selectivity can be estimated indirectly by fishing multiple mesh gill nets with equal effort and comparing catch frequencies within each of the mesh sizes. We will use a maximum-likelihood approach to estimate parameters of log-linear selectivity models (Normal location, Normal scale, Gamma, Lognormal, and Bi-normal), also known as the SELECT method (Share Each Length’s Catch Total), and correct the length data prior to the calculation of length-based stock assessment indices.

Age bias resulting from gill net selectivity can be corrected by applying the proportion of fish of each age within each length bin to the bias-corrected length data. To increase sample sizes of small (young) and large (old) Burbot, supplemental sampling with alternative gears will be conducted. We will conduct supplemental sampling concurrent with FWIN using towed trawls (to capture small fish), modified cod traps (large fish), and trammel nets (large fish) – all of which have been used to sample Burbot populations.

Recruitment variability can complicate the management of a fishery. If we can identify density-independent factors that are affecting recruitment, then we may be able to address them in a manner that would provide for more stable recruitment at a higher level. We will investigate the factors that are responsible for the apparent recruitment variability in the Lake Roosevelt Burbot population. Multivariate statistical techniques will be utilized to explore relationships between mean age 2 catch per unit effort in FWIN gill nets and the abiotic and biotic conditions in the year of birth for each year class.

The purpose of this project is to improve understanding of the factors that influence population abundance of burbot (Lota lota) in Lake Roosevelt. The Lake Roosevelt population, which is considered healthy, is an important fishery resource for local stakeholders, but interest in expanding fishing opportunities for the species has raised concern about the potential effects of increased harvest rates on population productivity. Although catch rate and distribution data for burbot in Lake Roosevelt have provided baseline information regarding population status, important knowledge gaps regarding the age, spawning frequency and spawning habitat of the adult population remain. These knowledge gaps complicate the efforts of the resource managers to identify the appropriate age structure and biomass for annual harvests that will help ensure a sustainable fishery. Moreover, Lake Roosevelt is subject to wide variation in flow and water level elevation that may affect spawning success and recruitment. The development of reliable techniques to determine age-specific contributions to recruitment and identify critical spawning habitat will provide managers valuable tools to implement policies and regulations that aid in effective management of burbot in Lake Roosevelt.

Burbot are a popular recreational fish in many North American lakes and rivers. Although protected at the state level in some areas, the population of burbot in Lake Roosevelt has been described as stable or increasing (Bonar et al. 2000; Polacek et al. 2006). However, information regarding the productivity of the population is limited. Data collected through the Lake Roosevelt Fishery Evaluation Program from 1988 to 2002 and reported by Polacek et al. (2006) showed that catch rates for burbot ranged widely over this time period (~ 10-fold), peaking in the late 1990’s. The reasons for these different catch rates are varied, but Polacek et al. (2006) suggested that both an increase in prey availability (hatchery releases of kokanee and rainbow trout) and favorable spawning conditions may have been contributing factors. Moreover, while they note that nothing is known about quantity or quality of burbot spawning habitat in Lake Roosevelt, other studies have shown that burbot move into shallow areas in the fall as part of a pre-spawning feeding migration (McPhail and Paragamian 2000). Catch rates in Lake Roosevelt were also highest in bays and inlets of tributary streams including the San Poil River, Colville River and Hawk Creek, particularly in the fall, suggesting these areas may be similarly important for pre-spawning and possibly spawning fish. Identifying use of these tributary habitats may be important for understanding population dynamics of burbot in Lake Roosevelt because the reservoir is subject wide annual variation in flow and water level elevation, which can alter movement and impact spawning success. Bergeresen et al. (1993), for example, showed that extreme drawdown eliminated access to virtually all suitable spawning and nursery habitat for burbot in Bull Lake (WY). Krueger and Huber (1997) also cited water level fluctuations as a probable factor limiting spawning success of burbot in the Big Horn/Wind River drainage in Wyoming.

Spawning frequency of sexually mature burbot is also currently unknown, and is assumed that these fish can skip spawning cycles in a given year (Pulliainen and Korhonen, 1994). In order to determine whether a burbot will successfully spawn, or skip spawning, a comprehensive endocrine profile is needed that examines not just plasma levels of the main reproductive hormones (estradiol and testosterone), but prior steroid precursors and indicators of oocyte maturation, yolk production, and spermatogenesis. Fish that skip a spawning year may arrest spawning progression at several stages, and a recent study has shown that burbot raised in a hatchery can proceed to the vitellogenic age stage prior to absorption of eggs (Foltz et al, 2012). Rideout and Tomkiewicz (2011) examined the phenomenon of skipped spawning in cod, and found that endocrine profiles showed distinct annual expression levels that were indicative of spawners compared to non-spawners. This characterization of annual endocrine profiling can be used to determine spawning frequency for use in subsequent stock assessments and biomass contribution for a managed fishery.

The age structure of the burbot population in Lake Roosevelt is presumably dominated by fish = age 5 (i.e. 81% of fish collected between 1999 and 2001, range 3-10). Whether this relates to the sampling methods or locations is uncertain, but the lack of younger fish in the population could indicate poor recruitment for the 1997-1998 year classes. The methods used for ageing burbot (whole otoliths for fish = age 5, burned thin sections for fish > age 5) are also likely to influence the age structure of the population, particularly if they do not provide reliable estimates in older fish (Edwards et al. 2011). Finally, measures of population productivity may be further confounded by the fact that not all adults spawn every year (McPhail and Paragamian, 2000). Burbot populations vary widely in the proportions of fish that do not spawn throughout their natural range, from only 1.5% in Lac Sainte Anne, Alberta (Boag 1989) to 30% in Bothnian Bay, Finland (Pulliainen and Korhonen 1990), and if this variation also occurs on an annual basis within populations, productivity estimates will be even more uncertain. Interestingly, Pulliainen and Korhonen (1994) found that otolith formation differed among regular and irregular spawning populations of burbot in Finland, which could provide an additional method to determine spawning frequency from population surveys. Calcium demand increases significantly in spawning fish for gamete formation (e.g. vitellogenin is a calcium binding protein) (Carragher and Sumpter 1991, Armour et al. 1997, Persson et al. 1998), an event that can potentially alter the calcium isotopic composition between soft and mineralized tissues (Skulan and DePaolo 1999, Skulan et al. 2007). If this isotopic variation is recorded in otoliths it could be used to determine the frequency of spawning among and within individual burbot. The uptake of other elements such as strontium (Sr) and barium (Ba) into otoliths have shown temperature dependence in a number of species (Townsend et al. 1992; Martin et al. 2004; Elsdon and Gillanders 2004) and may provide an additional measure of spawning frequency due to seasonal variation in water temperature. For other species (e.g. Atlantic salmon) Sr uptake appears to be independent of temperature and possibly a function of maturity state (Clarke and Friedland 2004). This relationship between maturation and Sr concentration in otoliths also has been experimentally confirmed (Mugiya and Tanaka 1995), suggesting that element uptake may respond directly to endocrine effects.

Taken together, the limited information on age structure, spawning frequency and habitat use by burbot in Lake Roosevelt will continue to complicate management efforts to maintain a healthy population and sustainable fishery. To improve understanding of burbot productivity in Lake Roosevelt, the Pacific Northwest National Laboratory (PNNL) will assist the Confederated Tribes of the Colville Reservation (CTCR) in conducting a stock assessment to help identify the critical factors affecting population abundance. PNNL is uniquely suited to provide support for this effort with laboratory facilities and equipment such as molecular physiology laboratories for assessing reproductive hormone profiles and a geosciences laboratory to age and characterize movement and habitat use with otoliths. The investigators associated with this project have expertise in the methods listed in the objectives below with several publications and/or current research using the facilities and capabilities proposed for this project. The proposed project will be conducted in two phases:

Phase 1: Pilot studies to develop baseline information for endocrine markers and techniques for ageing and geochemical assessment of otoliths.

Phase 2: Implement the results from Phase 1 pilot study to confirm the utility of endocrine markers to predict spawning frequency through a controlled laboratory experiment, and confirm the use of otolith structure and geochemistry as a tool to verify spawning frequency and habitat. Analyze archival otolith samples of Lake Roosevelt burbot using the most appropriate ageing techniques developed in Phase 1.

Project Tasks for Phase 1. The tasks described below are the only activities proposed in year 1 under this contract. The Phase 1 studies are intended to provide proof of concept data only. Specifically, Task 1 will provide an initial measure of reproductive markers at a single time point and to validate use of assays for plasma endocrine expression in spawning burbot. The results from Task 2 and Task 3 will be respectively used to refine the techniques and protocols to increase throughput of archived otoliths for age structure assessment of burbot in Lake Roosevelt, and provide a potential basis to increase accuracy and precision of age estimates in older (> 7-8 years) burbot.

Task 1. Measure blood estradiol and vitellogenin in N = 50 female burbot and blood 11-keto-testosterone in N = 50 male burbot. The efficacy of characterizing additional blood assays such as gonadotropin releasing hormone (GnRH), dopamine, maturation inducing steroid (MIS), and gonadotropic hormones will also be tested. Blood samples will be collected by CCT from reproductively mature burbot during spawning surveys in March-April 2015 and shipped to PNNL for analysis. The protocols for collection and sample preservation will be provided to CCT by PNNL.

Task 2. Determine the age from thin sections and light microscopy in N = 25 burbot otoliths. Otoliths of burbot collected by CCT for Task 1 will be shipped to PNNL and analyzed for age following the protocols described above (2.a).

Task 3. Evaluate the chemical composition in the otoliths used for age determination of N = 15 burbot by LA-ICP-MS to determine if annulus formation relates to variation in elemental (e.g. Sr, Ba, Mg, Cd, U) and (or) isotopic (87Sr/86Sr) signals.

Phase 1 Deliverables: A draft report will be prepared to describe findings of the pilot studies and their application to the second phase studies of the project. The results will include all data for the sex specific endocrine markers, the estimated ages of burbot from their otoliths, and elemental and isotopic composition in these otoliths. The report will provide the CCT with preliminary information to determine whether the Phase 2 studies should proceed and will include the experimental design, objectives and tasks for these additional studies and the basis for conducting future stock assessment of burbot in Lake Roosevelt.

Based on the results of the Phase 1 pilot studies, the CCT will decide whether to proceed with Phase 2 investigations, which would occur under subsequent project contracts. The CCT will also determine if Phase 2 activities can be conducted internally (by CCT staff) or contracted out. If any aspect of the Phase 2 studies are deemed to require contracted support, CCT will seek competitive bids or conduct a market survey to ensure reasonable cost.