Draft to BPA, 2/22/05
Lake Pend Oreille Fishery Recovery Project
Statement of Work and Budget FY2005

BPA Project Number:  1994-047-00
BPA Project Title: Lake Pend Oreille Fishery Recovery Project
Contract Number:  16828
Contract Title: Lake Pend Oreille Fishery Recovery Project
Performance/Budget Period:  March 1, 2005 to February 28,2006

Technical Contact Name:  Melo Maiolie
Technical Contact Title: Principal Fishery Research Biologist
Idaho Department of Fish and Game
16805 Limekiln Rd. PO Box 806
Bayview, ID     83803
Technical Contact Phone: (208) 683-9218 / Fax: (208) 683-3054
mmaiolie@adelphia.net

James Lau
Chief of Administration
Idaho Department of Fish and Game
600 S. Walnut St, PO Box 25
Boise, ID 83707
Contracting Contact Phone: (208) 334-3781 / Fax: (208) 334-2148
jlau@idfg.state.id.us

Financial Contact Name: Conan Chiu
Financial Specialist
Idaho Department of Fish and Game
600 S. Walnut St, PO Box 25
Boise, ID 83707
Financial Contact Phone: (208) 287-2813 / Fax: (208) 334-2148
cchiu@idfg.state.id.us

This project addresses the considerable impacts of the federal hydropower system on kokanee, and the indirect effects to bull trout, and rainbow trout in Lake Pend Oreille (Figure 1).  It also addresses the losses of warm water fish habitat that occurred in the Pend Oreille River above Albeni Falls Dam.  Drawing the lake down to the maximum extent possible each fall (elevation 2051'msl) caused major impacts. This reduced the amount of gravel along the shoreline of the lake that is usable for kokanee spawning. It also draws the Pend Oreille River down into the river channel causing the loss of much of the warmwater fish habitat.  Fish populations declined commensurate with these habitat losses.  
In 1996 the Northwest Power Planning Council (NWPPC) requested that the Corps of Engineers change lake levels to benefit fisheries.  Results of the first 5 years of changed lake levels showed a 150% increase in egg-to-fry survival during the years of elevated lake levels (excluding the flood year of 1997)(p=0.06).  The ISRP recommended increasing the length of the study to 10 years to make the results more definitive.  The USFWS, in its draft Biological Opinion for Bull Trout in Lake Pend Oreille, utilized the previous study's findings to recommend that the Corps change its management of lake levels.  These changes were requested to increase kokanee abundance and thereby increase the forage for bull trout.  Although not directly intended, a side benefit was that warm water fish habitat in the Pend Oreille River would increase seven fold.  The goal of this proposal is to recover these impacted fish populations and to protect the threatened bull trout population. We propose to document the response of the fish populations to these new lake levels.  We also propose to monitor changes in the shoreline spawning habitat to see that it is maintained in good quality.  Information gained by this study will be presented to the USFWS in its annual review of lake level changes.  This allows flexibility in the management of lake levels in future years.  However, the kokanee population remains low.  The stocking of hatchery-produced kokanee may be necessary to keep kokanee from disappearing in the short term.  Past stocking, however, has not produced the anticipated increase in adult kokanee.  Since all hatchery kokanee for the last 5 years have been marked, we have the unique opportunity to examine and compare the survival of each year class of hatchery and wild fish.   By doing so we can determine when survival is relatively the poorest, and gain insights into how to improve the hatchery program.
Overall problem: Lake Pend Oreille was the most productive resident fishery in the state of Idaho until the mid-1960's.  An average of one million fish were caught annually between 1952 and 1966 (Maiolie and Elam 1993).  Albeni Falls Dam was built on the outflow of the lake in 1952 and Cabinet Gorge Dam was built on the inflow the same year.  Blocking the Clark Fork River (by Cabinet Gorge Dam) caused immediate declines in bull trout and cutthroat trout abundance since spawning areas became limited.  The lake remained an exceptional fishery for kokanee and Kamloops rainbow trout until the mid-1960's when Albeni Falls Dam began lowering the lake 11 feet each fall for additional power production.  This started a down turn in the kokanee population that continues to this day.
Maiolie and Elam (1993) reported that deeper drawdowns in 1957, and 1961 reduced the kokanee fishery five years later.  Conversely, higher water levels in the winter of 1958, 1959, and 1960 produced better fisheries five years later.  They also found that somewhat higher lake elevations in 1982, 1983 and 1985 produced 39% more kokanee recruited to the fishery than years before or after these years.  Fredericks et al. (1995) found that the two best years on record for egg-to-fry survival were 1982 and 1983 when the lake was 3 to 4 feet higher than normal during the winter.  A second important factor is that the lake cannot be lowered once kokanee spawning has taken place (Hassemer 1984, and Maiolie and Elam 1993).  Drawdowns of less than 1 m can have a noticeable effect on the resulting fisheries if they occur during the egg incubation period; mid-November to June.
The mechanism for why lake levels control fish abundance can be explained rather simply.  Shoreline gravels around the lake rely on wave action to remain clean and well sorted from sand and cobble.  At whatever elevation the lake is held, gravel bars appear just above this elevation.  Fredericks et al. (1995) documented that quality spawning gravel was relatively sparse below the low pool elevation likely because wave action had not affected this part of the shoreline since 1953 when the dam became operational.  The consistency of drawing the lake to the same elevation each year prevented clean wave washed gravel from forming below the low pool elevation.
The previous studies funded by BPA demonstrated the effect of changing lake levels. At the beginning of these studies, surveys of potential spawning areas around the lake found that an additional 1.8 million square feet of gravel would be available for kokanee spawning if the lake were held 4 feet higher throughout the winter (Fredericks et al. 1995).  Fredericks estimated only 380,000 square feet of gravel were below the low pool elevation.  Improvements in kokanee survival were noted after raising the lake elevation during winter for four consecutive years.  Kokanee egg-to-fry survival increased from 1.4% in 1995 to 10.0% in 2000 (Figure 2).
Previous studies also monitored the quality of gravel in shoreline spawning areas. Good spawning gravel (a high percentage of gravel, low in fines and cobble) declined after being underwater for a period of years. Degradation was not due to an increase in the amount of fine material, but rather, due to an increase in the percent of cobble (material over 32 mm)(Maiolie et al 2000).   After three years, degradation was very noticeable.  This led researchers to the recommendation to lower the lake once every four years.
A different situation exists on the Pend Oreille River.  This 23 mile long river bottom is flooded during the summer and contains good habitat for warm water species such as bass and perch.  Drawdowns during winter drop the water into the old river channel.  This area then becomes a cold flowing river with little over-winter habitat for warm water species.  Bennett and DuPont (1993) calculated over-winter habitat would increase by 7.5 times if the winter elevation were 5 feet higher than normal.   Experimentally changing the lake elevation appeared to have benefited largemouth bass in the river (Karchesky and Bennett 2000).  Good year classes were produced during the winters of high water and these year classes continue to remain strong.  Drawdowns to an elevation of 2053' appeared to cause lower over-winter survival of young-of-the-year bass.  
A review of this project by the ISRP recommended that this project extend the lake level changes for 10 years instead of three.  This would allow more "sample" years to be collected and strengthen our confidence in the results.  In 2000 the USFWS Biological Opinion recommended lake levels be changed to benefit kokanee and thereby help bull trout.  Their plan was consistent with the projects previous findings, the recommendations of the ISRP, the Subbasin Summary, and the direction of the Council's program.  We therefore have the opportunity to continue to study the effect of lake level changes should this proposal be funded.
The draft Biological Opinion called for a low pool elevation during the winter of 2001-2002, with a 4 ft higher lake level during 2002-2003.   The low level gave wave action a chance to resort shoreline gravel.  During the winter of 2002/03, lake levels were higher giving kokanee a chance to spawn in clean gravel.  A full drawdown to 2051' was done during the winter of 2003-2004 followed by a higher winter pool level in 2004-2005.  We propose to monitor kokanee survival changes and annually compare them to lake elevations.  Information will then be reviewed by the USFWS and used to formulate the next year's lake management strategy.
The Lake Pend Oreille Predation Research Project, 2002-009-00, was combined with this project in 2004.  It attempts to develop the means to balance predator and prey in the lake.  Sport fisheries in the lake have been severely impacted by hydropower development on both the inflow and the outflow of the lake (Maiolie and Elam 1993, Fredericks et al. 1995 , Paragamian et al.1991).  Consistently drawning the lake down 11 feet annually has taken  the adult kokanee population from an estimated 5 million fish (able to support a 1 million fish harvest) to a population unable to support a sport fishery.  The low kokanee population has declined to the point where it can no longer support the populations of rainbow trout, bull trout, and lake trout in the lake.  (The diet of all three of these species is 60 to 80 % kokanee.)  Until kokanee recover, it appears the predator population needs to be reduced.  To our knowledge, no literature exists which describes the balance point between a kokanee population and its predators.  However, it appears we have passed this point, since kokanee survival rates from age 1 to age 2 have dropped from 80% to less than 20%, and are only now beginning to improve.  This bottleneck in the population has prevented good year classes of kokanee from recruiting to the older age classes and benefitting the next generation.  Reductions in survival of this magnitude could extirpate kokanee from the lake.  Should kokanee collapse, bull trout, rainbow trout, and lake trout will be in direct competition for a limited forage base.  This consitutes enough of a threat to the bull trout population that the USFWS in their draft Biological Opinion has recommended that the action agencies "shall evaluate the effects of varying winter lake level elevations on all life stages of kokanee in Lake Pend Oreille, and predator/prey dynamics."
This project proposes to develop methods to quantify pelagic predators in the lake (rainbows and bull trout) by using hydroacoustics.  The uncertainty we propose to solve is to identify the species of fish  seen on the echosounder. This will be done through cluster analysis of the echograms, and tagging studies, to show the habitat use of each species.  By using hydrosonic tags and/or recording tags, we can learn the depth and temperature used by these predators at night, and then enumerate them on the hydroacoustic surveys. Secondly, the project proposes to define the balance point between predators and prey.  Previous bioenergetic studies, with their many assumptions, have failed to estimate this point with much confidence. The kokanee population showed sharp declines in survival once the ratio of  predator to prey rose above 1 kg of predators to 6 kg of prey.  Therefore we will use this as a starting point for our work.  Lake Pend Oreille at the current time is a real world example of the conditions that exist as a kokanee population collapses. We can gain valuable empirical data about the balance of predator and prey by monitoring both predator and prey abundance, enhancing the kokanee population, and reducing the predator population.  We can then document kokanee survival rates and abundance to determine the balance point.  Balancing predator and prey will be considered from the standpoint of mass balance and production versus yield.
Lake Pend Oreille, Idaho, is nationally recognized for its trophy fisheries.  It holds the title for producing the world record rainbow trout (37 pounds), the State record bull trout (32 pounds), and numerous lake trout over 30 pounds.  These trophy species feed on what was once an abundant population of kokanee.  Kokanee are nearing the point of collapse and populations are too low to sustain the current level of predation.  Kokanee have been declining for the last 34 years due to the consistently deep winter drawdowns of the lake (Maiolie and Elam 1993).  
Currently there is no accurate method to estimate kokanee predators.  Studies by Vidergar (2000) used mark and recapture to estimate predator numbers.  Estimates took two years to complete and had confidence intervals of +/- 80%.  We propose to develop a hydroacoustic method that will give an accurate estimate of the pelagic predators on an annual basis. Hydroacoustic estimates of kokanee have been conducted annually since 1995, and demonstrate the feasibility of this approach.  However, a more intensive survey design is needed, which covers a great deal more water.  Also, large fish on the echograms need to be identified to species.  This could be accomplished by knowing the nighttime habitat preferences of the various species at different times of the year.  Most hydroacoustic assessments require the partitioning of acoustically derived abundance estimates into various species based on auxiliary information (Thorne 1983).  The depth of the fish, water temperature, nearness to the bottom, and depth of water will be used in a cluster analysis to define the fish assemblages.   Tagging studies will then be used to define the habitat preferences for the lake's predatory species, starting with rainbow trout.  Trawl data on fish species will be collected by the Lake Pend Oreille Fishery Recovery Project and will help to define the distribution of kokanee.  Because of the low sample volume by the trawl net, however, collection of rainbow trout or bull trout is rare.
Numerous studies discuss the balance of predator and prey, dating back to Swingle's articles in the 1950's.  To our knowledge, no prediction of balance for a kokanee prey base exists in the literature.  Many recent articles utilized bioenergetics as an approach to finding the balance point. Some studies, such as Negus (1995), point out the many assumptions and erroneous results that occur.  Vidergar (2000) used bioenergetics to determine that predators were eating approximately 154 tons of kokanee per year (+ or - more than 80%), or roughly 60% of kokanee production.  In this project we propose to empirically define the balance point by monitoring both predator and prey as the situation improves. This will be done two ways.  First we propose a mass balance approach that compares pounds of pelagic predators to pounds of prey.  These ratios will be compared to kokanee survival rates to determine the balance point.  Secondly, we will estimate kokanee production and yield at different densities to determine the point at which kokanee production can no longer sustain the amount of yield.

Relation to subbasin summary-   The Subbasin Summary for the Mountain Columbia Province describes the history of kokanee declines in Lake Pend Oreille.  It also describes the loss of kokanee spawning habitat as being the main cause for the declines.  The subbasin summary describes the Council's past efforts on Lake Pend Oreille to address kokanee declines by changing the management of the lake levels.  This project is a continuation of previous efforts with additional support from the USFWS.  There are two limiting factors for kokanee defined in the summary.  They are the loss of spawning habitat and predation.    
Objective #1 under the Goal #1 was to recover kokanee to three quarters of their former abundance.  This is also the heart of this proposal and our tasks reflect the "recommended actions" in the subbasin summary including: 1) developing changes in lake level management, 2) monitoring kokanee by trawling, hydroacoustics, and fry netting, 3) conducting spawner counts, 4) documenting the changes in shoreline gravels as lake levels change, 5) annually review lake level management with USFWS, 6) conduct baseline limnology, and 7) research the potential for building artificial spawning areas.  In Strategy #2 under Goal #1, the summary stated to "evaluate the contribution of Cabinet Gorge Hatchery to the recovery effort".  We propose to do three of the recommended actions including: 1. mark all hatchery kokanee and compare the survival rate of wild and hatchery fish, 2. determine the percent return of hatchery kokanee to Sullivan Springs Egg Collection Station, and 3.  determine if hatchery kokanee are impacting the wild kokanee by examining wild kokanee survival during high and low stocking years.  
The predation work in this project addresses limiting factor of predation in the Subbasin Summary.  Because of the low kokanee abundance (due to drawdowns which affect spawning habitat) predation levels appear too high and have limited kokanee survival from age 1 to age 2 to less than 20% (Maiolie et al. 2000).  The summary's second objective under Goal 1 is to balance predator and prey populations to a 1: 6 biomass ratio within the next 3 years.  The summary's stated tasks are to develop a hydroacoustic method to determine pelagic predator abundance, monitor kokanee survival rates, and re-balance predator and prey if needed through active removal and changes in fishing regulations.
Limiting factors for bull trout are also stated in the Pend Oreille Subbasin Summary.  It states: "The sharp decline in the kokanee population will result in lost forage for top predators such as bull trout and rainbow trout, and it is anticipated that this will eventually limit predator populations if not reversed." Should this project succeed in balancing kokanee and allowing their population to rebuild, a significant threat to the bull trout population will be reduced.