Fish Tracking Technologies used by Grant County PUD

As required by its Federal Energy Regulatory Commission (FERC) license and other agreements (such as the 2006 Priest Rapids Project Salmon and Steelhead Settlement Agreement and the Washington Department of Ecology's 401 certification for the Priest Rapids Project), each year Grant County PUD must achieve a 93 percent survival rate for salmonid smolts out-migrating through each of the two reservoirs and dams. The method used to establish if the standard has been met is to conduct three consecutive years of studies for each salmonid species and take the average of those three years' study results.

To meet this standard, Grant County PUD continually strives to reduce smolt mortality caused by predation in the reservoirs and injuries when passing the dams. Until a 93 percent project survival rate is achieved through one dam and one reservoir in this project, Grant County PUD will continue to evaluate the smolt survival rate and behavioral patterns in the reservoir and through three passage routes at each dam: spillway, turbine, and fish bypass facilities.

Grant County PUD uses fish tagging technologies to study the movement of juvenile and adult salmon, including migration timing and survival through the dams and reservoirs. Tagging data are used to examine the effects of dam passage routes on fish survival, migration rate, and behavior. This data also is used to evaluate and design future fish bypass structures, including the future unit fish bypass at Wanapum Dam, which was completed in 2008.¹

Investigations typically involve collecting, tagging, releasing, and recapturing/detecting fish, then analyzing the data to estimate vital statistics. In the Columbia River Basin, the three primary tag types used to collect fish survival and passage data are passive integrated transponder (PIT), radio, and acoustic tags. Each of these tags have advantages and limitations.

PIT tags

PIT tags are glass-encapsulated, implantable radio-frequency identification (RFID) devices that contain integrated circuited chips. They are passive because they do not contain an internal energy source, such as a battery. Consequently, the tag remains functional for the life of the fish and beyond. Each PIT tag contains an individual alphanumeric code, similar to a scanning bar code in a grocery store. This code means there is no duplication of unique codes between tags.

A single PIT tag pre-loaded into a hypodermic syringe is injected into an open space in the chest cavity of the anesthetized fish. The PIT tag is inactive until it is energized by the electromagnetic field generated by low-frequency radio waves emitted by an antenna connected to a transceiver. A PIT-tag system consists of the tag, antenna, and transceiver.

Because of their relatively small size (12.5 millimeters long by 2 millimeters in diameter, 0.1 gram of weight in air), PIT tags can be implanted into salmonid smolts as small as 60 millimeters in fork length (length from the tip of the nose to the middle, or fork, of the tail). In addition, the antenna used to read the tag's modulation also is relatively small (3 feet by 3 feet), and thus the tag detection range is relatively short (measured in feet). This limitation imposed by the antenna size (although there is a 17-foot by 17-foot antenna being used at Bonneville Dam) and the resulting short detection range prevents the use of PIT tags to track fish passing through the spillway or turbines of a large dam and thus from collecting data on route-specific movement and survival through these passage routes.

The tag detection efficiency at a hydroelectric project on the Columbia River can vary greatly, again due to the small antenna size and limited detection locations. Grant County PUD estimates that the detection of PIT-tagged salmonid smolts at the 980-MW McNary project (the first juvenile PIT detection site downstream from the Priest Rapids Project) is about 20 percent. To maintain statistical adequacy with this low detection rate, the number of fish used in a project survival tagging study for the Priest Rapids project must approach 200,000 smolts.

One of the primary advantages of a PIT tag is that there is no battery to go dead. Thus, the tag is still violable in the salmonid when it returns from the ocean, migrating upstream to its spawning grounds. As the adult passes through the fish ladders at each dam, the PIT tag is read and the data is correlated with information collected when the fish was a juvenile, years before. It is only through this unique technology that smolt-to-adult return rates are calculated.

Additionally, PIT-tag recoveries on bird colonies provide estimates of mortality by avian predators.

PIT tags are cost-effective compared to other types of tags used for evaluating project smolt survival rates. Cost for an individual PIT tag is about $3, depending on the number of tags purchased.

Radio tags

Radio tags transmit a radio-frequency signal detectible through a system of one or more antennas and receivers. A radio telemetry system consists of tags, antennas, and receivers. The tags usually are surgically implanted into a fish and function much like conventional radio transmitters, only at different frequencies. Similarly, the antennas and receivers function much like conventional antennas and radios used to detect standard broadcast signals.

Radio tags work well to evaluate both juvenile and adult salmonid passage at Columbia River dams, particularly in assessing fish behavior in the near-dam environment. Radio tags are useful in evaluating project, dam, reservoir, and route-specific survival; passage efficiencies; forebay survival and delay; tailrace egress; travel times; avian predation; adults who return to a different stream (not their natal stream) to spawn; spawning distributions and timing; and adult fallbacks at dams.

One of the most significant advantages of radio tag technology in a fish tagging study is the ability to detect the tag from the air as well as in turbulent hydraulic environments. Radio-tagged fish can be tracked from a vehicle, boat, or airplane, which allows for efficient surveys of remote or large study areas, as well as the detection of tagged smolts consumed by predatory birds.

Radio tag receiving equipment can vary, but typically is either a sequential scanner programmed to scan for a set period of time and then move to the next frequency of interest, or a device that scans all frequencies simultaneously. The receiving equipment works similar to a "police scanner," seeking a signal emitted by a tag. Radio tag receiver systems make use of multi-element Yagi antennas or tuned loops when stationed above the water surface, and stripped coaxial, dipole, or quad-pole antennas when used underwater. Receivers above the water surface can detect a radio tag only to depths of about 30 feet. Underwater antennas are used to detect tagged fish that may be passing at depths greater than 30 feet.

Radio tag detection probabilities within the various passage routes at a Columbia River hydro project are 95 to 100 percent with the proper detection antenna arrays. This high detection rate allows for a much smaller number of fish needed to conduct a tagging study. For example, performing a survival study of the Priest Rapids Project using PIT tags and the McNary detection site would require about 180,000 tagged fish. Using radio tags and Grant County PUD's detection system, the sample size for the study would be about 3,700 tagged fish.

All radio tags used in the Columbia River Basin require an external trailing antenna, which may affect the swimming performance of juvenile fish or attract predators.

The pulse rates of the tags is relatively high, ranging from 1 to 2 seconds. This high pulse rate is needed to capture/detect a tagged fish as it moves through a passage route at the dam at a higher flow rate, such as going through a spillway gate. As a result, the battery life of the tag is relatively short (9 to 18 days). The pulse rate can be slower, e.g., once every ten seconds. This would significantly increase the tag life, making it more suitable for system-wide applications. However, slowing the pulse rate limits the tag's effectiveness in determining route-specific passage at the dam.

Radio tags used in the Columbia River Basin also have a limited code set compared to those available with PIT and acoustic tags. For example, Lotek manufactures most of the radio tags used in the Columbia River Basin. The Lotek transmitters are on one of 25 frequencies ranging from either 149.32 to 149.80 mega-Hertz (MHz) or 150.32 to 150.80 MHz. For each frequency, Lotek features 521 unique codes, for a total of 12,500 unique transmitters (code and frequency combinations). If more than 12,500 radio tags are used in the system, the individual code/channel combinations can be repeated and separated by time for a migrating salmonid. For example, using a specific tag code in the first week of a six-week study and then reusing this code in the sixth week allows a time separation for this code.

Radio tags can be used to study all species of adult salmonids, adult Pacific lamprey, and juvenile salmonids as small as 90 millimeters fork length. Cost is about $250 per tag, with the price depending on the number of tags purchased.

Acoustic tags

Acoustic telemetry systems use sound waves to relay information from a transmitter through the water to a hydrophone, and then ultimately to a data logger or receiver. Acoustic telemetry systems offer varying tag sizes, transmission life, frequencies, encoding schemes, and receiver capabilities. Fish movements, migration rates, duration of residency, and survival rates can be estimated using acoustic tags. This includes survival estimates of both reach- and route-specific movement through dam bypasses, spillways, and turbines, as well as migration rates.

The acoustic tag receiver array allows for precise location (x, y, and z axis) of tagged fish in three dimensions (e.g., near dams), and detection capability generally is not affected by fish depth.

Three types of acoustic telemetry systems are used with Columbia River salmonid smolts: HTI, Juvenile Salmon Acoustic Tracking System (JSATS), and VEMCO. As with radio tags, acoustic tags have limitations due to handling and surgical procedures for tag insertion and size and life of the tag. Surgical procedures are required to insert tags into the body cavity of the fish. However, because there is no external antenna on the acoustic tag transmitter, surgical implantation is less invasive than with PIT tags. This operation and the tag's weight may affect the fish's swimming and feeding behavior, growth, and susceptibility to infection and/or predation. Acoustic tags are implanted only in smolts larger than 85 millimeters fork length (depending on the size of tag being used).

The size (volume and weight) of the acoustic tag is a function of power need (battery size). Tags for juvenile studies range from 0.55 to 2.2 grams in weight. Transmitting life is directly correlated to available power, as well as the transmission ("ping") rate of the tag. The higher the "ping" rate of the tag, the shorter the life of the transmitter. Battery life in juvenile salmonid transmitters generally is 20 to 90 days, depending on the study objectives (i.e., tag ping rate needed).

Detection of the acoustic-tagged fish is accomplished using hydrophones placed in the dam forebay. With the placement of the hydrophones both high and low in the water column, the acoustic tag is able to provide accurate location of the fish, not unlike using a global positioning system (GPS) unit with four satellites.

With a hydrophone array properly deployed, detection ranges for acoustic-tagged fish can exceed 500 yards and have an efficiency of nearly 100 percent. While acoustic tags cost about $250 each (depending on the number of tags ordered), the nearly 100 percent detection rate means a much smaller sample size (number) of study fish. The detection capability and/or efficiency of the acoustic tag is reduced in high-velocity, turbulent environments, such as the tailwater of a spillway. The number of unique tag codes associated with acoustic tags generally exceeds 50,000. Each type of acoustic tag can only be decoded by its respective receiver, and the types are not compatible across detection platforms.

Grant County PUD's experience

Grant County PUD has used PIT tags over the past decade to evaluate project survival for out-migrating salmonid smolts at the Priest Rapids Project. However, the utility's dams lack PIT detection sites necessary for estimating dam passage routes smolt usage and survival rates. Thus, Grant County PUD turned to radio tag systems, which work well for evaluation of both juvenile and adult salmonid passage at the hydropower projects, particularly for assessing fish behavior in the near-dam environment, resulting in improvements to dam structures and operations.

In 2004, Grant County PUD switched to acoustic tag systems. These tags have the ability to track the vertical and horizontal distribution of tagged juveniles, giving accurate tracks and location of the tagged smolts as they encounter the dams. This data will allow the utility to design and construct more efficient and safer fish bypass structures.

However, no single technology is useful in all situations. The most effective strategy may be to use several tag technologies in combination to address all management needs.

Curt Dotson, senior biologist with Grant County Public Utility District (PUD), is project manager for all of Grant County PUD's main-stem Columbia River salmonid survival studies.

This article was reprinted with permission from Hydro Review as part of the PennWell Corporation Renewable Energy World Network and may not be reproduced without express written permission from the publisher.