Steve Anglea and Rich Townsend describe the use of passive integrated transponder tags to determine juvenile and adult salmonid survival and passage characteristics in the Columbia river basin


Figure 1

Determining juvenile and adult survival and passage characteristics of anadromous and resident salmonids is a major component of efforts directed towards recovering threatened and/or endangered fish populations in the Pacific Northwest region of the US. Historically, researchers identified groups of fish using marks and tags such as freeze-brands, fin clips, and coded wire tags, alone or in combination. Though useful, estimates of survival made using these techniques were typically restricted to a group of animals on an annual basis. Advancements, particularly in micro-processors and battery miniaturisation, have provided today’s researchers with tools that allow monitoring of individuals and determination of survival with greater resolution.

One of these technologies is the Passive Integrated Transponder or PIT tag. The US National Marine Fisheries Service (NMFS) initiated a study in 1983 to evaluate the biological and technical feasibility of implanting PIT tags in salmonids. Prentice et al. (1990) determined that marking fish with PIT tags was both biologically and technically feasible. The use of PIT tags to monitor fish populations has grown so much since the 1980s that hundreds of thousands of juvenile anadromous salmonids are marked each year in the Columbia river basin. The regional desire to monitor fish passage through larger orifices or passage routes at dams has provided the motivation and funding for the latest technology advancements. The US Army Corps of Engineers (USACE) and Bonneville Power Administration (BPA) are working together to design, construct, operate and maintain a PIT tag detector system at the Bonneville dam second power house corner collector. The current configuration comprises an antenna with an opening measuring 5m x 5m.

Passive integrated transponder

A passive integrated transponder (PIT) tag is a radio frequency identification (RFID) device. The tag consists of a small copper coil and an integrated microchip encased in a glass capsule (Figure 1). Each tag is programmed with one of approximately 34 x 109 tag codes. The most common PIT tag used in fisheries application is approximately 2mm x 12.5mm. The RFID system uses a signal transmitted between the microchip and a reading device, such as a ‘scanner’, ‘reader’ or ‘transceiver’. The RFID or electronic identification devices most widely used in fish are passive. Passive integrated transponders have no battery so the microchip remains inactive until passing through an electromagnetic field (EMF) emanating from an antenna or handheld scanner. A current is induced in the tag’s coil when it passes through the EMF providing the power needed to send its unique code back to the transceiver and positively identify the fish. Passive tags are designed to last the life of the fish providing a reliable, long term identification method. PIT tags are typically injected into the body cavity using a 12-gauge hypodermic needle and syringe (Figure 2). Implant location varies depending on the species being studied, animal size and in some cases the behaviour of the fish.

Detection of pit at columbia river dams

PIT tag detection systems have been established at several dams in the Columbia river basin (Figure 3). Detection systems have been incorporated into juvenile fish bypass systems at all dams in the lower Snake and Columbia rivers. PIT tag detector systems have also been established in the adult fishways at Bonneville, McNary, Ice Harbor, and Lower Granite dams in the lower Snake and Columbia rivers and at all dams up to Chief Joseph dam in the mid-Columbia river. PIT-tagged fish are automatically detected when passing through the systems, and record of their detection is stored in the Pacific States Marine Fisheries Commission’s PIT Tag Information System (PTAGIS). PTAGIS is a web-based data collection, distribution, and coordination project that allows researchers to remotely and in near-real time monitor detection of their study fish.

A case study – public utility district no. 2 of grant county

Juvenile survival

The Public Utility District No. 2 of Grant County (Grant PUD) owns, operates, and manages the Priest Rapids project in Washington State. The project includes the Wanapum and Priest Rapids developments on the Columbia river. As part of its operating licence with the Federal Energy Regulatory Commission (FERC), Biological Opinion for the Priest Rapids project, Hanford Reach fall Chinook Protection Program, and Anadromous Fish Settlement, Grant PUD has initiated several fish protection measures including: flow management; installation of an advanced hydro turbine system at Wanapum dam; construction of a future unit bypass at Wanapum dam; exploring alternative bypass options at Priest Rapids dam; development of a habitat protection; and development of hatchery production programmes.

Grant PUD used PIT tags to determine the survival rate of juvenile Chinook salmon (Oncorhynchus tshawytscha) migrating through the project in 2001, 2003, 2004, and 2005. Action 1 of NOAA Fisheries Biological Opinion on the operation of the Priest Rapids hydroelectric projects directs Grant PUD to make progress towards achieving a minimum 91% combined adult and juvenile salmonid survival at the Priest Rapids and Wanapum developments. This performance standard includes a 93% juvenile performance standard for each development (reservoir and dam). Grant PUD contracted with Biomark and the University of Washington to determine survival of PIT-tagged juvenile Chinook salmon. Summarised in Tables 1 and 2 are the results from survival studies conducted in 2003, 2004, and 2005. Biomark was responsible for the marking, rearing, and release of the study fish, and the University of Washington was responsible for survival analyses.

Biomark annually marked and released between 88,200 and 114,601 juvenile hatchery Chinook salmon for the study. Fish were marked in the Biomark tagging trailer and then transferred to one of seven holding raceways at the Priest Rapids Hatchery using PVC pipes (Figure 4). Fish were marked from one (2003) to five (2004 and 2005) months prior to the release period allowing sufficient time for healing of the tag wound.

The release period was initiated in late-April. Fish were released following a paired-release model (Burnham et al. 1987) to determine either development or project survival. In 2003 and 2004, fish were released into the tailrace of Rock Island, Wanapum, and Priest Rapid dams. Due to limited fish availability in 2005, fish were only released into the tailraces of Rock Island and Priest Rapid dams. Depending on the year, between 3200 and 3750 PIT-tagged fish were released each day. Annual release periods were approximately 30 days in duration and were partitioned into ten (2004) or twelve (2003 and 2005) replicate blocks.

A pre-determined number of fish were loaded into release containers via a pescalator for each daily release (Figure 5). Fish were scanned and enumerated automatically as they were loaded into the release containers using a series of custom antennas. The data files generated during the loading process were uploaded to the PTAGIS database after proofing and served as the basis for survival analyses. Release containers were transported from Priest Rapids hatchery to a staging area at Wanapum dam. The following day, release containers were flown one at a time to the specified release location via helicopter (Figure 6). Study fish had the potential of being detected in the juvenile bypass systems at McNary, John Day, and Bonneville dams. Study fish detected at Bonneville dam and the NOAA Fisheries experimental PIT-tag trawl were pooled for all of the analyses. The PTAGIS database was queried to assemble detection histories at each downstream observation site for each fish. Detection and survival probabilities were determined for each release group and combined to determine development and project survivals.

Project and development survival estimates were similar among years (Table 1). Estimates of survival through the Priest Rapids development met the juvenile project survival performance standard (93%) identified in the Biological Opinion. Survival estimates through the Wanapum development were slightly lower than the performance standard. In 2003 and 2004, project survival estimates were similar to the product of the two development survival estimates and also to that estimated in 2005 when two released locations were used. The similarity among project survival estimates suggests that development survival in 2005 was likely similar to that estimated in 2003 and 2004. Grant PUD has continued its effort to monitor dam and project survival using a variety of technologies including PIT, radio, acoustic, and HI-Z balloon tags.

Adult detection

Biomark installed PIT tag detector systems for Grant PUD in both fishways at Priest Rapids dam in spring 2003. Establishment of a PIT-tag detection system provides fisheries managers with an additional site to monitor survival and passage characteristics of salmon and steelhead and any other aquatic organisms marked with PIT tags (i.e. lamprey and bull trout) in the Columbia river basin. Two weirs, each having two orifices, in the non-overflow upper section of each fishway were equipped with antennas (Figure 7). Antenna openings are approximately 0.6m x 1.2m. Attaching the antennas to the upstream face of the weirs eliminated the need to modify the concrete weir and facilitated installation. Each antenna is controlled by a unique reader or transceiver, and all transceivers are connected, via fibre optic cable, to a single data collection computer. All diagnostic and fish detection data is automatically uploaded to the PTAGIS database.

PIT tag detections at Priest Rapids dam in 2003 to 2005 were used to determine run characteristics (i.e. composition, timing, and migration rates) and detection efficiency of the PIT tag detector systems. Detections of PIT tagged fish were obtained by querying the PTAGIS database. Estimation of detection efficiency for the Priest Rapids dam adult fishway followed the methods presented by Townsend et al. (2002) for the two-detection weir PIT tag arrays at Wells dam, Columbia river. In general, this entails determining the detection history for each fish at each weir and assigning a ‘0’ if the fish was not detected at a weir and a ‘1’ if it was detected. Basic assumptions are that fish are travelling in an upstream direction and fish detected on the downstream weir continue travelling upstream. Detections of the same fish occurring greater than twelve hours apart were treated as unique passage events for the purpose of determining detection efficiency.

In 2005, the first detections of fish occurred in April and continued until the end of November (Figure 8). Spring Chinook salmon were the first group detected followed by summer/fall Chinook salmon and then steelhead in August and September. The majority of detections were of Chinook salmon followed by steelhead and then coho salmon (Table 2). The difference in the number of Chinook salmon and steelhead detected is more an indicator of research focus rather than survival to adulthood of PIT tagged fish. There is a myriad of fish releases made by private, state, federal, and tribal groups each with their own research programmes.

Migration rates between adult observation sites in the lower and mid-Columbia river were determined for Grant PUD study fish released in 2003 and 2004. Some fish from these releases are now returning as adults after spending one or two years in the Pacific Ocean. Migration rates were similar from Bonneville dam to McNary dam and from McNary dam to Priest Rapids dam for the two release years (Table 3). Migration rates for both groups decreased between McNary and Priest Rapids dams as the fish migrated through the free-flowing Hanford Reach of the Columbia river.

Detections of all species were combined to generate annual estimates of detection efficiency for each fishway. Detection efficiency for all years was consistently higher at the west bank fishway (0.99) compared to the east bank fishway (0.93-0.96). The difference in detection between fishways is likely the result of slightly larger antenna openings in the east bank fishway and also higher background levels of electrical interference in the vicinity of the east bank fishway. The increase in detection efficiency at the east-bank fishway over the past three years is likely the result of improved tag performance of returning fish. In spring 2003, a better performing tag (i.e. greater detection distance) was introduced by Digital Angel. The percentage of returning adult salmonids implanted with the better performing tag has increased over the past couple of years, and the benefit is being realised with higher detection and improved efficiency of fishway detector systems.

Past, present, and future

PIT tags have become more prevalent as the technology advances and new biological questions are being asked. Applications have broadened from a tool used to evaluate juvenile salmonid survival through mainstem dams and reservoirs to monitoring movement of resident fishes in small tributary streams. As mentioned earlier, USACE and BPA are working together to monitor PIT tags passing through the Bonneville dam corner collector. This effort has resulted in the development of better performing PIT tags and readers capable of being detected using a 17.5m x 17.5m antenna. Though developed to monitor dam passage, the new PIT tags and readers are being utilised by researchers conducting work on much smaller scales as well. The direction of future technological advancements will continue to be driven by the research community, whether it is to evaluate a new fish collection facility, a new turbine design, or a habitat restoration effort on your local stream.

Author Info:

Steve Anglea is a fisheries biologist with Biomark Inc and Rich Townsend is a Biometrician with the University of Washington. For further information, email:

Related Articles
Tribe signs Priest Rapids agreement

Table 1
Table 2
Table 3
Table 4