9710 - Laserfiche WebLink

Home Browse Search

PERFORMANCE OF PIT TAG INTERROGATION SYSTEMS 403 been performed to determine the efficiency of interrogation systems for detecting PIT-tagged fish. When studies have been done, they have generally used fish that have a high propensity to move downstream or upstream in relative unison during some part of their life history. Efficiency has been calculated for stationary interrogation systems using downstream traps to confirnl that fish have passed an interrogator, or by having a known number of fish tagged upstream of a detector and then assuming all emigrate past the intelTOgator. There have been substantial efforts to document efficiency using experimental channels and dam facilities. Nunnallee et al. (1998) evaluated efficiency of a PIT tag interrogation system in a fish collection channel using a direct method whereby a known number of fish passed the detector, and by an indirect method whereby detections at other antennas were compared with detections at the system being evalu- ated. They calculated efficiencies for detecting PIT- tagged salmonids to be 97% using the direct method and 99% using the indirect method. Using similar direct and indirect methods. Axel et al. (2005) found detection efficiencies of a tour-antenna system around a large bypass pipe (91.4 cm diameter) to be close to 100% for tagged salmonids. In an experimental fishway study, Castro-Santos et aI. (1996) used four arrays of one antenna each and found the detection efficiencies for three c1upeid species known to have passed their detector system to be 96% in a Denil-type fishway and 88% in a Steeppass-type fishway. In a study of juvenile Atlantic salmon Salmo salar in an artificial channel off the River Itchen in the United Kingdom, Riley et al. (2002) found detection efficiency of downstream-moving fish to be 70%, but detectors at each of three exit points (two exit points had two antennas, one exit had one antenna) were combined to determine an overall efficiency rate. Though upstream movement was detected, they were not able to calculate efficiency for upstream-moving fish. Using captures of fish at a trap downstream of a detection site (two 4-m X 1.2-m, side-by-side upright antenna~), Zydlewski et aI. (2001) found downstream detection efficiency to be 93% for juvenile Atlantic salmon. For cases when a known number of fish have passed an interrogator, Zydlewski et al. (2006) described a method for calculating the detection efficiency, but not the variance. In general, these findings indicate that stationary interrogation systems have potential to be highly effective in modified chlmnel systems, but alternate methodologies for estimating detection effi- ciency and its variability have been lacking. In most of the previous studies referred to above, detector systems were placed where flow was restricted by pipes or fish ways or at stream pinch points such as bridges or culverts. In some cases, researchers have modified the stream channel to force fish through or near antennas (Greenberg and Giller 2000; Riley et aI. 2003; Zydlewski et aI. 2006). While it may be possible to direct all water and fish at specific sites, we saw the need to develop an intelTOgation system that could be adapted to free-flowing streams in remote locations without reliance on existing structures (e.g., culverts and bridges) or modifying the channel. Despite attempts to direct fish past instream PIT tag antennas, tag detection efficiency is likely to be less than 100% for a number of reasons. Fish behavior can change with changes in stream conditions, and alternate passage routes can provide fish opportunities to pass beyond a detection field. The electrical properties of a PIT tag interrogation system can change with changes in water level, which may partially or completely expose an antenna to air, and with changes in water temperature, conductivity, and air tempemture. These changes can compromise a system's ability and consistency to detect tags. However, this latter problem can be partially or completely solved by using transceivers that automatically change their settings (self tune) to changing environmental conditions, thus improving petformance. A system's ability to read tags can also be compromised by ambient electromagnetic fields (EMFs) of similar frequency, which can be generated by nearby power lines, electric fences, pumps, or electrical devices in homes or businesses (Zydlewski et a1. 2006). This intetference can be steady or changing depending upon the noise source (Horton et aI. 2(07). Because the present systems cannot read two tags at once, multiple fish swimming through or holding in the detection field at the same time can compromise the ability to detect a tag (Greenberg and Giller 2000). Because of these and possibly other factors, investigators may need to determine detection efficiencies during discrete periods of differing condi- tions (Horton et al. 2007). The objectives of our study were to (1) describe a protocol for identifying active juvenile and adult salmonid migrants, (2) estimate the magnitude and variance of detection efficiency, (3) evaluate the effect of the direction of fish movement and stream flow on detection efficiency, and (4) explore the effect of antenna configuration on detection efficiency. We describe the tag-reading efficiencies, with estimates of variability, achieved by two similar PIT tag interroga- tion systems designed to (1) maximize detection of tagged fish, (2) distinguish between downstream and upstream movements, (3) be readily adaptable to remote stream sites, and (4) not be dependent on full-stream coverage. We describe an indirect method for deriving