Laserfiche WebLink
<br />PERFORMANCE OF PIT TAG INTERROGATION SYSTEMS <br /> <br />407 <br /> <br /> <br />Creek, we divided low and high tlow at a depth of 16 <br />cm over the top of the most embedded antenna <br />(equivalent to a stage height of 1.48 m and a flow of <br />0.38 m3fs). This depth corresponded to the maximum <br />read distance for BE-type PIT tags (the weakest tag <br />used in the watershed) at low EMF noise levels and <br />optimum orientation. In Beaver Creek, we reasoned <br />that low flow should be categorized as 22.9 cm or less <br />from the top of the most embedded antenna (equivalent <br />to a stage height of 1.69 m and a flow of 0.57 m3fs), <br />which was based on the water column height that <br />corresponded to the readable range of ST-type PIT tags <br />(the weakest tag used in the watershed) at low EMF <br />noise levels and optimum orientation for all six <br />antennas. While this method to separate low flow and <br />high flow ba'ied on water depth and read distance of the <br />weakest PIT tag type increased a'isurance that equal <br />probability of reading tags was achieved during periods <br />oflow flow, it did not assure it for periods of high flow, <br />and it did not incorporate enhancements for orientation <br />and EMF noise issues that later tags incorporated. <br />To help separate the events in which a fish wa'i likely <br />to have moved pa'it the entire interrogation system from <br />those when a fish did not complete the pa'isage (Figure <br />3), we developed criteria to select events suitable for <br />use in efficiency calculations (Table I). Based on the <br />frequency of time to pass the system. we selected a <br /> <br />I I <br /> <br />Fish 1 Fish 2 FiSh:> Fish 4 Fish 5 Fish .6 <br /> <br />TABLE I.-Rules for delemlining when the detection of a <br />PIT tag qualifies as a fish-detection event and a fish-passage <br />event tor the pUl]Jose of estimating tlIe probability of detecting <br />a fish passing a PIT tag inlelTogation system composed of <br />tlrree arrays with two antennas in each array. <br /> <br />l. Eliminate detections of PIT tags in fish that were captured, <br />tagged, and released within 50 m of the antennllS. <br />2. \\-'hen a lish is detected at only one array. assume that it passed <br />all three aITays bnt was . not detected at the other two. <br />3. If a PIT-tagged fish is detected at more than one alllenna and the <br />time between the lirst and last detections does not exceed 18 min. <br />treat it as a tish-detection event. (The 18-min value correspOllds <br />to the 90th percentile of all potential fish-derection events for the <br />Rattlesnake Creek interrogation site.) <br />4. If the direction of mavement cannot be reasonably detemlined <br />from previous or later derections. do not use the detection event. <br />5. If a fish-detection event meets all of the criteria above, treat it as <br />a fish-passage event <br />6. Do not use a fish-passage event if the same fish is detected on <br />any antenna t2 h before or after this event. <br /> <br />value of 18 min, which corresponded to the 90th <br />percentile value, to be the time frame within which a <br />fish had to pass the interrogation system. To avoid <br />using fish that were swimming back and forth over the <br />antennas, we eliminated a fish passage event if the same <br />fish was detected by any antenna within 12 h previous <br />or subsequent to the first passage event. When we <br />compared this 12-h criteria (protocol 1) with a much <br />more restrictive criteria of I month (protocol 2), <br /> <br />Flow <br /> <br />l <br /> <br />Array A <br /> <br />Array a <br /> <br />Array C <br /> <br />FIGURE 3.-Possible routes and detections of PIT-tagged tish moving across a tlrree-array PIT tag interrogation system. A <br />straight vertical line that crosses a horizontal array line represents a successful fish detection. The movemenb of fish 1-5 would <br />be classified as tish passage events (see Table 1); that of fish 6 would not. The movements illustrated by the circles and ovals <br />would probably not be considered fish passage events. <br />