9713 - Laserfiche WebLink

Home Browse Search

and Marsh 1996), but is thought to be at a minimum during the fall and winter months (Douglas and Marsh 1996, Valdez and RyeI1995). If HBC emigrate from the LCR or die between sampling events, it is assumed that both marked and . I unmarked fish are lost at the same rate. The Chapman-Petersen estimator can still be used in this circumstance, but the population estimate will be germane for the population during the marking event. Additionally, if HBC immigrate into the LCR between the two events, then the population estimate will be germane for the population during the recapture event. If both additions and losses (i.e., such as immigration and emigration) occur between the events, there is no possible correction and the estimate will overestimate HBC abundance. Finally, all fish captured during both mark-recapture efforts were handled with utmost care to avoid injury or stress related mortality. It was not possible to directly test the second assumption that capture and handling during the first event affected the recapture probability in the second event. However, results of the tests examining violation of the third assumption provided indirect evidence of whether the second assumption was violated. Again, careful handling of the fish throughout the study should have minimized problems of violating this assumption. The third assumption addresses equal capture probability of all fish. This assumption can be violated if the capture gear (i.e., hoop nets) is highly size selective. To determine if the probability of capture varied due to fish size, Kolmogorov-Smirnov tests were applied to the length frequency data collected during both the mark and recapture events. The first test compared the length frequency distributions of marked fish [M] with those captured during the recapture event [C]. The second test compared the length frequency distributions of fish marked during the marking event [M] with those recaptured during the recapture event [R]. Capture probability can also differ by location (i.e., along the LCR river corridor). During marking and recapture events, sampling was equally distributed throughout the entire 13.6 km study area. To validate whether all fish had an equal probability of capture during the marking event regardless of their location, a contingency table analysis was used to test whether the "mark rate" differed among sampling reaches and sub-reaches (Seber 1982). This was performed by dividing the number of recaptured fish [R] by the number of fish captured [C] within each geographic reach, and comparing the results in the contingency table analysis. Similarly, a "recapture rate" can be used to validate whether all fish had an equal probability of capture during the recapture event. This was performed by dividing the number of recaptured fish [R] by the number of fish marked [M] within each geographic reach, and comparing the results in the contingency table analysis. The results the above tests suggested if modifications to the Chapman-Petersen estimator were necessary to minimize bias (Bernard and Hansen 1992). These modifications included stratifying the abundance estimates by length, by geographic reach, or both, if necessary. The fourth assumption (potential tag loss) has proven to be more problematic to address. During the spring trips of 2001, a dorsal fin punch was used as an 16