Masters Thesis

An investigation into the efficiency of observers to visually detect adult salmon spawning in the Prairie Creek watershed, Humboldt County, California

I estimated the observer efficiency of spawning surveyors conducting foot surveys for live coho and Chinook salmon in Prairie Creek, Humboldt County, California, during the 2007-2008 spawning season, and in Prairie Creek and its main tributary, Lost Man Creek, during the 2008-2009 spawning season. A dual-mark recapture design was utilized whereby adult salmon caught at weirs below the spawning grounds were tagged with both passive integrated transponder (PIT) tags and externally-visible Peterson disc tags. An array of in-stream PIT interrogation antennae was used to monitor movement of tagged individuals into and out of survey reaches. Individuals that died within survey reaches, and whose exact date of mortality was unknown, were removed from the pool of observable tagged individuals through a process that involved estimating a range of probable dates of mortality from confidence intervals constructed around survey life estimates; the process produced two counts of tagged fish available to be observed – one based on removing individuals at the beginning of the estimated range of mortality, and one based on removing them at the end of the range. Survey life of tagged individuals was estimated as a linear function of date of arrival at the weir, condition upon arriving at the weir, and sex. Surveyors counted the numbers of tagged and untagged fish they observed on each spawning survey. Observer efficiency was calculated in two ways: observer efficiency was calculated as the ratio of the number of tagged fish observed on a given survey to the number of tagged individuals estimated to be present in the survey reach on the survey date from the antenna data ('comparison' method); the second method involved alternately substituting the weir-based mark-recapture escapement estimate, the lower- and upper-bounds of its confidence interval, and the lower- and upper bounds of the confidence interval of the survey life estimate into the area-under-the-curve escapement estimation formula and solving for the observer efficiency term ('solution' method). Due to a mid-season funding freeze and resulting lack of data, observer efficiency calculations were not possible for the 2008-2009 spawning season. Observer efficiency estimates obtained with the comparison method from the surveys conducted during the 2007-2008 spawning season ranged from 0 to 1, with a median value of 0.167 based on the low count of tagged individuals available to be observed, and 0.114 based on the high count. The solution method observer efficiency estimates ranged from 0.130, as calculated from the upper bounds of the escapement estimate and survey life confidence intervals, to 0.491 as calculated with the lower bounds of the escapement estimate and survey life confidence intervals.

Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.