Variability in Aerial Counts of Spawning Salmon
A study of spawning ground surveys for pink salmon (Oncorhynchus gorbuscha) was made in two streams on Kodiak Island. An experimental design is described which permits replication of observers' counts of spawning salmon. The variance in an observer's estimate was found to be proportionate to the size of the estimate. The experiments indicated that an observer will detect differences in population size of plus or minus 50%. The relationship between counts of one observer and another changes within different streams, but within each river the observations of one observer were correlated with those of another. The results of the experiments are summarized in recommendations for aerial surveys of spawning salmon.