Abstract
A roving-clerk creel survey with nonuniform probability sampling was implemented on three Norwegian rivers with different species of fish to determine if the method was applicable in streams with a low sportfishing intensity. The rivers are not stocked. The estimated total yield varied from 4.9 to 20 kg/hectare year and the fishing effort varied from 69.4 to 231.9 h/hectare year. Coefficients of variation for monthly estimates of yield and effort were used to measure the precision of the sampling design. Precision was acceptable for estimated fishing intensity during the summer peak season (June-August), when fishing intensity was 563-1,548 h/month, corresponding to 17-105 h/hectare, but decreased markedly outside the peak season, when fishing intensities were less than 20 h/hectare month. Precision also was comparatively lower for the estimated total yields. Sample sizes were 3-15 sampling units per month in the peak season and 1-6 sampling units per month outside the peak season. The method is ...

This publication has 0 references indexed in Scilit: