作者: Paul E. Pulley , Robert N. Coulson , John L. Foltz , William C. Martin , Claude L. Kelly
DOI: 10.1093/EE/6.5.607
关键词: Statistics 、 Stage (hydrology) 、 Standard deviation 、 Sampling (statistics) 、 Probability density function 、 Stratified sampling 、 Biology 、 Dendroctonus frontalis 、 Variable (computer science) 、 Sample (statistics)
摘要: The practice of quantitatively sampling within-tree populations Dendroctonus frontalis was investigated by comparing the relationships between intensity, informational content samples, and precision in estimation. Random stratified random plans were compared contrasted to selective or purposeful using tree geometry model-probability density function (TG-PDF) estimation procedure. techniques required large quantities data therefore less suitable than TG-PDF latter procedure incorporates knowledge vertical distribution D. life stages much for a given level precision. The evaluated on basis obtained technique different numbers sample levels position levels. Precision as judged refined standard deviation estimate σR increased number sampled increased. For levels, σR, which associates covariables with always produced better basic σB assumed equal information from each level. Positioning along infested bole influenced obtained. Dispersing over while avoiding extremes generally resulted stage. best where C individual similar. The optimal location varied stage being sampled. Generally 2 3 provided an adequate most requirements. emerging adult variable more difficult costly sample.