VIEW ARTICLE

Ecology and Epidemiology

Efficiency of Multistage Sampling for Estimation of Intensity of Leaf Spot Diseases of Alfalfa in Field Experiments. James A. Duthie, Graduate research assistant, Department of Plant Pathology, North Carolina State University, Raleigh 27695, Current address: Department of Plant Pathology, University of California, Davis 95616; C. Lee Campbell(2), and Larry A. Nelson(3). (2)Professor, Department of Plant Pathology, North Carolina State University, Raleigh 27695; (3)Professor emeritus, Department of Statistics, North Carolina State University, Raleigh 27695. Phytopathology 81:959-964. Accepted for publication 5 March 1991. Copyright 1991 The American Phytopathological Society. DOI: 10.1094/Phyto-81-959.

In each of two field experiments, efficiency of multistage sampling was evaluated for estimation of intensity of leaf spot diseases during 10 periods of alfalfa growth over 2 yr. Disease was assessed visually on sections of the alfalfa canopy with the aid of descriptive rating scales. In experiment A, severity of disease was assessed every 7 days on four canopy sections (0.5 × 0.5 m) per replicate plot (3 × 3 m) with 12 plots per treatment. In experiment B, incidence of diseased leaves and severity of disease on diseased leaves were assessed every 2–4 days on four canopy sections (1.0 × 0.17 m) per subplot (1.0 × 0.68 m) with two subplots per replicate plot (10.0 × 4.1 m) and five plots per treatment. In each experiment, efficiency of sampling was evaluated based on estimates of costs of sampling per unit, measured in units of time, and on estimates of components of error variance. Efficiency of sampling could not be improved by changing the rate of sampling during periods of alfalfa growth because none of the variance components changed consistently over time. When the total time allocated to sampling was constrained to C min per treatment, severity of disease was estimated most efficiently in experiment A by sampling four canopy sections in each of 0.417C plots. In experiment B, incidence of diseased leaves was estimated most efficiently with two canopy sections, two subplots, and 0.323C plots, and severity of disease on diseased leaves was estimated most efficiently with three canopy sections, two subplots, and 0.270C plots.