Comparison of Sweep-net and Ground-cloth Sampling Methods for Estimating Arthropod Densities in Different Soybean Cropping Systems

Abstract

Methods involving a sweep net, ground cloth, and ground cloth plus visual observation were evaluated for sampling arthropod pests and predators in three soybean cropping systems. The ground-cloth technique provided a good estimate of absolute density for most of the species sampled. In conventionally planted soybeans with rows spaced 81 cm apart, the ground cloth alone sampled from 82 to 97% of the true arthropod density as measured by the ground cloth plus visual observation method. In drill-planted soybeans and soybeans planted in small grain stubble, these ranges in efficiency were 75 to 96% and 73 to 90%, respectively. The lower percentages in ground-cloth sampling were primarily due to individuals remaining on the soybean foliage after the plants had been shaken. A power model was used to transform sweep-net counts to equivalent ground-cloth counts using regression analyses on mean densities of nine arthropod species encountered in 150 fields. In 11 of the 27 species/cropping system combinations, a linear model adequately described the relationship between the two methods. In those regressions that showed nonlinear relationships, the sweep net was more efficient than the ground cloth at high pest densities. Our equations provide estimates of absolute densities of arthropods in different cropping systems using different sampling methods and can be used to convert action thresholds between systems and methods.