Accuracy and precision of 4 sampling methods for predaceous natural enemies of aphids in sorghum

Several species of predatory arthropods and hymenopteran parasitoids inhabit grain sorghum, Sorghum bicolor (L.), fields in the U.S. Great Plains and play a role in biological control of key aphid pests of sorghum. Important taxa include predatory Syrphidae, Chrysopidae, Hemerobiidae, Coccinellidae, Anthocoridae, and Araneae and aphelinid and braconid parasitoids. We undertook a 4-yr study to evaluate 4 sampling methods for estimating predator density: D-vac suction sampling, sweepnet sampling, timed counting, and enumeration of predaceous arthropods on 2 sorghum leaves per plant (2-leaf counts) to quantify predator populations and used 2 of methods, 2-leaf counts and timed counting, to quantify relative density of mummies of aphelinid parasitoids. We evaluated the statistical accuracy and precision of density estimates for samples obtained using each method, comparing them to population estimates from complete enumeration of natural enemies on plants, which were considered the reference for unbiased population estimates. All 4 sampling methods provided biased estimates for at least some natural enemies. Timed counts and 2-leaf counts were biased for fewer taxa than were sweepnet and D-vac samples. Statistical precision measured by the coefficient of variation (CV) for each sampling method differed among taxa with no method being the most precise for sampling all taxa. Overall, 2-leaf counts and timed counts were the sampling methods exhibiting the greatest precision and smallest bias for estimating relative population density or population intensity for natural enemies of aphids in sorghum.