Abiotic and biotic variables affect aphid natural enemy population estimates on sorghum

Aphidophagous arthropods inhabit grain sorghum, Sorghum bicolor (L.), fields in the US Great Plains and provide naturally occurring biological control of aphids. We evaluated 4 sampling methods for estimating population densities of aphidophagous arthropods on sorghum: suction sampling, sweepnet sampling, counting natural enemies for a fixed time, and counting natural enemies on 2 sorghum leaves per plant. We compared population estimates obtained by these methods to estimates of absolute population achieved by complete enumeration of natural enemies on sorghum plants. Sorghum plant growth stage, aphid population intensity, and some meteorological variables were correlated with relative population and population intensity estimates for some natural enemies. Aphid population intensity and sorghum plant growth stage were frequently correlated with natural enemy relative density and population intensity. Significant partial correlations tended to be specific to particular natural enemy categories and sampling methods. Regression models were constructed to relate absolute population density (D) of natural enemies to relative population density or population intensity (R). Regression models for coccinellid adults and large larvae had relatively large R2 value, ranging from 0.66 to 0.80 for all sampling methods. Regression models for Aphelinus nigritus Howard from 2-leaf counts and timed counts had a small R2 value of 0.16 and 0.33. Regression models for lacewing larvae were not significant. Relative density (R) was the only variable that entered significantly into regression models for absolute population density (D) for most natural enemy categories. Results indicate that useful estimates of an absolute population of natural enemies in sorghum can be obtained using relative sampling methods.