A Sequential Sampling Plan for Counts of Adelges tsugae on Individual Eastern Hemlock Trees

Abstract

A sequential sampling plan was developed to estimate density of sistens of the invasive exotic hemlock woolly adelgid, Adelges tsugae (Annand), by counting adelgids on new shoots of four branch tips sampled from the lower half of the live crown of individual, asymptomatic eastern hemlock trees. The y-intercepts and slopes for the relationship between observed mean and variance of A. tsugae counts were similar for North Carolina and for 2 years of data collected from Virginia; thus, data were pooled to create one regression equation that was used to develop a count-based sequential sampling plan. Validation data sets were obtained in West Virginia and by randomly selecting half of the trees sampled in North Carolina and Virginia, trees not used previously to develop the sampling plan. Validation of the combined North Carolina–Virginia sampling plan showed that the model performed as specified. This plan should allow for estimation of adelgid population changes on a tree over time, provided the tree has not declined due to A. tsugae infestation. This plan can also be used to assess chemical or biological treatment effects on A. tsugae, provided such treatments do not fundamentally alter the mean–variance relationship.