Comparison of sampling procedures for estimating botrytis bunch rot incidence in New Zealand vineyards

Abstract

Quantifying botrytis bunch rot (BBR) in vineyards is essential to the New Zealand wine industry Simple random sampling (SRS) inverse sampling (IS) and adaptive cluster sampling (ACS) were compared using simulated disease incidence at two levels (1 and 3) three clustering factors (random low and high) and small (100 bays) or large (2500 bays) vineyard sizes Sampling relative efficiency (re) was highest using ACS when disease was highly clustered at low incidence Inverse sampling was investigated by repeatedly sampling from the simulated vineyards which resulted in total sample number and variance of IS being greatest in highly clustered BBR at low incidence IS resulted in a lower final sample number with less variance than ACS using simulated BBR that was randomly dispersed at 3 IS or SRS using a sample size calculated from desired sample confidence can provide practical and accurate botrytis bunch rot sampling for disease management decision support