Investigating time and economic costs of botrytis bunch rot sampling using interpolated data

Abstract

Botrytis cinerea causes botrytis bunch rot (BBR) disease in wine grapes. Small-scale labour-intensive visual disease assessments may not adequately represent an entire vineyard but larger assessments add cost without necessarily improving accuracy or financial returns. BBR-severity data were collected on three dates from two sites and spatially interpolated. Balanced acceptance sampling (BAS) and simple random sampling (SRS) were compared using sample sizes of 2 to 200 vines. Assessment times were calculated for both walking (rows ignored) and driving (rows impassable) and costs compared with assessment error and effects on crop value. Overall, BAS performed better than SRS. Driving was faster than walking except when sample distribution necessitated travelling down every row regardless of sample size. Annual crop losses of up to NZ$2578 per hectare could result from short assessment times and subsequent inaccurate estimates of BBR severity. Spatial interpolation was shown to be a useful and promising technique for studying BBR sampling strategies in vineyard blocks. Travel was not a substantial component of assessment time. An 80-minute-long assessment could substantially reduce economic losses because of errors in BBR assessments.