Abstract
Broccoli crops are produced throughout the year. They are grown under a wide range of environmental conditions, which results in large variation in the time to head initiation and in head growth after initiation. Misscheduling broccoli crops has financial consequences. Models to predict the harvest dates of broccoli crops allow growers to adapt their commercial strategies to maintain an appropriate supply of broccoli throughout the year in a cost-effective manner. The work described here developed models for the phenology and yields of Scottish broccoli crops using data from field trials and commercial crops that allow the prediction of harvest dates from observations or forecasts of daily temperatures and radiation. The models were validated using independent data obtained from commercial broccoli growers and used to explore the local, seasonal and year-to-year variation in the predicted harvest dates of broccoli in Scotland. The date of transplanting seedlings to the field had the largest effect on the time to harvest, although there was also considerable annual variation. Geographical location had least effect on the time to harvest.