Abstract

AbstractA growing challenge in canola (Brassica napus L.) production globally is the management of aphid pests, particularly species that are resistant to insecticides. Aphid pests of canola damage plants through direct feeding and virus transmission, with turnip yellows virus being particularly economically damaging. Integrated Pest Management, a strategy now employed by many growers to reduce the risk of insecticide resistance, requires forward planning and monitoring. Improved risk predictions can be used to help growers limit insecticide spraying by targeting high‐risk regions and/or periods. Within Australia, autumnal aphid flights coincide with the critical risk period for virus infestations in canola. In this study, we used an extensive database accumulated from 6 years of surveys collected from more than 200 canola fields across southern Australia with supervised machine learning models to predict aphid movements in autumn‐early winter as a function of environmental factors. We found: (i) our models achieve very high predictive accuracy when validated on untrained data; (ii) aphid movements are influenced by a combination of daily temperature and wind regimes as well as ‘green bridge’ effects mediated by summer rainfall patterns; and (iii) higher aphid capture rates in sticky traps are correlated with a higher probability of the aphids being carriers of turnip yellows virus. Taken together these results suggest that growers can use the outputs from predictive models to forecast aphid outbreaks in the early growing season and derive useful rules of thumb around the environmental conditions during which canola crops are at a greater risk of turnip yellows virus transmission.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call