In‐field classification of herbicide‐resistant Papaver rhoeas and Stellaria media using an imaging sensor of the maximum quantum efficiency of photosystem II

Reliable in‐season and in‐field tools for rapidly quantifying herbicide efficacy in dicotyledonous weeds are missing. In this study, the maximum quantum efficiency of photosystem II (F_v/F_m) of susceptible and resistant Papaver rhoeas and Stellaria media populations in response to treatments with acetolactate synthase (ALS) inhibitors were examined. Seedlings (4–6 leafs) were transplanted into the field immediately after the application of the ALS inhibitors florasulam, metsulfuron‐methyl and tribenuron‐methyl. The F_v/F_m values were assessed 1–7, 9 and 14 days after treatment (DAT). Based on the F_v/F_m values of all fluorescing pixels in the images of herbicide‐treated plants, discriminant maximum‐likelihood classifiers were created. Based on this classifier, an independent set of images were classified into ‘susceptible’ or ‘resistant’ plants. The classifiers’ accuracy, false‐positive rate and false‐negative rate were calculated. The F_v/F_m values of sensitive P. rhoeas and S. media plants decreased within 3 DAT by 28–43%. The F_v/F_m values of the resistant plants of both species were 20% higher than those of the sensitive plants in all herbicide treatments. The classifier separated sensitive and resistant plants 3 DAT with accuracies of 62–100%. False‐positive and false‐negative classifications decreased with increasing DAT. We conclude that by the assessment of the F_v/F_m value in combination with the classification sensitive and resistant P. rhoeas and S. media populations could be separated 3 DAT. This technique can help to select effective control methods and speed up the monitoring process of susceptible and resistant weeds.