Background: Centaurea diluta Aiton (North African knapweed) is a major weed concern in Spain as a result of the limited herbicides capable of controlling it, and the limited knowledge of its biology hinders the development of integrated weed management strategies.
Results: The current study presents results from two experiments that aimed to: (i) determine the effect of seed burial on seedling emergence; and (ii) model its phenology progression using sigmoidal (SRM) and artificial neural network models (ANN) based on different cohort emergence times. In the first experiment, burial at 2 cm and 5 cm decreased C. diluta emergence by 54% and 90%, respectively, compared to the emergence at 0 cm. In the second experiment, without crop-weed competition conditions, the emergence delay led to reductions in leaf number, rosette diameter, plant height and dry biomass by 63%, 50%, 59% and 93%, respectively. Seed production per plant exceeded 21 469. According to the growth model, leaf number was the most consistent morphological trait and critical for timing weed control actions, so it was used to compare SRMs and ANNs. On average, ANNs increased the precision in 5.72% (± 2.4 leaves) compared to SRMs. This slight performance of ANNs may be valuable for controlling C. diluta because control methods must be applied at the 4-leaf stage to achieve good efficacy.
Conclusion: Seed burial at 5 cm depth is an effective method reducing C. diluta emergence. ANNs accurately predicted the leaf number employing environmental variables can help increase the efficiency of C. diluta control actions and reduce the risk of escapes. © 2023 The Authors. Pest Management Science published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.
Keywords: North African knapweed; artificial neural networks; growth modelling; non-linear regression; seedling emergence.
© 2023 The Authors. Pest Management Science published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.