From field experiments to modelling of herbicide resistance evolution

The continuous spread of herbicide resistance urges the implementation of Integrated Weed Management (IWM) to establish more sustainable cropping systems. An important question is whether application of IWM strategies can provide a suffi cient level of weed control, and in addition, favour a reversion of the resistance evolution process. The PURE project tackled this issue through a combination of experiments and modelling. Two case studies were considered: Echinochloa crus-galli resistant to ALS inhibitors in maize and Alopecurus myosuroides resistant to ACCase inhibitors in wheat. Field and greenhouse experiments showed that the resistant (R) biotypes of the two species had diff erent fi tness costs. In E. crus-galli a delay in plant development and lower seed production were registered for the R biotype. Instead, in A. myosuroides a delay in speed of seed germination and seedling emergence rate was observed, in particular under stress conditions. The implications of these fi tness costs were investigated using a generic model for simulation of annual weed population dynamics in monoculture cropping systems. Simulations were performed to investigate the dynamics of weed populations in herbicide-free fi eld situations and the potential of IWM to replace herbicide-intensive weed management, thereby reversing the selection for herbicide resistance. For E. crus-galli, the R fi tness penalty in reproduction resulted in a stable population density with a population growth rate close to zero, under the assumption that other still eff ective herbicides generated eff ective control (98%- 99%). Whereas for A. myosuroides, the slight fi tness penalty detected in emergence rate resulted in no perceptible diff erence in population growth rate between the biotypes, implying that a herbicide-free cultivation approach would not help to control the R biotype. The combination of experimentation and modelling proved eff ective in translating detailed observations on fi tness costs at plant level to agronomic implications at fi eld level.