“…These field reports also pointed to the main short-coming of the simulation-based method, i.e., FlorSys only considers annual weed species. This is not a problem in tillagebased systems as shown by the evaluation (Colbach et al, 2015, submitted) but certainly leads to underestimating weed harmfulness in no-till systems which tend to favor perennials (Buhler et al, 1994;Tørresen et al, 2003;Norsworthy, 2008;Trichard et al, 2013), indicating that the low or no-till strategies identified here as optimal for reconciling production and biodiversity would only remain efficient if there were no risk perennials. Pending the development of a perennial weed dynamics model, an indicator of perennial weed spread depending on cultural practices could be developed from expert opinion.…”