Modelling the correlation between plant phenology and weed emergence for improving weed control

The efficiency of weed control practices could be improved if their timing is linked to weed emergence dynamics. A study was conducted in a pre-alpine valley in northern Italy to evaluate whether the phenological phases of some perennial plants could serve as reliable indicators of time of weed emergence and thus be an alternative to bioclimatic models for supporting management decisions. Weed emergence was observed from 2003 to 2004 in five sites at different altitudes. The emergence dynamics of the main weeds were modelled with a Gompertz model. The phenology of 10 common shrubs and a tree was monitored by visual assessment. The flowering and fruit-ripening phases of the most useful shrubs were modelled with a lognormal model. Correlation analysis between the two functions was used to study the correspondence between plant phenology and weed emergence. Flowering and fruit-ripening phases of the shrubs were well described by the lognormal model. The correlation analysis between the lognormal model and Gompertz model showed correspondences between the phenological phases and emergence dynamics of the main weeds. The proposed method can be used to examine shrub phenology–weed emergence correlations and consequently for supporting weed management, under certain conditions.