Separating the confounding effects of farming practices on weeds and winter wheat production using path modelling

Optimal crop yield can be achieved directly by optimizing farming practices to increase crop growth and indirectly by optimizing pest management to decrease pest pressure. The aim of this study was to quantify the indirect effect of farming practices on yield through a change of the weed pressure and, thereby, disentangle the effect of farming practices on yield and weeds. Between 2006 and 2012 in Burgundy, France, 152 winter wheat fields were surveyed for weeds and farmers were interviewed about their farming practices and yields. Data were analysed using partial least square path modelling (PLS-PM). A path model that related farming intensity (fallow management, sowing, chemical pest control and fertilization), crop productivity (yield), and weed pressure was designed and validated. It was then used to assess the relationships between the identified variables (β path coefficients) and compare groups of fields varying in, the preceding crop, herbicide use and weed pressure in the field. Farming intensity had a positive effect on crop productivity (β = 0.32). Weed pressure negatively impacted crop productivity (β = −0.12). Farming intensity decreased weed pressure and had a sufficiently negative effect on weeds (β = −0.19) to counteract the negative impact of weeds on crop productivity. Therefore, the indirect effect of farming intensity on crop productivity through a change of weed pressure was positive and accounted for 7% of the total (direct + indirect) effect of farming intensity on crop productivity. The indirect effect of farming intensity on crop productivity varied by preceding crop (3.6% and 23% with a winter and spring/summer preceding crop, respectively) and herbicide use (14.1% and 2.1% when herbicide use was less and more than the regional reference, respectively) and weed pressure (0.5% and 2.6% when the total weed abundance after weeding was less and more than 2 individuals m−2, respectively). From the path model, we quantified the direct and indirect effects of farming intensity on crop productivity to show that effective weed management can sustain crop production in cropping systems with reduced herbicide use.