Mechanical weed control in onion seed production

Two field experiments with onion (Allium cepa L.) cultivated by the "seed-to-seed" and "bulb-to-seed" methods for the organic seed production were carried out in central Italy in order to evaluate the effects of some mechanical methods (harrowing, hoeing, hoeing-ridging, split-hoeing, finger-weeding, manual weeding) on: i) weed control; ii) selectivity to the onion; iii) onion seed production. The results showed that the choice of the seed production method and then the mechanical treatments require to be managed very carefully in order to maximize the weed control, reducing seed yield losses. The seed-to-seed method seemed to be not advisable in the Mediterranean areas, because there were great difficulties to control weeds by mechanical method mainly owing to the long crop cycle with more fluxes of weed emergence, the very low competitiveness of onion at the early growth stages and the weather conditions during the crop cycle that has greatly affected the timeliness, repeatability and effectiveness of mechanical control. In these conditions, competition of uncontrolled weeds caused heavy seed yield losses or the crop failure and only manual weeding ensured a satisfactory weed control and seed yield, although with a very high hand labour. The bulb-to-seed method was the only advisable for onion seed production because allowed to control weeds by mechanical methods, obtaining good levels of seed yield. In particular, combined intra and inter-row methods, such as hoeing-ridging or split-hoeing thorn finger-weeding, allowed a high effectiveness against weeds (ranged from 63% to 92%) and a good selectivity to the crop, reducing weed-crop competition and assuring satisfactory onion seed production. Unlike seed yield, seed quality was not affected by weed competition: weight, moisture and germinability of onion seeds were always over of the standard levels required for commercialization, regardless of weed competition levels.