Effects of soil disturbance on weed seedling emergence and its long-term decline

Résumé: Zusammenfassung The long‐term effect of different depths of soil cultivation on weed seedling emergence from naturally occurring populations of weed seeds in the soil was examined in four experiments on land previously in pasture, raspberry canes or intensive vegetable production. At approximately monthly intervals, weed seedlings were counted and then killed with contact herbicides, after which plots were cultivated to 250 or 10mm, or were left undisturbed. The treatments were continued for 7 years. One experiment was then discontinued whilst the previously uncultivated plots on the other three were cultivated to 150 mm at approximately monthly intervals for a further 4 years in one experiment, and for 8 years in the other two. After the first year, very few seedlings emerged in the uncultivated and shallow cultivated plots, and seedling numbers declined slowly in the deep‐cultivated plots. Under repeated deep cultivation, seedling emergence of almost all species declined exponentially. Different species declined at different rates, with Rubus idaeus L., Plantago lanceolata L., Veronica arvensis L. and Ranunculus spp. being the most rapidly declining group. Rates of decline for individual species were similar to those observed in Europe.Juncus bufonius L. behaved differently from the other species, and showed no decline in seedling numbers. In the initial 7‐year period, 28 000 weed seedlings per m−2 emerged from the deep‐cultivated plots on soil previously cropped with vegetables. Over the same period, less than 11 000 seedlings emerged in shallow‐cultivated plots, and just over 4000 in uncultivated plots. In the second phase of the experiments, fewer seedlings of most species emerged than in the first phase, and the decline in numbers of Coronpous didymus (L.) Sm. seedlings was slower. Résumé Les effets à long terme de différentes profondeurs de travail du sol sur la levée des mauvaises herbes ont étéétudiés sur des populations naturelles dans des parcelles dont les précédents étaient: paturage, framboisier ou maraichage intensif. À des intervalles d'environ un mois, les jeunes plantes de mauvaises herbes étaient dénombrées puis détruites à l'aide d'herbicides de contact, aprés quoi les parcelles étaient travaillées sur 250 ou 10 mm, ou laissées sans intervention. Cette premiére phase a duré 7 ans. Une expérience a alors été arretée et les parcelles précédemment non cultivées des trois autres ont été travaillées à une profondeur 150 mm, à des int