SummaryField experiments were conducted to study weed population shifts in long‐term conservation tillage systems. The objectives of this study were to determine weed community abundance, diversity and composition on conventional tillage (CT), minimum tillage (MT), no‐tillage with paraplow (ZT) and no‐tillage (NT) systems, and to identify species that are associated with specific tillage systems. The paraplow is a subsoiling technique that results in a deep loosening of the soil, in order to alleviate compaction in certain soils where NT is practiced. The results showed significant differences in both the composition and the abundance of weeds, depending on the tillage systems. Weed diversity, species richness and Shannon's diversity and evenness indices were higher under the conservation tillage systems than in the CT system. In addition, various weed species were associated with reduced tillage systems. For instance, Anthemis arvensis, Hirschfeldia incana and Lolium rigidum became more prevalent in the NT system, whereas Chenopodium album and Filago pyramidata dominated in the ZT system. Therefore, the application of a paraplow treatment changed the weed community in the NT system. Other weed species, such as Capsella bursa‐pastoris and Torilis nodosa, dominated in all three conservation tillage systems, whereas soil disturbance by mouldboard ploughing favoured species such as Polygonum aviculare and Phalaris paradoxa.