Size and composition of weed seedbank in long-term organic and conventional low-input cropping systems

Long-term effects of organic (ORG) and conventional low-input (CONV) farming systems on size and composition of weed seedbank were assessed in 2007 in central Italy, in an on-farm experiment set up in 1996. In this experiment, six rotations (R1–R6) were considered, basically consisting on the same crop sequence with different starting points, i.e. (1) legume crop (soybean/field bean/common pea), (2) vegetable crop (pepper/melon), (3) winter cereal (soft/durum wheat), (4) summer cereal (maize) (5) industrial vegetable (processing tomato), (6) winter cereal (soft/durum wheat). All rotations were established both in ORG and in CONV, in strict adherence to EU regulations (ORG: EU reg. 2092/91; CONV: EU reg. 2078/92). Intercrops with green manuring purposes were included in ORG, while weed control was performed mechanically in ORG and chemically/mechanically in CONV. Weed seedbank was determined on 10 soil samples (0–0.30 m depth) in each plot and seeds were directly extracted, identified and counted. In all rotations, the adoption of ORG resulted in a significant increase in weed seedbank density, particularly in the case of summer weed species (Portulaca oleracea L., Amaranthus retroflexus L. and Chenopodium album L.), which are more competitive and difficult to control in summer crops under organic farming systems in the Mediterranean climates. Small differences between ORG and CONV were found in terms of number of weed species (23 in ORG and 20 in CONV, on average), but the index of diversity in CONV was on average higher than in ORG. Furthermore, the most dominant weeds in CONV represented a lower percentage of total seedbank (40%, 23% and 5%, respectively, for P. oleracea, A. retroflexus and C. album in CONV and 56%, 32% and 4% for the same three weeds in ORG). These results confirm that the wider availability of effective weed control methods in integrated low-input farming systems (CONV) is helpful to maintain a low seedbank size, with a lower dominance structure. However, the adoption of ORG systems based on long rotation cycles, very competitive crops and accurate weed control, especially at the beginning of the ORG management, may be sustainable in the long run, in terms of potential weed infestation levels.

► Weed seedbank in organic (ORG) and conventional (CONV) systems was determined.
► After a 12-year trial, seedbank size in ORG was 2.2 times higher than in CONV.
► Weed community biodiversity in ORG was slightly lower than in CONV.
► Dominance of broadleaved summer weeds increased in ORG.
► Weed management in summer crops is the ‘bottleneck’ in ORG systems.