Agroecosystems in Europe provide not only a production function but also a recreation and nature conservation function. Weed science therefore is able to contribute to agronomic and ecological issues. This paper describes the population dynamics of weed in integrated an organic farming systems with emphasis on the impact of conservation tillage on the relationship between seed bank and the above ground weed flora. In 1992 a field trial comprising an integrated and organic farming system was initiated at the Scheyern Research Farm in southern Germany. Field plots were set up on formerly conventionally ploughed arable land. Three tillage treatments were applied in the integrated farming system: (1) minimum tillage, where soil disturbance is restricted to seed drill, (2) tillage with a chisel to a depth of 10 cm and (3) conventional ploughing to 20 cm working depth. In the organic farming system two tillage regimes were analysed: (A) reduced tillage omitting deep ploughing in four of seven years in rotation, (B) conventional tillage with a plough to a depth of 20 cm each year in rotation. Seven years after the application of different soil cultivation vegetation was recorded from 1999 to 2002. Vegetation records contained seed bank analyses, repeated weed counts during the vegetation period and recording of weed biomass and flowering plants. Results for the integrated farming system show that the tillage treatment influenced weed dynamics to the greatest extend among factors included in the analysis (crop type, previous crop, tillage treatment, year of observation). Density of the weed seed bank increased in both reduced tillage systems five to six times in relation to conventional ploughing. Percentage emergence in spring and over the total vegetation period increased under conservation tillage, resulting in an increase of established plants in spring and especially summed over the whole vegetation period. As a result of improved weed establishment the biomass and number of flowering plants increased causing greater additions to the seed bank than in the conventional ploughing system. Species number increased both for seed bank and actual weed flora under conservation tillage. Therefore species diversity was promoted by conservation tillage. Those species that occur under conservation tillage but not under conventional tillage are common species of grasslands and field margins but not known as typical weeds. Species number in conservation tillage systems is increased by species that occur in low frequency and abundance and species evenness is decreased because of a few, especially troublesome, species which became dominant. Species that were promoted by conservation tillage are Chenopodium album, Elymus repens, Epilobium ciliatum, Galium aparine, Poa annua and Matricaria recutita. Conservation tillage contributes to the diversity on ecosystem level but a promotion of typical weed assembly can not be expected. Especially under conservation tillage the requirements to predict weed development are quite good. Percentage emergence of single species is more suitable as predictor than percentage emergence of the whole weed populations. Concerning the organic farming system the reduction of tillage intensity showed no influence on population dynamics of weeds. But species richness and diversity were changed when tillage was omitted. Grass weeds were promoted by reduced tillage. The tillage system of the organic part of the Scheyern research farm provides an economical and ecological alternative to conventional ploughing. The cropping of grass-ley mixtures reduces weed density in the following crops. Thus the percentage of grass-ley mixtures in rotation is an important tool of weed management in organic farming systems. In comparison to the integrated farming system the weed seed bank, the weed density, the number of flowering plants and weed biomass is significantly increased in the organic farming system. Percentage emergence differs not between the two farming systems. In both farming systems the long term population dynamics of weeds can be predicted by seed bank, plant density, density of flowering plants and weed biomass.
«