Persistence of grasses varieties and legumes mixtures selected for use grazing in organic farming.

Studies were carried out in the years 2006-2010 in the Experimental Farm of the ITP in Biebrza on permanent meadow situated on mud-peat-muck organic soil. Experiments were set up in a random block system in four repetitions on 15 m plots. A mixture (the so-called model mixture P1) designed for pasture use in an organic farm was compared with the purchased mixture P2, control (simplified) mixture P3 and an indicator species Lolium perenne L. 4n Barista (P4). Experiment was fertilized with 50 kg P2O5/ha in spring and with 60 kg K2O/ha (30 kg in spring and summer). Sward was mown 5 times. Botanical composition was analysed with the approximate method, sward height was measured on the day of mowing, dry weight yield and its quality was analysed with the method of Filipek (1973). Trifolium repens L. Astra and Riesling, Festuca pratensis Huds. Skra and Agrostis gigantea Roth Mieta sown in mixtures disappeared after 5 years of simulated pasture use. The smallest changes were observed in P1 mixture and the greatest – in the sward of P4 variant (Lolium perenne L. Baristra). Poa pratensis L. Skiz, Phleum pratense L. Skala and Festuca rubra L. Atra appeared most persistent among sown species. Most important from among unsown grasses were Poa pratensis L. and Poa annua L. and from among weeds a stoloniferous Ranunculus repens L. more tolerant of excessive moisture. Mixtures grew best in the summer period (the 3 and 4 cut) without significant differences between mixtures. Sward height depended more on moisture conditions and cuts than on mixture types, especially in subsequent years of grassland use. Total yields of particular mixtures were similar (from 18.69 to 20.76 t/ha); slightly larger for simplified (P3) and model (P1) and smallest for P4 variant (Lolium perenne L. Baristra) in pure sowing due to the highest soil degradation. A greater differentiation of yields was found among cuts and years with the most important effect of thermal and moisture conditions. Fodder value measured with the number of utility value was very good and good with the exception of P4 sward (Lolium perenne L.) which, due to a strong degradation in the last study year was mean (Lwu 5.92). Sward of the model mixture P1 was of very good quality (Lwu 8.0-9.0) during the whole study period which was a result of minor changes in its botanical composition. Observed changes evidence diminishing differences resulting from the composition of sown mixtures and the development of communities floristically adapted to habitat conditions. However, proper selection of species and their varieties for mixtures may slow down this process.