The influence of herbicides, nitrogen fertilizer, seed rate and method of sowing, on the establishment and long-term composition of a perennial ryegrass ley.

Changes in ihe species composition of a ryegrass pasture were examined over a 2 year period from a spring establishment. Seed of S23 perennial ryegrass was sown in main plots, eilher broadcast or drilled, at rates of 15, 25 or 35 kg/ha in a field experiment at Oxford. The whole area was oversown withfour weed grasses (Agrostis siolonifera.Holcus lanatus. Poa annua and P trivialis). Five cultural and three herbicide treatments were applied across the plots. These consisted of three levels of nitrogen (0. 50 and 100 kg N per ha), two companion species (Italian ryegrass and white clover) and 2.4-D, benazolin and methabenzihiazuron applications. Ten weeks after sowing, broad-leaved weeds accounted for nearly half the lotal dry-matter yield, except where controlled by the three herbicides. Grass yields were nol increased by herbicide treatment. The only cultural treatment checking broad-leaved weeds was ihe 35 kg/ha seeding rale. Seven months after sowing, the only treatment which had decreased the abundance of weed grasses was melhabenzthiazuron especially againsi A. stolonifera and Poa spp. Two years afler sowing, perennial ryegrass had only 34% ground cover and had been replaced largely by H. lanatus (38%), Poa spp, (15%) and A. stolonifera (7%), Where methabenzthiazuron had been used the ryegrass ground cover was 48%. with less P. trivialis and A. stolonifera. but H. lanatus was nol altered. White clover decreased H. lanatus cover. In two supplementary pot experiments, methabenzthiazuron sprayed pre-emergence damaged perennial ryegrass unless the seed was covered with at least 4 mm of soil. Only H. lanatmand F. rubra out of sixteen grasses were relatively resistant to methabenzihiazuron when applied at ihe tillering stage, confirming that this herbicide has potential for preventing the ingress of many unsown grasses in new leys.