1Plant Biology Department, and 2Food and Farming Systems Department, Scottish Agricultural College (SAC), Auchincruive, Ayr KA6 5HW, UK

Background and objectives
Ryegrass mosaic virus (RMV) is a major pathogen of Lolium multiflorum Lam. and L. perenne L. in the UK. Yield losses can be substantial where swards of single susceptible cultivars are treated with high rates of nitrogen. Grass swards for livestock are generally a mixture of several species and cultivars. The impact of RMV infection on competition between different component cultivars of swards has not been examined in relation to nitrogen rate. The effect of rate of nitrogen on competition between a very susceptible L. multiflorum cultivar and a moderately resistant L. perenne cultivar was examined over two seasons in a field experiment. In a glasshouse experiment, cultivars of L. perenne, L. multiflorum and a hybrid of the two were used.

Materials and methods
The field swards, which were a mixture of cv. Barverdi (L. multiflorum) and cv. Peramo (L. perenne), were sown in June 1994. The four treatments comprised a 2x2 factorial of nitrogen regimes (170 or 340 kg N/ha/year)xRMV inoculation (or not). Plots were first inoculated with a severe strain of RMV in August 1994. The trial was harvested on six occasions between autumn 1994 and autumn 1996 and the total and cultivar yields determined. In the glasshouse, three cultivars, i.e. Barverdi, Merlinda (L. perenne) and Dalita (a L. multiflorum/L. perenne hybrid) were grown in simulated mixtures. Individual cultivars were grown in separate pots to facilitate identification at harvest but the pots were grouped in sand beds to allow competition for nutrients and water. The six treatments comprised a 3x2 factorial of nitrogen regimes (600, 1200 and 1800 mg N/l of compost)xRMV inoculation. Inoculation with RMV took place 6 weeks after sowing. Yields were assessed 6 and 12 weeks after inoculation.

Results and conclusions
In the field experiment, RMV infection reduced the contribution of the very susceptible L. multiflorum to sward yield. There was no effect of RMV infection until the first harvest of 1996, at which time the reduction in RMV-infected swards compared with non-inoculated swards was similar for both rates of nitrogen. However, at the October cut in 1996, RMV infection caused a reduction in the yield of L. multiflorum at the higher rate of nitrogen only. The effect of RMV infection was to prevent the yield response of L. multiflorum to the higher rate of nitrogen. Whereas in 1995 the percentage contribution of L. multiflorum to yield increased between the first and last harvests to greater than 80% for all four treatments, a similar value was obtained during 1996 only for the non-inoculated plots treated with 340 kg nitrogen. The main impact of RMV followed an extended period of severe frost at the end of 1995. There is evidence that the frost severely restricted the yield of RMV-infected L. multiflorum tillers in 1996. Where RMV reduced the yield of L. multifiorum, there was compensatory growth by the less susceptible L. perenne. The compensatory growth was incomplete, especially at the higher rate of nitrogen, resulting in a reduced total yield for the virus-infected sward. The effect of rate of nitrogen on interactions between RMV and the components of mixed ryegrass swards not containing L. multiflorum was also examined in a field experiment [1].

In the glasshouse experiment, RMV infection reduced the yield of L. multiflorum where the two higher rates of nitrogen were applied. However, at these rates the yield of the other less susceptible cultivars increased most to compensate, and therefore RMV infection did not significantly reduce the total yields of the mixtures. As in the field experiment, compensatory growth by the less susceptible cultivars was generally not complete and for the two lower rates of nitrogen the losses of total yield were 4.6 and 6.0%.

1. Bain RA, Raynor EL, Baker LJ, Bagnall A, 1988. Annals of Applied Biology (in press).