Growth components of Lolium perenne and Agrostis spp. tillers and Trifolium repens stolon apices were compared for one year under set stocking and rotational grazing with sheep on hill pastures which had had low and high levels of fertiliser input. Leaf appearance rates of L. perenne and Agrostis spp. were unaffected by grazing management or fertiliser level. Leaf appearance in T. repens was also largely unaffected by treatments except in summer when rotational grazing allowed some relief from defoliation stresses and leaf appearance rates were greater than under set stocking during 3 grazing rotations. T. repens leaf appearance rates were slower in winter and increased later in spring than the grasses. For most of the year, tiller bud development in L. perenne was well below the maximum possible because of mutual shading in dense swards (mean total tiller density 29700 tillers/m 2 under set stocking and 20 too/m2 under rotational grazing). L. perenne lamina lengths, tiller lengths, and leaf extension rates were approximately twice those of Agrostis spp. Leaf extension rates were faster under rotational grazing for L. perenne and Agrostis spp. but tiller densities of both species were greater under set stocking. Total leaf growth per unit area is, therefore, likely to be similar under both managements. Slightly greater L. perenne growth and less Agrostis spp. growth per unit area under rotational grazing than under set stocking is indicated by a comparison of individual tiller growth and tiller density data. The results are discussed in relation to published findings from the same systems.

Leaf and tiller or stolon death of Lo/ium perenne, Agrostis spp., and Trifolium repens in set-stocked and rotationally grazed hill pastures AbstractLeaf death rates, leaf longevity, and tiller and stolon death rates of Lolium perenne, Agrostis spp. and Trifolium repens were compared for one year on hill pastures which were set-stocked and rotationally grazed by sheep. Leaf death rates and leaf longevity were largely unaffected by grazing management, and the number oflive leaves per tiller or stolon did not differ greatly between managements. Agrostis spp. had more leaves per tiller than L. perenne. Leaf longevity was least in spring and summer when leaf death rates were fastest, and was greatest in winter when leaf death rates were slowest. In autumn, spring, and summer, mortality of leaves occurred within 3 weeks of their appearance, resulting in the loss of up to 28% of leaves which had been formed soon after one grazing before the next grazing under the rotational management. The long rotation in winter (63 days) resulted in the death of many leaves before they had a chance to be grazed. Leaf utilisation data showed that 20-46% of L. perenne leaf tissue, 45-63% of Agrostis spp. leaf tissue, and 34-60% of T. repens leaf production was ungrazed and subsequently died and decayed. Grazing pressure is critical in determining the amount of leaf tissue entering the litter pool, and has a considerable influence on tiller and stolon death.

No abstract

(Festuca arundinacea Schreb.), established successfully at most sites. Findlay-Wilkinson analysis of annual total herbage accumulation (THA) of four sites, found that Maru production was above average and resident production was below average. Variation in THA between sites and years was related to variation in annual rainfall (621-1484 mmI year). Wana was the most persistent grass at all sites (60% of THA after 4-6 years), followed by Nui (51 % of THA after 4-6 years). Maru was only persistent at Southland, Wairarapa, and Taupo (averaging 39% ofTHA after 4-6 years). Matua was only persistent at Wairarapa (32% of THA over 5 years), and Roa was only persistent at Taupo (58% of THA over 5 years). At most sites, at least one grass species (in many instances more) outperformed the resident sward.

The influence of six pre-sowing herbicide treatments (no spray, two rates of glyphosate, and three rates of paraquat) on oversowing success of three species (,Grasslands Wana' cocksfoot, 'Grasslands Maru' phalaris, and 'Grasslands Tahora' white clover) were compared in autumn at two southern Hawke's Bay hill country sites (Ballantrae and Dannevirke) and in two seasons (spring and autumn) at the Ballantrae site. Pasture suppression by herbicide was compared using point analysis. Initial results from oversowing were compared by regular counts of seedling appearance and long -term results were determined from the contribution of introduced species to total herbage accumulation (RA) during the following 3-4 years. Paraquat and glyphosate were similar, at equal rates of active ingredient, for the establishment and performance of Wana or Maru, and were best at the highest rate of 0.7-0.8 kg a.i./ha. The contribution of Wana increased over 3 years under continuous sheep grazing to be 13-27% of total RA depending on the site, year, and herbicide combination, and was maintained during a fourth year. Wana did not significantly increase seasonal or annual total HA. Maru failed to establish and produce successfully. The establishment of Wana and Tahora was significantly and linearly related to rainfall in the first week and first 2 weeks respectively, following sowing, but unrelated to rain before sowing or 3 weeks after sowing. Spring oversowing atBallantrae was unsuccessful.