OVERSOWING OF 5 GRASS SPECIES AND WHITE CLOVER ON A TAUPO HILL-COUNTRY PUMICE SOIL

Methods of establishing perennial ryegrass (Lolium perenne L.), prairi e grass (Bromus willdenowii Kunth), phalaris (Phalaris aquatica L.), c ocksfoot (Dactylis glomerata L.), tall fescue (Festuca arundinacea Sch reb.), and white clover (Trifolium repens L.) by oversowing on a hill country pumice soil near Taupo were investigated in spring 1982 and 19 83. Four pre-sowing sward treatments (paraquat herbicide application, hard grazing by sheep, treading by sheep, or treading by cattle) were applied in two experiments. Populations of seedlings, tillers, and pla nts, and herbage growth, were monitored for 2 years after each sowing. Degree of sward suppression resulting from sward treatments was the m ain factor influencing establishment success. Hard grazing resulted in minimal suppression of resident sward growth and hence poor establish ment (2% of sown grass seed). Sheep treading yielded an average establ ishment rate of 4% of sown seed. Cattle treading removed most of the r esident vegetation and resulted in high seedling emergence rates, but seedling survival was poor and final establishment was low (3% of sown seed). Paraquat herbicide application gave the best grass establishme nt (13%) through a combination of high seedling emergence and good see dling survival. Large increases in the content of oversown grasses in swards occurred with time so that, in several instances, apparent esta blishment failures were rated as successful oversowings 2 years later. Grass species showed different responses to sward treatments and diff erent rates of early growth, factors which should be considered when e stablishing grasses in this environment. Emergence of white clover cul tivars Grasslands Tahora and Grasslands Huia averaged 60% of sown seed , but, with time, levels of white clover did not differ between sown a nd unsown plots.