THE USE OF AN UNENCLOSED FIELD FUMIGATION SYSTEM TO DETERMINE THE EFFECTS OF ELEVATED OZONE ON A GRASS CLOVER MIXTURE

A mixture of Trifolium repens L. (var. Grasslands Huia) and Lolium per enne L. (var. Melle) was established in an area with low ambient ozone concentrations and exposed to episodes of ozone using a simple, unenc losed field fumigation system. Control plots were exposed to ambient a ir. There were two cutting frequencies to simulate grazing and silage production respectively. This paper reports on the performance of the system and the effects on yield and species composition in the first y ear of fumigation. The system operated satisfactorily when the wind sp eed was between 1 and 5 m s(-1). When the wind dropped suddenly below 1 m s(-1) there were short periods with high concentrations but these events were rare and did not last more than a few minutes. Above 5 m s (-1) there was lateral downwind spillage of ozone and at higher wind s peeds it was not possible to maintain the target concentration of 50-7 0 nl 1(-1). The plots were fumigated only when the weather was suitabl e for ozone formation and when the wind speed was stable. Because the weather was poor, fumigation was restricted to 22 d between July and S eptember. The mean 4-7 h concentrations were 50-70 nl 1(-1). Seven-hou r mean ambient concentrations over the same period were 10-28 nl 1(-1) . There were significant effects of ozone on yield and herbage composi tion. The effects varied between harvests, probably because of differe nces in the timing of cutting relative to the episodes, and to the dif ferences in phenology between the species. Most of the loss of yield w as due to effects on Trifolium, which was much more sensitive than Lol ium. There was a persistent effect of ozone on the stolen density of T rifolium which lasted into the following year. It is suggested that th is arose because the clover replaced damaged leaves very quickly, main taining the canopy at the expense of the stolons. Persistent effects o n stolons might lead to poorer winter survival and lower nitrogen fixa tion. Gaps in the sward as a result of loss of Trifolium might allow r apid weed invasion. There was no interaction between cutting frequency and ozone, which may have been due to the fact that the poor summer r estricted the number of cuts and therefore decreased the difference be tween the two cutting frequencies.