The post-emergence herbicides, difenzoquat, flamprop-methyl and dichlofop methyl were evaluated for their effect on wild oats (Avena sterilis and A. fatua) and selectivity in wheat and barley crops on the Darling Downs, Queensland. Barban was included as a standard treatment. These herbicides generally did not provide complete control of wild oats. Wild oat control as measured near harvest-time, was equally as good with tillering or pre-tillering applications, with only two exceptions. Production of wild oat spikelets was reduced by all herbicides. However, although spikelet numbers were reduced to a very low level occasionally, spikelet production was usually quite substantial (500 to 1500 m-2). The most consistent increases in grain yield were obtained from dichlofop methyl (0.75 kg ha-1), difenzoquat (0.75 kg ha-1) and barban (0.175 kg ha-1) when applied prior to the tillering stage of the wild oats. With a few exceptions, dichlofop methyl, difenzoquat, and flamprop-methyl (0.45 kg ha-1) applied during tillering did not increase grain yield. Difenzoquat and flamprop-methyl were also evaluated at 1.5 and 0.9 kg ha-1 respectively. Apart from a small yield increase in two trials there was no advantage from these higher rates. For dichlofop methyl a rate of 0.5 kg ha-1 was less effective than 0.75 kg ha-1 in one of two trials in 1976. In weed-free trials, flamprop-methyl reduced the height and grain yield of the cultivar Kite, but had little or no effect on five other wheat cultivars. The damage occurred when flamprop-methyl was applied at the end of tillering, but not at the early tillering stage. These trials show that an increase in grain yield can occur when the effect on the wild oats has been only moderate, so herbicide performance should not be judged only by the effect on the wild oats.
