Abstract
The significance of physical soil factors in controlling the growth of subterranean clover under irrigation was assessed in a field experiment involving small plots scattered over a range of soils. Nutrient differences were reduced by a complete fertilizer dressing, so that as far as possible vegetative growth was controlled only by moisture supply and soil aeration under two watering treatments, one approximating the normal regional practice and the other a more frequent watering. Yield was increased by the latter and variation between sites due to soil differences was reduced. Spring yields from the plots watered at the normal regional level were significantly correlated with water increments at irrigation but not with available water capacity. Fifteen other soil properties showed significant simple correlations with spring yield and in multiple regression relationships a high proportion of yield variation could be accounted for in terms of the variation in selected properties. The best pair of variables for yield prediction, accounting for 59 per cent of variability, were per cent exchangeable sodium at 1-4 inches, and aggregate density at 1-4 inches. These variables were highly correlated with water increment at irrigation. On the more frequently watered plots, waterlogging induced in late winter depressed yield slightly and this, apparently, was not caused by poor aeration, but possibly by leaching of nutrients. Oxygen diffusion rates measured one day after irrigation were negatively correlated with yield. Spring yields from these plots were unrelated to total and air-filled porosity and air permeability. The evidence again pointed to moisture supply being responsible for the yield variation between sites.