SOIL PREPARATION

in every second inter-row while the soil was moist during autumn i.e. post- harvest. Inspection of profile pits showed no obvious compact layers down to a depth of 90 cm, an observation that was later confirmed by bulk density measurements. The ripper was set to cut the roots 60 cm from the vines to a depth of 30 cm. During the following year the tines were moved 30 cm towards the middle of the implement so as to cut the roots in the same inter-row, 90 cm from the vine row. This procedure was repeated annually until the whole inter-row was treated after four years. The implement was also set to loosen the soil progressively deeper to reach a maximum of 60 cm in the middle of the row in the fourth year. At the same time, the soil in the adjacent inter-rows was loosened and the roots cut in the reverse order i.e. the loosening action started in the middle of the inter-row. The treatment was repeated after four years, but the order of the ripping in the inter-rows was reversed. Intensive root pruning in this experiment had a serious negative effect on grapevine vegetative growth and yield. It can be concluded that the cutting of roots on both sides of vine rows simultaneously will seriously decrease the performance of a vineyard and should therefore be done only in alternate rows and only between the tractor tracks. Root pruning on the other side of the row should only be done in the following season. Furthermore, root pruning should be reserved for vineyards where root growth and consequently also overall vineyard performance is restricted by soil compaction. Van Huyssteen (1981) also recommended that it should not be done more regularly than once every five years. Root pruning experiments aimed at reduction of vigour of Shiraz in Australia (Dry et al ., 1998) and Cabernet Sauvignon in the United States (Giese et al ., 2015) also pointed to the importance of pruning severity and timing and even to an interaction between these two factors (Dry et al ., 1998). Application of this technique at bud burst on both sides of the vine row produced a moderate decrease in vegetative vigour without a significant lowering in grape yield. One- sided root pruning at bud burst was not severe enough to affect shoot growth, while the cutting of roots later in the season at flowering only had a minor effect because a large proportion of shoot growth had already been formed by then. The biggest grapevine response was found with two-sided root pruning conducted on the same vines every year. This drastic action not only achieved the desired decrease in vegetative growth, but unfortunately grape yields were also reduced significantly. Such a decline in yield can only be justified by a significant improvement in grape quality which was neither evident in the Australian nor in the American results. Giese et al. (2015) concluded that root pruning cannot be recommended as a commercial practice to control vegetative growth on the deep soils used in their experiments. A series of three more recent experiments in Australia investigated the effects of ripping and surface mulching on grapevine performance. In two

56 | ROOT RESPONSE TO SOIL CONDITIONS