Implications of grass–clover interactions in dairy pastures for forage value indexing systems. 5. Southland

This experiment, implemented at a Southland site representative of the lower South Island, New Zealand, and monitored for three full years from 2012/2013 to 2015/2016, was one of four throughout New Zealand that investigated the yield and nutritive value rankings of perennial ryegrass cultivars sown with or without white clover and with high (225 kg N/ha/annum) or low (50 kg N/ha/ annum) application rates of nitrogen fertiliser. High inputs of nitrogen (N) fertiliser increased annual total dry matter (DM) yields and seasonal yield in 12 out of 18 measurement occasions. Adding white clover increased both annual dry matter yield and seasonal yield in 7 out of 18 measurement occasions. An interaction between clover and N input occurred on 10 occasions. Over time, the low N minus clover treatment became progressively lower yielding compared with the other three treatments. Ryegrass cultivar by clover interactions occurred in the latter half of the experiment. When analysed using ryegrass characteristics, earlyand late-heading ryegrass cultivars produced similar yields when sown without clover, but late-heading ryegrass cultivars had a higher yield than early-heading pastures when sown with white clover (P < .001). Tetraploid ryegrasses had lower yields than diploid ryegrasses when sown without clover (P = .001). This difference was reduced when white clover was included. There was no significant interaction with ryegrass morphology when sown with white clover, as dense ryegrass cultivars always had a higher yield than open ryegrass cultivars (P = .005). Interactions occurred between N input levels and ryegrass cultivar. These occurred in early spring or late spring of each year. Analysis by ryegrass characteristics identified consistent interactions. Ryegrass cultivars with a late-heading date were higher yielding than early-heading cultivars at low N inputs (P = .029), but they were not different at high N inputs. Diploid and tetraploid ryegrass cultivars were similar with low N inputs, but diploid cultivars out-yielded tetraploid cultivars at high N inputs (P = .005). Significant interactions between N inputs and white clover addition affected herbage nutritive value but this did not extend to consistent interactions with ryegrass cultivar. Invertebrate pest numbers were generally low and unlikely to have influenced clover/ ryegrass interactions or cultivar ranking. Results from this site suggest that ranking of ryegrass cultivars needs to account for both the inclusion of clover and variation in N fertiliser input.