The effect of seedbed treatment, cutting frequency and selective grass defoliation on the production and botanical composition of experimental swards of Urochloa mosambicensis and Bothriochloa pertusa mixed with Stylosanthes

Two experiments are reported that examine the competition of tropical grasses and legumes in simulated swards grown in boxes in the glasshouse. These were designed to explore reasons for the performance of swards of Urochloa moscambicensis cv. Nixon (Nixon) and Bothriochloa pertusa cv. Bowen (Bowen) in association with Stylosanthes hamata cv. Verano (Verano) under grazing at high stocking rate. In the first, Verano and S. scabra cv. Seca (Seca) were oversown into swards of Nixon or Bowen or into bare soil. After 8 weeks, legume dry weight, height, leaf number, leaf area and shoot number were greatly reduced by the associated grasses while the time to flowering was increased. Verano greatly outyielded Seca in all treatments by a factor of about 3. Top growth of the legumes grown in mixtures was <26% of their growth in bare soil. Root weights of legumes in mixtures were only about 18% of those in pure stands and comprised only 4% of the total root mass of the grass-legume mixtures. When grown with Bowen, both legumes had lower uptakes of N, P, K and S than when grown with Nixon. For Verano, uptake of Ca and Mg were also lower when grown with Bowen, suggesting that Bowen was more competitive for these nutrients than was Nixon. In the second experiment, swards of Nixon-Verano and Bowen-Verano were cut every 3 or 6 weeks for 12 weeks. At each cutting, either the grass only or the grass + Verano was cut. After 12 weeks, all swards were cut to 5 cm and regrowth measured after 4 weeks and again at 8 weeks to assess residual effects of previous treatments. Yields of grass were higher with 6-weekly cutting for both defoliation strategies. When brass + Verano was cut, Verano yield was also hi-her with 6-weekly cutting. Where grass only was cut, Verano yields with Bowen were similar at the 2 cutting frequencies, whereas with Nixon, Verano yield was higher with 3-weekly cutting. Under 3-weekly cutting, yields of Verano and total yields were higher in association with Nixon for both defoliation strategies. In the first 4-week regrowth to measure residual effects, there was little difference between grasses in yield of grass or legume, but Nixon was twice as tall as Bowen. For the 8-week regrowth, Nixon swards had higher yields and % legume than Bowen swards. Yields following 6-weekly cutting were higher than following 3-weekly cutting and Verano yield and shoot numbers were higher with Nixon. Root weights were consistently lower following treatments where the grass + Verano was defoliated and with this treatment root yields tended to be lower after 3-weekly than after 6-weekly cutting. The chemical composition of the components suggested that Bowen was more competitive than Nixon in association with Verano for certain nutrients. Verano grown in association with Bowen was less productive than with Nixon, particularly under frequent defoliation. Nutrient competition, especially for S, may explain the lower Verano yield in association with Bowen. Cumulative yield for the 3-weekly, grass + legume cutting treatment over the 20 weeks of the experiment showed that Verano yield was higher (50%) with Nixon than with Bowen. Grass yield was similar for the 2 grasses. This result reflected the field result obtained under grazing at high stocking rate at Lansdown (Jones and Kerr 1993).