A field trial was carried out at the Estação Experimental Agronômica/UFRGS (30º05'52"S; 51º40'08"W) to evaluate the effect of irrigation (with and withoutirrigation), two clipping dates (E1- removing about 50% of the apical meristems of the primary tillers; E2- removing 75%), and four rates of nitrogen (0, 50, 100, and 150 kg.ha-1) on seed yield of pearl-millet (Pennisetum americanum (L.) Leeke). The experiment was arranged in a randomized split-split-plot design with four replications. Seeding was performed by direct drilling, on December 29, 2000. Nitrogen rates were equally split and broadcast, half in January 16, 2001 and half one week after clipping. The clippings were done at 20cm height, (E1=02/09 and E2=02/19) and the seed was harvested in April 30 and May 01, 2001. Nitrogen applications influenced seed yield components such as: total number of panicles.m-2, number of panicles with seeds.m-2, length of panicles, weight of panicles and number of seeds.panicle-1. The component panicles with seeds were affected by the interactions between irrigation and nitrogen rates, and clipping and nitrogen rates. Higher weights of panicles and thousand seeds were registered in early clipping (E1). The viable pure seed yield at the maximum technical efficiency of N (120kg.ha-1) was 769kg.ha-1. The seed yield components that correlated significantly with seed yield were number of seeds.panicle-1 and weight of one thousand seeds. The total dry matter yield was influenced by the nitrogen rates, which was described by a positive linear regression. Irrigation did not affect any of the studied variables.
Pennisetum americanum; yield components; annual pasture
