The objective of this work was to evaluate nutrient cycling by cover crops and its influence on the performance of the rotation between upland rice and soybean. The crops were sown under no-tillage, in a Rhodic Haplustox, in the Cerrado region of the state of Goiás, Brazil. The cover crops were sown mechanically, also under no-tillage, after the harvest of soybean (3/25/2008) and rice (4/7/2009). A randomized complete block design was used, split-plotted in time, with four replicates. The treatments consisted of the following cover crop species, besides fallow: Urochloa ruziziensis, U. brizantha, Pennisetum glaucum and U. ruziziensis + Cajanus cajan. The evaluation of the remaining nutrient in straw was performed at 0, 15, 30, 60, 90, and 120 days from the desiccation date of the cover crops. Urochloa ruziziensis and U. brizantha were the most efficient in the accumulation and release of nutrients, especially potassium. Urochloa ruziziensis is the most suited species for use as a cover crop previous to upland rice cultivation under no-tillage. However, none of the cover crop species significantly affect soybean grain yield.
Brachiaria brizantha; Brachiaria ruziziensis; Pennisetum glaucum; upland rice; crop rotation; crop succession
