Dinâmica da matéria orgânica do solo em áreas de integração lavoura-pecuária sob sistema plantio direto

Abstract

The cattle breeding and crop production integrated systems in the south of Brazil is based mostly on grain production with summer crops and beef cattle production over winter pastures. This activity have been expanded quickly in the state of Rio Grande do Sul, due the increase of soybean cultivation over native grass fields areas and the requirement to improve the livestock productivity index in that state. However, few research works exists until this moment to show the best management systems to this areas mainly due to its distinct character of agriculture and livestock when are isolated activities. With this meaning, this work has the purpose to evaluate the effect in increase use intensity of winter pastures and different summer crops systems over the soil organic matter (SOM) dynamics and its granulometric fractions particulate and mineral associated SOM on this areas and the grain and beef cattle yield achieved. For that, a field experiment was carried out over a Palehudult soil, on the city of Jari RS, with complete randomized blocks experimental design on factorial treatment distribution, four replications, with the utilization of three winter pastures management systems: no grazing (NG), grazing even 28 days (G28) and grazing even 14 days (G14), composing in this way increasing grazing frequencies over that pastures; and three summer crops systems: soybean monoculture (Mon-S), corn monoculture (Mon-C) and soybean/corn rotation (Rot-S/C). The beef cattle yield during the winter was not significantly affected by the increase of the grazing frequencies, but the biomass production to soil cover by the pastures and the grain yield of summer crops was reduced when the use intensity of the winter pastures by the animals was increased. The C addition to the soil by vegetal residues varied between 2.92 and 7.59 Mg ha-1 year-1 to the treatments G14 Mon-S and NG Mon-M, respectively. So, the main SOM fraction affected by the employed management in each treatment was the particulate fraction, expressed by the POM-C pool. This pool showed be 49 times more sensitive to the applied management over the winter pastures than the mineral associated fraction, expresses by the AMOM-C pools. The application of the Hennín & Dupuis s mathematical model to the prediction of each pools of SOM showed higher turnover rates of POM-C, in relation to the SOC and AMOM-C pools. The k1 coefficient (humification rate of the C additioned to the soil) obtained by the mathematical model was 0.1215, 0.0936 and 0.1249 year-1, to POM-C, AMOM-C and SOC, respectively, while the k2 coefficient (mineralization rate of de SOC) was 0.1292, 0.0161 and 0.0185 year-1, to the same pools. Using this data, the POM-C and SOC simulated pools on stability were 2.75 and 19.73 Mg ha-1, respectively, to the treatment G14 Mon-S, and 7.13 and 51.21 Mg ha-1, respectively, to the treatment NG Mon-M. The t½ (half-life time) for POM-C, AMOM-C and SOC were 5, 43 and 37 years, respectively, and the MRT (mean residence time) were 8, 62 and 54 years, to the same pools, respectively. The treatments that not received grazing during the winter showed good potential to the atmospheric CO2 sequestration, while the increase of the grazing intensity reduced this potential, passing to actuate like a sink of atmospheric CO2 when the G28 and G14 frequencies had been applied. The POM-N pools had been equally more sensitive to the grazing and crop management applied than the AMOM-N or TN pools, being that the higher pools had been observed when soybean was utilized on monoculture over the areas that not received grazing and the lowers pools had been observed when the corn was cultivated on monoculture over areas that received grazing even 14 days during the winter.