Pastures store approximately 30% of the global terrestrial carbon (C) and are vital for the provision of several ecosystem services such as forage and climate regulation. In Brazil, sustainable management systems are being proposed to minimize environmental damage by carbon dioxide (CO₂) emissions and increase soil C and nitrogen (N) stocks under pasture. Simulation models are important tools in the long-term assessment of C and N stock changes in pastures under different management practices. In this study, the Century model, version 4.0, was employed to simulate C and N stocks (0–20 cm layer) as well as forage biomass production in areas of pastures with different ages and different management practices. Carbon and N stocks simulated in the soil organic matter (SOM) equilibrium state under native vegetation were used in the adjustment of the deforestation model and subsequent establishment of pasture. Eight production scenarios for 2050 were simulated in comparison to the current situation depicted by absence of management practices (M), namely: pasture managed without periodic renewal (M1); pasture with chemical control of spontaneous plants (M2); pasture with only initial fertilization (M3); pasture with initial and maintenance fertilization (M4); pasture with initial and maintenance fertilization associated with chemical control of spontaneous plants (M5); silvopastoral system with 30% tree cover (M6); pasture managed with fire (M7); and pasture under intensive management (M8). Simulated scenarios revealed significant increases in C and N stocks, besides similar biomass production. Chemical control of spontaneous plants, initial fertilization at the time of pasture planting, and silvopastoral systems proved to be alternatives to maximize C sequestration. Simulated values of C stocks are strongly related to clay contents (SS2-18.2 t ha⁻¹, SS3-60.0 t ha⁻¹ and SS4-29.7 t ha⁻¹). Similarly, N stocks also were higher in these areas (SS2-1.7 t ha⁻¹, SS3-1.8 t ha⁻¹ and SS4-2.1 t ha⁻¹). Additionally, considering different scenarios, forage biomass production is conditioned by adopting management practices, such as control of spontaneous plants and initial fertilization. These results can assist national initiatives associated with management strategies which enhance pasture sustainability in Brazil. Modelling allows supply of information for the gap of data and appoint necessities of studies, especially for the Amazon region where Cerrado-forest transition areas are under strong pressure for agricultural occupation.

牧场储存了大约 30% 的全球陆地碳 (C)，对于提供多种生态系统服务（如饲料和气候调节）至关重要。在巴西，正在提出可持续管理系统，以尽量减少二氧化碳（CO ₂) 排放并增加牧场下的土壤碳和氮 (N) 库。模拟模型是长期评估不同管理实践下牧场碳和氮库变化的重要工具。在本研究中，Century 模型 4.0 版用于模拟 C 和 N 库（0-20 cm 层）以及不同年龄和不同管理实践的牧场地区的草料生物量生产。模拟在原生植被下土壤有机质 (SOM) 平衡状态下的碳和 N 库用于调整毁林模型和随后的牧场建立。与缺乏管理实践（M）所描述的当前情况相比，模拟了 2050 年的八个生产情景，即：牧场管理不定期更新（M1）；对自发植物进行化学控制的牧场（M2）；仅初始施肥（M3）的牧场；初始施肥和维持施肥（M4）的牧场；与自发植物的化学控制相关的初始和维持施肥的牧场（M5）；具有 30% 树木覆盖率的林牧系统 (M6)；用火管理的牧场（M7）；和集约化管理的牧场（M8）。模拟情景显示，除了类似的生物量生产外，C 和 N 库显着增加。自发植物的化学控制、牧草种植时的初始施肥和林牧系统被证明是最大化碳封存的替代方法。C 库的模拟值与粘土含量密切相关 (SS2-18.2 t ha 初始施肥和维持施肥（M4）的牧场；与自发植物的化学控制相关的初始和维持施肥的牧场（M5）；具有 30% 树木覆盖率的林牧系统 (M6)；用火管理的牧场（M7）；和集约化管理的牧场（M8）。模拟情景显示，除了类似的生物量生产外，C 和 N 库显着增加。自发植物的化学控制、牧草种植时的初始施肥和林牧系统被证明是最大化碳封存的替代方法。C 库的模拟值与粘土含量密切相关 (SS2-18.2 t ha 初始施肥和维持施肥（M4）的牧场；与自发植物的化学控制相关的初始和维持施肥的牧场（M5）；具有 30% 树木覆盖率的林牧系统 (M6)；用火管理的牧场（M7）；和集约化管理的牧场（M8）。模拟情景显示，除了类似的生物量生产外，C 和 N 库显着增加。自发植物的化学控制、牧草种植时的初始施肥和林牧系统被证明是最大化碳封存的替代方法。C 库的模拟值与粘土含量密切相关 (SS2-18.2 t ha 用火管理的牧场（M7）；和集约化管理的牧场（M8）。模拟情景显示，除了类似的生物量生产外，C 和 N 库显着增加。自发植物的化学控制、牧草种植时的初始施肥和林牧系统被证明是最大化碳封存的替代方法。C 库的模拟值与粘土含量密切相关 (SS2-18.2 t ha 用火管理的牧场（M7）；和集约化管理的牧场（M8）。模拟情景显示，除了类似的生物量生产外，C 和 N 库显着增加。自发植物的化学控制、牧草种植时的初始施肥和林牧系统被证明是最大化碳封存的替代方法。C 库的模拟值与粘土含量密切相关 (SS2-18.2 t ha^-1 , SS3-60.0 吨公顷^-1和 SS4-29.7 吨公顷^-1 )。类似地，这些地区的 N 库也较高（SS2-1.7 t ha ^-1、SS3-1.8 t ha ^-1和 SS4-2.1 t ha ^-1）。此外，考虑到不同的情况，牧草生物质的生产取决于采用管理实践，例如控制自发植物和初始施肥。这些结果可以帮助与提高巴西牧场可持续性的管理战略相关的国家举措。建模允许为数据差距提供信息并指定研究的必要性，特别是对于亚马逊地区，那里的塞拉多森林过渡区面临着农业占领的巨大压力。