ABSTRACT Agroforestry systems have the potential to increase soil organic matter, with effects on soil carbon and nitrogen contents, but information on the application of these systems in semi-arid regions is still scarce. This study aimed to analyze soil carbon and nitrogen stocks in the conversion of native forest from the Caatinga Biome into integrated agriculture systems in the Brazilian semi-arid region. We evaluated the following management systems in the Haplic Inceptisol (Cambissolo Háplico eutrófico): (1) Intercropping area, cultivated with corn and Massai grass; (2) Caatinga (natural vegetation); (3) AFS10: agroforestry system with native woody forest rows occupying 33 % and agriculture occupying 66 % of the total area; and (4) AFS20: agroforestry system presenting inverse proportions of AFS10. The agroforestry systems were intercropped with sorghum or millet, pigeon pea and Massai grass. We collected disturbed and undisturbed soil samples at the layers of 0.00-0.10; 0.10-0.20; 0.20-0.40, 0.40-0.60 and 0.60-1.00 m for analysis of carbon (SOC), nitrogen (N), soil bulk density, and calculation of SOC and N stocks and C/N ratio, two years after the conversion of natural vegetation to the agricultural area (intercropping) and agroforestry system (AFS10 and AFS20). We applied principal component and cluster analysis to explore the data, and confidence interval to compare the means of accumulated SOC and N stocks up to 1 m soil depth. No differences exist for the properties analysed in superficial layers (0.00-0.10 and 0.10-0.20 m), regardless of land-use systems. AFS20 increase the SOC content and, consequently, SOC stock, in subsurface layers; on the other hand, intercropping increases N content and N stock. AFS20 presented higher accumulated SOC stocks up to 1.00 m (114.97 Mg ha-1). Agroforestry systems management is an alternative for increasing carbon sequestration under the conversion from Caatinga to agricultural areas.

RESUMO A produção de bananas requer estratégias que melhor identifiquem a variabilidade das áreas fertirrigadas, considerando que a produção pode ser reduzida mediante cultivo consecutivo em solos com restrições químicas e de diferentes altitudes. Objetivouse estudar a variabilidade espacial dos atributos químicos e estoque de carbono do solo e relacioná-los à altitude de um bananal fertirrigado por microaspersão. O estudo foi realizado em perímetro irrigado de Nupeba (Riachão das Neves/BA), em área cultivada há 18 anos com banana prata. Em abril de 2017, amostras simples de solo foram coletadas em área com altitude mínima de 440,6 m e máxima de 445,8 m, nas profundidades de 0-0,20 m e 0,20-0,40 m, com uma malha experimental composta por 40 pontos georreferenciados e espaçados a cada 10 metros. Foram determinados: pH, carbono orgânico total (COT), matéria orgânica do solo (OM), nitrogênio (N), fósforo (P), potássio (K+), cálcio (Ca2+), magnésio (Mg2+), alumínio (Al3+), acidez potencial (H+Al), soma de bases (SB), capacidade de troca catiônica (CEC), saturação por base (V%) e estoque de carbono (Est C). Os dados foram avaliados pelos métodos da estatística descritiva e geoestatística e apresentou distribuição normal. Houve variabilidade espacial, permitindo a elaboração de mapas de fertilidade. Os mapas apontaram que a altitude influencia principalmente a OM, COT, N e Est C, mostrando também a distribuição e a concentração dos nutrientes no solo, o que possibilita minimizar a variabilidade por meio da aplicação de fertilizantes em quantidades variadas, visando sua oferta homogênea às bananeiras.

ABSTRACT Banana production requires strategies to assess the variability of fertigation areas since production can be reduced by consecutive cultivation in chemically-restricted soils and at different elevations. From this perspective, this study aimed to investigate the spatial variability of soil chemical attributes and the soil carbon stock and relate these parameters to the elevation of a banana plantation fertigated by a micro-sprinkler system. The study was conducted at the Irrigation Perimeter Nupeba (Riachão das Neves/BA), in an area cultivated with the banana cultivar Prata for 18 years. In April 2017, single soil samples were collected in an area with a minimum elevation of 440.6 m and a maximum of 445.8 m, at the depths of 0-0.20 m and 0.20-0.40 m, with a sampling grid composed of 40 georeferenced points and spaced 10 meters. The following parameters were determined: pH, total organic carbon (COT), soil organic matter (OM), nitrogen (N), phosphorus (P), potassium (K+), calcium (Ca2+), magnesium (Mg2+), aluminum (Al3+), potential acidity (H+Al), sum of bases (SB), cation exchange capacity (CEC), base saturation (V%), and carbon stock (Est C). The data were evaluated by descriptive statistics and geostatistics and showed normal distribution. There was spatial variability, allowing the construction of fertility maps. The maps showed that the elevation especially influences the OM, COT, N, and Est C, also showing the distribution and concentration of nutrients in the soil, allowing to minimize the variability through the application of varied nutrient contents aiming at their homogeneous supply to banana plants.