Dynamics of surface liming in a typic haplorthox

Alterations in soil chemical properties after liming depend on the frequency of soil sampling and rainfall. This study evaluated such alterations in a Typic Haplorthox (Rhodic Ferralsol) with a sandy-loam superficial layer. The experiment had a completely randomized block design, with four dolomitic lime rates (0, 2.4, 4.8, and 7.2 t ha(-1)) and three replicates. The experiment was carried out in Alto Parana in northwestern Parana state (southern Brazil). Lime was broadcast on the soil surface over Brachiaria decumbens grass residues in-between rows of a young orange plantation. The experimental plots consisted of an area of 310 m(2) with 15 orange plants each (3 rows and 5 plants per row). Soil samples were collected in-between the rows of 3 central orange plants at depths of 0-0.05; 0.05-0.10; 0.10-0.20; 0.20-0.40; and 0.40-0.60 meter at 4,12,18, 24, 31, and 37 months after liming. This sampling times corresponded to alternated rainy and dry seasons of the region. The analysis of variance was performed based on a complete randomized block design in a split-split-plot scheme with repeated measures over time (years, rainfall season and lime rates), which were used in the regression analyses through orthogonal polynomial models. Results showed that maximum alterations in the soil chemical properties were obtained with lime rates twice as high as those calculated to raise the soil base saturation to 70% at a depth of 0.20 m. Soil sampling after the high rainfall period proved important to adjust the base saturation data in the 0-0.05 m layer. Ca concentration only varied in the 0-0.10 m layer, while those of Mg varied progressively from the first to the third year in the 0-0.60 m layer. Mg concentrations were associated to the soil acidity correction (pH and H + Al) and to the increment of the base saturation values in the 0.10-0.20 m layer, where it was found the lowest Ca/Mg ratio (1.06). The most pronounced response to surface liming was observed in the third year after surface liming (0-0.10 m deep) identified by quadratic regression models. At the same time linear regression models identified the establishment of a alkalinization front in the 0.10-0.20 m soil layer. The methodology of split-split-plot design with repeated measures over time allowed the description of the superficial lime dynamics along the time.