Soil chemical attributes in areas under conversion from forest to pasture in southern Brazilian Amazon

Abstract

The south of the Brazilian Amazon is one of the largest cattle-producing regions in Brazil, however, most of the pastures are in low fertility soils. Thus, cattle breeders compensate for the low production of pastures, increasing the size of the areas, generating more deforestation and burning. These practices increase the chemical degradation process of Amazonian soils, making them increasingly infertile when improperly managed. With this, the objective of the work was to evaluate the impacts caused in the chemical attributes of soils, in areas under forest-to-pasture conversion, in the south of the Brazilian Amazon. The study was carried out in the district of União Bandeirantes, in an area of forest and two areas with pastures (brachiaria and mombaça grass). In the field, soil samples were collected at two depths (0.00-0.10 and 0.10-0.20 m), to carry out chemical analyzes. Further, uni, bi and multivariate statistical analyzes were carried out, besides geostatistical analyzes were carried out to study spatial variability and management zones. The conversion of forest to pasture increased the pH and exchangeable bases levels, reducing the availability of exchangeable aluminum and potential acidity, however, it induces losses of phosphorus and organic carbon from the soil. Among the pasture environments, the mombaça grass area presented higher fertility. Greater spatial variability of chemical attributes was observed in the environment with mombaça grass, indicating greater heterogeneity in the distribution of attributes in the area. We attribute this behavior to the higher grazing intensity and the micro-reliefs in the area that direct the flow of water and nutrients.

Figure 1

Boxplot and mean test of soil acidity components in different environments, in Porto Velho, Rondônia. The green line indicates the mean value of the data. Different lower case letters represent that there were significant differences by the Tukey test at 5%.

Figure 2

Boxplot and mean test of exchangeable soil bases in different environments, in Porto Velho, Rondônia. The green line indicates the mean value of the data. Different lower case letters represent that there were significant differences by the Tukey test at 5%.

Figure 3

Boxplot and mean test of cation exchange capacity and saturation by soil bases in different environments, in Porto Velho, Rondônia. The green line indicates the mean value of the data. Different lowercase letters represent that there were significant differences by the Tukey test at 5%.

Figure 4

Boxplot and mean test for phosphorus and total organic carbon of the soil in different environments, in Porto Velho, Rondônia. The green line indicates the mean value of the data. Different lower case letters represent that there were significant differences by the Tukey test at 5%.

Figure 5

Factorial plane of the chemical attributes of the soil in forest and pasture areas, in Porto Velho, Rondônia.

Figure 6

Semivariograms of the chemical attributes of the soil in a forest area in Porto Velho, Rondônia. The values in parentheses are respectively: range, R², Degree of Spatial Dependence (DSD) and cross-validation.

Figure 7

Semivariograms of the chemical attributes of the soil in a brachiaria grass area in Porto Velho, Rondônia. The values in parentheses are respectively: range, R², Degree of Spatial Dependence (DSD) and cross-validation.

Figure 8

Semivariograms of the chemical attributes of the soil in a mombaça grass area in Porto Velho, Rondônia. The values in parentheses are respectively: range, R², Degree of Spatial Dependence (DSD) and cross-validation.

Figure 9

Kriging maps of soil chemical attributes in a forest area in Porto Velho, Rondônia.

Figure 10

Kriging maps of soil chemical attributes in a brachiaria grass area in Porto Velho, Rondônia.

Figure 11

Kriging maps of soil chemical attributes in a mombaça grass area in Porto Velho, Rondônia.

Figure 12

Scaled semivariograms for the environments studied in Porto Velho, Rondônia. The values in parentheses indicate respectively: range, degree of spatial dependence, R² and cross-validation.

Figure 13

Experimental semivariograms fitted to the principal component scores for the different environments studied in Porto Velho, Rondônia. The values in parentheses are, respectively: range, R2, degree of spatial dependence (DSD) and cross-validation.

Figure 14

Kriging maps of the principal component scores for the different environments studied in Porto Velho, Rondônia.

Figure 15

Location and elevation map of the study areas.