The profiles of the three samples of each treatment revealed great similarity. The analyses of the structure of the bacterial communities (Figure 2) showed that these were significantly impacted by both the use (cultivation of sugarcane) and the management (burnt versus
green cane) of the soil, according to pairwise comparisons (MRPP analysis; p < 0.03). The ordering generated by the NMS grouped the replicates of each treatment in a distinct region, and the three treatments (centroids) practically equidistant from Gefitinib each other. The sensitivity of soil bacterial communities to changes in land use and management has already been shown by different authors in various settings [11, 54–56], including DGGE analyses SB203580 solubility dmso carried out in Brazilian Cerrado soils [20]. Figure 2 NMS ordination
of the DGGE profiles of 16S rRNA gene fragments (total bacteria) amplified from the soil samples (0–10 cm) collected from the treatments Control (C), Green cane (GC) and Burnt cane (BC). The fraction of total variance that accounts for each axis is indicated in parentheses. The angles and the length of radiating lines indicate the direction and strength of the relationship between the chemical and biological variables with the ordination scores. Several factors correlated with the NMS ordination. In particular, the total P and exchangeable Mg contents and soil density were associated with the bacterial community structures in the control soil, while the (reduced) C and N contents were correlated with the bacterial communities in the green cane treatment. Finally, the (decreased) value of the sum of bases (SB), the degree of saturation of the bases (V), the cation exchange capacity (CEC) and exchangeable calcium (Ca) were correlated with the communities from the burnt cane treatment (Figure 2). The soil properties that correlated with the segregation of the bacterial community structures were consistent with observations from Atlantic
forest soils under different agricultural production systems [11, 17, 20]. The amoA gene based DGGE (ammonia oxidizing bacteria) showed relatively simple profiles in all treatments (4–10 bands), with relatively similar patterns between the triplicates. The control soil revealed a higher number of bands in comparison to the green and burnt cane soils. The analysis of these communities indicated Phosphatidylinositol diacylglycerol-lyase a diffuse distribution, with some within-treatment variability (Figure 3). However, as reflected in the X axis, these communities responded significantly to the change in land use management (MRPP < 0.05), being the burn treatment a factor that exacerbated the response. Figure 3 NMS ordination of the DGGE profiles of amoA gene fragments (ammonia oxidizing bacteria) amplified from the soil samples (0–10 cm) collected from the treatments Control (C), Green cane (GC) and Burnt cane (BC). The fraction of total variance that accounts for each axis is indicated in parentheses.