Agrosilvipastoral systems represent agro-forestry practices that allow greater diversity and sustainability of a cropping system. Soil organic matter (SOM) is one of the main indicators of the sustainability of a cropping system, especially in tropical soils. The objectives of this study were to investigate the impact of eucalypt cultivation on different C and N of SOM fractions on agrosilvipastoral systems over time. To this end, a chronosequence consisting of six implantation times of an agrosilvipastoral system was sampled. Time zero was represented by the adjacent native vegetation (Cerrado), followed by 2, 3, 4, 7 and 10 year periods since the agrosilvipastoral system implantation in previously Cerrado land. The layers 0-10, 10-20, 20-40, 40-60, and 60-100 cm were sampled in each area, within and in-between eucalypt rows. The samples were chemically and physically analyzed to determine: total organic carbon (TOC), total nitrogen (TN), C in the light organic matter (LOM), N in the LOM, and C oxidizable by potassium permanganate, the labile carbon (LC). From these data, the CMI proposed by Blair et al. (1995) was calculated (CMI LC), as well as the modified CMI, which is based on C in the LOM (CMI LOM). In the planting row, the TOC stocks in the 0-100 cm layer remained stable over time, while in-between the rows the TOC stocks initially decreased and then recover over time, but have yet to return to the level found in the native vegetation soil. The trend for the TN stock (0-100 cm) was similar for both in and between the eucalypt rows. After a sharp initial decrease, the TN stocks recovered over the years. However, this increase was not sustained and at the end of the cycle the TN stocks were lower than the original values found for the soil under native vegetation. Major changes were observed in the free LOM fractions, confirming its great sensitivity to management changes. More free LOM C and N were lost in-between than in the rows. Labile C, as well as the CMI, were found not sensitive to describe SOM changes along the implantation time, whereas the CMI CLOM adjusted well for all soil layers. In addition, it was best correlated with TOC and is therefore more adequate to compare the studied cropping system. These results show the importance of studying deeper soil layers; since the loss of C at greater soil depths can be relatively high and difficult to recover. The agrosilvipastoral systems induce a decrease in soil organic C in the upper layers and the time to recover the original stocks was beyond the study period. The fraction of free light organic matter was most sensitive to the management.
carbon balance; carbon stock; light organic matter; nitrogen; labile carbon; C management índex
