The primary sources of carbon loss during the crop-establishment period in a subtropical Oxisol under contrasting tillage systems

Abstract

The physical protection of mineralizable carbon (C) in aggregates has been identified as the primary mechanism of soil C stabilization. Therefore, it is possible to hypothesize that the disruption of aggregate by soil tillage is a key process driving C losses during the crop-establishment period. However, these findings are based on studies performed in temperate soils. Limited information is available for studies performed in subtropical and tropical soils, especially in Oxisols, which are rich in oxides that provides chemical C stabilization. This study was performed in southern Brazil in a long-term soil-management experiment carried out in a clay Typic Haplorthox in Cruz Alta (RS). During the 22nd year of the experiment, carbon dioxide (CO2-C) emissions, temperature, and soil moisture were intensively evaluated over a 21-day summer crop-establishment period using a closed infrared CO2-flux chamber. The cropping system investigated was an intensive crop rotation following the soil input of winter-cover crops (black oat (Avena strigosa Schreb) + common vetch (Vicia sativa L) under two contrasting tillage systems, conventional tillage (CT) and no-till (NT). The apparent contributions to CO2-C losses by resident soil C associated with aggregate disruption and recent crop-residue C input were assessed in treatments with crop-residue input (+R) and with crop-residue removed (-R). An exponential-decay model was used to fit the differences in CO2-C flux between CT - R and NT - R (apparent aggregate-disruption effect) and between CT + R and CT - R (apparent recent crop-residue C input effect). As expected, the CT + R showed an increase of 72% in CO2-C losses relative to NT + R. During the three-week crop-establishment period, crop-residue C input was the primary source of CO2-C emissions under CT. The CO2-C losses under CT were equivalent to 65% of the aboveground C input by winter cover crops, whereas this value decreased to 35% in NT. Exponential-decay modeling of the data for the first week showed that approximately 20% of the CO2-C losses under CT were related to the exposure of mineralizable resident soil C due by tillage operations. The analysis showed that this value decreased to only 2% for the three-week period. The CO2-C emissions exhibited a positive linear relationship with soil temperature and soil water-filled porosity under NT, but a similar relationship was found only with soil temperature under CT. For this Oxisol during the crop-establishment period, the physical aggregate disruption induced by long-term CT played a secondary role in CO2-C losses relative to the recent crop-residue C input from tillage operations. (C) 2011 Elsevier B.V. All rights reserved.