Influence of Soil Texture and Management Practices on Soil Organic Carbon Stocks.
S.K. Gami, J.M. Duxbury, and J.G. Lauren. Cornell Univ, 915 Bradfield Hall, Ithaca, NY 14853
It is well known that substantial Soil Organic Carbon (SOC) losses occur from cultivated agricultural soils and that SOC levels are influenced by soil texture. In general, fine-textured soils maintain higher levels of SOC than coarse-textured soils when other factors are constant. Quantitative relationships between SOC stocks to a depth of 60 cm and texture (silt+clay) were established for native and cultivated land in Nepal and Bangladesh. A total of 100 soil samples from native ecosystem sites and 30 from cultivated sites were collected in 15 cm depth increments to 60 cm. Surface layer (0-15 cm) samples were collected from an additional 100 cultivated sites. Most of the samples were collected from forestland (native) and from rice-wheat fields (cultivated). Regression analyses showed linear increases in SOC with increasing silt+clay content for all depths for both native and cultivated soils. Overall, cultivation of forests resulted in a 36% and 31% reduction in SOC stocks in the 0-15 and 15-30 cm depths, respectively. There was no significant loss of SOC below 30 cm. Long-Term oil Fertility Experiments (LTFEs) provide an opportunity to understand the effects of nutrient and organic inputs on SOC stocks under conventional tillage. Three LTFEs (25 years old) located at Bhairahawa, Parwanipur and Tarhara in the Nepal terai were studied. The sites differed somewhat in soil texture and were cropped annually to rice-rice-wheat or rice-wheat. Three treatments; control (unfertilized), recommended NPK and Farm Yard Manure (FYM; 4 t dry wt./crop) were selected at all sites. Two additional treatments; FYM (4 t ha-1 yr-1) plus half recommended NPK, and chopped wheat straw (5 t ha-1 yr-1) plus half recommended NPK were also included at Parwanipur. An unfertilized and harvested grassland adjacent to 2 of the LTFEs was also sampled. The grassland was established at the initiation of the LTFE's and can be considered to represent a no-tillage control treatment. Soil samples were collected at 15 cm intervals to a depth of 60 cm. In general, SOC stocks were highest in the FYM treatment and the no-tillage grass control and lowest in the cultivated control and NPK treatments. Treatments with residue return and reduced FYM inputs had intermediate SOC stocks. Again, differences in SOC stocks were only found in the top 30 cm of soil. Soil organic carbon stocks were highest at the Bhairahawa site, which had the highest silt+clay content. At this site, the SOC stocks to a depth of 60 cm were 41.6, 43.3 and 59.1 t ha-1 for the control, NPK and FYM treatments, respectively. The SOC stock in the no-tillage grass control was 57.1 t ha-1, similar to that in the FYM treatment. The predicted SOC stock for the native forest ecosystem at the Bhairahawa site was 67.4 t ha-1. Thus, the SOC stock in the NPK treatment was 24.1 ha-1, or 36%, less than that predicted for native forest. Similar trends in SOC stocks were found at the other two sites. The results show 1) essentially no benefit of NPK fertilization on increasing SOC stocks when crop residues are removed; 2) that no-tillage provides a SOC benefit equivalent to adding 4 t ha-1 dry wt (10 t ha-1 fresh wt) of FYM/crop; 3) return of about 50-60% of the normal production of crop residues increases SOC stocks less than simply stopping tillage; and 4) that there is still potential to increase SOC stocks above the level achieved in the no-tillage control. Trends in animal populations and FYM availability in the study areas and throughout the Indo-Gangetic plains make it impractical to consider using FYM to increase SOC stocks. Shifting to no-tillage agriculture presents the best option for increasing SOC stocks.