Five surface soils of differing chemical and mineralogical compositions were subjected to either a sequence of dithionitelcitrate extractions in which the soil: citrate ratio was varied or to a sequence of 1% or 2% HF extractions. The 2% HF treatment resulted in the removal of the highest Fe concentrations (79-95%) while the dithionitelcitrate extractions were less effective in removing Fe from the same soils (18-75%). The Fe remaining after HF treatment appeared to be mostly associated with ilmenite crystals which were only slowly attacked by the dilute acid. During the 2% HF treatments, some organic carbon was lost (8-17%). but this loss had no significant effect on the organic chemistry of the samples as determined by CP/MAS 13C n.m.r. spectroscopy.The total 13c signal recovered after the various treatments was found to be correlated, in order of decreasing significance, with the mineral fraction present in the sample, the organic carbon/Fe ratio and the mass magnetic susceptibility. The expression (organic carbon/Fe) -0.147(mineral fraction present in the sample) +0.043(l/mass magnetic susceptibility), accounted for 85.3% of the variation in the relative visibility of the 13C signal.Prior to solid state CP/MAS 13c n.m.r. analysis, the recommended pretreatment for surface soils containing Fe involves a sequence of 2% HF extractions in the order 5x1 h, 2x16 h and 1x64 h. For soils high in Fe, this procedure allows CP/MAS 13c n.m.r. spectra to be acquired that would otherwise be difficult to obtain. It also results in a significant increase in sensitivity and in resolution of the 13c n.m.r. spectra of soils with moderate Fe contents.

A variant of solid-state 13C CPMAS NMR demonstrated the presence of carbon in spatially distinct domains within three de-ashed samples of humin from a forest soil profile under Douglas-fir near Victoria, British Columbia. Differences in 1H T1 values indicated incomplete spin diffusion in the solid state and separation of organic matter in domains with dimensions of tens of nanometres or more. These relaxation differences were used to separate subspectra of carbons associated with the slower- and faster-relaxing protons. The former were dominated by methylene chains, while the latter contained more aromatic and O-alkyl carbon. Most of the organic matter appears to remain associated with iron even after extensive de-ashing with HCl/HF, but the aromatic and carbohydrate structures appear to be more closely associated with iron than methylene groups in long chains. This NMR technique offers a novel approach to probing the structure of insoluble soil organic matter. Key words: Humin, CPMAS NMR, relaxation, iron, forest soil, Brunisol

The effects of increasing cropping and soil compaction on aggregate stability and dry-sieved aggregate-size distribution, and their relationship to total organic C (TOC) and the major functional groups of soil organic carbon, were investigated on 5 soils of contrasting mineralogy. All soils except the allophanic soil showed a significant decline in aggregate stability under medium- to long-term cropping. Mica-rich, fine-textured mineral and humic soils showed the greatest increase in the mean weight diameter (MWD) of dry aggregates, while the oxide-rich soils, and particularly the allophanic soils, showed only a slight increase in the MWD after long-term cropping. On conversion back to pasture, the aggregate stability of the mica-rich soils increased and the MWD of the aggregate-size distribution decreased, with the humic soil showing the greatest recovery. Aggregate stability and dry aggregate-size distribution patterns show that soil resistance to structural degradation and soil resilience increased from fine-textured to coarse-textured to humic mica-rich soils to oxide-rich soils to allophanic soils. Coarse- and fine-textured mica-rich and oxide-rich soils under pasture contained medium amounts of TOC, hot-water soluble carbohydrate (WSC), and acid hydrolysable carbohydrate (AHC), all of which declined significantly under cropping. The rate of decline varied with soil type in the initial years of cropping, but was similar under medium- and long-term cropping. TOC was high in the humic mica-rich and allophanic soils, and levels did not decline appreciably under medium- and long-term cropping. 13C-nuclear magnetic resonance evidence also indicates that all major functional groups of soil organic carbon declined under cropping, with O-alkyl C and alkyl C showing the fastest and slowest rate of decline, respectively. On conversion back to pasture, both WSC and AHC returned to levels originally present under long-term pasture. TOC recovered to original pasture levels in the humic soil, but recovered only to 60–70% of original levels in the coarse- and fine-textured soils. Aggregate stability was strongly correlated to TOC, WSC, and AHC (P < 0.001), while aggregate-size distribution was moderately correlated to aggregate stability (P < 0.01) and weakly correlated to AHC (P < 0.05). Scanning electron microscopy indicated a loss of the binding agents around aggregates under cropping. The effect of the loss of these binding agents on soil structure was more pronounced in mica-rich soils than in oxide-rich and allophanic soils. The very high aggregate stabilities of the humic soil under pasture was attributed to the presence of a protective water-repellent lattice of long-chain polymethylene compounds around the soil aggregates.

No abstract

No abstract