Reassessing the use of loss‐on‐ignition for estimating organic matter content in noncalcareous soils

Publisher Website
Google Scholar
Abstract
Surface samples were collected from soils used for cropland, pasture, and woodland and formed in three parent materials in northwestern Washington (USA) and British Columbia (Canada). These samples were analyzed for organic matter content by loss‐on‐ignition and by automated train analysis using a high‐frequency induction furnace (Leco carbon analyzer). The coefficient of determination between loss‐on‐ignition and train analysis for all samples combined was 0.98 indicating that loss‐on‐ignition may be a useful estimator of organic matter in noncalcareous soils. Although loss‐on‐ignition (LOI) has been dismissed widely as a crude and inadequate estimator of organic matter in noncalcareous soils6,9, it has been used in many ecological studies for the analysis of organic layers3,4. It is easy and rapid; 40 samples can be analyzed in about 2 hours. In soil studies, LOI overestimates the amount of organic matter since heating to temperatures above 150° C will drive off hygroscopic H2O and intercrystalline H2O from crystalline clays and allophane . Other positive errors are due to CO2 release from carbonates in calcareous soils and the loss of H2O from hydroxyl groups in sesquioxides. Therefore errors are dependent on soil properties, such as the amount and type of clay and the amount of carbonates and sesquioxides. Sometimes the variability in the organic matter content within the soil may hide these differences. In a study of soils with a range of organic matter and clay contents formed in different parent materials, Ball1 found excellent correlation between LOI and organic C. He determined LOI at 850°C and at 375°C and found a correlation coefficient for the 850° determination of 0.99 between LOI and organic C as determined by titration (Walkley‐Black method). The accuracy is greater for the 375° ignition, although no correlation was given. Parent material and clay mineralogy did not significantly affect the regression models. The purpose of this study was to determine the accuracy of the loss‐on‐ignition method for estimating organic matter content in noncalcareous soils in British Columbia and Washington in comparison to the automated train analysis using a high‐frequency induction furnace (Leco carbon analyzer).