The diversity of methoxyphenols released by pyrolysis-gas chromatography as predictor of soil carbon storage.

The variable extent to which environmental factors are involved in soil carbon storage is currently a subject of controversy. In fact, justifying why some soils accumulate more organic matter than others is not trivial. Some abiotic factors such as organo-mineral associations have classically been invoked as the main drivers for soil C stabilization. However, in this research indirect evidences based on correlations between soil C storage and compositional descriptors of the soil organic matter are presented. It is assumed that the intrinsic structure of soil organic matter should have a bearing in the soil carbon storage. This is examined here by focusing on the methoxyphenols released by direct pyrolysis from a wide variety of topsoil samples from continental Mediterranean ecosystems from Spain with different properties and carbon content. Methoxyphenols are typical signature compounds presumptively informing on the occurrence and degree of alteration of lignin in soils. The methoxyphenol assemblages (12 major guaiacyl- and syringyl-type compounds) were analyzed by pyrolysis-gas chromatography-mass spectrometry. The Shannon-Wiener diversity index was chosen to describe the complexity of this phenolic signature. A series of exploratory statistical analyses (simple regression, partial least squares regression, multidimensional scaling) were applied to analyze the relationships existing between chemical and spectroscopic characteristics and the carbon content in the soils. These treatments coincided in pointing out that significant correlations exist between the progressive molecular diversity of the methoxyphenol assemblages and the concentration of organic carbon stored in the corresponding soils. This potential of the diversity in the phenolic signature as a surrogate index of the carbon storage in soils is tentatively interpreted as the accumulation of plant macromolecules altered into microbially reworked structures not readily recognized by soil enzymes. From a quantitative viewpoint, the partial least squares regression models exclusively based on total abundances of the 12 major methoxyphenols were especially successful in forecasting soil carbon storage.