Particulate Organic Matter at the Field Scale: Rapid Acquisition Using Mid-Infrared Spectroscopy

• Published in: Soil Science Society of America journal Vol. 74; no. 4; pp. 1147 - 1156
• Main Authors: Bornemann, L, Welp, G, Amelung, W
• Format: Journal Article
• Language: English
• Published: Madison Soil Science Society 01.07.2010; Soil Science Society of America; American Society of Agronomy
• Subjects: Acetal resins; Agronomy. Soil science and plant productions; Assessments; biogeochemical cycles; Biological and medical sciences; Calibration; Carbon; Cellulose; chemical structure; Chemical, physicochemical, biochemical and biological properties; Correlation analysis; Crop rotation; Degradation products; Earth sciences; Earth, ocean, space; Exact sciences and technology; Fractionation; Fundamental and applied biological sciences. Psychology; Heterogeneity; Infrared radiation; Infrared spectroscopy; least squares; lignin; Mathematical models; model validation; new methods; Organic matter; particle size; Particulate organic matter; Physics, chemistry, biochemistry and biology of agricultural and forest soils; Pools; Prediction models; regression analysis; simulation models; soil organic carbon; soil organic matter; Soil science; Soils; spatial variation; spectral analysis; Spectroscopy; Spectrum analysis; Statistics; Studies; Surficial geology; temporal variation; electromagnetic waves; physical methods; organic materials; Particulate organic matter; Soil science; utilization; Spatial scale; Mid infrared radiation; Earth science; Acquisition; optical methods; infrared spectroscopy
• ISSN: 0361-5995; 1435-0661
• DOI: 10.2136/sssaj2009.0195

Modeling global C cycles requires in-depth knowledge about small-scale C stocks and turnover processes, yet different soil organic C (SOC) pools reveal considerable spatiotemporal heterogeneity at the field scale, which is scarcely known due to the considerable workload associated with traditional fractionation procedures. We investigated the potential of mid-infrared spectroscopy combined with partial least squares regression (MIRS-PLSR) for rapid assessment of different particulate organic matter (POM) pools and their spatial heterogeneity at the field scale. Locally calibrated prediction models estimated the contents of SOC, POM of three size classes (POM1: 2000–250 µm; POM2: 250–53 µm; and POM3: 53–20 µm), and lignin contents for 129 locations in a 1.3-ha test field. Relations between the parameters were described using correlation analysis and fuzzy- statistics. All parameters were predicted successfully by applying local calibrations for MIRS-PLSR (R2 = 0.77–0.96). The prediction model for POM1 chiefly relied on specific signals of lignin and cellulose; contents of POM2 were estimated by spectral bands assigned to degradation products as aliphatic C–H groups and aromatic moieties; carboxylic groups essentially contributed to the prediction of POM3. There was a close spatial relation between the coarse POM1 and POM2 fractions and lignin ( = 0.77), which largely also explained variations in bulk SOC. In contrast, POM3 exhibited a less deterministic pattern in the field, thus contributing little to the spatial variation in SOC content.