| Couteaux, M-M; Berg, B; Rovira, P: Near infrared reflectance spectroscopy for determination of organic matter fractions including microbial biomass in coniferous forest soils, Soil Biology Biochemistry, 35, 1587-1600 (2003) | |
| Abstract: The objective of this work was to investigate the usefulness of near infrared reflectance spectroscopy (NIRS) in determining some C and N fractions of soils: labile compounds, microbial biomass, compounds derived from added 13C- and 15N-labelled straw. Soil samples were obtained from a previous experiment where soils were labelled by addition of 13C- and 15N-labelled wheat straw and incubated in coniferous forests in northern Sweden (64–60°N) and south France (43°N). The incubation lasted three years with 7–9 samplings at regular time steps and four replicates at each sampling (204 samples). Samples were scanned using a near infrared reflectance spectrophotometer (NIRSystem 6500). Calibrations were obtained by using a modified partial least squares regression technique with reference data on total C and N, 13C, 15N, control extract-C, -N, -13C and -15N, fumigated extract-C, -N, -13C and -15N, biomass-C, -N, -13C and -15N contents. Mathematical treatments of the absorbance data were first or second derivative with a gap from 4 to 10 nm. The standard error of calibration (SEC)-to-standard deviation of the reference measurements ratio was 0.2 for 10 models, namely total C and N, 13C, 15N, control extract-C, fumigated extract-C and -N, biomass-C and -N and biomass-15N models and therefore considered as very good. With an R2=0.955, the fumigated extract-15N model is also good. The standard error of performance calculated on the independent set of data and SEC were within 20% of each other for all the best equations except for the biomass-15N model. The ability of NIRS to detect 13C and 15N in total C and N and in the extracts is noteworthy, not because of its predictive function that is not really of interest in this case, but because it indicates that the spectra kept the signature of the properties of the organic matter derived from the straw even after two- or three-year decomposition. The incorporation of the 13C in the biomass was less well predicted than that of the 15N. This could indicate that the biomass derived from the straw was characterised by a particular protein or amino acid composition compared to the total biomass that includes a large proportion of dormant micro-organisms. The predictive ability of NIRS for microbial biomass-C and -N is particularly interesting because the conventional analyses are time consuming. In addition, NIRS allows detecting analytical errors. Author Keywords: Near infrared reflectance spectroscopy; 13C; 15N; Soil organic matter; Microbial biomass; Forest soil |
