Abstract

In this study the near-infrared reflectance (NIR) spectra signals (750–2,500 nm) of soil samples was compared with the NIR signals of the biogenic aggregates produced in the lab by three earthworm species, i.e., Aporrectodea rosea (Savigny 1826), Lumbricus friendi Cognetti, 1904 and Prosellodrilus pyrenaicus (Cognetti, 1904) from subalpine meadows in the Central Pyrenees. NIR spectral signatures of biogenic aggregates, root-aggregates, and non-aggregated soil were obtained together with soil carbon (C), nitrogen (N), NHþ 4 and NO 3 determinations. The concentrations of C, N and C:N ratio in the three types of soil aggregates identified were not statistically significant (ANOVA, p>0.05) although non-macroaggregated soil had slightly higher C concentrations (66.3 g kg-1 dry soil) than biogenic aggregates (earthworm- and root-aggregates, 64.9 and 63.5 g kg-1 dry soil, respectively), while concentrations of NHþ 4 and NO 3 were highest in the root-attached aggregates (3.3 and 0.31 mg kg dry soil-1). Total earthworm density and biomass in the sampled area was 137.6 ind. m-2, and 55.2 g fresh weight m-2, respectively. The biomass of aggregates attached to roots and non-macroaggregated soil was 122.3 and 134.8 g m-2, respectively, while biomass of free (particulate) organic matter and invertebrate biogenic aggregates was 62.9 and 41.7 g m-2, respectively. Multivariate analysis of NIR spectra signals of field aggregates separated root aggregates with high concentrations of NHþ 4 and NO 3 (41.5% of explained variance, axis I) from those biogenic aggregates, including root aggregates, with large concentrations of C and high C:N ratio (21.6% of total variability, axis II). Partial Least Square (PLS) regressions were used to compare NIR spectral signals of samples (casts and soil) and develop calibration equations relating these spectral data to those data obtained chemical variables in the lab. After a derivatization process, the NIR spectra of field aggregates were projected onto the PLS factorial plane of the NIR spectra from the lab incubation. The projection of the NIR spectral signals onto the PLSR models for C, N, NHþ 4 and NO 3 from casts produced and incubated in the lab allowed us to identify the species and the age of the field biogenic aggregates. Our hypothesis was to test whether field aggregates would match or be in the vicinity of the NIR signals that corresponded to a certain species and the age of the depositions produced in the lab. A NIRS biogenic background noise (BBN) is present in the soil as a result of earthworm activity. This study provides insights on how to analyse the role of these organisms in important ecological processes of soil macro-aggregation and associated organic matter dynamics by means of analyzing the BBN in the soil matrix