Decomposer faecal food web and C sequestration in soil. Can near infrared spectroscopy describe transfers and transformations from fresh organic inputs to protected forms in soil aggregates?

This study sought to characterize the transfer of agricultural organic inputs by macrofauna to the biogenic aggregate compartments of the soil and test the existence of a faecal food web process based on coprophagic feeding behaviours. In an experimental plantain crop field of Colombia, we applied the agroecological FBO (Fertilisation Bio Organique ®) technology, a nucleation technique which consists in planting perennial plants in 1.0 x 0.4 × 0.4 m deep trenches where low- and high-quality organic materials are added in a specific design, and endogeic earthworms are inoculated. We evaluated the effectiveness of Near Infrared Spectroscopy (NIRS) to track this flow of organic matter through different natural stages of decomposition represented by soil macroaggregates produced by soil macroinvertebrates as faecal pellets and casts. After one year, great part of the organic inputs had been either mineralized or incorporated into macroaggregates of different sizes and origins. Using NIRS and physicochemical laboratory analyses, we assessed the quality and quantity of organic matter in the different types of aggregates separated manually according to their origin: large biogenic aggregates (LBA), medium sized biogenic aggregates (MBA) and small biogenic aggregates (SBA); physical aggregates (PA), root aggregates (RA) and residual soil (RS).

Earlier studies had shown that these structures were presumably formed by the activity of macroinvertebrates of three different functional groups, Diplopoda and Isopoda litter transformers for SBA, epidendogeic earthworms for MBA and mesohumic endogeic earthworms for LBA, organized in a feeding succession.

A Coinertia analysis between showed significant covariation among the two sets of physico chemical (22 variables) and NIR spectral (100 different wavelength absorbances) characterization of 135 samples representative of the three classes of biogenic macroaggregates, physical aggregates and residual non macroaggregated soil, (RV = 0.55; p < 0.001). This analysis clearly separated biogenic structures and ranked them according to their size, from small SBA to medium sized MBA and large LBA. Physical PA aggregates and RS residual soil were projected close to the LBAs in the coinertia factorial plan. Multiple combinatorial data analysis CDA, associated 5 specific wavelength absorbance ranges with aggregate types and residual soil. Along the sequence from small to large biogenic aggregates, residual soil and physical aggregates, wavelengths associated to easily decomposed substrates (in ranges 1708–1716 nm; 1796–1948 and 2164–2316 nm) had progressively decreasing absorbances. Substrates associated to slowly decomposing aromatic, alkane and phenolic substrates either increased (1420–1436) or decreased (1284–1380) along this sequence.

These results are compatible with the hypothesis of a progressive transformation and transfer of organic residues first into small biogenic aggregates that are presumably faecal pellets of Isopoda and Diplopoda, then to medium sized biogenic aggregates identified as casts of epiendogeic earthworms and finally to large biogenic aggregates expected to be the casts of the large endogeic earthworms present in these soils. A definitive verification of the hypothesis and the involvement of specific macroinvertebrates will require direct experimentation in controlled laboratory conditions.