Abstract
The concept of feedback loops between changes in chemical quality of decomposing organic residues and changes in faunal communities was employed in studying how such feedback loops, representing distinct ecological successional stages, determine decomposition dynamics in soils. A 52-week litterbag decomposition study was superimposed onto an 18-year long term field experiment. Four types of organic residues contrasting in chemical quality (i.e., nitrogen (N), lignin, polyphenols, cellulose) were incorporated into soil annually to assess decomposition and associated meso- and macrofauna communities. In the first 4 weeks after residue incorporation (loop #1), the abundances (densities) of both mesofauna and macrofauna were positively influenced by labile cellulose and N. The mesofauna Collembola and Acari contributed 70-100% and 0-30% to the decomposition, respectively, while the macrofauna beetles and flies contributed 20-90% and 10-66%, respectively. The abundances were highest under groundnut (high N, low lignin) ([1.35 and 0.85 individual number (g dry litter)(-1)] for mesofauna and macrofauna, respectively). The presence of macrofauna at week 2 led to a mass loss (R(2) = 0.67**), indicating that macrofauna preceded mesofauna in degrading residue. In week 8 (transition of loop #2 to #3), only macrofauna (beetles dominated contributing 65%) played an important role in lignin decomposition (R(2) = 0.56**), resulting in a mass loss (R(2) = 0.52**). In week 52 (loop #4) macrofauna, ants (Formicidae) replaced beetles as the dominant decomposers showing a feedback reaction to availability of protected cellulose. The Formicidans contributed 94% to the decomposition and influenced losses of mass (R(2) = 0.36*) and N (R(2) = 0.78***). The feedback loop concept provides a more comprehensive "two-sided" view into decomposition, as regulated simultaneously by two factors, than earlier "one-sided" approaches to soil fauna-mediated decomposition.