The dynamics of soil micro-food web structure and functions vary according to litter quality

• Micro-food web interactions decreased with increasing litter recalcitrance.
• Interactions networks revealed microbial but not faunal links.
• Biota strategies for C acquisition differed according to the initial litter quality.
• Enzymatic production was not explained by the microbial structure.
• L. terrestris changed enzyme dynamics without altering the micro-food web structure.

Anthropogenic pressures on agricultural soils are known to alter the properties of soil food webs, which may affect ecosystem functions and the capacity to deliver services. Cropping systems fuel belowground biota by supplying litter, but litter quality varies from year to year. In this study, we hypothesized that (i) the structure and function of the soil biota may respond strongly to the quality of crop residue, and (ii) this response may vary with the complexity of the food web. To test this hypothesis, a 3-month incubation experiment was performed using soil columns filled with a silty loam agricultural soil (Estrées-Mons, Northern France), and plant litter was incorporated into the 0–5-cm layer. The quality of the litter was modulated by using either high quality (maize leaves) or low quality materials (maize roots), and two levels of biotic complexity were tested by introducing or not introducing an ecosystem engineer (Lumbricus terrestris, anecic earthworm) into the columns. We then investigated litter decomposition and enzymatic activities, as well as the biomass and diversity of the bacterial, fungal and nematode communities. The C mineralization and enzymatic profiles varied according to litter quality, and the most obvious differences were observed in the hydrolytic enzymes and phenol oxidase. Micro-food web interactions and structure were mainly impacted by litter type and the time since litter incorporation, and as expected, we observed a temporal succession of cellulolytic organisms followed by lignolytic organisms. This progression was mirrored by the nematode community; the ratio of fungal feeders/bacterial feeders increased between 35 and 91 days. Initial litter quality primarily impacted bacterial and fungal community structure, and the development of the bacterial channel was greater with the addition of leaf litter. This was related to an increase in specific, favored phyla, Bacteroidetes and Proteobacteria, whereas Actinobacteria increased more with the addition of root litter. However, we could not statistically link the changes in the structure of the microbial communities over time with those in enzyme activity, which led us to assume that other factors, such as top-down regulation, are more important than microbial community structure in the determination of the enzyme secretions in the soil. Finally, the addition of earthworms did not significantly impact litter decomposition and had a low impact on micro-food web structure and interactions. However, the presence of L. terrestris did modify the enzymatic secretions, probably through its endemic microbial population, as well as the stimulation of microbial activity through predation.