Agro-ecosystems are a major contributor to global emissions of the greenhouse gas nitrous oxide (N₂O). It is known that earthworm activity may raise N₂O emissions from these systems, depending on the earthworm species present. Rather than emitting N₂O themselves, earthworms are thought to enhance soil microbial activity (nitrification, denitrification, and nitrifier denitrification) by changing physico-chemical properties, excreting mucus, and increasing available carbon. Here we focus on the link between earthworm activity and microbial N₂O emission pathways in different crop residue incorporation systems. Our hypotheses are that earthworm-induced N₂O emissions i) reflect earthworm ecological strategy; and ii) can be traced to changes in microbial activity. We initiated two 90-day mesocosm studies with combinations of the soil-dwelling Aporrectodea caliginosa (Savigny; endogeic) and the topsoil-dwelling Lumbricus rubellus (Hoffmeister; epigeic), all at realistic densities for Dutch agro-ecosystems. Crop residue was placed on top or incorporated in both experiments. Mesocosms in experiment 1 contained a loamy soil; in experiment 2 a sandy soil. In both experiments N₂O fluxes and soil N dynamics were measured throughout the incubation. In experiment 2, microbial biomass and denitrification enzyme activity (DEA) were measured, as well as gene copy number for the amoA, nirK, nirS and nosZ enzymes (using qPCR).

Results/Conclusions

Experiment 1: when residue was incorporated, only A. caliginosa significantly increased cumulative N₂O emissions (from 1350 to 2223 µg N₂O-N kg^-1 soil; p<0.05). When residue was applied on top, L. rubellus increased N₂O emissions from 524 to 929 µg N₂O-N kg^-1 soil (p<0.001), and a significant interaction between both earthworm species increased emissions to 1397 µg N₂O-N kg^-1 soil (p<0.05). Experiment 2: cumulative N₂O emissions increased only when L rubellus was present and residue was applied on top (from 64 to 128 µg N₂O-N kg^-1 soil; p=0.047). When residue was applied on top, DEA significantly increased in the presence of L. rubellus (from 27.1 to 45.0 µg N₂O-N hr^-1 kg^-1 soil; p<0.001), and a trend towards higher nosZ gene copy numbers was observed. This coincided with a positive correlation between DEA and N₂O fluxes (p=0.007). We conclude that i) earthworm-induced N₂O emissions follow earthworm ecological strategy: epigeic earthworms can increase N₂O emissions when residue is applied on top; endogeic earthworms when residue is incorporated into the soil by other earthworms species or humans, and ii) increased earthworm-induced microbial (denitrification) activity results in increasing N₂O emissions. So far, this effect can only be associated with epigeic earthworm species.