Recent studies show that earthworms can considerably increase emissions of the greenhouse gas nitrous oxide (N₂O). However, the extent of this effect varies widely with earthworm ecological group, and exact pathways remain unclear. Here we determined to what extent the earthworm-induced N₂O effect is related to the incorporation depth of residue by different ecological groups. We hypothesized that (i) N₂O emissions reduce with incorporation depth in the absence of earthworms, as more N₂O is reduced to N₂ while diffusing upwards; (ii) effects of a deep-burrowing (anecic) earthworms that are confined to various burrowing depths show a similar pattern; and (iii) the effects of an anecic and a topsoil-dwelling (epigeic) earthworm species are therefore identical when they are both confined to the topsoil. In a mesocosm (50 cm depth, sandy soil) study we measured cumulative N₂O emissions. In experiment A, maize residue was manually incorporated at a depth of 0, 10, 30 and 50 cm. In experiment B, ¹³C-labeled maize residue was applied on top, but individuals of the anecic species Lumbricus terrestris (L.) were confined to 10, 30 or 50 cm with a nylon mesh, and individuals of the epigeic species Lumbricus rubellus (Hoffmeister) to 10 cm.

Results/Conclusions

After 83 days, L. terrestris biomass was decreased with 25.3%, without a significant effect of confinement depth. L. rubellus had lost significantly more weight (48.0%; p=0.015). In experiment A, N₂O emissions decreased with incorporation depth (p<0.001) from 4.91 to 2.69 mg N₂O-N kg^-1 soil. In experiment B, N₂O emissions differed with earthworm treatment (p<0.001). Highest emissions were observed for the L. rubellus treatment at 10 cm confinement (5.05 mg N₂O-N kg^-1 soil). Although confining L. terrestris to 10 cm resulted in lower emissions (3.85 mg N₂O-N kg^-1 soil), this was not significantly different from L. rubellus. Confinement depth significantly affected emissions of L. terrestris (p<0.001), but lowest emissions were found for 30 rather than 50 cm incorporation depth. With regard to our hypotheses, we postulate that (i) N₂O emission decreases as expected with residue incorporation depth in the absence of earthworms; (ii) in the presence of earthworms this effect was obfuscated by changed gas diffusion properties through burrowing activity; and (iii) no significant differences were observed between an epigeic and anecic earthworm that were both confined to the topsoil. We conclude that earthworm-induced N₂O emissions are indeed related to residue incorporation depth, albeit in more intricate ways than previously assumed.