Soil deposition and UV radiation influence litter decomposition in a shrub-invaded dryland ecosystem

Background/Question/Methods

Decomposition is a fundamental driver of biogeochemical cycling that controls nutrient availability, carbon storage, and plant community composition.  Although mechanistic models successfully predict decomposition dynamics in many different systems, predictions in drylands remain problematic due to a poor understanding of driving variables.  Recent studies suggest two previously overlooked drivers of decomposition in drylands: UV radiation and mixing of litter with wind/water-transported soils.  We evaluated the relative importance of these drivers by experimentally quantifying radiant energy–soil-litter mixing–decomposition interactions in the context of woody plant encroachment into grasslands, a globally extensive vegetation change.  We tested the hypothesis that soil deposition into litter would negate UV effects and have a net positive effect on decomposition rates of shrub (mesquite) and grass (Lehmann lovegrass) litter by enhancing microbially-mediated decomposition. A factorial field experiment using open-topped litter boxes varied levels of soil deposition [0%, 50%, 100% of litter covered by soil], levels of  UV radiation [near ambient UV-A+UV-B, no UV-B, and no UV-A or UV-B], and levels of total radiant energy (open vs. under-shrub). Samples were collected after 0, 1, 3, 6, and 12 months and analyzed for mass loss and litter chemistry.

Results/Conclusions

Initial carbon (C) concentrations in grass litter (42.6%) were slightly, but significantly higher than those in shrub litter (40.9%; P<0.001), while initial nitrogen (N) concentrations in grass litter (0.6%) were significantly lower than those in shrub litter (2.3%, P<0.001).  Shrub litter decomposed more rapidly than grass litter (P<0.001).  Litter coverage by soil (%) was a significant predictor of mass loss and C and N content for shrub (P<0.05 for all) and grass litter (P<0.01 for all). Soil cover*time interactions were significant for C and N content, but not mass loss in shrub litter (P<0.001); for grass litter, soil cover*time interactions were significant for litter mass loss (P<0.05) and C and N content (P<0.001).  Ambient UV treatments resulted in lower C and N concentrations in shrub litter compared to treatments that blocked UV.  Carbon (R² = 0.61) and nitrogen (R² = 0.63) concentrations declined linearly with declines in litter mass.  Results indicate that soil-litter mixing accelerates decomposition as soil accumulation occurs, with rates of decomposition increasing with increasing soil coverage of litter.  These results suggest that soil-litter mixing is an important driver of decomposition in drylands and should be considered when examining decomposition dynamics in systems with low vegetative cover and high rates of aeolian and fluvial soil movement.