Precipitation pulse size and frequency controls on dryland litter decomposition rates

Abstract:

Drylands are an important component of the global carbon (C) cycle, accounting for 40% of the land area and 20% of the soil organic C globally. Litter decomposition is a key biogeochemical process, controlling C and nutrient cycling. While simple decomposition models successfully predict decomposition rates in many systems based on climate variables, there is a disconnect between the modeled and measured rates decomposition in drylands. This disconnect may stem from abiotic factors of importance in drylands, such as photodegradation and soil-litter mixing, not being taken into account. Soil-litter mixing can accelerate decomposition, but the underlying mechanisms are poorly understood. Potential mechanisms include microclimate buffering, physical abrasion, and enhanced microbial colonization. Recent work suggests that litter decomposition is remarkably insensitive to climate variables, at least when variables are presented as long temporal-scale values (e.g., annual precipitation). We hypothesized that decomposition would be more strongly affected by litter moisture content than total precipitation (PPT) alone. Thus, frequent, small PPT pulses would accelerate decomposition more than larger, but infrequent pulses. Furthermore, soil-litter mixing would enhance decomposition by buffering litter moisture content. To test the combined influence of soil-litter mixing and PPT pulses on decomposition, we incubated litter and soil in a semi-controlled greenhouse which simulated dryland summer temperatures. Two litter types (grass and shrub) were incubated under two levels of soil-litter mixing (no mixing and complete soil-litter mixing) and with 16 different PPT treatments (a factorial combination of four PPT pulses sizes and four PPT frequencies). We measured instantaneous CO2 flux throughout the 30 day incubation and mass loss at the end of the incubation. Shrub litter decomposed faster than grass litter. Flux rates generally peaked at day 8 and declined thereafter. CO2 flux rates were largely a function of time since last PPT pulse, with more stable CO2 flux rates for more frequent pulses than infrequent pulses. Soil-litter mixing altered responses to PPT pulses. Together, these data support the idea that decomposition is more related to stability of litter moisture than total long-term PPT.