Differential responses of microbial substrate-use in dryland shrub, grass, and bare ground patches

Background/Question/Methods

Drylands (arid and semiarid ecosystems) cover about 40% of the Earth’s surface, dominate the interannual variability observed in terrestrial carbon sequestration, and are considered natural models for a future world that is drier and warmer. Soils in drylands have inherently high spatial heterogeneity that has been enhanced by shrub encroachment and proliferation that has occurred in many areas over the past 150 years. This shrub-induced spatial heterogeneity has been quantified in terms of soil physical and biogeochemical properties; however, the spatial variability of microbial substrate-use and how it varies temporally is largely unknown. This knowledge gap is manifested in the poor performance of the biogeochemical components of Earth System Models in drylands. Here, we assessed microbial substrate-use in surface soils across three different vegetation cover types and their relationships to soil physicochemical properties. Our study focused on the three dominant vegetation patch types within a semi-desert grassland savanna: 1) shrubs (Prosopis velutina, mesquite), 2) perennial bunch grasses, and 3) bare soil. From the vegetation cover patches, we collected soil samples to estimate soil organic carbon (SOC), total nitrogen (TN), texture, bulk density, pH, and micro/macro aggregates. We also used multiple substrate-induced respiration to quantify induced CO2 fluxes associated with contrasting substrates (i.e., carbohydrates, carboxylic acids, and amino acids) from soils of each patch type.

Results/Conclusions

Vegetation composition contributed substantially to the spatial heterogeneity of biogeochemical and soil properties. We found a significant patch type effect (P<0.01) on CO2 fluxes induced by the different substrates, with soils associated with mesquite canopies having greater fluxes than soils associated with grasses or bare ground. Moreover, induced respiration by most of the substrates exhibited a strong relationship (P<0.01) with SOC, TN, pH, and litter mass, while the influences of micro/macro aggregates and silt were equivocal. Variability in substrate use in different patch types across drylands should be further investigated since new state-of-the-art mathematical models are being developed incorporating different soil substrates, as well as their kinetics. Integrating the spatial variability of biogeochemical processes, including the kinetics of different substrate use, will improve representation of dryland ecosystem-scale carbon fluxes in Earth System Models.