Woody plant encroachment impacts on soil carbon and microbial processes: results from a hierarchical Bayesian analysis of soil incubation data

Belowground processes and associated plant–microbial interactions play a critical role in how ecosystems respond to environmental change. However, the mechanisms and factors controlling processes such as soil carbon turnover can be difficult to quantify due to methodological or logistical constraints. Soil incubation experiments have the potential to greatly improve our understanding of belowground carbon dynamics, but relating results from laboratory-based incubations to processes measured in the field is challenging. This study has two goals: (1) development of a hierarchical Bayesian (HB) model for analyzing soil incubation data and complementary field data to gain a more mechanistic understanding of soil carbon turnover; (2) application of the approach to soil incubation data collected from a semi-arid riparian grassland experiencing encroachment by nitrogen-fixing shrubs (mesquite). Soil was collected from several depths beneath large-sized shrubs, medium-sized shrubs, grass, and bare ground—the four primary microsite-types found in this ecosystem. We measured respiration rates from substrate-induced incubations, which were accompanied by measurements of soil microbial biomass, soil carbon, and soil nitrogen. Soils under large shrubs had higher respiration rates and support 2.0, 1.9, and 2.6 times greater soil carbon, microbial biomass, and microbial carbon-use efficiency, respectively, compared to soils in grass microsites. The effect of large shrubs on these components is most pronounced near the soil surface where microbial carbon-use efficiency is high because of enhanced litter quality. Grass microsites were very similar to bare ground in many aspects (carbon content, microbial biomass, etc.). Encroachment of mesquite shrubs into this semi-arid grassland may enhance carbon and nutrient cycling and increase the spatial heterogeneity of soil resource pools and fluxes. The HB approach allowed us to synthesize diverse data sources to identify the potential mechanisms of soil carbon and microbial change associated with shrub encroachment.