Thursday, August 9, 2007 - 2:50 PM

COS 130-5: Is soil microbial ecology predictable: A cross-ecosystem study

Benjamin Colman, University of California, Santa Barbara, Joshua P. Schimel, University of California, Santa Barbara, and Noah Fierer, University of Colorado-Boulder.

While many studies have examined rates of microbial C and N mineralization in soil, they are generally site, ecosystem-type, or region specific.  To evaluate whether microbial process rates and community characteristics are predictable at a larger scale, based on soil and site specific characteristics, we sampled 84 soils from established research sites across North America representing a wide variety of soil, vegetation, and ecosystem types.  For all soils we characterized pH, CEC, base saturation, C and N content, soil texture, and water holding capacity.  These physicochemical properties, as well as site characteristics such as vegetation type, and mean annual precipitation (MAP), were then compared with the microbial community characteristics such as fungal:bacterial ratio, substrate induced respiration (SIR) biomass, and measurements of C mineralization and net N mineralization.  We found that soil microbial ecology is predictable, but improved inference comes from using SIR biomass as a descriptive variable. We found strong relationships between soil characteristics and process rates/microbial community characteristics. For example, we found that we could explain: 61% of fungal:bacterial ratio with C:N ratio, MAP, SIR biomass, and pH; 77% of variability in C mineralization rate with SIR biomass, clay content, and C:N ratio; and 88% of variation in N mineralization with soil C and soil C:N ratio.