Soil microbial communities play a central role in major ecosystem processes, including the transformation of energy and nutrients and the production of greenhouse gases. There is a critical need to understand how soil microbial community composition and physiology influence ecosystem functioning. The advent of new technologies such as molecular and isotopic techniques is accelerating the acquisition of knowledge of unknown biota and their linkage to function. Understanding the dynamics of microbial communities and how they control particular ecosystem processes is of central importance to ecosystem sustainability.
The broad objectives of my dissertation were to 1) develop techniques that link microbial communities to the degradation of different pools of soil organic matter (SOM), 2) quantify effects of environmental change on microbial community characteristics, and 3) determine how changes in community composition and physiology affect the degradation and utilization of different soil carbon pools.

In the first chapter, I reviewed what is known about microbial communities and C cycling, especially focusing on data that show when explicit knowledge of microbial community composition help to explain process rates. I discussed two useful techniques, 13C-PLFA, which directly links microbial communities with SOM utilization, and soil enzyme activities, a biological determinant of macromolecular SOM degradation. I reviewed the controls on microbial community composition and function, identified cases where changes in community composition affect soil processes, and determined the gaps in knowledge in soil carbon cycling that may be filled by improved knowledge of the dynamics of microbial communities.
If changes in the composition of the soil microbial community alter the physiological capacity of the community then such changes may have ecosystem consequences. In the first experiment (Chapter 2), I examined the relationships among community composition (PLFA), microbial biomass (CFDE), substrate utilization profiles (Biolog), lignocellulose degrading enzyme activities (b-glucosidase, cellobiohydrolase, b-xylosidase, phenol oxidase, peroxidase), and nutrient releasing enzyme activities (phosphatase, sulphatase) in a Tropeptic Haplustol soil. The soils supported a tropical forest and pineapple plantations of varying ages that were at different stages within the management cycle. Conversion from forest to agriculture significantly decreased %C and %N of the soil by 50-55%, microbial biomass by 75%, ß-glucosidase by 54%, sulphatase activity by 85%, decreased Ca, Mg, and Mn availability, and produced compositionally and functionally distinct microbial communities. Total enzyme activities were generally correlated with %C, %N, microbial biomass, and occasionally with community composition. I calculated the specific activities of the enzymes assayed (enzyme activity per unit microbial biomass C) in order to normalize activity to the size of the microbial community. Values for enzyme specific activity were more highly correlated with community composition than were total enzyme activities. In addition, Biolog was not correlated with community composition or enzyme activities. I concluded that enzyme activities and specific activities may provide a useful linkage between microbial community composition and carbon processing.

The mechanisms underlying changes in microbial utilization of soil organic carbon under altered temperature, moisture, and N fertilization regimes are poorly understood. The purpose of the Chapter 3 was to determine how changes in temperature, soil N, and soil moisture affect the microbial utilization of young and old soil carbon pools, and how this is mediated by changes in microbial community composition and physiology. In order to differentiate young and old soil carbon, I used a soil from a CAM pineapple plantation (d13C litter = -14‰) that had previously been C3 tropical forest (d13C soil carbon = -28‰). Thus forest derived carbon represented ‘old’ carbon and plantation inputs represented ‘new’ carbon. In an incubation experiment, I separated utilization of young (<14yr) and old (>14yr) soil carbon in respired CO2 and in microbial phospholipid fatty acids (PLFAs). Soil temperature, moisture content, and N content were manipulated in a 103 day lab incubation during which I monitored CO2 and 13C-CO2 production. I measured community composition (PLFA and Intergenic Transcribed Spacer (ITS) analysis of community DNA), lignocellulose degrading enzyme activities (b-glucosidase, cellobiohydrolase, b-xylosidase, phenol oxidase, peroxidase), and nutrient releasing enzyme activities (phosphatase and sulfatase). I found that at a lower temperature (5ºC) microbial communities utilized younger SOM and at higher temperatures (20 and 35ºC ) microbial communities utilized older, presumably more recalcitrant organic matter. Nitrogen fertilization stimulated the utilization of older soil carbon. ITS and PLFA patterns revealed clear changes in microbial community composition due to temperature, but not in response to irrigation or N fertilization. Enzyme activities measured at the end of the incubation decreased with increasing incubation temperature possibly due to the depletion of their substrate pools. Oxidative enzyme activities were lower in the high soil moisture treatment and higher in the N fertilization treatment compared to control. Nitrogen fertilization and irrigation significantly altered the isotope ratios of several bacterial PLFAs. These data challenge assumptions that the sizes and ages of substrate pools of C in soil are constant at different temperatures and fertilization regimes. Changes in the composition and physiology of the microbial community point to a plausible mechanism for alterations in C resource utilization patterns.

In the next experiment, I examined the degradation of C substrates and SOM in high and low fertility soils in order to determine whether microbial communities differ in their ability to degrade different C sources. Soils were located underneath oak canopies and in open grassland plots near Hopland, California. Carbon substrates examined were 13C-labeled starch, hemicellulose, vanillin, and pine litter. Substrate degradation differed between the two soils. Addition of substrate to soil enhanced the degradation of indigenous SOM, traditionally termed the priming effect. The enhancement of SOM degradation differed between the two soils: in the open grassland soil it was high and permanent while generally small, shortlived, or even negative in oak canopy soils. The stimulation of enzyme activity following substrate addition was highly related to the degradation of added substrate, but not SOM. Different microbial communities utilized the C substrates in the two soils. In general, the biomarker 16:0 and the Gram - biomarkers utilized added substrates. The fungal biomarker primarily utilized pine litter. 13C-PLFA data revealed that the fungal community in oak canopy soils was not as actively degrading SOM as the fungal community in the grassland soils, potentially explaining why oak canopy soils had a lower priming effect than open grassland soils. This experiment showed that differences in microbial community composition and activity can alter the degradation rates of different pools of soil C, thus demonstrating the critical importance of microbial community ecology in C cycling research.

In a field experiment in a California oak-grassland ecosystem, I transplanted soil cores between oak canopy and open grassland soils in order to determine the response of the microbial community to changes in plant communities over a one-year period. 13C-PLFA with a universal carbon substrate measured the activities of individual microbial biomarkers. Microbial community characteristics were associated with different plant communities, changed seasonally, and responded to the transplantation treatment. Microbial communities responded to the transplant treatment only when cores were transplanted from the oak canopy soil environment into the open grassland soil environment. Data supported the hypothesis that microbial community composition shifts when microbes are exposed to unfamiliar environmental conditions. Plant community control of microbial community composition was primarily due to plant effects on soil water content while the changes in microbial community composition seasonally appeared to be due to carbon availability and carbon quality. Microbial community composition and metabolic profiles (13C-PLFA) were strongly related to interannual variability in soil enzyme activities and soil respiration. This study showed that microbial communities are seasonally dynamic, sensitive to new environmental regimes, and central to C cycling process rates in soil.
Overall, linkages between microbial community composition, physiology, and soil processes were possible using a quantitative measure of community composition (PLFA), genetic fingerprinting, stable isotopes, and enzyme activities. These data and approaches identified several examples in which microbial communities affected rates of C cycling in soil (e.g. following disturbance, under different temperature and N fertilization regimes, during the degradation of macromolecular C substrates, seasonally in oak canopy and grassland soils), and determined some of the various factors that affect or control community composition and physiology. These data support the hypothesis that microbial communities are not functionally identical with regard to the degradation of macromolecular C substrates. The central importance of this complex array of interacting organisms to ecosystem health and sustainability is coming to light as the black box of belowground biodiversity is unearthed and unveiled.