Monday, 7 November 2005
1

New Statistical Approach to Analyzing Metabolic Diversity Data.

Melissa A. Dulinski, Cris San Miguel, and Robert L. Tate, III. Rutgers University, 14 College Farm Rd., Environmental and Natural Resources Building, New Brunswick, NJ 08901

Metabolic diversity profiles, such as those obtained from the BIOLOGTM system, have been used for many years to identify differences in soil microbial communities. The carbon substrates on the BIOLOG plates fall into several chemical groups including carbohydrates, carboxylic acids, amino acids, amines/amides, polymers, and miscellaneous compounds. Although an optical density (OD) value is obtained for each substrate at each time point that presumably could be used to evaluate actual soil microbial capability, soil sample variability precludes their use via currently accepted procedures. Thus, data analyses are usually limited to Principal Component Analysis (PCA) or related procedures. These statistical procedures do not supply information about microbial community structure or function. The objective of this study was to develop a simple, user-friendly technique which could be used to relate the BIOLOG data to field microbial capabilities. The method would be particularly useful for investigating specific soil environments, using custom designed BIOLOG plates for testing specialized substrates such as petroleum hydrocarbons. The procedure involved examination of multiple replicates of soil samples. To compare these communities, an ANOVA analysis of the ODs for each substrate was completed to determine if any particular substrate was metabolized to a greater extent by either soil community. The data from ANOVA was then visualized graphically to compare the utilization of all substrates present in the BIOLOG plates. This analysis is a progressive technique when attempting to demonstrate the selective utilization of a particular substrate in a soil community to show significant differences in metabolic diversity. The number of replicates needed to accurately reflect the variability of the microbial communities was a minimum of ten sample replicates. This type of data analysis may be useful when a study could benefit from custom designed BIOLOG plates.

Back to S 3 Graduate Student Poster Competition
Back to S03 Soil Biology & Biochemistry

Back to The ASA-CSSA-SSSA International Annual Meetings (November 6-10, 2005)