Principal sources of soil-derived fugitive dust need to be identified to reduce airborne PM10 in California's Central Valley. As a means to differentiate soils and possibly identify sources of fugitive dust, we have developed methods in our laboratory to produce fingerprints from microorganisms in soil. Fingerprinting methods described in this report are based on two classes of biochemical material, fatty acids and nucleic acids, which we extract from soil or dust prior to chemical analysis. Fatty acid analysis can be based either on phospholipid fatty acids (PLFAs), found only in cell membranes of living organisms, or on soil fatty acid methyl esters (SFAME) obtained from whole cells and nonliving biological material. Similarly, nucleic acid analysis can be based either on DNA sequences from individual groups of organisms or on DNA from the entire microbial community. PLFA or SFAME fingerprints consist of percentages of different fatty acids detected as peaks in gas chromatograms, while DNA fingerprints consist of band patterns in laboratory gels used to separate DNA fragments. Both types of fingerprinting methods generate multivariate data (fatty acid percentages or DNA band identities), which can be used in principal component analysis (PCA) to assess similarities among samples. Appendices in this report contain protocols for the extraction and analysis methods we developed.
Approximately 300 soil samples from California's Central Valley, representing numerous land use categories, soil types, crops and other variables important in determining sources of air-borne dust, were analyzed for their PLFA fingerprints. PCA plots of PLFA fingerprints under different vegetation and agricultural management showed a clear separation between microbial communities in poorly drained and well- drained soils. Redundancy analysis revealed that both soil texture and crop type were significantly correlated with variation in PLFA fingerprints across soils. The relative importance of environmental variables in governing the composition of microbial communities could be ranked in the order: soil type > time > specific farming operation (e.g., cover crop incorporation or sidedressing with mineral fertilizer) > management system > spatial variation in the field. Similar conclusions could be drawn from these PLFA data when they were analyzed with artificial intelligence (AI) programs in a research collaboration with a chemometrics laboratory.
SFAME fingerprints, advantageous because they utilize smaller sample sizes and require one third of the time needed for PLFAs, were performed on a smaller subset of soil samples (approximately 20), because we expected their chromatograms to be less reliable due to overlapping peaks that were difficult to identify. SFAME differentiated the 20 soils similarly but not identically to PLFA. We also evaluated the similarity between PLFA and SFAME fingerprints of source soils and bulk dusts collected from the surfaces of field equipment in two locations. In both cases, PLFA fingerprints of bulk dust and source soils were similar. SFAME fingerprints from bulk dusts and their source soils were more dissimilar to each other than were PLFA fingerprints, although SFAME fingerprints grouped together on PCA plots. Even though these findings cannot be directly extrapolated to establish similarities between airborne PMo and source soils, this study provided the first step in evaluating fingerprints obtained from dust and a single- source soil. More sophisticated multivariate analysis methods, such as AI programs, will be needed to interpret fingerprints of dust derived from multiple sources.
Our research on nucleic acid-based methods focused on identifying and improving methods to extract DNA from soil and on testing several approaches for analyzing the extracted DNA. It was possible to extract high quality DNA from soils representing a wide range of properties. The RAPD (Randomly Amplified Polymeric DNA) method for analyzing DNA, adapted from methods developed for fingerprinting individual species, proved to be inadequate for fingerprinting the extraordinarily diverse microbial communities in soil. Although TGGE (thennal denaturing gel electrophoresis) showed promise as a means to generate DNA fingerprints from specific groups of soil microbes, more research is needed to optimize this method for fingerprinting whole-community DNA. Further development of DNA-based methods is needed to provide taxonomic explanations for differences in fatty acid fingerprints and to supplement fatty acid fingerprinting in cases where more specific methods are required.
This report indicates that PLFA and SFAME fingerprinting of soil microbial communities will differentiate soils in a reproducible manner, although DNA fingerprinting requires additional development. We found that fingerprints of bulk dusts and their source soils were sufficiently similar to warrant adaptation of fingerprinting technology to PM10 studies of fugitive dust. We describe in the final section of this report how fingerprinting methods could be applied to PM10 samples collected on filters in the field. We also describe how AI programs could enhance statistical analysis of fingerprint data for source apportionment studies of PM10 in the field.
For questions regarding this research project, including available data and progress status, contact: Research Division staff at (916) 445-0753
Stay involved, sign up with CARB's Research Email Distribution List