| Summary: | Thesis (M.S.) University of Alaska Fairbanks, 2004 The purpose of this study was to better understand the link between dissolved organic matter (DOM) in soil leachate and different vegetation attributes. Soil cores were collected from the Caribou Poker Creeks Research Watershed (CPCRW) and subjected to a laboratory leaching procedure. The leachates were then subjected to a number of analytical tests, including pyrolysis-gas chromatography/mass spectrometry (py-GC/MS). Py-GC/MS is a 'molecular fingerprinting' technique that was used to help determine similarities and differences in organic matter leached from soils with different vegetation attributes. Numerous statistical tests were performed including Student-t, analysis of variance, principal components analysis, and partial least squares regression (PLS). Results from Student-t tests indicated that local vegetation plays an important role in the character of the DOM in soil leachate. Additionally, a principal components test revealed relationships between soil leachates and vegetation attributes. A prediction model was created using PLS to predict components of leachate DOM based on vegetation attributes. This model, while in its early development, was able to predict 70% of the total molecular fingerprint of leachate DOM based on cover vegetation.
|
|---|