Dissolved Organic Matter Characterization Using Multiway Spectral Decomposition of Fluorescence Landscapes

Abstract

Dissolved organic matter (DOM) plays an important role in many soil ecosystem functions. Multidimensional fluorescence spectroscopy of DOM with parallel factor analysis (PARAFAC) of the resulting spectral landscape has been successful in characterizing DOM from a variety of aquatic sources. This study was conducted to assess the multiway PARAFAC approach for quantitatively characterizing the fluorescent landscapes of DOM from aqueous extracts of soils and soil amendments. The DOM was extracted from plant biomass representative of crop, wetlands, and tree species; animal manures; and soils from controlled studies of cropping systems with known histories of organic amendments. The fluorescence landscape spectra were collected in the excitation range from 240 to 400 nm and emission range from 300 to 500 nm in 3‐nm increments. The excitation and emission spectra modeled from the PARAFAC analysis showed that the plant biomass, animal manure, and soil DOM contained five fluorescing components: tryptophan‐like (peak location at excitation 270 nm, emission 354 nm), tyrosine‐like (273/309 nm), and three humic‐substance‐like components (>240/465 nm, 306/405 nm, and 315/447 nm). Principal component analysis of the concentration loading showed that the soil‐derived DOM was very similar despite the different types and quantities of organic amendments incorporated in the different cropping systems. This study shows that PARAFAC analysis of multidimensional fluorescence spectra can model the chemical profile of terrestrial DOM in a chemically meaningful way. This represents a significant advance over current approaches to interpreting the complex DOM fluorescence spectra.

Fluorescence characteristics and biodegradability of dissolved organic matter in forest and wetland soils from coastal temperate watersheds in southeast Alaska

“…PARAFAC modeling of Xuorescence EEMs was conducted with MATLAB using the PLS_toolbox version 3.7 (Ohno and Bro 2006) following the procedures described in Stedmon et al (2003;Stedmon and Markager 2005). PARAFAC can take overlapping Xuorescence spectra and decompose the data into score and loading vectors that are quantitative estimates of the relative concentrations of the components.…”

“…Using PARAFAC modeling, we identiWed a total of nine unique components within the Xuorescence EEMs (Table 2). We validated our PARAFAC model using core consistency diagnostics (Ohno and Bro 2006) followed by a split plot analysis (Stedmon and Markager 2005). The core consistency provides a quantitative measure of how well the spectral loadings represent variation in data.…”

“…All nine components identiWed by our model have been previously identiWed as either part of a PARAFAC model (Stedmon et al 2003;Stedmon and Markager 2005;Ohno and Bro 2006) or through visual analysis of EEMs (Coble 1996;Baker 2001). Of the nine components identiWed by the PARAFAC model, we focused our analyses on the following four components: component 1 (humic-like Xuorescence), component 4 (fulvic-like Xuorescence), component 8 (tryptophan-like Xuorescence) and component 9 (tyrosine-like Xuorescence).…”

“…PARA-FAC decomposes the Xuorescence spectra of DOM into independent components whose abundance can be related to diVerences in composition and source material. PARAFAC analyses have been used successfully in soil DOM studies to diVerentiate between diVerent terrestrial sources (Ohno and Bro 2006) and to investigate the sorption of DOM onto mineral soils (Banaitis et al 2006).…”

Calibration, standardization, and quantitative analysis of multidimensional fluorescence (MDF) measurements on complex mixtures (IUPAC Technical Report)

“…With the physical measurement comes the increased use of chemometric data analysis, in essence a "mathematical chromatography" [1,107]. These include, amongst others, the characterization of natural organic matter in soil and aquatic ecosystems [72,108,109], oil-spill forensics [110] and oil characterization [6,111], monitoring of food quality and production [67,71,112], and diagnostics on body fluids e.g. [113,114].…”