Applying EEM spectra with absolute principal component coefficients to explore characteristic fingerprints and fractions of DOM from diverse land-use soils in a large-scale irrigation area

Soil dissolved organic matter (DOM) is crucial for soil formation as well as the global biogeochemical cycle. Land use plays a substantial impact on the composition structure of soil DOM, while the linkage between soil DOM fractions and land use is still rarely studied. Here, sixteen topsoil samples were collected from cropland (CRLD), xylophyta community (XYCY) and halophyte community (HACY) in Hetao irrigation area of China. Three-dimension excitation-emission matrix fluorescence spectroscopy (EEM) conjunct absolute principal component coefficients (PC coefficients), parallel factor analyses (PARAFAC) and structural equation modeling (SEM) was employed in characterizing fingerprints and fractions of DOM from the topsoil samples. Five fluorescent components were extracted from DOM via EEM-PARAFAC, i.e., tryptophan-like (C1), microbial humic-like (C2), fulvic-like (C3 and C4), and humic-like (C5). The microbial humic-like was the primary component in the CRLD and HACY topsoils, while fulvic-like was the major component in the XYCY topsoil. Total organic carbon had substantial negative relationships with soil moisture (r = −0.885, p < 0.001), electrical conductivity (r = −0.746, p < 0.001) and exchangeable sodium percentage (ESP) (r = −0.695, p < 0.05), and displayed a high consistency among the different topsoils. Based on SEM, ESP had an impact on C1 in the CRLD, but primarily influenced C1 and C2 both in the XYCY and HACY, which in turn affected the degree of soil humification. EEM spectra combined with absolute PC coefficients is an effective approach to characterize soil DOM fractions not only in irrigation area but also in broad regions.