Characteristics and controlling factors of soil dissolved organic matter in the rainy season after vegetation restoration in a karst drainage area, South China

Abstract

Dissolved organic matter (DOM) has strong reactivity and migration ability, which affects biogeochemical processes such as nutrient cycling. Karst ecological vulnerability has been widely discussed. However, the characteristics of soil DOM, the main energy matrix of the soil, in the process of rocky desertification control have rarely been reported. This study examined the characteristics of DOM in shallow soil after vegetation restoration in a karst area. Soil samples of 5 vegetation restoration types (abandoned land, AL; grassland, GL; peanut cultivated land, PCL; Zanthoxylum bungeanum land, ZBL; forest, FS) and water samples from the drainage were collected. DOM in soils of different vegetation restoration types was characterized by UV–Vis absorption spectroscopy and fluorescence excitation emission matrix (EEM)-parallel factor analysis (PARAFAC), and the fluorescent components of DOM in the soil and water samples were identified. Redundancy analysis (RDA) was used to identify the controlling factors of optical soil DOM characteristics. The results showed the following: (1) the DOC/SOC < 0.40 ± 0.04%, most HIX values < 3.0, most BIX values < 1 and the protein-like component ranged from 40 ± 0.03% to 54 ± 0.02%. These results indicate that the DOM content in the study area is low, and the soil has weak aromaticity and low humification. (2) Karst hydrodynamic processes control the migration and loss of DOM (the explanatory degree accounts for 23.53%). (3) In AL, the DOC/SOC is the lowest, and C3 (the molecular weight is the highest among the three components) accounted for the highest proportion compared to the other vegetation restoration types. Taken together, our results indicate that environmental and vegetation factors have important impacts on DOM migration. In rocky desertification areas with weak aromaticity and a low degree of humification, there is great potential for the soil to lose lightweight DOM. These results can help predict future changes in carbon allocation across landscapes due to changes in land use and hydrodynamic processes.