1 Introduction

Dissolved organic matter (DOM) is one of the key variables that drive soil and aquatic ecologies. The change of DOM in quantity and composition can directly affect carbon cycle, pollutant behavior, and microbial activity (Lapierre et al. 2013; Lou et al. 2018; Stegen et al. 2018). Recently, the change of DOM due to the frequent wildfires in forest watersheds was reported to pose a threat to water quality and affect drinking water treatment (Hohner et al. 2016, 2019). The changes of DOM in quantity and composition could affect the water treatment process and increase the challenge of coagulation and filtration. The further formed disinfection byproducts that are regulated in drinking water may pose threat to human health. Pyrogenic carbon (e.g., biochar), pyrolyzed from biomass during the wildfires and production, has accumulated or applied in the soil and surrounding ecosystem, which gained broad interest due to its significant effect on carbon sequential and soil amendment (Rudolf et al. 2013; Bistarelli et al. 2021; Zhang et al. 2021). The pyrogenic carbon in the soil and aquatic system may affect the DOM pool. Since biochar is easy to produce and apply, biochar is often used to represent pyrogenic carbon in laboratory and field research. Biochar may change the characteristics of soil DOM. For example, in a batch experiment, pyrogenic carbon increased the release of DOM from the soil due to the increased pH, and the composition of the released DOM correlated with the size exclusion effect of pyrogenic carbon pore structure (Smebye et al. 2016). Similar results were confirmed in the agricultural field experiment (Zhang et al. 2017; Liu et al. 2011). For the black soil and fluvisol, the active surface of biochar colloids might have exchanged the humic acid-like DOM through physicochemical processes, after which the exchanged DOM was leached out from the soils.

Sand column experiments were conducted to verify the mechanisms of enhanced release of humic acid-like substance during biochar colloid transport. Compared with the sand column with only HA transport, the DOC of the sand column with HA transport followed by pristine biochars injection was increased from 3.71 mg L−1 to 5.31 mg L−1 (Fig. 3a). The result indicated that more HA was released during biochar colloid transport. In addition, the SEM images also showed that after biochar colloid transport (Fig. 3c), less HA was retained on the sand surface than that without biochar colloids (Fig. 3b) transport, revealing the enhanced release of HA. The paddy soil had a high content of Fe element (Table S2), with a high isoelectric point (Wang et al. 2012), making the paddy soil surface less negative than the black soil and fluvisol. Due to the relatively weaker electrostatic repulsion between biochar colloids and paddy soil, the biochar colloids showed weak transport in the paddy soil column and exhibited little effect on DOM release. In addition, the large amount of Fe oxides in the paddy soil has a strong precipitation and immobilization ability for soil DOM, which might have also reduced the release of DOM. The sand column experiment also showed that Fe oxides decreased the release of HA. The DOC content in the Fe sand column was lower than that in the clean sand column (5.31 mg L−1 < 2.35 mg L−1) (Fig. 3a), which indicated that iron oxides decreased the release of HA. The SEM image also showed that much HA was retained on the sand surface in the column with Fe oxides (Fig. 3d). The less negative and rougher sand surface with Fe oxides than that without iron oxides decreased the release of HA.

Fig. 3
figure 3

The dissolved organic carbon (DOC) in the effluent (a) and the SEM images of quartz sand (b-d) in sand column experiments at pH 7. A: 10 mg L−1 humic acid (HA) in clean sand; B: 100 mg L−1 biochar colloids in clean sand; C: 10 mg L−1 HA followed by 100 mg L−1 pristine biochar colloids in clean sand; D: 10 mg L−1 HA followed by 100 mg L−1 pristine biochar colloids in iron oxides coated sand (Fe sand)

Full size image

3.3.3 Molecular characterization of released DOM

The FT-ICR-MS can provide detailed structural information on every molecule and the dynamics of the DOM released from soils (Sun et al. 2020; Liu et al. 2021; Hu et al. 2023). In biochar colloid transport experiments, relatively higher O/C, DBE, AImod, and NOSC values, and lower H/C values were observed in all soil leachate (Table 1), indicating that the transport of biochar colloids increased the unsaturation, aromaticity, and oxidation of the leached DOM. All formulas of the molecules mainly contained four subcategories based on their elemental composition: CHO, CHON, CHOS, and CHONS. CHO (relative abundance of 53%–64%) and CHON (relative abundance of 22%–38%) were the dominant formulas in the soil leachate (Fig. 4a and Table S4).

Table 1 Element (C, H, O, N, S) mass fraction, H/C, O/C, AImod, DBE, NOSC derived from assigned molecule of FT-ICR-MS from soil column leachate
Full size table
Fig. 4
figure 4

Relative abundance of formulas element classes (a) and compound classes (b) derived from van Krevelen diagram of DOM from soil column leachate. BS, FS, and PS represent black soil, fluvisol, and paddy soil, respectively

Full size image

The van-Krevelen diagram was used to present a graphic plot of compound distribution (Fig. S6), which revealed the changes in DOM molecules caused by the transport of biochar colloids. Lipid-like, protein-like, carbohydrate-like, lignin-like, tannin-like, condensed aromatic-like, and unsaturated hydrocarbon-like components were detected in the van-Krevelen diagram. The lignin-like component had the highest relative abundance in all soil leachate probably because of the degradation of terrestrial-plants sources (Fig. 4b; Ohno et al. 2010; Liu et al. 2021). The transport of pristine and aged biochar colloids increased the relative abundance of condensed aromatic-like and tannin-like compounds in the leachate from the black soil and fluvisol. The relative abundance of condensed aromatic-like compounds in the BS soil leachate increased from 9.44% to 15.24% and 15.69% with the transport of pristine and aged biochar colloids, respectively (Fig. 4b and Table S5). In the paddy soil, the relative abundance of the seven compounds showed minimal change, consistent with the slight change shown by the EEM spectra. Nevertheless, the relative abundance of condensed aromatic-like and tannin-like compounds in the paddy soil leachate slightly increased from 3.31% to 4.43% and 5.71%, and from 5.51% to 6.04% and 8.48% for pristine and aged biochar colloids, respectively. These results revealed that the transport of both pristine and aged biochar colloids enhanced the leaching of condensed aromatic-like and tannin-like compounds in all soils.

The condensed aromatic-like and tannin-like compounds could originate from biochar colloids and soils. To elucidate the sources of the increased DOM compounds in the leachate, we compared the number and types of molecules before and after biochar colloid transport in the three soils. Compared with the column experiments without biochar colloid transport, the molecule number of DOM in the leachate from pristine and aged biochar colloid transport experiments showed 1898, 1602, 2561, 1315, 1888, and 1768 newly emerged molecules, 1929, 1705, 2583, 2031, 1811, and 1472 vanished molecules, and 6724, 5923, 6336, 6026, 6369, and 7105 resistant molecules (Fig. S7). These molecules plotted in van Krevelen diagrams are shown in Fig. 5a-f. Since the molecules in pristine and aged biochar colloids were removed during the selection processes, these “New/Increased” and “Vanished/Decreased” molecules were from soils, not biochar colloids.

Fig. 5
figure 5

The van Krevelen diagrams (a-f), the number of “Vanished/Decreased” (g) and “New/Increased” (h) molecule in soil column leachate after biochar colloid transport. “New”, “Increased”, “Decreased”, and “Vanished” represent the newly emerged, intensity increased, intensity decreased, and vanished molecule after biochar colloid transport; BS, FS, and PS represent black soil, fluvisol, and paddy soil, respectively

Full size image

The “New/Increased” molecules with high O/C and low H/C ratios and “Vanished/Decreased” molecules with low O/C and high H/C ratios revealed the increased release of molecules with large molecular weight. The molecules released from soils with low H/C ratio suggested their low lability (Liu et al. 2009; Gu et al. 2019), which could change the soil organic carbon pool and alter the microbial communities since carbon lability affects microbial metabolic process. In addition, the results revealed that the transport of biochar colloids increased the release of soil DOM molecules with large molecular weight. This was different from the results in the previous studies, which reported that molecules with low molecular weights were easily leached out during precipitation and irrigation (Sun et al. 2022; Hu et al. 2023). This might be because the biochar colloids change the soil DOM toward molecules with larger sizes by sorbing smaller aliphatic species (Smebye et al. 2016). The elemental composition and compound distribution showed 191 “Vanished/Decreased” molecules (Fig. 5g) and 520 “New/Increased” molecules (Fig. 5h) in all soils after biochar colloid transport. CHO and CHON were the main formulas in “New/Increased” and “Vanished/Decreased” molecules. The lignin-like compound was the most abundant in “New/Increased” and “Vanished/Decreased” molecules. Condensed aromatic-like and tannin-like compounds had more “New/Increased” molecules with few “Vanished/Decreased” molecules, indicating that they were the most released DOM compounds during biochar colloid transport.

The optical characteristics of DOM in the EEM spectra were related to its molecular composition in the FT-ICR-MS. The EEM-PARAFAC model decomposed the EEM spectra into three individual characterized fluorescence components of DOM: C1 (at Ex/Em of 250 nm/ < 450 nm), C2 (at Ex/Em of < 250 nm/ < 450 nm), and C3 (at Ex/Em of < 270 nm/480 nm). C1 and C2 were microbial humic acid-like species with anthropogenic origin usually found in agriculturally dominated watersheds, while C3 was a typical terrestrial humic acid-like component primarily found in soil-derived DOM (Murphy et al. 2014; Peleato et al. 2016). The correlations between the three components (C1, C2, and C3) and the relative abundance of “New & Increased” and “Vanished & Decreased” formulas are presented in Fig. S10. There were 491, 281, and 421 “New & Increased” formulas significantly correlated to C1, C2, and C3 components, of which 64.6%, 6.1%, and 92.9% were positively correlated to C1, C2, and C3 components, respectively. This indicated that these formulas were mainly positively correlated with C1 and C3 components. For the C1 and C3 components, the positively correlated formulas were mainly distributed in the regions with high O/C and low H/C ratios and consisted of lignin-like (57.0% and 60.1%), condensed aromatic-like (22.8% and 19.4%), and tannin-like (18.4% and 18.9%) compounds. There were 297, 251, and 191 “Vanished & Decreased” formulas significantly correlated to C1, C2, and C3 components, of which 44.1%, 96.0%, and 2.1% were positively correlated to C1, C2, and C3 components, respectively. This showed that the formulas were mainly positively correlated with C1 and C2 components. For C1 and C2 components, the positively correlated formulas were mainly distributed in the regions with low O/C and high H/C ratios and contained lignin-like compounds (79.4% and 77.2%). The strong correlation between the optical characteristics and the molecular composition of DOM indicated the increased leaching of humic acid-like substances that mainly consisted of lignin-like, condensed aromatic-like, and tannin-like compounds. The decreased leaching of humic acid-like substance showed that the main component was a lignin-like compound.

3.4 Environmental implications

The increased release and downward movement of DOM may affect the soil carbon pool, microbial communities, and water quality, which is tightly related to composition and structure of DOM (Kellerman et al. 2015; Roth et al. 2019). Generally, the DOM enriched in components with a high H/C ratio (e.g., proteins and sugars) is considered more labile, whereas DOM composed of high O/C ratio structures (e.g., aromatic-like) is less labile (Sun et al. 2020). The most labile fractions of DOM are usually consumed firstly by the microbes, followed by less labile fractions. Our study showed that the DOM components with higher aromaticity were released during the transport of biochar colloids (Fig. 6), and these released DOM could not be metabolized by microbe or further change the microbial community in short term. However, the long-term effect of DOM with higher aromaticity should not be neglected. For example, an incubation experiment showed that after the initial faster consumption of labile DOM, the aromatic DOM tends to be more selective for the microbial community (Zhou et al. 2021). In addition, the degradation processes of DOM in lakes preferentially remove aromatic compounds rather than reduced and aliphatic compounds (Kellerman et al. 2015), and the light penetrates water preferentially degrading aromatic compounds (Stubbins et al. 2010). Thus, the different degradation process in aquatic systems compared with soils indicates that more attention should be paid to the watershed and water quality in regions with biochar application and frequent wildfires. Additionally, the high aromaticity and strong electron transfer ability of released DOM could change the transport and degradation of organic pollutants in aquatic systems (e.g., antibiotics) (Lou et al. 2018), and the preferential adsorption of aromatic compounds on mineral surfaces (Coward et al. 2019) may further affect pollutant behaviors and carbon sequestration in soils.

Fig. 6
figure 6

Schematics of the enhanced soil DOM release by biochar colloid transport

Full size image

4 Conclusion

This study showed the optical and molecular evidence highlighting the vital role of biochar colloid transport in soil DOM release. Aging process alters the properties of biochar colloids and slightly enhances their transport in black soil and fluvisol. The significant DOM release in three soils differed slightly in response to pristine and aged biochar colloid transport, revealing the critical role of both pristine and aged biochar colloids in soil DOM release. The minor effect of chemical aging on soil DOM release enhanced by biochar colloids revealed that chemical aging was not a detrimental factor contributing to the increased soil DOM release. Various aging processes may coexist in the natural environment, and further research is needed to investigate their effect on soil DOM release. Our study could provide a reference for evaluating the environmental aging process on the soil DOM release enhanced by biochar colloids during rainfall and irrigation. The different results between the adsorption and transport experiments indicated that the prediction of DOM release based on adsorption experiments might be inadequate in real fields where precipitation and irrigation frequently occur. The effect of biochar application on soil DOM release may be underestimated in areas with high rainfall and irrigation since soil DOM release could be enhanced during biochar colloids downward movement. The increased release of humic acid-like substance had large molecular weights and contained lignin-like, condensed aromatic-like, and tannin-like compounds, which had important implications for the soil organic carbon pool, microbial communities, and groundwater systems. Thus, more attention should be focused on the increasing accumulation of biochar colloids in the soil and aquatic environments due to frequent wildfires and artificial biochar application. The long-term effect of biochar field application on DOM release remains largely unknown and deserves further attention.