1 Introduction

Rice is one of the most significant food crops in the world, with a large cultivation area that is critical to ensuring global food security (Food and Agriculture Organization of the United Nations 2021). China has a large rice cultivation area, with more than 29% of the grain crops (i.e., rice, wheat and corn) acreage (National Bureau of Statistics 2021). However, agricultural waste such as rice straw is produced after harvest, and the utilization of straw resources has received increasing attention in China. Straw returning is a highly efficient way to use rice straw and is advantageous in reducing fertilizer use and pollution (Yin et al. 2018; Zheng et al. 2023). In China, the comprehensive utilization of crop straw was about 80%, and the use of straw returning reached about 400 million tons each year (Ministry of Agriculture and Rural Affairs 2021). Rice straw decomposition by soil microorganisms can improve the physical, chemical, and biological properties of soil (Huang et al. 2017; Singh et al. 2007; Chen et al. 2014, 2017). As an essential source of organic fertilizer, rice straw is rich in N, P, K, and other nutrients, as well as considerable amounts of soil organic matter (SOM) (Chen et al. 2022; Hao et al. 2008; Sun et al. 2023). Therefore, the straw returning can improve rice yields by increasing organic matter, bioavailable nitrogen and potassium content in the soil (Xu et al. 2010; Zhou et al. 2020). In addition, the decomposition of straw will produce dissolved organic matter (DOM), which affects largely the photochemistry of paddy water with poorly explored underlying mechanisms in the field (Ren et al. 2021; Liu et al. 2023).

In the aquatic environment, nitrate (NO3−), nitrite (NO2−), and DOM all generate reactive intermediates (RIs) such as •OH, 1O2 and 3DOM* under light (Vione et al. 2014; Zafiriou and True 1979b, a). Unlike these commonly aquatic systems, it has been reported that DOM mainly accounted for RIs generation in paddy water (Zeng et al. 2021). During the straw returning process, the decomposition of straw not only increased DOM concentration but also changed the structure of DOM, ultimately altering RIs generation in paddy water under light (Zeng et al. 2022). Since the properties of DOM formed from straw decomposition significantly differed from those of DOM dissolution from soil organic matter. For example, the straw-derived DOM generally has higher phenolic hydroxyl groups with larger molecular size than that of natural organic matter, which was favorable to form H2O2 as a photosensitizer (Ma et al. 2015). A recent study reported that straw returning would increase DOM concentration in paddy water during rice cultivation, which promoted RIs production and enhanced the photo transformation of organic pollutants (Liu et al. 2021). However, most of the previous studies were conducted in stimulation conditions, and the field experiments regarding the photochemical process of paddy water influenced by straw returning were poorly understood.

As a typical nicotinic pesticide, imidacloprid (IMD) is commonly utilized in rice agriculture to kill pests (Stevens et al. 2008). However, IMD easily enters soil and paddy water after several years of application (Wamhoff and Schneider 1999; Thuyet et al. 2011), leading to a toxic level of IMD concentration (≥ 200 μg/L) to aquatic arthropods in agricultural water (Zeng et al. 2023). The toxicity of IMD is transformation-dependent. The IMD can be degraded via direct photolysis, microbiological degradation, and biotransformation in the naturally occurring environment (Acero et al. 2019; Pang et al. 2020; Anderson et al. 2015). Some IMD transformation products such as IMD-desnitro, 6-chloronicotinic acid and IMD-olefin desnitro were more toxic than IMD (Xue et al. 2023; Malev et al. 2012). It has been reported that RIs such as •OH, 1O2 and 3DOM* exhibited high reactivity with IMD of second-order reaction rate constants of 2.52 × 109 M−1 s−1, 3.78 × 106 M−1 s−1, and 3.77 × 108 M−1 s−1 (Zeng et al. 2023). Therefore, it is reasonably hypothesized that the impact of RIs by straw returning will ultimately affect IMD transformation in paddy water. The underlying mechanisms during the abiotic transformation process of IMD is rarely explored.

Therefore, the main purpose of this study was to explore the photochemistry of paddy water as affected by rice straw returning during the different stages of rice growth through a field experiment. Ultraviolet–visible spectroscopy (UV–vis), fluorescence emission-excitation matrix (EEM) spectroscopy-parallel factor analysis (PARAFAC), and Fourier Transform Ion Cyclotron Resonance Mass Spectrometry (FT-ICR MS) were used to characterize DOM properties. Chemical probes combined with the kinetic model were used to determine the quantum yields and steady-state concentration of RIs. Structural equation model was used to investigate the dominant pathways of straw returning impact on RIs formation. In addition, we evaluated the effect of rice straw returning on indirect photodegradation of IMD transformation in paddy water. LC–MS/MS and ECOSAR program were applied to characterize IMD degradation products and their potential toxicity.

2 Materials and methods

2.1 Chemicals

All chemicals used were listed in Text S1.

2.2 Experimental site and weather details

The field experiments were performed at the experimental station of Guixi, Yingtan, Jiangxi, China (177°10′51.75″ N, 28°14′39.52″ E) from June to September 2022. The climate is categorized as subtropical with a monsoon zone, with a mean temperature of 27.9°C and a relative humidity of 74.5% during the experiment.

2.3 Experimental design and field management

The field experiments were conducted to investigate the effect of rice straw returning on paddy water properties and photochemical processes. The properties of the soil and rice straw used in this field study are shown in Tables S1 and S2. The total application dose of nutrient element was 15 kg ha−1 including organic source (rice straw, RS), and chemical fertilizer (CF) containing compound fertilizer (N: P2O5: K2O = 25: 10: 16). All treatments had the same amount of nitrogen fertilizer with different ratios of RS and CF as follows: CK (100% CF); 25%RS (25% RS + 75% CF); 75%RS (75%RS + 25%CF) (Table S3). Each treatment was repeated three times, and the area of each replicated plot was 4 m2 (2 m × 2 m).

The Zhongzu45, a medium late-maturing Geng rice widely consumed in the local area was used in this study. The fertilizer was applied to the paddy before the rice was transplanted, and then the rice was transplanted in rows 20 cm × 20 cm apart. Paddy water was collected on the 10th, 20th, 30th, 50th, and 70th day after fertilization. All samples were passed through a 0.45 µm membrane and placed in a refrigerator at 4 °C for subsequent analyses and experiments.

2.4 Paddy water parameters

The concentration of dissolved organic carbon (DOC) was measured by a Vario Select TOC analyzer (Elementar, Germany). The UV–vis absorption and fluorescence EEM spectra were obtained using a UV/Vis spectrophotometer (Shimazu, Japan) and F–7000 spectrofluorometer (Hitachi, Japan), respectively. The EEM-PARAFAC model was used to calculate the DOM bulk characteristics. FT-ICR MS (Bruker, Germany) was used to examine molecular structure of DOM with an in-house TRFu tool (Fu et al. 2020). More information can be found in Text S2-4.

2.5 Photochemical experiments

All photochemical experiments were conducted in a photoreactor equipped with a 500 W xenon lamp and a 290 nm filter. Circulating cooling water was used to maintain a constant reaction temperature. 2,4,6-Trimethylphenol (TMP, 4 mM), furfuryl alcohol (FFA, 2 mM), and terephthalic acid (TPA, 10 mM) were used to probe 3DOM*, 1O2 and •OH, respectively. Periodically, probe concentration was analyzed by high-performance liquid chromatography (Agilent 1200, USA) with reaction times of 0, 0.25, 0.5, 1 and 1.5 h. The quantum yields and steady-state concentrations ([RI]ss) of 3DOM* (fTMP),•OH (\({\Phi }_{{\text{OH}}}^{\cdot }\)), and 1O2 (\({\Phi }_{{}^{1}{{\text{O}}}_{2}}\)) were determined using chemical probes combined with the kinetic model. The detailed information on RI analysis was presented in Text S5 and Text S7.

Batch degradation experiments were conducted in a photoreactor with 10 mL of paddy water and IMD concentration of 10 μM, which is much higher than agricultural water. The solution was irradiated under a 500 W xenon lamp with a 290 nm filter. The concentration of IMD was determined using high-performance liquid chromatography outfitted with an LC-18 column and a diode array detector (Agilent, USA). A C18 column was used to separate degradation products, and an X500R LC-quadrupole-time-of-flight Premier mass spectrometer (Applied Biosystems, USA) was used for analysis. Toxicological predictions for IMD and its degradation products were evaluated using the ECOSAR software.

2.6 Data analysis

Statistical analysis was accomplished using Origin 9.8 (OriginLab Corporation, USA) and MATLAB R2018b (MathWorks Corporation, USA). The degradation of IMD and probes were fitted the pseudo-first-order kinetic equation: ln (Ct/C0) = kobs, where C0 and Ct are the concentration of target compounds at time 0 and t, respectively, and kobs is the corresponding pseudo-first-order rate constant ( h−1). Structural equation model (SEM) was established by IBM SPSS Amos 26.0 (International Business Machines Corporation, USA), and detailed information on SEM was presented in Text S8.

3 Result and discussion

3.1 Effects of straw returning on the formation of RIs in paddy water

The steady-state concentrations of RIs ([RI]ss) of each treatment in paddy water during rice cultivation were determined (Fig. 1a-c). Compared with the application of chemical fertilizer alone, straw returning greatly enhanced RIs generation, and it was observed that [•OH]ss, [3DOM*]ss and [1O2]ss increased by 14.84%—247.98%, 7.71%—125.68%, and 24.22%—252.07%, respectively. But the significant variations of RI concentration were also observed among the different straw dosages and rice growth stages. For instance, on the 10th day of rice growth during the seedling stage with straw returning, [•OH]ss, [3DOM*]ss and [1O2]ss increased by 32.30%—65.02%, 28.33%- 40.76% and 39.14%—48.11%, respectively. However, on the 20th day of the seedling stage, the change of [•OH]ss was insignificant, while [3DOM*]ss and [1O2]ss continuously increased by 7.71%—39.23% and 78.79%—99.53%, respectively. From the 40th day of the tillering stage, the RI concentration was decreased, but the straw treatment still increased RI concentration by 24.22%—252.07%. Moreover, the increase in straw dosage from 25 to 75% could not change RI concentration significantly, especially for [•OH]ss at the early seedling stage. The quantum yields of different RI were also determined (Fig. 1d-f). The values of fTMP, \({\Phi }_{{\text{OH}}}^{\cdot }\), and \({\Phi }_{{}^{1}{{\text{O}}}_{2}}\) in paddy water varied in the ranges of 9.59—70.81, (0.41—6.37) × 10–5, and (0.73—5.80) × 10–2, respectively. These results suggested that quantum yield and steady-state concentration of RIs have different variation trends. During 10—20 days after fertilization, the quantum yield of RIs first increased and then decreased obviously. These results suggest that straw returning greatly enhanced RI generation, but the changes in RI concentration were different, which was dependent on rice growth stage and RI species. The probable reason was that the changes in paddy properties during rice growth with straw returning, and will be discussed in detail in following section.

Fig. 1
figure 1

Steady-state concentrations ([a]: •OH; [b]: 3DOM*; [c]: 1O2) and quantum yields ([d]: •OH; [e]: 3DOM*; [f]: 1O2) of RIs in paddy water with different treatments. Conditions: CK (no treatment), 25% rice straw (25%RS), 75% rice straw (75%RS)

Full size image

3.2 Effects of straw returning on DOM properties in paddy water

3.2.1 Bulk DOM properties

The bulk DOM properties of paddy water were investigated by UV–vis and EEM, including fluorescence index (FI), biological index (BIX), humification index (HIX), E2:E3, CDOM/DOM, and SUVA254. As shown in Fig. S4b-c, FI and BIX of DOM were mainly distributed in the ranges of 2.0–2.5 and 0.7–0.9, respectively, which suggested a vital contribution of autochthonous (i.e., microbial/algae) sources to DOM (Jiang et al. 2017; Birdwell and Engel 2010; Li et al. 2019). The molecular mass of DOM negatively correlated with E2:E3 (Wang et al. 2009), and CDOM/DOM ratio could be used as an index of CDOM (Chromophoric Dissolved Organic Matters) content (Rochelle-Newall and Fisher 2002). It was observed that rice straw returning increased the molecular size of DOM and the relative content of CDOM in paddy water, which was ascribed to the decomposition of lignin and fiber from rice straw (Yadvinder et al. 2005). Meanwhile, DOM in the paddy water derived from rice straw had a higher DOM molecular mass with more humified, which is not favorable for the electron transfer processes for RIs formation (Maizel and Remucal 2017), and thus the lowest values of fTMP, \({\Phi }_{{\text{OH}}}^{\cdot }\), and \({\Phi }_{{}^{1}{{\text{O}}}_{2}}\) were observed on the 30th day of rice growth. In addition, the ratio of CDOM gradually declined after 30 days due to the instability of organic chromophores under solar irradiation (Sulzberger and Durisch-Kaiser 2009), which resulted in the decrease of RIs as observed in Fig. 1a-c.

SUVA254 could be used as an index of aromatic compounds in DOM (Weishaar et al. 2003), and it was observed that the decomposition of rice straw resulted in forming of aromatic organic molecules with a high level of SUVA254 (Fig. S4d). After 20 days of the seedling stage, SUVA254 decreased markedly due to the consumption and hydrolysis of aromatic compounds during the straw decomposition. It has been reported that DOM with charge-transfer complexes of aromatic groups was thought to be caused the long-wavelength absorption, which did not induce the triplet-state photochemistry (Sharpless and Blough 2014). Therefore, after 30 days of the seedling stage, the quantum yield of RIs increased with time as observed in Fig. 1d-f.

In addition, [DOC] of paddy water followed the order: 75%RS (56.22 mg/L) > 25%RS (38.64 mg/L) > CK (34.26 mg/L) on the 10th day, which indicated that the application of rice straw efficiently increased DOM concentration due to straw decomposition. However, [DOC] decreased dramatically from 12.78 mg/L to 33.28 mg/L on the 20th day, while increased significantly from 35.70 mg/L to 75.92 mg/L on the 30th day. The probable reason was the rapid increase of soil microbial community upon straw return, leading to the consumption of organic carbon by microorganisms (Yan et al. 2020; Chen et al. 2010; Zhu et al. 2023). Similar results were also observed in the previous studies (Wu et al. 2022; Zhang et al. 2020). Furthermore, the high DOM concentration would not account for high RI generation on the 30th day, since RI generation was not only dependent on DOM concentration but also significantly influenced by DOM properties (Page et al. 2014; Maizel and Remucal 2017). These combined results suggested that DOM changed significantly during the rice growth with straw returning, which markedly influenced the generation of RIs.

3.2.2 EEM-PAPAFAC compounds

The EEM-PARAFAC approach was used to further investigate how the DOM characteristics changed during the straw returning process. As shown in Fig. S5a-d, four fluorescence components were separated, and C1 [Ex/Em = 240(325)/415 nm] corresponded to terrestrial-derived UVA (Stedmon et al. 2003) and anthropogenic UVC (Stedmon and Markager 2005) humic acid-like compounds, respectively. The C2 [Ex/Em = 435/500 nm] represented soil-like fulvic acid compounds from irrigated agriculture in freshwater environments (Singh et al. 2010). The C3 [Ex/Em = 250(380)/444 nm] for humus-like substances, was related to degraded humus-like substances (Cory and McKnight 2005). The C4 [Ex/Em = 240 (285)/346 nm] could indicate tryptophan-like fractions, which are indicative of intact proteins or less catabolic peptides (Murphy et al. 2006).

The Fmax values of four components were shown in Fig. S5e–h. It was observed that Fmax value of 75% RS treatment was higher than that of CK during the 10–20 days, while it was insignificant between 25% RS treatment and CK. On the 30th day, the Fmax value of C1-C4 with 75% RS was significantly higher than that of 25% RS or CK. Meanwhile, the soil organic matter content of 75% RS treatment was higher than 25% RS treatment, owing to the positive correlation between the amount of rice straw returning and SOM accumulated (Table S1) (Lao et al. 2003), which impacted DOM properties in paddy water. After the 30th day, the Fmax values of each component gradually decreased due to the consumption of soil nutrients by plant growth and microbial activities (Zhang et al. 2020). Humus and fulvic acid had a strong absorption in the visible range, which was favorable to RI formation by absorbing photos efficiently (Frimmel 1994; Sakkas et al. 2002). These results indicated that the rice straw returning favored the accumulation of humus, fulvic acid, and tryptophan-like substances, which would enhance DOM photochemistry in paddy water.

3.2.3 Chemical composition of DOM

FT-ICR MS was used to analyze variations of DOM property at the molecular level with straw returning processes. It was observed that 3885 molecular formulas were identified for CK on the 10th day, while 5610 and 4291 molecular formulas were identified in 25% RS and 75% RS, respectively. These results suggested that the straw returning increased the diversity of DOM molecular composition in paddy water. The identified molecular formulas could be classified into different groups by the molar ratio of H/C and O/C, including carbohydrates, amino sugars, saturated compounds, tannins, lignins, unsaturated hydrocarbons, and condensed aromatic structures (Fu et al. 2020; Wu et al. 2023). The molecular formulae only observed in 25% RS and 75% RS are highlighted by red dots in Fig. 2. On the whole, straw returning greatly improved DOM abundance of amino sugars, saturated compounds, tannins and lignin components. However, the application of 25% RS was more efficient than that of 75% RS in improving DOM quantity of lignin components. The reason was attributed to the fact that sufficient nitrogen inorganic fertilizer was used in 25% RS to keep the balance of N utilized by microorganisms, and thus accelerated the mineralization of rice straw in the stage of concentrated straw decomposition (Fang et al. 2018a, b). As shown in Table S9, DOM intensity-weighted average parameters were also used to evaluate the effects of straw returning for DOM molecular properties. H/CW, O/CW, S/CW and P/CW ratios could indicate elemental ratios of DOM (Liu et al. 2011). The modified aromaticity index (AImod) was used to indicate aromatic formulas (Lu et al. 2015). The double-bond equivalence (DBEW) could represent the number of unsaturated compounds (Zhou, Liu, et al. 2019). Compared with CK, 25% RS had a higher value of DBEW. However, 75% RS had a lower value of DBEW than CK. Those indicated application of a low amount of rice straw would produce more aromatic formulas with appropriate C/N for microorganisms. Meanwhile, straw returning significantly enhanced S/CW and P/CW ratios in DOM formulas, owing to the decomposition of straw that produced more heteroatoms (e.g., S and P), which was rich in N, P, K, and other nutrients (Wang et al. 2022b). Briefly, straw returning would produce more amino sugars, saturated compounds, tannins and lignin components. Meanwhile, DOM would change into more aromatic formulas with more heteroatoms. These results further confirmed that rice straw returning in paddy fields considerably altered the amount and property of DOM in paddy water, which impacted RIs formation.

Fig. 2
figure 2

Van Krevelen of classification diagram of molecular formulas in paddy water with different rice straw application amounts. The cyan dots represent the molecules in CK-10d, and the red dots represent the molecules only in 25%RS-10d (a) or 75%RS-10d (b)

Full size image

Spearman correlation analysis was used to identify the relationship between the formation of RIs and the molecular properties of DOM (Fig. S7). The results showed that H/Cw ratio was negatively correlated with \({\Phi }_{{\text{OH}}}^{\cdot }\) (Spearman's R = -0.99, p < 0.05). As shown in Table S9, 25% RS with a higher \({\Phi }_{{\text{OH}}}^{\cdot }\) had a lower H/C ratio than CK and 75% RS on the 10th day. DOM with a high H/C ratio had more aliphatic formulas, which was not favorable to electron transfer for formation of •OH (Valle et al. 2018; Vione et al. 2014). AImod and DBE were positively correlated with \({\Phi }_{{\text{OH}}}^{\cdot }\) (Spearman's R = 0.96 and 0.99, p < 0.05). DOM had more aromatic formulas for the charge-transfer pathway of DOM photochemistry (Gonsior et al. 2009; Gaberell et al. 2003). Meanwhile, the application of a low amount of rice straw would produce more aromatic formulas in paddy water, which speeded up formation of RIs through promoting electron transfer.

3.3 Effects of straw returning on the mechanism of RIs generation in paddy water

The implications of various amounts of rice straw on the mechanism of RIs generation were investigated, as determined by the analysis of the relationship between DOM properties and RI generation. One important source of •OH generation in the aquatic environment is thought to be the photolysis of NO3− and NO2− (Vione et al. 2014). However, the application of rice straw had little impact on NO3−/NO2− concentration in paddy water, since the NO3−/NO2− generated from the decomposition of straw is limited (< 0.12 mg/L), which would not have a significant contribution to •OH generation as reported in the previous study (Zeng et al. 2022). In addition, it had been reported that there was no significant difference in NO3−/NO2− concentration in paddy water of each straw returning treatment during rice growth stages because most of the nitrogen applied to the soil was absorbed by plants or utilized by microbial activities (Liu et al. 2023). Therefore, DOM was the predominant source of RIs in this study.

The structural equation model was used to identify the main factors controlling RIs generation in paddy water during rice growth with straw returning processes. The rice growth stage (cultivation time) was significantly negatively correlated with FI, with a coefficient of -0.521, indicating that the FI decreased as rice growth (Fig. 3a). In addition, 25% rice straw was significantly negatively correlated with E2:E3, with correlation values of -0.548, indicating that DOM produced by a low amount of rice straw was macromolecular size. This might be due to the suitable C/N ratio of 25% RS treatment for microbial activity (Siedt et al. 2021). Both the 25% and 75% rice straw were significantly positively correlated with CDOM/DOM, with correlation values of 0.531 and 0.417, respectively, suggesting that straw returning increased CDOM content in paddy water. Compared with 75% RS treatment, the application of 25% RS had a relatively stronger enhancement. This might be attributed to the fact that a low amount of rice straw led to intense microbial activities with the appropriate C/N, which was beneficial for CDOM generation through microbial mineralization of organic matter in paddy water (Nelson and Siegel 2013; Fang et al. 2018a, b). 75% RS with a high C/N ratio can limit microbial activities, due to the limited production of certain enzymes for microorganisms through ammonia metabolite repression (Fog 1988). Another reason was that 75% RS with an exogenous carbon supply could decrease carbon priming effects for microorganisms by alleviating nutrient mining, which decreased the diversity of bacterial and fungal communities (Fang et al. 2018a, b). Figure 3b showed that incubation time was adversely associated with CDOM/DOM, with correlation values of -0.516. This indicated that the content of CDOM in paddy water decreased gradually with incubation time due to photobleaching (Helms et al. 2013).

Fig. 3
figure 3

Structural equation models of [RI]ss ([a]: ·OH; [b]: 3DOM*; [c]: 1O2) and standard total effects on [RI]ss (d)

Full size image

Furthermore, the structural equation model results showed the negative correlation between •OH formation and E2:E3, with the correlation coefficient of -0.58, which indicated that macromolecule DOM contributes to the formation of RIs due to the light absorption rate of macromolecule organic matter with a high humification degree (Maizel et al. 2017; Bianco et al. 2014). [1O2]ss and [3DOM*]ss were significantly positively correlated with [DOC], with a coefficient of 0.47 and 0.35, respectively, attributing to the fact that the photochemical processes of DOM in paddy water were an essential source of RIs (Guerard et al. 2009). Figure 4a showed that [1O2]ss was significantly negatively correlated with CDOM/DOM, with a coefficient of 0.39 (P < 0.05), because CDOM was an important component of paddy water that absorbed photons and subsequently, mediating a series of photochemical reactions (Vione et al. 2014). For example, Wu et al. (2022) examined the photochemical changes of DOM in farmland soil after straw return and found that the quantum yields of 3DOM* and 1O2 increased with the decomposition of straw (Wu et al. 2022). These results corresponded to the changing pattern of the influence of straw returning on RIs quantum yield in paddy water in a rice field.

Fig. 4
figure 4

a Experimental first-order rate constants of IMD phototransformation in irradiated paddy water samples, (b) kinetics of IMD degradation in the sample 25%RS-10d subjected to different treatments (CH3OH 100 mM; FFA 5 mM; TMP 5 mM)

Full size image

The standard direct and indirect effects of RIs by incubation time and various straw returning treatments were shown in Fig. 3d. The standard total effects of 25% RS and 75% RS were 0.19–0.28 and 0.20–0.30, respectively, which suggested that straw returning could increase the formation of RIs in paddy water. The 75% RS treatment had a net contribution to [•OH]ss of less than 0.1, suggesting that DOM produced by using a higher proportion of rice straw had a weaker impact on •OH formation. In conclusion, the organic replacement ratio of 25% rice straw treatment can significantly promote the formation and accumulation of RIs. In comparison, the organic replacement ratio of 75% rice straw treatment only strongly enhanced the formation of 3DOM* and 1O2.

3.4 Effects of straw returning on IMD photodegradation in paddy water

3.4.1 Kinetics of IMD Degradation

The straw returning influenced RIs generation and thus would also impact IMD indirect degradation mediated by RIs. Therefore, in order to investigate the effect of straw returning on indirect photodegradation of IMD, the IMD (10 μM) was added to paddy water on the 10th and 30th days after fertilization. The contribution of RIs to IMD degradation was measured by the shading coefficient \({S}_{\lambda }\), and the calculation methods of \({S}_{\lambda }\) and the indirect photodegradation rate constant \({k}_{obs,IMD}\) were shown in Text S6. As shown in Fig. 4a, the photodegradation rate of IMD in paddy water without straw returning was 1.72 × 10–2 h−1, while it was markedly increased to 6.28 × 10–2 h−1and 5.56 × 10–2 h−1 on the 10th day for 25% RS and 75% RS, respectively, which indicated that RIs promoted the indirect photodegradation of IMD in paddy water. On the 30th day, kobs,IMD increased by 1.87 and 1.38 times for 25%RS-30d and 75%RS-30d, respectively, although [RIs]ss decreased. These combined results indicated that rice straw application significantly accelerated IMD degradation.

Quenching experiments were further used to investigate the contribution of each RIs to IMD degradation. Our previous study found that 3DOM*, •OH and 1O2 were dominant ROS in paddy water under irradiation (Zeng et al. 2021). Although other ROS such as H2O2 and O2•– would be generated during the reaction, they were quickly converted to •OH under irradiation (Vione et al. 2014). Therefore, we only evaluated the variation of 3DOM*, •OH and 1O2 under straw returning during the different stages of rice growth, and the contribution of each RIs to IMD degradation. Methanol was an effective quencher of •OH with a second reaction rate constant of 9.7 × 108 M−1 s −1 (Fang et al. 2013). Figure 4b showed that the addition of CH3OH inhibited IMD degradation, revealing that •OH contributed about 24.7% of IMD degradation. FFA and TMP were chosen as quenchers for 1O2 and 3DOM* with second reaction rate constants of 8.3 × 107 and 8.1 × 108 M−1 s −1, respectively (Zhou et al. 2017). We estimated that the contributions of 1O2 and 3DOM* for degradation of IMD were about 28.5% and 75.7%, respectively. The indirect photodegradation of IMD was nearly inhibited by the addition of TMP (quencher of 3DOM*), demonstrating that 3DOM* was a crucial RI to IMD degradation. DOM could also induce the formation of •OH and 1O2, and it absorbed photons to DOM+•, and subsequently induced the production of •OH through Fenton-like reactions, whereas O2 and 3DOM* could combine to generate 1O2. These results showed that the DOM photochemical process was essential for controlling pollutant degradation in paddy water (Fig. 5).

Fig. 5
figure 5

Possible degradation pathways of IMD degradation in paddy water

Full size image

3.4.2 Determination of IMD degradation products

Liquid chromatography-time-of-flight tandem mass spectrometry (LC-TOF–MS /MS) was used to detect IMD degradation products after 8 h irradiation. The indirect photodegradation of IMD by RIs would form mono-hydroxy IMD (P1, m/z = 271) (Wang et al. 2022a, b). Zeng et al. examined IMD degradation products by different RIs and found further oxidation products of P1 with 3DOM* (Zeng et al. 2023). However, a further hydroxylation product of P1 was not found in this study, which was attributed to the complex photochemical process in paddy water. As a precursor of RIs, DOM produced •OH and 1O2, which limited the generation of 3DOM*, and could not induce the formation of a further hydroxylation product of P1. Other oxidation products such as nitroso derivative of IMD (m/z = 241) and IMD-olefin (m/z = 209) were also identified in this study, while these products were not detected in the previous study with a similar process by a laboratory study (Zeng et al. 2023). In addition, IMD was transformed into a primary nitroso derivative (P2, m/z = 241), and RIs attack caused the N–NO bond of P2 to break to form a cyclic guanidine derivative (P3, m/z = 212), and then P3 could form IMD-olefin (P5, m/z = 209) by hydrogen extraction (Xue et al. 2023). The amidine group and the C-2/C-1 atom in the imidazole ring of P3 were further oxidized to the carbonyl group to form P4 (m/z = 226). The further oxidation of P3 produced another oxidation product, 6-chloronicotinal (P6, m/z = 141). These combined results suggested that IMD degradation had a similar mechanism to the previous studies, but the degradation products were different.

3.4.3 ECOSAR Analysis of the Degradation Products of IMD

ECOSAR program was used to assess the potential toxicity of IMD photodegradation products. These results were provided in Table S8. Half-lethal concentration (LC50) of fish and daphnia and half-effective concentration (EC50) of green algae were used to evaluate the toxicity of IMD photodegradation products. Green algae represented aquatic plants, while fish and daphnia were indexes of aquatic animals. Each photodegradation product's toxicity toward fish, daphnia, and green algae varied. The LC50 of IMD for fish and daphnia was 147 mg/L and 121 mg/L, respectively, while EC50 for green algae was 73.40 mg/L. Briefly, hydroxyl derivatives of IMD had less toxicity for aquatic animals and plants than parent IMD. Nitroso derivatives and their further oxidation product P3 were more toxic to aquatic animals but less toxic to aquatic plants. Although P5 formed by the hydrogen extraction reaction of P3 was still harmful to aquatic animals, P4 and P6 formed by P3 oxidation were less toxic to aquatic animals and plants. This study showed that P1, P4, and P6 were less toxicity than IMD. P2, P3, and P5 were less dangerous to aquatic plants than IMD but more toxic to aquatic animals. In conclusion, the intermediate compounds produced by the indirect photodegradation of IMD were less harmful to aquatic plants but posed a greater risk to aquatic animals.

4 Conclusion

Straw returning is widely used to increase soil fertility worldwide, particularly in China. The influence of straw returning on the photochemical process of paddy water in a field was largely unknown. This study presents a field experiment to examine how straw returning affected RIs formation and IMD transformation during rice growth for the first time. Compared with the application of chemical fertilizer alone, straw returning significantly enhanced the photochemical process of paddy water. It was found that applying a low amount of rice straw was more beneficial to RIs accumulation in paddy water due to generated DOM being favorable to electron transfer with low aromaticity. The steady-state concentration of RIs steadily reduced with rice growth, owing to the photobleaching and plant consumption of DOM. Moreover, straw returning also contributed considerably to the transformation of IMD in paddy water, mainly attributed to 3DOM* and 1O2. Applying a low amount of rice straw had the best impact on accelerating IMD degradation. The intermediate products posed a more significant risk to aquatic animals but were less harm to aquatic plants than IMD. Furthermore, the finding of this study also suggests that long-term straw returning will alter the properties of DOM of paddy water, which ultimately influences RIs generation, pesticide transformation and element cycle (e.g., N and C) in paddy soil. Meanwhile, we would also eliminate the pollutants in paddy soil by regulating RIs formation in paddy water with strawing returning. This study was the first to investigate the effect of straw returning on the photochemical process of paddy water through a field experiment, providing a new idea for the fate of neonicotinoid insecticides in farmland systems.