The study was structured as a multi-location farmers’ field trial focused on different sugarcane cropping sequences (Supplementary Table 1). The research location is an important agricultural zone within the North West Alluvial Plain Zone, characterized by a semi-arid to sub-tropical climate. Agro-climatic Zone-IV (as defined by the planning commission, now NITI Ayog), known as the “Middle Gangatic Plains Region,” encompasses the entire state of Bihar, including the district of Samastipur (Figure 1). The soils of Samastipur, Bihar, are primarily alluvial in nature, classified as Entisols and Inceptisols. They have a loamy to clay-loam texture, moderate fertility, and slightly acidic to neutral pH, making them well-suited for crop cultivation. The location has a sub-tropical, semi-arid climate according to the Köppen-Geiger classification, characterized by an average annual temperature of 25.2 °C and annual rainfall of 1,236 mm (Figure 2). Summers are hot, winters are cool, and there is moderate rainfall, with the driest month being November and the highest precipitation occurring in July. The sugarcane field, managed by local farmers under Magadh Sugar and Energy Ltd., Hasanpur (Bihar), India (25.710118, 86.19065), follows standard procedures: fertilization, weed control, pest management, and leaf removal.

A total of 380 individual soil samples were combined into 38 composite samples (each comprised 10 individual soil samples) collected from Hasanpur (18), Rosera (08), Warisnagar (05), Bibhutipur (04), and Khanpur (03) blocks in the agricultural landscape of Samastipur. Soil sampling was conducted at various farmers’ fields within these blocks, covering two (0–15 and 15–30 cm) soil depths. Sampling sites within each field were randomly selected to ensure comprehensive coverage of soil variability (Figure 3).

Oxidizable organic carbon was estimated by titration (Walkley and Black, 1934; Jackson, 1973), and its various fractions (labile and non-labile) were determined using a modified Walkley and Black method (Chan, 2001). For total organic carbon (TOC) determination, 0.25 g of the soil sample was mixed with 20 mL of 0.4 N chromic acid solution, heated at 155–160 °C for 20–25 min, and then cooled to room temperature (Jackson, 1973). Active carbon was estimated by quantifying potassium permanganate (KMnO₄) oxidation with a spectrophotometer (Weil et al., 2003). Carbon stock and its active/passive pools were estimated (Equations 1–3), while lability and recalcitrance indices were calculated as per Datta et al. (2015) (Equations 4–6).

where, SOC = soil organic carbon; TOC = total organic carbon; BD = bulk density; D = depth of soil

where, CAP = carbon active pool; VLC = very labile carbon; LC = labile carbon

where,

CPP = carbon passive pool; LLC = less labile carbon; NLC = non-labile carbon

where,

LI = lability index; VLC = very labile carbon; LC = labile carbon; LLC = less labile carbon; TOC = total organic carbon; NLC = non-labile carbon; RI = recalcitrance index.

The study was structured as a multi-location farmers’ field trial, where ongoing demonstrations served as replications across different sugarcane-based cropping sequences. Data were collected over cropping years in the North West Alluvial Plain Zone of Bihar. Descriptive statistics for various soil and crop parameters were generated using IBM SPSS (v26). Location maps of the study sites were prepared in ArcGIS 10.8.2. To evaluate multi-location effects, analysis of variance (ANOVA) was performed. Replication and multi-location effects (treatment) effects across the experimentation were modeled as random factors, and variance components were estimated using Restricted Maximum Likelihood (REML). Degrees of freedom were adjusted using the Satterthwaite approximation to improve accuracy. Treatment means were compared using the critical difference (CD) test at the 5% significance level (p ≤ 0.05), and standard errors (SE) were calculated for all comparisons. For transparency, F-values, degrees of freedom, p-values, and effect sizes (95% CI) are reported in the results.

The data revealed significant variation in oxidizable SOC content across different soil depths within the studied agricultural landscape (Table 1). At a depth of 0–15 cm, mean SOC content was 5.23 g kg⁻¹. It ranged from 3.17 to 8.01 g kg⁻¹, and exhibited a slight positive skewness (0.30) and leptokurtic distribution (kurtosis = 0.73). In contrast, at a deeper soil depth of 15–30 cm, the mean SOC content decreased (4.56 g kg⁻¹), and it ranged from 2.50 to 6.90 g kg⁻¹, with skewness (0.19) and lower kurtosis (0.33). Overall, the SOC content decreased (12.8%) at 15–30 cm depth. SOC content decreased with increasing soil depth, indicating a clear vertical gradient in its distribution. This pattern is critical for understanding nutrient dynamics and assessing soil health in agricultural ecosystems (Table 1). Continuous addition of above-ground biomass and decaying roots significantly influences C-levels. Research by Dotaniya et al. (2014) highlighted that organic amendments enhance C-content and stimulate microbial activity. The observed slight increase in integration of organic manure (OM) and mineral fertilizer application can be attributed to improved crop growth and higher yield, leading to increased residue and root exudate inputs that decompose and enrich the SOC. Similar findings have been reported by More et al. (2007), Bokhtiar et al. (2008), Jha et al. (2017), Sinha et al. (2017a), and Mthimkhulu et al. (2016), reinforcing the positive impact of organic inputs on soil fertility and C-dynamics.

Data showed that the mean TOC content was 7.64 g kg⁻¹. It ranged from 5.68 to 11.4 g kg⁻¹, showing a positively skewed distribution (skewness = 1.57) and slightly leptokurtic shape (kurtosis = 1.11) at 0–15 cm. Meanwhile, at 15–30 cm, the mean SOC content decreased to 7.10 g kg⁻¹ and ranged from 3.85–11.4 g kg⁻¹, exhibiting a similar positively skewed distribution (skewness = 1.38) and lower kurtosis (kurtosis = 0.74) compared to the 0–15 cm depth. A 7.1% decline in mean TOC content was observed at the 15–30 cm soil depth (Table 1). Understanding the factors influencing variations in TOC levels, such as land use practices, soil management strategies, and climatic conditions, is crucial for optimizing sustainable agricultural practices and C-sequestration. OM application to soil, as observed with SOC, plays a significant role in enhancing C-content. This phenomenon is supported by findings in the literature (Dotaniya et al., 2014; More et al., 2007; Bokhtiar et al., 2008; Jha et al., 2017; Sinha et al., 2017b; Mthimkhulu et al., 2016), which highlight the positive impact of organic inputs on C-dynamics. By elucidating these relationships, researchers can better formulate strategies to enhance SOC levels and improve soil health, thereby contributing to broader environmental sustainability goals (Table 1).

Results showed that KMnO₄-C content in the soil varied between depths in the agricultural landscape (Table 1). At 0–15 cm, the mean KMnO₄-C content was 263.26 mg kg⁻¹, and it ranged from 145 to 382 mg kg⁻¹, showing a negatively skewed distribution (skewness = −0.20) and platykurtic shape (kurtosis = −0.13). While at 15–30 cm, the mean KMnO₄-C content decreased to 235.69 mg kg⁻¹ and ranged from 122 to 356 mg kg⁻¹, displaying similar skewness (skewness = −0.19) and kurtosis (kurtosis = −0.05) compared to the 0–15 cm depth (Table 1). The variation in the soil depth was 10.5%. This observed decline in KMnO₄-C levels with increasing soil depth suggests a gradient in C-availability that is pivotal for understanding SOC dynamics and nutrient cycling in agricultural systems. Further exploration of factors influencing KMnO₄-C, such as management practices and soil physical properties, is essential to optimize C-sequestration and enhance soil health strategies. Higher KMnO₄-C in the 0–15 cm layer may be attributed to enhanced root biomass, rhizodeposition, and microbial activity. Monitoring labile organic matter fractions offers a sensitive and effective approach for evaluating soil quality dynamics (Needelman et al., 1999; Ding et al., 2006; Benbi et al., 2015; Meena et al., 2021).

Data showed that a significant variation was found in VL-C and L-C content across different depths within the agricultural landscape (Table 2). At 0–15 cm, the mean C-VL content was 1.56 g kg⁻¹ and ranged from 1.02 to 2.60 g kg⁻¹, showing a positively skewed distribution (skewness = 0.63) and platykurtic shape (kurtosis = −0.58). However, at 15–30 cm, the mean C-VL content decreased to 1.31 g kg⁻¹, and ranged from 0.90 to 1.98 g kg⁻¹, exhibiting a similar positively skewed distribution (skewness = 0.68) and slightly lower kurtosis (kurtosis = −0.31) compared to the 0–15 cm depth. The decline in mean C-VL content from 0–15 cm to 15–30 cm depth was 16% (Table 2).

Meanwhile, C-L in the soil showed variations across different depths (Table 2). At 0–15 cm, the mean C-L content was 1.50 g kg⁻¹ and ranged from 1.13 to 2.08 g kg⁻¹, showing a positively skewed distribution (skewness = 0.33) and slightly platykurtic shape (kurtosis = −0.32). In contrast, at 15–30 cm, the mean C-L content decreased to 1.22 g kg⁻¹ and ranged from 0.43 to 1.97 g kg⁻¹, exhibiting a similar positively skewed distribution (skewness = 0.13) and leptokurtic shape (kurtosis = 0.88) compared to the 0–15 cm depth. The decrease in mean C-L content from 0–15 cm to 15–30 cm depth was approximately 18.7% (Table 2). These higher VL-C and L-C contents signify enhanced microbial activity and OM turnover (Meena et al., 2018; Meena et al., 2021). Monitoring these fractions provides insights into the responsiveness of soil carbon pools to management practices, facilitating sustainable agricultural strategies aimed at enhancing soil health and C-sequestration efforts (Table 2). Integrated studies on sugarcane highlight that sustainable trash management (Gadge et al., 2017; Shanthy et al., 2020), conservation tillage (de Oliveira et al., 2022; Turmel et al., 2015), and optimized fertigation (Ravikumar et al., 2011) are key to enhancing productivity, improving soil health, and ensuring long-term ecological balance.

Results showed that less labile carbon (LL-C) content at 0–15 cm soil layer ranged from 0.93–3.33 g kg⁻¹, with a mean value (2.16 g kg⁻¹), showing a slightly positively skewed distribution (skewness = 0.14) and mesokurtic shape (kurtosis = 0.39). At 15–30 cm, the mean LL-C content slightly decreased to 2.04 g kg⁻¹ and ranged from 1.00–3.45 g kg⁻¹, exhibiting a similar positively skewed distribution (skewness = 0.48) and higher kurtosis (kurtosis = 0.85) compared to the 0–15 cm depth. The slight decrease in mean LL-C content from 0–15 cm to 15–30 cm depth was approximately 5.6% (Table 2).

Meanwhile, in the case of NL-C at 0–15 cm, the mean was 2.41 g kg⁻¹ and ranged from 1.48 to 3.41 g kg⁻¹, showing a positively skewed distribution (skewness = 0.41) and a highly leptokurtic shape (kurtosis = 4.70). At 15–30 cm, the mean NL-C content slightly increased to 2.54 g kg⁻¹, ranging from 1.35–4.54 g kg⁻¹, exhibiting a similar positively skewed distribution (skewness = 0.98) and lower kurtosis (kurtosis = 0.82) compared to the 0–15 cm depth. The slight increase in mean NL-C content from 0–15 cm to 15–30 cm depth was 5.4% (Table 2). Sugarcane trash, being less prone to microbial decomposition, tends to accumulate as passive SOC over time. This phenomenon underscores the role of agricultural practices in influencing C-dynamics, where annual additions of sugarcane trash to the surface layer may enhance the active C-fraction compared to the deeper layer. Crop residues support microbial proliferation, which in turn stimulates nutrient cycling and enhances soil fertility (Ensinas et al., 2015; Mir et al., 2023). Additionally, Majumder et al. (2008) support the idea that management options influence the balance between C-inputs (e.g., root biomass, litter fall) and outputs (e.g., heterotrophic respiration), thereby affecting soil carbon pools. These findings align with studies by Mir et al. (2023) and Benbi et al. (2015), which emphasize the importance of understanding carbon dynamics in agricultural systems to optimize carbon management strategies and mitigate climate change impacts (Table 3).

Carbon stock in the soil varied across different depths within the agricultural landscape (Table 3). Results showed that the mean C-stock (17.00 Mg ha⁻¹) ranged from 12.6 to 25.9 Mg ha⁻¹ at 0–15 cm and showed a positively skewed distribution (skewness = 1.18) and leptokurtic shape (kurtosis = 2.19). At 15–30 cm, the mean C-stock slightly decreased to 16.72 Mg ha⁻¹ and ranged from 10.2 to 24.5 Mg ha⁻¹, exhibiting a similar positively skewed distribution (skewness = 0.62) and slightly lower kurtosis (kurtosis = 0.59) compared to the 0–15 cm depth. The decrease in mean C-stock from 0–15 cm to 15–30 cm depth was 1.6% (Table 3). The decline in mean C-stock from 0–15 cm to 15–30 cm depth highlights significant variations in organic carbon storage, crucial for soil fertility, C-sequestration, and overall soil health in the region (Table 3). Recent studies by Ensinas et al. (2015) emphasize the role of cropping systems, such as sugarcane cultivation, in influencing C-stocks and GHG mitigation. Further research on TOC dynamics, soil properties, and management practices is essential for optimizing carbon management strategies in agricultural landscapes (Benbi et al., 2015; Majumder et al., 2008; Sherrod et al., 2005; Mir et al., 2023).

Data showed that active and passive carbon pools varied across the soil depths (Table 3) within the agricultural landscape. At 0–15 cm, the mean active pool was 3.07 g kg⁻¹, and it ranged from 2.15 to 4.68 g kg⁻¹, showing a moderately positively skewed distribution (skewness = 0.56) and a slightly negative kurtosis (kurtosis = −0.30). At 15–30 cm, the mean active pool decreased to 2.53 g kg⁻¹, and it ranged from 1.43 to 3.65 g kg⁻¹, indicating a similar positively skewed distribution (skewness = 0.36) and a more negative kurtosis (kurtosis = −0.44) compared to the 0–15 cm depth. Overall, these findings indicated a decrease (17.59%) in the mean active pool of soil carbon from the 0–15 cm to the 15–30 cm soil depths. However, the passive pool of soil carbon showed a negligible change of approximately 0.22% from the 0–15 cm to the 15–30 cm soil depths (Table 3). These findings underscore the complex dynamics of soil carbon storage and turnover, crucial for optimizing C-sequestration and soil fertility management strategies (Lal, 2004; Paustian et al., 2000; Regnier et al., 2022; Tubiello et al., 2021; Wieder et al., 2015; Fan et al., 2023; Chen et al., 2022; Batjes et al., 2020).

The data indicated that LI, RI-I, and RI-II values varied significantly across different soil depths (Table 4). At the 0–15 cm depth, the mean LI was 1.28 and ranged from 1.10 to 1.47. The distribution showed a slight positive skewness (skewness = 0.08) and exhibited a platykurtic shape (kurtosis = −0.88), indicating a relatively flat distribution compared to a normal distribution. At the 15–30 cm depth, the mean LI slightly decreased to 1.19 and ranged from 1.04 to 1.38, with higher positive skewness (skewness = 0.22) and more negative kurtosis (kurtosis = −1.55), suggesting a more peaked and narrower distribution compared to the 0–15 cm depth (Table 4).

Meanwhile, in the case of RI-1 at 0–15 cm depth, the mean was 1.52 and varied from 1.23 to 2.32 with positive skewness (skewness = 1.27) and exhibited leptokurtic shape (kurtosis = 2.25). At 15–30 cm depth, the mean was increased to 1.83 and ranged from 1.33 to 2.53. Distribution showed a lower positive skewness (skewness = 0.18) and negative kurtosis (kurtosis = −0.79), indicating a more flattened and broader distribution compared to the 0–15 cm depth (Table 4). Similarly, in the case of RI-2 at 0–15 cm depth, the mean was 0.32 and ranged from 0.22 to 0.44, indicating a relatively narrow distribution. The distribution showed a slight positive skewness (skewness = 0.14) and exhibited a slightly leptokurtic shape (kurtosis = 0.61), suggesting a moderately peaked distribution compared to a normal distribution. At 15–30 cm depth, the mean RI-2 increased slightly to 0.36, and values ranged from 0.28 to 0.46, indicating a slightly wider distribution compared to the 0–15 cm depth. The distribution showed a positive skewness (skewness = 0.21) and negative kurtosis (kurtosis = −1.18), suggesting a flatter and broader distribution compared to the 0–15 cm depth (Table 4). These findings underscore the importance of understanding C-stability dynamics in soil profiles, crucial for implementing effective soil management strategies to enhance C-sequestration and mitigate climate change impacts (Hassink, 1997; Poeplau and Don, 2015; Yang et al., 2024; Nie et al., 2024; Lu et al., 2024; Adhikari et al., 2024; Meena et al., 2024).

The correlation matrix revealed significant relationships among soil parameters across different soil depths within the studied agricultural landscape (Tables 5, 6). OC refers to organic carbon in soil organic matter, while TOC is a broader term that includes all forms of organic carbon, both particulate and dissolved, providing a more comprehensive measure. At the 0–15 cm soil depth, pH was negatively correlated with TOC (r = −0.732**) and KMnO₄-C (r = −0.710**), indicating reduced carbon levels with increasing pH. In contrast, EC showed strong positive correlations with TOC (r = 0.750**) and KMnO₄-C (r = 0.757**). Nitrogen exhibited the highest correlation with TOC (r = 0.903**) and KMnO₄-C (r = 0.906**). Sulfur and water-stable aggregates (WSA) were also strongly associated with KMnO₄-C (r = 0.888** and r = 0.806**, respectively), suggesting their key roles in carbon stabilization (Table 5). Meanwhile, at 15–30 cm depth, nitrogen showed strong positive correlations with TOC (r = 0.826**) and OC (r = 0.863**). Sulfur and phosphorus were also highly correlated with TOC (r = 0.744** and r = 0.715**, respectively). In contrast, bulk density was strongly negatively correlated with TOC (r = −0.734**) and OC (r = −0.763**). Unlike in the surface layer, KMnO₄-C showed no significant associations, indicating its lower sensitivity at this depth (Tables 5, 6). Recent studies underscore the importance of sustainable sugarcane cultivation using efficient irrigation (Gunarathna et al., 2018), trash and tillage practices (Surendran et al., 2016; Solomon, 2016; Suma and Savitha, 2015), and precision nutrient management (Basanta et al., 2003; Choudhary et al., 2017). These strategies improve yield, enhance soil health, and boost resource use efficiency (Powar et al., 2021).