In the post-green revolution epoch, the global agricultural community has experienced a remarkable escalation in food productivity, despite rising land values and increasing land scarcity. To achieve a higher return in a shorter span, agricultural intensification features within arable fields and as a “bonus”, it costs elevated environmental menaces, namely, groundwater depletion, increased greenhouse gas emissions, water pollution, and a decline in soil organic carbon (C) and microbial diversity. Agricultural intensification leads to multiple adverse consequences for natural resources, and farmers ace varying degrees of climatic risk, biotic invasion, and economic uncertainty. Thus, a paradigm shift in farming practices is urgent, by expelling unsustainable limbs, that is, conventional agriculture (too much plowing/tilling, exonerating all organic material, and monoculture), for future productivity gain, along with conserving natural resources. Adoption of conservation agricultural (CA) practices (comprising three key principles, that is, minimal mechanical disturbance of soil, permanent soil cover, and diversified crops) (FAO, 2008) has been recognized as a major strategy for sustainable intensification of agricultural production systems.

Aggregate stability is the most frequently used indicator of soil structure. Higher aggregate stability signifies improved soil productivity (Rieke et al., 2022) and sustainability, due to congenial structural arrangement and resistance against external disruptive forces. Furthermore, soil organic carbon (SOC) is considered an imperative aggregate binding agent, as it has a proximal relationship with aggregate formation and stabilization processes in association with soil biota (Six et al., 2002). Principally, macroaggregates protect the SOC, but disruption via tillage exposes the microaggregate C pool to decomposers and thereby amplifies subsequent mineralization. Thus, different aggregate size classes exhibit differential means of physical protection of associated C against microbial decomposition (Xie et al., 2017). Furthermore, the three principles of CA, applied in tandem, could alter the aggregation status of soil and its corresponding C dynamics. However, the amplitude of change varies with the nature of the soil (parent material, texture, depth, etc.) (Palm et al., 2014), prevailing climatic conditions (Li et al., 2020), and cultivation and crop management practices (Huang et al., 2018) as CA impacts are highly site-specific. Moreover, zero-tillage (ZT) with residue cover moderates soil temperature, conserves soil moisture (Six et al., 2000), increases mean weight diameter, water stable aggregate content, improves pore-size distribution (He et al., 2009), enhances macroaggregate-occluded microaggregates, and protects SOC better than those under CT (Barreto et al., 2009). Continuous inputs of organic materials form heterogeneous component crop species, and their divergent rhizospheric zones generate a diverse group of aggregating agents, such as fungal hyphae, microbial bioproducts (Haynes and Francis, 1993), and root exudates (Guggenberger et al., 1999).

Rice-based cropping systems are predominant in the lower Indo-Gangetic Plains of India, and conventional management (puddling) hastens issues like depletion of soil moisture and SOC, ultimately deteriorating the physical health of the soil (Hobbs, 2021). Recent research across South Asia has emphasized that the impact of CA on aggregate carbon distribution is strongly influenced by the type of rice-based cropping system employed. For instance, the inclusion of oilseed crops, such as mustard, contributes residues with faster decomposition rates and lower quality carbon inputs, which often lead to reduced macroaggregation and weaker carbon stabilization in soil aggregates (Amin et al., 2020). Although legumes contribute to nitrogen cycling, their influence on soil physical properties and carbon stabilization may not be as significant as cereal-based sequences (Islam et al., 2022; Mishra et al., 2024). Tillage intensity critically influences these dynamics, as Modak et al. (2020) observed that SOC within macroaggregates was approximately 30 and 25% higher under ZT than CT at 0–5 and 5–15 cm soil depths, respectively. Similarly, Nisar and Benbi (2024) reported that, at 0–7.5 cm depth, no-tillage without mulch increased macroaggregate-associated carbon by 45.5% compared to CT without mulch, while no-tillage with mulch led to a 67.9% increase over conventional tillage with mulch. There is an enduring debate about the impact of CA practices on soil structure and the ecosystem properties mediated by this structure. However, there is scarce reporting on the impact of CA on soil structural attributes of the lower Indo-Gangetic Plains, especially in soils with very high clay content during the initial years of adoption. The majority of previous research works on the lower Indo-Gangetic Plains have been predominantly focused on coarse-to-medium-textured soils, often overlooking the unique physical dynamics of very fine-textured soils, where movement of water, aeration, and microbial activity differ significantly. More often, those studies lack robust integration of region-specific cropping system diversity and residue management interactions within the CA framework. By identifying the range to which CA can influence the aggregation and SOC dynamics in the earlier stages of adaptation, it will facilitate our ability to determine its collective impact. Therefore, we hypothesized that three components of CA—minimum soil disturbance, residue retention, and crop diversification—along with their interactions with soil type and crop rhizosphere dynamics, may significantly modify the soil aggregation status and structural stability, and influence the physical environment of the soil rhizosphere. Furthermore, we anticipated that the maximum retention of rice residues under conservation tillage would result in higher levels of stable aggregates, along with an array of associated carbon, in the heavy clayey soils of the lower Indo-Gangetic Plains. We further hypothesized that the magnitudes of these effects would vary across cropping systems and that the inclusion of legumes and cereals in rotation would enhance soil structural indices and partitioning of SOC compared to less diverse systems.

To address these hypotheses, an experiment was conducted in a heavy clayey soil of the lower Gangetic plains, by cultivating different rice-based cropping systems along with different combinations of tillage and residue management, with the following objectives, (i) to access the impact of short-term CA practices on structural attributes of soil, (ii) to examine the impact of such management practices on C associated with different aggregate fractions. Accordingly, we can quantify the aggregate-size-specific carbon content and mass, offering a detailed picture of the carbon stabilization pathway of very fine-textured (clayey) soils in the under-researched lower Indo-Gangetic Plains. Further, this could identify the best-fit combination of tillage, residue, and region-specific cropping systems under CA.

The amount and quality of water-stable soil aggregates pithily regulate soil productivity plus quality, being the pavement of soil structure, energy conservation, metabolism, and resilience (Zheng et al., 2018; Kundu et al., 2021). Our findings suggested that conservation agriculture practices (CA), even within a relatively short span of 3 years, promoted greater aggregate stability by retaining C mainly in the macroaggregate fractions of the surface soil. This experiment will provide knowledge for the better adoption of conservation agriculture practices at the regional level for achieving sustainability of rice-based cropping systems in heavy clayey soils of the lower Indo-Gangetic plains.

Our results demonstrated that, irrespective of the cropping system, an increased proportion of CMac was observed under CA practices, which was positively associated with enhanced aggregation indices such as geometric mean diameter (GMD) and aggregate ratio (AR) (Mondal et al., 2021). Other relatively smaller aggregate fractions (Meso, CMic, and S + C) were dominated under CT due to mechanical jeopardization of the soil. Our findings were consistent with earlier work by Nandan et al. (2019), who observed a 1.1 and 1.28 times hike in MWD, and a 1.4 times and 1.7 times hike in AR, due to residue retention and ZT over residue removal and CT plots, respectively, after 6 years of practice. Under CA practices, aggregate stability increases with increasing SOC as it acts as an aggregate binding agent. Retention of 100% rice residue resulted in the most impressive effect on the distribution of water-stable aggregates and aggregation indices (Li et al., 2019; Sharma et al., 2022). Addition of crop residue improves the aggregate stability in two separate ways, one by supplying carbon, which acts as an aggregate binding agents (namely, polysaccharides, humic substances, etc.) or acts as a hotspot of microbial activity, which impart beneficial consequences on soil structure (Choudhury et al., 2014). Literature suggests that biogenic agents are critical drivers of aggregate stabilization, especially in the early years of CA, as microbial exudates, fungal hyphae, and root-derived polysaccharides foster aggregate formation and stability (Baumert et al., 2021). Second, by providing protection against various disruptive forces (namely, rain splash, wind force, machinery, surface runoff, etc.) (Schmidt et al., 2018), moderated temperature-moisture fluctuations create favorable conditions for microbial binding agents. In our experiment, the second explanation holds good as a significant enhancement in structural indices was noticed without significant enrichment of SOC after residue retention. The non-significant impact of residue management on SOC might be due to (i) short-term (3 years) CA practice which did not provide the necessary time for decomposition of crop residues, (ii) residues retention of residues rather incorporation, (iii) heavy textural class of the experimental soil might subdue the effects of residue addition, (iv) reduced fertilization under residue retained plots. However, Oicha et al. (2010) observed that plant residue incorporation resulted in a significant hike in SOC under CA management in a short-term study (1 year) which leads to a greater (but non-significant) degree of soil aggregation in a heavy clayey soil (73% clay, 24% silt, and 3% sand).

Among the different cropping systems, cereal–cereal–legume-based RWG was reported to have better GMD and AR which was mainly due to higher SOC (Pi et al., 2019; Nandi et al., 2020). Zuber et al. (2015) reported significantly lower levels of water-stable aggregates due to the inclusion of oilseed in cereal-based crop rotation. The significant influence of crop rotation on SOC was most probably from the differences in underground biomass and contribution from leaf litter in short-term rotations (Dube et al., 2012). Cumulative carbon inputs (CCI) based on actual residue retention rates (100% in R2 and 50% in R3), root C, and rhizodeposition are provided in Table 4 and Supplementary Table 3. These added C inputs (up to 16.14 Mg ha⁻¹) served as the primary organic input, and in conjunction with minimal soil disturbance under CA, enhanced macroaggregate formation and increased aggregate-associated carbon. This supports early-stage SOC stabilization via physical encapsulation and reduced decomposition, as reflected by improvements in GMD, AR, and PCA analysis, favoring ZT-R2 treatments. Notably, the RWC system recorded significantly the highest CCI (13.62 Mg ha⁻¹), which was 1.3 times greater than RLF (10.6 Mg ha⁻¹), primarily due to greater root biomass and rhizodeposition contributions (Table 4). Higher fibrous root biomass of cereal crops accounted for significantly higher rhizodeposition which possibly acted as temporary and transient binding agents (Tisdall and Oades, 1982), and ultimately triggered stable aggregate formation. Rasse et al. (2005) emphasized the relevance of C input from roots of component crops. They estimated that the root-derived C has 2.4 times the mean residence time of shoot-derived C. Thus, differences in component crop in rotation under CA also have the potential to impact SOC values over conventional practices and trigger stable aggregate formation. A significant positive correlation was also observed between SOC and CCI at each soil layer (Figures 9a, b), implying that enhanced C input contributes to SOC accumulation which improves aggregate stability. However, a relatively weaker positive correlation existed between SOC and geometric mean diameter (GMD) (significant only at the surface layer). This might be due to the short duration of the study, and stronger associations may emerge over a longer timescale (Parihar et al., 2020). Jahangir et al. (2021) reported significantly higher values of GMD due to the inclusion of wheat in the rice–rice cropping system over the inclusion of legume and mustard in rotation at the sub-tropical rice ecosystem of Bangladesh in a short-term study.

TS, being a fundamental property of soil aggregates, is very sensitive to soil management, and we observed a trend under tillage systems of RT > CT > ZT (Abid and Lal, 2009). Jeopardization of soil leads to loss of SOC which hampers the stability of aggregates and as a result, higher TS was experienced. In clayey soils, high TS often translates to poor friability and restricted root penetration from an agronomic standpoint. Our results show that ZT, combined with residue retention, significantly reduced TS (up to 13% compared to CT), which implies improved soil workability and better seedling emergence. This is especially beneficial for intensive cropping in heavy clayey soils, where physical constraints often hinder crop establishment. The inverse correlation between CMacAC and TS further supports the role of SOC in improving aggregate friability under CA systems (Blanco-Canqui et al., 2005).

The concentrations of aggregate-associated C followed the order of S + CAC > CMicAC > MesoAC = CMacAC. The smallest particle size of “silt+ clay” fraction offers copious reactive sites within which C could be riveted through strong ligand exchange or polyvalent cation bridges (Majumder et al., 2007). Thus, greater surface area for S + C and their intimacy with decomposed C were accountable for sustaining a good amount of C. Accumulation of C in silt-clay fractions highlighted the importance of clay and aggregation in C sequestration and stabilization (Singh et al., 2018). However, the highest aggregate-associated C mass within coarse macroaggregates was encountered mainly due to two reasons. First, CA practices enhanced proportion of CMac from breaking down, and C was kept well protected within it. Second, the clay fraction itself acts as an aggregating agent, and heavy clayey soil accounts for a very high level of macroaggregates (Pinheiro et al., 2004). Transfer of C into the CMac mass was highest under ZT and lowest for CT, as CT hastens the oxidation of C by destroying the macroaggregates (Guo et al., 2013). Higher concentration of aggregated associated C under straw-retained plots of ZT/RT treatments in the surface layer may be related to the interactive effect of tillage and straw returning (Gathala et al., 2015). The relationship between CCI and aggregate-associated C across different aggregate size fractions of the 0–10 cm soil layer revealed a significant positive linear trend at each aggregate size class (Figure 10a). In contrast, the trend was weaker in the 10–20 cm layer (Figure 10b), indicating reduced sensitivity at the subsurface soil. Principal component analysis revealed that practicing ZT with 100% rice-residue retention was conducive to stable aggregate formation, as the majority of the structural attributes of experimental soil were positively influenced, irrespective of the cropping systems. Further, PCA demonstrated the importance of the inclusion of another cereal (wheat) into rice-based cropping systems which triggers the formation of water-stable aggregates in heavy clayey soils of the lower Indo-Gangetic plains.

The results clearly indicated that conservation agriculture practices exerted a net positive influence on soil aggregation and aggregate stability, thereby contributing to an increase in soil organic carbon content in the clay-rich soils of the lower Indo-Gangetic Plains. Minimum soil disturbance combined with residue retention increased the amount of C retained within the macro- and microaggregates of surface soil and thereby offered a higher level of structural stability. Although the concentration of aggregate-associated carbon was highest for the silt + clay fraction, the highest mass of aggregate-associated C was noticed in the case of the coarse macroaggregate-associated fraction. Zero tillage with 100% rice residue retention and 75% RDF fertilization has been proven to be the best conservation practice for maintaining structural attributes of soil. Moreover, the inclusion of wheat in a rice–legume cropping sequence provided better aggregate stability and associated carbon due to higher underground biomass and rhizodeposition. However, these short-term (3-year) findings may not fully capture long-term trends in soil carbon dynamics and aggregate stability. Furthermore, only rice residue was retained, and that was also limited to the surface, which may have limited the potential carbon enrichment in deeper layers of soil. Integrating both reduced RDF and residue retention as sub–sub plot treatment made it difficult to explicitly identify the individual effects of these two factors. Despite these constraints, current study will contribute practical insights for global agricultural community and encourage the adoption of conservation agriculture in rice-based cropping systems of this region. Future research should focus on a long-term monitoring involving diverse agroecological zones, deeper soil layers, and alternative residue incorporation strategies to capture the whole trajectory of organic carbon stabilization in very fine-textured soil to further validate the scalability and robustness of these findings.