Impact of Conservation Agriculture and Cropping System On Soil Organic Carbon and Its Fractions in Alluvial Soils of Eastern Gangetic Plains

A conservation agriculture-based sustainable intensication (CASI) practices have been proposed as a potential alternative management strategy for achieving the food, water and energy security while sustaining the soil health and climate resilience. In this study, we evaluate the performance of CASI technologies under two cropping systems on carbon (C) dynamics in the soils of recent and old alluvial nature of West Bengal in Eastern Alluvial Ganga Plains.


Introduction
The growing concern of global warming and climate change impacts on the community have spurred interest in enhancing the sequestration of atmospheric carbon dioxide (CO 2 ) in terrestrial ecosystems (Dolman et al., 2004;Lal, 2015;Sarkar et al., 2020). According to the Intergovernmental Panel on Climate Change (IPCC), about 22% of global anthropogenic greenhouse gas (GHG) emissions are contributed by agriculture, forestry, and/or other land uses (IPCC, 2019). Cultivation of arable lands leads to the substantial loss of soil organic matter (SOM) and increase emissions of CO 2 from soil to the atmosphere, thereby increasing the CO 2 concentration in the atmosphere (Ladha et al., 2015). The SOM is made up of dead plant residues, particulate organic C (POC), humus C, and recalcitrant C. It plays a major role in maintaining the fertility, productivity, and overall quality of soil (Larson and Pierce, 1994;Kang et al., 2005), besides having other important environmental functions (Fageria, 2012). The relative proportion of these fractions re ects in the soil ecosystems, including agricultural and non-agricultural soils, which can directly impact the microbial activity and C dynamics in soil. The labile C fractions in soil are the important component that determines the soil quality Although these fractions constitute a relatively smaller fraction of TOC and have a very short turnover times in soil which are highly sensitive to land management changes (Weil and Magdoff, 2004;Duval et al., 2018). The composition of these C fractions varies depending on the stage of decomposition, but they have critical role in soil functioning and health (Belay-Tedla et al., 2009). Improved land management practices should not only increase TOC stock, but ideally would optimize the proportion of C in these various TOC fractions. Any system that produces rich source of organic material, will have greater amounts of residue SOC. Thus, the study of TOC has increasingly focussed on identifying fractions of TOC that are related to how labile the C is. Fractions such as hot-water soluble C (HWEC), POC, and mineral associated organic C (MAOC) are used because they are indicative of residence or turnover times (Rakesh et al., 2020). These parameters also have been used as indicators for soil quality (Cambardella and Elliott 1992;Blair et al., 1995;Bolinder et al., 1999;Duval et al., 2018). Distribution of TOC-fractions and their stocks, in the soil pro le, or C strati cation, helps in identifying the variations in the quality of SOM of topsoil (Álvarez et al., 2011;Zhao et al., 2015).
Conservation agriculture-based sustainable intensi cation (CASI) management practices involving minimum soil disturbance, e cient crop rotations, and increased crop residue retention provides a means of increasing TOC (Johansen et al., 2012;Sharma et al., 2019). Tillage and residue management may in uence C sequestration, microbial activity, and also play an important role in affecting the soil physicochemical and biological properties (Jat et al., 2019;Choudhary et al., 2018). Crop residues have numerous bene cial effects as these are not only a source of organic C and nutrients (Yadvinder et al., 2009) but also form a mulch that conserves the soil moisture (Aulakh et al., 2012; Gathala at al., 2017; Sarkar et al., 2020). Moreover, it is estimated that the application of best management practices in agriculture has the potential of offsetting GHG emissions in the range of 1.1-4.3 Gt CO 2 -e yr −1 (UNEP, 2013). Thus, CA practices have the potential to reduce the 10 to 60 percent CO 2 -e emissions over energy intensive conventional systems, depending on the layering of CA practices implemented (Ladha et al., 2015;Gathala et al., 2020;Jat et al., 2020). Addition of organic materials to agricultural soil is important for replenishing the annual C losses and for improving both the biological and chemical properties of the soils (Goyal et al., 1999;Choudhary et al., 2018). Zero tillage (ZT) for crop production has been identi ed as an important practice to increase soil aggregation and C sequestration (Six et al., 1998 . The present study was undertaken to assess the effect of different tillage and crop residue management practices on soil C fractions. The investigation was conducted in selected farmers' elds of an on-going ACIAR-SRFSI research project which was initiated in 2013 to demonstrate the bene ts of CA over the conventional system. We hypothesized that changes in tillage and crop establishment techniques, along with crop residue retention and management practices under different cropping systems, may have a differential impact on the accumulation, and distribution of TOC and different TOC fractions in soil at different depths. Furthermore, that soils (old alluvial soils i.e., Inceptisols of Malda and recent alluvial soils i.e., Entisols of Coochbehar district) with different physicochemical characteristics may not have uniform response to tillage, crop residue retention, and cropping systems. The overall objective of the present investigation was to assess the impact of tillage (ZT and CT) in terms of C distribution in the soil of RW and RM cropping systems. The speci c objectives were: (i) to assess the response of different tillage practices and cropping systems on TOC and its fractions after four years (eight seasons) of cultivation, (ii) to explore the strati cation of TOC and its fractions at different soil depths, and (iii) to determine the relationships between TOC and its fractions with soil properties in two different agro-ecological regions.

Description of eld sites
The study was conducted in selected farmer's elds spread across the two districts, Coochbehar (26.3452° N, 89.4482° E) and Malda (25.0108° N, 88.1411° E) of the northern alluvial plains in West Bengal, sub-tropical eastern India. These districts present different soil and edapho climatic conditions. Field experiments were initiated in 2014-15 with RW and RM cropping systems, for 4 years until 2018-19, altogether 3 rice and 4 wheat or 4 maize crops were grown. These cropping systems were selected on the basis of the existing cropped area, as well as the potential of these systems for improvement of farming livelihoods (Dutta et al., 2020;Mitra et al., 2019;Gathala et al., 2020).This study formed part of an ongoing larger research project entitled 'Sustainable and Resilient Farming System Intensi cation (SRFSI)' being maintained by the Uttar Banga Krishi Viswavidyalaya (UBKV) in collaboration with the Australian Centre for International Agricultural Research (ACIAR), and the International Maize and Wheat Improvement Centre (CIMMYT) since 2013. The distance between the two eld sites / districts was approximately, 400 km. The eld sites selected for this study were historically used for growing rice in rotation with other dry-season crops using intensive tillage practices.
The study area has an overall warm humid subtropical climate with unimodal monsoonal rainfall, moderate to hot summer and cold winter, although wide variations existed between the Coochbehar and Malda districts with respect to annual precipitation and air temperature. The Coochbehar site receives a mean (30-year average) annual rainfall of 2357 mm and maximum temperature of 28.2°C and minimum of 20.0°C, while the Malda site has a mean annual rainfall of 1358 mm and maximum temperature of 30.6°C and minimum of 20.2°C.
The experimental trials were conducted as on-farm participatory trials (backed by researchers and managed by farmers; Islam et al., 2019). A total of seven eld experimental sites were identi ed, three (3) in Coochbehar and four (4) in Malda. At each of these eld experimental sites collaborating farmers were identi ed, and a factorial experiment of two cropping systems (RW and RM) and two tillage practices (ZT and CT) with three replications was established to study the effect of CASI practice on C dynamics. Due to the small land area available to each individual farmer, each trial was distributed across six farmer's elds, with each farmer implementing the two tillage practices (ZT and CT) on one cropping system (either RW or RM). Thus the full experiment was a 2x2 factorial conducted at seven sites and replicated three times. The soils at the Coochbehar sites are on recently deposited alluvium Entisol, acidic in reaction, with a light sandy texture (Sarkar et al., 2017). In contrast, the soils of Malda sites are on old alluvial material Inceptisol, neutral to alkaline in reaction, and silty loam to clay loam in texture. The pH (1:2.5 H 2 O), TOC, total nitrogen(N), soil texture, and bulk density of the experimental soils are given in  Crop and management practices

Collection of soil samples
Before the start of the experiment, soil samples (0-20 cm) were collected from each experimental site for the determination of the initial physicochemical properties of the soils. The samples were collected from 5 to 6 random spots in an individual farmer's eld using an auger of 5 cm diameter and mixed thoroughly to form a composite sample; three composite samples from each eld site were collected. The soil samples were air-dried and ground to pass through 2 mm sieve (for analysis of soil properties) and a subsample was further ground to pass through 0.5 mm sieve (for determination of soil organic C).
Processed soil samples were stored in sealed polythene containers till analyses were completed.
After the nal harvest of wheat (April, 2018-19) and maize (May, 2018-19) crops at the end of 4 years of the experiment, soil samples were collected at 0-5, 5-10 and 10-20 cm depths from each plot in each eld, with 5-6 random spots, mixed to one composite sample for each soil depth. Air dried composite samples were ground to pass through 2mm sieve for general analysis and a portion ground to pass through 0.5 mm sieve for SOC and its fractions and kept in sealed polythene containers until analyses were completed.

Soil properties
The pH of soil suspension in a soil: water ratio of 1:2.5 was determined with a pH meter, as described by Jackson (1967). Bulk density (BD) of soil samples was estimated using a core sampler of dimensions 5 × 5 cm (height × diameter) following the method of Cresswell and Hamilton (2002). The proportion of sand, silt, and clay in soil samples was determined by the Bouyoucos hydrometer method (Bouyoucos, 1962). The texture of the soils was ascertained from the particle-size distribution of sand, silt, and clay using soil texture triangle.

Carbon fractions Total organic C (TOC)
A modi ed Walkley and Black method (Baker, 1976) was followed for the analysis of TOC in soil determined by colorimetric method using sucrose as a standard. Brie y, one gram of soil sample was digested in the presence of 20 ml of 5% K 2 Cr 2 O 7 and 10 mL of concentrated H 2 SO 4 . After cooling for 30 minutes, 50 mL of 0.4% BaCl 2 was added and allowed to stand overnight. The intensity of the yellow/orange colour was read at 600 nm wavelength using a UV-visible spectrophotometer.
Hot water extractable C (HWEC) HWEC was determined by hot water extraction method (Ghani et al., 2003). The air-dried soil sample of 3 g was weighed into 50 mL centrifuge tube, 30 mL of de-ionized water was added, and the suspension was shaken for 30 minutes at 30 rpm and at room temperature. Then, it was centrifuged for 20 minutes at 3000 rpm, thereafter the supernatant was discarded to remove the cold-water soluble C. A further 30 mL of de-ionized water was added to the same residue and placed on a hot water bath at 80°C for 16 hours. After cooling down, the tubes were then shaken and centrifuged at 3000 rpm for 20 minutes. The supernatant was ltered through cellulose nitrate membrane (0.45 µm). The C concentration in the extract was determined by Nelson and Sommers (1982) method. A 4 mL of sample was oxidised with 1 mL of 0.066 M K 2 Cr 2 O 7 and 5 mL of concentrated H 2 SO 4 at 150°C for 30 minutes. Samples after cooling, were titrated against 0.033 M ferrous ammonium sulphate with 2-3 drops of o-phenanthrolene indicator until the colour turned from greenish violet to brick red.

Particulate Organic C (POC)
For the POC fraction (Cambardella and Elliott. 1992), 25 g of air-dried soil was dispersed in 100 mL of 0.5% sodium hexa-metaphosphate in reciprocating shaker for 16 hours. Suspension was then passed through 0.53 mm sieve followed by washing with de-ionised water to collect the >0.53 mm, the POM remained on the sieve. The POM was then dried and powdered. The C concentration of the POM was determined by following the modi ed Walkley and Black method (Baker, 1976).

Strati cation ratio (SR)
The strati cation ratio of a soil property is de ned as the ratio of its value at the soil surface to that at a lower depth (Franzluebbers, 2002). This ratio for a C fraction for 0-10 cm depth was calculated by dividing its value at 0-5 cm to that of its 5-10 cm depth. Similarly, for 0-20 cm depth, the value of 0-5 cm depth was divided by its C concentration at 10-20 cm soil depth.

Soil Organic C Stock
The C stock in soil was calculated considering soil depth (m), bulk density (BD, Mg m −3 ) and concentration (%) of TOC fraction using the following equation. There was no gravel in the soil samples at any of the sites.

Data analysis
Prior to performing statistical analysis, the normality assumption of analysis of variance (ANOVA) was tested using Shapiro-Wilk test (1965) using JMP statistical software (V9 software, Buckinghamshire, UK). Since the normality assumption of ANOVA was met, the data were not transformed. The data were analysed using proc GLM (general linear model) in SAS. We considered district (D), cropping system (CS), tillage treatments (T) and their interactions (CS x T; D x CS; D x T; D x CS x T) as xed effects and farmer (replication) as a random effect in the t-ANOVA model. The three-way interaction (D x CS x T) were not signi cant for any of the parameters at any depths except for TOC, MAOC, BD and TOC stock observed at 10-20 cm soil depth. The treatment means for all parameters were compared using Tukey's honest signi cant difference (HSD) test.
As the soil depth interval is a non-randomized factor, a mixed procedure with repeated measures was used for each experiment and analysed separately for each site. A correlation test was performed to determine correlations among soil organic C fractions with key important soil attributes at 0-5, 5-10, and 10-20 cm depths at the Malda and Coochbehar sites.

Results
Effect of cropping system and tillage on the concentrations of total organic C(TOC), hot water extractable C (HWEC), particulate organic C (POC) and mineral associated organic C (MAOC) TOC concentration varied widely between experimental sites and districts ( Table 2). Table 2 Effect of environment (District), cropping systems and tillage on total organic C (TOC) and hot water extractable C (HWEC) concentrations at different soil depths.
TOC concentration (g kg −1 ) HWEC concentration (mg kg −1 ) respectively which were 16.8 and 9.8 % greater than CT system (Table 2). However, at lower depth (10-20 cm), comparatively higher amount of TOC concentration was found in the soil under CT system (13.24 % higher) in comparison to ZT system.
Depth wise TOC concentration was found to gradually decrease with the increase in depths in soils in general; but the critical perusal of the Table 2 reveal that the depth distribution of TOC concentration differed in proportion when compared among the two districts, cropping systems and the tillage treatments. The concentration of the TOC was found to be more (26%) in 0-5 cm in comparison to 5-10 cm depth in Malda whereas in Coochbehar the same was found to be only 10% higher. Similar comparison of the subsequent depth distribution between 5-10 and 10-20 cm reveal that more or less uniform difference in TOC concentration between the two layers existed in both Malda (22%) and Coocbehar (24%) districts ( Figure 1). The trend of the distribution of TOC in the three layers were same in the both the cropping systems with only difference in concentration in each layers was found, where higher quantity of TOC was recorded in RM than the RW system ( Among the two labile pools of organic C (POC and HWEC), a signi cant (p<0.05) increment in HWEC concentration was noticed in Malda (284 mg kg −1 ) over Coochbehar (219 mg kg −1 ) at 0-5 cm depth but in the subsequent depths, there was no signi cant difference between the districts was observed in HWEC ( Table 2). According the HWEC classes for sandy and loamy soils given by Körschens and Schulz (1999), the HWEC content of FS-1(Malda) and FS-6 (Coochbehar) is low (<200 mg kg −1 ) and greater than the 400 mgkg −1 HWEC indicate high concentration was recorded in FS-4 (data presented in supplementary table   S2). Practice of RM system signi cantly (P≤0.05) increased the HWEC to the tune of 15.9, 5.9 and 37.9 % at 0-5, 5-10 and 10-20 cm depths respectively compared to RW. The signi cant interaction between the district and cropping system at 0-5 cm depth ( Figure 2) indicated that the concentration of HWEC was 25% higher for RM than RW system in Malda. While, in Coochbehar, variation was relatively similar between the cropping systems.
While ZT system improved the HWEC (23.7 %) over CT at 0-5 cm but at lower soil depth (10-20 cm) CT increased the same by 17%. The interplay of D x T showed maximum change at lowermost depth (10-20 cm) (Figure-2). CT system in both the districts improved the HWEC at 10-20 cm but in Coochbehar, the HWEC concentration between CT and ZT failed to attain signi cant difference. The contribution of HWEC fraction to TOC was recorded to be 1.5 to 2.4 % (Figure 4).
Field site and cropping system, signi cantly affected the POC concentration in both the districts at all depths, while tillage had in uenced the POC concentration at selected soil depths (  g kg −1 at 0-5, 5-10 and 10-20 cm depths respectively) as compared to RW ( Table 3). Implementation of CASI under ZT management signi cantly (P<0.05) enhanced the POC at 0-5 cm depth and CT improved the same at 10-20 cm; while in 5-10 cm soil depth, there was no signi cant variation observed between the two. However, the interaction effect of CS x T indicated that adoption of ZT under RM system improved the POC at surface depths (0-5 and 5-10 cm) but at the lower depth (10-20 cm), CT showed higher increments. Tillage systems did not affect POC under RW system when referred to 0-5 cm depth, however, in the lower depths CT improved the POC over ZT ( Figure 5). D x CS showed that there was a signi cant improvement in POC under RM system in Coochbehar compared to RW, but there was a very less difference in concentration noticed between the CS in Malda when referred to 10-20 cm soil depth ( Figure 5). The contribution of POC to TOC varied from 18 to 32% in both the Coochbehar and Malda soils ( Figure 4).
Concentration of MAOC was found to follow the same trend as that of TOC. The MAOC concentration was more in the soils of Malda (12.2, 9.25 and 7.48 g kg −1 at 0-5, 5-10 and 10-20 cm depths respectively) which were 33.4, 8.58 and 12.1 % higher than Coochbehar soils (  Similarly, CT in Coochbehar and ZT in Malda increased the MAOC under RW when referred to the depth 10-20 cm (Figure 3). The contribution of MAOC to TOC varied from 65 to 80% (Figure 4).

Effect of cropping system and tillage on soil BD at different soil depths
In both the districts, soil BD values were increased with the depth (Table 4) showed that only the effect of tillage was observed to be signi cant on soil BD but cropping system failed to show any such variation on BD. CT system signi cantly decreased the BD over ZT (Table 4)   The TOC stocks were signi cantly higher under the ZT than CT at 0-5 and 5-10 cm; while at 10-20 cm depths it was maximum in CT soils (   (Table 6B) between POC and sand (r= -0.59*, -0.54*, and -0.43*, p<0.05), but these were positively correlated in the Coochbehar soils (r= 0.73**, 0.78** and 0.85**,) at 0-5, 5-10 and 10-20 cm depths, respectively. It was also observed that silt was positively correlated with POC concentration in the Malda soils, but negatively correlated in the Coochbehar soils. Relationship of TOC with texture showed that in the Malda soils it was highly correlated with clay content (r= 0.78**) but in the Coochbehar soils, it was strongly correlated with sand content (r= 0.53*) (Table 6C).

Discussion
In this short-term (4 year) study adoption of CASI practice under ZT management signi cantly in uenced the concentration of soil C fractions, TOC stock and its depth-wise distribution compared to CT. The variability of TOC concentration in different sites of Coochbehar and Malda was observed is due to the background TOC content and difference in crop management practices adopted by the different farmers.
This is further corroborated from the statistical study which indicate that the effect of district and villages (sites) were signi cant.
TOC concentration was found to be signi cantly (p<0.05) higher in the RM cropping system compared to RW. Since the addition of C substrate is more in the RM (10 Mg ha −1 ) than RW (5. cm) were slightly greater in full inversion tillage than in no-tillage treatment. A similar trend of the soil under CT having higher TOC concentration at the lower depth (10-20 cm) by 18 % than ZT, was also reported by Zhu et al. (2014). Although there was no residue incorporation in the ZT, the higher TOC stocks were recorded at lower soil depths, which may be ascribed to the similar soil textural characteristics (low clay and high silt content) in the soils at these three sites (FS-1: 16 and 66 %; FS-6: 10 and 62 %; FS-7: 12 and 60 %, respectively) ( Table 1). This may have enhanced the movement of organic carbon fractions which constitute and are the component of the TOC, into the lower layers.
Implementation of ZT in the Malda soils, improved the concentration of TOC fractions at the upper depths as the soils were rich in clay content (28%) ( Table 1).
Therefore, in this study there is relatively higher TOC concentration in the surface (0-5 cm) in comparison to the subsequent depths (5-10 and 10-20 cm). Depth-wise distribution of TOC fractions decreased with increasing depth. Somasundaram et al., 2018 also observed a marked decrease in the very labile C fraction with increasing soil depth. However, the TOC stocks increased with depth re ecting higher soil bulk density. The variation in the pattern of the distribution of the TOC and its stock was in uenced by both the textural differences and the tillage treatments which in uenced the distribution of the residue within and on the soil. Irrespective of the site characteristics, the RM system added more C input than the RW system in the soil resulting in higher TOC concentration in all the three soil depths.
We found a signi cant positive relationship between clay and TOC in the Malda soils but a positive correlation with the sand and negative correlation with silt in the Coochbehar soils (Table 7C). This is a peculiar phenomenon and is the characteristics of the Coochbehar soils, which are low in clay content. Therefore, the role of clay may vary by region (Oades, 1988;Goncalves et al., 2017). Hassink (1997) attributed this relationship due to the formation of organo-mineral complexes between the organic matter particles and the clay forming bond which stabilizes the C in the soil. The process and the extent of binding always varies among the different clay types (Blanco-Canqui and Lal, 2004). The clay content and mineralogy also regulate the TOC pools by in uencing the sensitivity of soil C to microbial attack (Percival et al.,2000;Kumari et al., 2011;Choudhary et al., 2018), therefore, we expect higher rate of TOC turnover in the Coochbehar soils (recent alluvial Entisols) than the Malda (old alluvial Inceptisol) soils. Moreover, the moisture content in the soils of Coochbehar is relatively higher than that in the Malda soils, which allows the labile pools to move down the soil pro le with the water movement.
A strong and positive correlation of HWEC with TOC (r=0.76, n=84) ( Figure 7) showed that the concentration and the distribution of HWEC is directly in uenced by the organic C concentration of the soil. Similar relationship between TOC and HWEC was also reported elsewhere (Spohn and Giani 2011;Haney et al., 2012;Vladimiret al.,2016).This labile form of carbon is also related to the microbial biomass (Sparling et al., 1998) and micro-aggregation (Puget et al., 1999) in soil and is therefore an important indicator of soil quality (Ghani et al., 2003).The concentration HWEC of the two sites (FS-1 and FS-4), according to the classes given by Körschens and Schulz (1999) is highly depleted in organic carbon (<200 mg kg −1 ). This variation in the concentration at different sites under is due to the background TOC content of the soils of these regions. Intervention with respect to the tillage and cropping system in the long run would improve the concentration in soil and also its quality. Depth-wise distribution of HWEC followed a similar trend as that of TOC; while an increased level was observed in the 10-20 cm soil depth under CT practice over the ZT practice; this was certainly due to the incorporation of residues in these treatments. He et al., (2009) observed a signi cant difference between tillage systems at soil depths down to 20 cm, but not deeper in the soil pro le.
Concentration of POC was signi cantly higher under the RM system than RW; this was ascribed to the fact that the C input rates are basically higher in the RM system over the RW, which substantially increased the concentration of POC. After 6 years of the experimentation, Jat et al. (2019) also reported that the maize-wheat system had higher TOC and POC than rice-wheat system under no-till condition; this was attributed due to 10 t ha −1 more biomass added in the fomer than the latter system. Addition of higher biomass increased the level of POC, which was also reported by Mapfumo et al., (2007 Interestingly, our study showed that the POC concentration was strongly in uenced by the texture of the soils (speci cally sand) at 0-20 cm depth. We observed a negative correlation (Table 7B)  Wakindiki 2012) that the fresh POM is sequestered by the sand particles whereas clay physically and chemically protect the decomposed POM in soils. Therefore, a negative correlation of silt and clay fraction with POC concentration in the Coochbehar soils indicated that the strong association or stabilization of C may not occur in the Coochbehar due to the lower amount of clay (11.7%) ( Table-1 Concentrations of MAOC at 0-5 cm depth was maximum under the ZT practice, due to residues remained at the soil surface as compared to the CT practice, where the residues were incorporated in the tilled layer (0-20cm). These results corroborate the ndings of Somasundaram et al. (2017). A strong relationship between MAOC and TOC found in our study ( Figure 7) indicated a favourable increment of MAOC with TOC. The MAOC is formed upon binding of organic matter (OM) to clay and silt particles (Mikutta and Kaiser 2011). Plant-derived labile compounds act as a main source of C binding agent to the mineral fraction (Cotrufo et al.,2015). Such labile compounds bind to the mineral fractions or incorporated into microbial biomass (Castellano et al., 2015). Promotion of soil C accumulation under ZT occurs from organic C input from residue retention. This is also observed in other studies. A signi cant contribution of crop residues to TOC at 0-10 cm occurred in the silt and clay fractions, which indicated that most of the young TOC was protected in the form of MAOC fraction under a long (10-year) no-tillage study (Saet al., 2001).
In the present study, the strati cation ratio (SR) of TOC, HWEC, and POC progressively increased with increase in the soil depth, due to the decrease in the TOC concentrations along the soil depth (Table 6).
Compared with CT, the strati cation of TOC with depth is a spontaneous process which is mainly induced by the continuously higher input of C at the soil surface and less in the subsequent soil layers under ZT treatment, irrespective of the site and the cropping system. Similar results were reported by Sa and Lal (2009) and Ferreira et al. (2013). Franzluebbers (2002) reported that strati cation ratios of soil C and N pools for the four soils in Alberta and British Columbia were minimally, and variably affected by tillage system, which was unlike that observed for soils in Texas and Georgia. He observed that there was high strati cation ratio of soil C and N pools under CT in Alberta/British Columbia, clearly indicated that the soil degradation with inversion tillage may not have been affected in comparison to the other regions. He also considered other factors and among them variability in climate, played a signi cant role in Alberta.
In the context of present study, similar variation in TOC, HWEC, and POC was due to variable mixing of the residues in CT at the three sites (FS-1,4 and 5), which resulted in comparatively lower amount of TOC, HWEC and POC at 5-10 cm than ZT. Similarly, in FS-5, the SR was found to be higher in CT than ZT in 0-5/10-20 cm (Table 6). Melero et al. (2012), also observed that the relative proportion of variation within the factors contributing to the variations (tillage 54±15%, soil depth increment 25±14%, crop rotation 13±7%, and N fertilizer rate 8±3%) in the SR of TOC. The results presented in our study, indicate that the SR of the Malda soils was higher for TOC, HWEC and POC under the ZT than that under CT (Table 6), due to the inherent higher stock of TOC and the heavier soil texture in the soils of old alluvial Inceptisol (Malda) than the recent alluvial Entisol (Coochbehar) ( Table 5). Note: * and ** represent that correlation is signi cant at the 0.05 and 0.01 level (2-tailed) respectively; HWEC = hot water extractable C; TOC = total organic C; POC = particulate organic C The overall scenario with respect to the status and distribution of TOC and its fractions, in the seven sites was found to vary due to the tillage, cropping system, and soil types. From the interaction effect (CS × T), our results showed that the RM system in CASI under ZT improved the TOC and its fractions over CT

Conclusions
At the end of seven cropping seasons, the study showed a strong correlation of C fractions with TOC under both the CT and ZT practices and the RM and RW cropping systems. The result indicates that the labile C fractions represent a portion of TOC with different turnover rates and are important in judging the soil quality. The CASI practice under ZT increases the soil sequestration of C due to the addition of residue in the soil, though this increase varied with cropping system while the RM proved superior to RW.
The RM system increased the C turnover rate in both soil types and the amount of clay in these soils in uenced the stabilization/storage of C. Placement of residues on the surface results in slow decomposition of the residues and hence gradual loss of the added organic matter helps in reducing the loss to the environment. The concentration of TOC and its fractions increased with increasing soil depths.
The contribution of C fractions to TOC were in the order: MAOC (65-80%)>POC (18-32%)>HWEC (1.5-2.4%). The heavier textured Inceptisols could accumulate more C fractions compared to light textured Entisols and the former soils has a strong association or stabilization of C which corroborated from positive correlation (p<0.05) of TOC and POC with clay. Strati cation of the C in 0-5 cm soil depth may result in an imbalance in the distribution of C which is more prominent in clayey soils (old alluvial Inceptisol) than the sandy soils (recent alluvial Entisol). The input of organic material is critical to the long-term maintenance of SOM. Therefore, the residue management practices are likely to affect organic matter content in different soil types under different tillage and cropping systems.