Carbon Sequestration in Soil Aggregates under Different Cropping Patterns of Bangladesh

Land use change and cropping patterns are important factors for controlling carbon sequestration in soils and they may also change the relative importance of different mechanisms of soil organic matter stabilization. The study was conducted to investigate the state of carbon sequestration in soil aggregates under different cropping patterns of Khulna, Jessore and Chapainawabganj districts in Bangladesh. Thirty-six soil samples were collected from (0 100 cm depth) above mentioned regions of three physiographic regions: Ganges Meander Floodplain, Ganges Tidal Floodplain and High Barind Tract. The texture of the samples varied within three soil texture groups, Silt Loam, Silty Clay Loam and Silty Clay. The highest NSI value (0.89) was found under Wheat-Fallow-T. Aman cropping pattern in Silty Clay soils (sample No 15) and lowest value (0.59) was found Vegetables/Mustard-Fallow-T. Aman cropping pattern in Silt Loam soils (sample No 17). The highest value (735.20 mg∙kg) of active C was observed under Chickpea/mustard-T. Aman (Sample No 31) and the lowest value (619.23 mg∙kg) was found in case of Wheat-Fallow-T. Aman cropping pattern (Sample No 30). The highest SOC stock (1.62 Kg C m) was found in Silty Clay Loam soil under Mungbean/Ash gourd-T. Aman cropping pattern (Sample no 4) and the lowest SOC stock (0.35 Kg C m) was found in Silt Loam soil under Cauliflower/Pumkin/Spinach-T. Aman Cropping pattern (Sample No 2). Soil organic carbon associated with different size aggregates was the highest (3.14%) under Mungbean/Ash gourd-T. Aman (Sample No 20) and was the lowest (0.36%) under Cauliflower/Pumkin/Spinach-T. Aman cropping pattern (Sample No 2). Organic carbon content in aggregate size ranges > 2000 μm (SOC1), 2000 250 μm (SOC2), 250-53 μm (SOC3), and <53 μm (SOC4) varied from 0.36% 1.90%, 0.52% 2.10%, 0.50% 2.60% and 0.50% 1.62%, respectively. The percentages of SOC associated with <53 μm aggregates were higher than those of >2000 μm, 2000 250 μm and 250 53 μm, aggregates. Significant positive How to cite this paper: Amin, Md.S., Khan, Md.Z., Laskar, T. and Rabbi, S.M.F. (2020) Carbon Sequestration in Soil Aggregates under Different Cropping Patterns of Bangladesh. Open Journal of Soil Science, 10, 459-485. https://doi.org/10.4236/ojss.2020.1010024 Received: August 18, 2020 Accepted: October 17, 2020 Published: October 20, 2020 Copyright © 2020 by author(s) and Scientific Research Publishing Inc. This work is licensed under the Creative Commons Attribution International License (CC BY 4.0). http://creativecommons.org/licenses/by/4.0/ Open Access


Introduction
Global warming is a major threat to the environment and soil management practices are now believed to contribute significantly to changing environmental conditions [1]. Attention has been increasingly paid to soil organic carbon (SOC) pool and its dynamics in land use changes concerning terrestrial ecosystem carbon sink and the uprising atmospheric carbon dioxide [2]. The global soil carbon (C) pool of 2500 gigatons (Gt) includes about 1550 Gt of SOC and 950 Gt of soil inorganic carbon (SIC). The soil C pool is 3.3 times the size of the atmospheric pool (760 Gt) and 4.5 times the size of the biotic pool (560 Gt). The SOC pool to 1-m depth ranges from 30 tons/ha in arid climates to 800 tons/ha in organic soils in cold regions, and a predominant range of 50 to 150 tons/ha. The SOC pool represents a dynamic equilibrium of gains and losses. Conversion of natural to agricultural ecosystems causes depletion of the SOC pool by as much as 60% in soils of temperate regions and 75% or more in cultivated soils of the tropics. Severe depletion of the SOC pool degrades soil quality, reduces biomass productivity, and adversely impacts water quality, and the depletion may be exacerbated by projected global warming [1]. Carbon stock occurring in the topsoil is the most important, because this is the portion more influenced by external environmental and human factors and therefore the mainly susceptible to mineralization or synthesis processes. On the contrary, in deep layers, soil organic carbon is more stable and less liable to transformation. Soil surface layers account for the main part of the soil carbon stocks: on average, the 47% of the carbon stored in soils is held in the first 30 cm; about 2/3 are stored within a depth of 50 cm and the 80% within one meter. Soil carbon storing capability depends on many factors, such as pedoclimatic conditions, crop management practices and the starting soil carbon levels and on the interactions among them.
Urgency of meeting increased demand for agricultural produce is rapidly degrading soil quality and exacerbating degradation. Most agricultural soils have low soil organic matter (SOM) reserves due to fertility-mining practices and widespread problem of soil degradation. This decline is also attributed to removal of crop residue, and changes in cropping systems etc. Crop yield and use efficiency of input are also adversely affected by low levels of SOC pool [3].
Land use practices that result in a net accumulation of SOM in the soil are said to be C sequestering because they result in a net removal of C from the at-Md. S. Amin et al. Open Journal of Soil Science mosphere [4]. Several studies in temperate and tropical regions reported that no-tillage practices substantially increase SOM storage and improve soil aggregation [5]. Soil aggregation can increase SOC storage by reducing loss by erosion and from mineralization. Soil organic matter can be physically protected from microbial mineralization through sorption to clay minerals and enclosure within soil aggregates [6]. Improvement in soil aggregation is an important factor influencing SOC sequestration in soils [7]. Soil aggregation is an important process of C sequestration and hence a useful strategy to mitigate the increase in concentration of atmospheric CO 2 [8].
Tillage causes loss of soil organic matter through destruction of macro-aggregates and microbial mineralization of the physically protected SOM pool. Micro-aggregates and their associated SOC are stable to tillage, and this "passive or chemically protected" SOC pool represents the minimum level of SOC. Rice-wheat surface soils (0 -15 or 0 -20 cm) from Bangladesh increase in SOC as their silt + clay fraction increases. This variability may reflect differences in manure use, years under cultivation, tillage, sampling depth or length of time that soils are flooded [9]. Significant amounts of OC will not accumulate in soils in these environments without the protective effects of interactions with mineral surfaces (as provided by silt/clay) and the formation of aggregates. Hopelass et al. [10] reported that there was a trend of increase in concentration of SOC with decreasing aggregate size, but significant differences in these parameters in different aggregate size fractions were found only in few cases. The SOC concentration was higher in >0.25 mm than in <0.25 mm aggregates. The SOC sequestration rate by judicious use of inorganic fertilizer was the greatest in the grain-meadow rotation, while that by application of FYM was the greatest in the all grain rotation. Shrestha et al. [8] studied water stability of soil aggregates (WSA) and SOC associated with aggregates and primary particles of cultivated and forest soils. They concluded that microaggregates (<0.5 mm) were abundant (56% -63%) in cultivated soils but cultivated soils contained higher amounts of clay but less clay-associated SOC than forest soils.
Bangladesh is a tropical agrarian country with high population density and astounding food demand. Overexploitation of agricultural soils causes severe soil degradation in Bangladesh. Depletion of SOM is a widespread problem on croplands and grazing lands in the country. Most soils have low levels of SOC contents, ranging from 8 to 10 g/kg. Low external input of organic amendment causes depletion of SOC pool because nutrients harvested in agricultural products are not replaced, and are made available through mineralization of SOM. In some cases, soil is burnt to release nutrients contained in SOM. It is estimated that SOC loss due to agricultural activities in Bangladesh between 1967 and 1995 was 16.2 Mg C/ha, with a range of 3.8 to 30.5 Mg C/ha [3]. To meet the increased food demand cropping intensity must be increased and cropping pattern should be changed. Rabbi et al. [11] reported that in silt loam soil highest percent %SOM was estimated under Fallow-Fallow-T. Aman (2.90%) and lowest under Vegetable-Sugarcane/Jute pattern (0.86%). The soils under Vegetable, Open Journal of Soil Science Vegetable-Vegetable, Fallow-Jute-T. Aman, Fallow-Aus-Vegetable and Rabi-Jute-T. Aman patterns had relatively lower %SOM than other patterns. The %SOM under Sesame-Fallow-T. Aman pattern varied in different agricultural fields and it was 1.22% to 2.11%. In loam soils highest %SOM was obtained under Rabi-Jute-Turmeric (1.54%) and lowest under Boro-Fallow-T. Aman pattern (0.55%). Soils under Fallow-Fallow-T. Aman, Boro-T. Aman/Fish and Fallow-Jute-T. Aman patterns also had higher %SOM whereas Soils under Boro-Shrimp had lower %SOM in loam soil. The %SOM in silty clay loam soil under Boro-Fallow-T. Aman varied between 0.96% and 1.54%. The highest %SOM (2.05%) was obtained under Wheat-Jute-T. Aman pattern in silty clay soil and lowest (0.76%) in clay loam soil under Vegetable-Aus/Jute/Vegetable pattern. Soils under Fallow-Fallow-T. Aman in silt loam soil had higher %SOM than in loam soil under same pattern. The higher %SOM was obtained under Boro-Fallow-T. Aman pattern in silty clay loam soil than loam soils. It was reported that micro-aggregation of soils was also influenced by cropping patterns of the studied agricultural fields. The aggregate stability of soils of Ganges Floodplain is closely related with soil organic matter [12]. Therefore, the objectives of the research project were to evaluate the state of SOC sequestration in the agricultural soils under different cropping patterns and estimate the carbon stock of soils.

Methods and Materials
The study was conducted on agricultural soils of Jessore Sadar and Bagherpara  Table 1. Thirty-six soil samples (Core and Bulk sample) were collected from 0 -100 cm soil depth and all soil samples were kept in sealed plastic bags. Marking and labelling was performed with a detailed description of the selected sampling site on both the soil-plastic bags, and preserved in plastic bags until arrival at the laboratory for sample preparation. Then samples were air dried by spreading the soils on separate sheet of papers. After drying in air, the larger aggregates were broken through gentle crushing with a wooden hammer. A portion of the crushed soils was passed through by using different size sieves. The ranges were >2000 µm, 2000 -250 µm, 250 -53 µm and <53 µm aggregate size range. The sieved soils were then preserved in plastic bags and labeled properly. These samples were later used for various chemical analyses.

Soil Physical Properties
The particle size analyses of the soils were carried out by combination of sieving and hydrometer method as described by Gee and Bauder [13]. Textural classes were determined by using Marshall's Triangular Coordinator system. Bulk density of soil was determined by core method as described by Blake        . Sampling sites at Nachol upazila. [14]. Particle density was determined by pycnometer meter as described by Blake and Hartge [14] and total porosity was calculated from bulk and particle density.

Aggregate Stability (Normalized Stability Index) by Wet Sieving of the Aggregates
The stability of aggregates was determined by the method as described by Six et al. [15]. For the determination of aggregate stability soil samples were air dried and crushed by a wooden hammer. The crushed soils were then sieved through 8 mm sieve. The air-dried soils that were passed through 8 mm sieve but retained on 2 mm sieve divided into 8 -2 mm, 2 -0. 25 [5].
The whole soil disruption level (DL) was calculated as: where, n = number of aggregate size classes. i = 1 for the smallest size class.
The disruption level of a size class upon slaking (DLS i ) was calculated by the following formula: where, DLS i = disruption level for each size class i; P io = proportion of total sample weight in size class i before disruption (i.e., rewetted); P i = proportion of total sample weight in size class i after disruption (i.e., slaked); S io = proportion of sand with size i in aggregates of size i (=aggregate-sized sand) before disruption; S i = proportion of sand with size i in aggregates of size i in aggregates after disruption.
The whole soil DL (max) was calculated by the following formula: [DLS i(max) ] was calculated with the following formula: P p = primary sand particle content with the same size as the aggregates size class after complete disruption of the whole soil.

Soil Chemical Properties
The soil pH was measured with the help of a glass electrode pH meter using soil water suspension ratio of 1:2.5 as described by Jackson [16]. Electrical conductivity (EC) of soil was estimated by EC meter (Jackson EC meter). Maintain the ratio with of soil to water of 1:5 and the result was converted to the ratio of 1:1 (soil:water) as suggested by USDA [17]. The soil organic carbon was determined by wet oxidation method of Walkley and Black's method as described by Jackson [16]. The soil organic carbon associated with >2000 μm, 2000 -250 μm, 250 -53 μm and <53 μm aggregates was determined. Stock of soil organic carbon was evaluated by the method of Ellert et al. [18]. Active soil organic carbon was determined by the method as described by Weil et al. [19].

Particle Size Distribution
The  (Table 2). The texture of the samples varied within three soil texture groups, Silt Loam, Silty Clay Loam and Silty Clay ( Table 2). This result resembles to several findings by many other researchers. SRDI staff [21] found that on the young Ganges meander floodplain, soils of ridges and inter-ridges depressions are silt loam to silty clay; and on the old Ganges meander floodplain, soils of ridges and depressions are loamy to clay in texture. They found that most ridge soils in old Brahmaputra floodplain are silt loam to silty clay loam and in inter-ridge depressions they are mostly silty clay loam to silty clay subsoil. Joshua and Rahman [22] found that the soils of the Tista floodplain, in general, contained a high percentage of silt and this property appears to be characteristic for the floodplain.

Bulk Density, Particle Density and Porosity
Bulk density varied in the range of 1.18 to 1.84 g/cc with mean value of 1.40 g/cc, particle density in the range of 2.44 to 2.59 gm/cc, porosity in the range of 25.88% -52.98% in soil samples under different cropping patterns ( Table 2). The variations in the particle density, bulk density and porosity values under different cropping patterns were significant (p < 0.01). To a remarkable degree, increased organic matter can counteract the ill effects of too much clay or too much sand. Increasing the SOM content usually increases total porosity and therefore decreases bulk density [23]. At very high levels of SOM, additional OM has little further effect on soil aggregation and influences bulk density mainly Open Journal of Soil Science because of its low particle density [24].

Normalized Stability Index (NSI)
In this research we found that the NSI varied from 0.51 to 0.89 under different cropping patterns (Appendix Table A1). The highest NSI value was found un-   Rabbi et al. [12] reported that aggregate stability of soils of Ganges River and Tidal Floodplains of Khulna region increased with increasing clay percentage of soils.
The value of NSI can vary between 0 to 1 [15]. The lower NSI indicated that the aggregates were not water stable. Earlier investigation by Rabbi et al. [12] indicated that water stability of aggregates of silt loam texture was low. The mineralogy of soils may play important role in aggregate stability [15]. However, in the present study mineral identification was not done. In the present study the NSI increased with decreasing cropping intensity. Soils under Fallow-Fallow-T.
Amon cropping pattern had higher NSI. The cultivation of rice in winter and monsoon season also decreased NSI of soils. Six et al. [25] reported that conventional tillage with high cropping intensity caused maximum destruction of soil.
It has been reported that stable aggregate of tilled soils is lower than that of no-tilled soils [26] due to aggregate breakdown in tillage process.
The cultivation of rice twice a year accelerates the aggregate destruction rate.
Adiku et al., [27] concluded that soils under rice base cultivation were more prone to degradation. The inclusion of shrimp in land use markedly decreased the NSI may be due to application of saline water in soil. Rabbi et al. [11] re-  decomposition by its incorporation into soil aggregates [28]. Soil organic matter can be: 1) physically stabilized, or protected from decomposition, through micro aggregation, or 2) intimate association with silt and clay particles, and 3) can be biochemically stabilized through the formation of recalcitrant SOM compounds [5]. Hence, Soil aggregation is an important process of C sequestration [8].

Soil Chemical Properties
The pH of the soils under different cropping patterns varied from 5.99 to 7.85 (Table 3) Table   A1). The highest value was observed under Chickpea/mustard-T. Aman (Sample      This fraction rarely comprises more than 10% to 20% of the total soil organic matter [29].

Analyses of Different Carbon
The content of soil organic carbon (SOC) can be affected by many factors, such as forest types [30], soil moisture, soil type, temperature and precipitation [31]. In our result, highest active carbon was under Chickpea/mustard/tomato-T.
Aman. It may be the cause of addition of fresh plant and animal residues, soil moisture, soil type, soil organic carbon stock and organic carbon %.
In Silt Loam soils (Figure 12  High cropping intensity under conventional cultivation commonly has decrease in organic carbon content, because the amount of organic material returned to the soils is considerably lower and tillage enhance the decomposition of native soil organic matter [32]. Paddy soils with periodic submergence were slightly higher in organic carbon content than upland soils.
The average SOC associated with aggregates was in the order of SOC4 > SOC3 > SOC2 > SOC1. The fresh organic carbon that was derived from the crops were first incorporated to larger aggregates and then shunted to micro aggregates after decomposition by soil microbes and this process was stimulated by disturbance by conventional tillage [33]. The percentage of SOC associated with 2000 -250 µm, 250 -53 µm and <53 µm aggregates increased with increasing clay percentages of soils. Wiseman and Puttmann [34] described the importance of specific surface of clays rather than percentage of clays in SOC sorption. Wattel-Koekkoek et al. [35] showed that smectites have large sorptive capacity for SOC. The existing reports on clay minerals of Ganges River Floodplain of Bangladesh concluded that illite is the dominant clay mineral of this floodplain [36].
The surface area of illite is about 70 -120 m 2 •g −1 . So, the capacity of illite to sorb SOC at may play an important role and it requires further research to conclude.

Soil Organic Carbon Stock (SOC Stock)
In order to estimate the potential of carbon sequestration in soils the original C stocks in soils need to be determined. Batjes [37] discussed the total soil C stock distribution by major ecological    zones, developed by FAO, can constitute a reference framework to evaluate and monitor soil C storage in soils.

Relationship among Soil Properties
Significant positive correlations were found between SOC stock and SOC1, SOC stock and SOC2, SOC stock and SOC3, SOC stock and SOC4 (Appendix Table   A3). The SOC content in all four size ranges under this study was positively correlated with each other.

Conclusion
The distribution of aggregate size classes was influenced by land cultivation, as tillage destroyed especially aggregates > 1000 µm. Average SOC stocks in the soils under investigation were in order of Silty Clay Loam > Silty Clay > Silt Loam. The average SOC associated with aggregates was in the order of SOC4 > SOC3 > SOC2 > SOC1. The aggregate associated SOC concentration was greater for microaggregates (<250 µm) than macroaggregates (>250 µm) in the soils.

Data Availability
The data used to support findings of this study are available from the corresponding author upon request.

Acknowledgements
This study was made possible with funding from University Grants Commission of Bangladesh for climate change issue. The support rendered by Khulna University towards implemented of the study is acknowledged. We would like to express our sincere gratitude to our supervisor Associate Professor Md. Sadiqul Amin for his continuous supervision, guidance, inspiration and thoughtful sug- gestion during completion of the work. Individual efforts alone can never contribute in totally to a successful completion of any venture. We would be failing in our duty if we did not state our gratitude and appreciation to the following individuals who have made the valuable contribution towards the review work.