Comparison of slag-based gypsum with commercial gypsum as a nutrient source on soil properties, nutrient uptake and yield of rice (Oryza sativa L.) under aerobic and wetland conditions

ABSTRACT Slag-based gypsum (SBG) is produced from LD slag (Linz-Donawitz slag) which is a byproduct of the steel industry. As the constituents of SBG are almost similar to the commercial gypsum (CG), two field experiments on acidic and alkaline soil under aerobic and wetland conditions were conducted to know their performance on rice. The treatments included recommended dose of fertilizer (RDF) as control, 300, 450, 600 and 750 kg ha−1 of SBG and CG along with RDF. Application of 750 kg SBG ha−1 recorded higher grain (6.15 and 9.22 t ha−1, respectively) and straw yield (10.80 and 12.71 t ha−1, respectively) in acidic and alkaline soils. The application of 750 kg SBG ha−1 increased the pH of acidic soil from 4.56 to 5.00 and from 9.04 to 9.37 in alkaline soil. Available nutrients in the post-harvest soil increased with the increase in the application rates of SBG and CG in soil. There was an increase in total uptake of the nutrients by the rice crop with an increase in application rates of SBG and CG. Application of a graded levels of SBG recorded higher grain and straw yield, nutrient uptake and nutrient availability in comparison to CG.


Introduction
Gypsum is a soft sulphate mineral composed of calcium sulphate dihydrate, with the chemical formula CaSO 4 · 2H 2 O. The use of gypsum in agriculture was reported as early as the beginning of the 20 th century (Kelley and Arany 1928). Compared with limestone, gypsum is 200 times more soluble when applied to soil at a neutral pH (Electric Power Research Institute (EPRI) 2006). Gypsum's solubility allows movement of Ca (calcium) and S (sulphur) through the soil profile into rooting zones (Chen and Dick 2011). Therefore, in addition to supplying Ca and S for plant nutrition, gypsum helps to prevent soil particulate dispersion, decreases surface crust formation, aids in seedling emergence, increases water infiltration, and decreases the loss of soil and nutrients due to surface runoff and erosion (Chen and Dick 2011), and reduces reactive phosphorus (P) losses from agricultural fields . High soil acidity in the subsurface layer may restrict root growth and decrease water and nutrient uptake, thereby leading to low crop yields (Dalla Nora and Amado 2013;Zandoná et al. 2015). In such soils, gypsum application favours root growth in deep soil layers, both by supplying nutrients (Ca and S) and by causing a decrease in Al 3+ activity in the soil subsurface and alleviating its phytotoxic effects (Caires et al. 2016).
Gypsum, as a source of Ca, improves plant growth (Clark et al. 2001). However, naturally mined gypsum poses environmental threat such as air pollution, soil erosion etc. Therefore, synthetic gypsum produced from industrial waste such as LD slag (Linz-Donawitz slag) can be useful in this scenario by effective conservation of natural gypsum. Slag-based gypsum (SBG) is synthesized from the -60 mesh LD slag fines at the Chemical Laboratory at Tata Steel Limited, Jamshedpur (Ashrit et al. 2015). The significance of SBG in agricultural use can be well understood by the fact that it can be an excellent replacement for mined gypsum with its better plant nutritional value in terms of sulfur, calcium, phosphorus, iron, and silicon (Si), and also the traces of micronutrients like manganese (Mn), copper (Cu), boron (B), nickel (Ni), molybdenum (Mo), etc . Hence, there is need to evaluate SBG as a possible alternative for natural mined gypsum as nutrients source for different crops.
Rice (Oryza sativa L.) is the staple food for more than 50% of world population and its yearly production governs the world food security (Dass et al. 2015). It is the staple food for more than 65% of Indian population contributing approximately 40% to the total food grain production, thereby, occupying a pivotal role in the food and livelihood security of people. The majority (75%) of Asia's rice is produced in irrigated lowland fields, where irrigation requirements are often high (Kreye et al. 2009). Ensuring food security of the country without exhausting natural resource is one of the major aims of the researchers. Therefore, growing aerobic rice could be a solution in this regard (Jinger et al. 2021). Water-saving production systems such as aerobic rice may provide viable adaptation strategies for farmers who want to continue growing rice under water limited conditions. Application of gypsum in wetland rice cultivation as nutrients source was studied by many researchers (Lee et al. 2002;Singh et al. 2016). However, the efficiency of gypsum in aerobic rice cultivation has not been studied so far. Hence, the present investigation aims to study the effect of SBG and CG as sources of gypsum in both acidic and alkaline soil under two different rice cultivation condition on soil properties, nutrient uptake and yield of rice.

Experimental location
Two field experiments were carried out in two contrasting soils under different conditions. The two fields were of contrasting pH with Hassan soil belonging to acidic and Chamarajanagara soil belonging to alkaline pH in Karnataka, South India. The Hassan field was situated in the Southern Transition Zone of Karnataka between N 12°57`0.24`` and E 075°58`18.0`` with an altitude of 916 m above mean sea level and Chamarajanagara field was situated at Southern Dry Zone of Karnataka between N 11° 55ʹ 34.12" and E 076° 56ʹ 37.43" with an altitude of 709 m above mean sea level. The initial soil properties of the experimental locations were presented in Table 1.

Crop husbandry
The treatments included recommended dose of fertilizers (RDF) as control and four graded doses (300, 450, 600 and 750 kg ha −1 equivalent to 0.68, 1.01, 1.35 and 1.69 kg per plot at Hassan and 0.60, 0.90, 1.20 and 1.50 kg per plot in Chamarajanagara, respectively) of SBG and CG along with RDF. The RDF (100: 50: 50 as N, P 2 O 5 , K 2 O kg ha −1 ) as per the package of practice was used at both locations. Uniform dose of 50 kg ha −1 each of P 2 O 5 (di ammonium phosphate) and K 2 O (muriate of potash) were applied at the time of sowing/planting. N (nitrogen) in the form of urea was applied at 100 kg ha −1 in three splits viz., 50% at the time of sowing as basal, 25% each at 30 and 60 days after sowing/ planting. At harvest stage, above ground portion of rice plants were harvested from each experimental plot (22.5 m 2 in Hassan and 20 m 2 in Chamarajanagara) and threshed to separate the grains. After threshing, the grain and straw were kept for sun drying to attain constant weight. After sun drying, grain and straw samples were randomly collected plot wise from experimental field and weighed for moisture content till a constant weight was obtained. After attainment of the uniform weight, grain and straw samples were weighed for the yield calculation.

Sources of gypsum and its composition
Two different sources of gypsum (SBG and CG) were used to carry out the experiment. SBG was procured from Tata Steel Ltd. Jamshedpur, India. The gypsum materials were powdered using ball mill grinder and 0.1 g of fine powdered sample was predigested with 7:2:1 ratio of nitric acid, hydrogen peroxide and hydrogen fluoride. The predigested samples were digested using a microwave digester (Milestone-START D) at 150°C with following steps: 1200 W for 15 minutes, 1200 W for 10 minutes and venting for 10 minutes. The digested samples were volume made to 50 ml using deionized water and analysed using ICP -OES (Thermofisher -model 7000 series) with the help of multi nutrient standard solution  S and 1.37% of Si as SiO 2 . The particle size of SBG varies from 1.8 to 500 μm; the volume under 1.8 μm is 5.41% and that under 500 μm is 99.99% (Ashrit et al. 2015). Whereas, the particle size of mined gypsum ranges from 2000 to 4000 μm (Chen and Dick 2011). Both SBG and CG sieved with 200 mesh was used in the present investigation.

Aerobic rice cultivation at Hassan
The experiment was conducted in aerobic condition with variety Anagha in a randomized block design with nine treatments and three replications. Calculated dose of SBG and CG as per treatment details were broadcasted to each experimental plot and were mixed manually with the soil 15 days before the sowing. The seeds were directly sown with a spacing of 30 × 10 cm in a plot size of 4.5 × 5 m.

Wetland rice cultivation at Chamarajanagara
The experiment was conducted in wetland condition with variety Gangavathi Sona in a randomized block design with nine treatments and three replications. Calculated dose of SBG and CG as per treatment details were broadcasted to each experimental plot and were mixed manually with the soil 15 days before the transplanting. One rice seedling of 21 days old was transplanted with a spacing of 20 × 10 cm in a plot size of 5 × 4 m.

Plant and soil analysis
Ten rice plants randomly harvested at maturity stage of crop were thoroughly washed with deionized water. Grain and straw were separated and oven dried at 70°C to obtain constant weight, powdered and stored for analysis. The final moisture percentage obtained for grain was 7.41%. Powdered plant sample (0.1 g) was pre-digested with 7 mL HNO 3 (70%) and 3 mL H 2 O 2 (30%) in PTFE (Poly Tetra Fluoro Ethylene) vessels and later digested using a microwave digester (Milestone-START D) at 150°C with following steps: 1200 W for 15 min with a ramp rate of 7°C per min, 1200 W for 10 min and venting for 10 min. The digested samples were stored in clean plastic tubes of 50 mL capacity, after making up the volume to 50 mL using deionized water (Campbell and Plank 1998).
The digested samples were used to determine nutrient content following standard procedures. Si content of plant samples were determined following the procedure of (Ma and Takahashi 2002) with UV-visible spectrophotometer at 600 nm. Soil samples from each plot were collected from 0 to 15 cm depth after the harvest of the rice, air dried, powdered and passed through 2 mm sieve and used for further analysis. Soil pH and EC were measured in a suspension with a 1:2.5 soil: water ratio (Jackson 1973). Soil textural class was determined by following the international pipette method (Jackson 1973). Plant available nitrogen (N) was determined by alkaline potassium permanganate method (Subbiah and Asija 1956). Estimation of available phosphorus pentoxide (P 2 O 5 ) in acidic soil was done by Bray's 1 method (Bray and Kurtz 1945) and in alkaline soil by Olsen's method (Olsen et al. 1954). Available potassium (K), exchangeable calcium (Ca) and magnesium (Mg) were extracted by using neutral normal ammonium acetate. Potassium (K) was analysed by using flame photometry (Jackson 1973) and exchangeable Ca and Mg were determined by complexometric titration method (Baruah and Barthakur 1997). Plant available sulphur (S) was analysed by turbidimetric method (Williams and Steinbergs 1959). Plant-available Si content in soil was extracted by 0.01 M CaCl 2 (Haysom and Chapman 1975) and 0.5 M acetic acid (Korndorfer et al. 2001;Narayanaswamy and Prakash 2010) and estimated using UV visible spectrophotometer.

Nutrient use efficiency
Agronomic efficiency and apparent nutrient recovery have been used here to study the nutrient use efficiency by rice crop with the application of 300, 450, 600 and 750 kg ha −1 of SBG and CG respectively, along with RDF. Agronomic efficiency is defined as the additional economic yield acquired per unit nutrient applied and apparent nutrient recovery reflects the plant's ability to acquire applied nutrient from the soil (Fixen et al. 2015).
Agronomic efficiency ðkg grain kg nutrient applied Where, Y = yield of harvested portion of crop (kg ha −1 ) with nutrient applied; Y0 = yield with no nutrient applied (kg ha −1 ); F = amount of nutrient applied (kg ha −1 ); U = total nutrient uptake in above ground crop biomass with nutrient applied (kg ha −1 ); U0 = nutrient uptake in aboveground crop biomass with no nutrient applied (kg ha −1 ).

Statistical analysis
The statistical analysis of the data was carried out by using the standard statistical method of analysis of variance (Panse and Sukhatme 1985). The parameters that were significantly different in ANOVA were then used for multiple comparisons between all treatments using least significant difference (LSD) test at 0.05 probability level (P ≤ 0.05).

Effect of SBG and CG on the yield of rice grain and straw
Application of SBG and CG significantly increased the grain and straw yield of rice in both acidic (aerobic rice) and alkaline (wetland rice) soil when compared to RDF ( Figure 1). Among the treatments, application of 750 kg SBG ha −1 recorded higher grain yield in acidic (6.15 t ha −1 ) and alkaline soil (9.22 t ha −1 ). Significantly lower grain yield in acidic (4.59 t ha −1 ) and alkaline soil (7.52 t ha −1 ) was recorded in the control plot (RDF). Application of 750 kg SBG ha −1 recorded 33.85 and 22.49% increase in aerobic and wetland rice grain yield respectively when compared to control (RDF alone) treatments ( Figure 1). Similarly, significantly higher straw yield in aerobic and wetland rice (10.80 and 12.71 t ha −1 , respectively) was recorded with the application of 750 kg SBG ha −1 which accounts to 31.31 and 51.11%, respectively over control. While, lower straw yield in acidic soil (8.23 t ha −1 ) and alkaline soil (8.41 t ha −1 ) were recorded in RDF applied plots. In terms of grain yield, the higher per cent increase was recorded in aerobic rice (22.49%) while, the higher per cent increase of straw yield was recorded in wetland rice (51.11%) with the application of 750 kg SBG ha −1 .

Effect of SBG and CG on nutrient uptake by rice grain and straw
There was a significant increase in the uptake of macro nutrient (N, P and K) (Figure 2), secondary nutrient (Ca, Mg and S) ( Figure 3) and Si (Figure 3) with the application of a graded levels of SBG and CG to both aerobic and wetland rice. Total uptake of all the macro nutrients, secondary nutrients and beneficial element Si uptake were recorded to increase significantly with the application of a graded levels of SBG in comparison to CG in both aerobic and wetland rice. In general, application of 750 kg SBG ha −1 recorded higher nutrient uptake in both aerobic and wetland rice. However, lower nutrient uptake was recorded with the application of RDF alone.

Soil pH
Application of graded levels of SBG and CG recorded a significant effect on pH of acidic and alkaline soil ( Table 2). The pH of both acidic and alkaline soils increased in proportion to the application levels of SBG and higher pH of post-harvest acidic (5.00) and alkaline soil (9.37) was recorded with the application of 750 kg SBG ha −1 . However, acidic soil recorded a higher magnitude of increase (0.52) in soil pH when compared to alkaline soil (0.25).

Soil EC
Application of graded dose of SBG and CG significantly increased the EC in both acidic and alkaline soil ( Table 2). EC of acidic soil ranged from 0.11 to 0.24 dS m −1 and that of alkaline soil ranged from 0.32 to 0.44 dS m −1 . Higher EC in acidic and alkaline soil was recorded with the application of 750 kg SBG ha −1 . Lower EC in acidic and alkaline soil (0.11 and 0.32 dS m −1 , respectively) was recorded with the application of RDF alone. In acidic and alkaline soil, 300 kg SBG ha −1 and 450 kg CG ha −1 had the same EC (0.14 and 0.35 dS m −1 , respectively).

Effect of SBG and CG on major nutrients of soil
Application of various levels of SBG and CG significantly increased the available N content in alkaline soil but recorded no significant effect in acidic soil (Table 3). Higher available N in alkaline soil (179.2 kg ha −1 ) was recorded with the application of 750 kg SBG ha −1 which was found to be on Available P 2 O 5 content was significantly increased with the application of various levels of SBG and CG in acidic soil, whereas there was no significant effect in alkaline soil (Table 3). Though nonsignificant, higher available P 2 O 5 (48.73 kg ha −1 ) in alkaline soil was recorded with the application of 750 kg SBG ha −1 . Application of 450 kg SBG ha −1 and 600 kg CG ha −1 recorded an equivalent amount of available P 2 O 5 content (44.46 kg ha −1 ) in alkaline soil. Contrarily, higher available P 2 O 5 (137.6 mg kg −1 ) in acidic soil was recorded with the application of 750 kg CG ha −1 and was found to be on par with 750 kg SBG ha −1 .
Application of graded dose of SBG and CG significantly increased the available K 2 O in acidic soil whereas, there was no significant effect in alkaline soil (Table 3). Higher available K 2 O (254.5 kg ha −1 ) in acidic soil was recorded with the application of 750 kg SBG ha −1 which was found to be on par with 600 kg SBG ha −1 .

Exchangeable Ca and Mg
Compared to the RDF applied plots, there was significant increase in exchangeable Ca and Mg in acidic soil with the application of graded levels of SBG and CG, while there was nonsignificant effect in alkaline soil (Table 4). Exchangeable Ca and Mg in acidic soil ranged from 1.99 to 2.55 c mol (p + ) kg −1 and 1.18 to 1.48 c mol (p + ) kg −1 , respectively. Higher exchangeable Ca and Mg in acidic soil was recorded with the application of 750 kg SBG ha −1 . Application of 300 kg SBG ha −1 , 300 kg CG ha −1 and 450 kg CG ha −1 recorded an equal amount of exchangeable Ca content (11.59 c mol (p + ) kg −1 ) in alkaline soil. Whereas, application of 600 kg SBG ha −1 and 750 kg CG ha −1 recorded same exchangeable Mg content (15.08 c mol (p + ) kg −1 ) in alkaline soil.

Available S
Available S in acidic and alkaline soil significantly increased with the application of graded levels of both SBG and CG (Table 4). There was a proportional increase in available S of both soils with the application rate of SBG and CG. Application of 750 kg SBG ha −1 recorded higher S content in both acidic (41.00 mg kg −1 ) and alkaline soil (65.08 mg kg −1 ). In comparison to the initial value, the available S content in both soils decreased after the rice cultivation. However, the plots receiving higher level of SBG recorded an increase over the initial level as well as control plots in both soils.

AASi and CCSi
Application of graded levels of SBG and CG significantly increased acetic acid (AASi) and calcium chloride extractable Si (CCSi) in both acidic and alkaline soil (Table 4). Higher AASi in acidic and alkaline soil (40.83 and 90.33 mg kg −1 , respectively) were recorded with the application of 750 kg SBG ha −1 which was found to be on par with the application of 450 kg SBG ha −1 , 600 kg SBG ha −1 and 750 kg CG ha −1 . In general, available Si (AASi and CCSi) increased after rice cultivation in both soil when compared to the initial level in soil. However, CCSi in alkaline soil was recorded to decrease after rice cultivation in comparison to the initial level. Furthermore, alkaline soil recorded substantial increase in available Si (AASi and CCSi) over control when compared to acidic soil under aerobic rice cultivation.

Agronomic efficiency
Agronomic efficiency of N, P and K were recorded higher with the application of SBG in comparison to CG in both soils (Figure 4). A significant positive correlation was found between agronomic efficiency and nutrient uptake of the rice grain in both acidic and alkaline soils. However, alkaline soil recorded higher agronomic efficiency in comparison to acidic soil with application of both sources of gypsum. This result explains that with the increase in agronomic efficiency, the nutrient uptake was also recorded to increase. This further indicates an increase in nutrient uptake by the crop with addition of graded levels of both sources of gypsum.

Apparent nutrient recovery
A similar trend was observed in apparent nutrient recovery ( Figure 5). The apparent nutrient recovery has shown significant correlation with the addition of graded levels of SBG and CG. In general, apparent nutrient recovery of N, P and K were recorded higher in alkaline soil when compared to acidic soil and SBG performed better in increasing the apparent nutrient recovery of these major nutrients. This result suggests that the application of either source of gypsum improves nutrient use efficiency in rice crop. However, SBG has shown better nutrient use efficiency when compared to CG.

Effect of SBG and CG on the yield of rice grain and straw
Significant increase in yield of rice grain and straw with the application of graded levels of SBG over CG in both the soils can be attributed to the increase in availability of essential nutrients like Ca and S in addition to micronutrients (Fe, Mn and Zn) as well as beneficial nutrients such as Si. Application of SBG might have created a favorable condition of acidic soil under aerobic rice by reducing the Al from exchangeable sites and increasing Ca content of the soil. In addition, the Ca released from SBG dissolution might have enhanced the soil flocculation, structural stability and water retention in soil. Chen et al. (2001) reported a significant increase in yield of alfalfa (Medicago sativa L.) with the application of FGD (flue gas desulfurization) gypsum in an acidic soil (Wooster silt loam). Application of SBG promoted better root growth of the rice by increasing the concentrations of primary (P) secondary (Ca, Mg and S), micro (Fe, Mn and Zn) and beneficial nutrients (Si). Zhao et al. (2018) also noticed an increase in yield of rice with the application of FGD gypsum in saline-sodic soil with pH 9.8. Similarly, Lee et al. (2002) observed that application of fly ash and gypsum mixture at 40 Mg ha −1 increased the rice yield in an acidic soil having pH 5.2.

Effect of SBG and CG on nutrient uptake by rice grain and straw
Higher nutrient uptake with the application of SBG in comparison to CG can be attributed to its higher solubility owing to the smaller size of SBG (500 μm). Bolan et al. (1991) opined that the industrial gypsum like phospho-gypsum and FGD gypsum were more soluble in comparison to mined gypsum because the mined gypsum contains impurities like CaCO 3 coating which hampers its dissolution. Higher yield responses to gypsum addition in acidic soils, in most cases, were due to either improvement of Ca status or to a decrease in bioavailable Al in subsurface soil, which in turn stimulated root growth (Farina et al. 2000). Decreased Al 3+ saturation in acidic soils under aerobic rice in the deeper layers with the application of CG as well as SBG boost root development (Caires et al. 2016) and hence increase water and nutrient uptake, thereby increasing the yield. This finding is further supported by Chen et al. (2008) who observed that the application of FGD gypsum (33 kg S ha −1 ) with N (0-233 kg N ha −1 ) promoted corn growth and uptake of N in silt loam soil of Ohio, U.S.A. Also, synergistic relation between N and S was reported by Salvagiotti et al. (2009) which might be a reason for an increased N uptake through gypsum application. Higher Si uptake in alkaline soil in comparison to acidic soil can be attributed to higher grain and straw yield obtained in the alkaline soil (9.22 t ha −1 and 12.71 t ha −1 ) in comparison to acidic soil (6.15 t ha −1 and 10.80 t ha −1 ), respectively. Further, higher initial available Si (76.25 mg kg −1 and 53.05 mg kg −1 , respectively) AASi and CCSi in alkaline soil might have contributed in higher uptake of Si by the rice crop grown in alkaline soil in comparison to acidic soil with (14.56 mg kg −1 and 23.22 mg kg −1 , respectively) AASi and CCSi.

Soil pH and EC
The increase in pH can be attributed to the alkaline nature of SBG (pH 8.15). It may also be due to ligand exchange between SO 4 2from SBG and OH − on soil surface, and thereby releasing OH − ion into the soil solution further contributing for increase in the pH of soil. Prakash et al. (2020) reported an increase in pH of the soil with the application of SBG on maize crop in acidic and neutral soil. Further, Caires et al. (2006) noticed increase in pH of acidic soil with the application of gypsum on soybean crop. Further, Lee et al. (2008) reported an increase in soil pH with the application of fly ash and phospho-gypsum mixture in two different soils having pH 5.5 and 5.8 with rice crop. Zhao et al. (2018) reported a decrease in the pH of soil by 14.6% with the application of FGD gypsum in saline alkali soil with rice crop. Similar result was reported by Zhao et al. (2019) with the application of FGD gypsum in an alkaline soil after 17 years. Contrarily, application of SBG increased the pH of alkaline soil in the present investigation. Increase in pH of alkaline soil can be due to the infiltration of SO 4 2anion formed after dissolution of gypsum and moved downward or out of the field with drainage water (Zhao et al. 2018). An increase in pH was recorded in both acidic and alkaline soil with the application of both CG and SBG in the present investigation. Although, dosage of application of both SBG and CG for both aerobic and wetland rice was similar, no negative effects were recorded in terms of yield and nutrient availability.
Increase in EC of the post-harvest soil of both aerobic and wetland rice with the application of SBG may be attributed to its higher residual salts in comparison to CG. The increase in EC of both soils might be due to an increase in electrolyte concentration due to dissolution of gypsum and its higher Ca and S content (Tang et al. 2013). This result is corroborated by Ilyas et al. (1997) who noticed that application of gypsum to different crops (alfalfa, sesbania and wheat) on a saline-sodic soil increased the EC in top 20 cm soil. Further, increase in EC of soil with gypsum application was noticed by Khan et al. (2006) with wheat-rice crop in alkaline soil, Prakash et al. (2020) with maize crop in acidic and neutral soil and Laxmanarayanan et al. (2020) with groundnut crop in acidic and neutral soil.

Effect of SBG and CG on major nutrients of soil
A higher increase in available N content of post-harvest soil of both acidic and alkaline soil with the application of SBG can be attributed to the presence of Si (3.41% as SiO 2 ) in SBG as Si and N are reported to have a synergistic relationship. Similarly, White et al. (2017) reported that the application of silicate slag as Si source for wheat crop in acidic soil increased the plant-available N. Laxmanarayanan et al. (2020) reported an increase in available N in soil with the application of SBG for groundnut crop in two acidic soils having pH 6.93 and 5.77.
Increase in available P 2 O 5 content of soil in post-harvest soil of wetland rice field in alkaline soil with the application of SBG might be due to its higher content of P as P 2 O 5 (0.32%). Increase in available P 2 O 5 with the application of SBG can also be attributed to the presence of Si in the SBG as Si addition significantly increases P mobilization, by mobilizing Fe(II)-P phases from mineral surfaces (Schaller et al. 2019) and Si also decreases the P sorption in soils (Shariatmadari Hand Mermut 1999). Further, application of SBG could release more electrolytes, with a favorable ionic strength and Ca 2+ concentration in the soil solution attributing for improved chemical properties of soil which resulted in stronger phosphate adsorption. Norton and Rhoton (2007) reported that FGD gypsum application reduced P losses by 67% when compared to control. Jaakkola et al. (2012) also noticed that FGD gypsum application reduced P loses by 44% into the field scale simulation model. A higher available P 2 O 5 in acidic soil with application of CG over SBG in the present investigation can be attributed to higher uptake of P by the aerobic rice grain and straw with the application of SBG in comparison to CG.
Increase in available K 2 O content of the post-harvest soil in the present investigation could be due to the overall improvement of the chemical properties of soil with the release of Ca and SO 4 2ions through dissolution of SBG. Khan et al. (2006) noticed an increase in K content of the soil with the application of gypsum, however found to be non-significant in both wheat and rice crop in saline soil having pH 8.0.

Effect of SBG and CG on secondary nutrient
Increase in exchangeable Ca and Mg content of soil with the application of both sources of gypsum in acidic soil was expected because of both being a good source of Ca as well as Mg. There was 0.56 cmol ( + ) kg −1 soil increase in exchangeable Ca and 0.30 cmol ( + ) kg −1 soil increase in exchangeable Mg in acidic soil. Similar results were recorded by Zhao et al. (2019) with the application of FGD in a long-term field trial in alkaline soil, and Caires et al. (2006) with soybean crop grown in an acidic soil. Application of Fly ash and gypsum mixture also increased the exchangeable Ca and Mg content of rice crop grown in an acidic soil (Lee et al. 2002). Furthermore, higher increase of exchangeable Ca and Mg in alkaline soil in comparison to acidic soil could be due to the anaerobic condition of alkaline soil owing to wetland rice cultivation (Sahrawat 2012).
Application of both sources of gypsum increased the available S in post-harvest soil because of its coexistent with Ca in the gypsum. CrusciolA et al. (2010) reported that the application of gypsum increased the SO 4 2--S levels down the soil profile having acidic pH grown with rice and bean. Similar finding was reported by Caires et al. (2003) with soybean in acidic soil.

Effect of SBG and CG on plant-available Si
Significant higher plant-available Si in post-harvest soils in both aerobic and wetland rice was recorded with the application of SBG over CG application because of higher Si content (3.47%) in SBG than CG (1.37%), which could have contributed directly to its higher availability. This result is corroborated by Laxmanarayanan et al. (2020) who also noticed higher AASi and CCSi in post-harvest soils of groundnut crop grown in an acidic soil with the application of SBG. Further, Lee et al. (2002) recorded that a combination of fly ash and gypsum increased the SiO 2 content of the acidic soil grown with the rice crop. Narayanaswamy and Prakash (2010) noticed that acetic acid extracted Si content is higher in comparison to calcium chloride extracted Si. However, in the present investigation we noticed a higher CCSi over AASi in acidic soil. This might be due to a very low pH of the soil and also its highly weathered condition (Meunier et al. 2018). Further, levels of Si extracted by calcium chloride or acetic acid depends on the complex interaction of the molecules at mineralwater interface and may be controlled by dissolution and/or desorption (Hiemstra et al. 2007).

Agronomic efficiency
Higher agronomic efficiency with the application of SBG in comparison to CG can be ascribed to increased rice yield which resulted in increased nutrient uptake. A higher agronomic efficiency with the application of SBG in comparison to CG can be due to an increase in the availability of macronutrients in the soil which brought a significant increase in the straw and grain yield (Singh and Saha 1995). Further, higher agronomic efficiency of N, P and K in alkaline soil can be attributed to the higher grain yield obtained with application of graded dose of SBG in the alkaline soil in comparison to acidic soil.

Apparent nutrient recovery
Higher apparent nutrient recovery with the application of graded dose of SBG can be attributed to the increased uptake of the major nutrients. Increase in uptake of major nutrients (N, P and K) with the application of graded dose of SBG was also reported by Prakash et al. (2020) in maize crop grown in an acidic and neutral soil. Higher apparent nutrient recovery in alkaline soil in comparison to acidic soil can be attributed to the higher grain and straw yield obtained in the alkaline soil which increased the nutrient uptake by rice crop in comparison to acidic soil.

Conclusion
Application of SBG in varied levels increased available nutrients in the soil as well as their uptake and yield of rice in acidic and alkaline soils under both aerobic and wetland conditions. Application of 750 kg SBG ha −1 has performed better in both soils and rice growing conditions. Agronomic efficiency and apparent nutrient recovery by rice crop was higher with the application of 750 kg SBG ha −1 in both wetland and aerobic condition. In general, SBG performed better when compared to CG in both aerobic and wetland condition. Further, long-term studies are necessary to understand its effect on yield of rice and nutrient build up in soils.