Conservation agriculture based integrated crop management sustains productivity and economic profitability along with soil properties of the maize-wheat rotation

Field experiments were conducted to evaluate eight different integrated crop management (ICM) modules for 5 years in a maize-wheat rotation (MWR); wherein, ICM1&2-ˈbusiness-as-usualˈ (conventional flatbed maize and wheat, ICM3&4-conventional raised bed (CTRB) maize and wheat without residues, ICM5&6-conservation agriculture (CA)-based zero-till (ZT) flatbed maize and wheat with the residues, and ICM7&8- CA-based ZT raised bed maize and wheat with the residues. Results indicated that the ICM7&8 produced significantly (p < 0.05) the highest maize grain yield (5 years av.) which was 7.8–21.3% greater than the ICM1-6. However, across years, the ICM5-8 gave a statistically similar wheat grain yield and was 8.4–11.5% greater than the ICM1-4. Similarly, the CA-based residue retained ICM5-8 modules had given 9.5–14.3% (5 years av.) greater system yields in terms of maize grain equivalents (MGEY) over the residue removed CT-based ICM1&4. System water productivity (SWP) was the highest with ICM5-8, being 10.3–17.8% higher than the ICM1-4. Nevertheless, the highest water use (TWU) was recorded in the CT flatbed (ICM1&2), ~ 7% more than the raised bed and ZT planted crops with or without the residues (ICM4-8). Furthermore, the ICM1-4 had produced 9.54% greater variable production costs compared to the ICM5-8, whereas, the ICM5-8 gave 24.3–27.4% additional returns than the ICM1-4. Also, different ICM modules caused significant (p < 0.05) impacts on the soil properties, such as organic carbon (SOC), microbial biomass carbon (SMBC), dehydrogenase (SDH), alkaline phosphatase (SAP), and urease (URE) activities. In 0.0–0.15 m soil profile, residue retained CA-based (ICM5-8) modules registered a 7.1–14.3% greater SOC and 10.2–17.3% SMBC than the ICM1-4. The sustainable yield index (SYI) of MWR was 13.4–18.6% greater under the ICM7&8 compared to the ICM1-4. Hence, this study concludes that the adoption of the CA-based residue retained ICMs in the MWR could sustain the crop yields, enhance farm profits, save water and improve soil properties of the north-western plans of India.


Results
Five years' trends and pooled maize grain and stover yields. During the initial year, the maize grain yield did not differ significantly among the ICM modules, although the highest yield was recorded under the ICM 8 . Nevertheless, from the second year onwards, the different ICM modules had the significant (p < 0.05) impacts on the maize grain yield (Fig. 1a). The ICM 7 consistently produced the highest yield across the years, which was closely followed by the ICM 8 . Similarly, the highest stover yield across the years was recorded with ICM 7 , except first year (Fig. 1b). The highest pooled grain (5.2 Mg ha −1 ) and stover (8.7 Mg ha −1 ) yields were recorded with the ICM 7 , being close to the ICM 3-6&8 . On an average, the ICM 7&8 had produced 5.9-21% and 5.8-18.4% greater grain and stover yields, respectively, over the ICM [1][2][3][4][5][6] (Table 1).
Five years' trends and pooled wheat grain and straw yields. The different ICM modules did not impact the wheat grain yield significantly during the first three years. While, at the fourth year, the ICM 5 had the highest yield, being significantly higher than the ICM [1][2][3] , and subsequently in the fifth year, it was ICM 8 which outperformed significantly (p < 0.05) over the ICM 4 (Fig. 1c). Similarly, the straw yield did not differ significantly among the ICM modules in the initial three years, but significantly a greater yield was registered with the ICM 8 in the fourth and fifth years (Fig. 1d). However, the mean grain and straw yields under the ICM 5-8 (CA-based ZT) was 8.4-11.5% and 7-14% greater than the CT-based residue removed (ICM [1][2][3][4]  www.nature.com/scientificreports/ System yields in terms of maize grain equivalents. The ICM modules had a significant impact on the maize grain equivalents (M GEY ) across the years, except during the initial two years (2015- 16 and 2016-17), wherein the ICM 7 produced the highest yield during the 2017-18 and 2019-20, which was significantly greater than the ICM [1][2][3][4] to the tune of 19-22% and 17-26%, respectively. While, in 2018-2019, the highest yield was recorded with the ICM 8 , which was significantly higher than the ICM 1-4 by 16-22%. Averaged across the five years, the ICM 5-8 had 6-15% system M GEY advantage over the ICM [1][2][3][4] (Table 2).
Soil properties. The ICM modules had a significant impact on the variable soil properties i.e., soil organic carbon (S oc ), microbial biomass carbon (S MBC ), dehydrogenase activity (S DH ), alkaline phosphatase (S AP ) and soil urease (U RE ) activities (Fig. 4, Table 3).

Soil organic carbon (S oc ).
In the top 0.00-0.05 m soil depth, the highest S oc was recorded with the ICM 7, which was significantly higher than the ICM 2&4 . The increment in S oc under the ICM 7&8 over the ICM [1][2][3][4] was to the tune of 10.2-16.2%. Further, in the 0.05-0.15 m soil depth, the highest S oc was recorded with the ICM 6 , wherein it was significantly more than the ICM 3, but statistically (p < 0.05) similar to the ICM 1,2,4,5,7&8 . While, there were no significant differences among the ICM modules, with respect to the S oc , in the 0.15-0.30 m soil depth (Fig. 4a).

Soil microbial biomass carbon (S MBC ).
The highest S MBC in the 0.00-0.05 m soil depth was observed under the ICM 8 , wherein it was similar to the ICM 5&6 , but significantly higher than the ICM 1-4&7 . The ICM 8 had 6-23% greater S MBC than the ICM [1][2][3][4] . While, in the 0.05-0.15 m soil depth, the highest S MBC was recorded in the ICM 6 , being significantly greater than the ICM 1-5 to the tune of 12-22.8%, but similar to the ICM 7&8 . In contrast, at lower soil depth (0.15-0.30 m), the highest S MBC was observed under the ICM 3, and being greater than that of the ICM 1,2&4-8 (Fig. 4b).        www.nature.com/scientificreports/ Soil alkaline phosphatase (S AP ). The highest S AP in the top 0.00-0.05 m soil depth was recorded with the ICM 8 , being significantly higher than the ICM 1-5, but similar to the ICM 6&7 . Indeed, the ICM 7&8 resulted in 8.3-32.3% higher S AP compared to the ICM [1][2][3][4][5] . While, in the 0.05-0.15 m soil depth, the highest S AP was observed with the ICM 6 , where it was significantly more than the ICM 1-4&8 , but at par with the ICM 5&7 . Further, at 0.15-0.30 m, no significant difference in S AP was noticed among the ICM modules (Table 3).

Soil urease (U RE ).
The U RE in the 0.00-0.05 m soil depth was the highest with the ICM 8 , in which it was similar to the ICM 4-7 , but significantly greater than the ICM 1-3 . The increment in U RE under ICM 7&8 over the ICM 1-4 (CT modules) was to the tune of 12.7-27.2%. Similarly, in the 0.05-0.15 m, the highest U RE was recorded with the ICM 8 , which was significantly greater than the ICM 1-4, but similar to the ICM [5][6][7] . As expected, the ICM 7&8 produced 8-27% higher U RE compared to the ICM 1-4 . However, in the lowest soil layer (0.15-0.30 m), no significant differences in U RE were observed among the ICM modules (Table 3).

Sustainable yield index (S YI ).
Among the ICM modules in the maize, the ICM 7 had the greater S YI , but being at par to the ICM 1,5,6&8 , which was 12-15.2% greater than the ICM 2-4 . Again, S YI in wheat was the highest under the ICM 7 , similar with ICM 5&8 , being 17.9-25.3% greater than the CT-based ICM 1-4 modules. In the case of M WR , the S YI was the highest under the ICM 7&8 , which was 13.4-18.6% higher than the ICM 1-4 , and similar to ICM 5 (Fig. 5).

Discussion
The rice-wheat is the commanding rotation in northern India's ecologies. However, of late, from the resource exploitation to their judicious use for sustained yield, save water and improve soil-based properties is the focus 19,26 , besides achieving SDGs 12 . Seeing the degradation of natural resources, stagnation in crop yields and other constraints in adoption of rice-wheat rotation (R WR ), it is thus noteworthy to identify the alternative crops and cropping rotations to sustain the food security. Maize ˈQueen of cerealsˈ being a C 4 plant, has wider adaptability under the diverse climate, thus could be a striking substitute of rice 2 . Every year, in the rice-wheat belt of north western India, the ground water falls off by 0.30-0.40 m 27 , and therefore, acreage under maize is likely to increase with the time. It is clearly evident that rice is the main water consumer 28 , maize could be a potential choice for accompanying wheat in this area, as it saves irrigation water, fulfils demand for palatable fodder and industries. Rice residue burning rather than returning to the soil, is another concern which not only deteriorates the air quality, but also have acute effects on human health 8 . Thus, the M WR has a potential to replace the water guzzling rice under the R WR . The CA-based ICM practices in M WR would intend for sustainable residue recycling, improve soil properties 19,29 and sustain long-term production 30 .
Our findings confirmed the yield gains (14.6%, maize; 11.2%, wheat) under the CA-based ICM 5-8 over the ICM 1-4 , however, the M GEY enhanced by 12.3% (5 years' av.). The ICM 5-8 , proved superior because of ZT, crops residue, and eventually the efficient use of inputs 31,32 along with L BFs consortia and A MF . Most soil organic matter (S OM ) originates from the residues, and crops produce is positively linked with S OM 33 ; crops residue retention helps S OM build up, soil temperature moderation, improved water holding capacity, microbial and enzymatic activities, and nutrients mobilization in the rhizospheric zone 34,35 . In cereals, A MF has extraordinary importance in boosting the yields 36 , and has capacity to acquire immobile nutrients beyond the radius of roots through their hyphal network 37,38 owing to greater nutrients/water taken up 39,40 , ultimately improve yields 41,42 .
ICM 5-8 increased 0.49 Mg ha −1 pooled wheat yield, but was 0.73 Mg ha −1 in maize, whereas, the yield advantage was more (0.96 Mg ha −1 ) with ZT bed planted maize (ICM 7-8 ) than to the ICM 1-4 (Table 1, Fig. 1). Excess (heavy rains) and deficit (longer dry spells) moisture are the common obstacles in the rainy season maize ecologies, but such variability does not exist during winters (wheat season). Residue retention in the ICM 5-8 infiltrate more water (Fig. 6d), and creates better aeration for the maize crop, bed planted maize (ICM 7&8 ) combining residues recorded yield advantage. Some meta-analysis studies have shown that the A MF helps to tolerate such stresses 43,44 . The L BFs fixes atmospheric-N and helps in solubilizing the insoluble P compounds which facilitate nutrient uptake, and improves the soil fertility, thereby, reduces the rate of chemical fertilizers up to 25%.
Water productivity (W P ) is the crop yield unit −1 of water consumption. Five years' results delineated that the ICM 5-8 could save ~ 7% irrigation water, compared to the ICM 1&2 (Fig. 2a). Long-time ZT tilled conditions where residues are retained, not only conserve the soil water, but facilitate better moisture regimes in the effective rhizosphere, and resulted in greater W P 32,45 . In the ICM 5-8 modules, the surface residues could reduce the losses of water vapour and retained moisture for the longer period, thus requiring lesser irrigations. Further, the bed planting coupled with the crops residues has twin benefits of greater infiltration and lower water application rates 4,46,47,80 . In 2017-18 and 2018-19, the higher W P was associated with the least water input coupled with greater yields than in other years (Fig. 2b).
Modules ICM 5-6 being lesser expensive, on account of lesser tillage operations involved and thus saved labor costs in various physical field operations, whereas, the ICM 7-8 were relatively costlier as these involved extra expenses in reshaping the beds (Fig. 3a). While, the ICM 1&4 incurred the highest cost owing to more trafficking in different tillage operations 48 . The sequential tillage included the extra fuel cost, eventually these modules gave lower yield, as indicated in the inclination of economic net returns 5 . Of course, the timely sowing of the succeeding wheat under the ZT conditions gave yield advantage 49,50 with the improved economic returns 48 . These results also reinforce the earlier research work in the adjoining ecologies 32,49,51 .
The ICM based agronomic management have vital role in the soil profile activities, and sustaining the soil health in the long-run 52 . Continuous crop residues recycling significantly improves the S OC fractions 53 and total S OC 45 . These CA-based practices have been widely analyzed for improving the S OC and the microbial population size 54 . Interestingly, over the years, the ZT + residues could increase the S OC , particularly by releasing the  40 . Our results showed that the S OC changed remarkably in the top soil layers, and ICM 5-8 increased the S OC storage by 12.1% in the top soil layer over the CT-based ICM 1-4 (Fig. 4a), as intensive tillage operations facilitate the loss of S OC, which is undesirable for the global C balance 45,55 . The S MBC is the living component (i.e., bacteria and fungi) of S OM , being the key indicator for S OC . In spite of small size, being a labile pool of S OM 56 , it contributes to the transformation or cycling of S OM 57,58 . In this study, the CA-based residue retained modules had 13.7% greater S MBC in the 0.0-0.15 soil layers than the modules where residues were removed (Fig. 4b), as regular residue addition accumulated the soil C that enhanced the S MBC and other microbial activities 46,59 . Moreover, the ZT conditions with sufficient crops residue are more conducive for the fungal hyphae growth, with additional supply of A MF along with L BFs further enhanced the fungal population and diversity, which could play an important role in the C / N cycling through their hyphal networks 60 . The S DH is the most intuitive bioindicators, describing the soil fertility 61 . It is associated with the S OM oxidation, and its activity depends on the microorganisms' abundance and activity 62 . Current results showed a 10.1% improvement in the S DH activity under the CA-based ICM 5-8 modules, over CT-based practices ( Table 3). The S MBC and S DH activities are directly associated with the recycling of the organic amendments, such as, the crops residues 46,63 .
Phosphatase activity is needed for P-mineralization and release of the PO 4 3− for the plant uptake. Often it is stated that the phosphatase activities (alkaline / acid) are greater in the P deficient soils 65 , and the current study soils are alkaline in nature (pH 7.9) with only 13 kg ha −1 available P. The P deficiency, residue addition and stoichiometric changes 66 would exhilarate the phosphatase activity under the CA-based modules. The urease activity responsible for the N mineralization and NH 3 release through hydrolysing the C-N bond of the amides 67 Figure 5. Effect of ICM modules on sustainable yield index (S YI ) of the maize, wheat and maize-wheat rotation. Means followed by a similar lowercase letter within a bar are not significantly different at p < 0.05 using Tukey's HSD test. www.nature.com/scientificreports/ and the soil biological properties, and hence the soil fertility. We conclude that the CA-based residue retained modules of M WR improved crops yields, farm economic profitability, and conserved the soil moisture. Such practices could also supplement the nutrients, sustain the crop yields, conserve natural resources, especially water and boost up the soil microbial functions for the long-term sustainability.

Conclusions
The five years' results clearly indicated the superiority of the CA-based residue retained ICM 5-8 modules, which produced 9.5-14.3% greater system maize grain equivalents (M GEY ) over the CT-based modules (ICM 1-4 ). Further, the ICM 2-8 saved 6.5-8.0% irrigation water, and ICM 5-8 recorded 10.3-17.8% higher system W P than the residue removed (ICM 1-4 ) modules. Of course, the conventional modules (ICM 1-4 ) were expensive, however,  [1][2][3][4] , thus the effect of residue retention was more pronounced in the upper soil layers than in lower depths. Further, there is a need to change regional growers' perceptions towards the adoption of maize, it could be a potential choice for accompanying wheat in this area. Therefore, the ZT residue retained modules either ICM 7&8 or ICM 5&6 could be acceptable for their adoption in the M WR for improving the yields, economic profitability and soil biological properties in the north western India and probably in other similar agro-ecologies.

Materials and methods
Experimental site, location and climate.  Table 4). The crop residues were completely removed in the CT-based modules (ICM 1-4 ), while in the ICM 5-8 modules, in-situ wheat (~ 3 Mg ha −1 on dry weight basis)) and maize (~ 5 Mg ha −1 , on dry weight basis) residues were retained on the soil surface during all the seasons of crops cultivation (Footnote Table 4, Fig. 6a,b).  www.nature.com/scientificreports/ In the ICM 1-4 modules, the field preparation was carried out by sequential tillage operations, such as, deep ploughing using the disc harrow, cultivator/rotavator twice (0.15-0.20 m), followed by levelling in each season. In the ICM 3-4 , the raised beds of 0.70 m bed width (bed top 0.40 m and furrow 0.30 m) were formed during each cropping cycle using the tractor mounted bed planter, and simultaneously wheat sowing was done (Fig. 6c). In the case of maize, ridges (0.67 m length) were prepared using the ridge maker. In the CA-based ICM 5-8 modules, the tillage operations, such as, seed and fertilizer placement were restricted to the crop row-zone in maize and wheat both. In the ICM 7&8 , the permanent raised beds (0.67 m mid-furrow to mid-furrow, 0.37 m wide flat tops, and 0.15 m furrow depth), were prepared (Fig. 6d). However, these beds were reshaped using the disc coulter at the end of each cropping cycle without disturbing the surface residues. The sowing was accomplished using the raised bed multi-crop planter.
Cultural operations and the fertilizer application. During every season, the maize (cv. PMH 1) was sown in the first week of July using 20 kg seed ha −1 . The wheat (cv. HD 2967) crop was sown in the first fortnight of November using the seed-cum fertilizer drill (ICM 1-2 ), bed planter (ICM [3][4] ) and zero-till seed drill (ICM 5-8 ) at 100 kg seed ha −1 . The chemical fertilizers (N, P and K) were applied as per the modules described in the footnote of Table 4. At sowing, the full doses of phosphorous (P) and potassium (K) were applied using the di-ammonium phosphate (DAP) and muriate of potash (MOP), and the nitrogen (N) supplied through DAP. The remaining N was top-dressed through urea in two equal splits after the first irrigation and tasseling / silking stages in maize, and crown root initiation and tillering stages of wheat. In the modules receiving ¾ fertilizers (ICM 2,4,6,8 ), the seeds were treated with the NPK liquid bio-fertilizer (L BFs ) (diluted 250 ml formulation 2.5 L of water ha −1 ), and an arbuscular mycorrhiza (A MF ) was broadcasted at 12 kg ha −1 as has been described by 75 . This L BFs had the microbial consortia of N-fixer (Azotobacter chroococcum), P (Pseudomonas) and K (Bacillus decolorationis) solubilizers, procured from the commercial biofertilizer production unit of the Microbiology Division, ICAR-Indian Agricultural Research Institute, New Delhi (Patentee: ICAR, Govt. of India). Weeds were managed by integrating the pre-and post-emergence herbicides, and their combinations along with the hand weeding-mulching, as mentioned in the concerned modules (Footnote Table 4). However, in the CA-based modules (ICM 5-8 ), the non-selective herbicide glyphosate (1 kg ha −1 ) was used 10 days before the sowing. The need-based integrated insect-pests and disease management practices were followed uniformly across the modules.
Soil sampling and analysis. Before start of the experiment, the soil sampling was done from 0.0-0.15 m depth. Afterwards, five random samples from each module from 0.0-0.30 m soil depth were collected at the flowering stage of 5th season wheat. These samples were taken from the three soil depths (0.0 to 0.05, 0.05-0.15 and 0.150-0.30 m) using the core sampler. The ground, air-dried soil samples, passed through a 0.2 mm sieve were used for the determination of the Walkley and Black organic carbon (S OC ), as described by 76 . For the soil biological properties, the soil samples were processed, and stored at 5ºC for 18-24 h, then analyzed the soil microbial biomass carbon (S MBC ), dehydrogenase (S DH ), alkaline phosphate (S AP ) and the urease (U RE ) activities.
The soil microbial biomass carbon (S MBC ). The S MBC was measured using the fumigation extraction method as proposed by 71 . The pre-weighed samples from the respective soil depths were fumigated with the ethanol-free chloroform for the 24 h. Separately, a non-fumigated set was also maintained. Further, 0.5 M K 2 SO 4 (soil: extractant 1:4) was added, and kept on a reciprocal shaker for 30 min. and then filtered through a Whatman No. 42 filter paper. OC of the filtrate was measured through the dichromate digestion, followed by the back titration with 0.05 N ferrous ammonium sulphate. The S MBC was then calculated using the equation: where, EC = (C org in fumigated soil -C org in non-fumigated soil), and expressed in µg C g −1 soil.
The dehydrogenase activity (S DH ). The S DH activity (μg TPF g −1 soil d −1 ) was assessed using the method of 73 . The soil sample (~ 6 g) was saturated with 1.0 ml freshly prepared 3% triphenyltetrazolium chloride (TTC), and then incubated for 24 h under the dark. Later on, the methanol was added to stop the enzyme activity, and the absorbance of the filtered aliquot was read at 485 nm.
The alkaline phosphatase activity (S AP ). The A PA activity was estimated in 1.0 g soil saturated with 4 ml of the modified universal buffer (MUB) along with 1 ml of p-nitrophenol phosphate followed by incubation at 37 °C for 1 h. After incubation, 1 ml of 0.5 M CaCl 2 and 4 mL of NaOH were added and the contents filtered through Whatman No. 1 filter paper. The amount of p-nitrophenol in the sample was determined at 400 nm 72 and the enzyme activity was expressed as µg p-NP g −1 soil h −1 .
The urease activity. Urease activity was measured using 10 g soil suspended in 2.5 ml of urea solution (0.5%). After incubating for a day at 37 °C, 50 ml of 1 M KCl solution was added. This was kept on a shaker for 30 min and the aliquot was filtered through Whatman No. 1 filter paper. To the filtrate (10 ml), 5 ml of sodium salicylate and 2 ml of 0.1% sodium dichloro-isocyanide solution were added and the green color developed was measured at 690 nm 74 . These values are reported as µg NH 4 -N g −1 soil h −1 .
Water application and productivity. In experimental modules, water was given through the controlled border irrigation method. The current meter was fixed in the main lined rectangular channel, and the water www.nature.com/scientificreports/ velocity was measured. To get the flow discharge, then multiplied with area of cross section of the channel. The following formulae were used to calculate the applied irrigation water quantity and depth 3 : where, F is flow rate (m 3 s −1 ), t is time (s) taken in each irrigation in each module and A is area (m 2 ). The effective precipitation (E P , difference between total rainfall and the actual evapotranspiration) was calculated, and then E P was added to the irrigation water applied to calculate the total water applied in each module. Across the maize and wheat modules (ICM 1-8 ), irrigations were given at the critical growth stages, such as, knee high and silking / tasseling (maize) and crown root formation, maximum tillering, flowering, heading / milking (wheat) stages, and after long dry spell (≥ 10-days).
On the basis of the soil water depletion pattern (at the depth of 0.60 m), in each season, 3-6 irrigations were given to maize, while wheat received 5-8 irrigations per season or crop including the pre-sowing irrigation. The rainfall data were obtained from the meteorological observatory located in the adjoining field. The water productivity (kg grains ha −1 mm −1 of water) was measured as per the equation given below: Additionally, the systems water productivity (S WP ) was also estimated by adding the water productivity (W P ) of both maize and wheat crops grown under the M WR .
Yield measurements. In each season, the maize and wheat crops were harvested during the months of October and April, respectively, leaving 0.75 m border rows from all the corners of each module. The crops were harvested from the net sampling area (6 m × 3 m, 18 m 2 ) located at the center of each plot. Maize crop was harvested manually and the wheat by using the plot combine harvester. All the harvested produce was sun dried before threshing and the grain and straw / stover yields were weighed separately. The stover/straw yields were measured by subtracting the grain weight from the total biomass. To compare the total (system) productivity of the different ICM modules, the system yield was computed, taking maize as the base crop, i.e., the maize equivalent yield (M GEY ) using the equation 20 : where, Y m = maize grain yield (Mg ha −1 ), Y w = wheat grain yield (Mg ha −1 ), P m = price of maize grain (US$ Mg −1 ) and P w = price of wheat grain (US$ Mg −1 ).

Farm economics.
Under different ICM modules, the variable production costs and economic returns were worked out based on the prevailing market prices for the respective years. The production costs included the cost of various inputs, such as, rental value of land, seeds, pesticides, L BFs / consortia, A MF , labor, and machinery; tillage / sowing operations, irrigation, mineral fertilizers, plant protection, harvesting, and threshing etc. The costs for the crops' residues were also considered. The system total returns were computed by adding the economic worth of the individual crop, however, the net returns were the differences between the total returns to the variable production costs of the respective module. The Govt. of India's minimum support prices (MSP) were considered for the conversion of grain yield to the economic returns (profits) during the respective years. Further, the system net returns (S NR ) were worked out by summing the net income from both maize and the wheat in Indian rupees (INR), and then converted to the US$, based on the exchange rates for different years.
Sustainable yield index (S YI ). 77,78 described the S YI as a quantitative measure of the sustainability of agricultural rotation/practice. The sustainability could be interpreted using the standard deviation (σ) values, where the lower values of the σ indicate the greater sustainability and vice-versa. Total crop productivity of maize and wheat under the different ICM modules was computed based on the five years' mean yield data. S YI was calculated using equation 78 .
where, -ȳ a is the average yield of the crops across the years under the specific management practice, σ n-1 is the standard deviation and Y -1 m is the maximum yield obtained under the set of an ICM module.
Statistical analysis. The GLM procedure of the SAS 9.4 (SAS Institute, 2003, Cary, NC) was used for the statistical analysis of all the data obtained from different ICM modules to analyze the variance (ANOVA) under the randomized block design 79 . Tukey's honest significant difference test was employed to compare the mean effect of the treatments at p = 0.05. Authors have confirmed that all the plant studies were carried out in accordance with relevant national, international or institutional guidelines. Water productivity = economic yield kg ha −1 / total water applied (mm)