High-value crops’ embedded groundnut-based production systems vis-à-vis system-mode integrated nutrient management: long-term impacts on system productivity, system profitability, and soil bio-fertility indicators in semi-arid climate

The current study identified two new climate-resilient groundnut-based cropping systems (GBCSs), viz., groundnut–fenugreek cropping system (GFCS) and groundnut–marigold cropping system (GMCS), with appropriate system-mode bio-compost embedded nutrient management schedules (SBINMSs) for semi-arid South Asia. This 5-year field study revealed that the GMCS along with leaf compost (LC) + 50% recommended dose of fertilizers (RDF50) in wet-season crop (groundnut) and 100% RDF (RDF100) in winter-season crop (marigold) exhibited the highest system productivity (5.13–5.99 t/ha), system profits (US$ 1,767–2,688/ha), and soil fertility (available NPK). Among SBINMSs, the application of 5 t/ha leaf and cow dung mixture compost (LCMC) with RDF50 showed the highest increase (0.41%) in soil organic carbon (SOC) followed by LC at 5 t/ha with RDF50 and RDF100. Legume–legume rotation (GFCS) had significantly higher soil microbial biomass carbon (SMBC) and soil microbial biomass nitrogen (SMBN) than legume–non-legume rotations (groundnut–wheat cropping system (GWCS) and GMCS). Among SBINMSs, the highest SMBC (201 µg/g dry soil) and SMBN (27.9 µg/g dry soil) were obtained when LCMC+RDF50 was applied to groundnut. The SMBC : SMBN ratio was the highest in the GWCS. LC+RDF50 exhibited the highest SMBC : SOC ratio (51.6). The largest increase in soil enzymatic activities was observed under LCMC+RDF50. Overall, the GMCS with LC+RDF50 in the wet season and RDF100 in the winter season proved highly productive and remunerative with better soil bio-fertility. SBINMSs saved chemical fertilizers by ~25%’ in addition to enhanced system productivity and system profits across GBCSs in semi-arid regions of South Asia. Future research needs to focus on studying the potential of diversified production systems on water and environmental footprints, carbon dynamics, and energy productivity under semi-arid ecologies.


Introduction
Due to the burgeoning population, the pressure on natural resources and soil health is on a constant rise, specifically in densely populated South Asia, which is dominated by semi-arid agroecology.Consequently, a plateau has been witnessed in the productivity levels of major crops and food systems of this economically, ecologically, and demographically fragile region (Nath et al., 2017;Rajpoot et al., 2021).Groundnut (Arachis hypogaea L.) assumes a major role in the agrarian and agroindustrial economy of South Asia, which is a major legume oilseed crop rich in both protein and oils (IOPEP, 2017;Heba et al., 2021).Hence, the declining productivity and profitability of groundnut-based cropping systems (GBCSs), especially the conventionally grown groundnut-wheat (Triticum aestivum L.) cropping system (GWCS) in addition to deteriorating soil health, have alarmingly threatened the agricultural sustainability of GBCSs vis-à-vis the oil supply-demand scenario in the region.Globally, groundnut is grown in an area of ~32.7 million ha (m ha) with a production of 53.9 million tons (mt) and productivity of 1,648 kg/ ha (FAOSTAT, 2023).In contrast, India, the second largest groundnut producer, grows it in ~5.97 m ha area with a production of 10.2 mt and 1,716 kg/ha yield (FAOSTAT, 2023).India exported about 514,164 metric tons (MT) of groundnuts to the world, valued at US$ 629 million during 2021-2022, mainly for edible oil production and its by-product cake as protein-rich animal feed (APEDA, 2023).However, the prolonged dominance of single crop-based groundnut systems, unbalanced use of fertilizers, and reduced use of organic manures have acutely aggravated the production vulnerabilities in the region.To sustain soil health for achieving food and nutritional security, on a long-term basis, diversification of existing production systems with legumes and high-value crops and judicious nutrient management are being advocated (Ambast et al., 2006;Varatharajan et al., 2022).Among the major groundnut-based production systems, the conventional GWCS is a predominant system, particularly under irrigated semiarid ecologies of South Asia (Heba et al., 2016;Heba et al., 2021).Nonetheless, several reports have indicated productivity stagnation of this system, chiefly owing to deterioration of soil organic carbon (SOC) levels, multi-nutrient deficiencies, multiple pest and disease infestations, and concerns about the timely and cost-effective availability of chemical fertilizers (Noman et al., 2016;Choudhary et al., 2020).We hypothesized that among the diversification options of the GWCS, the high-value crops' embedded groundnut-fenugreek (Trigonella foenum-graecum L.) cropping system (GFCS) and groundnut-marigold (Tagetes erecta L.) cropping system (GMCS) could be more remunerative, climateresilient, and sustainable, owing to their greater market demands, lower nutrient and water requirements, and contrasting cropgrowth nature in semi-arid agro-ecologies.
Diversification of the existing GWCS vis-à-vis an effective nutrient management strategy for the diversification alternatives is equally necessary to enhance system productivity and soil health.In the recent past, escalating synthetic fertilizers' prices have emerged as a serious concern.Further, their injudicious applications have caused harmful effects on soil and environmental health and have resulted in groundwater pollution (Sharma and Singhvi, 2017).Hence, there is an urgent need to incorporate numerous organic nutrient sources for proficient nutrient management (Sreedevi et al., 2013;Choudhary and Rahi, 2018).A judicious amalgamation of chemical fertilizers and organic nutrient sources has been well-known and strongly advocated for yield and soil fertility gains while minimizing production costs and environmental footprints (Sharma et al., 2019;Selim, 2020;Harish et al., 2022b).Additionally, the preparation of organics using diverse organic wastes not only promotes appropriate organic waste management but also supports trade-offs balancing between soil properties, crop quality, and animal health (Doran and Zeiss, 2000;Doran, 2002).Moreover, the inclusion of organics under nutrient management programs ensures a smooth supply of micronutrients as well as major nutrients (Bana et al., 2018;Rajpoot et al., 2021).The usage of organics also bestows a favorable micro-environment for crop growth and development, principally by improving soil properties (Bana et al., 2012;Bana et al., 2016;Bana et al., 2018;Bhupenchandra et al., 2022a).
Further, soil biological functions, the vital subset of soil health, are characterized by diverse indicators including soil microbial biomass carbon (SMBC), soil microbial biomass nitrogen (SMBN), and activities of soil microbial enzymes (dehydrogenase, protease, acid, and alkaline phosphatase) (Van Bruggen and Semenov, 2000;Singh et al., 2020;Singh et al., 2021;Singh et al., 2022a;Singh et al., 2022b).Owing to the potential of soil enzymes in plant nutrient availability via nutrient transformations, their activity in the soil is considered the most potent ecological indicator for a soil health assessment (Karlen et al., 1997;Alkorta et al., 2003;Singh et al., 2022b).Moreover, the soil biological functions are enormously dynamic and extremely sensitive to agronomic management practices like cropping system diversification (Yusuf et al., 2009), nutrient management (Liu et al., 2010;Bana et al., 2015;Ghosh et al., 2017), residue recycling, and crop establishment methods (Srinivasarao et al., 2013;Naragund et al., 2020;Harish et al., 2022a).Therefore, it is crucial to have deeper insights and understanding of the effect of diverse cropping systems and contrasting nutrient management practices on SMBC, SMBN, and soil microbial enzymatic activities for designing more appropriate and economically and ecologically sound management practices for the long-term sustainability of food systems (Bhupenchandra et al., 2022a).
Relative comparisons of diverse cropping systems concerning yield, economics, and soil fertility have been documented well across the globe and specifically in South Asia (Das et al., 2013;Heba et al., 2016;Heba et al., 2021;Rajpoot et al., 2021), but there is a lack of understanding about the diversification effects on soil biological functions (Bana et al., 2023).A knowledge gap exists on the inclusion of fenugreek and marigold in rotation with groundnut and its effect on system productivity, profitability, and soil health.Similarly, it has been established that plant nutrition protocols considerably influence soil fertility and its physical activity in addition to its microbial activity.Constant usage of chemical fertilizers, however, differently affects the soil systems as compared to the integrated application of organics and synthetic fertilizers (Tamilselvi et al., 2017).The composition and amount of applied organic manures also differentially affect the soil microbial diversity, abundance, and soil enzymatic activity (Chinnadurai et al., 2014;Kumar et al., 2017;Nath et al., 2017).Hence, the amalgamation of organic manures, prepared using locally available farm resources with chemical fertilization, was hypothesized to influence soil health (both chemical and microbial) and system productivity positively.Furthermore, medium-or long-term effects of chemical fertilizers and the integrated application of organic and inorganic nutrients need thorough investigation.Considering the above, a medium-term study was undertaken to design diversification options for the GWCS for improved yield and soil health and to assess the impact of different practices on system productivity, farm profits, and soil biological and chemical properties.

Experimental site
The fixed plot field experiment was performed for five consecutive years from 2010-2011 to 2014-2015 at ICAR-Indian Agricultural Research Institute (IARI), New Delhi, India (latitude 28°4′N; longitude 77°12′E; altitude 228.6 m).Delhi falls under a semi-arid climatic zone where ~70%-80% of the annual rainfall (652 mm) is received from July to September and the rest during winter (Bana et al., 2016).The mean monthly rainfall received during the period of experimentation is presented in Figure 1.The soil of the experimental site was sandy-loam in texture (Inceptisol), slightly alkaline in reaction, poor in SOC and available N, and medium in available P and K. Detailed physico-chemical properties of experimental soil are presented in Table 1.

Treatment detail and crop management
The experiment was laid out in a split-plot design with three replications.In the main plots, three diversified cropping systems were allocated.Four nutrient management practices applied to groundnut were kept in sub-plots, whereas in sub-sub plots, two nutrient management treatments were applied to winter-season crops, hence with a total of eight system-mode bio-compost embedded nutrient management schedules (SBINMSs) under each cropping system (Table 2).All the treatments were kept on the same plots for 5 years under conventionally tilled conditions.Before the sowing of crops, a deep plowing using a soil turning Mean monthly rainfall received during the period of experimentation.Bana et al. 10.3389/fpls.2023.1298946Frontiers in Plant Science frontiersin.orgplow, followed by three passes of a 9-tyne cultivator, was performed to prepare the field for sowing.Groundnut variety 'GG-10' was sown in the first week of July during all 5 years of study, at 40-cm row spacing using 100 kg/ha seed rate.Gap filling and thinning operations were performed within 20 days of sowing.The recommended dose of fertilizers (RDF) for groundnut was 25 kg N + 60 kg P 2 O 5 + 40 kg K 2 O per ha.The chemical fertilizers were placed at the time of final tillage before the sowing of groundnut.Under SBINMSs, the organic manures were applied 3 weeks before groundnut sowing.The application rate of leaf compost (LC) and leaf and cow dung mixture compost (LCMC) (containing mean N, P 2 O 5 , and K 2 O content of 0.63%, 0.31%, and 0.67% and 0.57%, 0.33%, and 0.51%, respectively) was performed using 5 t/ha composts on an oven-dry weight basis.To control weeds in groundnut, pre-emergence application of pendimethalin was performed using 0.75 kg a.i.ha −1 in 500 L/ha spray solution.
Winter crops were grown as per the standard package of practices, recommended by ICAR-IARI, New Delhi.For wheat, the 'HD-2967' variety was sown using a seed drill in the second week of November at a row spacing of 25 cm.The RDF of wheat, 120 kg N + 60 kg P 2 O 5 + 60 kg K 2 O per ha, was applied as per the treatments (Table 2).Similarly, the RDF for fenugreek and marigold was 20 kg N + 60 kg P 2 O 5 + 60 kg K 2 O/ha and 120 kg N + 60 kg P 2 O 5 + 60 kg K 2 O/ha, respectively.For fenugreek, the 'RMt 305' variety was sown using a 25 kg/ha seed rate with 25-cm row spacing.
Likewise, 4-week-old seedlings of the 'Pusa Narangi Gainda' variety of marigold were transplanted in the second week of November.

Soil sampling and analysis
From the fixed plots, soil samples were taken using a core auger from 0-15-cm depth instantaneously after harvest and were transferred to the laboratory for microbial analysis.Soil acid and alkaline phosphatase enzymatic activities were determined using 16 mM of para (p)-nitrophenyl phosphate as substrate and reported as mmol p-nitrophenol•g −1 •h −1 (Tabatabai and Bremner, 1969).Dehydrogenase activity was determined using the rate of reduction of triphenyl tetrazolium chloride to triphenyl formazan and expressed as mg TPF•g −1 •24 h −1 .The chloroform-fumigation extraction method was employed for the analysis of SMBC and SMBN and expressed as µg/g of dry soil (Vance et al., 1987).The value of the efficiency of extraction of microbial biomass carbon (kEC) was 0.45 (Jenkinson and Powlson, 1976).The value of efficiency of extraction of microbial biomass nitrogen (kEN) was 0.68 (Brookes et al., 1985).The optical density at 485 nm was compared to that of the triphenyl formazan standard (Casida et al., 1964).Soil with 1% casein as the substrate was incubated for 2 h in 0.05 M tris-hydroxymethyl-aminomethane-hydrochloric acid buffer at pH 8 for protease enzyme determination.The released amino acid was analyzed through the Folin-Ciocalteu colorimetric method described by Ladd and Butler (1972) and expressed as µmol tyrosine

Soil nutrient analysis
Soil samples from 0-15-mm depth were collected using the core sampler to assess the treatment effects on the fertility status of the soil after harvest as suggested by Rana et al. (2014).Soil samples were analyzed to study available N, P, and K through modified Kjeldahl's method (Jackson, 1958), Olsen's method (Olsen et al., 1954), and the flame photometer method (Jackson, 1958), respectively.Before the initiation of experiments, the extractable Zn, Mn, Fe, and Cu were determined in soil samples using DTPA (Lindsay and Norvell, 1978).

System productivity and profitability
The yield of the crops was recorded from the net plot area and expressed in t/ha.For comparing the different production systems, the yield of winter crops was converted as groundnut equivalent yield (GEY).The GEY of winter crops was calculated using Equation 1: where Ye is GEY, Yw is the economic yield of winter crop, Pg is the market price of groundnut, and Pw is the market price of winter crop.To compute system productivity of each production system, the following equation was used (Equation 2): where Ps is the system productivity, Y is the yield of groundnut, and Ye is the GEY of winter crops.

Statistical analysis
The data on various parameters were statistically analyzed as per the procedure of analysis of variance (ANOVA) to determine treatment effects through Tukey's honestly significant difference test as a post hoc mean separation test (p< 0.05) using SAS 9.1 software (SAS Institute, Cary, NC, USA).Tukey's procedure was used where ANOVA was found significant.Correlation analyses and treatment means were compared at a 5% level of significance.

Soil biological and chemical health
Deteriorating soil health is an emerging issue in agriculture.The quality and quantity of food produced through agriculture largely depend on the status of soil health.One of the main objectives of this study was to analyze the effect of different GBCSs and SBINMSs on soil health.Soil health was assessed using the following soil health parameters: SOC, nitrogen (N), phosphorus (P), potassium (K), SMBC, SMBN, SMBC : SMBN, SMBC : SOC, acid phosphatase, alkaline phosphatase, dehydrogenase, and protease.Based on the measured levels of soil health parameters, treatments were clustered into four groups in a hierarchical manner, which is displayed in a dendrogram (Figure 2).The group that showed superior soil health consisted of four treatments of integrated nutrient management in the GFCS, followed by the integrated nutrient management (INM) treatments of the GWCS.Poor soil health was observed for all controls.Results for each soil health parameter are discussed below.

Soil organic carbon
Levels of SOC significantly varied among the cropping systems and nutrient management practices in wet-season crops, but they were non-significantly affected by fertilizer application in winterseason crops (Figure 3).The initial SOC content of the soil was 0.36% (Table 1).The GFCS exhibited a 0.02% increase in SOC content, whereas a decline of 0.03% in SOC was observed in the GMCS.The initial amount of SOC content was maintained as it is in the GWCS.In the long run, legumes have been observed to increase SOC (Nath et al., 2019;Choudhary et al., 2020).Among the three experimental crops, leaf-litter fall was the highest in fenugreek followed by wheat and then marigold.In addition to that, the marigold crop was uprooted after the season.Thus, nutrient and biomass recycling was less in the marigold.However, more crop residues of wheat and fenugreek remained in the field, leading to higher nutrient and biomass recycling.This might have resulted in higher levels of SOC (Saha and Ghosh, 2013).The highest decline of 0.06% in SOC was observed in control plots over initial levels of SOC.The application of 5 t/ha LCMC with 50% RDF showed the highest increase of 0.41% in SOC content, followed by LC at 5 t/ha with 50% RDF (0.37%) and 100% RDF (0.36%).The 100% RDF treatment had significantly higher SOC content (0.35%) than control (0.30%).These results indicate that an integrated approach to system-basis nutrient management through SBINMSs can efficiently increase the levels of SOC (Bana et al., 2016;Srinivasarao et al., 2021).

Soil fertility
After 5 years of experimentation, significant improvement in soil fertility concerning available NPK levels under various treatments was observed (Figure 4).The data show that the GFCS and GWCS rotations had statistically similar effects on soil available N and P levels.Compared to these rotations, the soil fertility was slightly lower in the GMCS.This could be correlated with lower biomass recycling in the marigold as mentioned previously.Available K was found statistically similar throughout all three experimental cropping systems.Higher improvement in soil fertility was observed after the application of LC and LCMC in groundnut compared to control and chemical fertilizers.The available N content was 191 kg/ha in control, which was improved by 32 kg/ha, 54 kg/ha, and 46 kg/ha, respectively, due to the application of 100 RDF, LC+RDF 50 , and LCMC+RDF 50 .The P content was found to be significantly higher when either leaf compost (14.9 kg/ha) or mixed compost (14.5 kg/ha) was applied.With the application of LC+RDF 50 , the available P content was enhanced by 1.18 kg/ ha and 5.57 kg/ha compared with 100 RDF and control, respectively.The soil K content ranged from 214 kg/ha to 261 kg/ha.The data revealed that available K was the least in control, while it increased with the addition of both inorganic and organic fertilizers.The highest K was recorded in LC +RDF 50 whose effect was statistically similar to that of the mixed compost.Fertilizer application in winter crops also affected the soil fertility status considerably.Significantly higher soil NPK content was observed in the F100 compared with F50.Leaf compost and mixed compost contain lower nutrients but have sufficient soil organic matter (SOM), hence improving soil health (Choudhary et al., 2020).Further, the SOM increases the water-holding capacity of the soils and also makes more nutrients available to the plants (Richardville et al., 2022).It can be inferred from the results that as LC acts as more of a soil conditioner, soil retains externally applied inorganic nutrients more efficiently and improves nutrient availability to plants (Bhupenchandra et al., 2022b;Gupta et al., 2022).

Soil microbial properties
The legume-legume crop rotation (GFCS) had significantly higher levels of SMBC and SMBN compared to legume-nonlegume rotations (GWCS and GMCS) (Figure 5).The practice of crop rotation has been observed to alter the status of soil nutrients through microbial immobilization (Balota et al., 2003;Bhupenchandra et al., 2022a;Bhupenchandra et al., 2022c).In this study, the highest SOC was noted for the GFCS.Borase et al. (2020) also demonstrated a positive Effect of groundnut-based cropping systems (GBCSs) and system-mode bio-compost embedded nutrient management schedules (SBINMSs) on soil fertility status in terms of available NPK in the soil.Effect of high-value crops' embedded groundnut-based cropping systems (GBCSs) and system-mode bio-compost embedded nutrient management schedules (SBINMSs) on soil organic carbon (SOC).
correlation between SOC, SMBN, and SMBC.Among the cropping systems, the highest values of SMBC (205 µg/g of dry soil) and SMBN (25.6 µg/g of dry soil) were observed in the GFCS followed by the GWCS.The lowest values of SMBC (159 µg/g of dry soil) and SMBN (20.4 µg/g of dry soil) were recorded in the GMCS.Nutrient management strategies in groundnut had a considerable impact on SMBC and SMBN.Compared to the control and 100% RDF, a noticeable increase in both SMBC and SMBN was observed in the INM approach.The highest SMBC (201 µg/g of dry soil) and SMBN (27.9 µg/g of dry soil) were obtained when 5 t/ha LCMC+RDF 50 was applied to groundnut.Higher microbial count in manure compost could likely be the reason for these higher values (Bana et al., 2012;Bhupenchandra et al., 2022a).Compared to F50, the F100 improved both SMBC and SMBN in the study.The legumecereal rotation had a higher SMBC : SMBN ratio than the legume-legume and legume-marigold rotations (Figure 6A).
The ratio of SMBC : SMBN was the highest in the GWCS (8.3), followed by the GFCS, and the least in the GMCS (8.0).The SMBC : SMBN ratio was maximum for control (9.8).In contrast, the decline in the SMBC : SMBN ratio was observed after the application of chemical fertilizers and organic manures.The plots receiving LCMC+RDF 50 recorded the least SMBC : SMBN (7.2).The addition of N through fertilizers was reported to reduce the microbial C:N ratio of the soil (N addition effect on microbes).The application of F100 to winter crops lowered the SMBC : SMBN ratio compared with F50.The ratio of SMBC : SOC was the highest for the GFCS (55.3) and lowest for the GMCS (46.5) (Figure 6B).Among the nutrient treatments, LC +RDF 50 (51.6)recorded the highest SMBC : SOC ratio, while RDF had the least SMBC : SOC ratio (49).
One of the prime indicators of healthy soil is biologically alive soil containing a healthy microbial community (Singh et al., 2021;Singh et al., 2022a;Singh et al., 2022b).Microbial

B A FIGURE 5
Effect of groundnut-based cropping systems (GBCSs) and system-mode bio-compost embedded nutrient management schedules (SBINMSs) on (A) soil microbial biomass carbon (SMBC) and (B) soil microbial biomass nitrogen (SMBN).

B A FIGURE 6
Effect of groundnut-based cropping systems (GBCSs) and system-mode bio-compost embedded nutrient management schedules (SBINMSs) on (A) soil microbial biomass carbon to soil microbial biomass nitrogen (SMBC : SMBN) ratio and (B) soil microbial biomass carbon to soil organic carbon (SMBC : SOC) ratio.
activity in soil is mostly correlated with available SOM (Zhong and Cai, 2007).Microbes primarily feed on SOM and make several nutrients available to plants, which are produced as their metabolic by-products, among which carbon and nitrogen are especially of importance for plant growth (Perelo and Munch, 2005;Six et al., 2006;Bhupenchandra et al., 2022a;Bhupenchandra et al., 2022c).Microbes, while feeding on SOM, also release many biostimulants and hormones, which again help in enhancing plant growth and productivity (Bhupenchandra et al., 2022c;Rajanna et al., 2023).The interactive effect of soil microbes on other soil components improves the soil structure.The type of plant root system directly influences the growth of these micro-organisms.Leguminous plants are well known to fix atmospheric N through their root nodules.They also have been observed to increase the levels of SMBN, SMBC, and SOC in the soil in several studies (Zahran, 1999;Mayer et al., 2004;Liu et al., 2011;Wang et al., 2018;Bhupenchandra et al., 2022a).Wheat roots are fibrous and usually go 100-200 cm deep down into the soil.In contrast, fenugreek and marigold root systems are shallow and remain confined largely into or adjacent to the plow layer only (0-15 cm deep).Therefore, wheat adds more SOM and improves the soil microbial activity.Wheat root exudates such as glues, gums, and waxes help in improving the soil structure by providing adherence to soil molecules, thus increasing the stability of soil aggregates (Bana et al., 2018), which again helps in better aeration and water infiltration (Choudhary and Rahi, 2018).The cumulative effect of all these factors creates a better environment for flourishing microbial growth (Mauchline et al., 2015).Manure compost is rich in nutrients as well as in microbial count; hence, it acts as microbial inoculum for the soil (Ren et al., 2019).The mineral-based inorganic fertilizers do not help in significantly increasing the microbial activity of the soil, compared to organic fertilizers (Lazcano et al., 2013;Kumar et al., 2021;Kumar et al., 2022).

Soil enzymatic activities
Soil enzymatic activities, as shown in Figure 7, follow a similar trend of SMBC and SMBN.Plots with the GFCS had higher soil enzymatic activities than the GMCS and GWCS.Previous studies have also reported similar results where the cultivation of legumes increased the enzymatic activity of soil (Singh et al., 2008;Singh et al., 2021;Singh et al., 2022a).Soil enzymatic activity is positively correlated with microbial activity, as soil enzymes catalyze the biochemical reactions that are critical for microbial functions (DeLuca et al., 2019).In this study, organic manures along with chemical fertilizers remarkably improved the alkaline and acid phosphatase, dehydrogenase, and protease activities.The largest increase in soil enzymatic activities was observed in LCMC+RDF 50 treatment, where leaf and cow dung-based compost was used as a nutrient source along with 50% RDF.The plots that received no fertilizers and manures (control) showed the lowest enzymatic activities.The F100 application to winter crops also increased soil enzymatic activities compared with F50.Hence, it is inferred that the integrated application of chemical fertilizers and organic manures, i.e., SBINMSs, is highly effective in enhancing the soil enzymatic activities (Bhupenchandra et al., 2022a;Bhupenchandra et al., 2022c).

System productivity
Another major objective of this experiment was to study the effect of high-value crops' embedded diversified cropping systems and different SBINMSs on system productivity.The overall effect of different treatments on system productivity is illustrated in Figure 8. Nine experimental treatments showed an overall decline in system productivity.It was lowest in 2012-2013.Among the three cropping systems, the GMCS recorded the highest system productivity (5.13-5.99t/ha) in terms of Effect of groundnut-based cropping systems (GBCSs) and system-mode bio-compost embedded nutrient management schedules (SBINMSs) on soil enzymes.Y-axis measuring units: alkaline phosphatase (µg p-nitrophenol•g −1 soil•h −1 ), acid phosphatase (µg p-nitrophenol•g −1 soil•h −1 ), dehydrogenase (µg TPF•g −1 soil•h −1 ), and protease (µg trypsin•g −1 soil•h −1 ).
GEY, which was 1.73 and 1.85 t/ha higher than that of the GWCS and GFCS, respectively (Figure 9).The market prices of the yield obtained from wheat, fenugreek, and marigold were taken into consideration to calculate the GEY.The GMCS also had higher system net returns over the years (Figure 10), with an average (5-year av.) system net returns of US$ 1,767-2,688/ha and system benefit:cost (B:C) ratio of 2.51-3.23 across different SBINMS practices despite of higher system cost of cultivation (US$ 1,135-1,184/ha) (Figures 11,12).Marigold received considerably higher prices in the market compared to wheat and fenugreek.Therefore, the system productivity of the GMCS was much higher (5.13-5.99t/ha) compared to that of the GWCS (3.68-4.10t/ha) and GFCS (3.48-3.88t/ha) (Figure 9).On average, the highest system productivity, system net returns, and system B:C ratio were observed under C3F3S1 (GMCS with the application of LC at 5 t/ha + 50% RDF in groundnut and 100% RDF in marigold crop) followed by C3F4S1 (GMCS with the application of LCMC at 5 t/ha + 50% RDF in groundnut and 100% RDF in marigold crop) (Figures 10-12; Supplementary Figure 1).The SBINMSs, particularly with LC, were found Effect of groundnut-based cropping systems (GBCSs) and system-mode bio-compost embedded nutrient management schedules (SBINMSs) on system productivity (GEY).Effect of groundnut-based cropping systems (GBCSs) and system-mode bio-compost embedded nutrient management schedules (SBINMSs) on system productivity (t/ha).
superior compared with sole fertilizer use due to better nutrient supply and soil biological properties (Bhupenchandra et al., 2022c;Varatharajan et al., 2022).The plots treated with LC+RDF 50 increased system productivity by ~80% and 12% compared with control and 100 RDF plots, respectively (Figure 10).Likewise, F100 also showed ~17% higher system productivity than F50.Among nutrient management practices, LC+RDF 50 had the highest B:C ratio compared with LCMC+ RDF 50 (Figure 12).The highest B:C ratio was observed for treatments C3F3S1 and C3F4S2, whereas the lowest was observed for C1F1S2 and C1F1S1 (Figure 12).This can be attributed to the lower prices of LC compared to manure compost.

Which-won-where
To analyze the system productivity, different types of GGE biplots were generated, the results of which are mentioned here (Gabriel, 1971;Yan and Tinker, 2006).The "which-won-where/ what" polygon was employed to identify the best treatment for each environment (year) (Figure 13A).Almost all of the variations among Effect of groundnut-based cropping systems (GBCSs) and system-mode bio-compost embedded nutrient management schedules (SBINMSs) on cost of cultivation (US$/ha) and gross and net returns (US$/ha) of various GBCSs (5-year av.).

FIGURE 10
Effect of groundnut-based cropping systems (GBCSs) and system-mode bio-compost embedded nutrient management schedules (SBINMSs) on system net returns (US$/ha).
different treatments related to system productivity were explained by only one factor, system yield (98.3%), and gross returns explained a further 1.5%.In terms of system yield and net returns, the GMCS significantly outperformed the GWCS and GFCS throughout the 5 years.Treatment LC+50 for the wet season and treatment F100 for the winter season were observed to be the most superior treatments for all the years.The polygon clustered the treatments in five different clusters, which are given in the order of performance ("a" performed the poorest and "e" performed the best): a) treatments 1, 2, 9, 10, 18, and 17 (all controls); b) treatments 4, 6, 8, 12, 14, and 16 (GFCS system with F50 in winter); c) treatments 3, 5, 7, 11, 13, and 15 (GWCS with F100 in winter); d) treatments 20, 22, and 24 (GMCS with F50 in winter); and e) treatments 19, 21, and 23 (GMCS with F100 in winter).These results suggested that the application of 100F in the winter season plays an important role in increasing the system-productivity yield and system net returns.Overall, it can be suggested that by adopting the SBINMS approach embedded with leaf compost, the annual use of inorganic fertilizers can be reduced by ~25% in the system mode, and on top of that, system productivity can also be enhanced appropriately.Hence, the selection of an appropriate cropping system adds to the benefit of higher system productivity.

Mean vs. stability
The mean vs. stability biplot was used to identify the performance stability of the treatments for the tested environments.Treatments were ranked along the average-tester axis (ATA).The circle indicates the average treatment.Based on the mean performance of system productivity and system net returns, the treatments above average were ranked as  Effect of groundnut-based cropping systems (GBCSs) and system-mode bio-compost embedded nutrient management schedules (SBINMSs) on system B:C ratio (5-year av.).13B).All other treatments gave below-average system productivity and system net returns.The highest yield stability and profitability were recorded for treatments 21, 23, and 19 in the years 201321, 23, and 19 in the years -201421, 23, and 19 in the years and 201421, 23, and 19 in the years -201521, 23, and 19 in the years . For the years 201021, 23, and 19 in the years -201121, 23, and 19 in the years , 201121, 23, and 19 in the years -201221, 23, and 19 in the years , and 201221, 23, and 19 in the years -2013, the highest yield stability and profitability were recorded for treatments 5, 13, and 7.The treatment stability is reciprocally related to the length of the projection (regardless of direction) away from ATA (Bana et al., 2022).Treatments 21, 23, and 19 were more stable compared to 22, 24, and 20.

Conclusions
The current study aimed at pinpointing the most appropriate high-value crops' embedded GBCS and system-mode INM schedules for achieving higher system productivity while preserving soil bio-fertility in semi-arid South Asia.It was suggested that the SBINMS approach of leaf compost (5 t/ha) along with 50% RDF in the GMCS had a significantly positive correlation with system productivity and soil biological and chemical health.The use of leaf compost in system mode (SBINMSs) proved a more cost-effective and efficient alternative than manure compost, which may reduce the annual use of inorganic fertilizers by ~25% while using LC+50% RDF in groundnut and 100% RDF in winter-season crops.The LC+50% RDF also enhanced the system productivity and soil biological and chemical properties.Overall, the high-value crop (marigold) embedded GMCS along with LC+50% RDF in wet-season crops and 100% RDF in winter-season crops proved more productive with system-productivity of 5.13-5.99t/ha and remunerative with system net returns of US$ 1,767-2,688/ha in addition to improved soil bio-fertility indicators in the semi-arid climate of South Asia.Understanding the effect of diversified production systems and innovative nutrient management schedules on environmental footprints (greenhouse gas emissions), water productivity, energy use efficiency, and carbon footprints would be an important future research area based on the work of the present study.Furthermore, calibration and validation of cropping system models from the present study to identify ideal future management of GBCS under changing climate and resourceavailability scenarios, and diverse representative concentration pathways, would be other future research areas that the present work could contribute to.

FIGURE 2
FIGURE 2 Dendrogram displays hierarchical clustering of experimental treatments based on the levels of soil bio-fertility indicators.Red color indicates higher values of health parameters, which gradually fades into blue color, indicating lower values.

FIGURE 11
FIGURE 11 FIGURE 13 GGE biplots.(A) Which-won-where polygon displays superior treatments, and (B) mean vs. stability biplot displays the performance of the treatment and its stability.X axis: principal component 1-system yield.Y axis: principal component 2-gross returns.

TABLE 1
Physical and chemical properties of the top 15-cm soil of the experimental field.