This study was conducted from 2021 to 2022 at the College Farm of Navsari Agricultural University, Navsari (Gujarat), India. The site is located at 20.9248°N latitude and 72.9079°E longitude at an altitude of 11.98 m (Figure 1). Physiographically, the area lies in the coastal plains of South Gujarat, characterized by nearly level to gently sloping terrain, with moderate drainage and medium to deep alluvial soils. The region is part of the sub-humid agro-ecological zone, influenced by the Arabian Sea, contributing to high humidity and seasonal monsoonal rainfall (Keniya et al., 2024). The weekly rainfall during the kharif season ranged from 0.0 to 248 mm and 0.0 mm to 517 mm in 2021 and 2022, respectively. In both years, the rainfall began in June and concluded in September. The mean annual rainfall and temperature during the research period was 248 mm and 24°C during 2021, 345 mm and 22°C during 2022, as depicted in Figures 2, 3, respectively. The soil taxonomy of the experimental site was classified under soil order “Inceptisols,” and soil series “Jalapor.” He initial soil physico-chemical properties were mentioned the Table 1.

The experiment was carried out in two seasons (viz., summer and kharif seasons) for 2 years, 2021 and 2022, in split-plot design with three replications. The treatment on the main plot was summer legumes (T) sown in summer season with four legumes viz^., T₁: Green gram (Vigna radiata), T₂: Cowpea (Vigna unguiculata), T₃: Dhaincha, and T₄: Fallow. In sub plots, there were six nutrient management practices (W) viz., W₁: 100% RDF (100 kgN + 30 kg P₂O₅ + 00 K₂O kg/ha), W₂: 75% RDF (75 kg N + 22.5 kg P₂O₅ + 00 K₂O kg/ha), W₃: 50% RDF (50 kg N + 15 kg P₂O₅ + 00 K₂O kg/ha, W₄: 75% RDF + 25% N from FYM, W₅: 50% RDF + 50% N from FYM, and W₆: No-fertilizer application. The main plot treatments, i.e., green gram and cowpea) were incorporated into the soil after the crop harvest, whereas dhaincha was incorporated into the soil at 50 % blooming stage aimed at optimizing nutrient cycling, decomposition rate, and nitrogen use efficiency (NUE) for the succeeding rice crop and their respective biomass, straw yield and nutrient content were mentioned in Tables 2, 3 (amount of biomass we have incorporated and the respective nutrient we have supplied through the summer legumes). A cropping sequence involving the cultivation of summer legumes green gram (Vigna radiata), cowpea (Vigna unguiculata), and dhaincha (Sesbania aculeata) was grown during the summer season, followed by transplanted rice (Oryza sativa) in the kharif season, with the land left fallow during the rabi season. This sequence was repeated on the same field in the subsequent year. After incorporation of the residues each main plot (summer legumes) was divided into six sub-plots and in each subplot rice crop of variety GNR-3 was grown with different nutrient management practices. Farmyard manure (FYM) was applied 15 days prior to rice transplanting. Inorganic fertilizers (NPK) were administered based on treatment protocols: phosphorus was applied entirely as a basal dose, nitrogen was split across basal, tillering, and at panicle initiation stages, while potassium was applied at basal and panicle initiation. Biomass from summer legumes (green gram, cowpea, and dhaincha) was incorporated into the soil post-harvest, allowing a one-month decomposition period before rice cultivation. Treatments with 100, 75, and 50% of the recommended NPK levels were applied to evaluate whether the added residue could help make up for the reduced fertilizer in the lower NPK treatments. This approach was used to assess the potential of the residue to partially replace chemical fertilizers. Straw incorporation was employed as a strategy to enhance soil fertility by supplementing inorganic nutrient inputs. After harvesting green gram and cowpea, residual biomass was incorporated into the soil through ploughing. In the case of dhaincha, biomass was incorporated at 50% flowering, followed by irrigation to facilitate decomposition. The chemical and other composition details in the experiment is given in Tables 2–5.

Soil samples were collected at the initiation of the experiment, after the soil fertility gradient stabilizing experiment, before and after the test crop experiment, and the verification trial experiment and analyzed for pH and EC (Jackson, 1973), organic carbon (Walkley and Black, 1934), available nitrogen (Subbiah and Asija, 1956), Olsens extractable phosphorus (Bray and Kurtz, 1945), and neutral normal ammonium acetate potassium (Knudsen et al., 1982), respectively. Nutrient content in grain and straw were obtained by total nitrogen by the modified Micro Kjeldhal method (Bremner, 1996), phosphorus by the vanadomolybdophosphoric yellow method (Morre, 1991), and potassium by the wet digestion method (Chapman and Brown, 1950), respectively. Nutrient absorption by grain and straw was calculated by multiplying grain yield (kg ha⁻¹) by nutrient concentration in the grain (%) and straw yield (kg ha⁻¹) by nutrient concentration in the straw (%), respectively. The total nutrient uptake by the crop is the sum of nutrient absorption in the grain and the straw.

The nutrient (NPK) uptake (kg/ha) of pods, grain and stover/straw of summer legumes (green gram, cowpea, dhaincha) and rice was worked out by using Equation 1, according to (Sunil Kumar et al., 2024).

The agronomic, physiological and apparent recovery efficiencies were estimated by using Equations 2–4 as stated by Congreves et al. (2021) and Sarkar et al. (2021).

where, Y = Grain yield (kg/ha) with applied nutrient; Y0 = Grain yield (kg/ha) with no applied nutrient, U = Total nutrient uptake (kg/ha) with applied nutrient; U0 = total nutrient uptake (kg/ha) with no applied nutrient.

On the other hand, the Partial factor productivity, nutrient efficiency ratio, and nutrient addition through straw, were calculated according to Equations 5–7.

The residual effect of summer legumes was approximately calculated based on yield and total uptake data obtained in the field experiments. Apparent recovery ratio was used as a tool to calculate nitrogen and phosphorus accumulation (residual effect). The proportion of nitrogen/phosphorus absorbed and derived from the applied nitrogen/phosphorus cannot be verified; thus, it is referred to as “apparent” nutrient recovery. The residual effect and absolute residual effect were calculated using Equations 8, 9.

where, A = Yield from residue incorporated plot with fertilizer, B = Yield from residue free plot with fertilizer, and C = Yield in residue free plot without fertilizer.

The statistical procedures outlined by Panse and Sukhatme (1967) were employed to analyze the data on a variety of variables. The ‘F’ test was implemented to compare the treatment effects on all the characters under investigation. The mean values of the summer legumes were presented, and the subsequent kharif rice was analyzed using a Split Plot Design. In the event that the ‘F’ test revealed significant differences among the interventions, the Critical Difference (CD) at 5% was calculated. Otherwise, the standard error of the mean was computed. Pooled analysis of the summer legumes and succeeding kharif rice analysed for two years was worked out as per the method described by Cochran and Cox (1957). Bertlett’s test was applied to examine the homogeneity of variance due to error.

The fundamental method of variance analysis may not be suitable for two distinct seasonal conditions, as the error variances between seasons and the treatment × season interaction could be substantial. Consequently, the method outlined by Cochran and Cox (1957) was employed to conduct an aggregated analysis of the summer legumes and subsequent kharif rice over a two-year period. The homogeneity of variance attributable to error was evaluated using Bartlett’s test. The presence or absence of a season × treatment interaction was determined by comparing the variance resulting from the season × treatment components to the pooled estimate of error variance.

From the pooled analysis, significantly higher total nitrogen uptake was noticed in rice grown in dhaincha incorporated plots (T₃, 80.88 kg ha⁻¹) with 37.43% increase over fallow. Whereas, the application of 100% RDF (W₁) exhibited significantly higher nitrogen uptake (97.09 kg ha⁻¹) with 121.73 percent increase over W6. The interaction effect was found to be significantly with highest total nitrogen uptake recorded in dhaincha incorporated plots along with application (109.29 kg/ha) with over a 113.39 percent increase over absolute control, which it remained consistent for 2 years (Table 7).

Rice sown after incorporation of dhaincha recorded significantly higher total phosphorus uptake (16.01 kg ha⁻¹) with 44.08 increases over fallow treatment. Application of 100% RDF has shown significantly higher total phosphorus uptake (21.12 kg/ha) with 233.65 percent over W6. Interaction effect i.e., dhaincha incorporation along with 100%RDF as 24.03 kg ha⁻¹, and the interaction effect with year was found to be non-significant (Table 8).

Pooled analysis results revealed that incorporation of dhaincha recorded significantly higher potassium uptake (100.04 kg ha⁻¹) with 39.06 percent increase over fallow. Among the nutrient levels applied to rice crops, significantly the highest total potassium uptake by rice (111.61 kg/ha) was found with 100% RDF with 95.72 percent increase over control. Interaction effect of dhaincha incorporation along with 100%RDF was found significant (129.38 kg ha⁻¹) with 166.01 percent increase over absolute control, but interaction with year was consistent over 2 years (Table 9).

Pooled analysis revealed that significantly highest agronomic use efficiency (kg grain/kg N applied through fertilizer + FYM) of rice crop was noted with dhaincha incorporation. In sub plots, the agronomic use efficiency of rice crops was significantly highest in 100%RDF. The interaction effect was found to be non-significant with highest obtained in treatment dhaincha + 100% RDF, which was found to be efficient in using nitrogen with agronomic use efficiency of 25.20 kg grain/kg N applied as depicted in Figure 4 and it remained consistent throughout 2 years. Nutrient applied through fertilizer and FYM was mentioned in Table 10.

The nitrogen use efficiency of rice in the study was expressed in terms of grain yield, i.e., (kg grain/kg N applied through fertilizer + FYM + residue) presented in Table 11. Among the different summer legumes, the highest nitrogen use efficiency (11.68 kg grain⁻¹ kg⁻¹ N applied through fertilizer + FYM + residue) of rice crop was registered with green gram incorporation and it is statistically at par with cowpea, dhaincha and summer fallow.

Nitrogen use efficiency of rice crops was significantly highest (16.41 kg grain/kg N applied through fertilizer + FYM + residue) in the treatment that received 100% RDF (W1) and percent increase over W6 was 269.59 percent. However, during the experimentation, the treatment of 75% RDF + 25% N from FYM (W₄) was at par with 50% RDF + 50% N from FYM (W₅) in the pooled study.

Rice grown in green gram-incorporated plots recorded significantly higher ARR-N (Figure 5) as depicted in 4,688 kg/ha 34.60%, Whereas ARR-P was not significant, but higher ARR-P was found with incorporation of green gram Significantly higher ARR-N (%) in rice was noticed in 100% RDF (45.72%, pooled basis) as denoted in Figure 5. The treatment 100% RDF (W₁) recorded a 42.67% apparent recovery ratio of phosphorus during pooled studies (Figure 6). From the pooled analysis, it can be revealed that the interaction effect of summer legume incorporation along with nutrient application was found to be non-significant for ARR-N and noted as significant for ARR-P, respectively.

Summer legume-incorporated plots have no significant influence on physiological efficiency of rice (Figure 7). However, higher physiological efficiency was found in rice grown in dhaincha plots (34.83 kg kg⁻¹). Higher physiological efficiency was found with the application of 50% RDF (W₃, 38.24 kg kg⁻¹), but it was statistically at par with W₂, W₅, W4, and W, which remained consistent for 2 years.

Significant variations in post-harvest soil N, P₂O₅, and K₂O levels were observed following summer legume incorporation and nutrient applications in kharif rice. Dhaincha, with its high nutrient content and rapid growth, outperformed other legumes, contributing significantly where N, P₂O₅, and K₂O become more available, underlining the critical role of legume incorporation in enhancing soil health, fertility, and overall ecosystem sustainability (Zhao N. et al., 2024). The plots that received 50% RDF + 50% N from FYM or 75% RDF + 25% N from FYM maintained soil fertility as effectively as 100% NPK treatments over 2 years and pooled data. The increase in soil nitrogen was linked to legume nitrogen fixation, residue incorporation, and fertilizer use. Phosphorus and potassium levels improved due to the breakdown of crop residues and organic matter, aided by inorganic fertilizers. The integration of FYM and inorganic fertilizers and green manure likely created favorable conditions for nutrient mineralization for increased accessible nitrogen and organic carbon enhancement (Sharma and Ghosh, 2000). Applying NPK along with organic manures (vermicompost, FYM, or green manure) significantly raised available phosphorus compared to NPK fertilizers alone, potentially due to the ability of organic inputs to complex cations that contribute to phosphorus fixation (Kamla et al., 2005; Bajpai et al., 2006). Organic acids released during green manure decomposition may have further boosted phosphorus availability (Alagappan and Venkitaswamy, 2016). Additionally, green manuring improved potassium availability through the release of exchangeable K during residue decomposition (Maiti et al., 2006; Upadhyay et al., 2011).

Incorporating summer legumes significantly increased N, P, and K uptake in rice, with dhaincha showing the highest uptake due to its incorporation at 50% flowering, unlike green gram and cowpea, which were incorporated post-harvest. The decomposition of green manures improved soil conditions, enhancing root growth and nutrient absorption (Talathi et al., 2009; Saraswat et al., 2010; Islam et al., 2014, 2019). Rice uptake of N, P, and K was significantly higher when inorganic fertilizers were combined with organic manures compared to when no fertilizer was used (W6). The highest N uptake was observed with 100% RDF, followed by 75% RDF + 25% N from FYM and 50% RDF + 50% N from FYM, may be due to optimal nutrient balance and immediate nutrient availability and similar trend observed for P and K uptake across both years and in pooled analysis. This correlation aligns with dry matter accumulation and yield per hectare across treatments and improved nutrient availability (Sunitha et al., 2010; Kumar et al., 2012, and Kumari et al., 2013).

Fertilizer contribution (Figure 8) was more in total uptake of NP, followed by others. It is because of the readily available nature of fertilizers. But in the long run, due to the decomposition of crop residues, the nutrient availability may increase after complete decomposition. Tarafdhar et al. (2016) reported that FYM applied to the crop will be available by about 30% in the first year of application and the rest will be available in subsequent years. A considerable amount of N in FYM is lost during its preparation and storage, mainly as NH₃ volatilization and leaching. Hence, in this study, the treatments with FYM were efficient only after 100% RDF chemical fertilizer application.

The dhaincha crop has a higher biomass nutrient concentration when compared to other legumes whose incorporation has shown better nutrient availability, which enhanced the growth and yield of rice in treatments involving dhaincha incorporation, ultimately improving the agronomic use efficiency (Irin et al., 2019). Enhanced agronomic use efficiency due to dhaincha crop incorporation was also reported by Chen et al. (2018), Islam et al. (2019), Walia et al. (2024), and Thulasi et al. (2024).

The apparent recovery ratio (%) of nitrogen varied from 17.67 to 50.69% during the pooled study. Among the summer legumes tested for incorporation, significantly higher ARR of nitrogen (50.69%) was found in green gram—100% RDF (T1W1) and it was followed by T2W1 (cowpea-100% RDF), and the lowest was found in T3W6 (17.67%) and T4W6 (0) in the pooled study (Figure 5). The data of ARR of phosphorus (Figure 6) showed that significantly higher ARR-P was found in the treatment when rice was grown in green gram incorporated plots with 100% application of fertilizer (T1W1) with ARRP of 48.17% during the pooled study. and the lowest was observed in T4W3 (Fallow-50% RDF, 6.77), besides absolute control (0%, T4W6). The apparent recovery ratio of phosphorus varied from 6.77 to 48.17% in pooled. The higher ARR-N was found in green gram-incorporated plots with 100% RDF when compared to dhaincha-incorporated plots with 100% RDF. Even though dhaincha supplied a larger amount of biomass, green gram biomass decomposed and released nutrients in such a way that it correlated with the uptake requirement of the kharif rice crop, which might be the cause for better apparent recovery efficiency of nitrogen and phosphorus, respectively (Peoples et al., 2017). Similar findings were supported by Rani et al. (2022), Govindasamy et al. (2023), and Vaziritabar et al. (2024).

Physiological efficiency (PE) indicates the effectiveness of nutrient accumulation and conversion from source to sink. In a pooled study, rice grown in dhaincha-incorporated plots achieved the highest PE (34.83 kg/kg), attributed to enhanced soil organic matter, nitrogen content, and microbial activity. Conversely, the lowest PE (33.08 kg/kg) was recorded in fallow plots, highlighting the role of green manuring in improving soil fertility (Figure 7). Among fertilizer treatments, the highest PE (38.24 kg/kg) was observed with 50% RDF (W3), although this was statistically comparable to other nutrient regimes (W2, W5, W4, W1). The absence of fertilizers (W6) resulted in the lowest PE (18.68 kg/kg), underlining the necessity of nutrient supplementation for optimal rice performance. The interaction of summer legumes and nutrient doses revealed that fallow plots with fertilizers yielded higher PEN, whereas legume-incorporated plots without fertilizers showed reduced efficiency (22.25–30.38 kg/kg). This suggests that while legume incorporation enhances soil fertility through nitrogen fixation and organic matter addition, it cannot entirely replace inorganic fertilizers. Instead, integrating legumes with moderate fertilizer application is more effective for optimizing PE. Overall, treatments involving summer legumes significantly outperformed the control (W6), reinforcing the synergistic benefits of green manuring and nutrient management for improved physiological efficiency in rice.