Groundnut (Arachis hypogaea L.) is the most important annual oilseed legume of the semi-arid tropics, valued for edible oil, protein-rich kernels and nitrogen-enriching residues. India is the second largest producer, but most of the area lies in rainfed, resource-poor tracts where shallow, light-textured soils, erratic rainfall and low native fertility constrain productivity (Directorate of Economics and Statistics, 2024). As a nodulating legume, groundnut obtains much of its nitrogen (N) through symbiotic fixation, yet this is often limited by scarce or poorly competitive native rhizobia and by phosphorus (P) deficiency, because P is indispensable for nodule initiation, bacteroid energy metabolism and nitrogenase activity (Sulieman and Tran, 2015).

P is required for nodule primordium formation, for the high respiratory and phosphorylation demand of bacteroids and for nodule acid phosphatases that mobilize internal P reserves when supply is limiting (Zhong et al., 2023). Legumes therefore have a high P requirement, aggravated in semi-arid calcareous and red soils where P is fixed as calcium or iron/aluminium phosphates (Chen et al., 2023; Zhong et al., 2023). Combined application of rhizobial inoculant and P fertilizer often gives synergistic gains in nodulation and yield, because adequate P amplifies the response to inoculation while active nodulation improves P acquisition (Asante et al., 2020; Míguez-Montero et al., 2020; Naveen and Senthilkumar, 2021).

Strain adaptation is a further, less-examined dimension. Commercial cultures do not always out-compete indigenous populations or tolerate local stresses as well as isolates from the same niche, so locally adapted strains may give better nodule occupancy, persistence and symbiotic effectiveness (Gouda et al., 2018; Mendoza-Suárez et al., 2021). Effective native groundnut rhizobia have been identified in Ghana (Osei et al., 2018) and native isolates of groundnut have also been characterized and evaluated (Chandrakanta et al., 2024). However, studies on Rhizobium-P interactions come largely from Africa (Asante et al., 2020; Kyei-Boahen et al., 2017) or have used consortium or phosphate-solubilizing cultures rather than a locally isolated Bradyrhizobium strain (Bajya and Sharma, 2024; Ravi et al., 2024) and quantitative data on how a locally adapted strain performs against a recommended culture across a graded P series and how this translates into nutrient-use efficiency and profitability, are meagre for the semi-arid red soils of Telangana.

This study therefore aimed at evaluating (i) the effect of Rhizobium inoculation on growth, nodulation and nutrient uptake of groundnut, (ii) the optimum P level, (iii) the Rhizobium × P interaction and (iv) the economics of the resulting combinations, in order to recommend a practical inoculant and P package for rainfed groundnut.

Site and soil A field experiment was conducted during the Kharif (rainy) season of 2025 at the Instructional and Research Farm, School of Agriculture, SR University, Warangal, Telangana, India (18.0°N, 79.6°E; 265 m above mean sea level). The climate is semi-arid tropical, with a hot, dry summer and monsoon rains (June-October) supplying most of the 950-1000 mm mean annual rainfall; soils are shallow to medium-deep red sandy loams (Alfisols). The experimental soil (0-15 cm) was sandy loam, neutral (pH 7.2, 1:2.5), non-saline (EC 0.28 dS m^-1), low in organic carbon (0.42%, Walkley-Black) and available N (188 kg ha^-1, alkaline permanganate) and medium in available P (16.4 kg P₂O₅ ha^-1, Olsen) and K (218 kg K₂O ha^-1, neutral normal ammonium acetate).
 
Treatments and crop management
 
The experiment followed a randomized block design with a 3 × 4 factorial arrangement and three replications (36 plots; gross plot 4.2 × 3.6 m). Factor A comprised three Rhizobium regimes: R1, uninoculated control; R2, seed treatment with a recommended commercial Bradyrhizobium sp. (Arachis) culture (200 g per 10 kg seed, ≥10⁸ viable cells g^-1) in jaggery slurry; and R3, seed treatment at the same rate with a locally isolated, phosphate-solubilizing native Bradyrhizobium strain multiplied in the host institution’s microbiology laboratory. Factor B comprised four P levels (P0, P20, P40 and P60 = 0, 20, 40 and 60 kg P₂O₅ ha_^-1) applied as single super phosphate at sowing. All plots received a basal 20 kg N and 30 kg K₂O ha^-1. Groundnut cv. Dharani (K-9), a bunch-type, semi-spreading, 105-110-day cultivar recommended for rainfed kharif cultivation in Telangana and Andhra Pradesh, was sown at 30 cm × 10 cm. Gap filling, weeding, plant protection and life-saving irrigation followed the recommended package of practices and were uniform across treatments.

Observations
 
Five randomly tagged plants per plot were used to record plant height, leaf area index (LAI; non-destructive leaf-area-ratio method), dry matter and SPAD chlorophyll readings (SPAD-502, Konica Minolta) at 60 days after sowing (DAS), when nodulation differences are most discernible. Roots were excavated and washed; nodules were counted and oven-dried at 70±2°C and root dry weight was recorded. Tissue N and P were estimated by the micro-Kjeldahl and vanado-molybdo-phosphoric yellow colour methods and multiplied by dry matter to obtain uptake (kg ha^-1). At harvest, pods plant^-1, pod weight plant^-1, shelling percentage and 100-kernel weight were recorded; pod and haulm yields were recorded plot-wise and expressed at 8% pod moisture. Kernel oil was determined by Soxhlet extraction and protein as kernel N × 6.25.
 
Efficiency and economics
 
Within each Rhizobium regime, agronomic efficiency of applied P (AE, kg pods kg^-1 P₂O₅) was calculated as (pod yield at a given P level - pod yield at P0)/P₂O₅ applied and additional P and N uptake per kg P₂O₅ were derived likewise from uptake means. Cost of cultivation was worked out treatment-wise from prevailing input costs (seed, inoculant, fertilizer, labour) and gross return from prevailing prices of pods and haulm. Net return was gross return minus cost of cultivation and benefit:cost ratio (BCR) was gross return divided by cost of cultivation.
 
Statistical analysis
 
Data were subjected to analysis of variance for a two-factor factorial in randomized block design (Gomez and Gomez, 1984), with replication, Rhizobium, P level and their interaction as sources of variation. Means were compared using the critical difference (CD) at P = 0.05 from the pooled error mean square; standard errors of the mean (SEm±) are also reported. Computations were done in Python 3.11 (pandas, statsmodels).

Growth and physiological attributes
 
Rhizobium inoculation and P level significantly influenced all growth attributes at 60 DAS (Table 1). Plant height, LAI, dry matter and SPAD index were highest with the locally adapted strain (52.02 cm, 3.38, 51.08 g plant^-1 and 43.13), followed by the recommended culture (48.78 cm, 3.15, 47.48 g and 41.03) and the uninoculated control (44.19 cm, 2.81, 43.03 g and 39.01). Each P increment further improved growth; 60 kg P₂O₅ ha^-1 increased plant height by 21.9% and dry matter by 35.4% over no P. The R × P interaction was significant only for plant height and SPAD index, so LAI and dry matter responded additively. Higher SPAD readings indicate better leaf N status, consistent with greater N fixation and uptake (Zhong et al., 2023).



Nodulation and root growth
 
Nodulation was the most responsive trait (Table 1). Relative to uninoculated plants (40.7 nodules and 77.7 mg nodule dry weight plant^-1), the recommended culture increased nodule number by 32.5% and nodule dry weight by 30.1% and the local strain by 56.0% and 54.2%, respectively. Nodule number and dry weight rose by 80.8% and 70.3% from 0 to 60 kg P₂O₅ ha^-1. The interaction was significant (Table 3): the nodule advantage of the local strain over uninoculated plants widened from 17.9 plant-1 at P0 to 27.3 at P60 and the maximum (80.5 nodules and 149.8 mg plant^-1) occurred with R3P60. Even without added P, the local strain (46.1 nodules) out-nodulated uninoculated plants given 20 kg P₂O₅ ha^-1 (37.2). Root dry weight followed the same trend (Table 1).

This synergy is well documented: P is required for nodule development and nitrogenase function and P starvation impairs nodule development and N fixation (Wang et al., 2020; Zhong et al., 2023). Positive inoculant × P interactions on nodulation have likewise been reported in groundnut (Asante et al., 2020) and cowpea (Kyei-Boahen et al., 2017), as has greater nodulation with P plus a microbial culture in groundnut (Ravi et al., 2024) and with combined N and P in bambara groundnut (Hasan et al., 2021). The local strain out-nodulated the commercial culture at every P level (by 7.5-10.8 nodules plant^-1), probably reflecting better compatibility with the host and local soil and its phosphate-solubilizing ability; competitiveness and persistence of inoculant strains are decisive for nodule occupancy (Mendoza-Suárez et al., 2021). Strain occupancy was not, however, assessed here.
 
Nutrient uptake
 
N and P uptake increased significantly with rhizobium and P level (Table 1). Mean N uptake was 24.1% (recommended) and 41.0% (local strain) higher than in uninoculated plants and P60 raised it by 43.5% over P0; corresponding P uptake gains were 9.9%, 15.9% and 38.7%. Maximum N (138.0 kg ha^-1) and P uptake (10.9 kg ha^-1) occurred with R3P60 (Table 3), 101.7% and 57.3% above the control (R1P0). Greater N uptake by inoculated plants reflects enhanced N fixation, while the larger root systems (Table 1) and the phosphate-solubilizing strain (Bajya and Sharma, 2024) probably improved P acquisition, for which root growth is central in legumes (Chen et al., 2023).
 
Yield attributes and yield
 
Pods plant^-1, pod weight, shelling percentage and 100-kernel weight all increased with inoculation and P (Table 2); the interaction was significant only for pods plant^-1 and pod weight and R3P60 produced the most pods (43.1 plant^-1; Table 3). Mean pod yield was 2818, 3313 and 3666 kg ha^-1 for R1, R2 and R3 and 2665, 3063, 3457 and 3877 kg ha^-1 at 0, 20, 40 and 60 kg P₂O₅ ha^-1; haulm yield followed a similar trend (Table 2). The interaction was significant (Table 3): R3P60 recorded the highest pod yield (4370 kg ha^-1) and haulm yield (4492 kg ha^-1), 86.7% and 56.3% above the uninoculated, unfertilized control. The local strain out-yielded the recommended culture at every P level (by 256-422 kg ha^-1, or 9.5-11.2%; CD = 89.5) and R3P40 (3901 kg ha^-1) was at par with R2P60 (3948 kg ha^-1), so the local strain saved 20 kg P₂O₅ ha^-1. Even R3P20 (3437 kg ha^-1) exceeded uninoculated plants given 60 kg P₂O₅ ha^-1 (3312 kg ha^-1). Comparable gains from combined inoculation and P have been reported in groundnut and cowpea (Asante et al., 2020; Kyei-Boahen et al., 2017; Naveen and Senthilkumar, 2021; Ravi et al., 2024) and here reflect more nodules, greater N and P uptake and better pod set.




 
Nutrient-use efficiency
 
AE was 15.5-16.2 kg pods kg^-1 P₂O₅ without inoculation, 19.9-20.8 with the recommended culture and 23.6-24.1 with the local strain and did not decline with P dose; inoculation thus raised AE by 26-30% (R2) and 46-53% (R3). N uptake per kg P₂O₅ was likewise stable with the local strain (0.69-0.72 kg). In contrast, additional P uptake per kg P₂O₅ fell from 0.064 kg at P20 to 0.057 and 0.053 kg at P40 and P60 with the local strain, indicating diminishing recovery of applied P, as expected in soils that fix P (Lambers, 2022). Thus, although P recovery declined at higher doses, yield efficiency did not and the response had not plateaued at 60 kg P₂O₅ ha^-1.
 
Kernel quality
 
Kernel oil and protein rose significantly but modestly with inoculation (oil 47.13-48.08%, protein 23.92-24.90% from R1 to R3) and P (oil 47.01-48.19%, protein 23.97-24.80% from P0 to P60), without interaction (Table 2), consistent with improved N and P nutrition.
 
Economics
 
Net return and BCR increased with both factors (Table 2). Relative to uninoculated plants, inoculation raised mean net return by Rs 27,875 (recommended) and Rs 48,162 ha^-1 (local strain) while adding only Rs 900-1,200 ha^-1 to the cost of cultivation and 60 kg P₂O₅ ha^-1 added Rs 67,644 over no P. R3P60 gave the highest net return (Rs 202,422 ha^-1) and BCR (4.62), 129.4% above the control (Rs 88,232) and R3P40 (Rs 175,988; BCR 4.20) was at par with R2P60 (Rs 177,861; BCR 4.20). Each 20 kg P₂O₅ ha^-1 increment added Rs 26,093-27,162 ha^-1 with the local strain, so 60 kg P₂O₅ ha^-1 remained the most profitable dose tested, in agreement with Ravi et al., (2024), who recorded the highest net return and benefit:cost ratio in groundnut with 60 kg P ha^-1 plus a microbial culture.

Rhizobium inoculation and P nutrition acted synergistically on nodulation, nutrient uptake and productivity of rainfed groundnut on a semi-arid red sandy loam. The locally isolated phosphate-solubilizing Bradyrhizobium strain outperformed both the uninoculated control and the recommended commercial culture at all P levels and with 40 kg P₂O₅ ha^-1 it matched the commercial culture at 60 kg P₂O₅ ha^-1. Nodulation, nutrient uptake, pod and haulm yield and net return rose with each P increment up to 60 kg P₂O₅ ha^-1 without any decline in agronomic efficiency, so the local strain with 60 kg P₂O₅ ha^-1 gave the highest pod yield (4370 kg ha^-1) and net return (Rs 202,422 ha^-1; BCR 4.62), 86.7% and 129.4% above the control. Where P supply is limited, the local strain with 40 kg P₂O₅ ha^-1 is an economical alternative. Because the study covered one season and site and response had not plateaued at 60 kg P₂O₅ ha^-1, multi-location, multi-season validation with higher P rates and nodule occupancy assessment is needed.

The authors thank the School of Agriculture, SR University, Warangal, for providing field and laboratory facilities. The present study received no external funding.
 
Disclaimers
 
The views and conclusions expressed in this article are solely those of the authors and do not necessarily represent the views of their affiliated institutions. The authors are responsible for the accuracy and completeness of the information provided, but do not accept any liability for any direct or indirect losses resulting from the use of this content.
 
Informed consent
 
This study was a field crop trial and did not involve human participants, human data or animals; ethics approval and informed consent were therefore not applicable.

The authors declare that there are no conflicts of interest regarding the publication of this article. No funding or sponsorship influenced the design of the study, data collection, analysis, decision to publish, or preparation of the manuscript.