The present investigation was carried out in the Experimental Orchard of the Division of Fruits and Horticultural Technology, ICAR- IARI New Delhi, which is located at 77°07’E longitude, 28°41’N latitude, and an altitude of 227.4 m above mean sea level during the year 2021-24.The treatment consisted of various grafting combinations involving two nterstocks: no interstock, Amrapali (Dashehari x Neelum) as an interstock, and Mallika as an interstock (Neelum x Dashehari), with Olour used as both the rootstock and scion. The grafting combinations used in the experiment included both double and single grafts, as outlined in Table 1 . In mango rootstock research in India, a major limitation is the absence of standardized clonal rootstocks and the lack of reliable methods for identifying true-to-type nucellar seedlings for grafting. This has resulted in inconsistency and non-uniformity across studies, leading to variable performance. Additionally, there is a notable gap in understanding the physiological and anatomical changes occurring at the graft union, especially when interstocks are used - an area that remains largely unexplored in mango. In polyembryonic mango rootstocks, one sexual embryo and multiple nucellar embryos are typically present, with the nucellar embryos possessing the same genetic makeup as the mother plant (Cordeiro et al., 2006). These adventitious embryos arise directly from the maternal nucellar tissue surrounding the embryo sac, which contains the developing zygotic embryo. Therefore, identifying the zygotic embryo is of critical importance in mango. LMMA8 primer was used in this study to detect polymorphism and to distinguish between nucellar and zygotic seedlings derived from stones of Olour rootstock. The forward primer sequence of LMMA8 marker was 5’-CATGGAGTTGTGATACCTAC-3’, and the reverse primer sequence was 5’-CAGAGTTAGCCATATAGAGTG-3’. The primer was designed to amplify at an annealing temperature of 47 °C. The identical SSR profiles of seedlings and their mother plants (Olour) confirm their nucellar origin, while any deviation from the mother plant’s SSR profile indicates a zygotic origin ( Figure 1 ). Identified nucellar seedlings like seedling no. 152,155,167 etc. were subsequently grafted by interstocks and scions whereas zygotic seedlings like seedling no. 117, 112 etc. were discarded after screening from LMMA8 marker. The selection of Olour as a rootstock for this study is based on its well-documented salt tolerance, dwarfing ability, and genetic uniformity, making it a valuable choice for improving commercial mango cultivars in diverse environmental conditions (Dubey et al., 2007). Additionally, Olour exhibits good compatibility with most improved mango cultivars, including Langra, Himsagar, and Alphonso, providing greater flexibility in grafting combinations. One of Olour’s most significant traits is its polyembryonic nature, which ensures the production of nucellar seedlings that are genetically uniform and true-to-type. This study addresses this gap by selecting two contrasting mango hybrid cultivars–Amrapali (Dashehari x Neelum), known for its dwarfing habit, and Mallika (Neelum x Dashehari), recognized for its vigorous growth - as both interstocks and scions on a common rootstock (Olour). Amrapaliis precocious, distinctly dwarf, highly regular and prolific in bearing. It is suitable for high-density orcharding (Singh et al., 1972; Majumder et al., 1982). Mallika tree is semi-vigorous. It is medium to heavy cropper, and has a strong tendency to bear regularly. The fruits have an attractive appearance and the average fruit weight is 307 g (Singh et al., 1972). Both these hybrids original mother plants are present in ICAR- IARI experimental field (Both the hybrids were developed at ICAR-IARI). By using Amrapali and Mallika which are present on their own root system as well as in combination with Olour rootstock, the study aims to investigate how these genotypes influence scion vigour or dwarfness and to evaluate the graft-transmissible effects across the graft union. Initially ripened Olour fruits were harvested and then the stones were extracted. The extracted stones were sown in July 2022, with germination occurring within 30–40 days. After 3–4 months of seedling growth, molecular screening using SSR markers was conducted to identify true-to-type nucellar seedlings and eliminate zygotic ones. The first grafting was carried out in May 2023, followed by a second grafting in September 2023. Physiological, anatomical, and biochemical observations were recorded in 2024, ensuring sufficient growth period post-grafting- ( Figure 2 ). Softwood grafting was performed in this study, and the plants were cultivated in the open field. The grafting was carried out on newly emerged flushes. Scion wood, having the same thickness as the terminal shoot, was defoliated 10 days prior to grafting. The grafts were securely tied using a 1.5 cm wide, 200-gauge polyethylene strip. In this technique, the graft union was positioned higher above the ground, and the rootstock sprouts were regularly pinched off as they have the potential to interfere with scion growth to ensure better quality. This foundational step validates the uniformity and reliability of the experimental material. For the double-grafted plants, the length of the interstock was uniformly maintained at 10 cm to ensure consistency across treatments. All cultural practices, including irrigation, weeding, hoeing, and fertilization, were performed uniformly across the treatments to ensure consistency and eliminate variability due to management practices.

The physicochemical analysis of soil from the experimental site revealed a nutrient-rich profile ( Table 2 ). Macronutrient levels were high, with nitrogen at 2.78%, phosphorus 1.004%, calcium 3.85%, potassium 1.75%, and magnesium 0.45%. Micronutrients included zinc (369.74 ppm), copper (300.98 ppm), iron (15,878.14 ppm), manganese (381.13 ppm), boron (340.17 ppm), and sodium (19,889.11 ppm). Trace elements such as cobalt (4.97 ppm), aluminium (5938.86 ppm), chromium (32.57 ppm), nickel (16.68 ppm), cadmium (3.04 ppm), strontium (212.29 ppm), and lead (16.48 ppm) were also present. These values indicate that the soil was nutritionally sufficient to support healthy plant growth and allowed for a reliable assessment of grafting effects without confounding limitations from nutrient deficiencies.

The physical parameters of seedlings and leaves, viz., internodal length, leaf length, leaf width, leaf ratio, leaf fresh and dry weight, leaf area, no. of leaves, rootstock and scion girth, and stock: scion ratio were measured following the standard procedures. Girth of the rootstock was measured at 5 cm below the first graft union, while girth of the interstock and scion were measured at 5 cm above the first and second graft union, respectively with the help of vernier caliper. The stock-scion ratio was calculated by dividing the rootstock girth by the scion girth. To measure the average internodal length, the first 5 internodes from 60 cm height above the base of each individual plant were selected. The length of each internode was measured in cm using a measuring tape. Then, the lengths of all 5 internodes were added together to get the total length. Finally, the total length was divided by 5 to calculate the average internodal length for each plant, which was expressed in cm. Five fully mature leaves were collected from grafted plants, and their length and width were measured using a vernier caliper, with the results expressed in cm. The leaf ratio was determined using the measured values of leaf length and width. Leaf area was measured with WINFOLIAPro2020, aleaf area meter, and expressed in cm². Leaf fresh weight and dry weight (leaves were dried at 65˚C in hot air oven for 72 hours until the weight becomes constant)were measured using an electronic balance. Number of leaves was counted manually from each grafting combination.

Chlorophyll (‘a’, ‘b’, total) and carotenoids were estimated using the method described by Hiscox and Israelstam (1979). A fully opened mature leaf was taken as the experimental sample for chlorophyll estimation. Leaf samples weighing 50 mg were collected, then cut into small pieces and placed in a test tube. 10 ml of dimethyl sulphoxide (DMSO) was added to it, and the solution was kept in an oven at 60-62˚C for 4 hours. The following day, readings were taken at 480 nm, 649 nm, and 665 nm, respectively, through UV UV-Visible spectrophotometer (UV PLUS, MOTRAS SCIENTIFIC). Chlorophyll ‘a’, ‘b’, and total carotenoid content were estimated according to the formula given by Wellburn, 1994. The formulas are given below:

Gas exchange measurements were conducted on fully opened mature leaves from each genotype, with a minimum of five observations recorded between 09:00 and 11:30 h. The observations were made under ambient light and CO₂ levels. Mature leaves were placed inside an infrared gas analyzer (IRGA) (LICOR, Lincoln, NE, USA) until the relative humidity (RH) and internal leaf CO₂ concentration (Ci) reached stable values. Net photosynthetic rate (A) (μmole CO₂ m^-2 s^-1), transpiration rate (E) (m.mole H₂O m^-2), stomatal conductance (gs) (mole H₂O m^-2s^-1), and internal leaf CO₂ concentration (Ci) (μmole CO₂ m^-2 s^-1) were measured. Data were transferred to a computer for analysis. Observations on gas exchanges were recorded under a photoperiod of 11 hr, optimum day temperature of 32.7°C & night temperature, 18.9°C, and relative humidity of 71.9%. Using the Sairam et al. (1994) approach, the membrane stability index (MSI) was computed by measuring the electrical conductivity of leaf leachates in double-distilled water at two different temperatures: 40°C and 100°C. Small pieces of mature leaf, weighing 0.5 g, were chopped up and placed in two sets of test tubes containing 10 ml of double-distilled water. The electrical conductivity of both sets, denoted as C1 and C2, was measured using anelectric conductivity meter. One set of test tubes was maintained at 40°C for 30 minutes, while the other set was heated to 100 °C in a boiling water bath for 15 minutes. The electrical conductivity measurements before and after heating provide insights into the membrane stability of the leaves. The membrane stability index (MSI) was then calculated using the formula:

The experiment was conducted using a Randomized Block Design (RBD) with five treatments and five replications (n=5), with each replication containing a single seedling. The data were subjected to statistical analysis of variance (ANOVA) using R software, and significant differences were compared, followed by DMRT at p ≤ 0.05. The analysis of data was used to interpret the results and draw valid conclusions. Correlation of coefficients, clustering, heat maps, and PCA among the physical parameters, stomatal parameters, leaf nutrient contents, photosynthetic pigments, physiological and biochemical parameters were calculated and prepared using the R software.

Analysis of the data presented in Table 3 reveals that rootstock girth, scion girth and internodal length were significantly affected by the different grafting combinations. However, no significant differences were observed in the stock-to-scion ratio across the various grafting combinations. Rootstock girth ranged from 7.11 mm in the Olour/Amrapali/Olour combination to 10.28 mm in the Mallika/Olour combination, with an overall mean of 8.23 mm. The Mallika/Olour combination also recorded the highest scion girth (6.33 mm), whereas the Amrapali/Olour/Amrapali combination exhibited the lowest (4.32 mm), with a mean value of 5.01 mm across different grafting combinations. The stock-to-scion ratio remained relatively stable, ranging from 1.63 in Mallika/Olour to 1.65 in other grafting combinations, with a mean of 1.64. Internodal length varied notably among combinations, with the Mallika/Olour grafts showing the greatest length (1.67 cm), indicating enhanced vegetative vigour, while the Amrapali/Olour grafts recorded the shortest internodes (0.93 cm), reflecting a tendency toward dwarfing.

Table 4 further highlights significant differences in leaf morphology and biomass attributes among the grafting combinations, such as leaf length, width, ratio, fresh and dry weight, area, and number of leaves. The longest leaves (17.85 cm) were observed in the Mallika/Olour combination, while the shortest (9.67 cm) were recorded in Olour/Olour, with an average length of 12.45 cm. Leaf width ranged from 2.28 cm in Olour/Olour to 3.71 cm in Olour/Mallika/Olour, with a mean of 3.18 cm. The highest leaf length-to-width ratio (5.17) was foundin Mallika/Olour, and the lowest (3.38) in Olour/Mallika/Olour. The Mallika/Olour treatment also exhibited the greatest leaf fresh weight (3.03 g) and dry weight (1.11 g), whereas Olour/Amrapali/Olour had the lowest values (1.26 g and 0.40 g, respectively). In terms of leaf area and number of leaves, the Mallika/Olour combination again outperformed the others with values of 56.48 cm² and 20.20 leaves, while Olour/Mallika/Olour recorded the lowest no. of leaves i.e.6.20 leaves and minimum leaf area was reported in Amrapali/Olour (26.83 cm²). These findings underscore the substantial impact of grafting combinations on vegetative growth and physiological performance in mango.

The data presented in Table 5 demonstrates significant differences in physiological parametersnamely, intercellular CO₂ concentration, stomatal conductance, net photosynthesis, transpiration, and leaf membrane stability index (MSI)among the various mango grafting combinations. The Olour/Mallika/Olour combination exhibited the highest intercellular CO₂ concentration (356.20 µmol m^-2 s^-1), nearly 1.7 times greater than the lowest value observed in Olour/Amrapali/Olour (209.00 µmol m^-2 s^-1). The overall average across grafting combinations was 260.00 µmol m^-2 s^-1. A similar trend was evident in stomatal conductance, with Mallika/Olour showing the highest rate (0.070mol m^-2 s^-1), compared to just 0.014 mol m^-2 s^-1 in Olour/Amrapali/Olour. The net photosynthetic rate followed a similar pattern, with Olour/Mallika/Olour recording the highest value (8.51 µmol m^-2 s^-1) and Olour/Amrapali/Olour the lowest (3.50 µmol m^-2 s^-1). Transpiration rates varied from 0.36 mmol m^-2 s^-1 in Olour/Amrapali/Olour to 1.98 mmol m^-2 s^-1 in Olour/Mallika/Olour, with a mean of 1.00 mmol m^-2 s^-1. The leaf membrane stability index (MSI), an important indicator of cellular integrity and resilience, varied significantly among the grafting combinations. The Olour/Olour combination recorded the lowest MSI value (56.02%), whereas the Olour/Amrapali/Olour combination exhibited the highest (68.43%), indicating greater membrane stability and potential adaptability under suboptimal conditions. However, despite its higher MSI, the Olour/Amrapali/Olour combination showed comparatively lower photosynthetic efficiency, which may correlate with reduced vegetative vigour and a more compact growth habit. In contrast, the Olour/Mallika/Olour combination demonstrated superior performance in key physiological parameters, including higher rates of photosynthesis and transpiration, which are typically associated with enhanced plant vigour and biomass accumulation.

Biochemical analysis of the leaves demonstrated significant variations among grafting combinations in terms of total soluble protein, sugars, phenols, and proline content ( Table 6 ). The Olour/Mallika/Olour combination recorded the highest levels of total soluble protein (4.34 mg/g FW) and total sugars (119.05 mg/g FW), indicating greater metabolic activity and potentially vigorous growth. In contrast, the Olour/Amrapali/Olour combination showed markedly higher levels of phenols (3067.53 mg/100 g) and proline (1.06 μg/g FW) in the apical buds. These elevated levels are often associated with reduced cell division and elongation, and may play a regulatory role in growth inhibition. This biochemical profile aligns with the dwarfing trait observed in this grafting combination, suggesting that higher accumulation of phenolic compounds and proline could be contributing factors to its compact and restrained growth habit. The Olour/Mallika/Olour combination recorded the lowest proline (0.36 μg/g FW) levels and the lowest phenolic content (1014.31 mg/100 g). These findings underscore the influence of rootstock and interstock on the biochemical composition of mango leaves. Variations in nutrient uptake efficiency, and overall physiological performance likely contribute to the differences observed. The elevated protein and sugar levels in the Olour/Mallika/Olour combination suggest enhanced photosynthetic efficiency and nutrient assimilation, both of which are indicative of higher metabolic vigour and active growth. Such a biochemical profile typically supports robust vegetative development and overall plant vitality.Conversely, the Olour/Amrapali/Olour combination exhibited increased phenol and proline levels, which are known to influence growth regulation by modulating cellular processes such as lignification, osmotic adjustment, and cell expansion. The accumulation of these compounds may contribute to a more controlled or restrained growth pattern, resulting in reduced plant vigour. These findings emphasize the potential of specific grafting combinations for optimizing biochemical composition, which could be harnessed to improve plant performance under varying environmental conditions.

As presented in Table 7 , the levels of chlorophyll and carotenoids varied significantly among the grafting combinations, highlighting the influence of rootstock and interstock on foliar pigment composition. The Olour/Mallika/Olour combination consistently exhibited the highest concentrations of chlorophyll a (3.07 mg/g FW), total chlorophyll (4.04 mg/g FW), and carotenoids (0.22 mg/g FW), indicating enhanced photosynthetic efficiency and a greater capacity for light absorption. In contrast, the Olour/Amrapali/Olour combination showed the lowest levels of chlorophyll A (1.15 mg/g FW), total chlorophyll (1.65 mg/g FW), and carotenoids (0.13 mg/g FW), suggesting a relatively lower photosynthetic potential.Chlorophyll B content varied from 0.48 mg/g FW in Olour/Amrapali/Olour to 0.92 mg/g FW in Olour/Mallika/Olour, with the chlorophyll A:B ratio ranging between 2.41 and 3.32. These differences reflect variation in light-harvesting capacity and energy transfer efficiency among the grafting treatments. The superior pigment content observed in the Olour/Mallika/Olour combination suggests a physiological advantage that may contribute to improved growth and productivity. Figure 3 complements the findings on foliar photosynthetic pigments, depicting that the Olour/Mallika/Olour combination has significantly higher levels of chlorophyll ‘A’, chlorophyll ‘B’, total chlorophyll, and carotenoids. These pigments are directly associated with photosynthetic performance, indicating that this graft combination is physiologically more efficient.

Anatomical analysis of stomatal traits ( Figure 4 ; Tables 8 , 9 ) revealed significant differences among the grafting combinations, underscoring the influence of rootstock and interstock on stomatal structure and potential gas exchange efficiency. The Mallika/Olour combination exhibited the most prominent stomatal pore characteristics, with the longest pore length (78.31 µm), greatest pore width (72.90 µm), and the largest pore area (5708.57 µm²). In contrast, Olour/Amrapali/Olour recorded the smallest values for all three parameters - 65.60 µm length, 59.18 µm width, and 3883.55 µm² area - suggesting reduced gas diffusion potential. These variations reflect the influence of rootstock and interstock combinations on stomatal morphology, which may play a role in regulating gas exchange and water use efficiency. Regarding stomatal complex dimensions, the Mallika/Olour combination exhibited the largest complex length (260.32 µm), width (248.95 µm), and area (64806.98 µm²), suggesting greater potential for efficient gas exchange and photosynthesis. In contrast, the Olour/Olour combination showed the smallest dimensions (191.65 µm length, 179.76 µm width) and the smallest stomatal complex area (34454.50 µm²). The highest stomatal density was recorded in Olour/Mallika/Olour (746.00 mm^-2), which also demonstrated the greatest potential stomatal conductance index (0.00031), indicating enhanced capacity for CO₂ diffusion, transpiration and increased plant vigour. The perimeter of the stomata was found to be highest in Mallika/Olour (237.48 µm) and lowest in Amrapali/Olour (199.76 µm).

These findings indicate that stomatal anatomical traits are significantly influenced by grafting combinations, with direct implications for photosynthetic activityand water-use efficiency. The larger pore size and complex dimensions observed in Mallika/Olour highlight its suitability for maximizing gas exchange, while the higher stomatal density recorded in Olour/Mallika/Olour underscores its potential for promoting physiological efficiency and contributing to increased plant vigour.

As illustrated in Figure 5 and detailed in Table 10 , significant variability in macronutrient concentrations was observed among the different mango grafting combinations, reflecting the influence of rootstock and interstock on nutrient uptake and assimilation.The Mallika/Olour combination consistently demonstrated the highest concentrations of key macronutrients. Nitrogen content peaked at 1.86%, well above the overall mean of 1.22%, while the lowest value was recorded in Olour/Amrapali/Olour (0.70%). Similarly, phosphorus levels ranged from a high of 0.42% in Mallika/Olour to just 0.08% in Olour/Amrapali/Olour, indicating a pronounced disparity in phosphorus uptake efficiency. Calcium concentrations followed a comparable trend, with Mallika/Olour again leading at 3.02%, compared to the lowest value of 1.13% in Olour/Amrapali/Olour. Potassium content varied from 0.86% in Mallika/Olour to 0.41% in Olour/Olour. In contrast to the other nutrients, magnesium content was highest in Amrapali/Olour and Olour/Amrapali/Olour (both at 0.35%), and lowest in Olour/Mallika/Olour (0.16%). These results clearly indicate that the Mallika/Olour combination possesses a superior capacity for macronutrient absorption, supporting vigorous vegetative growth and efficient metabolic function. On the other hand, the Olour/Amrapali/Olour grafts consistently exhibited the lowest nutrient concentrations, suggesting limited nutrient uptake or internal transport efficiency.As presented in Table 11 , leaf micronutrient concentrations varied significantly among the grafting combinations, indicating differential nutrient uptake and distribution influenced by rootstock–interstock interactions. The Mallika/Olour combination consistently exhibited superior micronutrient profiles. Zinc content reached its highest level in this treatment (47.43 ppm), while the lowest concentration (28.02 ppm) was found in Olour/Amrapali/Olour. Iron and manganese followed similar trends, peaking in Mallika/Olour at 161.21 ppm and 93.16 ppm, respectively, and reaching minimum values in Olour/Olour (58.68 ppm) and Olour/Amrapali/Olour (23.05 ppm). Copper levels deviated from this pattern, with the highest concentration observed in Olour/Olour (68.28 ppm) and the lowest in Olour/Amrapali/Olour (21.86 ppm). Boron content ranged from 48.03 ppm in Mallika/Olour to 23.21 ppm in Olour/Amrapali/Olour, reinforcing the trend of improved micronutrient accumulation in the Mallika-based grafts. Sodium concentration was notably highest in Amrapali/Olour (309.08 ppm) and lowest in Olour/Mallika/Olour (123.77 ppm), suggesting variation in ion regulation and possibly stress response mechanisms among treatments. These findings underscore the enhanced micronutrient uptake capacity of the Mallika/Olour combination, which likely supports critical physiological processes such as enzyme activation, chlorophyll synthesis, and stress resilience. In contrast, the consistently lower micronutrient levels observed in Olour/Amrapali/Olour suggest potential limitations in root function or nutrient translocation efficiency. Heavy metal accumulation in leaves ( Table 12 ) further demonstrated the differential impact of grafting combinations. Cobalt concentrations ranged from 0.05 ppm in Olour/Amrapali/Olour and Olour/Mallika/Olour to 0.19 ppm in Olour/Olour. Aluminium levels were highest in Amrapali/Olour (51.86 ppm) and lowest in Olour/Mallika/Olour (12.22 ppm). Chromium content ranged from 1.29 ppm in Olour/Amrapali/Olour to 3.73 ppm in Olour/Olour, and nickel from 1.18 ppm in Olour/Amrapali/Olour to 5.46 ppm in Olour/Olour. Cadmium levels were highest in Amrapali/Olour (0.72 ppm) and lowest in Olour/Olour (0.13 ppm). Strontium and lead also followed a similar trend, with the highest concentrations in Amrapali/Olour (97.05 ppm and 1.71 ppm, respectively) and the lowest in Olour/Mallika/Olour (37.42 ppm and 0.73 ppm, respectively). These observations suggest that certain combinations, particularly Amrapali/Olour and Olour/Olour, may be more prone to translocating heavy metals to aerial parts. In contrast, the Mallika/Olour combination demonstrated limited heavy metal accumulation, indicating its effectiveness in restricting toxic element uptake and improving selectivity in nutrient absorption. Figure 5 correlates with the results on mineral nutrient and heavy metal accumulation in leaves. It highlights that the Olour/Mallika/Olour combination not only facilitates better uptake of essential macro and micronutrients such as nitrogen, phosphorus, potassium, calcium, and magnesium but also shows the lowest accumulation of toxic heavy metals like aluminum and cobalt. This suggests superior selective absorption and nutrient utilization capacity, potentially may be due to improved root system architecture and physiological compatibility provided by the interstock. Conversely, the Amrapali/Olour combination, while enhancing nutrient content, showed elevated heavy metal levels, possibly indicating less efficient exclusion or detoxification mechanisms.

The significant variations in rootstock girth, scion girth, stock-to-scion ratio, and internodal length among different grafting combinations highlight the influence of stock-scion compatibility on vegetative growth. The superior performance of the Mallika/Olourcombination in terms of rootstock girth (10.28 mm) and scion girth (6.33 mm) suggests enhanced vigour, likely due to efficient nutrient uptake and water transport facilitated by the rootstock’s robust vascular connections (Perez et al., 1993). Additionally, the relatively consistent stock-to-scion ratio across combinations could suggest strong graft compatibility, a critical factor for sustained growth and productivity (Hartmann et al., 2002). The greater internodal length in the Mallika/Olour combination (1.670 cm) further corroborates its superior vegetative potential, consistent with findings by Perez et al. (1993). This is supported by recent studies indicating that compatible graft combinations, often those with superior girth development, also exhibit increased internodal length, suggesting more vigorous shoot growth (Ismail and Ebeed, 2013). Studies have demonstrated that different rootstock genotypes significantly influence scion vigour and tree architecture in mango (Ali, 2023).

The observed variability in leaf morphological traits, such as leaf length, width, fresh weight, and leaf area, underscores the role of rootstock and interstock in determining the photosynthetic and assimilatory potential of grafted plants. The Mallika/Olour combination, exhibiting the highest leaf area (56.48 cm²) and fresh weight (3.03 g), aligns with reports by Perez et al. (1988) which associate larger leaf areas with higher photosynthetic efficiency and biomass accumulation. Furthermore, the superior physiological traits in Mallika/Olour, such as stomatal conductance (0.070 mol m^-2 s^-1) and net photosynthesis (8.51 μmol m^-2 s^-1), emphasize its potential for enhanced carbon assimilation. These results align with earlier studies indicating that rootstock-scion combinations significantly influence gas exchange and water use efficiency in fruit crops (Chauhan, 2006; Perez et al., 1993). It has been shown that photosynthesis variables, including stomatal conductance, intercellular CO₂ concentration, and transpiration, were all significantly influencedby the rootstock (Koepke and Dhingra, 2013). Losciale et al. (2008) found that ‘Bosc’ pear grafted onto quince EMC rootstocks allocated fewer photons to photosynthesis, increasing light dissipation and PSII damage under stress. Size-controlling rootstocks, with reduced light absorption and limited CO₂ assimilation under water stress, are more prone to photo-oxidative damage, ultimately reducing photosynthetic efficiency. Similarly, Minja et al. (2017) observed that Alphonso grafted onto Ngwangwa rootstock exhibited the greatest leaf development—recording the highest number of leaves (3.17), longest leaf length (16.83 cm), and greatest leaf width (24.42 cm) - four months post-grafting, further affirming the critical role of rootstock in influencing early vegetative growth and leaf morphology. Seedling evaluations highlight the critical role of genotype in vegetative vigour and physiological traits. Abirami et al. (2011) found that among polyembryonic mango genotypes, ‘Nekkare’ showed the highest seedling height, vigour, and leaf area, while in monoembryonic types, Amrapali recorded the lowest vigour and leaf area. Dayal et al. (2016) further reinforced the physiological impact of polyembryonic rootstocks, reporting that Olour rootstock increased total chlorophyll content and photosynthetic rate in Pusa Surya, Mallika, and Dashehari, while K-5 enhanced photosynthesis in Amrapali and Dashehari. Additionally, Olour improved transpiration rates in Pusa Arunima, Pusa Surya, and Amrapali. According to Hartmann et al. (2011), dwarfing rootstocks can modulate key physiological parameters including photosynthetic rate, transpiration rate, stomatal conductance, and osmotic potential. Similarly, Zhou et al. (2021) observed that the use of dwarfing interstocks in ‘Red Fuji’ apple scions leads to a downregulation of photosynthetic capacity. Srivastav et al. (2009) examined the association between physiological traits and vigour indices in 16 polyembryonic mango genotypes. Their findings identified chlorophyll ‘a’ content, net CO₂ assimilation rate, and total leaf phenol content as key physiological markers for evaluating seedling vigour at the nursery stage.

Biochemical analysis revealed significant differences in protein, sugar, phenol, and proline content across grafting combinations. The highest total soluble protein (4.34 mg/g FW) and total sugars (119.05 mg/g FW) in the Olour/Mallika/Olourcombination could suggest enhanced photosynthetic efficiency (showing high leaf net photosynthesis) and nutrient assimilation, consistent with findings by Sampaio and Simao (1996). Elevated levels of phenols (3067.53 mg/100 g) and proline (1.06 μg/g FW) in the Olour/Amrapali/Olour combination could suggest a physiological orientation towards growth regulation, which may contribute to its dwarfing nature. These compounds are known to play critical roles in modulating cell division, lignification, and osmotic adjustment, processes that, while enhancing stress resilience, can also result in restricted vegetative growth and reduced vigour (Yong et al., 1983). These variations reflect the metabolic adaptations imparted by rootstock and interstock combinations, crucial for optimizing performance under varying environmental conditions. Iyer (1991) suggested that higher phenolic concentrations in the apical bud correlate with reduced vigour and dwarfing in mango which are in line with the findings. Earlier, Hayat et al. (2022) reported that rootstocks significantly influence the soluble protein content in scion trees of ‘Shatangju’ mandarin. Similarly, Satisha et al. (2007) observed a notable impact of rootstocks on leaf protein content in grapevines.Phenolic compounds are abundantly present in the bark and leaves of fruit trees. Studies on apple rootstocks have indicated a link between bark phenols and indole-3-acetic acid (IAA) metabolism. Metabolomic analysis by Cui et al. (2021) revealed that the use of ‘OHF51’ (‘Old Home’ × ‘Farmingdale’) as an interstock altered metabolite profiles in both scion and rootstock, with phenolic acids and their derivatives identified as key contributors to scion dwarfism. Interstocks can induce physiological and biochemical changes in the scion. Fayek et al. (2022) reported that Paulsen 1103, used as an interstock with Flame Seedless grape under drought stress, enhanced leaf proline, phenol, and total sugar content, resulting in notable physiological and anatomical modifications in the scion. Sridevi et al. (2021) assessed physiological and biochemical traits in eight mango half-sib populations at IIHR, Bengaluru. Goa Mankurad showed the highest phenol content (197.81 mg/100 g FW).

The significant differences in photosynthetic pigments across combinations, with Olour/Mallika/Olourexhibiting the highest chlorophyll A (3.07 mg/g FW) and total chlorophyll (4.04 mg/g FW) contents, highlight its superior capacity for light harvesting and energy transfer (Feungchan et al., 1992). The chlorophyll A:B ratio, ranging from 2.41 (Olour/Amrapali/Olour) to 3.32 (Olour/Mallika/Olour), indicates variations in the efficiency of photosystem II, consistent with findings by Avila et al., 1993. These findings align with previous studies emphasizing the value of photosynthetic pigments. Abirami et al. (2011) identified chlorophyll fractions, along with phenolic content, as reliable predictors of rootstock vigour at the seedling stage. Further supporting this, Abou-Ellail et al. (2014) demonstrated wide genetic variability in Egyptian mango genotypes, with chlorophyll content ranging from 202.5 µg/g to 386.9 µg/g and carotenoid content from 19.9 µg/g to 106.2 µg/g. These results align with those of Dayal et al. (2016), who reported that the Olour rootstock induced the highest total chlorophyll content in cultivars Pusa Surya, Mallika, and Dashehari, reinforcing its role in improving pigment concentration and photosynthetic performance. Deepak et al. (2017) assessed three polyembryonic mango rootstocks - Olour, Vellaikolumban, and Nekkare - and found that Nekkare had the highest chlorophyll a, b, and total chlorophyll content, classifying it as a vigorous rootstock. Olour showed intermediate chlorophyll levels (semi-vigorous), while Vellaikolumban had the lowest, indicating the least vigour. Similarly, Mog and Nayak (2018) evaluated fifteen cashew accessions based on photosynthetic pigments and reported significant histological differences. ‘Bhaskara’ exhibited the highest chlorophyll a and total chlorophyll content, whereas ‘Selection-2’ recorded the highest carotenoid concentration, highlighting the role of pigment traits in assessing vigour. The anatomical analysis further revealed larger stomatal dimensions in the Mallika/Olour combination, including pore area (5708.57 μm²), which may facilitate enhanced gas exchange and transpiration. These results support the hypothesis that stomatal traits are modulated by rootstock-scion interactions, influencing water use efficiency and photosynthetic capacity (Hartmann et al., 2002; Mukherjee and Das, 1980). Stomatal density has been closely associated with plant vigour in mango.In mango, Chakladar (1967) and Bang (1970) reported a positive correlation between stomatal density and plant vigour, with vigorous rootstocks exhibiting higher stomatal densities. Similar associations were observed in apple (Beakbane and Majumder, 1975) and plum (Pathak et al., 1977), as well as in guava (Saroj et al., 1997). Majumder et al. (1972, 1981) even proposed a classification system for mango rootstocks based on stomatal traits.Earlier Zhou et al. (2020) also reported variation in stomatal density in scion varieties of apple and found more stomatal density in dwarfing rootstocks M.9 and B.9 than in vigorousrootstock Baleng on apple scion variety ‘Red Fuji’. Sridevi et al. (2021) reported that Goa Mankurad showed the highest stomatal density (1244/mm²), Malanji recorded the largest stomatal size (252.16 µm), and Mandoor Khatta exhibited the highest total chlorophyll content (6.69 mg/100 g FW).

The superior macronutrient uptake in the Mallika/Olour combination, particularly nitrogen (1.86%) and calcium (3.01%), underscores its enhanced ability to support growth and metabolic processes. The higher micronutrient levels, such as zinc (47.43 ppm) and iron (161.20 ppm), further indicate the role of rootstock in improving nutrient assimilation, consistent with earlier findings (Willis and Marler, 1993). Conversely, the Olour/Amrapali/Olour combination exhibited lower nutrient concentrations and higher heavy metal accumulation, reflecting reduced nutrient uptake efficiency and potential environmental stress susceptibility. The ability of the Mallika/Olour combination to limit heavy metal translocation to leaves aligns with studies by Jones (1984), emphasizing its potential for sustainable cultivation in contaminated soils. These findings are consistent with extensive research across perennial fruit crops, which has established that rootstocks play a pivotal role in modulating mineral uptake, transport, and assimilation. In apple (Malus domestica), rootstocks such as M9 and MM106 exhibit distinct efficiencies in the absorption of key nutrients, including nitrogen (N), potassium (K), calcium (Ca), phosphorus (P), manganese (Mn), and iron (Fe) (Amiri et al., 2014). Similarly, in grapevine (Vitis vinifera), rootstock architecture significantly influences nitrogen use efficiency (NUE) and phosphorus acquisition (Zamboni et al., 2016). In stone fruit species, rootstocks have been shown to affect the macro and micronutrient composition of vegetative and reproductive tissues, as reported in cherry (Prunus avium) (Hrotko et al., 2014; Jimenez et al., 2004), peach (Prunus persica) (Mestre et al., 2015; Zarrouk et al., 2005), and plum (Prunus domestica) (Reig et al., 2018), emphasizing the central role of rootstocks in nutrient dynamics across diverse fruit crops. Lazare et al. (2020) demonstrated that avocado rootstocks influence tree nutritional status through differential mineral transport, as reflected in the leaf ionome of the scion. Rossdeutsch et al. (2020) found that in grapevines, less vigorous rootstocks exhibited greater nitrate uptake and assimilation following nitrate resupply compared to more vigorous rootstocks, likely due to higher root carbohydrate reserves. Sarkhosh et al. (2021) explored the influence of rootstocks on scion nutrient status and highlighted their role in cultivar improvement.

The heat map and hierarchical clustering analysis ( Figure 6 ) revealed distinct groupings of grafting combinations based on morpho-physio-chemical, nutrient contents, and anatomical parameters. Each row represents a specific grafting combination, while each column corresponds to a measured parameter, such as sugars, stomatal characteristics, chlorophyll content, mineral nutrients, etc. The colour scale reflects the relative magnitude of each parameter, with dark blue or purple shades indicating lower values, green representing intermediate values, and yellow denoting higher values. This colour gradient enables quick visual identification of which combinations performed better or worse for each trait.Grafting combinations with similar traits were clustered together, indicating shared physiological and biochemical attributes. For instance, combinations such as Olour/Mallika/Olour and Mallika/Olour were closely grouped due to their higher total protein and sugar contents, while Amrapali/Olour and Olour/Amrapali/Olour formed a separate cluster characterized by elevated phenol and proline levels. This clustering reflects the influence of rootstock and interstock combinations on the overall biochemical, physiological and anatomical profiles, suggesting that specific grafting combination scan be targeted to optimize desired traits. These findings support earlier studies emphasizing the role of grafting in modulating biochemical profiles (Hayat et al., 2022; Cui et al., 2021; Fayek et al., 2022; Yong et al., 1983).

The correlation analysis highlighted significant relationships among the measured parameters ( Figure 7 ). Leaf total soluble protein showed a strong positive correlation with total sugars (r > 0.8, p< 0.05), suggesting that protein synthesis may be linked to carbohydrate metabolism in grafted plants. Conversely, phenol content exhibited a negative correlation with total sugars and protein, indicating a trade-off between primary and secondary metabolite pathways, consistent with the findings of Cui et al. (2021); Satisha et al. (2007) and Yong et al. (1983), who observed similar metabolic adjustments in grafted plants. Proline content was positively correlated with phenols but negatively correlated with protein and sugars, reflecting its role in vigour control and osmoprotection under variable physiological conditions. These correlations highlight the complex interplay of metabolic pathways influenced by grafting combinations.

The principal component analysis (PCA) revealed that the first two principal components (PC1 and PC2) accounted for the majority of the variance in the dataset, cumulatively explaining over 80% of the variability ( Figure 8 ). PC1 was primarily associated with parameters like total protein, sugars, and phenols, while PC2 was strongly influenced by proline and anatomical traits. Vigorous combinations like Olour/Mallika/Olour were positioned prominently in the PCA biplot, associated with traits favouring photosynthetic efficiency and metabolic vigour. In contrast, Olour/Amrapali/Olour exhibited a clear separation based on traits related to dwarfness, including higher proline and phenol levels, which contribute to stress tolerance and reduced vegetative growth. These results indicate that rootstock and interstock combinations contribute significantly to the phenotypic and metabolic diversity in grafted plants, allowing the identification of specific combinations for tailored traits, as supported by earlier findings (Yonemoto et al., 2007).

These results align with existing research showing that grafting combinations influence tree size, nutrient uptake efficiency, and stress responses (Karlidag et al., 2014). Vigorous interstocks, such as those used in the Olour/Mallika/Olour combination, promote faster canopy development and efficient nutrient utilization, which are essential for maximizing light interception and photosynthetic output. In contrast, dwarfing combinations like Olour/Amrapali/Olour are advantageous for high-density planting systems, as their compact growth habit facilitates management and maximizes space utilization (Yonemoto et al., 2007).

The integration of heat map clustering, correlation analysis, and PCA underscores the multidimensional nature of grafting-induced variations in plants. The clustering results align with the observed biochemical trends, confirming that rootstock and interstock combinations influence nutrient allocation and metabolic adjustments. Correlation analysis highlights the interdependence of key physiological pathways, such as the trade-offs between primary and secondary metabolites, which are crucial for optimizing plant performance under different environmental and nutritional conditions. PCA further strengthens this understanding by visually separating grafting combinations based on their unique trait profiles, providing a robust framework for selecting combinations tailored to specific traits. These findings emphasize the importance of multivariate approaches in understanding the complex interactions that govern plant responses to grafting.