Introduction

In natural conditions, plants are exposed to a multitude of biotic (e.g. insects, nematodes, viruses, fungi, bacteria) and abiotic (e.g., drought, salinity, extreme temperatures, heavy metals, imbalances in plant nutrients) stress factors. These factors can have a detrimental effect on plant growth, development and productivity, from the initial stage of seed germination. In response to these stress factors, plants initiate defense mechanisms and undergo metabolic restructuring. Abiotic stress factors, in particular, have been shown to have serious limiting effects on global food security, food quality, and plant productivity1.

Drought, salinity, extreme temperatures, and low soil fertility are widely considered to be the most important abiotic stress factors limiting agricultural production in the world. Drought is a major problem, especially in developing countries with agriculture-based economies. Research indicates that the effects of drought will intensify in the future, adversely impacting agricultural production2,3. A study by Gaur and Squires4 found that 41% of the world’s land and 36% of agricultural land are currently affected by drought. Despite the increase in irrigation water use worldwide, only 20% of agricultural land is irrigated, and 40% of irrigated land provides agricultural production5. The increasing importance of drought and water deficit is especially pronounced in regions where irrigable land is decreasing and agricultural production is more dependent on rainfall.

Water stress is one of the most significant limiting factors in vegetable cultivation. Solanum melongena L., commonly known as eggplant, is a member of the Solanaceae family that requires moderate water and is sensitive to drought stress. Water restriction has been shown to have a negative impact on eggplant yield and quality6. In order to mitigate the deleterious effects of drought stress on eggplant, strategies such as the development of resistant varieties through classical or biotechnological methods and genetic modifications are being implemented6. In addition, the use of tolerant rootstocks has been shown to enhance plant tolerance to stress7. These methods have been shown to enhance plant growth and yield8,9.

However, there are instances where grafting proves ineffective in long-term and severe drought conditions. The stress tolerance provided by grafting is only effective in mitigating the immediate effects of drought; it does not offer more enduring solutions. Consequently, there is a need for environmentally friendly and integrated strategies to ensure more sustainable drought management. In this context, VC stands out as a viable solution. The use of VC has been shown to enhance soil moisture retention, thereby reducing plant dependency on water10. Additionally, VC has been demonstrated to enhance soil productivity and quality, thereby increasing plant stress tolerance11,12,13. Previous studies have reported that both VC application and grafting, when evaluated separately, can enhance crop production under drought stress conditions14,15,16. For instance, studies focusing solely on VC have demonstrated improvements in soil water-holding capacity and plant growth16, while those investigating only grafting have reported increased yields linked to stronger root systems and improved water uptake under drought14,15,17. However, there is a lack of systematic studies exploring the potential synergistic effects of combining these two approaches.

This study aims to fill this research gap by investigating the combined effects of grafting and VC application on eggplant yield, quality traits, and soil parameters under drought stress. In doing so, it addresses the absence of integrated assessments regarding the synergy between VC and grafting on drought-tolerant rootstocks. Accordingly, the hypothesis was proposed that combining vermicompost with grafted rootstocks would significantly enhance eggplant yield, fruit quality, and soil properties under both mild and severe drought stress conditions, compared to their individual application.

In this study, (i) Whether the use of grafted plants and VC under water constraint conditions has an effect on improving drought tolerance and improving yield and quality characteristics in eggplant, (ii) The changes in soil properties caused by applications in limited water conditions and how these changes were reflected in the stress tolerance of eggplant were examined in both greenhouse and open field conditions.The concrete outputs of the study, through quantitative data obtained from a greenhouse and open field trials, include increases in plant yield, improvements in factors such as soil moisture and mineral content, and improvements in quality parameters (antioxidant capacity, phenolic compounds). These findings have the potential to contribute to integrated strategies for drought management.

Results

Findings obtained under greenhouse conditions

The greenhouse experiment served as a preliminary study to determine the vermicompost dose to be used in the field trial. Therefore, only the key greenhouse findings relevant to dose selection are presented.

Changes in yield, yield components, and some fruit quality characteristics

In the greenhouse environment, grafting, VC doses, drought stress, and their interactions significantly affected various yield- and quality-related traits. The three-way interaction (G x VC x DS) was significant for yield per plant (p < 0.05), FDM (p < 0.01), and TSS (p < 0.05) (Table 1; Fig. 1a), but not forfruit diameter or average fruit weight (p > 0.05) (Table 1). Grafted plants and VC treatments consistentlyincreased yield per plant compared with the control (0% VC and N–G). Although drought reduced yield under both mild (MS) and severe (SS) stress, grafting combined with VC effectively mitigated these losses. Under MS conditions, the highest yield was obtained from the ‘G x 2% VC’ (93.55%) and ‘G x 3% VC’ (97.85%) combinations, which did not differ statistically. Under SS conditions, the ‘G x 2% VC’ combination resulted in a 144.12% increase (Fig. 1a). Since 2% and 3% VC produced comparable yield responses, 2% VC was selected for the field experiment as a more cost-effective and practical dose for large-scale applications.

Drought stress increased FDM in fruits, whereasVC application to grafted plants moderated this increase at both stress levels. The greatest reductions in FDM relative to the control were recorded in ‘G x 3% VC’ (14.35%) under MS and in ‘G x 3% VC’ and ‘G x 2% VC’ (24.70% and 22.85%) under SS (Table 2). Drought stress also significantly increased TSS content, and the three-way interaction was significant (p < 0.05) (Table 1). Grafting and VC tended to counteract stress-induced increases in TSS, with ‘G x 3% VC’ showing the largest reductions under both MS and SS (Table 2).

Metabolic responses to drought revealed increases in total phenolics, antioxidant activity, and total flavonoids, with the magnitude of increase depending on stress severity (Table 2). The combined use of grafted plants and VC further enhanced these increases, and the ‘G x VC x DS’ interaction was significant for all three traits (p < 0.01) (Table 1). Compared with the control, the ‘G x 2% VC’ treatment increased total phenolics by 97.85% under MS and 144.12% under SS, yielding values of 23.06 µg GAE ml⁻¹ and 28.19 µg GAE ml⁻¹, respectively (Table 2). Similarly, antioxidant activity increased by 51.31% (144.38 µg GAE m⁻¹) under MS and by 45.51% (117.51 µM Trolox g⁻¹) under SS. Total flavonoid content increased by 38.15% under MS and by 43.40% under SS (Table 2).

Table 1 Variance analysis of yield, yield component, and some fruit quality characteristics.
Full size table
Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Effect of grafted plant use and vermicompost (VC) application on yield under greenhouse (a) and field (b) conditions. MS: mild stress, SS: severe stress.

Table 2 Effect of ‘G x VC x DS’ interaction on average fruit weight, fruit diameter, fruit dry matter, total soluble solid, total phenolic matter, antioxidant activity, and total flavonoids.
Full size table

Correlation analysis, polar heatmap dendrogram, and principal component analysis in greenhouse data

Correlation analysis revealed that yield (Y) was positively correlated with fruit diameter (FD) and negatively correlated with TSS and fruit dry matter (FDM) (Fig. 2a), reflecting balance between high yield and concentration-based quality parameters. No significant correlation was found between yield and TPM and AA. Greater differences and clearer clusters were observed in grafted and vermicomposted plants, especially under SS conditions, compared to MS (Fig. 2b). The dendrogram showed direct clustering between Y, FD and average fruit weight (AFW) and TPM, TSS, and FDM. Moreover, similar clustering was also observed between AA, TPM, and TF, Y and FD.

PCA also summarised the treatment effects on yield and quality traits under different drought levels (Fig. 3). Under control conditions, the first two components explained 63.42% of the total variation (Fig. 3a), increasing to 76.24% under MS (Fig. 3b), and 72.90% under SS (Fig. 3c). Under MS conditions, PC1 is positively correlated with FD (0.810), AFW (0.760), and Y (0.658), and negatively correlated with FDM (-0.783) and TSS (-0.769), representing a gradient from high yield, more concentrated fruits. PC2, was strongly and positively correlation with antioxidant properties, particularly AA (0.867) and TF (0.807), as well also with Y (0.717). This component likely represents both plant defense mechanisms and productivity. Under SS conditions, PC1 was strongly correlated with Y (0.979), TPM (0.924), AA (0.833), and TF (0.802), indicating that it reflects both biochemical quality traits and yield. In contrast, PC2 is associated with the morphological yield components AFW (0.828) and FD (0.526). ‘G x 2% VC’ showed a strong positive correlation with Y, TF, AA, and TPM in MS and SS conditions. Conversely, ‘G x 1% VC’ and ‘G x 3% VC’ in MS exhibited weak correlations with these traits (Fig. 3b).

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Evaluation of ‘VC x DS’ interaction under greenhouse conditions by correlation analysis (a) and polar heatmap with dendrogram (b). N-G: non-grafted. G: grafted. VC: vermicompost. Y: yield. AFW: average fruit weight. FD: fruit diameter. TSS: total soluble solid. FDM: fruit dry matter. TPM: total phenolic matter. AA: antioxidant activity. TF: total flavonoids.

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

PCA of yield, yield components, and quality characteristics in eggplant under different drought conditions (control- a, mild stress- b, severe stress- c) in the greenhouse. N-G: Non-Grafted. G: Grafted. VC: Vermicompost. Y: Yield. AFW: Average fruit weight. FD: Fruit diameter. TSS: Total soluble solid. FDM: Fruit dry matter. TPM: Total phenolic matter. AA: Antioxidant activity. TF: Total flavonoids.

Effects of treatments on some soil properties under greenhouse conditions

Analysis of variance for the treatments and the soil organic matter and physical properties is presented in Table 3, along with the Duncan groupings. Vermicompost and irrigation levels had significant effects on soil organic matter (p < 0.01). Under drought stress, the highest soil organic matter was observed in the ‘3% VC x Control’ combination (0.92%), while the lowest was recorded in the ‘0% VC x Control–100% irrigation’ (0.43%) (Fig. 4a). VC doses significantly influence bulk density, field capacity, available water capacity, saturated hydraulic conductivity, aggregate stability, and porosity (p < 0.05). However, VC doses did not significantly influence wilting point, structure stability index, or mean weighted diameter (p > 0.05).. Volume weight decreased from 1.26 g cm− 3 in the control treatment (0% VC) to 1.19 g cm− 3 in the 3% VC treatment, indicating an improvement in soil physical conditions with increasing VC dose.

Table 3 Analysis of variance results of some soil properties and the effect of vermicompost treatment on some physical properties of soils under greenhouse conditions.
Full size table
Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Effect of ‘VC x DS’ interaction on organic matter of soils under greenhouse (a) and field (b) conditions.

Findings on field conditions

Changes caused by applications on yield, yield components, and some fruit quality characteristics

In the field experiment, the 2% VC dose determined in the greenhouse was tested under open-field conditions to assess its effectiveness and applicability. Increasing water limitation caused pronounced yield losses, with the greatest reductions observed under SS in non-grafted plants without VC (p < 0.01) (Fig. 1b). VC application significantly increased yield under drought stress, and these improvements were further enhanced in grafted plants. The highest yield increase (106.5%, 4734.8 kg ha⁻¹) was recorded in grafted plants treated with VC under SS, whereas under MS the increase was 101.1% (5641.7 kg ha⁻¹) (Fig. 1b).

The three-way interaction was not significant for average fruit weight, fruit diameter, or fruit length (p > 0.05) (Table 4), indicating that yield improvements were achieved without notable changes in these fundamental efficiency components.Under stress conditions, the ‘G x VC’ interaction was significant for EC, titratable acidity (TA), and peel hue angle (hue) (p < 0.01), but not for pH, total soluble solids (TSS), or color (chroma) (p > 0.05) (Table 4). EC, TA, and hue values were higher in plants grown under stress conditions compared to the control (Table 4). However, the ‘G x VC’ interaction substantially mitigated these increases, with grafting and VC reducing EC values by 15% under MS and 9.56% under SS. Similarly, stress-induced TA increases were reduced by the combined treatment, with decreases of 45.83% under SS and 25.53% under MS. Although hue increased under stress, the combined grafting and VC treatment limited this increase, resulting in hue increases of only 2.87% under MS and 47.48% under SS.

Table 4 Variance analysis of yield, yield component, and some fruit quality characteristics and the effect of ‘G x VC x DS’ interaction on average fruit weight, fruit diameter and length, EC, pH, TSS, TA, chroma and Hue.
Full size table

Correlation analysis, Polar heatmap dendrogram and principal component analysis (PCA) in field data

Correlation analysis identified various significant relationships between yield and quality characteristics (Fig. 5a).Positive correlations: pH and yield (Y); EC and Hue; Chroma and Hue. Negative correlations: TA and Chroma; TA and Hue. The polar heat map and dendrogram (Fig. 5b) highlight the effect of VC and grafting on the multivariate response model under different drought levels by grouping applications and characteristics according to these relationships.

PCA bi-plots (Fig. 6) summerized the joint variationin the yield, yield components, and quality traits under MS and SS. The total variance explained was 83.40% under control conditions (Fig. 6a), 78.20% under MS conditions (Fig. 6b) and 79.48% under SS conditions (Fig. 6c). Under MS conditions, PC1 was positively correlated with Y (0.956), AFW (0.948), and FD (0.920), and negatively correlated with TSS (-0.938), indicating a potential inverse relationship between high yield and TSS content. PC2 was associated with Hue (0.938), TA (-0.856), FL (0.705), and EC (-0.693), representing variation in quality and structural traits. Under these conditions, the ‘G x 2% VC’ treatment showed a strong positive association with Y, AFW, and FL. In particular, the narrow angle between Y and AFW supports the strong correlation between these two variables (Fig. 6b). Under SS conditions, PC1 was associated with FD (0.993), FL (0.972), pH (0.950), and AFW (0.843), while PC2 was associated with Hue (0.962), Y (0.748), EC (0.714), and Chroma (-0.624). ‘G x 2% VC’ treatment had strong positive association with Y and AFW; ungrafted and non-VC groups showed weak associations. Again, the narrow angle between Y and AFW highlights the strong positive correlation between these variables (Fig. 6c).

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

Evaluation of ‘VC x DS’ interaction in field conditions with correlation analysis (a) and polar heatmap including dendrogram (b). N-G: Non-Grafted. G: Grafted. VC: Vermicompost. Y: Yield. AFW: Average fruit weight. FD: Fruit diameter. FL: Fruit length. TSS: Total soluble solid. TA: Titratable acidity.

Fig. 6
Fig. 6The alternative text for this image may have been generated using AI.
Full size image

PCA of yield, yield components, and quality characteristics in eggplant under different drought conditions (control- a, mild stress- b, severe stress- c) in the field. N-G: Non-Grafted. G: Grafted. VC: Vermicompost. Y: Yield. AFW: Average fruit weight. FD: Fruit diameter. FL: Fruit length. TSS: Total soluble solids. TA: Titratable acidity.

Effects of treatments on some soil properties under field conditions

Analysis of variance for soil organic matter and physical properties in the field is presented in Table 5. The ‘G x VC x DS’ interaction was not significant for on soil organic matter or physical properties (p > 0.01). However, ‘VC x DS’ interaction significantly affected soil organic matter content (p < 0.01) (Fig. 4b). The highest organic matter content was 1.82% in the ‘2% VC x Control-100% irrigation’ combination, while the lowest organic matter content was 0.48% in the ‘0% VC x SS’ combination. VC treatments significantly improved bulk density, field capacity, available water capacity, hydraulic conductivity, aggregate stability, and porosity (p < 0.01), except for wilting point, structure stability index, and mean weighted diameter (Table 5). These changes demonstrated that even at the 2% dose, VC enhanced key soil physical properties relevant for water retention and root development under field conditions.

Table 5 Analysis of variance results of some properties of soils under field conditions and effect of vermicompost treatments on physical properties of soils under field conditions.
Full size table

Discussion

This study supports the hypothesis that applying VC to grafted eggplant plants under both mild and severe drought stress conditions provides significant improvements in yield, fruit quality, and soil properties. In particular, the combination of grafting and 2% VC exhibited stronger positive effects on morphological and biochemical parameters compared with either treatment alone. The 2% VC dose, identified as the most effective in greenhouse conditions, was also validated in the field, producing superior results in terms of yield and quality. Many traits-including yield, AA, TPM, TF, FDM, and TSS18-responded more favorably to this combined treatment, a pattern that likely reflects an association with improved soil physical properties and grafting-related root system characteristics under drought conditions10,19. Additionally, these results suggest that grafting together with VC may be associated with improved physiological performance under drought conditions .

The results from both greenhouse and field experiments further highlight that drought stress reduced total yield per plant, with MS causing less yield loss than SS. The integration of the grafting with 2% VC continuously mitigated the adverse effects of drought on plant performance. This treatment produced the highest improvements in fruit yield, FDM, TSS, TF, TPM, and AA in the greenhouse under both MS and SS, and a similar enhancement was evident in the field, especially under SS, where yield increased by nearly 144% relative to the severely drought-stressed, untreated control under the tested conditions. PCA consistently positioned the %2 VC-treated plants closest to yield-related vectors under both mild and severe stress conditions, suggesting a consistent association with yield-related traits under the tested conditions. Across the drought severity spectrum, PC1 primarily captured morphological and yield traits, while PC2 reflected biochemical and antioxidant responses, suggesting two distinct but related response patterns to drought stress. The enhanced vegetative and generative growth observed in grafted plants appears to be associated with the strong macro- and micronutrient uptake capacity of drought-tolerant rootstocks. The Köksal rootstock-a drought-tolerant Solanum incanum hybrid-likely facilitated greater water and nutrient acquisition under low water-potential conditions through its robust root system9. In compatible graft combinations, rootstock-derived hormones, metabolites, and signaling molecules can support the generative development of the scion during drought by improving vigor20. Incorporating VC into the growth medium may also have contributed to improved soil water retention and increased ion availability involved in osmotic adjustment. Soil water potential was not directly measured; therefore, improvements in soil physical properties should not be interpreted as direct evidence of enhanced plant water status. These structural improvements may have contributed to improved root-zone moisture conditions during drought. Additionally, VC application may be associated with improved root-zone conditions, supporting water and nutrient availability under drought stress. The presence of phytohormones such as cytokinins, IAA, gibberellins, and brassinosteroids in VC21 may also help explain the improved growth performance observed in grafted plants.Under limited irrigation, the accumulation of water and photosynthates in developing fruits is reduced23, and in our study both fruit weight and size declined accordingly. Although the ‘G x VC’ interaction moderated these reductions, it did not fully reverse them. This observation aligns with the negative yield–TSS relationship identified in the correlation analysis: drought decreases fruit water content, which elevates TSS, while simultaneously restricting yield through reduced cell expansion. Thus, higher TSS represents a physiological consequence of enhanced osmotic adjustment under drought. Although fruit DM and TSS increased linearly with drought severity, these increases were partially tempered by the ‘G x VC’ combination. While moderate drought may elevate DM and TSS without causing substantial yield penalties, even MS caused a notable (68%) yield decline in greenhouse conditions. This combined treatment reduced the severity of drought-related shifts in fruit composition while supporting overall productivity. Maintaining water uptake under stress would reduce the need for carbohydrate accumulation as osmolytes, thereby moderating fruit DM.Improvements in soil physical properties associated with VC indicate a more favourable moisture environment around the roots. Consequently, soluble sugar accumulation-which reflects osmotic adjustment processes involving glucose and fructose24-was reduced in ‘G x VC’ plants. The microbial and bioactive components of VC may have complemented these effects by influencing plant-water relationships and nutrient dynamics25. Microbial activity associated with VC can enhance rhizospheric nutrient mineralization, increasing nutrient accessibility and indirectly supporting plant water status18. These physiological adjustments are consistent with previous findings indicating that VC or microbial inoculants can reduce TSS levels under drought stress26,27.Phenolic compounds and flavonoids are key secondary metabolites that contribute both to fruit quality and to plant tolerance against drought-induced oxidative stress28. In the greenhouse experiment, antioxidant activity increased substantially in drought-stressed plants, consistent with elevated synthesis of total flavonoids and phenolics as stress severity intensified. The tight clustering of AA, TF and TPM in PCA biplots highlights their combined antioxidant functions, which intensify as drought severity increases. Increased ROS accumulation is known to trigger phenolic and flavonoid synthesis, and the PCA patterns support this mechanistic linkage by demonstrating interconnected biochemical responses. The ‘G x VC’ interaction further enhanced the accumulation of these antioxidant compounds. This combined effect suggests that biochemical processes (e.g., hormonal signaling and ROS regulation) and physical mechanisms (e.g., soil moisture buffering and microbial activity) in the root zone likely act simultaneously. While grafting improves water uptake efficiency via a more vigorous root system, VC provides complementary buffering of water and nutrient availability, collectively supporting antioxidant-related responses and strengthening antioxidant capacity under drought. These metabolites showed consistent convergence in PCA space and supported their shared roles in antioxidant defence caused by drought. The nearly 144% increase in phenolic content recorded under SS with 2% VC, relative to the severely drought-stressed untreated control, suggests an association with enhanced antioxidant-related responses under drought conditions. Although 3% VC improved some traits under control conditions, its effectiveness under drought was more limited. Previous studies similarly indicate that microbial inoculants can stimulate phenolic and flavonoid synthesis under drought29. Increased nutrient availability resulting from VC-associated microbial mineralization11,30,31-as well as processes such as PAL activation32 or enhanced precursor availability35-may also contribute to these responses, findings consistent with reports in cowpea16 and studies identifying phenolic compounds in VC leachate33,34.

In the field experiment, fruit EC and TA increased markedly under drought stress-particularly in ungrafted plants without VC and most strongly under SS. Similar responses have been reported in watermelon and eggplant under drought conditions9,36. The parallel increase in EC and TA, together with their close clustering in the PCA-especially under MS (Fig. 6b)-suggests that both parameters reflect enhanced mineral ion concentration associated with osmotic adjustment. This interpretation aligns with earlier findings showing that drought and salt stress elevate K, Na, and Cl levels in fruit tissues37. Conversely, the ‘G x VC’ combination reduced EC and TA, likely due to improved water uptake and the resulting dilution of solute concentrations. This moderating effect is consistent with the functional balance grafted plants maintain between root and shoot hydraulic systems38. VC may further contribute by enhancing soil physical structure, thereby increasing water availability, improving water-use efficiency, and indirectly supporting fruit water content10,39. These combined physiological and soil-mediated effects help explain the lower EC and TA values observed in grafted plants receiving VC under drought.

Hue values of the fruit increased with drought, varying by severity. This may be linked to anthocyanin synthesis, a flavonoid compound40. These compounds help plants cope with stress by scavenging free radicals, chelating ions, and modulating signaling pathways41. VC has been reported to enhance the accumulation of such phenolic compounds30,42. The joint effects of grafting and VC likely stimulated anthocyanin accumulation. Particularly under SS, Hue values rose substantially. This may be due to increased synthesis of phytohormones like ABA and ethylene under drought43,44, which are known to upregulate anthocyanin-activating genes45. ABA produced in the rootstock may contribute to drought tolerance by regulating stomatal movement and gene expression46, and VC has been shown to contain ABA33.

Such physiological responses are further supported by the relationships between morphological and quality traits revealed through correlation and cluster analyses conducted under both greenhouse and field conditions. In the greenhouse experiment, the correlation analysis and polar heat map highlighted biologically meaningful associations between yield and quality parameters. The clustering of yield, FD, and AFW with TSS, TPM, and FDM in the dendrogram suggests a coordinated response, indicating that morphological and biochemical traits are functionally interconnected. Moreover, the distinct cluster formed by AA, TPM, and TF-particularly in grafted and vermicompost-treated plants under SS-points to a shared antioxidant regulation mechanism contributing to enhanced stress tolerance. The tight clustering of these antioxidant-related traits in grafted and VC-treated plants under SS indicates that osmotic adjustment and ROS-scavenging capacity are simultaneously and synergistically enhanced, consistent with the observed improvements in drought tolerance. Together, these patterns point to an integrated morphological-biochemical response framework shaped by grafting and VC. Multivariate clustering clearly distinguished the ‘G x VC’ treatments from the ungrafted plants, indicating different physiological strategies under drought conditions. Yield-related traits showed a consistent trend in PCA, consistently aligning with the same direction as the ‘G x VC’ treatments and differing from the control groups under drought stress. Yield in various environments was closely related to fruit size characteristics; antioxidant markers, on the other hand, clustered under stress, reflecting two complementary dimensions of the drought response. This supports the interpretation that morphological and biochemical traits jointly contribute to drought tolerance, and that grafting combined with VC exerts a coordinated regulatory influence on this adaptive response. This study demonstrates the yield- and quality-enhancing effects of 2% VC on grafted plants under drought, but some limitations exist. First, it is uncertain whether the 2% VC dose will perform similarly across different soils and cultivars. Second, vermicompost quality may vary between batches or suppliers, affecting reproducibility. Root architecture was not directly measured in this study, limiting direct conclusions regarding the role of root system traits in drought tolerance. Finally, the results are based on a single season; multi-season validation is needed to confirm reliability.

The cluster analyses revealed that grafting combined with VC induced strong and similar stress-mitigation responses, whereas control and ungrafted plants exhibited weaker and more closely related stress profiles. The heatmap further confirmed that irrigation levels and treatments generated distinct and clearly separated physiological response patterns. However, another critical factor influencing the effectiveness of these biological responses under stress conditions is the supportive environment provided by the soil structure in which the plant grows. In this context, the contribution of soil physicochemical properties to plant–water relations and overall stress tolerance is equally important as the plant’s physiological responses and should be regarded as an integral component of plant adaptation to drought. VC application was associated with increased organic matter content and improved soil physical properties under both greenhouse and field conditions. Similar results have been reported, where increasing VC doses raised soil organic matter compared to control treatments10,47. Numerous studies confirm that compost and plant waste additions increase soil organic matter48. Organic matter is crucial for water retention, soil stability, and chelation. The increase in VC levels led to gradual improvements in various important physical properties of the soil, thereby improving root zone conditions.This may be due to VC-induced soil aggregation, which improves structure via its porous nature and water-holding capacity. Previous research corroborates these improvements10,11,47,48. Soil aggregation may support root development, gas exchange, and nutrient availability and is commonly associated with organic matter dynamics in soils Soil aggregation, which enhances root development, gas exchange, and nutrient availability, results primarily from microbial decomposition of organic matter49. Soil aggregates influence pore continuity, infiltration, and water retention50. Microaggregates, formed from primary particles, benefit from root hair proliferation in grafted plants. Enhanced aggregation improves hydraulic conductivity, respiration, and overall plant development51. Leonardite and VC have been shown to significantly improve aggregate stability52. Increased organic matter correlates with higher aggregate stability and lower bulk density, supporting root growth and boosting yield. Many researchers have reported that there are increases in aggregate stability values depending on the increase in the organic matter content of soils49. While the increase in total porosity due to the increase in organic matter content in soils decreases the volume weight values, the products released as a result of the mineralization of organic matter cause an increase in aggregate stability and average weighted diameter values. The decrease in the volume weight values of the soils likely facilitated the formation of macropores suitable for root development, which may have contributed to higher yield levels.

This study clearly demonstrates the role of vermicompost in regulating soil properties and its significant effects on plant growth and metabolic activity under drought stress in eggplant. The negative effects of drought stress are not limited to vegetable production; they also constrain the cultivation of ornamental plants. Moreover, as stated by Karagüzel53, research in this area is considerably less extensive for ornamental plants than for vegetables. Therefore, it is important to evaluate the potential contributions of vermicompost within ornamental plant production systems as well. Future studies should address this knowledge gap and assess the applicability of vermicompost across different production systems and environmental conditions.

Materials and methods

Materials used in experiments

The study was conducted and evaluated as two independent types of research: greenhouse and field studies. Seedlings grafted with ‘Aydın Siyahı’ on the rootstock ‘Köksal’ were used as plant material in both stages of the research. Non-grafted ‘Aydın Siyahı’ seedlings grown on their own roots constituted the control plants. Grafted seedlings were produced by Antalya Tarım Hybrid Seeds Company (Antalya, Türkiye). VC used as organic material was obtained from Ekosol Farming & Livestock Company (Manisa, Türkiye).

Greenhouse experiment

Experiment design and growing conditions

The experiment was carried out in a glass greenhouse under controlled conditions (temperature: 25–27 °C, humidity: 50–55%) within the Soil Fertilizer and Water Resources Central Research Institute (Ankara, Türkiye) (April 1-September 1, 2020). The factors in the research, which was designed according to the three-factor factorial experimental design in a random plots trial model, are as follows: Factor 1: grafted and non-grafted plants, Factor 2: vermicompost (VC) application (0, 1, 2, and 3%), and Factor 3: water deficit levels (full irrigation, 70%, and 30% irrigation mild and severe drought stress).

In the greenhouse experiment, soil brought from Antalya land, where the second phase of the experiment was established, was used. These soils were dried in air, passed through a 4 mm fine sieve and then transferred to 35 l PE pots (height: 39, diameter: 35 cm). VC in solid form was applied to a depth of approximately 10 cm, 1, 2, and 3% (w/w) of the weight of the potting soil, 15 days before planting the seedlings. At the end of this period, grafted and non-grafted seedlings with 4–5 true leaves were transplanted into pots (April 27, 2020), with one plant per pot. According to the soil and VC analysis results (Supplementary Table 1), 100 kg ha− 1 P2O5, triple super phosphate (TSP), and 70 kg ha− 1 N were given to each pot before planting seedlings. 30 kg ha− 1 N was added to each pot during flowering and harvest.

Water deficit conditions

Stress application started 15 days after planting seedlings (May 11, 2020). Until this time, the humidity in the pots was ensured to be at the FC (field capacity or pot capacity) level. To determine FC; two randomly selected pots from each subject were weighed, placed in basins filled with water, covered, and allowed to reach a saturation point by absorbing the water (24 h). At the end of this period, the pots were kept until no water drained from their bottoms and the FC values were determined by weighing them. While control (non-stressed) plants were irrigated at FC (full irrigation, 100%) level during stress, a MS was created with 70% of the water given to the control, and a SS was created with 30%. At the end of the stress treatments, all plants were examined for some morphophysiological characteristics. Stress treatments lasted 110 days, and at the end of this period, the plants were evaluated in terms of yield and quality characteristics.

Yield, yield components, and some fruit quality characteristics

The fruits collected from the first harvest to the last harvest date of the plants were weighed and the yield was calculated by adding the values ​​obtained (kg plant− 1). The fruit weight was calculated by dividing the weight of all fruit harvested from each application by the number of fruits. The fruits were harvested 5 weeks after the flowering date and their measurements were taken. The harvested fruits diameters were measured in millimetres at their midpoints using a digital calliper. Chopped fruit samples were weighed on a precision scale to determine their fresh weight. After drying in an oven set at 65 °C until they reached a constant weight, their dry weights were recorded by weighing. The % fruit dry matter (FDM) was calculated based on the fresh and dry weights. A few drops of fruit juice were analysed using a digital handheld refractometer (Hanna HI 96801) and the results were expressed as a percentage to determine the total soluble solids content (TSS-°Brix). Total phenolic content was determined spectrophotometrically at 750 nm after a 2-hour incubation period, during which Folin-Ciocalteu reagent, distilled water, and sodium carbonate were added to the pureed fruit samples54. The results were expressed as gallic acid equivalents (µg GAE g⁻¹). Antioxidant activity was calculated as Trolox equivalent (TEAC) according to the decrease in absorbance of the ABTS.+ radical measured at 734 nm55. To determine total flavonoids, eggplant extracts were reacted with AlCl₃ and measured at 415 nm after 10 min. The results were expressed as quercetin equivalents (mg QE g⁻¹)56. In the trial, biochemical measurement and analyses were conducted using three biological replicates per treatment, each replicate consisting of pooled samples from three.

Field experiment

Experimental area, experimental setup, and deficit application of water

The second phase of the research was carried out in the trial area of Antalya Tarım Hybrid Seeds Company operating in Türkiye (Altitude: 45, 36 96’ North latitude and 30 81’ East longitude) (April- August 2021). In the study, climate data for 2021 and long-term averages for the April-August periods are given in Supplementary Table 2. Other physicochemical soil properties and some properties of the irrigation water used are presented in Supplementary Table 3.

Experimental design, planting, and treatments

The experiment followed a split-plot design with three replications. Based on the findings from the greenhouse experiment, 2% vermicompost (VC)-identified as the most effective dose-was applied in the field trial. Yield and quality traits of grafted (on Köksal rootstock) and non-grafted eggplants were evaluated under three irrigation levels (SS = 0.30 Epan (SS: Severe drought stress), MS = 0.70 Epan (MS: Mild drought stress) and Control = 1.00 Epan) and two vermicompost (VC) treatments (0% and 2%). Based on previous greenhouse results, VC was incorporated into the soil (15 cm depth) along planting rows 15 days before transplanting (April 5, 2021). Irrigation was based on Class A pan evaporation data (recorded weekly), with 5.0 × 10.0 m plots. Fertilization was done before planting, according to soil analysis. Grafted and non-grafted seedlings were planted on April 20, 2021, at 0.50 × 1.20 m spacing (120 plants per plot). Deficit irrigation began on May 31 and ended on August 12, 2021.

Irrigation method and management

A surface drip irrigation system was used, with 16 mm PE drip lines along plant rows. Drippers (2 l h⁻¹, 40 cm spacing) delivered water based on weekly Class A pan evaporation data57. The irrigation amount for Control (1.00 Epan) was calculated using:

\({\text{I}}\,=\,{\text{A}} \times {\text{Epan}} \times {\text{Kpc}} \times {\text{P}},\)

where I = irrigation volume (L), A = plot area (m²), Epan = weekly pan evaporation (mm), Kpc = crop-pan coefficient (0.30–1.00 for different treatments), and P = wetted area percentage.

Initially, P was set at 35%58. due to low canopy coverage; actual percent cover values were used once they exceeded 35%, calculated from the crown width of five pre-selected plants divided by row spacing (1.20 m)59.

Evapotranspiration, water productivity, and irrigation water productivity, and yield response factor (Ky)

Evapotranspiration (ET) was estimated using the soil water balance method60.

\({\text{ET}}\,=\,{\text{I}}\,+\,{\text{R}}\, \pm \,{\text{DW }} - {\text{ D}}\)

where ET = evapotranspiration (mm), I = irrigation (mm), R = rainfall (mm), ΔW = change in soil moisture (mm), and D = deep percolation, assumed negligible due to controlled irrigation.

Water productivity (WP) and irrigation water productivity (IWP) were calculated61.

\({\text{WP }}={\text{ }}\left( {{\text{Y }}/{\text{ ET}}} \right){\text{ }} \times {\text{ 1}}00\)

\({\text{IWP }}={\text{ }}\left( {{\text{Y }}/{\text{ I}}} \right){\text{ }} \times {\text{ 1}}00\)

WP: Water productivity (kg m− 3); IWP: Irrigation water productivity (kg m− 3); ET: Evapotranspiration (mm); I: Applied irrigation water (mm); and Y: Yield (kg ha− 1).

The yield response factor (Ky) was determined per Stewart et al.62, showing yield sensitivity to water deficit:

\(({\text{1}}\, - \,{\text{Ya}}/{\text{Yx}})\,=\,{\text{Ky}} \times ({\text{1}}\, - \,{\text{ETa}}/{\text{ETx}}),\)

where Ya = actual yield, Yx = maximum yield, ETa = actual ET, ETx = maximum ET.

Yield, yield components, and some quality characteristics

Total yield (tonnes per hectare) was calculated as the cumulative weight of fruit collected between the first and last harvests in each plot. Fruit weight (grams per fruit) was determined as the total fruit weight divided by the number of fruits. Diameter (mm) was measured using digital callipers and length (cm) using a tape measure on 15 fruits randomly selected from each plot. Electrical conductivity (EC) was measured using a Milwaukee MW170 MAX, and pH was measured using a WTW Inolab 7310. Total soluble solids (TSS) were measured using a handheld digital refractometer (Hanna HI 96801).TA values were calculated by adding 10 ml of distilled water to 5 ml of filtrate and titrating with 0.1 N NaOH until the pH reached 8.0163. Fruit colour (L*, a*, b*) was determined using a Minolta CR-400 colourimeter and hue (°) and chroma were calculated from a* and b* values64. Each plot was considered as one biological replicate, and 15 fruits were randomly sampled from each replicate for morphological measurements.

Some soil properties examined in greenhouse and field experiments

The air-dried samples were passed through 4- and 2-mm sieves prior to analysis. Soil reaction (pH) was measured using a pH meter on a 1:1 soil-to-water suspension. Lime content was determined using a Scheibler calcimeter65, while soil electrical conductivity (EC at 25 °C) was identified using an EC meter on the same suspension66. The modified Walkley–Black method was employed to determine soil organic matter (SOM) content66. Soil particle size distribution (i.e. sand, silt and clay content) was analysed using a hydrometer test67. The soil moisture content at field capacity (FC) and permanent wilting point (PWP) was determined using a pressure plate apparatus68. Available water content (AWC) was then calculated as the difference between FC and PWP. Soil bulk density (BD) and total soil porosity was determined as described by Tüzüner69.

The structural stability index (SSI) was calculated using hydrometer data70. Aggregate stability (AS) was determined using a wet-sieving apparatus71. The saturated hydraulic conductivity of the samples was calculated using a fixed-head permeameter, which is typically used for fine-grained soils72. The average weight diameter was calculated using dry sieving data (sieves of 2.00, 1.00, 0.50, 0.25, 0.106 and 0.053 mm)68.

Data analysis

Numerical data were subjected to variance analysis (ANOVA) using the Statistical Package for the Social Sciences (SPSS) version 11.0 (IBM Corp., Armonk, NY, USA), and statistically significant differences between treatments were identified. Normality (tested by the Shapiro-Wilk test) and homogeneity of variances (tested by Levene’s test) were verified prior to conducting ANOVA. Differences were expressed using letters at the 0.05 significance level based on Duncan’s multiple comparison test. OriginPro 2024 (OriginLab Corporation, Northampton, MA, USA) software was used to visualize correlations between measured variables and to generate a polar heat map with a dendrogram through hierarchical cluster analysis. Relationships among the examined features were analyzed using Principal Component Analysis (PCA) with the XLSTAT (Addinsoft, Paris, France) add-in for Excel.

Conclusion

This study highlights the synergistic effects of VC application and grafting onto tolerant rootstocks in improving yield and quality of eggplants under varying drought levels in both greenhouse and field conditions. The combined treatment was particularly effective under SS, enhancing phenolic content, antioxidant activity, and flavonoid levels in the greenhouse, while increasing fruit weight and size and reducing EC and TA in the field. VC also improved key soil properties, including water retention, infiltration, root development, and aggregate stability, while reducing bulk density. These findings suggest that VC and grafting can substantially enhance drought tolerance, yield, and soil health, especially in arid and semi-arid regions. However, translating these findings to the farmer scale requires careful evaluation of several practical limitations. The cost of VC and grafted seedling applications may constrain adoption, particularly among small-scale producers. Additionally, variability in VC quality between batches and the large quantities required for field application raise uncertainties regarding the long-term sustainability of the 2% VC dose. Therefore, although the biological effectiveness of the VC × grafting combination is evident, future research should incorporate multi-year trials across diverse agro-ecological zones, detailed cost–benefit assessments, and farmer-centered evaluations. Such efforts are essential for determining whether these practices can be feasibly integrated into commercial production systems as sustainable and scalable drought management strategies.