Physiological determinants of drought tolerance in groundnut (Arachis hypogaea L.): insights from multi-generational responses

Introduction

Groundnut (Arachis hypogaea L.) is a vital food and oilseed crop valued for its high oil (47-53%) and protein (25-36%) content. As a self-pollinated tetraploid (2n = 4x = 40) with a unique geocarpic fruiting habit, it adapts well to diverse agro-ecological regions and serves as a major source of edible oil, industrial raw material and nutrient-rich fodder. Its short duration and broad adaptability also support rapid crop improvement and multi-season cultivation. Globally, groundnut ranks as the fifth most cultivated oilseed crop, covering 27.9 million hectares with an annual production of 47 million tonnes.

India is the second-largest producer, with Gujarat, Rajasthan, Madhya Pradesh, Karnataka, Tamil Nadu, Andhra Pradesh and Maharashtra contributing significantly. Gujarat is leading state in area (169.53 lakh ha) followed by Rajasthan (89.06), during the year 2023-24 (Ministry of Agriculture and Farmers Welfare, Govt. of India). Global groundnut agriculture covers 4.96 million hectares and produces 10.03 million tons with a productivity of 1616 kg/ha (FAOSTAT, 2023) [3] . However, nearly 80% of the crop is grown under rainfed conditions, making productivity highly vulnerable to erratic rainfall. With global vegetable oil demand expected to double by 2040, improving groundnut performance under water limited environments is increasingly important. Drought is the most severe abiotic constraint reducing groundnut productivity, adversely affecting physiological processes such as photosynthesis, stomatal regulation, transpiration efficiency and water relations. These disruptions ultimately impair flowering, peg penetration and pod filling. Conventional selection for yield under drought is often unreliable due to strong genotype × environment interactions. Therefore, a deeper understanding of the physiological mechanisms governing drought tolerance is essential for developing resilient cultivars. Multigenerational evaluation provides a powerful approach to understand trait stability and the genetic control of key physiological responses under moisture stress. The present study focuses on physiological determinants of drought tolerance across generations to elucidate adaptive responses and identify traits with potential for use in breeding waterefficient, drought-resilient groundnut genotypes.

Materials and Methods

The present investigation was conducted at the field of AICRP on summer groundnut, Mahatma Phule Krishi Vidyapeeth, Rahuri, Ahmednagar (MS) during Kharif-2023, Summer-2024 and Kharif-2024. Six generations (P 1 , P 2 , F 1 , F 2 , BC 1 and BC 2 ) were developed from cross TAG-24 × ICG 4670 for drought tolerance and pod yield. The experimental material consisted drought tolerant parent (ICG 4670) as donor and recipient parent (TAG-24) is high yielding and wider adaptability of groundnut genotypes. The cross TAG-24 × ICG 4670 was effected in Kharif-2023 to produce the F 1 seeds and breeding material like F 1 , F 2 , BC 1 and BC 2 and advancement was attempted in Summer-2024. During Kharif-2024, final evaluation trial conducted for study of inheritance of drought tolerance related traits and generation mean analysis, for physiological traits under both irrigated and drought (rainout shelter) condition. The characters viz., SPAD chlorophyll meter reading, water-use efficiency (WUE), specific leaf area (SLA), number of stomata and stomatal conductance were analysed for this study. Each character was analysed separately and the ANOVA was constructed as per Panse and Sukhatme (1995) [13] . The data was analysed for mean performance of the generations. Gene action has been worked by generation mean analysis as per six parameter model given by (Hayman, 1958) [4] . Adequacy of additive dominance effect was detected by individual scaling test; three tests of scale were carried out to detect presence or absence of gene interaction by using formulae given by Mather (1949) [10] and Hayman and Mather (1957) [5] . Cavalli's (1952) [2] joint scaling test was applied to test the adequacy of additive dominance model. Whenever, the model was found inadequate, Hayman (1958) [4] six parameter model was used to estimate the different gene effects. Statistical analysis of data was done by using software package INDOSTAT.

Results and Discussion

Analysis of Variances

The analysis of variance for five characters is presented in Table 1 for two crosses under both irrigated and drought (rainout shelter) condition. Under irrigated condition (Table 1), the analysis of variance revealed significant differences among treatments in cross TAG-24 × ICG 4670 i.e., parents and their derived generations (F 1 , F 2 , BC 1 and BC 2 ) for majority of traits. This indicates the presence of substantial genetic variability among the parents and their progenies. However, trait such as water use efficiency exhibited non-significant differences, suggesting limited genetic variation for this character. Under drought condition (Table 1), the analysis of variance showed significant differences among treatments in cross TAG-24 × ICG 4670 i.e., parents and their respective generations (F 1 , F 2 , BC 1 and BC 2 ) for all traits studied the, indicating considerable genetic variability under moisture stress.

Mean performance of parents and generations

The mean performance of parents and their crosses for physilogical traits in groundnut are presented in Table 2.1 and 2.2 for irrigated and drought condition respectively. Higher values are desirable for all traits under study except for days to specific leaf area for which lower values are preferred. The trait wise results are discussed as below:

SPAD chlorophyll meter reading

Under irrigated condition, SPAD chlorophyll meter readings ranged from (45.44) to (49.36) across the parental genotypes and their derived generations. The BC 2 generation recorded the highest chlorophyll content (47.39), followed by F 1 (46.67), F 2 (46.54) and BC 1 (46.19). Among the parents, TAG-24 exhibited a SPAD value of (45.44), whereas ICG 4670 had a slightly higher reading of (49.36). The elevated SPAD values in BC 2 and F 1 indicate enhanced chlorophyll retention and potentially higher photosynthetic efficiency under optimal conditions. Under drought condition, SPAD chlorophyll meter readings declined across all generations except P 2 , with values ranging from 37.10 to 49.10. The BC 2 generation recorded the highest SPAD value (45.98), followed by F 2 (44.59), F 1 (42.71) and BC 1 (38.55). Among the parents, TAG-24 and ICG 4670 recorded (37.10) and (49.10) respectively. The reduction in SPAD readings under drought reflects a decrease in chlorophyll content due to stress induced degradation. However, the relatively higher values in BC 2 and F 2 suggest better chlorophyll retention and photosynthetic resilience highlighting their potential for use in drought tolerant breeding programs.

Under irrigated conditions, water use efficiency (WUE) among the parental genotypes and derived generations varied from (4.29) to (6.62). The highest WUE was observed in F 2 (5.40), followed by F 1 (5.13), BC 2 (5.13) and BC 1 (4.84). The parental lines TAG-24 and ICG 4670 recorded WUE values of (4.29) and (6.62) respectively. Under drought stress, water use efficiency (WUE) exhibited noticeable variation across generations, ranging from (2.56) to (6.17). The highest WUE was recorded in the BC 2 generation (6.17), followed by F 2 (4.65), F 1 (4.24) and BC 1 (3.74). Among the parental genotypes, TAG-24 and ICG 4670 showed WUE values of (2.56) and (6.02) respectively. Specific leaf area (cm 2 /g) Under irrigated condition, specific leaf area (SLA) values ranged from (168.73) to (197.83) cm 2 /g across the parental lines and their derived generations. The BC 1 generation exhibited the highest SLA (191.96 cm 2 /g), followed closely by F 1 (190.00 cm 2 /g), F 2 (185.11 cm 2 /g) and BC 2 (185.05 cm 2 /g). Among the parents, TAG-24 recorded a considerably higher SLA (197.83 cm 2 /g) compared to ICG 4670 (168.73 cm 2 /g). Higher SLA under irrigated conditions is generally associated with thinner leaves and rapid growth contributing to increased light interception and photosynthetic surface area. The elevated SLA in BC 1 and F 1 indicates vigorous vegetative development, which could be advantageous for biomass accumulation under optimal moisture conditions. Under drought stress, specific leaf area (SLA) values increased across all generations and parental lines except BC 2 , ranging from (170.34) to (214.97 cm 2 /g). The BC 2 generation exhibited the lowest SLA (170.34 cm 2 /g), followed by F 2 (191.37 cm 2 /g) indicating thicker and more compact leaves as a typical adaptive trait under water stress. In contrast, BC 1 (202.02 cm 2 /g) and F 1 (194.37 cm 2 /g) maintained relatively higher SLA values suggesting a less conservative water-use strategy. Among the parents, TAG-24 recorded an SLA of (241.97 cm 2 /g), whereas ICG 4670 showed a lower value of (173.31 cm 2 /g). The reduction in SLA under drought is indicative of structural adjustments aimed at reducing transpirational water loss and enhancing drought tolerance. Generations like BC 2 and F 2 with lower SLA values may possess better physiological resilience under limited water availability.

Number of stomata

Under irrigated condition, the number of stomata varied from (12.91) to (20.36) among the different generations and parental lines. The BC 2 generation recorded the highest stomatal density (17.68), followed by F 2 (17.40), F 1 (16.38) and BC 1 (13.81). Among the parents, ICG 4670 exhibited a higher number of stomata (20.36) compared to TAG-24 (12.91). A greater number of stomata under irrigated conditions reflects the enhanced gas exchange capacity, facilitating efficient photosynthesis and transpiration. The elevated stomatal numbers in BC 2 and F 2 suggest their potential for high photosynthetic efficiency and productivity when water is not a limiting factor. Under drought stress, the number of stomata showed a increase across generations, ranging from (14.46) to (20.51). The BC 2 generation maintained the highest stomatal count (19.86), followed by F 2 (19.70), F 1 (16.58) and BC 1 (15.73). Among the parents, ICG 4670 had a higher stomatal number (20.51) than TAG-24 (14.46). The relatively higher stomatal numbers retained in BC 2 and F 2 imply a balanced strategy, possibly enabling sustained gas exchange and moderate water conservation, thereby contributing to their better performance under moisture deficit conditions.

Under irrigated condition, stomatal conductance among the parental lines and their respective generations ranged from (0.249) to (0.587 μmoles/m 2 /s). The BC 2 generation exhibited the highest conductance (0.453 μmoles/m 2 /s), followed closely by F 1 (0.414), F 2 (0.345) and BC 1 (0.317). Among the parents, ICG 4670 recorded a conductance of (0.587 μmoles/m 2 /s), while TAG-24 had the lowest value (0.249 μmoles/m 2 /s). Higher stomatal conductance under well-watered conditions indicates efficient gas exchange and transpiration, supporting optimal photosynthesis and growth. Generations such as BC 2 and F 1 with elevated stomatal conductance reflect superior physiological performance under non stress condition and are potential candidates for maximizing productivity. Under drought condition, stomatal conductance values increased across all generations except P 1 , ranging from (0.233) to (0.640 μmoles/m 2 /s). The BC 2 generation maintained the highest conductance (0.529 μmoles/m 2 /s), followed by F 1 (0.418), F 2 (0.416) and BC 1 (0.328). Among the parents, ICG 4670 showed a higher value (0.640 μmoles/m 2 /s) compared to TAG-24 (0.233 μmoles/m 2 /s).

Reduction in stomatal conductance of P 1 reflects a typical drought induced response aimed at reducing water loss through transpiration. Notably, the ability of BC 2 and F 1 to sustain relatively higher conductance even under moisture stress suggests enhanced stomatal regulation and water use dynamics making them promising for drought tolerance breeding programs.

Estimates of scaling tests for detecting non-allelic interactions

The results on scaling test (A, B, C and D) in respect of physiological and drought tolerance related traits have been tabulated in (Table 3) under irrigated and drought stress (rainout shelter) condition. The result for cross: TAG-24 × ICG 4670 is mentioned under two conditions as follow:

Under irrigated condition (Table 3) SPAD chlorophyll meter reading For SPAD chlorophyll meter reading trait, all joint scaling tests were found non-significant showing additive dominance model was adequate to explain all the genetic variations.

Water use efficiency

For water use efficiency, scaling test 'B' (-1.48) was negatively significant while scaling test 'D' (0.83) was found positively significant. The joint scaling test was found highly significant for water use efficiency in the cross TAG-24 × ICG 4670 which indicates inadequacy of additive dominance model to explain all the genetic variation.

Specific leaf area

All joint scaling tests were found non-significant showing additive dominance model was adequate to explain all the genetic variations as the chi-square values of both the crosses were non-significant.

Number of stomata

In case of number of stomata, the scaling test 'B' (3.64) and 'D' (3.31) was positively significant. Joint scaling test was found highly significant which indicates the presence of epistasis.

Stomatal Conductance

Scaling test 'B' (-0.09), 'C' (-0.28) and 'D' (-0.08) was negatively significant in the cross TAG-24 × ICG 4670 for stomatal conductance. The joint scaling test for stomatal conductance was found significant, indicating inadequacy of additive dominance model to explain all the genetic variation.

Under drought condition (Table 3) SPAD chlorophyll meter reading

For SPAD chlorophyll meter reading, scaling tests 'C' (6.74) and 'D' (4.46) were significant, which indicates additive dominance model was inadequate to explain all the genetic variations.

Water use efficiency

For water use efficiency, scaling tests 'A' (0.67), 'B' (2.27) and 'C' (1.53) were positively significant and 'D' (-0.70) was negatively significant, showing additive dominance model was inadequate to explain all the genetic variations as the chi-square values of both the crosses were significant.

Specific leaf area

In case of specific leaf area, scaling tests 'A' (-32.31) and 'B' (-27.00) were negatively significant, showing additive dominance model was inadequate to explain all the genetic variations as the chi-square values of both the crosses were significant.

Number of stomata

For number of stomata, scaling test 'B' (2.63), 'C' (10.65) and 'D' (3.79) were found positively significant, which indicates inadequacy of additive dominance model to explain all the genetic variation.

Stomatal Conductance

Scaling tests 'C' (-0.042) and 'D' (-0.023) was negatively significant. The joint scaling test for stomatal conductance was found significant, indicating inadequacy of additive dominance model to explain all the genetic variation.

Estimates of gene effects for yield and yield contributing characters

The estimates of m (mean), major genetic effects additive [d] and dominance [h] and non-allelic gene interactions (i, j and l) based on six parameter model (Hayman, 1958) [4] for physiological traits for cross: TAG-24 × ICG 4670 (Table 4 and Table 5) under both irrigated and drought stress condition respectively. The parameter [m] was significant for all the characters under study.

Cross: TAG-24 × ICG 4670

The gene effects estimated by using perfect fit model in respect of traits associated with pod yield in groundnut has been presented in Table 4 and Table 5 for irrigated and drought condition respectively.

SPAD chlorophyll meter reading

Under irrigated condition, from estimates of genetic parameters, it was observed that significant mean (m) (46.39) effect was observed, but additive (d) and dominance (h) effects were non-significant, indicating an absence of non-allelic interactions for this trait. Under drought stress, from estimates of genetics parameters, it was observed that 'd' (-7.43) and 'h' (-9.68) was negatively significant. The interaction component 'i' (-9.28) was observed negatively significant, while 'l' (11.82) was estimated positively significant. Opposite sign observed for genetic component dominance 'h' and dominance x dominance 'l', revealing epistasis was predominantly of duplicate type for SPAD chlorophyll meter reading (SCMR). The duplicate type of epistasis was earlier noticed by Upadhyaya et al., (2011) [17] and Lal et al., (2005) [8] reported preponderance of additive gene action in the inheritance of SCMR.

Water use efficiency

Under irrigated condition, from estimates of genetic parameters, it was observed that 'h' (-2.01) was negatively significant. The interaction component 'i' (-1.67) was observed negatively significant, while 'l' (2.90) was estimated positively significant. Opposite sign observed for genetic component dominance 'h' and dominance x dominance 'l', revealing epistasis was predominantly of duplicate type for water use efficiency. Under drought stress, from estimates of genetic parameters, it was observed that 'd' (-2.52) was negatively significant and 'h' (1.36) was positively significant. The interaction component 'i' (1.41) was observed positively significant, while 'l' (-4.36) was estimated negatively significant.

Opposite sign observed for genetic component dominance 'h' and dominance x dominance 'l', with presence of duplicate epistasis for water use efficiency. Interestingly, John et al., (2012) [7] has found significant role of nonadditive gene actions for this trait.

Specific leaf area

Under irrigated condition, significant mean (m) (169.69) effect with high magnitude were observed. Additive (d) (14.54) effect was positively significant, but non-allelic interactions were absent, suggesting the preponderance of additive gene effects for specific leaf area. Under drought condition, from estimates of genetic parameters, it was observed that 'd' (31.67) was positively significant and 'h' (-34.03) was negatively significant. The interaction component 'l' (80.07) was estimated positively significant. Opposite sign observed for genetic component dominance 'h' and dominance x dominance 'l', revealing epistasis was predominantly of duplicate type for specific leaf area. John et al., (2012) [7] found predominant role of non-additive gene actions for this trait which is parallel with present finding. Meanwhile, based upon generation mean analysis Suriharn et al., (2005) [16] and Nigam et al., (2001) [12] reported predominance of additive gene effect in the control of SLA in groundnut. Similar findings were noticed by Jayalakshmi et al., (1999) [6] and Lal et al., (2005) [8] in diallel analysis. Interestingly, role of additive and nonadditive gene action for SLA were earlier reported by Venkateswaralu et al., (2007) [18] and Upadhyaya et al., (2011) [17] .

Number of stomata

Under irrigated condition, from estimates of genetic parameters, it was observed that 'd' (-3.86) and 'h' (-6.85) were negatively significant. The interaction component 'i' (-6.63) was observed negatively significant, however 'l' (9.62) was estimated positively significant. Opposite sign observed for genetic component dominance 'h' and dominance x dominance 'l', with presence of duplicate epistasis. Under drought stress, from estimates of genetic parameters, it was observed that 'd' (-4.13) and 'h' (-8.50) were negatively significant. The interaction component 'i' (-7.59) was observed negatively significant, however 'l' (4.54) was estimated positively significant. Opposite sign observed for genetic component dominance 'h' and dominance x dominance 'l', with presence of duplicate epistasis. Role of both additive and non-additive gene effects were significantly ascertained by John et al., (2012) [7] , Pitaloka et al., (2022) [14] . While, Ramya et al., (2021) [15] reported importance of non-additive gene action for this trait in wheat.

Stomatal Conductance

Under irrigated condition, from estimates of genetic parameters, it was observed that 'd' (-0.135) was negatively significant while 'h' (0.156) was positively significant. The interaction component, only 'i' (0.161) was observed positively significant, suggesting early selection benefits for this trait. Under drought stress, from estimates of genetic parameters, it was observed that 'd' (-0.201) was negatively significant and 'h' (0.028) was positively significant. The interaction component only 'l' (-0.051) was estimated negatively significant. Opposite sign observed for genetic component dominance 'h' and dominance x dominance 'l', with presence of duplicate epistasis for stomatal conductance. John et al., (2012) [7] reported predominance of both additive and nonadditive gene action for stomatal conductance. However, Ramya et al., (2021) [15] reported importance of non-additive gene action for this trait in wheat.

Conclusion

The present investigation demonstrated that drought stress significantly alters key physiological processes in groundnut, influencing chlorophyll stability, leaf structural efficiency and stomatal behaviour. The evaluation of five major physiological traits viz., SPAD chlorophyll meter reading, water-use efficiency (WUE), specific leaf area (SLA), number of stomata and stomatal conductance, across six generations revealed substantial variability for droughtadaptive traits under both irrigated and stress environments. The distinct performance of parents and derived generations highlighted the contrasting physiological mechanisms underlying drought tolerance, with the drought-tolerant parent and BC 2 generation exhibiting superior SPAD values, higher WUE and favourable stomatal adjustment under moisture stress. Generation mean analysis indicated that the additive dominance model alone could not sufficiently explain trait inheritance, necessitating a six-parameter model to capture the complexity of gene interactions. The predominance of dominance and dominance × dominance effects for SLA, stomatal conductance and WUE under drought stress suggests that non-additive gene action plays a crucial role in governing these traits. Such traits may be more effectively improved through hybridization followed by delayed selection. Conversely, significant additive and additive × additive effects observed for SPAD, WUE and stomatal number indicate the potential for early-generation selection when stable inheritance is present. Overall, the findings underscore the value of integrating physiological indicators into drought-breeding pipelines. Emphasizing traits such as SPAD chlorophyll meter reading, WUE and stomatal behaviour can enhance the precision of selection and accelerate the development of groundnut genotypes with improved water-use efficiency and resilience to moisture stress.

References

1. Rahuri: Mahatma Phule Krishi Vidyapeeth; 2012.
2. Cavalli LL. An analysis of linkage in quantitative inheritance. In: Rieve ECR, Waddington CH, editors. Quantitative inheritance. London: HMSO; 1952. p. 135-144.
3. FAOSTAT. FAOSTAT statistical database. Rome: Food and Agriculture Organization of the United Nations; 2023. Available from: http://www.fao.org/faostat/en/
4. Hayman BI. The separation of epistasis from additive and dominance variation in generation means I. Heredity. 1958;12:371-390.
5. Hayman BI, Mather K. The description of genetic interactions in continuous variation. Biometrics. 1957;11:371-390.
6. Jayalakshmi V, Reddy LG. Generation mean analysis for certain quantitative characters in groundnut. Agric Sci Dig. 2002;23(4):255-258.
7. John K, Reddy PR, Reddy PH. General and specific combining ability estimates of physiological traits for moisture stress tolerance in groundnut (Arachis hypogaea L.). Int J Appl Biol Pharm Technol. 2012;2:1-12.
8. Lal C, Hariprasanna K, Rathnakumar AL, Basu MS, Gor HK, Chikani BM. Identification of water-useefficient groundnut genotypes for rainfed situations through leaf morpho-physiological traits. IAN. 2005;25:4-7.
9. Mahon JD, Domey J. A lightweight battery-operated infrared gas analyzer for field measurement of photosynthetic CO₂ exchange. Photosynthetica. 1979;13:459-466.
10. Mather K. Biometrical genetics. New York: Dover Publications Inc.; 1949. p. 1-882.
11. Nigam SN, Upadhyaya HD, Chandra S, Nageswara Rao RC, Wright GC, Reddy AGS. Gene effects for specific leaf area and harvest index in three crosses of groundnut (Arachis hypogaea L.). Ann Appl Biol. 2001;139:301-306.
12. Panse VG, Sukhatme PV. Statistical methods for agricultural workers. New Delhi: ICAR; 1995. p. 145150.
13. Pitaloka MK, Caine RS, Hepworth C, Harrison EL, Sloan J, Chutteang C, et al. Induced genetic variations in stomatal density and size of rice strongly affect water-use efficiency and responses to drought stress. Front Plant Sci. 2022;13:1-12. doi:10.3389/fpls.2022.801706
14. Ramya KT, Bellundagi A, Harikrishna, Rai N, Jain N, Singh PK, et al. Gene action governing the inheritance of stomatal conductance in four wheat crosses under high-temperature stress. Front Plant Sci. 2021;12:658443. https://doi.org/10.3389/fpls.2021.658443
15. Suriharn B, Patanothai A, Jogloy S. Gene effects for specific leaf area and harvest index in peanut (Arachis hypogaea L.). Asian J Plant Sci. 2005;4(6):667-672.
16. Upadhyaya HD, Sharma S, Singh S, Singh M. Inheritance of drought-resistance-related traits in two crosses of groundnut (Arachis hypogaea L.). Euphytica. 2011;177:55-66.
17. Venkateswarlu D, Vasanthi RP, Raja Reddy K, Reddy PV, Hariprasad Reddy K, Eswar Reddy NP. Combining ability analysis for yield attributes in groundnut (Arachis hypogaea L.). J Oilseeds Res. 2007;24(2):274276.