INTRODUCTION Chickpea (Cicer arietinum L.) is a self-pollinating, diploid (2n=2x=16) pulse crop with a 738Mbp genome (Varshney et al., 2013). Chickpea seeds are a good source of carbohydrates and proteins for the vegetarian diets of resource-poor consumers. Globally chickpea covers 14.8 million ha (mha) area with an annual production of 15.1 million tons (FAO, 2021). In India, ‘Pulse Revolution’ is majorly contributed by chickpea to move the country towards self-sufficiency in pulses. An all-time high of 12.61 mt chickpea production recorded during 2020-21 (Dixit, 2021). Drought is being most detrimental abiotic stress by limiting production and productivity of crops more than other abiotic stresses (Shao et al., 2009). Drought mainly affects yield, membrane integrity, osmotic adjustment, pigment content and photosynthetic activity. In India, there has been substantial shift of region of chickpea cultivation from cooler Northern climatic conditions zones to hot southern Indian conditions limited to drought prone marginal and sub marginal tracts. That’s greatly affected chickpea yields of country over the past few years. Further, late onset of raining delayed chickpea sowing in rice fallows conditions and exposing chickpea to heat and drought stresses during reproductive stage as terminal heat and drought stresses (Sachdeva et al., 2017). The crop responses to various abiotic stresses are complex involving morpho-physiological, biochemical and gene regulatory mechanisms for drought resilience. Thus, this study was conducted for characterization of the chickpea genotypes on the basis of morpho-physiological responses under drought stress to select promising drought tolerant line. MATERIAL AND METHOD The experimental material consists of 40 chickpea genotypes including released varieties, identified donar and advance breeding lines (Table 1). The research trial was laid out in RCBD with three replications during rabi seasons 2020-21 and 2021-22 at field of Biotechnology center, Jawaharlal Nehru Krishi Vishwa Vidyalaya, Jabalpur. Three row of 1 m length was planted for each genotype with 10 cm of plant to plant and 45 cm of row to row distance. Standard agronomical practices have been implemented to maintain ten numbers of plants in each row. Five evocative plants are carefully chosen from each line for recording the observations on chlorophyll content index (CCI), plant height (PH), number of primary branches (NPB), number of secondary branches (NSB), biological yield per plant and seed yield per plant. Relative water content was calculated according Sachdeva et al. (2017). Saturation Water Deficit was calculated by subtracting RWC from 100. Canopy temperature depression was calculated by subtracting canopy temperature of plant from air temperature. Statistical analysis of pooled data of both seasons was done by using Window Stat 9.1 software. Genetic diversity was calculated using Mahalanobis’s D2 (Mahalanobis’s, 1936) while and clustering of genotypes was conducted according to Tocher’s method (Rao, 1952). RESULTS AND DISCUSSION In India, chickpea is third most important legume crop occupying 45% of total pulse production. Drought and heat both limit chickpea production critically. The mean of the studied characters indicate presence of moderate amount of variation in the tested genotypes. on the basis of pooled data analysis of both seasons, RWC, SWD, CTD, CCI, PH, NPB, NSB, biological yield per plant and seed yield per plant of the forty genotypes were recorded. Under normal condition, the average RWC value was recorded 73.5 + -4.6 with range from 65.2 to 79.5, mean SWD was found 31.7 + -6.3 with range from 20.6 to 43.0, mean CTD was obtained 3.3+- 0.4 with range from 2.6 to 4.0, mean CCI was observed 58.8 + -2.0 with range from 55.5 to 62.5, PH was recorded 50.9 + - 4.7 with range from 38.7 to 58.2 cm, mean NPB was recorded 3.0 + -0.2 with range from 2.6 to 3.6, mean NSB was found 8.5 + -0.8 with range from 7.5 to 10.3, mean biological per plant was observed 34.4 + -4.7 g with range from 25.5 g to 48.0 g and mean seed yield per plant was recorded 13.7 + -1.8 g with range from 11.7 to 21.7 g (Table 2). Under drought stress situations, the mean RWC value was recorded 68.3 + - 6.3 with range from 57.0 to 79.4, mean SWD was found 31.7 + - 6.3 with range from 20.6 to 43.0, mean CTD was obtained 1.7 + - 0.3 with range from 1.1 to 2.2, mean CCI was observed 54.8 + -2.1 with range from 51.2 to 58.9, PH was obtained 45.3 + - 4.5 with range from 33.5 to 51.5 cm, mean NPB was recorded 2.5 + -0.2 with range from 2.0 to 3.1, mean NSB was found 7.4+-0.7 with range from 6.6 to 9.3, mean biological per plant was observed 23.5 + -3.8 g with range from 17.4 g to 33.9 g and mean seed yield per plant was recorded 8.9 + -0.7 g with range from 6.1 to 9.9 g (Table 3). The dendrogram based on Tocher clustering grouped the forty tested genotypes into ten major clusters (Table 4, Fig. 1). The largest cluster, cluster I comprised of 13 genotypes (ICCV15118, JG2016-1411, JG32, JG24, JG33, JG28, JG2016-921814, ICCV181664, JG2016-45, ICCV15102, JG2003-14-16, JG2016-9651 and JG2021-1617) followed by cluster II, III and IV consisted with 9 (JAKI9218, JG63, ICC4958, JG11, JG16, JG2018-51, JG17, ICCV19616), 6 (JG42, JG28, JG2022-74, JG2016-36, JG2016-1614, JG2016-44) and 6 (JG14, JG74, JG226, JG2016-45) genotypes, respectively. Rest six clusters comprised with single-single genotypes (JG2021-6301, JG2021-1424, JG36, JG2016-74315, PG205, JG2016-634958 respectively). Bharadwaj et al. (2001) suggested that phenotypic and/or genotypic diversity per se is an inferential criterion so should not be used as as a direct measure of genetic diversity. It is may not be more useful for selecting the genotypes as parents for breeding program, generally done by most breeders. Numerous clustering techniques have been utilized by different researchers to quantify the genetic diversity in a given set of germplasm/ genotypes on the basis of collected data (Bharadwaj et al., 2011; Sachdeva et al., 2017; Katkani et al., 2022). Tocher clustering could clearly delineate the drought tolerant chickpea genotypes from the susceptible genotypes. In this study, Tocher clustering clearly grouped most drought tolerant genotypes into cluster II (JAKI9218, JG63, ICC4958, JG11, JG16, JG2018-51, JG17, ICCV19616) and discriminated from drought sensitive genotypes which were grouped into cluster VI (JG14, JG74, JG226, JG2016-45). Rest clusters contained with moderately drought tolerant chickpea genotypes. Sachdeva et al. (2017) also grouped chickpea genotypes on the basis of morpho-physiological traits dendrogram and found that Cluster IIa contained with most drought tolerant genotypes viz., ICC4958, ICCV10313, ICCV10 and ICCV97309 while cluster I and cluster III had the most susceptible chickpea genotypes. The clustering pattern of genotypes clearly depicted that considerable amount of diversity was present in the utilized material of study. This could be due to differential selection executed by breeders for selection of seed yield attributing and other traits which have been considered as genetic drift because of selection (Murty and Arunachalam 1966). Further, the intra and inter cluster Mahalanobis D2 values depicted wide range of intra cluster distance from 0.00 to 12.29 (Table 5). Cluster III demonstrated highest intra cluster D2 mean value (D2 = 12.29) followed by Cluster II (D2 = 10.14), Cluster IV (D2 = 9.80) and Cluster I (9.13), whereas remaining six clusters (Cluster V, VI, VII, VIII, IX and X) revealed zero value for Intra cluster distance due to having single genotype in each cluster. These monogenotypic clusters represented minimum diversity for the present study. The maximum inter cluster divergence distance was depicted between genotypes of Cluster II and Cluster IV (56.04) representing their highest suitability for utilizing in crossing programme. Outcomes of the study clearly specified the remarkable possibilities of incorporation of allelic resources existing in these genotypes by using a systematic breeding program. The mean of clusters for all studied traits in pooled data analysis are presented in (Table 6). Cluster IX (77.6) revealed highest mean for RWC while Cluster V was found with minimum cluster mean (61.44). Maximum SWD was recorded for Cluster V (38.56) while minimum SWD was observed in Cluster IX (22.40). Highest CTD was depicted by Cluster II (2.08) whereas lowest SWD was found in Cluster IV (1.12). Utmost superior CCI was demonstrated by Cluster Cluster IX while utmost inferior value recorded from Cluster VIII. Tallest plants were showed by Cluster VI (47.67) with shortest plants in Cluster V (33.50). Maximum NPB was recorded in Cluster XI (3.13) with minimum NPB in Cluster V (2.17). Highest NSB found in Cluster V (8.40) with lowest NSB in Cluster VI (7.02). Utmost high biological yield per plant (33.87 g) were noted down in Cluster VII while utmost low biological yield per plant (20.83 g) were noted in Cluster VIII. Maximum seed yield per plant was demonstrated by Cluster IX (9.50 g), while minimum was observed in Cluster VI with 8.13 g mean value. These findings approved in earlier research of Tiwari and Babbar (2017); Gediya et al. (2018); Ponnuru et al. (2019); Dar et al. (2020); Janghel (2020); Boparai et al. (2021); Katkani et al. (2022); Biswal et al. (2022).