Natural variability and heritability of root-nodulation traits in chickpea (Cicer arietinum L.) minicore

Abstract

The existence of large variations for nodulation traits in chickpea minicore was revealed and genetic materials for beneficial biological nitrogen fixation (BNF) traits like early nodulation, high nodulation, and delayed nodule senescence were identified. Early-nodulating genotypes viz. ICC12968, ICC7867, ICC13816, ICC867, ICC15264, ICC15510, and ICC283 produced > 10 nodule number per plant (NNPP) at 15 as well as 30 days after sowing (DAS). Maximum of 36 NNPP at stage 3 i.e., 253% higher than check cultivar were observed in Iran originated ICC6874. Chickpea minicore showed large variations for nodule mass that ranged up to 850 mg/plant at 60 DAS and 2290 mg/plant at 90 DAS. Strong positive correlation was found between nodule fresh weight and specific weight at stage 3 (0.69) and stage 4 (0.76). Besides these, few slight positive significant correlations were also observed viz., nodule number per plant at stage 3 and 4 (0.45), nodule fresh weight at stage 3 and 4 (0.39). Principal component analysis (PCA) indicated that dimensions 1 (21%), 2 (17.6%), and 3 (13%) accounted for a substantial portion of the phenotypic variance, each contributing more than 10%. Accessions viz. ICC1431, ICC13599, ICC13764, and ICC13863 with pink active root nodules and high nodule biomass at later crop growth stages are considered as genetic resources to extend the BNF support in chickpea. High broad-sense heritability values of 76.43 and 90.23 were observed for early nodulation and delayed nodule senescence, respectively. Hence, the identified genotypes for early nodulation and delayed nodule senescence can be used for improving symbiotic efficiency in chickpea.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13205-023-03908-1.

Introduction

Chickpea (Cicer arietinum L.) is considered globally as one of the most important grain legumes and cultivated in different bioclimatic regions from subtropical India and northeastern Australia to arid and semi-arid Mediterranean region of West Asia and North Africa (WANA), East Africa Basin and South Australia (Laranjo et al. 2008). The global production of chickpea rose up to 14.7 million MT, with Asian countries accounting for 85.5% of the total production (FAOSTAT 2021). India is the world's largest chickpea producer and consumer, sharing about 71.0% of the world’s acreage and production. Chickpea grains are highly nutritious with protein content of 22.12–24.42%. In addition to its valuable nutritional quality, chickpea restores and maintains soil fertility through symbiotic nitrogen fixation (SNF) and plays a key role in crop rotation to sustain agricultural productivity.

Biological nitrogen fixation (BNF) globally contributes 107 Tg of nitrogen in natural terrestrial ecosystems with a range of 19.8–107.9 Tg Nyr⁻¹ (Yu and Zhuang 2020). The symbiotic relationship between grain legumes with their host specific Rhizobium contributes nearly 10 Tg Nyr⁻¹ (Ogola 2015). Chickpea performs BNF in association with Mesorhizobium and nearly 80% of the nitrogen requirement of chickpea is obtained by symbiosis. Inoculation of chickpea with an efficient strain of Mesorhizobium supplied nitrogen to the soil between 80 and 141 kg ha⁻¹. Several researchers found that effective establishment of SNF increased the N status in grain, which is directly associated with high protein content of chickpea grains (Tagore et al. 2014; Kumar et al. 2014). Chickpea inoculated with Mesorhizobium was observed with 61.1% and 11.4% greater grain N and P contents, respectively (Kaur et al. 2015).

Besides the selection of superior Mesorhizobium strains for seed inoculation, strategies on developing elite crop varieties with high BNF potential through breeding program should also be considered. Availability of large variations for a particular trait in plant genetic resources provides the basis and the raw material that plays a fundamental role in success of crop improvement programs. Allito et al. (2021) reported nodulation behavior and biomass production in three varieties of fava bean with six different Rhizobium strains and found that inoculated plants produced higher nodulation, shoot and root dry weight compared to un-inoculated control plants. Fava bean cultivars i.e., Dosha and Gora produced maximum nodules viz. 50.3 and 45.3 NNPP, respectively, with higher nodule dry weight.

Till now, limited germplasm sets of pulses have been screened for SNF and resulted in identification of host accessions that fix more atmospheric N₂ compared to others under similar conditions. Certain genotypes of common bean germplasm showed higher SNF (127.7 mg N₂/plant) compared to other plants (47.8 mg N₂/plant) under limited P supply i.e., 70 m mol P/plant (Vadez et al. 1999). Large germplasm collections must be genotyped and phenotyped for characteristics of interest to provide usable information to plant breeders. Phenotypic analysis of all accessions is not feasible in many cases. International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) (Patancheru, India) contains several thousands of chickpea lines and to characterize those lines in a short period of time a “mini core collection” concept was developed in which selection of 1% of the entire collection of the germplasm using all the available information (Upadhyaya and Ortiz 2001).

Tellawi et al. (2007), Mensah and Olukoya (2007), and Gallani et al. (2005) have reported variations in nodulation among chickpea genotypes. Several different types of nodulation groups viz. high nodulating, low nodulating, and non-nodulating types have been identified within chickpea cultivars (Rupela 1994). In general, high nodulating selection grew better than the non-nodulating and low nodulating selections of a given cultivar. Chickpea cultivar G 130, a high nodulating selection, produced 31% more grains at low soil N level than its low nodulating selection (Venkateswarlu and Katyal 1994).

Biabani et al. (2011) reported fourfold differences in the total N₂ fixed (0.02–0.84 g/plant) in a subset of genetically diverse USDA chickpea core collection accessions, with accession ILC 235 from Iraq being the greatest N₂ fixer. Average nodules per plant ranged from a minimum of 21 on accession number (ACNO) 268376 from Afghanistan to 101 on ACNO 451161 from Iran. The maximum proportion of N-fixed (PNF) was 78.1% from ACNO 451112 (Iran) while the minimum was 47.5% from ACNO 360439 (Iran). PNF had a positive, significant correlation with total plant biomass (Pearson r = 0.20, P < 0.01) indicated that plants with greater biomass tended to support more N fixation. Larger plants may need more N to support the larger biomass, provide a stronger N sink, and have more available photosynthates resources to support the energy-intensive N fixation process. Similarly, Plett et al. (2021) found that among six chickpea genotypes, two genotypes i.e., Moti and Yorker produced higher root nodules in moderate available N in comparison to higher available N.

Even though there is a huge genotypic variability in the number of nodules and nodule mass in legumes, efforts have been limited to use this variability for improved N₂ fixation in breeding. Early nodulation and delayed nodule senescence are considered as desirable traits since they can extend the duration of support to the plants through symbiotic nitrogen fixation. Hence, the present investigation aims to reveal the phenotypic variability on nodulation traits of chickpea minicore. Potential and contrasting chickpea genotypes identified in this study for early nodulation, late nodule senescence, and high nodulation can be used for developing elite cultivars with high symbiotic efficiency.

Materials and methods

Field trial to reveal the nodulation potential of chickpea

BNF is energy-intensive process in terms of ATP requirement, and identification of chickpea genotypes that produce higher nodulation potential under P-limited conditions will be helpful to improve BNF in resource poor farmer’s field. Hence, the field experiment was laid out in the low-P field strip using an augmented design (Federer and Searle 1976) with chickpea minicore lines (n = 214) during Rabi 2018–19 at B-16 field (latitude 26°31′10.2″N and longitude 80°15′02.2″E), New Research Farm of ICAR-Indian Institute of Pulses Research (ICAR-IIPR), Kanpur-208 024, India. Field soil was characterized as Inceptisol with 0–15% granularity, 0.159% organic carbon, 224.7 kg N/ha, 117.60 kg K/ha, 10.06 kg S/ha, 1.68 ppm Fe, and 1.872 ppm of Zn. Field strips with low available soil-phosphorus (7 ppm) were developed at B-16 field by avoiding the application of phosphoric fertilizers to the crops, stabilized, and properly monitored by determining Olsen P after cultivation of every crop. No fertilizer was applied to the field strip with a purpose to identify chickpea lines that potentially nodulate in low fertility soils. The augmented design is commonly used to evaluate the performance of large numbers of genotypes using non-replicated new test treatments and replicated check treatments. Experimental design consisted of 11 blocks each containing maximum of 20 chickpea test lines and 2 randomly allotted checks via cv Subhra and RSG 888. Check cultivars were used to normalize the data before analysis to neutralize the variations caused due to soil conditions of large field. One row with a length of 4 m for each chickpea genotypes was maintained with row-to-row spacing of 30 cm and plant-to-plant spacing of 10 cm. Furrow sowing of chickpea seeds and standard agronomic practices were followed to grow the chickpea minicore lines. Samples were collected from three randomly selected plants of each genotypes at four different crop growth stages viz. stage 1 (S1: 15 days after sowing-DAS) to select early nodulating genotypes, stage 2 (S2: 30 DAS), stage 3 (S3: 60 DAS) to identify germplasm lines with high nodulation potential, and stage 4 (S4: 90 DAS) to select genotypes with delayed nodule senescence. The observations were made on BNF-related traits viz., number of nodules (S1–S4), nodule fresh weights (S3 and S4), plant dry weight (S1–S4), and nitrogen content (S3).

Determination of nitrogen in plant samples

The plant samples collected from the field were dried in an oven (50 °C) and ground into homogenous powder. A known weight of plant sample (0.2 g) was taken in a digestion tube, and 2.5 ml of concentrated sulphuric acid solution and catalyst mixture (K₂SO₄ + CuSO₄ + Selenium) were added and mixed gently. Boiling tubes with sample mixture were placed in the Tecator block digester and pre heated at 100 °C for one hour. Three ml of H₂O₂ was added to each tube and heated at 350 °C for 3 h. Following digestion, the tubes were removed from the block digester and left to cool until they were able to be handled. The digest was first made up to 25 ml with deionized water and shaken well to dissolve the entire digest. After mixing, the digest was transferred into Kjeldahl flask, added with 10 ml of 40% NaOH, 6 ml of 4% boric acid and bromo-cresol green indicator and then allowed for distillation. After distillation, 0.1 N HCL was added to the indicator mixture through burette, followed by readings noted down, and finally % of nitrogen was calculated (Kjeldahl 1883).

Variability analysis

Phenotypic data on nodulation and other SNF-related parameters were analyzed using R package “augmented RCBD” for error correction, ANOVA, descriptive statistics, and for variability analysis, such as phenotypic variation (PV), genotypic variation (GV), environmental variations (EV), phenotypic coefficients of variation (PCV), genotypic coefficients of variation (GCV), broad-sense heritability (H²), genetic advance (GA), and genetic advance as percentage of mean (GAM). Phenotypic (${\sigma }_{\text{p}}^2$), genotypic (${\sigma }_{\text{g}}^2$), and environmental variance (${\sigma }_{\text{e}}^2$) were calculated using the formula as described by Federer and Searle (1976):

$${\sigma }_{\text{g}}^2={\sigma }_{\text{p}}^2-{\sigma }_{\text{e}}^2.$$

Phenotypic and genotypic coefficients of variation (PCV and GCV) were determined as described by Burton (1951, 1952).

$$\text{GCV}=\frac{{\sigma }_{\text{g}}^2}{\sqrt{\overline{x}}}\times 100,$$

where x is the mean.

PCV and GCV for chickpea minicore lines were further categorized (Sivasubramaniam and Madhavamenon 1973) as follows:

$$\begin{array}{cc}& \text{CV}(\%)\text{Category}\hfill \\ \hfill & x<10\text{Low}\hfill \\ \hfill & 10<x<20\text{Medium}\hfill \\ \hfill & >20\text{High}.\hfill \\ \hfill \end{array}$$

The broad-sense heritability (H²) is calculated as described by Lush (1940) using the following formula:

$$H^2=\frac{{\sigma }_{\text{g}}^2}{{\sigma }_{\text{p}}^2}.$$

The estimates of broad-sense heritability (H²) are categorized (Robinson 1966) as follows:

$$\begin{array}{cc}& \text{H}2\text{Category}\hfill \\ \hfill & x<30\text{Low}\hfill \\ \hfill & 30<x<60\text{Medium}\hfill \\ \hfill & >60\text{High}.\hfill \\ \hfill \end{array}$$

Genetic advance (GA) was determined (Johnson et al. 1955) as follows:

$$\text{GA}=k\times {\sigma }_{\text{g}}\times \frac{H^2}{100},$$

where the constant k is the standardized selection differential or selection intensity. The value of k at 5% proportion selected is 2.063 as described by Falconer and Mackay (1996). Genetic advance as percent of mean (GAM) was determined and categorized (Johnson et al. 1955) as follows:

$$\text{GAM}=\frac{\text{GA}}{\overline{x}}\times 100.$$

$$\begin{array}{cc}& \text{GAM}\text{Category}\hfill \\ \hfill & x<10\text{Low}\hfill \\ \hfill & 10<x<20\text{Medium}\hfill \\ \hfill & >20\text{High}.\hfill \\ \hfill \end{array}$$

The error corrected phenotypic data were further used to test the strength of relationship between variables following correlation analysis with R packages “Corrplot” and “Performance Analytics”. The principal component analysis (PCA) of various traits with corrected phenotypic data was performed with “Factoextra” and “FactoMine” packages in the R programming language.

The genotypes were clustered based on adjusted phenotypic variables. The K-means algorithm, implemented using the “factoextra” and “cluster” packages in the R programming language, was utilized to predict the number of clusters present in the studied materials. To determine the optimal number of clusters in the studied materials, both the silhouette and the gap statistic methods were employed. Gap statistic method was performed with 500 bootstrap iterations.

Results

Chickpea minicore lines showed large variations for nodule number per plant (NNPP). Variation for nodulation at early crop growth stage i.e., 15DAS (S1) is in the range of 0–18 NNPP (Table 1). A total of 11 minicore lines did not produce any nodule at S1 and further 9 genotypes recorded the values of < 1 nodule per plant as mean of three replications. Most of the minicore lines produced less than 10 nodules per plant, while only 14 genotypes produced more than 10 nodules per plant i.e., 66% higher nodulation than check cultivars (Fig. 1). Genotypes viz. ICC1356, ICC7819, and ICC283 with high nodulation potential i.e., 18, 15, and 12 NNPP, respectively at 15 DAS are considered as early nodulating type, which is a preferred trait for early establishment of chickpea seedlings. However, ICC1356 with 18 NNPP at stage 1 did not produce any nodules at later stages of the crop growth. At stage 2 (S2: 30 DAS), a single genotype was observed with no nodulation, while 59 genotypes produced high nodulation i.e., more than 10 NNPP (Fig. 2). ICC506 produced 29 NNPP at stage 2; however, it did not produce any nodule at stage 1. Seven genotypes viz. ICC12968, ICC7867, ICC13816, ICC867, ICC15264, ICC15510, and ICC283 are common under high nodulation groups (> 10 NNPP) of stage 1 as well as stage 2.

Table 1.

Descriptive statistics of nodulation traits chickpea minicore genotypes grown in phosphorous deficient field environment

Trait	Mean	Overall adjusted mean	Std error	Std deviation	Min	Max	CV
NNPP_S1 (units)	5.25	5.24	0.22	3.16	0	17.94	29.12
NNPP_S2 (units)	8.26	8.26	0.31	4.37	0	29.48	22.65
NNPP_S3 (units)	10.4	10.33	0.48	6.75	0	36.10	20.46
NNPP_S4 (units)	10.23	10.15	0.55	7.89	0	37.90	13.38
NFW_S3 (g/plant)	0.16	0.15	0.01	0.15	0	0.85	40.70
NFW_S4 (g/plant)	0.27	0.25	0.02	0.31	0	2.29	32.94
NSW_S3 (mg/plant)	17.42	17.35	1.88	26.68	0	326.04	16.10
NSW_S4 (mg/plant)	26.56	25.71	2.07	29.48	0	194.06	16.64
PDW_S1 (g/plant)	0.14	0.14	0.004	0.06	0.05	0.35	12.28
PDW_S2 (g/plant)	0.26	0.26	0.01	0.1	0.04	0.65	10.58
PDW_S3 (g/plant)	1.44	1.44	0.05	0.72	0.26	5.15	9.86
PDW_S4 (g/plant)	4.97	4.97	0.17	2.38	0.06	15.15	4.58
NC_S3 (%)	1.25	1.25	0.01	0.18	0.82	1.65	7.60

NNPP nodule number per plant, S stage, NFW nodule fresh weight, NSW nodule specific weight, PDW plant dry weight, NC nitrogen content

Fig. 1.

Early nodulation and delayed nodule senesces in chickpea: A ICC1052 is non-nodulation while ICC13816 and ICC16915 are early nodulating genotypes. B Nodules of ICC762 are degenerated at 90DAS while ICC13863 and ICC13764 had active nodules

Fig. 2.

Distribution frequency of chickpea genotypes based on nodule number at four crop growth stages

At stage 3 (S3: 60 DAS), 7 genotypes were observed with no nodulation while 17 genotypes produced more than 20 nodules/plant and the maximum number of nodules per plant (36) was observed in ICC6874 (Table 1). Ten genotypes viz. ICC16374, ICC13816, ICC9002, ICC12968, ICC6816, ICC11944, ICC4567, ICC506, ICC4533, and ICC14815 are common to the high nodulation category of stage 2 (> 10 NNPP or 28% higher than check) and stage 3 i.e., > 20 NNPP 96% higher nodulation than check cultivars. Two genotypes viz. ICC13816 and ICC12968 are common to high nodulating groups of all three stages i.e., > 10 NNPP at S1, S2, and > 20 NNPP at S3.

At stage 4 (90DAS), 16 genotypes were observed with no nodulation, while 23 genotypes produced more than 20 nodules/plant. ICC11121 produced the maximum number of 38 NNPP i.e., 283% higher nodulation than check cultivars (Table 1). However, it did not perform well in terms of nodulation at early stages of crop growth i.e., 1, 8, and 18 NNPP at S1, S2, and S3 respectively. Eight genotypes viz. ICC12307, ICC14402, ICC6874, ICC12155, ICC4463, ICC9002, ICC12968, and ICC3325 are common for the high nodulation category (> 20 NNPP) of stage 3 and stage 4. ICC12968 is ranked under high nodulating group of all 4 crop growth stages in comparison to the rest of the genotypes in the same category. ICC3325 was observed with high number of nodules per plant constantly with 33 and 35 NNPP during stage 3 and stage 4 respectively.

Chickpea minicore lines belong to “desi type” were considered as a good source of variations for early nodulation potential that ranges from 0 to 21 nodules per plant at stage 1, while kabuli chickpea produced maximum of 13 nodules per plant (Table 2). In general, the nodulation behavior at all 4 stages of crop growth is poor in pea-type chickpea lines followed by kabuli chickpea (Fig. 3). Many of the chickpea genotypes originated from Iran produced a greater number of nodules, which was indicated by comparatively higher values of adjusted median (Fig. 4). The highest number of NNPP (43) at stage 3 was observed in ICC6874 originated from Iran, while Indian originated chickpea lines produced higher NNPP at remaining stages of plant growth (Table 2). Chickpea genotypes viz. ICC7668 (Russian Federation), ICC12155 (Bangladesh), and ICC3325 (Cyprus) were produced > 25 NNPP at Stage 3 as well as stage 4. However, ICC12155 (Bangladesh) and ICC3325 (Cyprus) were considered as poor nodulators at early stages with < 10 NNPP. Chickpea genotypes originated from Ethiopia are considered as a poor nodulators.

Table 2.

Variability in nodule number among chickpea minicore genotypes in response to origin

Origin of chickpea genotype	Minimum	Maximum	Mean	Median	Standard Deviation	CV
Desi type
India (78)
S1	1.0	20.5	05.51	05.00	3.40	61.70
S2	1.5	31.0	07.95	07.00	4.68	58.86
S3	1.0	29.0	11.03	10.00	7.18	65.09
S4	1.5	38.0	12.39	10.00	8.86	71.50
Iran (38)
S1	1.0	18.0	05.92	05.50	3.55	59.96
S2	1.0	26.0	08.70	08.50	4.29	49.31
S3	2.0	43.0	12.01	09.50	8.53	71.02
S4	1.5	33.0	12.84	12.50	8.13	63.31
Ethiopia (13)
S1	2.5	05.0	03.85	03.75	0.81	21.03
S2	1.0	11.0	04.90	04.50	2.93	59.79
S3	1.5	22.0	06.77	05.00	5.88	86.85
S4	0.5	14.0	05.31	04.00	8.09	152.35
Others (26)
S1	1.5	12.0	05.21	04.50	2.80	53.74
S2	2.0	15.0	07.30	06.00	3.25	44.52
S3	2.5	35.0	11.29	09.75	7.42	65.72
S4	2.0	33.0	12.11	10.25	8.43	69.61
Kabuli type
Iran (13)
S1	2.5	12.5	05.38	04.00	2.79	51.85
S2	3.0	18.5	08.96	09.50	4.03	44.97
S3	5.5	15.0	10.50	10.00	2.73	26.00
S4	4.0	14.5	10.40	09.00	5.46	52.50
India (5)
S1	2.5	10.5	04.87	03.25	3.77	77.41
S2	7.0	10.5	09.25	09.75	1.65	17.83
S3	5.5	20.0	14.62	16.50	6.92	47.33
S4	5.5	34.0	14.37	09.00	13.18	91.71
Morocco (4)
S1	2.0	08.0	04.66	04.00	3.05	65.45
S2	3.5	21.5	11.00	08.00	9.36	85.09
S3	12.5	28.0	18.50	15.00	8.32	44.97
S4	7.0	20.0	11.83	08.50	7.11	60.10
Others (18)
S1	0.5	09.5	05.03	05.25	2.69	53.47
S2	3.5	17.0	08.12	06.75	3.90	48.02
S3	4.0	12.0	08.00	07.50	2.71	33.87
S4	2.0	21.0	08.87	08.25	5.05	56.93

Fig. 3.

Box plot explaining variations in nodule number of chickpea types at four stages of crop growth

Fig. 4.

Box plot explaining variations in nodule number of chickpea from different origins

Nodule fresh weight was recorded at stage 3 and stage 4 of crop growth. Chickpea germplasm lines showed large variations for nodule fresh weight that ranges from 0 to 850 mg/plant at stage 3 and 2290 mg/plant at stage 4 (Tables 1 and 3). Eight genotypes viz. ICC13764, ICC762, ICC13892, ICC13599, ICC1431, ICC13863, ICC9755, and ICC14077 produced > 500 mg fresh weight of root nodules per plant at stage 3. ICC12028, ICC1431, ICC13599, ICC3946, ICC13764, and ICC13863 were reported with nodule fresh biomass of > 1.0 g/plant at stage 4. Four genotypes viz. ICC1431, ICC13599, ICC13764, and ICC13863 were common for the high nodule biomass category of both stage 3 and stage 4. Since these genotypes were observed with pink nodules, they are considered to possess the trait of late nodule senescence.

Table 3.

Variability in nodule fresh weight among chickpea minicore genotypes in response to origin

Nodule fresh weight	Minimum	Maximum	Mean	Median	Standard deviation	CV
Desi type
India (78)
S3	0.006	0.72	0.13	0.08	0.12	92.30
S4	0.01	1.40	0.23	0.15	0.25	108.00
Iran (38)
S3	0.01	0.54	0.17	0.15	0.13	76.40
S4	0.02	1.40	0.31	0.20	0.30	96.70
Ethiopia (13)
S3	0.04	0.81	0.26	0.09	0.28	107.60
S4	0.05	2.10	0.38	0.09	0.79	207.80
Others (26)
S3	0.01	0.81	0.15	0.10	0.17	113.30
S4	0.02	0.92	0.30	0.23	0.27	90.00
Kabuli type
Iran (13)
S3	0.04	0.48	0.23	0.24	0.14	60.80
S4	0.04	1.68	0.32	0.17	0.47	146.80
India (5)
S3	0.33	0.25	0.13	0.12	0.09	69.20
S4	0.09	0.24	0.19	0.02	0.06	31.50
Morocco (4)
S3	0.04	0.24	0.14	0.15	0.10	71.40
S4	0.18	0.24	0.22	0.22	0.03	12.50
Others (18)
S3	0.03	0.42	0.12	0.08	0.11	91.66
S4	0.61	0.95	0.32	0.20	0.29	90.62

Nodule specific weight (NSW) for each genotype in chickpea minicore was calculated. Five genotypes viz. ICC8350, ICC10755, ICC8855, ICC13599, and ICC13863 were common for high NSW category of stage 3 and stage 4 with NSW > 50 mg/nodule. The above listed genotypes were observed with very few nodules ranging from 1 to 13 with relatively higher nodule biomass in the range of 0.17–2.29 g/plant. Median value for NSW at stage 3 is the highest in pea-type chickpea genotypes (23.67) in comparison to desi (12.9) and kabuli (14.7) (Fig. 5). Same trend was maintained by germplasm lines during stage 4, where the median for NSW is the highest in pea-type (34.75) in comparison to desi (17.11) and kabuli (22.21). Maximum NSW at stage 3 (326.04 mg/nodule) and stage 4 (194.03 mg/nodule) was reported in ICC14077 and ICC13863, respectively. Chickpea genotypes from Ethiopia recorded the highest variation for NSW at stage 3 ranging from 1 to 326.04 mg of nodule specific weight with median value of 17.93. Indian genotypes possess narrow range for NSW from 0 to 61.54 mg with median value of 11.49 mg. Similar trend was observed with range of variations in nodule specific weight at stage 4.

Fig. 5.

Box plot explaining variations in nodule specific weight of chickpea

Plant dry weight of minicore lines observed at any particular growth stage showed strong positive correlation (0.26–0.59) with that of other stages. NNPP at stage 1 did not show any correlation with that of other stages i.e., S2–S4 (Figs. 6 and 7). However, NNPP at S2 showed strong positive correlation (0.27) with S3 and NNPP at S3 showed strong positive correlation (0.45) with S4. NNPP at S2 showed week positive correlation (0.11) with NNPP at S4. NNPP and NFW are positively correlated at S3 (0.25) as well as stage 4 (0.33). NSW is strongly correlated with NFW at stage 3 (0.69) as well as stage 4 (0.76), but negatively correlated with NNPP at stage 3 (− 0.10) and not correlated with NNPP at stage 4. NNPP showed positive correlation with PDW at stage 1 (0.35) and stage 2 (0.29). However, at later crop growth stages, no correlation was observed among these parameters. Nitrogen content of the plant at stage 3 showed negative or no correlation with other observed parameters. Different variability parameters viz., mean, range, phenotypic variance, genotypic variance, environmental variance, coefficient of variation (CV), genotypic coefficient of variation (GCV), phenotypic coefficient of variation (PCV), heritability (BS), genetic advance (GA) and genetic advance as a percentage over mean of 200 genotypes are represented (Table 4). Contribution of genotypic variability to the phenotypic variations related to BNF traits is much higher than environmental variability. Particularly, contribution of environmental variation is too low for the early nodulation and late nodule senescence traits of certain chickpea genotypes. Phenotypic and genotypic coefficient of variations falls under high category (> 20) for BNF traits except nitrogen content that comes under Low for GCV (8.49) and medium for PCV (11.42). High broad-sense heritability with the value > 60 as described by Robinson (1966) was observed for BNF traits except nitrogen content that falls under medium category with the value of 55.32. PCA analysis indicated that dimensions 1 (21%), 2 (17.6%), and 3 (13%) accounted for a substantial portion of the phenotypic variance, each contributing more than 10%. These three dimensions, when considered collectively, accounted for an impressive 51% of the total observed phenotypic variance, showcasing their significant contribution (Fig. 8). Dimension 1 exhibited a contribution of more than 10% from variables, such as PDW stage 4, PDW stage 1, NFW stage 3, NFW stage 4, and NSW stage 4 (Fig. 9A). Similarly, dimension 2 demonstrated a contribution of more than 10% from variables PDW stage 1, PDW stage 2, NSW stage 4, NFW stage 4, NSW stage 3, and PDW stage 3 (Fig. 9B). Additionally, PCA analysis revealed the presence of a positive correlation between PDW stage 1 and PDW stage 2, as well as a positive correlation of NFW stage 3 with both NFW stage 4 and NSW stage 4 (Fig. 10). Consistent with the findings in correlation analysis, the PCA analysis also revealed a negative correlation between nitrogen content and plant dry weight at stage 1. Interestingly, both silhouette and gap statistic methods consistently predicted the presence of two major clusters in the dataset. Cluster 1 comprises 126 genotypes, while cluster 2 consists of 74 genotypes (Fig. 11). Cluster 1 is dominated by “desi chickpea” lines while cluster 2 is with pea-type chickpea. All chickpea genotypes originated from Turkey are grouped under cluster 2 (Supplementary Table 1).

Fig. 6.

Correlation values explaining the relationship among nodulation traits in chickpea:

Fig. 7.

CorrPlot explaining the relationship among nodulation traits in chickpea

Table 4.

Variability analysis for nodule number and fresh biomass, nodule specific weight, plant dry weight and nitrogen content of chickpea minicore

S. no.	Symbiotic nitrogen fixation-related traits	Crop stage	PV	GV	EV	GCV	PCV	ECV	hBS	GA	GAM
1	Nodule number (unit/plant)	Stage 1	10.34	7.9	2.44	53.54	61.24	29.73	76.43	5.07	96.56
		Stage 2	18.57	15.00	3.57	46.89	52.17	22.87	80.79	7.18	86.95
		Stage 3	41.95	37.17	4.77	58.65	62.30	21.01	88.62	11.84	113.90
		Stage 4	61.03	59.08	1.95	75.17	76.40	13.66	96.80	15.60	152.57
2	Nodule fresh weight (g/plant)	Stage 3	0.02	0.02	0.0046	84.60	94.80	42.78	79.64	0.25	155.75
		Stage 4	0.09	0.08	0.01	107.01	112.66	35.22	90.23	0.56	209.69
3	Plant dry weight (g/plant)	Stage 1	0.0029	0.0026	0.0003	37.39	39.44	12.57	89.84	0.10	73.11
		Stage 2	0.01	0.01	0.0007	36.41	37.92	10.59	92.19	0.19	72.12
		Stage 3	0.49	0.47	0.02	47.52	48.60	10.19	95.61	1.38	95.87
		Stage 4	5.75	5.70	0.05	48.08	48.30	4.58	99.10	4.90	98.74
4	Nitrogen content (%)	Stage 3	0.02	0.01	0.01	8.49	11.42	7.63	55.32	0.16	13.03

Fig. 8.

Scree plot with explained variances in each dimension

Fig. 9.

Variables contributing to Dimension 1 (A) and Dimension 2 (B)

Fig. 10.

Variable correlation plot PCA analysis for evaluating contribution of individual traits on nodulation and nitrogen fixing potential of chickpea germplasm

Fig. 11.

Grouping of chickpea minicore genotypes following k-mean clustering

Discussion

Nodulation potential of different grain legumes was assessed during past decades and found that nodule weight and nitrogen fixation varied significantly among the pulses grown across the years. The grain legumes like faba bean, field pea, and chickpea had the most nodules at early flowering i.e., 31.3, 27.9, and 27.5 NNPP, respectively (Hossain et al. 2016). Similarly, 16 green gram genotypes were assayed for variation in nodulation and found that number of nodules and nodule weight varied from 16.5 to 28.5 and 60.23 to 154.35 g per plant, respectively (Mishra et al. 2014). Several attempts were made by global researchers to assess the nodulation potential of chickpea. For example, a total of 15 chickpea varieties that include 3 kabuli and 12 desi types were evaluated for nodulation traits at the Gangetic New Alluvial Zone of West Bengal, India (Priyadarsini et al. 2017). Verma and Waldia (2013) screened 6 diverse genotypes belonging to desi, kabuli, non-nodulating genotypes along with their 15 F₁ derivatives. A genetically diverse subset consisting of 39 global accessions, from the USDA global chickpea core collection and commercial cultivar UC-5 was assayed for BNF potential in a greenhouse experiment (Biabani et al. 2011). These studies have their limitations since the assessment of limited number of germplasm lines did not reveal the true variability for a trait and generally chickpea cultivars represent poor genetic variability due to frequent use of same/similar set of parent lines in breeding programs. Few research groups have studied the nodulation potential of chickpea using significantly large numbers of germplasm lines. Local germplasms consisting of 200 Moroccan lines of chickpea were screened for nitrogenase activity, dry weight, nodule mass, and total nitrogen under different level of salt concentrations (Sadiki and Rabih 2001). In another attempt, Gopalakrishnan et al. (2017) screened chickpea mini-core consisting of 211 accessions developed from 1956 accessions of the core collection of chickpea, representing 16,991 accessions available in the ICRISAT gene bank (Upadhyaya and Ortiz 2001). Above studies were carried out using either nitrogen-free sand (Sadiki and Rabih 2001) or pot mixture consisting of equal proportion of sand and black soil (Gopalakrishnan et al. (2017) where the chickpea genotypes do not express their potential for nodulation. We had the opinion that above plant growth conditions along with introduced single or mixture of few Mesorhizobium strains might reveal under-representation of nodulation potential due to soil compaction, volume constraint in pots, and compatibility of introduced Mesorhizobium strain(s) with all minicore lines. It was evidenced by the large difference in nodulation potential of chickpea in pot experiments conducted at IIPR-Kanpur and ICRISAT-Hyderabad, India. A total of 39 chickpea accessions identified as non-nodulating type at ICRISAT produced nodules at IIPR, and interestingly 5 of them are considered as high nodulating genotypes with > 24 NNPP (Gopalakrishnan et al. (2017). By considering above facts, the present investigation was designed to screen chickpea minicore lines for nodulation potential under field strip having low soil P availability (7 ppm) with an additional aim to identify suitable chickpea genotypes that can be either used directly in the breeding program or utilized for revealing molecular basis of early/high nodulation potential.

Early nodulation and delayed nodule senescence are considered as desirable traits since they can extent the duration of support to the plants through symbiotic nitrogen fixation. Hence, the plant sampling was done at 15 and 90 DAS apart from 30 and 60 DAS and screened for nodulation status. A total of 14 chickpea lines were identified as early nodulators that produced more than 10 number of root nodules per plant. Due to abiotic stresses, many genotypes resulted in pre-mature senescence of nodules and observed with poor nodulation in subsequent crop growth stages. The contrasting genotypes viz., ICC 1356 (maximum NNPP i.e., 18 at S1 and 0 nodule at S2) and, ICC 506 (maximum NNPP i.e., 29 at S2 and 0 nodule at S1) along with other low as well as high nodulating genotypes can be used to reveal the variations in signaling mechanisms that determine early nodulation. Seven genotypes viz., ICC12968, ICC7867, ICC13816, ICC867, ICC15264, ICC15510, and ICC283 are common under high nodulation groups (> 10 NNPP) of stage 1 as well as stage 2. Biabani et al. (2011) reported a higher level of variation for nodule number that ranges from 21.0 to 101 NNPP in chickpea minicore of USDA global chickpea.

A total of 17 genotypes were identified as high nodulating type with more than 20 nodules/plant and the maximum number of nodules i.e., 36 per plant was observed in ICC6874. Two genotypes viz., ICC13816 and ICC12968 performed constantly and assigned under high nodulating groups at all three stages i.e., > 10 NNPP for S1, S2, and > 20 NNPP for S3. Eight genotypes viz., ICC12307, ICC14402, ICC6874, ICC12155, ICC4463, ICC9002, ICC12968, and ICC3325 are common for the high nodulation category (> 20 NNPP) of stage 3 and stage 4. However, these genotypes are not considered for selection since they produced poor nodule biomass ranging from 0 to 0.43 g/plant during stage 4 where the cut-off is > 1.0 g nodule biomass/plant. Four genotypes viz., ICC1431, ICC13599, ICC13764, and ICC13863 were common for the high nodule biomass category of both stage 3 (> 0.5 g/plant) and stage 4 (> 1.0 g/plant). The pink-colored nodules were observed even at 90 DAS in the above genotypes, and hence, they are considered as genotypes with delayed nodule senescence. Interestingly, none of the above listed genotypes except ICC1431 were considered as high nodulation group at all four stages of crop growth. The contrasting genotypes viz., ICC13599, ICC13764, and ICC13863 with high nodule biomass (0.53–2.29 g/plant) and ICC14402, ICC4463, and ICC3325 with low nodule biomass (0.09–0.28 g/plant) but having high nodule number per plant (> 20 NNPP) can be used to understand the variations in the mechanism of nodule organogenesis in chickpea. Screening of 200 Moroccan chickpea revealed the variability for nodule mass in the range of 60–120 mg/plant under normal environment and reduction of same due to salt stress is in the range of 30–67% (Sadiki and Rabih 2001). Biabani et al. (2011) reported that the variation for nodule mass ranges between 0.044–0.267 g/plant in chickpea minicore of USDA global chickpea. Another study indicated that Cluster-III consists of high nodulating genotypes recorded with 1.04 g of nodule weight and 33.24 nodules per plant (Verma and Waldia 2013).

Though it was an overestimation of numbers due to the process of collection, conservation, and exchange, global database indicated substantial collection and conservation of 85,865 cultivated and 1476 wild chickpea genotypes (Archak et al. 2016; Chandora et al. 2020). Core and minicore set of chickpea germplasm were developed at various gene banks located at ICRISAT, NBPGR, for the utilization of large collection of germplasm. Unfortunately, the below ground property of nodulation parameters was not considered while developing the core/minicore sets and hence a very meager portion of germplasm (0.24%) is assessed for variations in BNF-related traits. This study suggested a targeted approach to reveal large variations for a particular BNF trait. Desi chickpea is considered as a good source of variations for early nodulation with maximum of 21 NNPP at 15DAS (Table 1) while the NNPP at all 4 stages of crop growth is poor in pea shaped chickpea lines and followed by Kabuli-chickpea (Fig. 4). Priyadarsini et al. (2017) has also reported that desi types had better nodulating behavior including nodule number, nodule dry weight, and better economic yield potential in contrast to kabuli types. Hence, researchers can target desi genotypes for identifying better genetic resources to improve nodule number through introgression breeding. Many of the chickpea genotypes originated from Iran produced more number of nodules, which was indicated by comparatively higher values of adjusted median i.e., 6.11, 7.97, 9.85, and 10.90 at stage 1 to stage 4 respectively, while the large variability at all growth stages except stage 3 was observed with genotypes from India. Chickpea genotypes originated from Ethiopia is considered as poor nodulators.

In our study, a positive correlation was observed for nodule number per plant at different stages viz S2 vs S3 (0.27), S2 vs S4 (0.11), and S3 vs S4 (0.45). It also showed a positive correlation with plant dry weight only at early crop growth stages. NSW is strongly correlated with NFW at stage 3 (0.69) as well as stage 4 (0.76), but negatively correlated with nodule number. Biabani et al. (2011) have reported that nodule number at flowering stage was significantly correlated only with total nodule weight and nodule weight was correlated also with plant biomass. Other studies also indicated that no significant correlation of nodule number or weight with N fixation and preferred only limited numbers of nodules for maximum BNF since plant photosynthates may be most efficiently used for other biomass components (Hefny et al. 2001; Delić et al. 2010). Interestingly, the contribution of environmental variation is too low for the observed phenotypic variations in BNF traits and high broad-sense heritability with the value > 60 was observed for early nodulation, late nodule senescence, and high nodule number. Hence, identification of potential genetic resources for high nodulation, extended nitrogen fixation, and high symbiotic efficiency helps overcome the bottlenecks and makes grain legumes as nitrogen fixing factories to fertilize the soil in a sustainable way.

Conclusion

Present study indicated that desi chickpea as a good source of variations for early nodulation with the range of 0–21 NNPP at 15DAS. Accessions viz., ICC1431, ICC13599, ICC13764, and ICC13863 with pink active root nodules and high nodule biomass at later crop growth stages are considered as good genetic resources to extend the BNF support. Many of the chickpea genotypes originated from Iran are high nodulators, indicated by comparatively higher values of adjusted median. Contribution of genotypic variability to the phenotypic variations related to BNF traits is much higher than environmental variability. Particularly, contribution of environmental variation is too low for the early nodulation and late nodule senescence traits of certain chickpea genotypes. Hence, the identified chickpea lines for early high nodulation and delayed nodule senescence can be used to extend the BNF benefit to host plants under P-limited field condition.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization, designing experiments, data interpretation, and manuscript preparation were done by SKM and NKA. Field/pot experiments were carried out by RV. Genetic variability and other statistical analysis were done by PSS and RV. All authors reviewed the manuscript.

Funding

RV acknowledges Council of Scientific and Industrial Research-University Grant Commission- Junior Research Fellowship (Ref. No. 22/12/2013(ll)EU-V).

Data availability

Data that support the findings are included within this manuscript.

Declarations

Conflict of interest

The authors declare that there is no conflict of interest.