Evaluation of Drought Tolerance of Rice Landraces in Mid-Hills of Nepal Using Drought Tolerant Indices

doi:10.21203/rs.3.rs-4707154/v1

Download PDF

Research Article

Evaluation of Drought Tolerance of Rice Landraces in Mid-Hills of Nepal Using Drought Tolerant Indices

https://doi.org/10.21203/rs.3.rs-4707154/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Drought is a critical abiotic stress that affects rice yield. There is typically a limited and irregular pattern of rainfall in mid-hill regions, which leads to drought conditions that significantly impact rice production. Therefore, our study aimed to identify the best-performing landraces of rice and determine suitable drought indices. For these, 17 different rice landraces grown in the mid-hills were selected. A two-factorial (genotype and moisture stress) completely randomized design with two replications was carried out at the Lamjung Campus in the spring of 2023. Drought stress resulted in a wide range of yield reductions, from 10–69%. Our findings revealed significant differences in the agro morphological traits of rice. A strong and positive correlation was found between grain yield and drought indices, tolerance indices (TOLs), mean productivity indices (MPs), geometric mean productivity (GMP), harmonic means (HMs), yield indices (YIs), and stress tolerance indices (STIs) under various drought stress conditions. Among these indices, the Juhari landrace performed well in terms of MP, GMP, STI, HM, and YI. Principal component analysis (PCA) clustered drought indices into two groups (PC1 and PC2). Of these, PC2 was found to be better because of its high yield potential. Further screening using the ranking method and cluster analysis identified Juhari, Kathe, Manamure, and Pahele as high-quality drought-tolerant genotypes. Based on our findings, we recommend cultivating Juhari in drought-prone regions within the mid-hill region. In conclusion, our study highlights the importance of indices such as the MPI, STI, GMP, HM, and YI as important instruments for drought screening. Our study highlighted the importance of choosing suitable drought-tolerant landraces that exhibit improved yield and stability under adverse conditions.

Abiotic stress

Drought indices

Drought tolerance

Rice landraces

Yield loss

Drought is a prominent nonliving factor that diminishes rice production in regions reliant on rainfall and susceptible to drought. Rain-dependent rice production accounts for 57% of the overall rice-growing area in Nepal, and an estimated 0.8 million ha of land is regarded as drought-prone. (Yuan et al., 2022). Rice plants are exceptionally responsive to insufficient soil moisture, particularly during crucial growth phases, primarily before flowering and during grain filling. Drought, a noteworthy nonliving factor, leads to decreased rice production in regions dependent on rain and susceptible to drought (Adhikari et al., 2019; Hassan et al., 2023). Drought-induced losses are especially severe in South Asian countries, including Nepal, India, Bangladesh, and Myanmar (MoAD, 2015/16). In Nepal, approximately 44% of the regions dedicated to rice cultivation depend on monsoon rainfall for successful farming (Tripathi et al., 2019). The significant variability in the timing of rainfall across multiple decades has resulted in an annual decrease in the cultivated area, creating a significant obstacle to rice agriculture.

Plant responses to drought are extensively documented and thought to be complex, involving numerous physiological, biochemical, and molecular changes (Atkinson and Urwin, 2012; Bargaz et al., 2015). Tolerance to drought stress is thus the result of multiple traits expressed throughout the stress. As a result, it is doubtful that a single characteristic can greatly improve crop yield in response to water scarcity (Farooq et al., 2009; Kamoshita et al., 2008). A number of attributes associated with rice drought performance have been identified, including root structure, root penetration and distribution, leaf curling, reduced leaf area, early flowering and seed maturation, osmotic adjustment (accumulation of compatible solutes such as proline and soluble sugars), and increased production of abscisic acid (ABA), which leads to stomatal closure (Bimpong et al., 2011; Price et al., 2000). The identification and incorporation of these features into breeding programs could help to ensure long-term rice production in areas susceptible to drought (Nguyen et al. 1997).

Identifying drought-tolerant rice cultivars is critical for sustaining rice yield in drought-prone locations. A comparison of how different genotypes perform under conditions with and without drought stress can serve as a useful tool for identifying drought-resistant varieties suitable for drought-prone areas. Numerous indices of drought have been suggested, each based on mathematical relationships among yields under drought-affected and nonstressed conditions. TOL, MP, GM, STI, and SSI are among the indicators used to measure a cultivar’s drought tolerance. GM, HARM, GMP, STI, REI, YI, ATI, MRP, and SNPI are the best drought indices for selecting upland rice cultivars that are both drought-tolerant and high-yielding (Sabouri et al., 2022). Sixteen rice genotypes of medium maturity duration, known to possess QTLs associated with drought tolerance, were assessed to pinpoint drought-tolerant genotypes (Singh et al., 2018). This evaluation involved the utilization of drought tolerance indices, viz. DTE, TOL, SSI, and STI (Kumar et al., 2014). The STI provides additional benefits for selecting appropriate cultivars under both stressful and nonstressful conditions (Nouraein et al., 2013). Drought-tolerant rice cultivars have low SSI and TOL values, as well as high STI values (Adhikari et al. 2019). These indices categorize genotypes as either drought-resistant or drought-susceptible (Raman et al., 2012).

This research is driven by the recognition of several critical issues within the region's rice ecosystem. However, there is a lack of complete information on the drought tolerance of rice landraces adapted to the unique agroecological conditions found in the mid-hills. The difference in suggested drought indices for rice in the Lamjung District is another problem. Despite the diversity of rice genotypes cultivated in this region, the capacity of these plants to withstand drought conditions has not been fully explored. Therefore, drought tolerance indices that can be used to efficiently and reliably measure rice genotype susceptibility to drought stress are needed. The absence of standardized methodologies for evaluating drought tolerance in rice landraces further complicates the selection of suitable varieties for cultivation in this unique environment. Furthermore, the potential mismatch between the suitability of rice landraces and the local agroecological conditions of the mid-hills is a significant concern. These discrepancies can lead to suboptimal rice yields and, consequently, affect the livelihoods of farmers who heavily depend on rice cultivation. In addition to these challenges, the region has insufficient water resources and recurrent periods of low rainfall, both of which have profound impacts on vegetative growth rates and overall crop yields (Mostajeran and Eichi., 2009).

Thus, our research aimed to analyze rice tolerance to drought using drought tolerance indices. This approach assists in identifying the important morphophysiological traits and parameters influencing drought tolerance among the studied rice genotypes. The major purpose of this study was to determine the relationships between yields and various drought tolerance indices.

2.1 Research location

A pot experiment investigating rice drought tolerance at the Institute of Agriculture and Animal Science (IAAS), Lamjung Campus, Nepal, was conducted from February to August. The study site, which is situated at an elevation of 610 m above sea level, falls within the subtropical climate zone of Nepal and is characterized by an annual average rainfall of 203 mm. The geographical coordinates of the experiment were approximately 28.1448°N and 84.4120°S.

2.2 Experimental design

Seventeen landraces of rice were grown under soil field capacity (FC), 60% FC and nonstressed conditions (irrigated conditions) in two factorials completely randomized design (CRD) with two replications. The study was performed in 102 plastic pots (17 × 3 × 2). The two factors of treatment are as follows:

a. Soil moisture regimes

Saturation throughout the growing period (nonstressed condition)

Field capacity (FC) during the reproductive stage to harvest

60% of field capacity during the reproductive stage to harvest

b. Landraces: 17 landraces of rice

2.3 Treatment details

In this research, 17 rice landraces were collected from the Purkot Community Seed Bank, Tanahun and Ghanpokhara Community Seed Bank, Lamjung. Table (1) shows the landraces used in the experiment.

Table 1

The 17 landraces of rice used in the experiment
SN.	Name of the Genotype	SN.	Name of the Genotype
1.	Aapjhutte	10.	Kalo Patlae
2.	Kathe	11.	Pamali
3.	Lekali basmati	12.	Darmali
4.	Manamure	13.	Juhari
5.	Gaure	14.	Kalo Jhinuwa
6.	Kartike	15.	Pahele
7.	Rato Anadi Lamcho	16.	Pahelomansar
8.	Piyale	17.	Dalle
9.	Jhumka

2.4 Cultivation practices

A 2 m × 1 m long seedbed was prepared in the quarter area of the Lamjung campus, Sundarbazar, from 20th February to 10th July 2023. The bed was dug one week before sowing and kept open for soil solarization. The soil was fine and mixed with the well-rotted FYM. After completion of land preparation, 17 rows were prepared where presoaked rice seeds (for 24 hours) were sown in rows with a 5 cm row-to-row distance and 2 cm depth. Tags containing the names of the 17 different genotypes were prepared using ribbons and short bamboo sticks, which were subsequently placed at the ends of the respective rows. The first watering session was performed immediately after sowing, and the plants were watered daily until the plants were ready to be transplanted into pots (at the 3–5 leaf stage).

First, the bottom drainage hole of each pot was stuffed with putting to prevent loss of water during irrigation and maintain submerged conditions. The soil used as a planting medium for filling the pots was collected from a nearby field, and well-decayed FYM was obtained from the Lamung Campus animal farm. The soil and FYM were combined at a 2:1 ratio on an open surface, and each pot was filled with an equal amount (4 kg/pot) of the prepared media. Chemical fertilizers such as urea, DAP and MOP were added to each media-filled pot at proportions of 1 g, 0.78 g and 0.6 g, respectively. Later, at the peak tillering stage and panicle initiation stage, a split dose of nitrogen was applied. After 30 days of sowing, the nursery bed was irrigated before the plants were uprooted to avoid root injury. After that, single plants of each rice genotype were transplanted per pot on March 20, 2023.

Just after transplanting, water was applied to all 102 pots to maintain saturation. The plants were maintained for 15 days until the completion of the transplanting shock of the rice plants up to the panicle initiation stage. After that, water was added at 3 different moisture levels, namely, the soil saturation level, FC and 60% FC. The amount of water for these three conditions was determined by using a gravimetric method as follows:

In the gravimetric method, a pot was filled with a soil sample taken from the same field. The plants were saturated by adding water and kept in an open field covering the mouth with a plastic sheet for 24 hours. This process was performed to prevent evaporation loss. After 24 hours, the soil reaches FC after losing its gravitational water. The weight of the pot containing the soil was subsequently recorded.

The weight of the pot (Wp) = 145 gm = 0.145 kg

Weight of air-dried media (Wa) = 3.85 kg

Weight of oven-dried media (Wd) = 2.72 kg

Out of 3.85 kg (Wa) of our planting media, the weight of soil solid only (Wd) was 2.72 kg, which results in a weight of soil water (Ww) = Wa-Wd= (3.85–2.72) kg = 1.135 kg.

Weight of soil at FC (Wf) = 4.52 kg

where the soil solid concentration (Wd) = 2.72 kg

Therefore, the soil water at FC (Wc) = Wf-Wd= (4.52–2.72) kg = 1.8 kg

Therefore,

Amount of water to be added to bring the media to 100% FC (W1),

= Weight of soil water at FC (Wc)- weight of water already prevailing in the media (Ww)

= 1.8-1.135 kg = 0.67 kg = 670 gm

Similarly,

Amount of water to be added to bring our media to 60% FC (W2) = 60% of 670 gm

= 670×60/100 gm

= 402 gm = 0.4 kg

With a density of water of 1 g/cc, 0.67 litres of water was initially added to the soil. Therefore, the final weight of our pot media at FC became (Wp + Wa + W1) = (0.145 + 3.85 + 0.67) kg = 4.665 kg, and for 60% FC, it became (Wp + Wa + W2) = (0.145 + 3.85 + 0.4) = 4.395 kg. For subsequent irrigation, the pots were weighed, and the appropriate amount was added to the plants to obtain the final weight.

2.5 Observation and data collection

Plant height (PH), number of total tillers per plant (TT), number of effective tillers per plant (ET), number of noneffective tillers per plant (NET), panicle length (PL), number of total grains per panicle (TGP), number of filled grains per panicle (FGP), number of unfilled grains per panicle (UGP), 100-grain weight (100-GW), seed filling percentage (SFP), and grain yield per plant (GYP) were observed for the rice genotype.

Yield data were collected from each plant growing in a single plastic pot. The mature plants were collected manually. The moisture content was subsequently measured using a moisture meter. The grain yield per plant (grams) was determined under fresh moisture conditions using the equation below.

Grain yield (gm/plant) =\(\:\frac{Fresh\:weight\:of\:harvested\:rice\left(gm\right)\times\:\left(100-MC\right)\times\:Shelling\:Coefficient}{100-Desirable\:Moisture\:\%\:of\:grain\:for\:storage}\)

where MC\(\:=\)Moisture content of the grains at harvest (in %)

The shelling coefficient of rice (0.7), also known as the milling percentage, is the ratio of the weight of milled rice (the edible portion) to the weight of paddy rice (the whole, unprocessed rice grains). The percentage of the paddy rice that remained edible after the milling process was calculated. The desirable moisture percentage of the grains for storage is 12.5%.

The formulae in Table (2) were used to calculate eight drought stress tolerance indices, TOL, MPI, GMP, STI, YSI, SSI, HM, and YI, based on grain yield under nonstressed (Yp) and drought stress (Ys) situations.

Table 2

Drought tolerance indices
SN.	Drought tolerance indices	Equations	References
1.	Tolerance Index (TOL)	TOL = Yp −Ys	(Rosielle & Hamblin, 1981)
2.	Mean Productivity Index (MPI)	MPI=\(\:\frac{Yp+Ys}{2}\)	(Rosielle & Hamblin, 1981)
3.	Geometrical Mean Productivity (GMP)	MP =\(\:\sqrt{\left(Yp\right)\left(Ys\right)}\)	(Adhikari et al., 2019) (Fernandez, 1992)
4.	Stress Tolerance Index (STI)	STI=\(\:\frac{\left(Yp\right)\left(Ys\right)}{{\left(\stackrel{-}{Y}p\right)}^{2}}\)	(Fernandez, 1992) (Kandel et al., 2022)
5.	Yield Stability Index (YSI)	YSI = Ys/Yp	(Bouslama & Schapaugh Jr, 1984)
6.	Stress Susceptibility Index (SSI)	SSI =\(\:\frac{1-(Ys/Yp)}{SI}\) \(\:\text{S}\text{I}\hspace{0.17em}=\hspace{0.17em}1-\:(\overline{Y}\text{s}/\:\overline{Y}\text{p})\)	(Fischer & Maurer, 1978)
7.	Harmonic Mean (HM)	HM=\(\:\frac{2(Yp\times\:Ys)}{(Yp+Ys)}\)	(Doorenbos & Kassam, 1979) (Jafari et al., 2009)
8.	Yield Index (YI)	YI=\(\:\frac{{Y}_{s}}{{\stackrel{-}{Y}}_{s}}\)	(Gavuzzi et al., 1997)
Note: Yp-grain yield under non-stress conditions; Ys-grain yield under stress conditions; \(\:\overline{Y}\)p- average yield under non-stress conditions; \(\:\:\overline{Y}\)s- average yield under stress conditions

2.6 Genotype ranking

Different indices classify different genotypes as drought-resistant; thus, finding drought-tolerant genotypes based on a single criterion does not produce definitive results. To select promising drought-tolerant genotypes, the mean rank, standard deviation of rankings, and rank sum of the indicators were computed. The following relationship formula was used to determine the rank sum (RS) for screening drought-tolerant genotypes:

Rank Sum (RS) = Rank Mean (R) + Standard Deviation of Rank (SDR) (Farshadfar et al., 2011)

SDR= (S² i)^0.5

2.7 Statistical analysis

Several drought tolerance indices were calculated using the technique in Table (2), which is based on a mathematical relationship between yield in drought and nonstressed conditions. The data were subsequently entered into MS Excel (2019) and examined using RStudio software (version 2022.02.2). PCA and cluster analysis were performed by using Minitab (version 21.1, 64-bit). The correlations between drought tolerance indices and plant yield, as well as ANOVA, were computed using the RStudio program. The use of parametric tests and CRD allowed us to draw meaningful conclusions from our data while considering the interactions between two key factors of interest.

3.1 Mean separation of yield and yield-attributing characteristics

Plant Height (PH): All 17 rice landraces exhibited extremely significant variance in plant height, with a mean of 99.44 cm. The plant height ranged from 128 cm in Kartike to 55 cm in Pahelo Mansar. In terms of the influence of moisture level on plant height, the maximum plant height at saturation was determined to be 109 cm, which differed significantly from the plant height at FC. Similarly, a significant difference was found in plant height between the FC treatment and 60% of the FC treatment. However, the association between genotype and moisture conditions was determined to be nonsignificant. (Table 3).

Total Tiller Number per Plant (TN): All 17 rice landraces exhibited significant differences in tiller number, with a mean of 18. The number of tillers ranged from 30 in Pahelo Mansar, which is significantly on par with that in the genotype Aapjhutte, to the lowest mean value of 8.5 in Gaure. Regarding the effect of moisture level on tiller number, the maximum tiller number was found to be 19 at saturation, which was statistically at par with the tiller number at FC. No significant difference was found in the tiller number between the FC treatment and 60% FC treatment groups.

Effective Tiller per Plant (ET): All 17 rice landraces had significantly different tiller numbers, with a grand mean of 7. The number of effective tillers ranged from 18 in Juhari to 3 in Piyale. Regarding the effect of moisture level on the effective tiller, the maximum effective tiller was found to be 8 under saturated conditions. A significant difference was found in the number of effective tillers at saturation, at FC and at 60% FC. However, the interaction effect between genotype and irrigation conditions was found to be significant.

Non-effective Tiller per Plant (NET): All 17 rice landraces exhibited extremely significant variances in terms of non-effective tiller number, with a mean of 11. The number of non-effective tillers ranged from 22 in Aapjhutte to 3.33 in Juhari. Regarding the effect of moisture level on the non-effective tillers, there were 11 non-effective tillers in terms of saturation, FC and 60% FC. However, the interaction effect between genotype and irrigation conditions was not significant.

Panicle length (PL): All 17 landraces of rice exhibited highly significant differences in terms of panicle length, with a mean of 23.4 cm. The panicle length ranged from 34.4 cm in Kalo Jhinuwa to 12.6 cm in Pahelo mansar. In terms of the influence of moisture level on panicle length, the greatest panicle length was determined to be 26 cm at saturation, which differed significantly from the panicle length at 60% FC. Similarly, a significant difference was found in panicle length between the FC treatment and 60% of the FC treatment. However, the interaction effect between genotype and irrigation conditions was found to be significant.

Panicle weight (PW): All 17 landraces of rice exhibited highly significant differences in terms of panicle weight, with a mean of 2.7 gm. The panicle weight ranged from 3.6 g in Piyale to 1.5 g in Pahelo Mansar. Regarding the effect of moisture level on panicle weight, the maximum panicle weight was 3.1 gm at saturation, which was significantly different from the panicle weight at 60% FC. Similarly, a significant difference was found in the panicle weight at FC and 60% of FC. On the other hand, the interaction effect between genotype and moisture conditions was highly significant.

Total Grain per Panicle (TGP): All 17 landraces of rice exhibited highly significant differences in terms of total grain per panicle, with a mean of 199.99. The total grain per panicle ranged from 270.99 in Kartike to 117.22 in Pahelo Mansar. Regarding the effect of moisture level on the total grain per panicle, the maximum total grain per panicle was 230.92 at 60% FC, which was significantly different from the total grain per panicle at saturation and FC. However, no significant difference was found in the total grains per panicle at FC or saturation. On the other hand, the interaction effect between genotype and moisture conditions was highly significant.

Filled Grain per Panicle (FGP): All 17 landraces of rice exhibited highly significant differences in terms of filled grain panicle-1, with a mean of 117.52. The filled grain panicle-1 ranged from 155 in Kartikae to 69 in Pahelo Mansar. In terms of the influence of moisture level on the number of filled grains per panicle, the maximum number of filled grains per panicle was 127 at saturation, which was significantly different from the total number of grains per panicle at FC and 60% FC. However, no significant difference was found in Filled-grain Panicle-1 at FC or 60% FC. On the other hand, the interaction effect between genotype and irrigation conditions was highly significant.

Unfilled Grain per Panicle (UGP): All 17 landraces of rice exhibited highly significant differences in terms of unfilled grain panicle-1, with a mean of 82.47. The number of unfilled grains per panicle ranged from 116 in Kartike to 48.02 in Pahelo mansar. Regarding the effect of moisture level on the unfilled grain panicle-1, the maximum unfilled grain panicle-1 was found to be 120 at 60% FC, which was significantly different from the unfilled grain panicle-1 at FC. Similarly, a significant difference was found in unfilled-grain Panicle-1 at FC and saturation. On the other hand, the interaction effect between genotype and moisture conditions was highly significant.

Seed Filling Percentage (SFP): No significant difference was found between rice landraces in terms of fertility percentage. Regarding the effect of moisture level on the fertility percentage, the maximum fertility percentage was found to be 69.93 at saturation, which was statistically at par with the fertility percentage at FC. Similarly, a significant difference was found in the fertility percentage between the 60% FC and FC groups. However, the association between genotype and moisture conditions was determined to be nonsignificant.

100-GW (HGW): All 17 landraces of rice exhibited highly significant differences in terms of 100-GW, with a mean of 2.29 gm. The 100-GW concentration ranged from 3.39 gm in Rato Anadi Lamcho to 1.55 gm in Jhumka. Regarding the effect of moisture level on 100-GW, the maximum 100 GW was found to be 2.47 gm in saturation, which was significantly different from that of 100-GW at FC. Similarly, a significant difference was found in the 100-GW at FC and 60% FC. On the other hand, the interaction effect between genotype and moisture conditions was highly significant.

Grain yield/plant (GYP): All 17 landraces of rice exhibited highly significant differences in terms of grain yield plant-1, with a mean of 17.59 gm. The grain yield plant-1 ranged from 41.72 gm in Juhari to 4.46 gm in Kalo Jhinuwa. Concerning the influence of moisture level on grain yield/plant, the maximum grain yield per plant was determined to be 23.96 gm at saturation, which was significantly different from the grain yield per plant at FC. Similarly, there was a considerable difference in grain yield (GY)/plant between FC and 60% FC. On the other hand, the interaction between genotype and moisture conditions was determined to be highly significant.

Table 3

Agro-morphological and yield attributing traits of rice landraces under different moisture regimes
Factor A (Landraces)	Plant Height (cm)	Total Tiller Plant^− 1	Effective Tiller Plant^− 1	Non Effective Tiller Plant^− 1	Panicle Length (cm)	Panicle Weight (gm)	Total Grain Panicle^− 1	Filled Grain Panicle^− 1	Unfilled Grain Panicle^− 1	Seed filling Percentage (%)	100-GW	Grain Yield Plant^− 1
Appjhutte	82ⁱ	27^ab	5^e	22^a	18.6^gh	1.7^g	171.78^cd	99^cd	73^fg	59.69^abc	1.76ⁱ	9.35^f
Kathe	111^cde	27^abc	9^d	18^b	27.8^c	2.9^cde	217.74^b	130^b	87^cdef	60.79^ab	2.19^efg	24.55^c
Lekali Basmati	108^de	15^ef	8^d	7^efg	20.4^fg	3.0^bcd	241.21^b	132^b	109^ab	57.02^c	2.28^cde	34.97^c
Manamure	95^fgh	23^cd	10^c	13^cd	21.2^f	2.6^ef	189.50^c	112^c	77^def	60.05^abc	2.32^cd	26.80^c
Gaure	109^de	9^h	3^ghi	6^fg	21.6^f	2.9^cde	216.34^b	128^b	89^cde	60.35^abc	2.23^def	7.77^fg
Kartike	128^a	11^fgh	4^ef	7^efg	25.1^de	3.7^a	270.99^a	155^a	116^a	58.62^bc	2.37^c	15.60^e
Rato Anadi Lamcho	123^abc	13^efg	3^fgh	10^cde	22.8^ef	3.0^bcd	166.00^cde	92^de	74^efg	58.15^bc	3.39^a	8.76^f
Piyale	104^def	16^e	3^hi	13^c	34.3^a	3.6^a	231.32^b	139^b	93^cd	61.29^ab	2.62^b	9.14^f
Jhumka	65^j	11^fgh	5^e	6^fg	18.0^h	1.3^h	143.73^e	83^e	61^gh	59.07^abc	1.55^j	6.91^fg
KaloPatle	82ⁱ	15^efg	8^d	7^efg	20.5^fg	2.5^ef	159.40^de	99^cd	60^gh	62.08^a	2.57^b	19.32^d
Pamali	101^efg	14^efg	4^efg	10^def	30.6^b	3.1^bc	220.25^b	131^b	89^cde	60.04^abc	2.34^c	12.92^e
Darmali	92^fgh	24^bcd	7^d	17^b	18.6^gh	2.6^ef	184.87^cd	110^c	75^efg	60.27^abc	2.37^c	19.34^d
Juhari	87^hi	21^d	18^a	3^g	21.6^f	2.3^f	181.20^cd	109^c	73^fg	60.16^abc	2.12^gh	41.72^a
KaloJhinuwa	124^b	11^gh	1ⁱ	9^def	34.4^a	3.2^b	228.55^b	140^b	89^cde	60.94^ab	2.30^cd	4.46^g
Pahele	114^bcd	23^bcd	10^c	13^cd	24.4^de	3.2^bc	234.53^b	138^b	96^bc	59.90^abc	2.27^cde	33.63^b
PaheloMansar	55^j	30^a	13^b	13^c	12.6ⁱ	1.5^gh	117.22^f	69^f	48^h	59.66^abc	2.16^fgh	19.60^d
Dalle	111^de	16^e	5^e	11^cde	26^cd	2.7^de	225.28^b	131^b	94^c	59.62^abc	2.07^h	14.26^e
SEm (±)	3.91	1.33	0.46	1.17	0.76	0.12	8.54	4.66	4.87	1	0.032	1.24
LSD	11.11	3.79	1.31	3.33	2.17	0.33	24.25	13.23	13.81	2.84	0.09	3.52
CV	9.64	18.26	16.63	25.93	7.98	10.60	10.46	9.71	14.45	4.09	3.44	17.28
F test	***	***	***	***	***	***	***	***	***	NS	***	***
Factor B (Moisture level)
Saturation	109^a	19^a	8^a	11^a	26.0^a	3.1^a	180.71^b	127^a	54^c	69.93^a	2.47^a	23.96^a
Field Capacity (FC)	98^b	18^ab	7^b	11^a	23.3^b	2.7^b	188.35^b	115^b	74^b	61.22^b	2.31^b	16.81^b
60% of FC	92^c	17^b	6^c	11^a	21.0^c	2.3^c	230.92^a	111^b	120^a	48.44^c	2.09^c	12.01^c
SEm (±)	1.64	0.56	0.19	0.49	0.32	0.05	3.59	1.96	2.04	0.42	0.01	0.52
LSD	4.66	1.59	0.55	1.40	0.91	0.14	10.19	5.56	5.80	1.19	0.04	1.48
F test	***	**	***	ns	***	***	***	***	***	***	***	***
Interaction (A×B)	ns	ns	**	ns	*	***	***	***	***	ns	***	***
Grand Mean	99.44	18	7	11	23.4	2.7	199.99	117.52	82.47	59.86	2.29	17.59
Note: , and * significant at P = 0.05, P = 0.01 and P = 0.001, respectively, and ns = nonsignificant*

3.2 Comparing rice landraces based on drought tolerance indices

Various drought tolerance indices most likely evaluate comparable aspects of drought tolerance or resistance. Grain yield was measured in rice plots grown under both nonstressed (irrigated) and drought-stressed conditions (FC and 60% FC). According to Table (4) and Table (5), the mean grain yield under non-stress conditions was the greatest (23.97 gm/plant), indicating potential yield (YP). The average yield at FC (16.81 gm/plant) was 29.88% lower than that at YP, while at 60% FC (12.01 gm/plant), it was 49.91% lower.

3.2.1 Tolerance index (TOL)

Table 4 shows that rice plant genotypes such as Appjhutte (1.12), Gaure (1.21), Rato Anadi Lamcho (1.54), Piyale (2.29), Dalle (3.28), Lekali Basmati (4.02), Pahelo Mansar (4.06), Kalo Jhinuwa (4.18), and Jhumka (4.34) had lower TOL values. In contrast, genotypes such as Pahele (31.67), Juhari (20.56), and Kathe (10.02) exhibited higher TOL values, suggesting that these plants are unsuitable for drought conditions. Several researchers have observed similar results when choosing genotypes based on these indices (Pantuwan et al., 2002; Ouk et al., 2006; Sio-Se Mardeh et al., 2006).

As indicated in Table 5, lower TOL values were identified for the Kalo Jhinuwa (3.04), Appjhutte (4.15), Rato Anadi Lamcho (4.68), and Piyale (4.82) genotypes. Conversely, higher TOL values were observed for the Pahele (39.21), Juhari (31.61), Kathe (19.55), Manamure (17.36), and Pahelo Mansar (13.22) genotypes, signifying their limited suitability for drought conditions. All the other genotypes displayed moderate TOL values.

3.2.2 Mean Productivity Index (MPI)

As shown in Table 4, the Jhuwari (48.83), Pahele (41.42), Manamure (31.60), and Kathe (29.40) genotypes displayed higher values, indicating tolerance, while the Kalo Jhinuwa (4.78), Jhumka (8.21), and Gaure (9.34) genotypes exhibited lower values, indicating susceptibility to stress. As shown in Table 5, the Jhuwari (43.30), Pahele (37.65), Manamure (25.87), and Lekali Basmati (25.24) genotypes demonstrated higher values, suggesting tolerance, whereas the Kalo Jhinuwa (5.34), Gaure (7.29), and Jhumka (7.35) genotypes presented lower values, suggesting susceptibility to stress. All remaining genotypes fell within the intermediate category.

3.2.3 Geometric mean productivity (GMP)

As shown in Table 4, the Juhari (47.73), Pahele (38.27), and Manamure (31.46) genotypes exhibited the highest GMP values, indicating their ability to withstand drought conditions. Conversely, lower GMP values were observed in Kalo Jhinuwa (4.29), Jhumka (7.92), and Gaure (9.32), suggesting their susceptibility to drought stress. As shown in Table 5, the Juhari (40.32), Pahele (32.14), and Lekali Basmati (25.04) genotypes had the highest GMP values, demonstrating their resilience to drought. In contrast, lower GMP values were reported for Kalo Jhinuwa (5.12), Jhumka (6.69), and Gaure (6.80), indicating vulnerability to drought stress.

3.2.4 Stress tolerance index (STI)

Drought tolerance was demonstrated in association with increased STI levels. As shown in Table 4, a high STI suggests the potential to withstand moisture stress in plants of several genotypes, such as Juhari (3.97) and Pahele (2.55). Conversely, a low STI suggested diminished tolerance to moisture stress, as observed in Kalo Jhinuwa (0.03), Jhumka (0.11), Gaure (0.15), and Rato Anadi Lamcho (0.18). As shown in Table 5, a high STI signifies resilience to moisture stress in genotypes such as Juhari (2.83) and Pahele (1.80). On the other hand, low values denote reduced tolerance to moisture stress, as evidenced by Kalo Jhinuwa (0.05), Jhumka (0.08), Gaure (0.08), and Rato Anadi Lamcho (0.12).

3.2.5 Yield Stability Index (YSI)

As shown in Table (4), the genotypes Appjhutte (0.90) and Gaure (0.88), closely followed by Lekali Basmati (0.86) and Rato Anadi Lamcho (0.86), exhibited the highest YSI values, demonstrating their resilience to stress. These findings align with previous research by many researchers (Garrity and O'Toole 1995, Ouk et al., 2006, Kumar et al., 2008, and Raman et al. 2012). Similarly, as shown in Table 5, Lekali Basmati (0.77), Darmali (0.67), Kalo Patle (0.64), and Apppjhutte (0.63) had the highest YSI values, indicating their stability under stress conditions. However, lower levels of stress indicated a lack of stability, whereas the other genotypes exhibited moderate stress.

3.2.6 Stress Sensitivity/Susceptibility Index (SSI)

Genotypes with lower SSI values are more resistant to drought. As Table (4) illustrates, genotypes such as Appjhutte (0.34), Gaure (0.41), Lekali Basmati (0.47), Rato Anadi Lamcho (0.48) and Pahelo Mansar (0.54) exhibited the lowest SSI values, indicating increased resistance to drought. On the other hand, the Kalo Jhinuwa (2.04) and Pahele (1.85) genotypes were the most susceptible, followed by the Jhumka (1.40) and Pamali (1.31) genotypes. All the other genotypes exhibited drought resistance within the intermediate range. Table 5 shows that all the genotypes with lower SSIs were drought resistant, including Lekali Basmati (0.45), Darmali (0.67), and Kalo Patle (0.73).

3.2.7 Harmonic mean (HM)

According to Table 4, Juhari (46.66) and Pahele (65.37) emerged as the most appropriate genotypes for HM, whereas Kalo Jhinuwa (3.86) and Jhumka (7.64) had the lowest values for these indices. Similarly, according to Table 5, the most suitable genotypes for HM were Juhari (37.54) and Pahele (27.44), whereas Kalo Jhinuwa (4.91) and Jhumka (6.10) had the lowest scores on these indices.

3.2.8 Yield index (YI)

As shown in Table 4, regarding STI cross-testing for genotypes suitable for drought conditions, Juhari (2.29), Manamure (1.70), Pahele (1.52), and Lekali Basmati (1.45) exhibited higher YI values, indicating their suitability. Conversely, lower YI values were observed in Kalo Jhinuwa (0.16), Jhumka (0.36), Gaure (0.52), Piyale (0.55), and Rato Anadi Lamcho (0.55), indicating their susceptibility to drought. According to Table 5, the Juhari (2.29), Lekali Basmati (1.84), Pahele (1.50), Manamure (1.43), and Darmali (1.33) genotypes exhibited higher YI values, suggesting their suitability. Conversely, lower YI values were noted in Kalo Jhinuwa (0.32), Jhumka (0.36), Gaure (0.39), Rato Anadi Lamcho (0.51), and Piyale (0.56), indicating their susceptibility to drought. All the other genotypes in Table (6) also fell within the intermediate range.

3.3 Yield reduction by drought

Evaluating the influence of drought stress on rice landrace yield reduction is a useful tool for determining how a cultivar's yield changes in FC and 60% of FC vs non-stress situations.

3.3.1 Yield reduction percentage under FC

Crop stability is frequently connected with a cultivar's capacity to function efficiently under stress, with lower yield decreases suggesting greater drought tolerance. Table (4) shows that the most substantial yield reduction occurred in Kalo Jhinuwa (60.84%), followed by Pahele (55.31%), Jhumka (41.76%), and Pamali (39.15%). Conversely, Appjhutte, Gaure, and Lekali Basmaki exhibited remarkable performance under FC, with the lowest percentage of yield loss due to drought recorded for Appjhutte (10.07%), followed closely by Gaure (12.14%). The range of yield reduction caused by drought conditions in our study spanned from 10% to 61%, with an average of 27.55%. These findings align with those of Singh et al. (2018); however, for different cultivars, a 35% mean yield decrease was reported, which is comparable to our 27.55% average yield reduction. Rashidi et al. (2011) reported a greater range of yield reductions in wheat caused by drought, varying from 13% to 76%, with an average of 55%, which was slightly greater than that recorded in paddy fields. Hooshmandi (2019) reported an average decrease of 34.03% in the yield of wheat under stressful conditions, which is similar to the 27.55% average yield reduction in rice. Consequently, Appjhutte, Gaure, and Lekali Basmati are considered drought-tolerant landraces because they exhibited the lowest yield reduction under drought conditions.

3.3.2 Yield reduction percentage under 60% FC

Table 5 shows that the greatest degree of yield reduction was observed in Pahele (68.49%), followed closely by Pamali (63.62%) and Jhumka (58.43%). Conversely, Appjhutte, Gaure, and Lekali Basmaki exhibited outstanding performance under 60% FC, with the lowest percentage of yield loss due to drought recorded for Lekali Basmati (22.55%), which was trailed by Darmali (33.46%). The range of yield reduction attributable to drought conditions within our study spanned from 22% to 69%, with an average of 47.6%. Consequently, Lekali Basmati, Darmali, and Kalo Patle are recognized as drought-tolerant landraces because they exhibited the least yield decrease in response to drought stress.

Table 4: Mean grain yield and 10 drought tolerance indices for 17 rice genotypes grown under nonstressed conditions and FC during the reproductive stage

Genotypes	Yp	Ys	TOL	MPI	GMP	STI	YSI	SSI	HM	YI	Yield loss%
Appjhutte	11.11	9.99	1.12	10.55	10.54	0.19	0.90	0.34	10.52	0.59	10.07
Kathe	34.41	24.39	10.02	29.40	28.97	1.46	0.71	0.97	28.54	1.45	29.11
Lekali Basmati	28.45	24.43	4.02	26.44	26.36	1.21	0.86	0.47	26.28	1.45	14.14
Manamure	34.55	28.65	5.91	31.60	31.46	1.72	0.83	0.57	31.32	1.70	17.10
Gaure	9.94	8.73	1.21	9.34	9.32	0.15	0.88	0.41	9.30	0.52	12.14
Kartike	21.23	14.24	6.99	17.73	17.39	0.53	0.67	1.10	17.05	0.85	32.92
Rato Anadi Lamcho	10.83	9.29	1.54	10.06	10.03	0.18	0.86	0.48	10.00	0.55	14.25
Piyale	11.50	9.21	2.29	10.36	10.30	0.18	0.80	0.67	10.23	0.55	19.90
Jhumka	10.38	6.04	4.34	8.21	7.92	0.11	0.58	1.40	7.64	0.36	41.76
Kalo Patle	24.55	17.77	6.78	21.16	20.89	0.76	0.72	0.92	20.62	1.06	27.61
Pamali	19.66	11.96	7.69	15.81	15.33	0.41	0.61	1.31	14.87	0.71	39.15
Darmali	24.05	17.95	6.10	21.00	20.78	0.75	0.75	0.85	20.56	1.07	25.36
Juhari	59.11	38.55	20.56	48.83	47.73	3.97	0.65	1.16	46.66	2.29	34.78
Kalo Jhinuwa	6.86	2.69	4.18	4.78	4.29	0.03	0.39	2.04	3.86	0.16	60.84
Pahele	57.26	25.59	31.67	41.42	38.27	2.55	0.45	1.85	35.37	1.52	55.31
Pahelo Mansar	25.36	21.31	4.06	23.33	23.25	0.94	0.84	0.54	23.16	1.27	16.00
Dalle	18.21	14.94	3.28	16.58	16.49	0.47	0.82	0.60	16.41	0.89	17.99
Grand Total	23.97	16.81	7.16	20.39	20.07	0.70	0.70	1.00	19.76	1.00	29.88

Table 5: Mean grain yield and 10 drought tolerance indices for 17 rice genotypes grown under nonstressed conditions and 60% soil FC during the reproductive stage

Genotype	Yp	Ys	TOL	MPI	GMP	STI	YSI	SSI	HM	YI	Yield loss%
Appjhutte	11.11	6.96	4.15	9.03	8.79	0.13	0.63	0.75	8.56	0.58	37.37
Kathe	34.41	14.86	19.55	24.63	22.61	0.89	0.43	1.14	20.75	1.24	56.82
Lekali Basmati	28.45	22.03	6.42	25.24	25.04	1.09	0.77	0.45	24.83	1.84	22.55
Manamure	34.55	17.20	17.36	25.87	24.38	1.03	0.50	1.01	22.96	1.43	50.23
Gaure	9.94	4.65	5.29	7.29	6.80	0.08	0.47	1.07	6.33	0.39	53.23
Kartike	21.23	11.35	9.88	16.29	15.52	0.42	0.53	0.93	14.79	0.95	46.55
Rato Anadi Lamcho	10.83	6.15	4.68	8.49	8.16	0.12	0.57	0.87	7.84	0.51	43.22
Piyale	11.50	6.69	4.82	9.10	8.77	0.13	0.58	0.84	8.46	0.56	41.88
Jhumka	10.38	4.32	6.06	7.35	6.69	0.08	0.42	1.17	6.10	0.36	58.43
Kalo Patle	24.55	15.63	8.92	20.09	19.59	0.67	0.64	0.73	19.10	1.30	36.33
Pamali	19.66	7.15	12.50	13.40	11.86	0.24	0.36	1.27	10.49	0.60	63.62
Darmali	24.05	16.01	8.05	20.03	19.62	0.67	0.67	0.67	19.22	1.33	33.46
Juhari	59.11	27.50	31.61	43.30	40.32	2.83	0.47	1.07	37.54	2.29	53.47
Kalo Jhinuwa	6.86	3.82	3.04	5.34	5.12	0.05	0.56	0.89	4.91	0.32	44.27
Pahele	57.26	18.04	39.21	37.65	32.14	1.80	0.32	1.37	27.44	1.50	68.49
Pahelo Mansar	25.36	12.14	13.22	18.75	17.55	0.54	0.48	1.04	16.42	1.01	52.13
Dalle	18.21	9.63	8.59	13.92	13.24	0.31	0.53	0.94	12.59	0.80	47.15
Grand Total	23.97	12.01	11.96	17.99	16.96	0.50	0.50	1.00	16.00	1.00	49.91

3.4 Correlations between yield and drought tolerance indices

The most effective indices for drought treatment include those that strongly correlate with yield under both stressed and nonstressed conditions (Hooshmandi, 2019). The Pearson correlation coefficients between Yp and Ys and between many drought tolerance indices were analyzed to determine the most appropriate drought tolerance criteria. Quantitative calculations were performed for drought tolerance indicators, as shown in Figures (1) and (2). In simpler terms, conducting a correlation analysis between two or more variables can serve as a valuable method for evaluating the best cultivars and indices, as highlighted by Selamawit Abebe et al. (2021). Mitra (2001) emphasized that an appropriate measure must have a significant correlation with grain yield in both scenarios.

3.4.1 Correlations between yield and drought tolerance indices under FC

The yield (Yp) under nonstressed conditions had a very strong correlation (r=0.92) with the yield (Ys) under FC. Nouraein et al. (2013) studied wheat, and Rahimi et al. (2013) studied rice; both of these authors reported similar findings on the relationship between Yp and Ys. In addition, Yp and Ys had extremely significant positive relationships with drought indicators such as TOL, MPI, GMP, STI, and HM under FC. These findings showed that the parameters shown in Figure (1) were helpful for identifying high-yielding cultivars under a variety of drought conditions. Yp displayed a perfect positive correlation (r=1) with YI. In contrast, YSI (r=-0.26) had negative and nonsignificant correlations with Yp. Similarly, Yp and SSI (r=-0.06) also had negative and nonsignificant relationships. A comparable outcome was observed in a drought investigation involving a wheat cultivar (Hooshmandi, 2019). Under FC, the YSI and SSI displayed perfect negative correlations (r=-1) with each other. Furthermore, at FC, a perfect relationship was identified between the drought indices GMP and HM and between the drought indices and the MPI, which is similar to the findings of Mau et al. (2019). Another group of drought indices, namely, the MPI, HM, STI, GMP, and YI, exhibited strong correlations with each other.

3.4.2 Correlations between yield and drought tolerance indices at 60% FC

Figure (2) shows the correlation coefficients between grain yield and drought tolerance indices under 60% FC. The correlation coefficient (r) between Yp and Ys was 0.88, indicating a positive and highly significant relationship. This finding suggested that the selection of drought-stressed plants may be influenced indirectly by their performance under nonstressed conditions. The average yield under non-stress conditions (Yp) had a positive and significant relationship with TOL (r = 0.94), MPI (r = 0.99), GMP (r = 0.97), STI (r = 0.96), HM (r = 0.95), and YI (r = 0.88). In contrast, there was a negative and nonsignificant association (r = -0.33) between YSI and Yp/Ys (Figure 3). TOL (r=0.67), MPI (r=0.94), GMP (r=0.96), STI (r=0.91), HM (r=0.98), and YI (r=1.00) all had high and positive correlations with the mean yield under stress conditions (Ys). Figure (3) shows a nonsignificant, negative correlation between the SSI (r = -0.13) and Ys. Ys and YI had a 100% positive relationship under 60% FC. The MPI, GMP, and HM all showed a perfect positive correlation, whereas the SSI and YSI had a significant negative correlation (r=1). Toorchi et al. (2012) reported positive associations between MP and Ys and between GMP and Yp. Khalili et al. (2012) also reported that GMP, MP, and STI were strongly and positively correlated with stress yield.

3.5 Genotype ranking

The use of different indicators suggested that the various genotypes were drought-tolerant. As a result, using a single criterion to identify drought-tolerant genotypes yields mixed results. To identify suitable drought-tolerant genotypes, we computed the average rank, standard deviation of ranks, and rank sum for all indices.

3.5.1 Ranking of genotypes under FC

Juhari (6.33), Kathe (7.27), Manamure (7.57), and Pahele (7.70) had the best mean rank and relatively low standard deviation of rank, making them the most drought-tolerant genotypes. Kalo Jhinuwa (19.72), Gaure (18.02), and Jhumka (17.94) were shown to be the most susceptible genotypes (Table 6). These findings align with prior studies (Farshadfar et al., 2012, b; Khalili et al., 2012), which supports the validity of this technique.

3.5.2 Ranking of genotypes under 60% field capacity

As shown in Table 7, Juhari (6.14), Manamure (6.98), Pahele (8.04), and Kathe (8.85) exhibited the most favourable average rankings and relatively consistent rankings across different criteria, establishing them as the top performers in terms of drought tolerance. On the other hand, Kalo Jhinuwa (18.72), Jhumka (17.77), and Gaure (16.93) emerged as the most susceptible genotypes, exhibiting less impressive effects. Naghavi et al. (2013) reported similar findings.

Table 6: Rank (R), Rank mean and standard deviation of ranks (SD) of drought tolerance indices at FC.

Genotype	Rank Yp	Rank Ys	Rank TOL	Rank MPI	Rank GMP	Rank STI	Rank YSI	Rank SSI	Rank HM	Rank YI	Mean±SD of Rank	∑R
Appjhutte	13	12	17	12	12	12	1	17	12	12	12±4.15	16.15
Kathe	4	5	3	4	4	4	11	7	4	4.5	5.05±2.22	7.27
Lekali Basmati	5	4	12	5	5	5	3.5	15	5	4.5	6.4±3.65	10.05
Manamure	3	2	8	3	3	3	6	12	3	2	4.5±3.07	7.57
Gaure	16	15	16	15	15	15	2	16	15	15	14±4.02	18.02
Kartike	9	10	5	9	9	9	12	6	9	10	8.8±1.89	10.69
Rato Anadi Lamcho	14	13	15	14	14	13.5	3.5	14	14	13.5	12.85±3.15	16.00
Piyale	12	14	14	13	13	13.5	8	10	13	13.5	12.4±1.84	14.24
Jhumka	15	16	9	16	16	16	15	3	16	16	13.8±4.14	17.94
Kalo Patle	7	8	6	7	7	7	10	8	7	8	7.5±1.02	8.52
Pamali	10	11	4	11	11	11	14	4	11	11	9.8±3.06	12.86
Darmali	8	7	7	8	8	8	9	9	8	7	7.9±0.7	8.60
Juhari	1	1	2	1	1	1	13	5	1	1	2.7±3.63	6.33
Kalo Jhinuwa	17	17	10	17	17	17	17	1	17	17	14.7±5.02	19.72
Pahele	2	3	1	2	2	2	16	2	2	3	3.5±4.2	7.70
Pahelo Mansar	6	6	11	6	6	6	5	13	6	6	7.1±2.51	9.61
Dalle	11	9	13	10	10	10	7	11	10	9	10±1.48	11.48

Note: SD: standard deviation of rank; ∑R: rank sum (rank mean +rank SD)

Table 7: Rank (R), Rank mean and standard deviation of ranks (SD) of drought tolerance indices at 60% FC.

Genotype	Rank Yp	Rank Ys	Rank TOL	Rank MPI	Rank GMP	Rank STI	Rank YSI	Rank SSI	Rank HM	Rank YI	Mean±SD of Rank	∑R
Appjhutte	13	12	16	13	12	12.5	4	14	12	12	12.05±2.94	14.99
Kathe	4	7	3	5	5	5	14	4	5	7	5.9±2.95	8.85
Lekali Basmati	5	2	11	4	3	3	1	17	3	2	5.1±4.76	9.86
Manamure	3	4	4	3	4	4	10	8	4	4	4.8±2.18	6.98
Gaure	16	15	13	16	15	15.5	12.5	5	15	15	13.8±3.13	16.93
Kartike	9	9	7	9	9	9	8.5	10	9	9	8.85±0.71	9.56
Rato Anadi Lamcho	14	14	15	14	14	14	6	12	14	14	13.1±2.47	15.57
Piyale	12	13	14	12	13	12.5	5	13	13	13	12.05±2.41	14.46
Jhumka	15	16	12	15	16	15.5	15	3	16	16	13.95±3.82	17.77
Kalo Patle	7	6	8	6	7	6.5	3	15	7	6	7.15±2.9	10.05
Pamali	10	11	6	11	11	11	16	2	11	11	10±3.49	13.49
Darmali	8	5	10	7	6	6.5	2	16	6	5	7.15±3.55	10.70
Juhari	1	1	2	1	1	1	12.5	5	1	1	2.65±3.49	6.14
Kalo Jhinuwa	17	17	17	17	17	17	7	11	17	17	15.4±3.32	18.72
Pahele	2	3	1	2	2	2	17	1	2	3	3.5±4.54	8.04
Pahelo Mansar	6	8	5	8	8	8	11	7	8	8	7.7±1.49	9.19
Dalle	11	10	9	10	10	10	8.5	9	10	10	9.75±0.68	10.43

Note: SD: standard deviation of rank; ∑R: rank sum (rank mean +rank SD)

3.6 Principal component analysis (PCA)

Two distinct PCAs were performed: one to measure the effect of drought stress on genotypes and another to determine how genotype performance corresponds to specific drought indices. In the PCA biplot (Figures 3 and 4), a cluster was formed comprising yields and drought tolerance indices, viz. GMP, MPI, and STI reinforce the connection between them.

3.6.1 PCA of drought tolerance indices under FC

The results of PCA of the indices revealed that two main components (PC1 and PC2) accounted for 97.8% of the fluctuations, with an eigenvalue larger than 1 seen in the drought tolerance indices. PC1 accounted for approximately 74.78% of these variations, while PC2 explained approximately 23.02% (Table 8). Furthermore, the findings revealed that PC1 exhibited positive correlations with nearly all drought indices, except for the YSI, which showed a negative correlation. Hooshmandi (2019) and Kandel et al. (2022) documented analogous findings, with strong and positive correlations observed with Yp, Ys, STI, and MP. On the other hand, PC2 displayed positive associations with the drought indices TOL, STI, and SSI but exhibited negative relationships with MPI, GMP, YSI, HM, and YI. The biplot in Figure 3 shows that the drought indices TOL, STI, and SSI were the best indicators for selecting drought-resistant genotypes with high and steady grain yields, such as the Pahele genotype.

3.6.2 PCA among drought tolerance indices under 60% FC

To enhance the understanding of the relationships, commonalities, and distinctions among the drought tolerance indicators, biplots derived from principal component analysis (PCA) were generated. Researchers have widely employed biplot analysis for evaluating diverse criteria across various plant species. In Table 9, two principal components with eigenvalues exceeding one are depicted, revealing more than 98% of the variation. This study showed that the first two PCAs included 98.8% of the overall variance, with PC1 contributing 75.1% and having a positive relationship with Yp, Ys, TOL, MPI, GMP, SSI, STI, HM, and YI. Mau et al. (2019) reported that the STI, GMP, GM, HARM, ATI, MRP, REI, YI, and SNPI were the most appropriate drought indices for selecting drought-tolerant, high-yielding upland rice genotypes. PC2 contributed 23.7% of the variance and was positively correlated with Yp, TOL, and SSI. As a result, the first component may be understood to express yield potential and drought tolerance, whereas the second component falls within the stress susceptibility group. Hence, genotypes such as Pahele, Kathe, Pahelo Mansar, and Manamure, which had greater PC1 and PC2 values (Figure 4), represent superior choices for both stressed and nonstressed environments. These varieties also had the greatest levels of Yp, Ys, GMP, STI, and MP, which made them suitable for both environments. However, most of the genotypes that had small PC2 and PC1 scores were susceptible. Research by Miah et al. (2013) supports the use of PCA for identifying the most significant characteristics leading to drought tolerance in rice genotypes (Mau et al., 2019). The study revealed that the first two PCAs comprised 72.2% of the total variance, with PC1 contributing 47.2% and showing a positive correlation with grain yield in both drought and nondrought settings. PC2 contributed 25.0% of the total yield and had a positive relationship with grain yield under nondrought conditions.

Table 8: PCA for drought tolerance indices of rice plants grown under FC

Indices	PC1	PC2
Yp	0.363	0.043
Ys	0.348	-0.184
TOL	0.302	0.306
MPI	0.365	-0.044
GMP	0.363	-0.071
STI	0.358	0.004
YSI	-0.075	-0.640
SSI	0.075	0.640
HM	0.361	-0.097
YI	0.348	-0.184
Eigenvalue	7.478	2.302
Percent of variance (%)	74.78	23.02
Cumulative percentage (%)	74.78	97.8

Table 9: PCA for drought tolerance indices of rice plants grown under 60% FC

Indices	PC1	PC2
Yp	0.361	0.086
Ys	0.342	-0.220
TOL	0.323	0.280
MPI	0.365	-0.008
GMP	0.363	-0.059
STI	0.356	-0.012
YSI	-0.077	-0.633
SSI	0.077	0.633
HM	0.359	-0.111
YI	0.342	-0.220
Eigenvalue	7.51	2.37
Percent of variance	75.11	23.7
Cumulative percentage	75.1	98.8

3.7 Cluster analysis

3.7.1 Cluster analysis of genotypes under FC

Cluster analysis was performed on a collection of 17 rice genotypes, which were divided into two separate groups based on their drought tolerance indices. These genotype groupings aligned with the outcomes obtained from the PCA. Cluster 2 included two genotypes with high yield potential and related components, while Cluster 1 included the remaining 15 genotypes with poor yields under both environmental circumstances (Table 11). Cluster 2 had the largest degree of similarity, accounting for 64.42%, followed by Cluster 1 at 28.59%, with an overall similarity of 0% (Figure 5). The genotypes in Cluster 1 had lower values for TOL, MPI, GMP, STI, SSI, HM, and YI. As a result, the genotypes in Cluster 1 were classified as drought-prone, whereas those in Cluster 2 were recognized as drought-tolerant. Similar findings have been found in research involving numerous crops, including maize (Khodarahmpour et al., 2011), durum wheat (Golabadi et al., 2006), wheat (Mohammadi et al., 2011), and chilli pepper (Saendamuen et al., 2022). Genetic variability among genotypes has resulted in their clear differentiation into two distinct clusters, where genotypes within the same cluster exhibit a greater degree of genetic similarity. This analytical approach was employed to prevent the selection of parents from genetically uniform clusters and to ensure the preservation of a relatively diverse genetic pool. The largest cluster, Cluster 1, consisted of fifteen genotypes, suggesting that these genotypes share certain characteristics or common ancestry, which is why they were grouped. In contrast, Cluster 2 included only two genotypes.

3.7.2 Cluster analysis of genotypes under 60% FC

Cluster analysis revealed that the rice landraces were divided into two groups based on the indices: Cluster 1 and Cluster 2 (Figure 6). In this study, the second group had the greatest TOL, MPI, GMP, STI, SSI, HM, and YI, making it the most desirable cluster for both growth conditions (tolerant group). The first group had mean indicator values (the sensitive group).

Table 10: Distances between cluster centroids

	Cluster1	Cluster2
Cluster1	0.00000	6.75522
Cluster2	6.75522	0.00000

Table 11: Grouping of 17 rice genotypes under two clusters

Cluster	No. of genotypes	Genotypes
Cluster 1	15	Appjhutte, Kathe, Lekali Basmati, Manamure, Gaure, Kartike, Rato Anadi Lamcho, Piyale, Jhumka, Kalo Patle, Darmali, Kalo Jhinuwa, Pamali, Pahelo Mansar, Dalle
Cluster 2	2	Pahele, Juhari

Table 12: Cluster centroids

Variable	Cluster1	Cluster2
Yp	-0.295510	2.21633
Ys	-0.217328	1.62996
MPI	-0.271280	2.03460
GMP	-0.262149	1.96612
STI	-0.300098	2.25073

Table 13: Distances between cluster centroids

	Cluster1	Cluster2
Cluster1	0.00000	6.69142
Cluster2	6.69142	0.00000

Table 14: Grouping of 17 rice genotypes under two clusters

Cluster	No. of genotypes	Genotypes
Cluster 1	15	Appjhutte, Kathe, Lekali Basmati, Manamure, Gaure, Kartike, Rato Anadi Lamcho, Piyale, Jhumka, Kalo Patle, Darmali, Kalo Jhinuwa, Pamali, Pahelo Mansar, Dalle
Cluster 2	2	Pahele, Juhari

Table 15: Cluster centroids

Variable	Cluster1	Cluster2
Yp	-0.295510	2.21633
Ys	-0.211161	1.58370
MPI	-0.277142	2.07856
GMP	-0.263220	1.97415
STI	-0.301417	2.26063

Research has shown that screening based on drought indices results in the discovery of genotypes with considerably greater drought performance at the expense of somewhat reduced yields under normal irrigated circumstances. Among the 17 genotypes, a combination of five drought indices and deviations determined Pahele, Juhari, Lekali Basmati, and Manamure as four positive high-yielding drought-tolerant genotypes, as indicated by the excellent MPI, GMP, YSI, STI, and YI values, suggesting that these combinations could be used as alternatives to choose drought-tolerant genotypes with high yield performance under both stressed and unstressed conditions. Among the numerous tolerance indices studied, all indices except YSI and SSI exhibited good associations with grain yield under stress and nonstress conditions, showing that these indices are better suited for selecting resistant genotypes. The ranking approach and cluster analysis were used to screen drought-tolerant genotypes, and the landraces Juhari, Kathe, Manamure, and Pahele were found to be the most drought tolerant, while Kalo Jhinuwa, Gaure, and Jhumka were the most sensitive genotypes. As a result, these materials are suggested for use as parents for improving drought tolerance in other cultivars.

Considering the findings of both PCA and correlation analysis, we recommend that TOL, MPI, GMP, the STI, the SSI, HM, and YI be employed as key indicators for drought screening in rice. Despite the scarcity of studies on drought assessment in rice, this study could lead to the use of a useful resource for upcoming drought screening efforts in rice farming. Similarly, according to the results of PCA and drought indices analysis, the Juhari, Pahele, and Kathe genotypes have the potential for cultivation in drought-prone parts of Nepal's mid-hill region. As a result, farmer field experiments should be conducted on these genotypes before they are suggested for cultivating rice in Nepal's drought-prone mid-hill area.

TOL_Tolerance Index; MPI_Mean Productivity Index; GM_Geometric Mean Productivity; STI_Stress Tolerance Index; SSI_Stress Susceptibility Index HARM_Harmonic Mean Productivity, YSI_Yield Stability Index; YI_Yield Index; ATI_Abiotic Tolerance Index; DI_Drought Response Index; MRP_Mean Relative Performance; REI_Relative Efficiency Index; YI_Yield Index; SNPI_Stress/Non-Stress Production Index; DTE_Drought Tolerance Efficiency; PCA_Principal Component Analysis; Ys_Yield under stress; Yp_Yield under non stress; Ῡs_Yield mean under stress; Ῡp_Yield mean under non stress; DI_Drought Index; CSI_Composite Selection Index

Author contributions

The study design was finalized by all the authors. Material preparation and data collection were performed by Prabin Shrestha, Merina Dhakal, Ritika Karanjit, and Prajwal Shrestha.

The data were analyzed by Prabin Shrestha, Bishnu Prasad Kandel and Merina Dhakal, while the main manuscript was written by Prabin Shrestha. Proofing and suggestions to address the reviewer’s comments were provided by Ankur Poudel and Bishnu Prasad Kandel.

Acknowledgment

The completion of this research could not have been possible without the Ghanpokhara Community Seed Bank, Lamjung and Purkot Community Seed Bank, Tanahun. We thank for providing invaluable seeds of rice landraces.

Funding

No funding, grants or financial aid was received during the course of the entire research period or manuscript preparation.

Data availability

The datasets generated and analyzed in the current study are not publicly available for privacy reasons but can be made available by the corresponding author upon reasonable request.

Statements and declarations

Competing interests: The authors have no financial or nonfinancial interests related to the material submitted for publication, whether directly or indirectly linked.

Conflict of interest: The authors state that there are no conflicts of interest.

Ethical approval: This research contains no investigations on human or animal subjects conducted by any of the authors.

Plant Ethics: The field experiments on rice, including the collection of plant material, were conducted following all relevant institutional, national, and international guidelines and legislation. The methods, study, and protocols were carried out in accordance with the guidelines and regulations of the Nepal Agricultural Research Council (NARC) and the Ministry of Agriculture and Livestock Development (MoALD).

Adhikari, M., Adhikari, N. R., Sharma, S., Gairhe, J., Bhandari, R. R., & Paudel, S. (2019). Evaluation of Drought Tolerant Rice Cultivars Using Drought Tolerant Indices under Water Stress and Irrigated Condition. American Journal of Climate Change, 08(02), 228–236. https://doi.org/10.4236/ajcc.2019.82013
Atkinson NJ, Urwin PE (2012). The interaction of plant biotic and abiotic stresses: from genes to the field. J. Exp. Bot. 63:3523-3543
Bargaz A, Zaman-Allah M, Farissi M, Lazali M, Drevon J-J, Maougal RT, Georg C (2015). Physiological and molecular aspects of tolerance to environmental constraints in grain and forage legumes. Int. J. Mol. Sci. 16:18976-19008.
Bargaz A, Zaman-Allah M, Farissi M, Lazali M, Drevon J-J, Maougal RT, Georg C (2015). Physiological and molecular aspects of tolerance to environmental constraints in grain and forage legumes. Int. J. Mol. Sci. 16:18976-19008.
Bimpong IK, Serraj R, Chin JH, Ramos J, Mendoza E, Hernandez J, Mendioro MS, Brar DS (2011). Identification of QTLs for drought-related traits in alien introgression lines derived from crosses of rice (Oryza sativa cv. IR64) × O. glaberrima under lowland moisture stress. J. Plant Biol. 54:237-250.
Bouslama, M. and Schapaugh, W.T. (1984) Stress tolerance in soybean, Part I: evaluation of three screening techniques for heat and drought tolerance. Crop Sci. J 24:933-937
Doorenbos, J., & Kassam, A. H. (1979). Yield response to water. Irrigation and Drainage Paper, 33, 257.
Farooq M, Kobayashi N, Wahid A, Ito O, Basra SMA (2009). Strategies for producing more rice with less water. Adv. Agron. 101:351-388. Fischer RA, Maurer R (1978). Drought resistance in spring wheat cultivar: I- Grain yield response. Aust. J. Agric. Res. 29:897-912.
Farshadfar E, Farshadfar M, Dabiri S (2012). Comparison between effective selection criteria of drought tolerance in bread wheat landraces of Iran. Ann Biol Res 3(7):3381- 3389.
Farshadfar E, Mahmodi N, Yaghotipoor A (2011) AMMI stability value and simultaneous estimation of yield and yield stability in bread wheat (Triticum aestivum L.) Australian Journal of Crop Science 5: 1837–1844. Link: https://bit.ly/3mudxSm
Fernandes, G.C.J. (1992) Effective selection criteria for assessing plant stress tolerance. In: Kus EG (ed) Adaptation of Food Crop Temperature and Water Stress, Proceeding of 4th International Symposium, Asian Vegetable and Research and Development Center, Shantana, Taiwan. pp 257-270
Fischedr RA, Maurer R (1978). Drought resistance in spring wheat cultivars. I: Grain yield responses. Australian J Agric Res 29:879-912.
Garrity, D.P. and O’Toole, J.C. (1995) Selection for reproductive stage avoidance in rice, using infrared thermometry. Agron J 87:773-779.
Gavuzzi, P., Rizza, F., Palumbo, M., Campaline, R.G., Ricciardi, G.L. and Borghi, B. (1997) Evaluation of field and laboratory predictors of drought and heat tolerance in winter cereals. Can J Plant Sci 77: 523-531
Golabadi, M., Arzani, A., & Maibody, S. A. M. M. (2006). Assessment of drought tolerance in segregating populations in durum wheat. African Journal of Agricultural Research, 1(5), 162–171.
Hassan, M. A., Dahu, N., Hongning, T., Qian, Z., Yueming, Y., Yiru, L., & Shimei, W. (2023). Drought stress in rice: morpho-physiological and molecular responses and marker-assisted breeding. Frontiers in Plant Science, 14(July), 1–16. https://doi.org/10.3389/fpls.2023.1215371
Hooshmandi B. 2019. Evaluation of tolerance to drought stress in wheat genotypes. IDESIA (Chile). 37(2):37–43
Jafari A, Paknejad F, Jami M and Ahmadi AL (2009). Evaluation of selection indices for drought tolerance of corn (Zea mays L.) hybrids. Int. J. Plant Prod. 3: 33-38.
Kamoshita A, Babu RC, Boopathi NM, Fukai S (2008). Phenotypic and genotypic analysis of drought-resistance traits for development of rice cultivars adapted to rainfed environments. Field Crops Res. 109:1-23.
Kandel, B. P., Joshi, L. P., Sharma, S., Adhikari, P., Koirala, B., & Shrestha, K. (2022). Drought tolerance screening of rice genotypes in mid-hills of Nepal using various drought indices. Acta Agriculturae Scandinavica Section B: Soil and Plant Science, 72(1), 744–750. https://doi.org/10.1080/09064710.2022.2072382
Khalili M, Naghavi MR, Pour Aboughadareh AR, Talebzadeh J (2012). Evaluating drought stress tolerance based on selection indices in spring canola cultivars (Brassica napus L.). JAS 4(11):78-85.
Khodarahmpour, Z., Choukan, R., Bihamta, M. R., & Majidi, H. E. (2011). Determination of the best heat stress tolerance indices in maize (Zea mays L.) inbred lines and hybrids under Khuzestan province conditions.
Kumar, A., Bernier, J., Verulkar, S., Lafitte, H.R. and Atlin, G.N. (2008) Breeding for drought tolerance: direct selection for yield, response to selection and use of drought-tolerant donors in upland and lowland-adapted populations. Field Crops Res 107:221-231
Kumar, S., Dwivedi, S. K., Singh, S. S., Jha, S. K., S, L., Elanchezhian, R., Singh, O. N., & Bhatt, B. P. (2014). Identification of drought-tolerant rice genotypes by analyzing drought tolerance indices and morpho-physiological traits. Sabrao Journal of Breeding and Genetics, 46(2), 217–230.
Mau, Y. S., Ndiwa, A. S. S., Oematan, S. S., & Markus, J. E. R. (2019). Drought tolerance indices for selection of drought tolerant, high yielding upland rice genotypes. Australian Journal of Crop Science, 13(1), 170–178. https://doi.org/10.21475/ajcs.19.13.01.p1778
Mitra J (2001). Genetics and genetic improvement of drought resistance in crop plants. Curr Sci India 80:758-762.
MOAD (2016) Statistical Information on Nepalese Agriculture (2015/13). Government of Nepal, Ministry of Agricultural Development, Agribusiness Promotion and Statistics Division, Singha Durbar, Kathmandu
Mohammadi, M., Karimizadeh, R., & Abdipour, M. (2011). Evaluation of drought tolerance in bread wheat genotypes under dryland and supplemental irrigation conditions. Australian Journal of Crop Science, 5(4), 487–493.
Mostajeran, A., & Rahimi-Eichi, V. (2009). Effects of Drought Stress on Growth and Yield of Rice (Oryza sativa L.) Cultivars and Accumulation of Proline and Soluble Sugars in Sheath and Blades of Their Different Ages Leaves. & Environ. Sci, 5(2), 264–272.
Naghavi, M. R., Pour Aboughadareh, A., & Khalili, M. (2013). Evaluation of Drought Tolerance Indices for Screening Some of Corn (Zea mays L.) Cultivars under Environmental Conditions. Notulae Scientia Biologicae, 5(3), 388–393. https://doi.org/10.15835/nsb539049
Nguyen HT, Babu RC, Blum A (1997). Breeding for drought tolerance in rice: physiology and molecular genetics considerations. Crop Sci. 37:1426-1434.
Nouraein M, Mohammadi SA, Aharizad S, Moghaddam M, Sadeghzadeh B. 2013. Evaluation of drought tolerance indices in wheat recombinant inbred line population. Ann Biol Res. 4(3):113–122
Ouk, M., Basnayake, J., Tsubo, M., Fukai, S., Fischer, K., Cooper, M. and Nesbitt, H. (2006) Use of drought response index for identification of drought tolerant genotypes in rainfed lowland rice. Field Crops Res 99:48-58.
Pantuwan G, Fukai S, Cooper M, Rajatasereekul S and O'Toole JC (2002) Yield response of rice (Oryza sativa L.) genotypes to different types of drought under rainfed lowlands - Part 3. Plant factors contributing to drought resistance. Field Crops Research 73, 181-200.
Price AH, Steele KA, Moore BJ, Barraclough PB, Clark LJ (2000). A combined RFLP and AFLP map of upland rice (Oryza sativa L.) was used to identify QTL for root penetration ability. Theor. Appl. Genet. 100:49-56.
Raman, A., Verulkar, S., Mandal, N., Variar, V., Shukla, V., Dwivedi, J., Singh, B., Singh, O., Swain, P., Mall, A., Robin, S., Chandrababu, R., Jain, A., Ram, T., Hittalmani, S., Haefele, S., Hans-Peter, P, and Kumar, A. (2012) Drought yield index to select high yielding rice lines under different drought stress severities. Rice, 5:31-43
Rosielle, A.A. and Hamblin, J. (1981) Theoretical aspects of selection for yield in stress and nonstress environment. Crop Sci. 21:943-946
Sabouri, A., Dadras, A. R., Azari, M., Saberi Kouchesfahani, A., Taslimi, M., & Jalalifar, R. (2022). Screening of rice drought-tolerant lines by introducing a new composite selection index and competitive with multivariate methods. Scientific Reports, 12(1), 1–14. https://doi.org/10.1038/s41598-022-06123-9
Saendamuen, M., Chutteang, C., Chanprame, S., & Kongsamai, B. (2022). Drought tolerance indices for yield improvement in local varieties of chilli pepper. International Journal of Agricultural Technology, 18(3), 1233–1246.
Selamawit Abebe, G., Benjamin, D., Sylvia, H., & Fekadu, G. (2021). Evaluating Drought tolerance indices for selection of drought tolerant Orange Fleshed Sweet Potato (OFSP) genotypes in Ethiopia. International Journal of Agricultural Science and Food Technology, 7, 249–254. https://doi.org/10.17352/2455-815x.000115
Singh A, Kumar S, Kumar M, Nahakpam S, Sinha S, Smrity S, Sundaram P, Kumar S, Singh PK. 2018. Identification of drought-tolerant rice (Oryza sativa L) genotypes using drought tolerance indices under normal and water stress conditions. Int J Current Microbiol Appl Sci. 1(7): 4757–4766.
Sio-Se Mardeh, A., Ahmadi, A., Poustini, K. In addition, Mohammadi, V. (2006) Evaluation of drought resistance indices under various environmental conditions. Field Crops Res 98:222-229.
Toorchi M, Naderi R, Kanbar A, Shakiba MR (2012). Response of spring canola cultivars to sodium chloride stress. Ann Biol Res 2(5):312-322.
Yuan, S., Stuart, A. M., Laborte, A. G., Rattalino Edreira, J. I., Dobermann, A., Kien, L. V. N., Thúy, L. T., Paothong, K., Traesang, P., Tint, K. M., San, S. S., Villafuerte, M. Q., Quicho, E. D., Pame, A. R. P., Then, R., Flor, R. J., Thon, N., Agus, F., Agustiani, N., … Grassini, P. (2022). Southeast Asia must narrow down the yield gap to continue to be a major rice bowl. Nature Food 2022 3:3, 3(3), 217–226. https://doi.org/10.1038/s43016-022-00477-z.

No competing interests reported.

Download PDF

Editorial decision: Revision requested
05 Sep, 2024
Reviews received at journal
02 Sep, 2024
Reviews received at journal
30 Aug, 2024
Reviewers agreed at journal
26 Aug, 2024
Reviewers agreed at journal
26 Aug, 2024
Reviewers agreed at journal
23 Aug, 2024
Reviewers invited by journal
20 Aug, 2024
Editor assigned by journal
16 Jul, 2024
Submission checks completed at journal
15 Jul, 2024
First submitted to journal
08 Jul, 2024

You are reading this latest preprint version

Evaluation of Drought Tolerance of Rice Landraces in Mid-Hills of Nepal Using Drought Tolerant Indices

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials and methods

2.1 Research location

2.2 Experimental design

2.3 Treatment details

2.4 Cultivation practices

2.5 Observation and data collection

2.6 Genotype ranking

2.7 Statistical analysis

3. Results and discussion

3.1 Mean separation of yield and yield-attributing characteristics

3.2 Comparing rice landraces based on drought tolerance indices

3.2.1 Tolerance index (TOL)

4. Conclusion

Abbreviations

Declarations

References

Additional Declarations

Status:

Version 1