AMMI Stability Analysis for Yield of Black Cumin (Nigella sativa L.) Genotypes

Reader Impact Factor Score
[Total: 3 Average: 4.7]

Published on International Journal of Agriculture & Agribusiness
Publication Date: January, 2020

Beshir Hamido, Habtamu Zeleke & Tesfaye Letta
Adami Tullu Agricultural Research Center, P.O. Box 35, Ziway
Haramaya University College of Agriculture and Environmental Science, P.O. Box 138, Dire Dawa
Oromia Agricultural Research Institute, P.O. Box 81262, Addis Ababa
Ethiopia

Journal Full Text PDF: AMMI Stability Analysis for Yield of Black Cumin (Nigella sativa L.) Genotypes (Studied at East Shoa and West Arsi Zones, Oromia, Ethiopia).

Abstract
Black cumin is an erect annual herb cultivated for its seed, growing on all kinds of soils. In Ethiopia black cumin is cultivated as rain fed crop in the highlands from 1500 to 2500 meter s above sea level and the weather makes a suitable environment for the growth of black cumin seed. Genotype × environment interaction and yield stability analysis are important in measuring the genotypic yield stability and suitability for cultivation across seasons and ecological conditions. The objective of this study was to assess the stability of black cumin genotypes under the agro-ecological conditions of East Shoa and West Arsi zones. Fifteen black cumin genotypes were evaluated at six locations in randomized complete block design with three replications during 2018/19 cropping season. Analysis of variance for each location revealed the presence of significant differences among genotypes for seed yield. The combined analysis of variance over locations showed significant differences amongst genotypes. Gammachis variety was recorded with the highest mean yield (1.35 ton ha-1) followed by Dirshaye (1.26 ton ha-1) while genotype 90575-2 was recorded with the lowest (0.78 ton ha-1) mean seed yield. Additive Main Effects and Multiplicative Interaction analysis of variance had showed that most of the total variations (41.99%) was attributed to the environmental effects and the rest were attributed to the genotypic effects (31.96%) and the GEI (10.96%). AMMI stability analysis identified variety Dirshaye as the most stable genotype whereas genotype MAB-057 was the most unstable genotype. In general, genotypes (such as Dirshaye, Gammachis, Soressa and Derbera) should be used for cultivation at all the test locations since they perfomed well as compared to the other tested black cumin genotypes.

Keywords: Black cumin seed, Nigella sativa, Adaptation trials, Seed yield performance.

1. Introduction
Black cumin (Nigella sativa L.) is a diploid and an erect annual herb cultivated for its seed, growing on all kinds of soils (Jansen, 1981). It is a medicinal plant belonging to the family Ranunculaceae grown naturally in Southwest Asia and the Mediterranean region (Toncer and Kizil, 2004). It originated in Egypt and East Mediterranean countries, but is widely cultivated in Iran, Japan, China and Turkey (Shewaye, 2011). Nigella sativa is probably indigenous to the Mediterranean region and the Middle East up to India.
Black cumin is cultivated in the subtropical belt extending from Morocco to Northern India and Bangladesh, East Africa and in the former Soviet Union. In Europe, North America and South-east Asia, it is cultivated on a minor scale, mainly for medicinal use (Akhtar and Saha, 1993). It is also cultivated in sub-Saharan Africa particularly in Niger, and Eastern Africa especially Ethiopia (Iqbal et al., 2010).
In Ethiopia, black cumin is cultivated as rain fed crop in the highlands from 1500 to 2500 meters above sea level. In Ethiopia, Nigella sativa can be intercropped with barley and wheat (Ahmed and Haque, 1986). However improved production technology must be available (Ministry of Agriculture and Rural Development, 2003).
Black Cumin has a long history of uses for food flavors, perfumes and medicinal values. It is used as an essential ingredient in preparing soup, sausages, cheese, cakes and candies. Studies have shown that Nigella sativa seeds have high nutritional values: proteins content ranging from 20 to 27%, carbohydrates ranging from 23.5 to 33.2%, moisture content ranging from 5.52 to 7.43% and ash content ranging from 3.77 to 4.92% (Al-Jassir, 1992; Al-Ghamdi, 2001 and Nergiz and Otles, 1993).
Black cumin grows on a wide range of soils. Sandy loam soil rich in microbial activity is the most suitable for its cultivation. Areas with moderate rainfall and well drained soils with pH of 7-75 are quite suitable for black cumin production (Orgut, 2007). Some studies shown that black cumin is able to tolerate moderate water stress and responds well to soil fertilization (Mozzafari et al., 2000; Bannayan et al., 2008).
In Ethiopia, the weather makes a suitable environment for the growth of black cumin seed. Amara, Oromia, South Nations, Nationalities and Peoples, (SNNP) and Gambella regions are known in producing of the black cumin seed in Ethiopia (Atta, 2003 and Takrun et al., 2008).
The productivity of black cumin depends on the genetic potential of varieties and the suitability of environmental factors across production areas. Identification of high yielding and well adapted genotypes is achieved through analyzing the effect of genotype × environment interaction (GEI) and yield stability.
Genotype × environment interaction and yield stability analysis are important in measuring varietal stability and suitability for cultivation across seasons and ecological conditions (Romagosa and Fox, 1993). Globally, several stability analysis methods are available, and they have been used in different research efforts to determine the magnitude of GEI effects for many different crops by different researchers (Akcura et al., 2005).
Additive Main Effects and Multiplicative Interaction (AMMI) model has found to be more effective in selection of stable genotypes (Crossa et al., 1991; Haji and Hunt, 1999; Ariyo and Ayo-Vaughan, 2000; Taye et al., 2000). It is used to analyze multi-location trials (Gauch and Zobel, 1988; Zobel et al., 1988; Crossa et al., 1990). AMMI integrates the analysis of variance and principal component analysis into a unified approach (Bradu and Gabriel, 1978).
The genetic potential contained within the crop, the environmental effects and their interaction plays a great role in determining the performance and stability of the crop to a given environment. Therefore, the current study was conducted with the objective of evaluating and identifying the high yielding, stable, and adaptable black cumin genotypes at Eat Shoa and West Arsi zones.

2. Materials and Methods
2.1. Study Area
The experiment was conducted in East Shoa and West Arsi zones at six locations viz; two sites in East Shoa zone [Adami Tullu Agricultural Research Center (ATARC) on research station at Adami Tullu Jiddo Kombolcha district and Bekele Girrisa at Dugda district] and four sites in West Arsi zone (Ali Woyyo, Makko Oda, Bute and Umbure) during the 2018/19 main cropping season under rain fed condition. The locations are the representative for the diverse agro-ecologies of spice crops growing environments in East Shoa and West Arsi zones.

Table 1. Description of the test locations used in the study

Key: ‘NA’ stands for not available and ‘ATARC’ for Adami Tullu Agricultural Research Center

2.2. Breeding Materials, Experimental Materials and Managements
A total of fifteen black cumin genotypes [i.e., ten accessions (viz; AC-BC-4, AC-BC-9, AC-BC-10, AC-BC-19, MAB-042, MAB-057, 90575-2, 20750-1, 242834-1and 244654-1) along with five standard checks (viz; Derbera, Dirshaye, Eden, Gammachis and Soressa] that were obtained from Sinana Agricultural Research Center were used in this study. The materials were evaluated using Randomized Complete Block Design (RCBD) with three replications at six locations in the main cropping season of the year 2018/19.
The plot size for each experimental unit was 1.2m × 2m (4 rows, each 2m long). The total area of a plot was 2.4m2. The spacing between rows, plots and blocks were 0.35m, 0.5m and 1m, respectively. Sowing was done by hand drilling and covered slightly with the soil. Fertilizer rate of 46kg Di-Phosphorus pent-oxides (P205) ha-1 and 60kg Nitrogen (N) ha-1 was used to facilitate and increase root development and increases yield in black cumin (Champawat and Pathak, 1982).

2.3. Data Collection and Analysis
The following data were recorded: days to emergence, days to 50% flowering, days to maturity, plant height (cm), number of primary branches per plant, number of capsules per plant, seed yield per hectare (ton ha-1) and thousand seed weight (g). Data was collected from the middle two harvestable rows for traits estimated from a plot. Data were collected from ten randomly selected plants found on the central two rows and the average values calculated.
All the recorded data were subjected to analysis of variance following the standard procedure for each location and combined analysis of variance over locations were computed using the Gen-Stat 18th Edition Statistical Computer Software Programs. Bartlett’s chi-square test was used to determine the validity of the combined analysis of variance and homogeneity of error variances among environments. Then combined analysis of variance was carried out to estimate the additive effects of environment (E), genotype (G) and their interactions (GEIs).

2.4. Stability Analysis
The additive main effects and multiplicative interaction (AMMI) stability analysis was used to integrate the analysis of variance and principal component analysis into a unified approach. An initial analysis of variance was performed for each environment to verify the existence of differences among the genotypes. Thereafter, the homogeneity between residual variances was determined, and a joint analysis of variance was used to test the genotype and environment effects and the magnitude of the GEIs.
Different researchers have been working with AMMI stability analysis as stability measuring parameter for studying the stability of seed yield and quality of different crop genotypes, particularly wheat across various environments (Desalegn et al., 2004; Ferney et al., 2006; Mohammadi and Amri, 2008; Mohammed, 2009; Mut et al., 2010).
The AMMI model was given as:
Where;
Yij is the yield of the ith genotype in the jth environment,
μ is the grand mean,
Gi and Ej are the genotype and environment deviations from the grand mean respectively.
λk is the eigen value of the interaction principal component axis K;
αik and Yjk are genotype and environment principal component scores for axis K
eij is the error term.
AMMI Stability Value (ASV), IPCA1 and IPCA2 were computed to identify the stable genotype with consistence yielding performance across the testing environments. The degrees of freedom for the IPCA axes were also calculated based on the following method (Zobel et al., 1988).

Where;
df is degree of freedom
G is number of genotypes,
E is the number of environments and
n is the nth axis of IPCA
The AMMI stability value (ASV), was also calculated for each genotype and each environment as follows (Purchase et al., 2000):

Where,
ASV — is AMMI stability value
SS — is sum of squares and;
IPCA1 and IPCA2 — are the first and second interaction principal component axes, respectively
Accordingly, genotypes with the least AMMI stability value (ASV) were considered as the most stable genotypes, where as those which have the highest ASV were considered as unstable (Purchase, 1997).

3. Results and Discussions
The analysis of variance of an individual environment revealed that seed yield showed a highly significant difference (P ≤ 0.01) at all test environments (Table 2). This indicated that, genotypes might not express the same seed yield performance at a specified test location’s environmental conditions; or different genotypes may respond differently to a specified environment. Accordingly, at location Bute, the variety Soressa ranked 1st in its seed yield performance of 1.20 ton ha-1, while the same variety ranked 5th in its seed yield performance of 1.54 ton ha-1 at location Makko Oda.

Table 2. Analysis of variance for seed yield of fifteen black cumin

Key: ** = highly significant ( P ≤ 0.01 ) at 1% level of significance, df = degree of freedom
The highest and the lowest mean seed yield performance of the tested genotypes across the testing environments were 1.35 ton ha-1 and 0.78 ton ha-1, which were obtained from genotypes Gammachis and 90575-2 respectively (Table 3).

Table 3. Mean seed yield of fifteen black cumin genotypes tested at six locations

Key: ATARC = Adami Tullu Agricultural Research Center, GM = Genotypic means, EM = Environmental means, MSE = Mean square of error, SE (d) = Standard error of difference, LSD = Least Significant Difference. Values with the same letters in a column are not statistically significantly different.

On the other hand, the combined analysis of variance for seed yield revealed the presence of highly significant difference among genotypes, environments and GEI. This result is in agreement with the finding of Fufa (2018) with the result of combined ANOVA showing highly significant variation (P ≤ 0.01) among different genotypes evaluated across location for seed yield.
The mean seed yield values of genotypes averaged across the environments showed that genotype Gammachis had the highest mean yield (1.35 ton ha-1) followed by genotype Dirshaye (1.26 ton ha-1) while genotype 90575-2 had the lowest (0.78 ton ha-1) mean seed yield.

Table 4. Combined analysis of variance for mean seed yield of fifteen black cumin genotypes across locations

Key: * and ** stand for significant differences at (P ≤ 0.05) and (P ≤ 0.01), respectively; ns for non-significant difference, df = degree of freedom and G×L = Genotype by location interaction
This indicates that the test environments were highly variable and showed high contribution in varying the yield performance of black cumin genotypes. The presence of blocking and/or replicating within the testing environments could not influence the yield performance of the tested genotypes.
In addition to this, the combined analysis of variance across the locations for seed yield revealed that genotypes, environments, GEI, error variance and block within environments contributed 35.89%, 47.16%, 12.31%, 4.50% and 0.14%, respectively (Table 5).

Table 5. Percent contribution of genotypes, environments, GEI and error sum squares over locations

Key: Values with the same letters have no significant difference and the numbers in the brackets stand for the degree of freedom
D50%F = Days to 50% flowering, DM = days to maturity, PH = Plant height, NPB = Number of primary branches and SYPH = Seed yield per hectare

AMMI analysis of variance for seed yield of fifteen black cumin genotypes evaluated at six locations indicated that most of the total sum square of the model (41.99%) was attributed to the environmental effects and the rest were attributed to the genotypic effects (31.96%) and the GEI (10.96%) (Table 6).

Table 6. AMMI analysis of variance for seed yield of fifteen black cumin genotypes across locations

Key: *, ** represent significant at P ≤ 0.05 and P ≤ 0.01, respectively, ns for non-significance, df = degree of freedom, MS = Mean Square and TSS = Total Sum Square

The observed large sum of square and highly significant mean of square of location showed that the locations were highly diverse, with large differences among the location means causing most of the variation in seed yield.
AMMI stability value (ASV) was calculated for each of the fifteen black cumin genotypes. Accordingly, the variety Dirshaye was the most stable with an ASV value of (0.093) followed by genotypes 242834-1 and Soressa with their ASV value of (0.095) and (0.109), respectively. Genotype MAB-057 was the most unstable with its ASV value of 1.004 followed by genotypes AC-BC-9 and Gammachis with their respective ASV of (0.985) and (0.913) (Table 7).

Table 7. IPCA1 scores, IPCA2 scores and ASV scores of fifteen black cumin genotypes

Key: The number in the parenthesis represent the rank of the values; GM = Grand Mean, IPCA1 = Interaction principal component axis one, IPCA2 = Interaction principal component axis two and ASV = AMMI Stability Values
With AMMI analysis of variance, the location effect was found the most influential factor in discriminating the seed yield of fifteen black cumin genotypes that were evaluated at six locations, contributing about 41.99% as compared to that of the genotypic effect and GEI effect with their percent contribution of 31.96% and 10.96%, respectively.
Generally, AMMI stability analysis identified the genotypes Dirshaye, Soressa, AC-BC-10 and 242834-1 as the most stable genotypes. On the other hand, the location Ali Woyyo was identified as the most favorable black cumin growing environment.

4. Summary and Conclusion
The analysis of variance of an individual environment revealed that seed yield showed highly significant difference (P ≤ 0.01) at all individual test environments. This pointed out that genotypes might perform differently at a specified test environment.
After the significant difference of genotype × environment interaction and homogeneous residual variation were corroborated, combined analysis was computed and showed that there were highly significant differences among the black cumin genotypes, environments and GEI. The observed highest variation to the total variations was attributed to the environmental effects. This inturn shows that the environment had contributed a great influence in varying the seed yield of the genotypes. Accordingly, environment had contributed about 47.16% to the total variations.
The combined mean seed yield values of genotypes averaged across the environments showed that Gammachis had the highest mean yield (1.35 ton ha-1) followed by Dirshaye (1.26 ton ha-1) while genotype 90575-2 had the lowest (0.78 ton ha-1) mean seed yield. Most of the total sum of squares of the AMMI model (41.99%) was attributed to the environmental effects and the rest were attributed to the genotypic effects (31.96%) and the GEI (10.96%).
AMMI stability analysis identified the genotypes Dirshaye as the most stable with ASV value of (0.093) followed by the genotype 242834-1 and Soressa with their ASV value of (0.095) and (0.109), respectively. On the other hand, the location Ali Woyyo was identified as the most favorable black cumin growing environment. In general, genotypes (such as Dirshaye, Gammachis, Soressa and Derbera) should be used for cultivation at all the test locations since they perfomed well as compared to the other tested black cumin genotypes.

5. References
Ahmed NU and Haque KR (1986). Effect of row spacing and time of sowing on the yield of black cumin (Nigella sativa L.). Bangladesh Journal of Agriculture, 11:21-24.
Akcura M, Yaya Y and Tanner S (2005). Genotype × environment interaction and phenotypic stability analysis for grain yield of durum wheat in the Central Anatolian region. Turkish Journal of Agriculture, 29: 369-375.
Akhtar MS and Shah MU (1993). Elemental constituents of ant diabetic screening of a folkloric medicinal plant prescription. International Journal Toxic, Occupation Environment Health, 2: 46-47.
Al-Ghamdi MS (2001). The anti-inflammatory analgesic and antipyretic activity of Nigella sativa. Journal of Ethnopharmacol, 76:45-48.
Al-Jassir MS (1992). Chemical composition and micro flora of black cumin (Nigella sativa L.) seeds growing in Saudi Arabia. Food Chemistry, 45: 239-242.
Ariyo OJ and Ayo-Vaughan MA (2000). Analysis of genotype × environment interaction of okra (Abelmoschus esculentus (L) Moench). Journal of Genetic Breeding, 54:33-40.
Atta MB (2003). Some characteristics of black cumin (Nigella sativa L.) seed cultivated in Egypt and its lipid profile. Journal of Food Chemistry, 83:63–68.
Bannayan MF, Nadjafi F, Azizi M, Tabrizi L and Rastgo M (2008). Yield and seed quality of Plantago ovata and Nigella sativa under different irrigation treatments: Indian Crops Production, 27: 11–16.
Bradu D and Gabriel KR (1978). The bi-plot as a diagnostic tool for models in two way tables: Techno metrics, 20: 47-68.
Champawat RS and Pathak VN (1982). Role of nitrogen, Phosphorus and potassium fertilizers and organic amendments in cumin (Cuminum cyminum L.) with incites by Fusarium oxysporum f. sp. cumin. Indian Journal Agricultural of Science, 58 (9): 728-730.
Crossa J (1990). Statistical analysis of multiplication trials: Advances in Agronomy, 44: 55-85.
Crossa J, Fox PN, Pfeffer WH, Rajaran S, and Gauch HG (1991). AMMI adjustment for statistical analysis of an international wheat trial. Theoretical Applied Genetics, 81:27-37.
Desalegn D, Solomon G, Balcha Y, Bedada G, Berhanu M, and Debebe M, (2004). Grain Yield Stability of Bread Wheat Genotypes in Favorable and Stressed Environments: 76-85.
Ferney GH, Alexei M and Aigul A (2006). Evaluation of Grain Yield Stability, Reliability and Cultivar recommendations in Spring Wheat (Triticum aestivum L.) from Kazakhstan and Siberia. Journal of Central European Agriculture, 4:649-660.
Fufa M (2018). Agronomic Performance, Genotype × Environment Interaction and Stability of Black Cumin Genotypes Grown in Bale, Southeastern Ethiopia. Adv Crop Sci Tech. 6: 358. doi:10.4172/2329-8863.1000358
Gauch HG and Zobel RW (1988). Predictive and postdictive success of statistical analysis of yield trials. Theoretical Applied Genetics, 76: 1-10.
Haji HM, Hunt IA (1999). Genotype × environment interactions and underlying environmental factors for winter wheat in Ontario. Canadian Journal of Plant Science, 79:49-505.
Iqbal MS, Ghafoor A and Qureshi AS (2010). Evaluation of Nigella sativa L. for genetic variation and ex-situ conservation. Pak. Journ. Bot., 42: 2489-2495.
Jansen PCM (1981). Spices, condiments and medicinal plants in Ethiopia; Their taxonomy and agricultural significance, Addis Ababa.Center for Agricultural Publishing and Documentation: 111-120.
Ministry of Agriculture and Rural Development (2003). Animal and Plant Health Regulatory Directorate. Crop Variety Register, Addis Ababa, Ethiopia: 183.
Mohammadi R and Amri A (2008). Comparison of parametric and non-parametric methods for selecting stable and adapted durum wheat genotypes in variable environments. Euphytica, 159:419- 432.
Mohammed MI (2009). Genotype × Environment Interaction in Bread Wheat in Northern Sudan Using AMMI Analysis. American-Eurasian Journal of Agricultural and Environmental Science, 6 (4): 427-433.
Mozzafari FS, Ghorbanli M, Babai A, Faramarzi M (2000). The effect of water stress on the seed oil of Nigella sativa. Journ. Essential Oil Res., 12: 36–38.
Mut Z, Aydin N, Bayramoglu HO and Ozcan H (2010). Stability of some quality traits in bread wheat (Triticum aestivum L.) genotypes. Journal of Environmental Biology, 31: 489-495.
Nergiz C and Otles S (1993). Chemical composition of Nigella sativa L. seeds. Food Chemistry, 48: 259-261.
Orgut (2007). Market Assessment Study, Ethiopian Nile Irrigation and Drainage Project. Main Report and Annexes, Ministry Of Water Resources, Addis Ababa.
Purchase JL (1997). Parametric analysis to describe genotype × environment interaction and yield stability in winter wheat. Department of Agronomy, Faculty of Agriculture of the Orange Free State: Bloemfontein, South Africa.
Purchase JL, Hatting H and Vandenventer CS (2000). Genotype × environment interaction of winter wheat in South Africa: Stability analysis of yield performance. South African Journal of Plant and Soil, 17: 101-107.
Romagosa I and Fox PN (1993). Genotype × environment interaction and adaptation: Plant Breeding Principles and Prospects; Chapman & Hall, London: 373–390.
Shewaye L (2011). Antifungal Substances from Essential Oils: MSc. Thesis, Addis Ababa University.
Takrun HRH, Dameh MAF (2008). Study of the nutritional value of Black cumin seeds agriculture; Duhok University, Iraq. Mesopotamia Journal Agriculture, 36 (1).
Taye G, Getachew T, Bejiga G (2000). AMMI adjustment for yield estimate and classifications of genotypes and environments in field pea (Pisum sativum L.). Journal of Genetic Breeding, 54:183-191.
Toncer O and Kizil S (2004). Effect of seed rate on agronomic and technologic characters of Nigella sativa L. International Journal of Agriculture and Biology, 3: 529-532.
Zobel RW, Wright MJ and Gauch HG (1988). Statistical analysis of a yield trial. Agronomic Journal, 80: 388-393.