Genetic Variability and Association of Yield and Yield-Related Traits under Moisture Stress in Common Bean Genotypes ( Phaseolus vulgaris L.) at Melkassa and Miesso, Ethiopia

,


Introduction
Common bean (Phaseolus vulgaris L.) is an annual leguminous plant that belongs to the genus Phaseolus, family Fabaceae, with pinnately compound trifoliate large leaves. Particularly in areas of sub-Saharan Africa (SSA), it is a nutritious crop that promotes food security. Te annual per capita consumption of common bean is higher among low-income people who cannot aford to buy nutritious food stuf, such as meat and fsh [1]. Dry beans, often called the "meat of the poor," provide micronutrients to over 300 million people in the tropics and are the second most important source of calories following maize [2]. Common bean (dry bean) is predominantly produced in Latin America and eastern and southern Africa where seasonal rainfall is erratic and soil moisture defcit often limits its yield production [3]. Dry beans that do not meet human food quality standards are used as feed for livestock.
Te common bean could be grown as a seed legume in dry-land rotations with winter wheat to increase production diversity [4]. Beans are rich in a type of antioxidant called polyphenols [5]. Antioxidants fght the efects of free radicals, which are chemicals that afect a wide range of processes in the body, from physical aging to cancer and infammation. People who consume beans may be less likely to die of a heart attack, stroke, or other cardiovascular health problem. Health benefts of beans are generally acquired from their direct attributes, including their high content of proteins, dietary fbers, low saturated fat content, vitamins, minerals, and phytochemicals, as well as replacement in the diet, when they substitute for animal products [6]. In Ethiopia, common bean is most likely introduced by the Portuguese in the 16 th century. Tere is a wide range of common bean types grown in Ethiopia, including mottled, red, white, and black varieties.
Beans grow well under average rainfall ranging from 500 to 1500 mm above sea level with an optimum temperature range of 16-24°C. Usually, high temperatures do not afect it if adequate soil water is present, although high night temperature will inhibit pollination [7]. Tey do poorly in very wet or humid tropical climate because of their susceptibility to bacterial and fungal diseases. According to [8], the current national average yield of common bean is 1.77 t·ha −1 . However, this yield is far less than the attainable yield (2.5-3.6 t·ha −1 ) under good management conditions [9]. Tis low productivity is due to lack of high yielding varieties, poor cultural practices, and other impacts of climate change. Te main challenges from climate change to agriculture and food production are the more frequent and severe droughts and foods and the higher pressure from insects and diseases. Te most serious impediment in the production of common beans is intensive drought occurrence brought by climate change [10]. Tis problem is escalating from one year to another [11].
Climate change can change weather patterns, resulting in altered temperature and rainfall efects in diferent regions, which can have concomitant impacts on the suitability of crops for continued cultivation in climate change-impacted regions [12]. In particular, elevated temperatures (heat) and reduced rainfall (drought) can reduce crop yields [13]. Although it is documented that common bean is susceptible to moisture stress or water defcit, the production of this crop in many places of the world is carried out under moisture stress conditions due to insufcient water supply by rainfall and/or irrigation, which causes yield loses of more than >66% [14]. Te adaptation strategies that will be necessary to implement at scale may be incremental (e.g., breeding new bean varieties or using agronomic practices such as irrigation) or transformational (e.g., involving changing to a different protein or high-value crop species or fnding an alternative livelihood that is more climate-resilient) [15]. On the other hand, legume/common bean has a major role in mitigating climate change. Tese include lower fossil energy use than N-fertilized systems; lower greenhouse gas emissions than N-fertilized systems; contributing to C sequestration in soil; opportunities to replace petro problem ducts as a source of feedstock for biofuels and biorefneries [16].
Terefore, superior varieties generated through plant breeding are the most efcient approaches for agricultural production to increase or at the very least remain steady under new pressures from climate change. Te study of genetic variability under moisture stress areas enables selection of drought-tolerant genotypes that would give better yields to ensure food security, especially for poor rural farmers. Te choice of promising genotypes from a diverse genetic base and their subsequent utilization for hybridization is one of the strategies for improving the productivity of common beans [17]. Terefore, the assessment of variability parameters, viz., phenotypic and genotypic coefcients of variation, heritability, and genetic gain, is a prerequisite for the planning and execution of a breeding program for the improvement of diferent qualitative and quantitative traits in any [18]. Genetic variability studies have been conducted in Ethiopia by a considerable number of researchers on the common bean [19][20][21]. Moreover, even though several common bean varieties have been released for diferent agroecologies of Ethiopia, their level of drought tolerance has not been well documented. Indeed, only a few studies have tested breeding lines for drought stress and farmers perception of their drought tolerance [22,23]. Terefore, keeping the above facts in view, the present study was carried out to understand the nature and extent of genetic variability, heritability, and genetic advance in some important traits of common bean genotypes under moisture stress. Tus, the main objectives were to assess the genetic variability of common bean genotypes for yield and yieldrelated traits, estimate association of traits, and determine direct and indirect efects of traits on yield.

Experimental Design and Procedures.
Te experimental material comprising of 25 genotypes of common bean genotypes was obtained from the Melkassa Agricultural Research Center. Te experiment was conducted using 5 × 5 triple lattice design on plot sizes of 3.2 m 2 , each with 4 rows of 2 m length. Te spacing was 0.4 m and 0.1 m between rows and plants, respectively. Te spacing between replications was 1.5 m and 1 m between the incomplete blocks. Data were collected from the two central rows, leaving the plants at the ends of each row on both sides and the two outer rows for the border efect. All necessary cultural operations like weeding and hoeing were done as and when required during the growing period. In Table 1, the lists of the genotypes with their respective sources are presented.

Data Collection.
Data were collected on single plant and plot bases. On a plant basis, data were collected from fve randomly selected plants from each genotype in each replication, namely, plant height (cm), number of nodes on the main stem (count), internodes length (cm), number of pods per plant (count), and number of seeds per pod (count). While the data on plot basis were collected from the two central rows, leaving the two plants at the ends of each row on both sides, which include, days to 50% fowering, days to 90% maturity, grain flling period, hundred seed weight (g), and seed yield per plot (kg).

Data
Analysis. Data were analyzed for estimation of genotypic and phenotypic coefcients of variation according to the methods suggested by [25]. Heritability in the broad sense was calculated following a method adopted by [26]. Te expected genetic advance (GA) under selection was calculated assuming the selection intensity of 5% was calculated as proposed by [27]. Te genotypic coefcient of variation (GCV) and phenotypic coefcient of variation (PCV) were calculated as per [25] (Table 2). Trait associations between yield and yield-related traits were computed using the method suggested by [30].
Path coefcient analysis was performed using the correlation coefcients to know the direct and indirect efects of yield components on grain yield using the general formula of [31].
where i � coefcient of selection which is 2.06 at 5% selection intensity, σp � phenotypic standard deviation, and h 2 � heritability in broad sense. Genetic advance expressed as percentage over mean (GAM) in percent as suggested by Johnson et al. [27].
where GAM � genetic advance as percent of mean, GA � genetic advance, and X � genetic mean of the character.

Results and Discussion
Highly signifcant diferences among genotypes (P ≤ 0.01) were observed for 10 traits studied at individual location (Table 3), indicating the presence of variability for further improvement for yield and yield components of common bean. Both at Melkassa and Miesso, days to fowering, hundred seed weight, and seed yield showed highly signifcant (P ≤ 0.01) diferences among the genotypes. At Melkassa, grain flling period, plant height, pods per plant, and seeds per pod showed signifcant diferences (P < 0.05).
On the other hand, days to maturity, internode length, and number of nodes per plant did not show signifcant differences. While at Miesso, plant height, days to maturity, and grain flling period showed a signifcant diference (P < 0.05). However, internode length, number of nodes per plant, number of pods per plant, and seeds per pod did not show signifcant variations among the tested genotypes (Table 3). At Melkassa, the shortest days to fowering (35 days) were scored for genotype DRKDDRB-93 and only genotype DRKDDRB-34 matured earlier than the check. Te plant height ranged from 36.67 cm for genotype DRKDDRB-49 to 55 cm for DRKDDRB-33 with a mean of 45.87 cm. Te shortest internode length (2 cm) was found on genotypes (DRKDDRB-70, DRKDDRB-96, and DRKDDRB-66), , and DRKDDRB-66 (764 kg·ha −1 ). Generally, yield performance was better at Melkassa than Miesso, indicating its potential for common bean production.

Estimates of Coefcient of Variation.
Genotypic and phenotypic coefcients of variation are used to measure the variability that exists in a given population [25] and are essential in opening a breeding program and in developing better varieties. At Melkassa, the phenotypic coefcient of variation (PCV) values ranged from 0.81% for days to maturity to 30.75% for pods per plant, whereas the genotypic coefcient of variation (GCV) ranged from 0.28% for days to maturity to 18.18% for pods per plant. Moreover, next to pods/plants, other traits with high PCV values include internodes length (29.53%) and grain yield (21.02%). A similar result was reported by [32] the regarding high PCV and GCV values of pods per plant. Te values indicate the existence of variability in common bean genotypes for pods per plant, internodes length, and seed yield, indicating the possibility improvement for these traits through selection. However, internodes length had a low GCV (6.99%); thus, its relatively high PCV could be due to a high environmental efect, implying that there is limited scope for its improvement through selection. In addition to days to maturity, which had the lowest PCV and GCV values, other traits with low PCV include days to fowering (1.63%) and grain flling period (1.47%), while those with low GCV are plant height (8.27%), number of nodes (7.52%), and hundred seed weight (9.05%). It implies that those traits with low PCV and GCV values have limited scope for improvement through selection due to the high infuence of the environment. PCV values were generally higher than their corresponding GCV values for all the traits considered at Melkassa (Table 4). Tis indicates that the apparent variation was not only due to genotypes but also due to the infuence of the environment on the expression of the characters. Tis result agrees with what [33] reported.
At Miesso, the phenotypic coefcient of variability (PCV) values ranged from 2.65% for days to maturity to 26.51% for grain yield. On the other hand, the GCV values ranged from 1.60% for days to maturity to 22.64% for grain yield. Among all characters, high GCV and PCV were  Generally, high heritability shows the reliability with which a genotype can be recognized by its phenotypic expression [37], mainly due to the major role of genotypic factors in the expression of the characters. However, heritability alone provides no indication of the amount of genetic improvement that would result from the selection of individual genotypes. Hence, knowledge about heritability coupled with genetic advance and genotypic coefcient of variations are most useful. In this study, the expected genetic advance as a percent of mean (GAM) ranged from 0.20% for days to maturity to 30.75% for grain yield at Melkassa. Pods per plant also had high GAM (22.10%) next to seed yield at Melkassa, while moderate GAM was observed for plant height (10.18%), hundred seed weight (14.89%), and seeds per pod (15.33%). At Miesso, GAM ranged from 1.99% for days to maturity to 39.77% for seed yield, while the moderate GAM value was obtained for hundred seed weight (12.92%). In the present study, high heritability coupled with high GAM was observed for seed yield at both locations, while high heritability coupled with moderate GAM was observed for hundred seed weight at Melkassa. Generally, high heritability together with GAM provides a better result than heritability alone and enables considerable improvement in the characters by selecting the best individuals and predicting the results [38].

Association among Traits
Te analysis of the relationship among yield-related traits and their association with seed yield is essential to establish selection criteria [39]. Te result at Melkassa indicated that seed yield showed positive and signifcant association with days to fowering, internodes length, and pods per plant. Terefore, any improvement of these traits can result in a substantial increment on seed yield. An author [40] also testifed that seed yield was found to be positively correlated with number of pods per plant in common bean. Te Advances in Agriculture 5 positive association of number of pods per plant with seed yield is in agreement with previous report by [41]. Moreover, [42] reported a signifcant positive correlation between the number of pods per plant and seed yield in pigeon pea. However, the grain fling period showed a signifcant and negative correlation with seed yield. On the other hand, at Miesso, seed yield showed a positive and signifcant phenotypic association with seeds per pod (r � 0.40) and hundred seed weight, indicating that any improvement/ management that increases seeds per pod may result in improvement of seed yield to some extent. Te correlation study also showed signifcant associations among the yield related traits. At the Melkassa location, days to fowering showed a highly signifcant positive correlation with days to maturity (r = 0.49), and similarly, highly signifcant positive correlations were observed between days to maturity and grain flling period (r = 0.65) and internodes length with seeds per plant (r = 0.24) ( Table 6). While at Miesso (Table 7), positive and signifcant associations were observed between plant height and days to fowering (r = 0.48), days to fowering and days to maturity (r = 0.28), days to maturity and grain flling period (r = 0.87), and days to maturity and pods per plant (r = 0.26). As the result of path analysis showed that, internode length (0.243), pods per plant (0.216), and days to fowering (0.208) had high positive direct efects on seed yield at Melkassa (Table 8). A similar result was reported by [43]. Tese characters could be considered as main components of selection in a breeding program for obtaining higher seed yield. And also at Miesso (Table 8), days to maturity (0.245), seeds per pod (0.361), and hundred seed weight (0.211) showed a high positive direct efect on seed yield at the phenotypic level. However, grain flling period had a high positive indirect efect (0.214) through the days to maturity. Generally, the phenotypic residual values at Melkassa (0.089) and at Miesso (0.085) were low, indicating that the traits which were included in the phenotypic path analysis explained 91.1 and 91.5% of the variation in seed yield, and other factors not included in the study can explain only 8.9 and 8.5%, respectively. It is suggested that maximum emphasis should be given on the above traits in selecting common bean genotypes for seed yield in addition to the grain flling period at Miesso.

Conclusion
Since climate change can afect crop production, the study of genetic variability under moisture stress areas enables selection of drought-tolerant genotypes that would give better yields to ensure food security through the selection of traits based on high genetic parameters and keeping them for breeding purposes. Highly signifcant diferences among genotypes (P ≤ 0.01) were observed for ten traits studied at individual locations. However, the results at both locations were not consistent. Terefore, the recommendation of genotypes location-wise is crucial. Especially those genotypes which took shorter days to mature (early maturing genotypes) DRKDDRB-34, DRKDDRB-65, DRKDDRB-55, DRKDDRB-37, and DRKDDRB-70 could be recommendable for moisture stress area due to they need shorter rainy season and can scape moisture stress. Te highest yielding genotypes were DRKDDRB-32 (2243.3 kg·ha −1 ), followed by DRKDDRB-80 (1948.7 kg·ha −1 ) and DRKDDRB-55 (1752.3 kg·ha −1 ) at Melkassa, while at Miesso, DRKDDRB-94 (777 kg·ha −1 ), DRKDDRB-36 (775.67 kg·ha −1 ), and DRKDDRB-66 (764 kg·ha −1 ) could be recommended for a breeding program or as breeding material. Generally, yield performance was better at Melkassa than Miesso, indicating its potential for common bean production. Te experimental studies revealed that substantial amount of genetic variability among the genotypes under study. Genetic parameters in association with the correlation study indicated that primary emphasis should be given on days to fowering, internodes length, pods per plant, seeds per pod, and hundred seed weight for selection of superior genotypes. Hence, selection of those genotypes with high GCV, heritability, genetic advance, and positive correlation coefcient and direct efect on seed yield can be recommended for further yield improvement of common bean at respective location. Generally, legume/common bean has a major role in adaptation and mitigation of climate change because early maturing common bean has the chance to escape a rain shortage if they grow under moisture stress area.

Data Availability
Te raw data and additional information are available from the corresponding author upon request.

Disclosure
An earlier version was submitted to Haramaya University as MSc Tesis titled "Genetic Variability and Association of Morphoagronomic Traits in Selected Common Bean (Phaseolus vulgaris L.) Genotypes Grown under Moisture Stressed Environments in Ethiopia."

Conflicts of Interest
Te authors declare that there are no conficts of interest.

Authors' Contributions
Meseret Tola initiated the research, wrote the research proposal, conducted the research, performed data entry and analysis, and wrote the manuscript. Bulti Tesso and Berhanu Amsalu were involved in analysis, methodology,  supervising, writing, reviewing, and editing of the research proposal and manuscript and also read and approved the fnal manuscript.