INHERITANCE OF QUALITATIVE MORPHOLOGICAL TRAITS AND VARIATION OF QUANTITATIVE AGRONOMIC TRAITS OF ENSET ( ENSETE VENTRICOSUM (WELW.) CHEESMAN) CLONES OBTAINED FROM WOLAITA ZONE, SOUTHERN ETHIOPIA

. The present cultivated enset clonal landraces in Ethiopia originated from few wild progenitors. However, enset has a mixed mode of reproduction in which, the wild enset reproduces sexually through seeds, while cultivated enset is generally propagated vegetatively. The objective of this study was to understand the genetic structures of enset cultivars and estimate their genetic variability by evaluating the morphological data generated from progenies of cultivated and wild enset clones. Hence, seeds collected from six cultivated and four wild enset genotypes were used for this study. Data on four qualitative and six quantitative morphological traits were recorded from the progenies of the 10 enset genotypes. Progenies of seven enset genotypes segregated with 3:1 genetic ratio while progenies of the remaining genotypes segregated differently for the qualitative traits considered. With regard to the quantitative traits, the progenies of the 10 enset genotypes differed significantly for five of the six traits except pseudostem length. Generally the cultivated clones performed better than the wild types. This study demonstrated the possibility of creating genetic variation through selfing of the existing clones of enset for traits of interest and makes improvements either through selection or crossing the elite types to develop novel cultivars.


INTRODUCTION
Enset is considered mainly as an African crop that currently provides the staple food for one-fifth of Ethiopian population (Yemataw, et al., 2017;Borell, et al., 2019). It is a large perennial monocarpic herbaceous plant, similar in form to the related bananas of the genus Musa. Unlike to Musa species that has n=7, 10, and 11 set of chromosomes with various ploidy levels, Enset is a diploid plant with chromosome number 2n=18 with no record of polyploidy (Diro, et al., 2003). Ensete is geographically distributed in the wild in many parts of Sub-Saharan Africa and Asia with about 6 -7 species (Simmonds, 1962;Pursglove, 1972) in which Ensete ventricosum species is cultivated only in its native indigenous farming systems of south and south-western Ethiopia (Brandt et al., 1997). The highlands of southern part of the country form the geographical centre of the crop cultivation (Vavilov and Rodin, 1997) and the various ethnic groups in this region recognize and exploit many enset landraces.
The enset planting is complex, supports a denser population than any other farming system (Brandt et al., 1997). The crop has been domesticated and is cultivated for food, animal feed and fibre (Bezuneh et al., 1967), ensuring food security for about 20% of the human population in Ethiopia that depend on enset as one of the staple food sources. It is Ethiopia's most important root crop, a traditional staple crop in the densely populated parts of the country (George, 2004). This multipurpose culture crop used as source of large quantities of carbohydrate-rich food (Abraham et al., 2016), animal forage, fibre production, construction materials, as ornamental (Hölscher and Schneider, 1998). Moreover, products from enset are used in different forms in traditional medicine and a starch for textile, adhesive and paper industries is being produced (Diro and van Staden, 2005;Temesgen et al., 2014).
The present cultivated enset in Ethiopia originated from few wild progenitors.
However, enset has a mixed mode of reproduction in which, the wild enset reproduces sexually through seeds, while cultivated enset is generally propagated vegetatively. Naturally, vegetative propagation results in the genetic fixation, which could lead to loss of clones owing to diseases, and abiotic stress resistant due to selection pressures, or changes in land use systems. The wild Musaceae family have always been known for their broad genetic base and carry several desirable genes (Vuylsteke et al., 1995) which breeders should look in future. Seed propagation of enset might be one of the options to create variation and allow breeders to select the clones with desired traits with the knowledge of enset seed germination and seedling growing techniques to breed enset efficiently (Karlsson et al., 2012). So far, maintenance of the existing germplasm in the wild populations, as well as introduction of genes from wild or related species into the cultivated clones is useful to improve e.g. disease resistance and adaptation could have a major impact on future food security of Ethiopia.
Genetic diversity study on available genotypes either from molecular and phenotypic data may help to understand the extent of the variation in the species (Biswas, et al., 2020). The source of variation in enset crop lacks to pin point either due to cross pollination (recombination) or entirely due to ancestors' inherent genetic makeup.
The information generated from such researches explain the variation is due to the individual genetic constitutes which can help the breeders to design exploitation of genetic diversity in the species as a whole but not able to provide information how much is the breeder can create variation. Unlike to most vegetatively propagated species that are known to be polyploidy in nature and have homogenous plants in their clones with heterozygous loci in their genome, little is known about the genetic structure of the diploid species of E, ventricosum that produces morphologically uniform/ homogenous plants when multiplied by vegetative propagation. The improvement of cross pollinated crops exploits the variation within and between the family that can be manipulated by planned hybridization or recombination breeding. However, before suggesting the possibility of applying recombination breeding to exploit the within and between families variation, it is necessary to understand the extent of phenotypic variation inherited to the progenies since, the extent of variation within a seed cohort is not known. Morphological comparisons of genotypes within seed cohorts can help much to understand the extent of genetic variation achieved through seed propagation. Generating such information is needed to launch crossing program and selection of clones from natural outcrosses to develop new enset cultivars. Hence, the present studies has undertaken detailed morphological characterization on the progenies of each mother plant with the objectives of determining the number and types of qualitative morphological traits, and estimate the variability parameters for quantitative traits present among the enset genotypes.

Description of the Study Area
The study was conducted in Wolaita Sodo University field research station located in Wolaita Sodo town, Wolaita Zone, SNNPR region 315 km away from Addis Ababa.
The specific location of the experimental area lies at elevation of 1891 meter above sea level and its geographic coordinates are 37 0 45'08" E longitudes and 6 0 50'00" N latitude. Wolaita Zone covers an altitude range of 800 to 3500 meter above sea level.
The area experiences bimodal type of rainfall. The shortest rainy season stretches from March to April and the main rainy season extends from June to September. The 12 years average annual rainfall data (2003 to 2015 cropping years) was 1580 mm.
Minimum and maximum average annual temperature was 12.7 0 C and 23.7 0 C respectively and major soil type of the area was reported to be Nitosols (Fanuel et al., 2017) having well drained sandy loam textural class with low organic carbon content (Hailu et al., 2017).

Plant Material
Progenies of the mother plants of clonal landraces of enset cultivated in Wolaita zone and wild plants of enset genotypes collected from natural forests found in Dawuro and Keffae areas were used for this study. The enset collections used for this study were constituted from six commonly cultivated landraces and four wild plants of enset genotypes (Appendix Table 1). The progenies of each clone were generated from the seeds of the respective mother plant.

Design and Layout of the Field Experiment
Each of the progenies of the mother plants (the 10 clones) was planted in a single row of 16 plants using nested design. The spacing was 3 m between plants and 4 m between rows. All the management practices such as weeding, hoeing, mulching, watering and fertilizer application were properly and uniformly applied to all plots using the recommended practices of Areka Agricultural Research Center.

Data Collection
The data included both qualitative and quantitative parameters. Data for qualitative parameters were collected from all available plants in each plot. While for the quantitative parameters data were collected from four plants per plot. List of qualitative and quantitative parameters are depicted in Appendix Table 2.

Data Analysis
Chi-squared analyses were conducted to test the goodness of fit of the observed segregation to the theoretically expected ratios for a given genetic model to determine the number of genes involved in the inheritance of the qualitative characters.
The formula for calculating chi-square analysis ( 2 ) is described below: For a recombinant inbred (RI) population a 1:1 ratio is expected for a single gene.
However for an F2 (2 nd filial generation) population a single dominant gene is expected in a 3:1 ratio, and for a co-dominant single gene the expected genetic ratio will be 1:2:1.
Analysis of variance was computed using nested design for each quantitative character in order to estimate the variability among accessions for each trait. Hierarchical classification was used for the partitioning of the variation into different sources of variations (Table 3). The ANOVA was constructed by considering the experimental units (the four enset plants within each clone) as factor B nested within levels of factor A (the 10 clones) (Sokal and Rolf, 1969). The differences between treatment means was compared using least significant difference (LSD) test at 5% level of significance when the ANOVA showed the presence of significant difference between genotypes.

Variability Analysis
The phenotypic and genotypic variances of agronomic traits at each location were estimated using the following formula described by Burton and Devane (1953)

Estimation of heritability in broad-sense (h 2 b) and genetic advance (GA)
Broad-sense heritability (h 2 b): was calculated as the ratio of the genotypic variance to the phenotypic variance, using the following formula described by Allard (1960); Where, =heritability (in broad-sense) = genotypic variance =phenotypic variance Genetic advance (GA): was computed using the formula adopted from Johnson et al. (1955) and Allard (1960): Where: -GA= genetic advance at 5% selection intensity, K = the selection intensity (K= 2.06 at 5% selection intensity), σp is the phenotypic standard deviation and h 2 b is heritability in broad sense.
Genetic advance as percent of mean: GAM5% = Where: -GAM5%=Genetic advance as percent of mean at 5% selection intensity, GA = genetic advance, and X=mean value of the trait.

Variation for Qualitative Morphological Traits
Enset plant is usually propagated vegetatively through corms. Plants propagated through corms are genetically uniform, hence they are said to be clones. However, most asexually (vegetatively) reproducing plants when propagated through seeds (sexually) their progenies show genetically diverse genotypes. Similarly, the enset progenies considered in this study demonstrated genetic diversity in both qualitative and quantitative traits as they were propagated through seeds obtained from each of the ten mother plants. The data for all four qualitative traits showed single gene segregation confirmed by chi-squared analysis for single gene (non-significant for χ 2 < 3.841 at P = 0.05 and 1 d.f.) at F2 generation with genetic ratio = 3:1 for the eight landraces (Appendix tables 1, 2, 3, 4). On the other hand the cultivated landrace Gefetanuwa-1 didn't show segregation for all qualitative traits, while Gefetanuwa-2 segregated for a single gene with genetic ratio of recombinant inbred lines (1:1 at P = 0.05 and 1 d.f.) ( Table 4). The three qualitative traits; pseudostem color, petiole color and mid-rib color exhibited segregation for two distinct types of color classes for each trait (Table 4)

Analyses of Variances
Univariate analysis of variance computed for the quantitative agronomic traits showed significant differences (P<0.05) among the enset genotypes except for pseudostem length that displayed non-significant mean square for genotypes (Table   5). This study demonstrated the presence of significant variation among the genotypes for the agronomic traits that witnessed progress/improvement can be made for the traits considered through selection and breeding efforts. Leaf length exhibited presence of highly significant (P<0.01) difference between the genotypes (Table 5) indicating that this trait is the most varied among the quantitative traits.

Mean Performance of Genotypes
Estimated mean performances of the 10 enset genotypes for the sixth agronomic morphological traits are presented in Table 6. The result showed presence of significant differences for five of the traits viz. leaf length, leaf width, number of leaves, plant height, and pseudostem circumference at 5% probability level that further confirmed by mean comparison tests using the respective LSD values. The mean data indicated that mainly the wild genotypes had inferior performances compared to the cultivated clonal landraces with the exception of the genotype 'Wild 15' that showed average or competitive performance in all the traits evaluated ( Table   6). The genotype 'Wild 15' performed better than 'Alageena' and 'Gamo Gofa71' clones for majority of agronomic traits and also ranked second next to 'Arkia' for traits such as leaf width, number of leaves, pseudostem length and plant height. The cultivated clonal landrace 'Arkia' is the top performer for majority of traits except for pseudostem length on which 'Wild 15' was the top performer, whereas 'Wild 11' was the least (Table 6). The enset genotypes showed unique performances with respect to pseudostem length though statistically not significant; for instance, the least performing genotypes 'Wild 11 and Wild 10' performed better than the cultivated ones 'Alageena and Gamo Gofa71' suggesting that the wild enset genotypes can also contribute to the improvement of kocho yield apart from quality traits and stress resistance.

Estimates of Variance Components
The results of estimated variance components, phenotypic (PCV) and genotypic (GCV) coefficients of variation, broad sense heritability (hb 2 ), genetic advance (GA) and genetic advance as percentage of mean (GAM%) were calculated for the six traits investigated using the ANOVA computed between the tested genotypes and presented in Table 7.

Phenotypic and Genotypic Coefficients of Variation
Both the PCV and GCV values computed for the six traits ranged from 21.49 to 33.88 and 11.40 to 20.04 for leaf width and pseudostem circumference, respectively (Table 7). The value of phenotypic coefficients of variation were generally higher than the corresponding value of genotypic coefficients of variation for all traits studied indicating the influence of environmental differences occurred across years was significant, particularly annual climatic (weather) changes were important. High PCV was observed along with moderate GCV values for all the six traits studied.

Heritability in Broad Sense
Broad sense heritability ( ), which is an estimate of the total contribution of the genetic variance to the total phenotypic variance ranged from 0.197 (pseudostem length) to 0.38 (leaf length). The heritability value estimated was moderate for half of the traits; namely, leaf length, plant height and pseudostem circumference which might be due to presence of relatively higher genotypic variations among the enset genotypes and less effect of environmental influence on the expression of these traits.
The remaining three traits leaf width, number of leaves per plant and pseudostem length exhibited low estimate of heritability (Table 7) implying the environmental influence in the expression of these traits was higher as compared to the genetic variation between the genotypes.

Genetic Advance
The genetic advance percent of means (GAM) expressed ranged from 11.11% for leaf length to 24.42% for pseudostem circumference. This refers to the improvement of the characters in genotypic value for the new population compared with the base population in one cycle of selection is within the range of 11.11 % to 24.42 % at 5% selection intensity. High GAM was observed for pseudostem circumference (24.42%) whereas moderate GAM was obtained for the rest of the traits that showed there is huge potential for improving the enset yield through selection and breeding using the available germplasm (Table 7). N.B. 2g = genetic variance, 2p = phenotypic variance, 2e = environmental variance, GCV = genotypic coefficient of variance, PCV = phenotypic coefficient of variance, h2b = heritability in broad sense, GA5% = genetic advance at 5% selection intensity, and GAM5% = genetic advance as percentage of the mean at 5% selection intensity

DISCUSSION
Enset is a perennial crop mainly cultivated in the highlands of southern and southwestern parts of Ethiopia, particularly in densely populated areas of the country (Yemataw, et al., 2014) such as, Gurage, Silte, Wolaita, Gedeo, Sidama and Gamo Gofa zones. It is a staple food for nearly one-fifth of the country's population. The crop represents 65% of the total crop production in the southern regions of Ethiopia.
The major food types produced from matured enset plant are Kocho, bulla and amicho. Kocho is fermented starch processed from scraped leaf sheaths and corms; it constitutes the major product of enset. Several food recipes can be prepared from this product depending on the cultures; kitta (leavened bread), burseme, kocho frfir, etc.
Bulla is a liquid, which, is obtained when leaf sheaths and corms are pulverized; the liquid starch is dried to make white powder, Bulla is usually used to make porridge.
Amicho is prepared from pieces of corm/rhizomes of enset plant and boiled and eaten similar to the other root crops (Brandt, et al., 1997). The byproducts of enset can be used for fiber production that can be further processed to make different products; bags, ropes, twines, cordage, and mat.
Though enset has several benefits to the society little progress has been made in terms of improving the crop through selection and breeding works to develop

CONCLUSIONS
Enset is one of the major staple food sources for Ethiopian population. It is a highly resilient crop with regard to environmental stresses such as drought and frost.
However, little attention has been given in terms of improving the productivity of the crop mainly due to its local importance and perennial nature of the crop. So far only six improved cultivars have been released to growers. This study gave insight that there is huge potential to improve this crop through hybridization and clonal selection methods since enset has viable flowers and can easily propagated either through seeds or vegetatively with its corm. Heterosis or hybrid vigour can be fixed once we develop superior gene combinations through crosses of elite enset clones.
We can also create genetic variability through selfing of the various clonal landrace collections that can be used as sources of genes for quality and yield improvement as well as stress (both biotic and abiotic) tolerance.

Data Availability
The data used to support the findings of this study are presented in the manuscript.
Additional data can be obtained from the corresponding author upon request Appendix