Assessing New Banana Genotypes for Relevant Traits: Implication for Variety Selection New

Understanding the performance of new crop genotypes for traits of relevance is important in selecting potential cultivars to satisfy end-users. The objective of this study was to determine the performance of new banana genotypes for bunch mass (BMS) and BMS-related traits, resistance to black Sigatoka and sensory attributes. Eight cooking banana genotypes consisting of six new hybrid genotypes selected from advanced breeding trials and two control cultivars were evaluated in a randomized complete block design for three crop cycles at three locations in Uganda. Genotype, location, crop cycle and their interaction effects were significantly different for most traits assessed. The overall top two genotypes (“9058K-2” and “8099K-16”) combining high BMS, resistance to black Sigatoka and acceptable sensory attributes were identified. These genotypes are suggested as potential commercial cultivars for release to farmers in Uganda and/or other East African countries with similar environmental conditions to those where the genotypes were tested. It should be noted that high BMS/yield and resistance to diseases are not the only factors to consider when selecting banana genotypes that meet end-user needs. A combination of high BMS/yield, acceptable sensory attributes and resistance to diseases essentially influence the selection process of banana cultivars.


Introduction
Banana (Musa spp.), a perennial and vegetatively-propagated plant, is an important crop in the tropical and subtropical regions of the world [1]. It is a source of food and income for millions of smallholder farmers, especially in Asia, South and Central America, and sub-Saharan Africa (SSA) [2] [3]. Majority of the banana production in SSA is done on small plots and backyard gardens [3].
Every part of the banana plant is useful although the fruit is the most used plant part. The mature fruit of banana can be utilized in many forms, such as dessert when fully ripe, food when cooked and mashed or eaten directly after boiling or steaming [4] [5]. When cooked or steamed, banana, especially the cooking East African Highland Bananas (EAHBs) "Matooke", are characterized by a unique flat taste and aroma, golden yellow colour and a soft texture [6]. Fruits, especially those of beer type of bananas "Embire" also the EAHBs can be squeezed when fully ripe to produce a natural sweet juice "Eshande" which upon fermentation can be processed into wine and/or "Waragi"; a tradition drink enjoyed in most east African countries.
The highest per capita consumption of banana in the world is in the East African region, where one-third of the people depend on this crop as a staple food-the crop occupies between 20% and 30% of the acreage under cultivation [7]. Annual production of bananas in the region is worth US$ 4.3 billion, accounting for about 5% of the region's gross domestic product [8]. The predominant type of bananas grown in East Africa is the EAHBs. These are a subgroup of triploid banana cultivars, with genome code AAA [9].
In Uganda, millions of people rely on banana for income and daily food, with approximately 75% of farmers cultivating the crop [10]. Banana occupies the third largest cultivated area amongst staple food crops in Uganda [1]. Despite the benefits of growing bananas in the country, the yield gap between the actual and potential is high. For example, over the last 30 years, banana yields have been declining with low yields of 5 to 10 t/ha/year compared to potential yield of 70 t/ha/year and bunch mass dropping from 60 to 10 kg/plant or less [1] [11]. The number of years that banana plantations continue to be productive in some parts of Uganda, especially the central region, has been reduced to only five years compared to 50 years of plantation productivity in the past. The leading factors responsible for the decline in yield and productivity life are largely due to susceptibility of local cultivars to pests [11], especially banana weevils and nematodes; and diseases, particularly black Sigatoka [12] [13] [14].
Black Sigatoka, also known as black leaf streak disease, caused by the fungus

Banana Germplasm
Eight cooking banana entries consisting of six new genotypes from advanced banana breeding evaluation trials of NARO and two control cultivars (improved and landrace) from farmers' fields in Uganda were used for this study (Table 1).
Selection of six new genotypes was based on their previous performance for bunch yield of over 10 kg/bunch and resistance to black Sigatoka. The control cultivars "Kabana 6H" and "Mbwazirume" were selected based on their predominance in cultivation as improved and landrace cultivars, respectively. In addition "Kabana 6H" is a black-Sigatoka-resistant improved cultivar [12] whereas "Mbwazirume" is a most farmer-preferred landrace cultivar due to its superior food quality attributes (good taste, aromatic, good mouth feel, golden yellow colour) but susceptible to black Sigatoka.

Experimental Sites
Experiments were conducted at three sites: Mbarara Zonal Agricultural Research

Experimental Design
Experiments at each site were laid out in a randomized complete block design

Data Collection
Data were collected on the number of standing leaves (NSL), youngest leaf spot-

Statistical Analysis
The data collected were subjected to analysis of variance (ANOVA) using GenStat, version 14 [27]. Means were separated using the Least Significance Differences (LSD) at P < 0.05. To identify and select high yielding banana genotypes with resistance to black Sigatoka and combining good sensory attributes, all the genotypes were sorted and ranked based on their performance for each trait of the 10 traits assessed across sites and cycles using Microsoft Excel [28]. Individual trait ranks were added; and a genotype with the lowest overall rank of one or close to one was considered the best genotype. Ranking was conducted on a scale of 1 to 8; where; 1 = Excellent and 8 = Worst.

Performance of Genotypes in Response to Locations
Genotypes and locations were significantly different for all the agronomic traits assessed (Table 2). Their second order interaction effects were similarly significantly different for all traits except for bunch mass (Table 2).

Bunch Mass
The highest mean performance for bunch mass across genotypes was recorded in Mbarara (28.0 kg/plant) and lowest in Kawanda (25.4 kg/plant) ( Table 2).
Genotype "8099K-16" produced highest bunch mass (31.4 kg/plant), followed closely by "9058K-2" (31.24 kg/plant) and "Kabana 6H" (30.4 kg/plant). The least mean bunch mass of 13.8 kg/plant was recorded by "10054K-1" ( Means with the same small letter within the same column do not differ significantly according to Fisher's test of least significant differences (P < 0.05). Means with the same Capital letter within the same row do not differ significantly according to Fisher's test of least significant differences (P < 0.05). JN = Jinja, KW = Kawanda, MB = Mbarara. LSD (0.05) = Least Significant Difference at 5%.
All genotypes had least performance for bunch mass in Kawanda except for "6880K-2" and "Mbwazirume" whose least performance was recorded in Jinja.

Total Number of Clusters
The highest mean performance for number of clusters across genotypes was recorded in Mbarara (9.7), while Jinja and Kawanda produced almost the same number of clusters of 8.0 and 8.1, respectively (Table 2). Genotype "9058K-2" produced highest number of clusters (9.4) trailed by "8099K-16" with 9.2 clusters, and "10072K-10", "6880K-2" and "Kabana 6H", all with 9.0 clusters. Genotypes "9019K-3" and "Mbwazirume" had the same trend of performance for the number of clusters in Jinja and Kawanda while "6880K-2" performed slightly better in Jinja than Kawanda although their performance for the trait was not significantly different. For the rest of the genotypes, performance for number of clusters was highest in Mbarara, followed by Kawanda and least in Jinja.

Performance Genotypes in Response to Crop Cycles
Genotypes were significantly different for all the agronomic traits assessed while plant cycles were significantly different for only bunch mass, number of clusters and fruits (Table 3). Genotypes × crop cycle interaction effects were significantly different for all the traits (Table 3).

Bunch Mass
The bunch mass of all genotypes except "Mbwazirume" increased gradually from plant crop to second ratoon crop (

Total Number of Clusters
The total number of clusters for all genotypes except "Mbwazirume" increased gradually from plant crop to second ratoon crop (Table 3). "Mbwazirume", attained the highest number of clusters (8.6) at plant and second ratoon crops and the least number of clusters at the first ratoon (7.8). Averaged across genotypes, the highest number of clusters was observed at the second ratoon crop (9.5) and the least observed at plant crop (7.8).

Total Number of Fruits
The genotypes performance for total number of fruits took the same trend, increasing from plant crop to the second ratoon crop cycles except for "Mbwazirume" (

Number of Standing Leaves
Averaged across genotypes, the numbers of functional leaves at plant and ratoon crop cycles were not significantly different (Table 3). However, the highest mean performance for the trait was observed at the plant crop phase. With the exception of "Mbwazirume" and "10072K-10" that attained the highest number of functional leaves at the plant crop phase, the rest of the genotypes produced the highest number of functional leaves at the second ratoon crop phase.

Index of Non-Spotted Leaf
Averaged across genotypes, the Indices of non-spotted leaf at plant and ratoon crop cycles were not significantly different (Table 3). Nevertheless, the highest mean performances for these traits were observed at plant crop. Genotypes

Performance of Genotypes for Sensory Attributes
Banana genotypes differed significantly for all the sensory attributes evaluated (Table 4). For food taste, all genotypes had a score above 5.0 and the most superior genotypes for the trait were "8099K-16", "Mbwazirume" and "9058K-2". For food aroma and mouth feel, the most superior genotype was "Mbwazirume", which was trailed closely by "8099K-16" and "9058K-2". For food colour and overall acceptability, "8099K-16" performed highly, followed closely by "Mbwazirume" that was also followed closely by "9058K-2".

Overall Performance of the Genotypes for All the Traits Assessed
To identify and select high yielding banana genotypes with resistance to black Sigatoka and combining good sensory attributes, genotypes were ranked for performance for each of the traits assessed (Tables 2-4). Top performers for a combination of the desired traits were identified. Accordingly, genotype "9058K-2" was ranked the most superior (Table 5). This was followed closely by "8099K-16".
There was not a single genotype that was superior in all the traits assessed. For instance, the best genotype for bunch mass was "8099K-16" while the best for total number of clusters was "9058K-2". For the total number of fruits, the best genotype was similarly "9058K-2", while the best for number of standing leaves was "6880K-2".  Means with the same letter within the same column do not differ significantly according to Fisher's test of least significant differences (P < 0.05). Hedonic scale of 1 to 6, where: 1 = dislike extremely, 2 = dislike, 3=like fairly, 4 = like, 5 = like very much, 6 = like extremely. LSD (0.05) = Least Significant Difference at 5%.

Discussion
Understanding the performance of new cropgenotypes for traits of relevance is important in cultivar selection to satisfy end-users. This ultimately facilitates faster adoption of new cultivars when right cultivars are selected. The objective  [15]. Vuylsteke et al. [33] revealed that resistance to black Sigatoka results from the interaction between a major recessive gene and two modifiers with additive effects.
The sensory attributes (taste, aroma, soft mouth feel, and colour) of "9058K-2" and "8099K-16" were consistently ranked high and closest to those of "Mbwazirume". "Mbwazirume" is one of the most consumer-preferred landrace cooking banana cultivars due to high quality attributes of its cooked food. It dominates the banana markets in Uganda and is therefore highly marketable. Adoption rates of improved banana cultivars are often low regardless of their economic importance in terms of higher yield and resistance to diseases [3]. Reasons given  [23]. It is therefore important that new banana cultivars in addition to high yield and resistance to pests and diseases should possess good sensory attributes to meet farmer needs and preferences for higher adoption. In the present study we selected and recommended two banana hybrid genotypes "9058K-2" and "8099K-16" because of their high BMS, resistance to black Sigatoka and good sensory attributes.

Conclusion
Banana genotypes evaluated in this study showed high degree of variation for all the traits assessed, implying high potential for selection among them. Significant differences among crop cycles for BMS, NCL and NFT, with higher mean performance for these traits observed at the second ratoon crop revealed that the selection for BMS and BMS-related traits can be perfectly done at the second ratoon crop for optimal results. Non-significant crop cycle effects for NSL and INSL showed that selection for these traits can be done at any crop cycle of banana without losing any valuable information. The top two banana genotypes that combined high BMS, resistance to black Sigatoka and acceptable sensory attributes were "9058K-2" and "8099K-16" (Table 2). These two genotypes are recommended as potential commercial cultivars in Uganda and/or other East African countries with similar environmental conditions to those where the genotypes were tested.