Variability, heritability and genetic advance in sugarcane (Saccharum spp. hybrid) genotypes

Abstract

The study was carried out with the aim of estimating variability, heritability, and genetic advance for 196 sugarcane genotypes. Results of the analysis of variance identified a very highly significant variation among the genotypes in all the traits evaluated. Except for stalk length and single cane weight, all of the traits’ lowest and highest mean values were present within the genotypes from fuzz, indicating the wide ranges of variability within the genotypes. Except for single cane weight, the highest mean values for all the traits were seen within the genotypes of fuzz, implying greater genetic potential for the genotypes of fuzz for these traits. The number of sprouted buds, number of tillers, and cane yield had high genotypic and phenotypic coefficients of variation, heritability, and genetic advance as a percentage of the mean. This indicates the presence of high levels of heritable genetic variations for these traits, implying that these traits could be improved through phenotypic selection. The number of stalks, plant height, stalk length, internode length, stalk diameter, single cane weight, brix, and purity had high to moderate heritability with moderate to low genetic advance as percentages of the means. This indicates that the phenotypic expression of these traits is highly influenced by non-genetic factors. Thus, it could be difficult to directly improve these traits via phenotypic selection; instead, efforts should be directed toward better management practices. This study presents the first evidence for the existence of variability in the newly introduced sugarcane genotypes that should be exploited.

Keywords:

• Variability
• heritability
• genetic advance
• genotype
• phenotype

PUBLIC INTEREST STATEMENT

Despite the suitability of Ethiopia’s agro-ecology for sugarcane production, crop productivity is declining rapidly, with one of the major reasons being a shortage of improved sugarcane varieties. Most of the introduced sugarcane varieties were unable to adapt to the diverse agro-ecologies of the country. In addition, the world climate is changing continually, influencing the phenotypic expression of important traits. Thus, the development of high-yielding adaptable sugarcane varieties and resilient to the changing climate is critical to the sugar industry’s profitability and sustainability. To do so, curated information on the presence of heritable genetic variation with high predictable genetic gain within the available germplasm is mandatory. Identification of traits with high heritable variation and genetic gain, and selection of individual plants with a high mean value of these traits help identify promising genotypes to be used for commercial purposes and as parents for crossing to further enhance the traits

1. Introduction

Sugarcane (Saccharum spp. hybrids), which belongs to the family Poaceae (Gramineae) and genus Saccharum, is one of the most efficient converters of solar energy into sugars with high sucrose accumulation features and the exceptionally highest biomass production potential among plant species (Gianotto et al., 2011;Hoang et al., 2015;Mirajkar et al., 2019). It is one of the earliest crops grown worldwide for sugar and more recently for bioethanol production. It has been cultivated by smallholder farmers in Ethiopia since the 16th century, preceded by commercial cultivation for the production of sugar in 1951 (T. G. Esayas et al., 2018). Since the commencement of commercial production, sugarcane has been exploited to produce additional valuable by-products, such as molasses and bagasse, in addition to sugar for household and industrial consumption. This has created enormous employment opportunities for the local population and made significant contributions to the national economy (T. Esayas et al., 2016a;Teklemariam, 1991).

In addition, the country’s climate and soil types have proven to be ideal for sugarcane growth and productivity (Anon., 2014;Semiea et al., 2019). However, the country is unable to meet national sugar demand, and sugar production per hectare per year in the major sugar mills falls from 166 to 84, 140 to 101, and 165 to 157 t/ha, respectively, for Finchaa, Wonji, and Metahara sugar estates from 1998 to 2019 harvesting years. One of the main factors identified as accountable for the decline in productivity has been the shortage of high-yielding, improved sugarcane cultivars that can be adapted to the different agro-ecologies of the Ethiopian sugarcane plantations (Kebede et al., 2013;T. Esayas et al., 2016a).

Development of an improved sugarcane variety requires reliable information on the availability of sufficient heritable genetic variation with predictable genetic advance for the desired trait within the available germplasm. Genetic variability, a measure of the tendency of individual plant’s traits in a population to differ from one another, is a major contributor to successful plant breeding as it allows breeders to further improve the traits to develop new varieties. The presence of a greater range of trait variation in the germplasm offers a better opportunity for trait enhancement through selection.

Knowledge of the mode of inheritance of traits and the expected genetic gain is of paramount importance in planning effective breeding experiments, especially for perennials like sugarcane. In addition, identification of the heritable and the non-heritable portions of the observed variability are essential to get evidence of the genetic control over the expression of a particular trait and its phenotypic reliability to predict its breeding value (Ullah et al., 2012). High heritability is not always associated with high genetic advance, and thus, heritability estimates and genetic advance should always be considered simultaneously (Amin et al., 2004). High heritability values should be combined with a high genetic advance to ensure effective selection for trait improvement and a high prediction of the expected genetic gains (Johanson et al., 1955;Udeh & Ogbu, 2011).

Careful planning of breeding experiments for the successful development of an improved variety requires a clear understanding of the heritable proportion of variation and the magnitude of the predicted yield improvement that could be attained through selection (S. Kumar et al., 2019). Studies on the genetic variability, heritability, and anticipated genetic advancement of sugarcane morpho-biochemical traits have been conducted in Ethiopia (Diribu et al., 2020;T. Esayas et al., 2016b; Feyisa et al., 2014).

However, the type of genetic material, the trait being measured, and the environmental conditions to which the material is exposed determine the extent of variability and heritability (Dabholkar, 1992;Falconer & Mackay, 1996). Additionally, 178 of the 184 exotic genotypes used in this study were F1 crosses that were recently introduced, and their performances were unknown. Therefore, the objective of this study was to assess the genetic variability, heritability, and genetic advance of 196 sugarcane genotypes.

2. Materials and methods

2.1. Description of the study site

The study was conducted at the Metahara sugar estate of the Ethiopian Sugar Industry Group during the growing season of 2021–2022. Metahara is situated 950 m above sea level at 8° 51“N latitude and 39° 52” E longitude. With an average annual rainfall of 554 mm and average low and high temperatures of 17.5°C and 32.6°C, respectively, the area has a semi-arid climate.

2.2. Experimental materials and design

The study consisted of 196 sugarcane genotypes, of which the majority, 179 (including 178 recently imported fuzz-derived germplasms and one standard check), were obtained from Barbados. From the remaining 17, 12 were local collections, while four and one were from Mexico and South Africa, respectively. The list of genotypes, years, and sources of introductions is presented in Table 1. Thus, the sugarcane genotypes chosen for this study were both uncharacterized and from old collections.

Table 1. Country of origin, number of genotypes from each country and list of the sugarcane genotypes used for the study

Geographical origin	Number of genotypes	Name of the genotypes
Barbados	179	B153–1, B154–5, B154–6, B360–3, B489–1, B489–10, B489–11, B491–7, B491–8, B491–17, B491–18, B498–9, B498–10, B516–5, B516–6, B516–15, B516–16, B516–25, B516–26, B516–35, B516–36, B516–45, B516–46, B516–55, B516–56, B516–65, B516–66, B517–1, B517–10, B517–11, B517–20, B517–21, B517–30, B517–31, B517–40, B517–41, B517–50, B517–51, B519–1, B519–10, B519–11, B519–20, B519–21, B527–1, B527–10, B527–11, B527–20, B527–21, B527–30, B527–31, B528–1, B528–10, B528–11, B528–20, B528–21, B528–30, B528–31, B528–11, B546–10, B546–11, B546–20, B546–21, B549–1, B549–10, B549–11, B549–20, B549–21, B549–30, B549–31, Local 435, B552–10, B552–11, B552–20, B552–21, B556–1, B556–10, B556–11, B556–20, B556–21, B558–1, B630–1, B558–10, B563–1, B563–10, B563–11, B563–20, B563–21, B564–1, B564–10, B564–11, B566–1, B566–10, B566–11, B566–20, B566–21, B568–1, B572–1, B572–10, B572–11, B594–1, B599–1, B628–1, B498–11, B635–1, B635–10, B635–11, B645–1, B651–1, B658–1, B658–10, B658–11, B658–20, B658–21, B644–1, B644–10, B668–1, B676–1, B688–1, B688–10, B688–11, B690–1, B690–10, B690–11, B690–1, B694–1, B694–10, B694–11, B685–1, B700–1, B707–1, B52–298, B708–9, B708–2, B707–2, B700–2, B685–2, B694–12, B694–9, B694–2, B691–9, B691–2, B690–12, B690–9, B930–2, B688–12, B688–9, B688–2, B678–2, B668–2, B644–12, B644–9, B558–10, B658–22, B658–10, B658–12, B658–9, B658–2, B651–2, B645–2, B644–2, B635–9, B635–2, B630–2, B628–2, B599–2, B594–2, B572–12, B572–9, B572–2, B568–2, B566–22, B566–19, B566–12, B566–9, B566–2, B564–12, B564–9, B690–2, B563–22, B563–19
Ethiopia	12	local − 434, Local 435, Yejima-shenkora, Kay shenkora,62-tikur ageda,72-Shenkora ageda, 93-Barabure Shenkora,111- dikala, Local −106, Local −189, Local − 53, Local 421
Mexico	4	MEXY52129, MEXY 57,197, MEXY54/245, MEXY53/142
South Africa	1	NCO-334

The experiment was arranged in a partially balanced lattice design with two replications. The spacing between adjacent plots, incomplete blocks, and border spacing were 1.5 m, 2.9 m, and 5 m, respectively. The size of each plot was 21.75 m2 (three furrows of 5 m in length with 1.45 m in width). The planting materials were seven-month-old cane cut into three bud setts with an end-to-end planting pattern. The experiment was carried out on the light soil (Cambisol) type of the sugarcane plantation fields. The sugarcane genotypes were the only treatments, while estate-recommended agronomic management practices were used throughout the growth period.

Data on the agro-morphological and biochemical traits of sugar cane were collected. The agro-morphological traits were counted, including the number of sprouted buds per plot 30 and 45 days after planting and the number of tillers per plot three, four, and five months after planting. In addition, plant height and number of stalks per plot ten months after planting, number of internodes per stalk, length of internodes, stalk diameter, weight of individual canes, and cane yield were recorded. Data on biochemical traits, namely, brix percent, pol percent, purity percent, estimated recoverable sucrose percentage, and sugar yield from cane juice analysis were also collected at harvest or 18 months after planting. The collected data were subjected to an analysis of variance following the procedure for a partially balanced lattice design with the variance component method using the Agricolae and Mass packages in the statistical software program R.

2.3. Variability, heritability, and genetic advance analysis

The variability among sugarcane genotypes was estimated using range, mean, standard error, phenotypic and genotypic variance, and coefficients of variation. The resulting components of variance were used to compute phenotypic and genotypic variability, broad sense heritability, genetic advances in absolute units, and genetic advances as a percentage of the mean.

Accordingly, genotypic (σ²g) and phenotypic (σ²p) variances were estimated based on the method suggested by Burton and De vane (1953). Thus, σ²g = (σ²t - σ²e)/r; and σ²p = σ²g + σ²e. Where σ²g = genotypic variance; σ²t = mean square of the particular trait; σ²e = mean square of error (environmental variance); r = the number replications; and σ²p = phenotypic variance.

The phenotypic coefficient of variation (PCV) and genotypic coefficients of variation (GCV) were computed using the formula described byR. K. Singh and Chaudhary (1999) as: GCV (%) = $\left(\sqrt{\sigma 2\mathrm{g}}/\mathrm{X}\right)$*100; PCV (%) = $\left(\sqrt{\sigma 2\mathrm{p}}/\mathrm{X}\right)$*100. Where x = the grand mean of the trait. Heritability in broad sense (H²) = (σ²g/σ²p) *100 was calculated for each of the traits by using the formula described byAllard (1960).

The expected genetic advance in absolute units (GA) under selection was computed as: GA = k *$(\sqrt{\sigma 2\mathrm{p}}$) * (σ²g/σ²p) = k*$\hspace{0.17em}(\sqrt{\sigma 2\mathrm{p}}$)* H² = k* σ* H²; and Genetic advance as percent of mean (GAM) = (GA/x) *100. Where k denotes the standardised selection differential at 5% selection intensity (2.063) and σ represents the phenotypic standard deviation as described byJohanson et al. (1955).

3. Result and discussion

3.1. Genetic variability of 196 sugarcane genotypes

Analysis of variance confirmed that all the traits considered had a very highly significant variation (Table 2). This indicates the presence of substantial variation among the sugarcane genotypes for all the traits tested. This implies the existence of sufficient variability and the suitability of phenotypic selection for sugarcane yield improvement.

Table 2. Analysis of variance for agro-morphological and biochemical traits for 196 sugarcane genotypes

Traits	Replication (1)	Genotype (195)	Block(Replication) (26)	Residual (169)	Mean	CV (%)
SPN30	1649.02***	137.58***	30.00^ns	28.30	20.86	25.50
SPN45	2011.59***	94.40***	82.00**	36.90	24.26	25.00
TL3MAP	12880.10***	1148.60***	447.20*	285.90	59.49	28.40
TL4MAP	16174.30***	1236.90***	370.40^ns	328.80	70.45	25.70
TL5MAP	14608.30***	1127.20***	396.90^ns	398.40	73.99	27.00
SC10MAP	12012.50***	558.00***	292.60^ns	223.10	64.96	23.40
PH10MAP (m)	16.87***	0.17***	0.12*	0.07	1.91	13.90
SL18MAP (m)	3.15***	0.23***	0.07^ns	0.06	2.82	8.40
NI	272.28***	24.22***	12.14^ns	11.25	30.29	11.10
IL (cm)	0.95^ns	3.41***	0.84^ns	0.86	9.42	9.90
SD (cm)	0.012^ns	0.13***	0.03^ns	0.04	2.65	7.70
SCW (kg)	2.07***	0.22***	0.14^ns	0.09	1.66	17.80
CY (t/ha)	1342.42***	38.33***	33.70**	17.32	14.88	28.00
Brix (%)	4.39^ns	2.75***	1.92^ns	1.54	19.68	6.30
Pol (%)	1.77^ns	3.62***	2.29^ns	1.82	18.33	7.40
Purity (%)	9.10^ns	7.61***	4.38^ns	4.67	93.03	2.30
RS (%)	0.42^ns	2.46***	1.49^ns	1.20	13.13	8.40
SY (t/ha)	461573***	14787***	13683**	7275	29.25	29.20

Notes: Numbers in parenthesis represent the degree of freedom of the respective sources of variations; SPN30 and SPN45 = number of sprouted buds per plot 30 and 45 days after planting, respectively; TL3MAP, TL4MAP and TL5MAP = number of tillers per plot three, four and five months after planting, respectively; SC10MAP = number of stalks per plot 10 months after planting; PH10MAP = plant height ten months after planting; SL18MAP = stalk length 18 months after planting; NI = number of internodes; IL = internode length; SD = stalk diameter; SCW = single cane weight; CY = cane yield; RS = recoverable sucrose; and SY = sugar yield. The efficiency factor as compared to RCBD was 88.24%; alpha = 1 % (Tukey HSD test). “***” = p < 0.0001; “**” = p < 0.001; “*” = P < 0.01 and “ns”= p > 0.05.

T. G. Esayas et al. (2018) reported a very highly significant variation among the evaluated sugarcane genotypes for the number of sprouted buds, stalk number, internode number, internode length, plant height, stalk length, stalk diameter, cane yield, brix, recoverable sucrose, and sugar yield, which agrees with the current study results. Feyisa et al. (2014) also found a highly significant variation in the number of stalks, stalk diameter, cane yield, recoverable sucrose, and sugar yield among the tested sugarcane genotypes, which is consistent with the findings of the current study.Alam et al. (2017) and Gowda et al. (2016) similarly discovered very high and significant variation in sugarcane for internode length, internode number, stalk length, brix, and single cane weight. Thus, the current study’s findings corroborate those researchers’ earlier results that there is a great deal of variability in sugarcane for the aforementioned traits.

The examined sugarcane genotypes showed a very highly significant variance for the number of sprouted buds per plot, the number of stalks per plot, the number of internodes per stalk, the internode length (cm), the plant height (m), the stalk length (m), the stalk diameter (cm), the cane yield (t/ha), the brix (%), the recoverable sucrose (%), and the sugar yield (t/ha). In their previous studies on sugarcane,T. G. Esayas et al. (2018) and Feyisa et al. (2014) also discovered a highly significant difference in the number of stalks, stalk diameter, cane yield, recoverable sucrose, and sugar yield.Alam et al. (2017) andGowda et al. (2016) also discovered a very high and significant variation for internode length, internode number, stalk length, brix, and single cane weight in sugarcane, which agrees with the results of the current study. The current study’s findings therefore support the earlier findings of those researchers that sugarcane exhibits high levels of variability for the aforementioned traits.

The mean performance, range, coefficient of variation, heritability, and genetic advance for tested sugarcane genotypes are presented in Table 3.

Table 3. Estimates of variability, heritability, and genetic advance for 13 agro-morphological and five biochemical traits of 196 genotypes of sugarcane

Traits*	Range	Mean ± SE	σ^2e	σ^2g	σ^2p	ECV	GCV	PCV	H^2	GA	GAM
SPN30	4.74–48.59	20.86 ± 3.78	28.54	54.52	83.06	25.61	35.40	43.70	65.64	12.32	59.09
SPN45	3.60–49.12	24.26 ± 4.63	42.93	25.74	68.67	27.01	20.91	34.16	37.48	6.40	26.37
TL3MAP	8.44–170.42	59.49 ± 12.40	307.41	420.58	727.99	29.47	34.47	45.36	57.77	32.11	53.98
TL4MAP	12.0–165.0	70.46 ± 12.93	334.33	451.27	785.60	25.95	30.15	39.78	57.44	33.17	47.07
TL5MAP	26.94–162.34	73.99 ± 14.11	398.20	364.50	762.70	26.97	25.80	37.33	47.79	27.19	36.75
SC10MAP	28.13–115.77	64.96 ± 10.78	232.37	162.83	395.20	23.46	19.64	30.60	41.20	16.87	25.97
PH10MAP(m)	1.10–2.63	1.91 ± 0.20	0.08	0.05	0.13	14.45	11.64	18.56	39.33	0.2871	15.04
SL18MAP(m)	1.76–3.99	2.82 ± 0.17	0.06	0.09	0.14	8.57	10.38	13.46	59.46	0.4653	16.49
NI	21.79–39.54	30.29 ± 2.38	11.37	6.43	17.79	11.13	8.37	13.93	36.12	3.1391	10.36
IL(cm)	5.85–13.83	9.42 ± 0.66	0.86	1.28	2.14	9.84	11.99	15.51	59.75	1.7992	19.09
SD (cm)	1.93–3.42	2.65 ± 0.14	0.04	0.04	0.08	7.53	7.87	10.89	52.17	0.3096	11.70
SCW(kg)	0.95–2.86	1.67 ± 0.22	0.09	0.06	0.16	18.39	15.05	23.77	40.09	0.3266	19.63
CY(t/ha)	1.11–29.86	14.8840 ± 3.12	19.50	9.41	28.92	29.67	20.61	36.13	32.55	3.61	24.23
Brix (%)	16.19–22.30	19.68 ± 0.89	1.59	0.58	2.17	6.41	3.87	7.48	26.77	0.8125	4.13
Pol (%)	14.98–21.33	18.33 ± 0.97	1.88	0.87	2.75	7.48	5.08	9.05	31.58	1.0788	5.89
Purity (%)	87.61–97.70	93.03 ± 1.52	4.63	1.49	6.12	2.31	1.31	2.66	24.30	1.2384	1.33
RS (%)	9.88–15.56	13.13 ± 0.79	1.24	0.61	1.85	8.48	5.94	10.35	32.89	0.9210	7.02
SY(t/ha)	2.39–56.21	29.2446 ± 6.38	81.30	33.29	114.59	30.83	19.73	36.60	29.05	6.41	21.91

Notes: σ2 g = genotypic variance; σ2e = environmental variance; σ2p = phenotypic variance; ECV (%) = environmental coefficient of variance; GCV (%) = genotypic coefficient of variance; PCV (%) = phenotypic coefficient of variance; H² (%) = broad sense heritability, GA = genetic advance, GAM (%) = genetic advance as percentage of the mean. SPN30 and SPN45 = number of sprouted buds per plot 30 and 45 days after planting, respectively; TL3MAP, TL4MAP and TL5MAP = number of tillers per plot three, four and five months after planting, respectively; SC10MAP = number of stalks per plot 10 months after planting; PH10MAP = plant height ten months after planting; SL18MAP = stalk length 18 months after planting; NI = number of internodes; IL = internode length; SD = stalk diameter; SCW = single cane weight; CY = cane yield; RS = recoverable sucrose; and SY = sugar yield.

With a range of 4.74–48.59, and a difference of 43.85, the average number of sprouted buds per plot 30 days after planting was 20.86. The sugarcane genotypes with the lowest and highest mean values for the number of sprouted buds per plot were B527–30 and B630–1, respectively, and both were germinated from introduced fuzz from Barbados. This indicates the presence of substantial variation among the tested genotypes for the number of sprouted buds. In contrast to the current result, Esayas et al. (2018) evaluated 400 sugarcane genotypes and found a mean value of 14.25 sprouted buds per plot 30 days after planting, with a range of 1 to 41 sprouted buds per plot. The genetic variation in sugarcane, or the genotypes, may be the cause of the difference.

Similarly, the mean for the number of sprouted buds per plot 45 days after planting was 24.26, with a range of 3.60 for genotype B516–25 to 49.12 for genotype B563–10, with a difference of 45.52 sprouted buds per plot. This result confirmed the existence of a large variation in this trait among the evaluated sugarcane genotypes. Generally, the highest and lowest values for the numbers of sprouted buds per plot both at 30 and 45 days after planting were observed for the genotypes from fuzz clones originating in Barbados. This indicates the presence of marked variability among the sugarcane genotypes for these traits. This implies that sugarcane genotypes from fuzz had a wide range of genetic variability for sprouting, and thus, sugarcane varieties with improved sprouting potential can be developed by selecting the best genotypes of this trait among the tested fuzz materials.

Three months after planting, there were 59.49 tillers on average per plot, ranging from 8.44 for genotype B516–66 to 170.42 for genotype B688–11, with a 162.0 difference between the two values. Similarly, four months after planting, there were 70.46 tillers on average per plot, ranging from 12.0 for genotype B516–66 to 165.0 for genotype B546–10, with a 153.0 difference between the two values. Five months after planting, the number of tillers per plot ranged from 26.94 for genotype B658–10 to 162.34 for genotype B516–45, with a mean value of 73.99 tillers per plot.

The minimum and maximum number of tillers per plot three, four, and five months after planting were observed for the sugarcane genotypes from F1 crosses or fuzz introduced from Barbados. This indicates that enormous genetic variability exists among the sugarcane genotypes for this trait. This suggests that improved sugarcane varieties with improved tillering potential can be developed via the selection of individual genotypes that produced the best number of tillers.T. G. Esayas et al. (2018) also reported the existence of variability in the number of tillers in sugarcane, which agrees with the current study result.

With a range of 28.13 for genotype B676–1 to 115.77 for genotype B685–1, representing a difference of 87.64 stalks per plot, the mean number of stalks per plot was 64.96. This indicates that the tested sugarcane genotypes exhibited substantial variation in the number of stalks per plot. This suggests that improved sugarcane varieties with a better stalk population could be developed through the selection of the best genotypes of this trait.T. G. Esayas et al. (2018) andP. Kumar et al. (2018) found the existence of substantial variability in the number of stalks per plot in sugarcane, which agrees with the current study result.

Plant height, stalk length, and internode length had mean values of 1.91 m, 2.82 m, and 9.42 cm, respectively. The observed plant height values ranged from 1.10 m for genotype B517–31 to 2.63 m for genotype B644–12. The stalk length values ranged from 1.76 m for genotype Local-435 to 3.99 m for genotype B528–30. In addition, the internode length values ranged from 5.85 cm for genotype B564–11 to 13.83 cm for genotype B517–30. The lowest and highest mean values for plant height and internode length were recorded in genotypes from fuzz, which were introduced from Barbados. The genotype with the shortest stalk length was from a local landrace from Ethiopia, while the longest genotype was from the fuzz introduced from Barbados.

This study revealed the existence of plant height and internode length variability within the sugarcane genotypes of fuzz introduced from Barbados and stalk length variability between the genotypes of fuzz obtained from Barbados and local landraces of Ethiopian origin. However, all the genotypes with the highest mean values for plant height, internode length, and stalk length were from the fuzz genotypes obtained from Barbados. This indicates the merits of the fuzz genotypes for improving these traits in sugarcane. Similar results were reported by different researchers.Gowda et al. (2016) observed stalk length variability, whileT. G. Esayas et al. (2018) found internode length variability in sugarcane.Gowda et al. (2016) and Abdul et al. (2017) also observed a high level of variability for plant height in sugarcane, which agrees with the current study result.

The number of internodes per stalk ranged from 21.79 for genotype B528–10 to 39.54 for genotype B563–10, with a mean value of 30.29 internodes per stalk. The fuzz genotypes imported from Barbados had the least and most internode numbers per stalk. Single cane weight values varied from 0.95 kg for genotype B528–31 to 2.86 kg for genotype Kay Shenkora, with a mean value of 1.67 kg. The lighter genotype came from Barbados-imported fuzz, and the heavier genotype came from Ethiopian landraces. This shows that the tested genotypes have a substantial amount of variation for these traits, suggesting the possibility of improving these traits by selecting the best genotypes of these traits from the evaluated genotypes of sugarcane.

The stalk diameter values ranged from 1.93 cm for genotype B517–30 to 3.42 cm for genotype B549–10, with a mean value of 2.65 cm. The fuzz introduced from Barbados produced both the thinnest and thickest girth genotypes.

Abdul et al. (2017) classified sugarcane stalk diameters as thin (2.0 cm), medium thin (2–2.5 cm), medium (2.5–3.00 cm), medium thick (3.0–3.5 cm), and thick (>3.5 cm). Based on this benchmark, the sugarcane genotypes evaluated had stalk thickness values ranging from thin to medium thick (1.93–3.42 cm). This shows that there is a substantial amount of variation in this trait across these materials and that thicker stalked varieties might be created by selecting the best genotypes of this trait.

The diameter of the sugarcane stalk is a varietal characteristic that is heritable and stable in different environments (Abdul et al., 2017). In line with this finding,T. G. Esayas et al. (2018) found that sugarcane stalk diameter varied widely, from 1.70 to 3.34 cm. In contrast to the findings of the present study, Abdulet al. (2017) identified a narrower range of stalk diameter variability in sugarcane, varying from 1.95 to 2.70 cm.

The cane yield values observed ranged from 1.11 t/ha to 29.86 t/ha, with a mean of 14.88 t/ha. The lowest cane yield was observed in genotype B552–10, while the highest was in genotype B658–22. Both genotypes with the lowest and highest cane yield values were from the fuzz imported from Barbados. This indicates the existence of variability in cane yield within the genotypes from imported fuzz. This suggests that sugarcane varieties with improved cane yield could be developed by selecting the best genotypes of this trait among the evaluated materials. The presence of variability in cane yield in sugarcane was reported (Gowda et al., 2016;P. Kumar et al., 2018;T. G. Esayas et al., 2018), which agrees with the current study result.

The brix values measured ranged from 16.20% for genotype B549–20 to 22.30% for genotype B564–10, with a mean value of 19.68%. Similarly, the pol values ranged from 14.30% for genotype B549–20 to 21.33% for genotype B564–10, with a mean value of 18.33%. In addition, the purity percent values ranged from 87.61% to 97.69% with a mean value of 93.03%. The lowest and highest values were for genotypes B516–36 and B694–12, respectively. Furthermore, the recoverable sucrose percent values ranged from 9.88% for genotype B549–20 to 15.55% for genotype B564–10, with a mean value of 13.13%.

The lowest and highest mean values for all the biochemical traits: brix, pol, purity, and estimable recoverable sucrose percent were observed within the genotypes from fuzz introduced from Barbados. However, the difference between the lowest and highest values in all the biochemical traits was narrow, indicating the presence of slight variation for these traits within the genotypes evaluated. This implies the need to improve the base population.P. Kumar et al. (2018) reported a limited range of biochemical traits variability in sugarcane, which corroborates with the current study results. However, in their previous studies in sugarcane,T. G. Esayas et al. (2018) reported a relatively wider range of biochemical traits variability, which is in contrary to this study result.

The estimated sugar yield values ranged from 2.39 (t/ha) for genotype B552–10 to 56.21(t/ha) for genotype B688–2, with a mean of 29.25 (t/ha). The genotypes with the lowest and highest sugar yield values were both from the fuzz imported from Barbados, suggesting the existence of a large variability among the Barbados introduced fuzz genotypes for sugar yield. Thus, higher sugar yield genotypes can be developed by selecting the best fuzz genotypes for this trait. In agreement with this result, a wide range of variability in sugar yield was reported (Gowda et al., 2016;P. Kumar et al., 2018;T. G. Esayas et al., 2018).

Among the evaluated sugarcane genotypes, there was little variation in biochemical traits despite the wide range of variation in agro-morphological traits. With the exception of stalk length and single cane weight, which were found between the genotypes from fuzz imported from Barbados and local landrace of Ethiopian origin, the minimum and maximum values observed for all the traits were found within the genotypes from fuzz. Information on genotypic coefficients of variation (GCV) and phenotypic coefficients of variation (PCV) is essential in determining trait responses or suitability for phenotypic selection, in addition to genetic variability. Consequently, the observed variability should be dissected into its constituent parts.

3.2. Phenotypic and genotypic coefficient of variability

Estimates for coefficients of variability, heritability, and genetic advance in absolute units and as percent of the mean are presented in Table 3. The results of the variability analysis revealed that the observed phenotypic coefficients of variability (PCV) values were generally greater than their corresponding genotypic coefficients of variability (GCV) values for all the considered traits. This indicates that every trait is under the influence of environmental or non-genetic factors. The GCV and PCV values ranged from 1.31% to 35.40%, and 2.66% to 43.70%, respectively. The lowest values for both GCV and PCV were observed for purity percent, while the highest values were for the number of sprouted buds per plot 30 days after planting.

According toShivasubramanian and Menon (1973); andDeshmukh et al. (1986), GCV and PCV values ranging from 0 to 10%, 10 to 20%, and greater than 20% are considered low, intermediate, and high, respectively. The number of sprouted buds, the number of tillers, and cane yield all had high GCV and PCV values based on this classification. This indicates that the genotypes tested have high levels of genetic variability for these traits. This implies that sugarcane varieties with an improved number of sprouts, the number of tillers, and cane yield potential could be developed via selection. High GCV and PCV values were observed for the number of sprouted buds (Chaudhary, 2001), number of tillers (Diribu et al., 2020; Esayas et al., 2016a), and cane yield (Esayas et al., 2016b; Feyissa et al., 2014;Anbanandan & Eswaran, 2018;Diribu et al., 2020).

Sugar yield (GCV: 19.73%; PCV: 36.60%), single cane weight (GCV: 15.05%; PCV: 23.77%), and the number of stalks (GCV: 19.64%; PCV: 30.60%) all had high PCV and moderate GCV values. Moderate GCV and high PCV values were observed for stalk number (Feyissa et al., 2014), single cane weight (Esayas et al., 2016), and sugar yield (Gowda et al., 2016), which agrees with the current study results. Contrasting results for stalk number and single cane weight (Anbanandan & Eswaran, 2018;Chaudhary, 2001;Mali & Patel, 2013); and sugar yield (Feyissa et al., 2014;Behou & Pene, 2019) were reported. This disparity could be attributed to genotype differences and the environmental conditions under which the genotypes were evaluated.

Low GCV and moderate PCV values were observed for recoverable sucrose percent (GCV: 5.94; PCV: 10.35), number of internode per stalk (GCV: 8.37; PCV: 13.93), and stalk diameter (GCV: 7.87; PCV: 10.89). Moderate GCV values coupled with high PCV values; and low GCV values with moderate PCV values indicate that these traits are highly influenced by environmental or non-genetic factors. This suggests that improvement of these traits through direct selection is difficult. The current result for the number of internode per stalk agrees with the previous findings ofT. Esayas et al. (2016b). Nevertheless, Feyissa et al. (2014);Chaudhary (2001); andBelwal and Ahmad (2020) found contrasting results for recoverable sucrose percent and stalk diameter.

The GCV and PCV values for plant height (GCV: 11.64, PCV: 18.56), stalk length (GCV: 10.38, PCV: 13.46), and internode length (GCV: 11.99, PCV: 15.51) were all moderate. This suggests that acceptable improvements to these traits could be possible via selection of the best genotypes for these traits.

This study found that brix percent had low GCV and PCV values (GCV: 3.87; PCV: 7.48), as did pol percent (GCV: 5.08; PCV: 9.05) and purity percent (GCV: 1.31; PCV: 2.66). This indicates that the tested sugarcane genotypes show little variability in these biochemical traits, suggesting the need to improve the base population to create variability in these traits. Low GCV and PCV values for brix percent, pol percent, and purity percent were reported (Esayas et al., 2016;Belwal & Ahmad, 2020;Gowda et al., 2016;Marwa et al., 2018), which agrees with the current study results.

Low GCV and PCV values were reported for brix percent (Alem et al. 2017;Mali & Patel, 2013); pol percent (Mali & Patel, 2013;P. Kumar et al., 2018; Japheth et al. 2019;); and purity percent (Behou & Pene, 2019;Mali & Patel, 2013) which agrees with the current study results. Furthermore, there was not much of a difference in GCV and PCV values for internode length, internode number, stalk length, plant height, and stalk diameter. This slight difference between the GCV values and their corresponding PCV values for these traits indicates that the environment or non-genetic factors have a relatively minimal impact on the phenotypic expression of these traits.

Allard (1960) andPoehlman and Sleper (1995) described that selection is more effective when the GCV value is greater than its corresponding ECV value. The GCV values for the number of sprouted buds per plot 30 days after planting, the number of tillers per plot three and four months after planting, cane yield (t/ha), stalk length (m), internode length (cm), and stalk diameter (cm) were all greater than their corresponding ECV values. This implies that genetic factors have a greater influence on the phenotypic expression of these traits than the non-genetic factors.

Wright (1921) proposed that GCV values in conjunction with heritability estimates provide a reliable prediction of the amount of genetic advance to be expected through phenotypic selection. Heritability is a good indicator for the transmission of characters from parents to their progeny and is helpful in predicting the expected progress to be achieved through the process of selection.Dabholkar (1992) classified heritability estimates as high (>30%), medium (10–30%), and low (5–10%). Based on this benchmark, the traits taken into consideration in this study were classified as having moderate-to-high heritability values. The observed heritability values ranged from 24.30% to 65.64%. The number of buds that had sprouted per plot 30 days after planting had the highest rating, while percent purity had the lowest heritability value.

High heritability values were found for all of the variables examined, with the exception of sugar yield, brix percent, and purity percent, which had moderate heritability values (Table 3).

A high heritability value means that non-genetic or environmental factors have a limited impact on the traits’ expression and thus, respond well to selection. However, the moderate heritability value demonstrates that the phenotypic manifestation of the traits is regulated by non-additive genetic influences and thus, under the heavy influence of environmental or non-genetic factors. This suggests that it would be difficult to improve them via direct selection. High heritability values were reported for the number of sprouted buds, number of tillers, stalk numbers, single cane weight, and stalk diameter (Belwal & Ahmad, 2020;Gowda et al., 2016;P. Kumar et al., 2018), which is consistent with the current study results. In line with the current study results,Gowda et al. (2016) observed high heritability values for stalk length, internode length, number of internodes, and cane yield. In contrast,Marwa et al. (2018) discovered high heritability values for sugar yield, brix percent, and purity percent.

The examined agro-morphological traits all had high heritability values, as did a few biochemical traits recoverable sucrose percent and pol percent. High heritability values, however, are rarely linked to important genetic advancements. Therefore, it is usually advisable to take into account both heritability estimates and genetic advance (Johanson et al., 1955). Genetic advance is an excellent predictor of the effective and efficient selection progress that may be anticipated when phenotypic selection is performed on the base population.

In one cycle of selection at specific selection intensity, genetic advance in absolute unit (GA) refers to the enhancement of traits in genotypic value for the new population as compared to the base population. Reliable estimates of the genetic advance should consider only traits with moderate-to-high variability and high heritability values. As a result, it is anticipated that the mean increment for the number of sprouted buds per plot 30 and 45 days after planting will be 29.40–36.96 and 26.03–35.29, respectively.

Similar to this, the expected mean genetic advance for the number of tillers per plot three, four, and five months after planting will be 79.20–104, 90.70–116.56, and 87.07–115.29, respectively. In addition, it is possible to anticipate a 15.37–21.61 rise in mean cane yield (t/ha). The plant height, stalk length, and internode length might all be enhanced to 2.0–2.40 m, 3.12–3.34 m, and 10.56–11.88 cm, respectively, from their current levels.

Johanson et al. (1955) classified the expected genetic advance as a percentage of the mean (GAM) as low (0–10%), medium (10–20%), and high (>20%). Thus, the number of sprouted buds, the number of tillers, stalk number, cane yield, and sugar yield all had high GAM values (Table 3). High GAM values for the number of sprouted buds, number of tillers, number of stalks, and cane yield were reported (Anbanandan & Eswaran, 2018; Esayas et al., 2016;Gowda et al., 2016).

A high GAM value for sugar yield was reported (Anbanandan & Eswaran, 2018; Esayas et al., 2016;Gowda et al., 2016; Behou and Pene, 2019), which supports this result. However, GAM values were moderate for stalk diameter, internode length, stalk length, and single cane weight. Moderate GAM values for internode length, stalk diameter, and plant height were reported (Esayas et al. 2016;Belwal & Ahmad, 2020;Gowda et al., 2016;P. Kumar et al., 2018), which is in agreement with the current study results.

Similarly to the current study results,Gowda et al. (2016) also reported moderate GAM values for internode length and stalk length, whileMali and Patel (2013),P. Kumar et al. (2018), andGowda et al. (2016) found moderate GAM values for single cane weight in their previous studies in sugarcane.

However, low GAM values were reported for the number of internodes per stalk (10.36%), recoverable sucrose (7.02%), pol (5.89%), brix (4.13%), and purity (1.33%). Similar results, low GAM values for brix percent, pol percent, purity percent, and recoverable sucrose percent (Belwal & Ahmad, 2020; Japheth et al. 2019; P. Kumar et al., 2018); brix percent and pol percent (Mali & Patel, 2013); and purity percent (Behou and Pene, 2018;Gowda et al., 2016) were reported.

According to (1955), high heritability values should be combined with high estimates of GCV and GAM values in order to obtain a high forecast of the expected genetic advance and to guarantee successful selection to improve the trait of interest.

The current study results revealed that the number of sprouted buds per plot 30 and 45 days after planting, the number of tillers per plot three, four, and five months after planting, and cane yield (t/ha) showed high heritability (H²) values coupled with high GCV and GAM values. The high GCV, H², and GAM values in the number of sprouted buds, number of tillers, and cane yield indicate that additive gene actions regulate the phenotypic expression of these traits. This implies that by using simple selection techniques, reliable selection might be carried out based on the phenotypic expression of these traits in the individual plants. Similar findings were also reported for cane yield (Esayas et al., 2016;Anbanandan & Eswaran, 2018) and the number of sprouted buds 45 days after planting (Chaudhary, 2001).

A high trait heritability value in conjunction with a modest genetic advance expressed as a percentage of the mean (GAM) can provide room for improving traits through selection (V. K. Singh et al., 2016). The number of stalks per plot, plant height (m), stalks length (m), internode length per stalk (cm), stalk diameter (cm), and single cane weight (kg) exhibited high heritability and moderate GAM values. High heritability and moderate GAM values for the number of stalks were reported (Mali & Patel, 2013;P. Kumar et al., 2018;Belwal & Ahmad, 2020; Esayas et al., 2016; andMali & Patel, 2013), which agrees with the current study results. Similar to the current study result, high heritability and moderate GAM values for stalk diameter were also reported (Belwal & Ahmad, 2020; Esayas et al., 2016;Gowda et al., 2016; Kumar et al., 2019).

Similar results with high heritabilities and moderate GAM values for internode length and stalk length were found (Gowda et al., 2016;Mali & Patel, 2013;P. Kumar et al., 2018). However, brix percent and purity percent showed moderate heritability and GAM values. This suggests that these traits’ phenotypic expression is mostly controlled by non-additive gene action and that direct selection cannot be used to improve these traits because the majority of variation is attributed to non-genetic or environmental factors. The environmental or non-genetic factors influencing a trait’s expression and selection could be heterogeneity in soil fertility, epigenetic factors, and other unpredictable factors. Thus, improvement of such traits requires improved management practices rather than selection.

In general, the sugarcane genotypes studied showed highly heritable variation with high expected genetic advance for the number of sprouted buds, number of tillers, and cane yield. Thus, the top 5% of the sugarcane genotypes with the greatest number of sprouted buds per plot were B563–10, B658–22, B688–2, B517–20, B528–11, B668–1, B658–10, Local-434, B516–45, and B644–9. Similarly, the sugarcane genotypes with a greater number of tillers per plot were B516–45, B558–1, B566–10, B694–2, B572–1, B694–11, B546–20, Local-189, 93-burabure shenkora, and B546–10, while genotypes B658–22, B688–2, local-434, Kay Shenkora, B528–30, B668–1,B645–1,B644–12,B516–45, and B635–2 exhibited better cane yield (t/ha) than the rest of the sugarcane genotypes evaluated. Therefore, the sugarcane genotypes selected for the greater number of sprouted buds, number of tillers, and cane yield could be used as potential parents for crossing to develop improved sugarcane varieties for these traits.

4. Conclusion and recommendation

This study has discovered that the genotypes of sugarcane that were examined exhibit a wide range of genetic variability. The observed genotypic and phenotypic coefficients of variability, broad sense heritability, and genetic advance as a percentage of the mean were 1.30–35.40%, 2.6–55.36%, 24.30–65.64%, and 1.33–39.09%, respectively. High genotypic coefficients of variation should be combined with high heritability and genetic progress as a percentage of the mean to ensure successful selection to improve the trait of interest. Thus, only the number of sprouted buds, tillers, and cane yield demonstrated high genotypic coefficients of variability along with high heritability and genetic advancement. This indicates that non-genetic factors have a limited impact on the traits’ expression and thus respond well to selection. Thus, the genotypes B563–10, B658–22, B688–2, B517–20, B528–11, B668–1, B658–10, Local-434, B516–45, and B644–9 exhibited a higher number of sprouted buds per plot. Similarly, the sugarcane genotypes with a greater number of tillers were B516–45, B558–1, B566–10, B694–2, B572–1, B694–11, B546–20, Local-189, 93-burabure Shenkora, and B546–10, while genotypes B658–22, B688–2, local-434, Kay Shenkora, B528–30, B668–1, B645–1, B644–12, B516–45, and B635–2 exhibited better cane yield per hectare. Thus, the sugarcane genotypes with a greater number of sprouted buds, tillers, and cane yield could be used as potential parents for crossing to develop improved varieties for these traits. In addition, the sugarcane genotypes with high cane yield could be tested over seasons for commercial use at the Metahara sugar estate.

Acknowledgments

We are grateful to the NORAD project, Hawassa University for its financial support. We are also greatly thankful for Metahara Research Centre and Metahara sugar factory for their valuable technical and material support, experimental field management and data collection.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/23311932.2023.2194482

Additional information

Funding

This research work was supported by the Institutional Collaboration Program between Hawassa University and the Norwegian University of Life sciences (NMBU) financed by the government of Norway.

Notes on contributors

Belay Tolera

Belay Tolera, Gemeda is a senior researcher in plant biotechnology at the Ethiopian Sugar Industry Group, Wonji Research Center, Ethiopia. He had done research on the agro-morphological and biochemical characterization of sugarcane and sugarcane tissue culture. His interests are in research related to genetic diversity, GWS, GS, creation of genetic variability, and development of multi-trait selection indices in sugarcane.

Andargachew Gedebo

Dr. Andargachew Gedebo, Abitea, has BSc in Crop Science and MSc in Crop Physiology from Reading University, UK; and a PhD, in Agriculture from the Norwegian University of Life Sciences, Norway. He is an Associate Professor of Plant Breeding and Agronomy at Hawassa University, Ethiopia, with more than 20 years of experience in teaching and research and published more than 25 research papers.

Esayas Tena

Dr. Esayas Tena, Gashaw, is a senior plant breeder and research program coordinator at the Ethiopian Sugar Industry Group, Wonji Research Center, Ethiopia. His team is working on the evaluation and maintenance of genetic resources and provision of adaptable, high-performing sugarcane varieties to the various sugarcane plantations of Ethiopia.