Selection in Sugarcane Germplasm under the Egyptian Conditions

Stage I

Each genotype of the evaluated germplasm (Table 1) was planted during the 1st week of March, 2014 in two rows of 5-m in length and 1-m width in a randomized complete block design with three replications. After 12 months from planting, the following data were recorded: 1. Millable cane height (cm) was measured from soil surface up to the top visible dewlap.

2. Millable cane diameter (cm) was measured at the middle part of stalks.

3. Brix% (total soluble solids percent) was determined using “Brix Hydrometer” according to A.O.A.C. [6].

4. Sugar content % was determined using the following equation of: Sugar content % = brix% x 0.69 [7].

5. Cane yield/fad (ton). (Fad = 4200 m2).

6. Theoretical sugar yield was assessed as follows: Sugar

yield = cane yield (ton) x sugar content %.

Stage II

A total number of 14 sugarcane genotypes from stage I were selected as the best ones based on sugar yield at 5% selection intensity, to be evaluated in stage II as a plant cane and its 1st ratoon crops. They were replanted during the 1st week of March, 2015 in three rows of 5-m in length and 1-m in a randomized complete block design with two replications. Two standard cultivars (the commercial variety GT 54-9 and new variety G2003-47) were used as checks. Planting was achieved by placing twenty five 3- budded cane cuttings in each row. The field was irrigated right after planting and all other agronomic practices were carried out as recommended by Sugar Crops Research Institute. Harvest took place at age of 12 months in both plant and 1st ratoon cane crops. In addition to the traits recorded at stage I, the following traits were determined for each genotype in stage II: 1. Sucrose percentage of clarified juice was determined using automated Sacharimeter according to A.O.A.C. [6].

2. Juice purity percentage was calculated according to the following equation of: Juice purity % = (sucrose %/ brix %) x 100 [8].

3. Sugar recovery % was calculated according to the following equation, described by: Sugar recovery % = [sucrose % - 0.4 (brix % - sucrose %)] x 0.73 [9].

4. Cane yield/fad (ton) was determined from cane weight of each plot (kg), which was converted into ton per fad.

5. Sugar yield/fad (ton) was calculated according to the following equation as described by Mathur: Sugar yield/fad (ton) = cane yield/fad (ton) x sugar recovery% [10].

Statistical Analysis

The collected data were statistically analyzed according to the procedures outlined by Snedecor and Chochran to estimate the following criteria [11]: a. Estimation of Variance Components Genotypic and phenotypic components of variance were estimated with the help of following formulae: Genotypic Variance (σ2 g) = gMS – eMS Phenotypic variance (σ2 p) = σ2 g + eMS b. Coefficient of Variability Both genotypic and phenotypic coefficient of variability were computed for each character as per method suggested by Burton and Devane (1953) Genotypic Coefficient of Variation (GCV) % was estimated as: (GCV) % = (δg / general mean) x 100. Phenotypic Coefficient of Variation (PCV) % was estimated as: (PCV) % = (δp / general mean) x 100.

Where: δg = genotypic standard deviation δp = phenotypic standard deviation Heritability (H) in broad sense was computed for each character as the ratio of genetic variance to the total variance as suggested by Hanson, et al_._ (1956) H = σ2 g / σ2 p x 100 Where, σ2g and σ2p are genotypic and phenotypic variances respectively. Expected genetic gains from selection as % of mean (GA %) were calculated following the formula adopted by Falconer: GA = K× H x δp GA as % of mean = (GA / general mean) x 100 Where: K= 2.06 for 5% selection intensity, δp = phenotypic standard deviation and H = broad sense heritability. To compare between treatment means, LSD at 5% level of significance was used according to Steel and Torrie [12]. All statistical analyses were performed using analysis of variance technique of (MSTAT-c) computer software package and Genes computer software package.

The First Stage

The results revealed higher range of variance among the evaluated 294 genotypes in the studied characters, i.e. stalk length, diameter, brix%, sugar content%, cane and sugar yields/fad (Figures 1-6). A wide variation was obtained in stalk length of genotypes at harvest, which ranged from 58 to 259 cm. Likewise, values of 1.22 to 3.62 cm were observed for stalk diameter of the tested genotypes. Juice brix showed a great variation ranged from 13 to 24%, while the recorded values of sucrose% ranged from 8.97 to 16.56%. Moreover, an extreme variance of 7.74-87.26 and 0.98-13.25 ton/fed was detected in cane and sugar yield, successively.

Click to enlarge

Click to enlarge

Click to enlarge

Table 2: Variance components, coefficients of variation, broad sense heritability and genetic advance for studied traits of 294 sugarcane germplasm in stage one. Genotypic and Phenotypic Variance After partitioning phenotypic variance, it was found that genotypic variance was higher than the environmental one for the six characters studied (Table 2). The magnitude of variance was the highest in stalk length (σ2g = 1114.47, σ2e = 1.96) followed by cane yield/fed (σ2g = 191.41, σ2e = 1.22). These results indicate that a negligible role was played by the environmental factors in the inheritance of these characters in sugarcane. The high genotypic variance in stalk height, diameter and brix was also reported by Singh and Singh [8].

Genotypic and Phenotypic Coefficients of Variation The estimates for phenotypic coefficient of variation (PCV) were closed with genotypic coefficient of variation (GCV) in all the traits, indicating lower influence of environment on genetic variation (Table 2). The highest phenotypic and genotypic coefficient of variation were observed for sugar yield (PCV = 33.24% and GCV = 32.97%) followed by cane yield/fed (PCV = 29.75% and GCV = 29.66%) and stalk length (PCV = 20.49% and GCV = 20.47%). High genotypic and phenotypic coefficients of variation for stalk height, stalk diameter, brix % and cane yield were reported by Singh RK, Singh GP [13]. Heritability Genotypic coefficient of variation is not a correct determine the heritable variation present and should be considered together with heritability estimates. Table, 2 illustrated that high heritability estimates were recorded for all characters, stalk length (99.82%), yield/fed (99.37%), sugar yield (98.43%), stalk diameter (98.16%), brix and sugar content (87.19%). These findings suggest that simple selection for these traits would be more effective. High heritability estimate for stalk diameter was reported by Chaudhary RR [14]. Similarly, Jamoza, et al. also found high heritability for stalk diameter and moderate heritability for cane yield [15].

Genetic Advance Heritability estimates along with expected genetic gain is more useful than the heritability value alone in predicting the resultant effect for selecting the best genotypes [16]. Genetic gain, expressed as percent of mean, came in the following descending rank: sugar yield (67.39%), cane yield/fed (60.90%), stalk length (42.13%), stalk diameter (33.76%), brix (20.47%) and sugar content (20.47%), indicating that there is a scope to improve cane yield to a considerable extent by adopting suitable breeding procedures (Table 2). High genetic advance in cane yield was also reported by Jamoza JE, Owuoche J, Kiplagat O, Opile W [15]. Total soluble solids and sugar content had high heritability with moderate genetic advance. Gravois and Milligan illustrated that moderate heritability with low genetic advance in sugar quality parameters indicate the presence of non-additive gene action [17]. Therefore, simple selection on phenotypic performance may not be effective. Gulzar et al. reported that moderate to high heritability coupled with high genetic advance were recorded for juice/cane, juice extraction % and sucrose per cent [18]. Tena, et al. found that moderate broad sense heritability estimates were found in sugar % and brix %, this suggests that a considerable proportion of the total variance is heritable and selection of these traits would be effective [19]. The results suggest that selection should be practiced on the basis of sugar yield, yield/fed, stalk Length and diameter for higher cane yield. Improvement in these traits would lead to a significant increase in yield in limited selection cycles.

The second Stage

Data in Tables 3-5 cleared that the 14 selected genotypes from the first stage differed significantly (P = 0.05) in all studied traits; stalk diameter, stalk length, stalk weight, brix%, sucrose%, juice purity%, sugar recovery%, cane and sugar yields, in plant cane and first ratoon. Stalk Diameter: Results in table 3 showed that the selected genotypes varied significantly in stalk diameter, in the plant cane and first ratoon, and their mean. The two genotypes namely SP72-5181 and G2006-36 recorded significantly higher than that of the two check varieties; GT54-9 and G2003-47 in the plant cane, first ratoon and their mean. Stalk Length: Data in Table 3 revealed that the tested genotypes differed markedly in stalk length, in the plant cane and first ratoon. Stalk length varied from 225 cm for the genotype M57-35 in the plant cane and 231.5 cm for the commercial variety GT54-9 in first ratoon to (352.5 and 358.5 cm) for the genotype EI58-37 in the plant cane and first ratoon, respectively. The five genotypes; F150, EI58-37, F161, G2008-64 and G2003-49 gave significantly taller stalks as compared to two check varieties GT54-9 and G2003-47, in the plant cane and first ratoon. These results are in agreement with those mentioned by Yousif, et al. Mehareb, et al, Mehareb and Abazied, Ahmed, who found that the tested varieties significantly differed in stalk height and diameter [20, 21, 22, 23]. Stalk Weight: Results in Table 3 indicated that the evaluated genotypes differed substantially in stalk weight, in the plant cane and first ratoon, and the mean of the two cane crops. In the plant cane, the genotype F161 recorded the highest mean value of stalk weight. Three genotypes namely F161, G2008-64 and G2006-36 had significantly higher than that recorded by the two check varieties; GT54-9 and G2003-47. In first ratoon, most of genotypes were significantly higher than the two check varieties in this growth trait. Moreover, F150 recorded the highest mean value of stalk weight in the first ratoon and overall mean of the two seasons. These results are in agreement with those reported by El-Shafai and Ismail, Yousif, et al., Mehareb, et al., Mehareb and Abazied, Mehareb et al., who found that the studied varieties significantly differed in stalk weight [20, 21, 22, 24, 25].

PC (plant cane), FR (first ratoon). Table 3: Mean performance of stalk diameter (cm), stalk length (cm) and stalk weight (kg) of the fourteen promising sugarcane genotypes at harvest in the plant and first ratoon cane crops. Brix %: Data in Table 4 revealed that genotype viz. EI58- 37 recorded the highest brix % in the plant cane, first ratoon and overall of both crops. Six genotypes namely EI58-37, SP 72-5181, G2008-64, G2003-49, L62-96, G2008-20 and G2006-36 had significantly higher values of this trait than that recorded by the two check varieties; GT54-9 and G2003-47 in the plant cane, without significant variance among them, in the first ratoon. These results are in harmony with those of Mehareb, et al., Mehareb, et al., Mehareb and Abazied, who found differences in brix % among the tested sugar cane varieties [21, 22, 25]. Sucrose %: Data in Table 4 reveal that sucrose% varied significantly among studied genotypes in the plant cane, first ratoon and overall. Two genotypes; EI 58-37 (19.20% and 20.20%) and G2003-49 (18.70% and 20%) were stable and gave significantly higher sucrose % than the two check varieties; GT54-9 (16.60% and 17%) and G2003-47 (17.30% and 18.50%). In the plant cane and the first ratoon, respectively. Purity %: Purity percentage presented in Table 4, the genotype SP81-3250 recorded the highest significantly mean purity % (89.75%) in plant cane but the genotype F 150 recorded the highest significantly mean purity % (88%) in the first ratoon and the genotype G2006-36 recorded the highest significantly mean purity % (85.03%) in the overall.

PC (plant cane), FR (first ratoon). Table 4: Mean performance of Brix %, Sucrose % and Purity %of the fourteen promising sugarcane genotypes at harvest in the plant and first ratoon cane crops. Sugar recovery %: Data presented in Table 5 showed that the evaluated genotypes differed significantly concerning sugar recovery% at the plant cane and first ratoon and over both crops. In the plant cane, the highest sugar recovery% was recorded for the genotype SP81- 3250 (13.55%) followed by the genotypes; G2006-36 (13.30%), EI58-37 (13.10%), G2003-49 (12.65%), and L62-96 (12.25%), respectively. All the previous genotypes gave significantly higher sugar recovery% as compared to the two checks GT54-9 (10.85%) and G2003-49 (11.45%). In the first ratoon, the genotypes; EH88/5-27 (13.20%), F150 (14.05%), EI58-37 (13.45), M57-35 (13.35%), SP72-5181 (14.45%), G2003-49 (13.65%) and G2008-20 (12.45%) significantly surpassed the two checks GT54-9 (11.45%) and G2003-47 (11.80%).

Cane yield: Data of the plant cane, first ratoon and overall of the two crops, in Table 5 revealed that the 14 genotypes selected from the first stage differed significantly in cane yield. In the plant cane, six genotypes; EH88/5-27(19.00 and 6.50), SP72-5181(16.10 and 3.60), G2008-64 (26.50 and 14.00) , G2008-20 (26.20 and 13.70), G2007-61 (24.50 and 12.00)and G2006-36 (16.00 and 3.50) produced significantly tons of canes higher than that obtained by GT54-9 and G2003-47 check cultivar, successively, corresponding to EH88/5-27(17.80 and 5.80), SP72-5181(13.00 and 1.00), G2008-64 (24.60 and 12.60) , G2008-20 (27.75 and 15.75), G2007-61 (29.70 and 17.70)and G2006-36 (15.50 and 3.50) tons of canes, in the 1st ratoon cane crop. The overall mean of the two seasons showed the superiority of seven genotypes viz. EH 88/5-27( 18.40 and 6.15), SP 72-5181( 14.55 and

2.30), G 2008-64(25.55 and 13.30), G 2008-20 (26.98 and 14.73), G 2007-61(27.10 and 14.85), SP 81-3250 and G2006-36(15.75 and 3.50), which out-yielded GT 54-9 and G 2003-47 cultivars by tons of canes, respectively.

Sugar yield: Results of Table 5 revealed that the sugar yield differed significantly among the evaluated genotypes in the plant cane, first ratoon and overall. In the plant cane, four genotypes; G2008-64 (3.50 and 1.75 ton/fad), G2008-20 (3.45 and 1.70 ton/fad), SP81-3250 (3.00 and 1.25 ton/fad) and G2006-36 (3.55 and 1.80 ton/fad) recorded significantly greater sugar yield values than the two checks GT54-9 (6.45 ton/fad) and G2003-47 (8.20 ton/fad). In the first ratoon, four genotypes; EH88/5-27 (3.40 and 1.80 ton/fad), SP72-5181 (1.95 and

0.20 ton/fad), G2008-64 (3.25 and 1.65 ton/fad) and G2008-20 (4.05 and 2.45 ton/fad) recorded significantly greater cane yield values than the two checks GT54-9 (7.05 ton/fad) and G2003-47 (8.65 ton/fad). In the over two seasons, six genotypes; EH88/5-27 (2.03 and 0.35), SP72-5181 (2.83 and 1.15 ton/fad), G2008-64 (3.38 and 1.70), G2008-20(3.75 and 2.08), G2007-61 (1.70 and 0.02), and G2006-36 (3.03 and 1.35) tons/fad recorded greater sugar yield values than the two checks GT54-9 (6.45 ton/fad) and G2003-47 (8.43 ton/fad). This result is in agreement with those reported by El-Shafai, & Ismail, Yousif, et al., Mehareb, et al., Ahmed, et al., Mehareb and Abazied and Mehareb, et al., who found that the tested varieties significantly differed in cane yield and sugar yield [20, 21, 22, 23, 24, 25].

PC (plant cane), FR (first ratoon). Table 5: Mean performance of sugar recovery %, cane yield (t/fad) and sugar yield (t/fad) of the fourteen promising sugarcane genotypes at harvest in the plant and first ratoon cane crops. Genetic Parameter: Among the cane characters, showed wide range of variation in plant cane as well as in the ratoon crop (Tables 6-8), providing wide scope of selection for this traits. Wide range of variations for number of millable canes and stalk height was reported earlier by Ghosh and Singh [26]. Genotypic and Phenotypic Variance: The wide range of genotypic (σ2g) and phenotypic (σ2p) variation was observed for stalk length (σ2g = 739.35 & σ2p = 850.42) in plant cane and (σ2g = 721.53 & σ2p = 836.93) in first ratoon followed by cane yield (σ2g = 68.12 & σ2p = 70.01) in plant cane and (σ2g = 85.25 & σ2p= 86.33) in first ratoon followed by purity (σ2g = 22.82 & σ2p = 23.89) in plant cane and (σ2g = 30.14 & σ2p = 36.06) in first ratoon. The lowest estimates of σ2g and σ2p were exhibited by stalk diameter (σ2g = 0.05 & σ2p = 0.06) in plant cane and (σ2g = 0.04 & σ2p = 0.06) in first ratoon (Tables 6-8). Similar findings were reported by Hapase and Hapase and Verma, et al., found variability of higher magnitude for number of shoots per plot, number of millable canes and cane yield [27, 28]. Also Kumar, et al., and Pawar, et al., found similar results for most of the cane yield and its contributing traits [29, 30]. A perusal of the estimates of environmental component of variance in relation to their genotypic counterpart revealed that the estimates of σ2g were higher than σ2e for all of the characters. The higher magnitude of genotypic variance suggested little influence of environments in the expression of genetic variability.

PC (plant cane), FR (first ratoon). Table 6: Components of variances, coefficients of variation, heritability for stalk length, stalk diameter and stalk weight in fourteen promising sugarcane genotypes for plant crop and first ratoon.

PC (plant cane), FR (first ratoon). Table 7: Components of variances, coefficients of variation, heritability for Brix%, Sucrose% and Purity% in fourteen promising sugarcane genotypes for plant crop and first ratoon.

PC (plant cane), FR (first ratoon). Table 8: Components of variances, coefficients of variation, heritability for Sugar recovery%, Cane yield and Sugar yield in fourteen promising sugarcane genotypes for plant crop and first ratoon Genotypic and phenotypic coefficient of variance. The genotypic coefficient of variability (GCV) was moderate for stalk weight followed by sugar yield, cane yield and sugar recovery in both plant and ratoon crops, whereas, brix was lowest also in both plant and ratoon crops (Tables 6-8). The slightly high phenotypic coefficient of variability (PVC) over GVC throughout the plant and ratoon crops indicated a good scope for selection of these characters. However, the differences between the estimates of GCV and PCV was comparatively narrow for all the characters, suggesting the possibility of affective selection of these traits and indicating high prospects for genetic progress through selection under the conditions of this investigation. The same finding was observed by earlier workers, viz. Nair, et al., Singh, et al., Verma, et al., Gosh and Singh and Bhaskar [26, 31, 32, 33, 34]. The success of a variety improvement program depends largely on the amount of genetic variability present in the population. The GCV values for stalk weight and its components like sugar yield, cane yield and sugar recovery were larger than the values for juice brix, purity and sucrose content. Singh and Singh also reported the large amount of genetic variation for stalk height, diameter and number of millable canes in clonal population and concluded that progress in breeding for higher sucrose yield can be made by emphasizing selection for high sucrose content at early ripening stage along with higher cane yield. Degree of Heritability: It is evident from Tables 6-8 that cane yield, sugar recovery, sugar yield, purity, sucrose percent and stalk length were highly heritable characters in both plant and first ratoon, while stalk diameter and stalk weight were highly only in plant crop and first ratoon, respectively. It might be due to high genetic variance or low environmental variance or both. Thus, high values of heritability suggested the possibility of improvement of these characters through selection Singh and Singh, which may indirectly help in the yield and quality improvement Gravois, et al. [35, 36, 37, 38, 39, 40].