Index

Abstract

Millet is grown mostly as the main grain crop in the drier Western Parts of Sudan (Darfur and Kordofan States) where climatic conditions permit only millet production. Also there is a limited cultivation of millet in the Eastern region of the country. Most species of Pennisetum are protogynous, but pearl millet is more conspicuous in this regard. This facilitates the introgression of characters from other annual penicillaries into pearl millet and hence has helped in the genetic enrichment of this species. In this study 20 pearl millet genotypes were collected from different parts of the Sudan with concentration on the western states. They were evaluated to assess the extent of variation among them in morphological and yield parameters, using a randomized complete block design with two replications. The twenty pearl millet genotypes exhibited significant differences in most of the characters studied. Out of the 19 parameters, only yield per tiller panicle and total yield per plant were non-significantly different.   There is a noticeable variation, not only among the different genotypes of the same species but also within the same genotype or cultivar. Such variation may be attributed to the open pollination system of this crop

Keywords: Pennisetum glaucum, Natural variability, Agronomic parameters.

Received: 25 March 2020 / Revised: 27 April 2020 / Accepted: 2 June 2020/ Published: 14 July 2020

Contribution/ Originality

This study will add basic information to the existing literature concerning Sudanese genotypes.   This study is one of very few studies which have investigated the Sudanese local Pennisetum glaucum genotypes, giving full botanical and agronomic characters. This study can elucidate the use of these genotypes in any future improving programs.

1. INTRODUCTION AND LITERATURE REVIEW

Pearl millet (Pennisetum glaucum (L.) R. Br.) is a diploid species (2n=14) believed to have originated in the Sahel Zone of West Africa, which extends from Western Sudan to Senegal (Gill, 1991). It ranks sixth among cereals in the world in terms of production, following wheat, rice, maize, barely and sorghum (Brunken, 1977). Millet is grown mostly as the main grain crop in the drier Western Parts of Sudan (Darfur and Kordofan States) where climatic conditions permit only millet production. Also there is a limited cultivation of millet in the Eastern region of the country (Salih, 1997; Salih., El-Tigani, & Abdalla, 1999). P. glaucum is the third in importance after sorghum and wheat in the Sudan. The grain is particularly devoid of paste during storage and has long storage and keeping quality (Chandra & Matta, 1990).

Millet grain has little commercial status and marketable surpluses as direct share of agricultural sales are low, millet indirectly contributes to the receipt, from livestock sales of the grain remain largely unexplored (Kumar, 1993).

 Pearl millet is protogynous, this helps the introgression of characters from other annual penicillaries into pearl millet and hence has helped in the genetic enrichment of this species as well as it can explain the non-homogeneity of many genotypes. Pearl millet responds very well to out breeding enhancement (heterosis breeding); Hazza (1994) indicated that the increase in grain yield of the hybrid over its parents could mount up to 74.2% due to out-breeding enhancement effect, thus hybrid seed production is possible (Jauhur, 1981). Concerning African pearl millet landraces, there is a long history of recombination among west-African subpopulations compared to other land races (Serba et al., 2019). Estimation of natural variability generally can provide more efficient way to identify potential accessions and help in the improvement of  different genotypes collections in plant breeding (Kumari et al., 2016).

In Sudan the production of pearl millet was 28.6 million tons and average grain yield was 772.0 kg/ha in 1998 (FAO, 1999). Pearl millet provides staple food to 3.5 million people of Western Sudan. Ninety one percent of the total millet is produced in the western states of the country (Kordofan and Darfur). In these states, however, yield levels are lowest. The annual rainfall in the millet growing areas varies between 250-750mm.

The low yield of pearl millet in the Western parts of the country can be referred to many reasons, the most important are:

After the drought season in 1991, seeds of many genotypes were imported from West Africa without enough precaution which might resulted in the loss of yield due to pests and diseases (Abuelgasim, 1999). Pests rather than diseases cause extensive damage to pearl millet crop.

In addition to the low yield and the factors mentioned, pearl millet faces many problems; these can be summarized as follows:

2. MATERIALS AND METHODS

Twenty accessions of pearl millet (Pennisetum glaucum), provided by Dr. Abdalwahab H. Abdallah, Department of Agronomy, Faculty of Agriculture, University of Khartoum, were used in this study. These accessions were collected from different parts of the Sudan with concentration on the western states. They were evaluated to assess the extent of variation among them Table 1.

Table-1. Twenty accessions of pearl millet used in the study and their origins.

No.
Accession
Origin
Notes
1
Shambat bulk
Bulk Selection
Selected
2
Bauda
Darfur/Sudan
Check
3
*JM 44/9
Darfur/Sudan
Selected
4
Darmasa
Darfur/Sudan
Collection
5
**SP2C2
Selection from SP2
Selected
6
Ugandi
Uganda/adopted
Check
7
*JM24/15
Darfur/Sudan
Selected
8
***RP2C2
Selection from Rawakeeb P2
Selected
9
*JM23/13
Darfur/Sudan
Selected
10
***RP1C2
Selection from RP1
Selected
11
*JM3/16
Darfur/Sudan
Selected
12
Madlkawia
Kordofan/Sudan
Collection
13
*JM21/2
Darfur/Sudan
Selected
14
*JM25/10
Darfur/Sudan
Selected
15
*JM30/13
Darfur/Sudan
Selected
16
*JM49/17
Darfur/Sudan
Selected
17
**SP1C2
Selection from Shambat P1
Selected
18
*JM36/11
Darfur/Sudan
Selected
19
*JM48/18
Darfur/Sudan
Selected
20
*JM45/7
Darfur/Sudan
Selected

Note:   * JM: Jebel Marra.  **SP: Shambat Population.   ***RP: Rawakeeb Population.

3. METHOD

A randomized complete block design with two replications was used to execute the experiment. Each genotype was grown in 4 × 5 m2/plot at the rate of 5-7 seeds/hill in ridges. Hill-to-hill and ridge-to-ridge spacing was 20 and 70 cm, respectively. Irrigation was applied at an interval of 12-14 days. Compensation was carried out a week after the sowing date. Thinning and weeding were conducted, 3 weeks after the sowing date. No pesticides were applied. For fertilization, nitrogen was applied at a rate of 3.5 kg per plot.

Ten randomly selected plants from each plot were used to estimate the following parameters:

  1. Plant height (cm): measured from the soil surface to the tip of the main panicle.
  2. Stem diameter (cm): measured at maturity on the main stalk at 10 cm above the soil level.
  3. Days to 50% flowering: The numbers of days from sowing to the time when 50% of the plants within a plot had fully exerted heads.
  4. Days to maturity: number of days from sowing to the day when all the heads in a plot had reached physiological maturity.
  5. Leaf area (cm2): the area of leaves of the internodes number 4-6 of the plant was measured and the leaf area was calculated as follows: maximum width × maximum length × 0.75.
  6. Number of leaves per plant: average number of leaves per plant after panicle exertion.
  7. Total leaf area (cm2): average leaf area times the average number of leaves per plant.
  8. Number of reproductive tillers per plant: average number of the panicle bearing tillers per plant.
  9. Number of reproductive branches per plant: average number of the panicle bearing branches on each plant.
  10. Main panicle length (cm): the length of the head beard on the main stem from the base to the tip of the panicle.
  11. Main panicle diameter (cm): the average maximum diameter of the main panicles of ten random plants was used, excluding bristles when present.
  12. Yield per main head (g): the average weight of grains produced by the main head.
  13. Yield per tiller head (g): the average weight of the grains produced by tillers multiplied by the average number of tillers per plant.
  14. Yield per branch head (g): the average weight of the grains produced by branches multiplied by the average number of branches per plant.
  15. Total yield per plant (g): this was calculated as the summation of steps 12, 13 and 14.
  16. 1000 seed weight (g): the weight of random samples of 1000 grains obtained from the grain yield of each plot.
  17. Number of grains per plant: calculated by dividing the total yield per plant by the 1000 seed weight for each accession.

In addition to the previous numerical parameters, the following qualitative characters were also studied:

  1. Anthers colour/colours: all the colours observed for each accession were recorded in order of their dominance: pale yellow, yellow, dark yellow, orange, brown, pale violet and violet.
  2. Main panicle shape: according to Khairwal, Ram, and Chhabra (1990) and Bono (1971) there are nine shapes for Pennisetum glaucum panicles: cylindrical, conical, spindle, club, candle shape, sauna, lanceolate, oblanceolate, globosely and goosy.
  3. Grain colour: these were scored as proposed by Murty, Upadhyay, and Marchonda (1967): white, yellow, deep yellow, brown, brownish grey, grey, deep grey, purple and purple black.
  4. Seed embedding: Exposed intermediate and enclosed (Khairwal et al., 1990).
  5. Glumes colour: light/dark.

3.1. Statistical Analysis

The collected data were subjected to analysis of variance then comparison among means was carried following Duncan’s Multiple Range Test as suggested by Gomez and Gomez (1984).

4. RESULTS

The result of agronomic and botanical parameters are displayed as follows:

4.1. Plant Height (cm)

Analysis of variance reflected significant difference (P=0.000) between the different genotypes in plant height Table 2. The average plant height ranged between 197.4cm for JM 45/7; as the highest genotype; and 106.3cm for Ugandi the shortest. Dwarf plants (less than 50 cm in height) appeared among genotypes Ugandi and JM36/11.

Duncan’s test for means grouped the 20 genotypes into 12 groups according to the plant height Table 3.

Variation within the population for plant height was observed in the following genotypes: Ugandi, JM24/15, RPIC2, JM25/10, JM36/11 and JM48/18.

4.2. Stem Diameter (cm)

Analysis of variance reflected significant difference (P=0.000) between the 20 pearl millet genotype in the stem diameter Table 2. The stem diameter ranged between 5.95cm for JM48/18 and 3.02cm for Ugandi i.e. the highest record is almost twice the smallest one.

Duncan’s test for means grouped the 20 genotypes into 10 groups according to the stem diameter measurement Table 4. High variation within the population was noticed in JM48/18.

Table-2. Output of Analysis of Variance, comparing 20 pearl millet genotypes in 19 characters.

Parameter
Df
Mean square
F
Sig.
Plant height
19
1099.801
43.862
.000
20
25.074
39
Stem diameter
19
1.460
108.989
.000
20
1.340E-02
39
Days to 50% inflorescence
19
38.074
9.639
.000
20
3.950
39
Days to maturation
19
29.637
49.395
.000
20
.600
39
Leaf area
19
4408.830
113.012
.000
20
39.012
39
Leaf count
19
3.627
106.663
.000
20
3.400E-02
39
Total leaf area
19
693222.342
169.769
.000
20
4083.325
39
Number of reproductive tillers
19
7.200
17.723
.000
20
.406
39
Number of reproductive branches
19
19.725
17.534
.000
20
1.125
39
Panicle length
19
22.069
4.951
.000
20
4.458
39
Panicle width
19
1.660
5.076
.000
20
.327
39
Main panicle yield
19
153.390
9.130
.000
20
16.801
39
Tiller yield
19
16.864
1.757
.110
20
9.601
39
Total tiller yield
19
4415.149
11.167
.000
20
395.382
39
Branch yield
19
21.155
4.305
.001
20
4.914
39
Total branch yield
19
5749.674
22.443
.000
20
256.193
39
Total yield
19
1574464.141
1.668
.132
20
943790.007
39
Seed weight
19
1.502
11.553
.000
20
.130
39
Seed count
19
2.293
5.018
.000
20
.457
39

Table-3. Duncan’s Test for means of plant height output.

Table-4. Duncan’s test output for stem diameter (cm).

Genotype
N
Subset for alpha = .05
1
2
3
4
5
6
7
8
9
10
Ugandi
2
3.0200
Jm3/16
2
3.3400
3.3400
Sbv
2
3.5500
3.5500
Bauda
2
3.6000
3.6000
Jm36/11
2
3.6500
3.6500
3.6500
Rp/c2
2
3.6700
3.6700
3.6700
Jm21/2
2
3.8500
3.8500
Sp/c2
2
3.8700
Jm49/17
2
3.9100
Rp2c2
2
4.1750
Jm23/13
2
4.1800
Jm30/13
2
4.2000
Jm24/15
2
4.2800
Sp2c2
2
4.3750
Madlakawia
2
4.8400
Jm44/9
2
5.2500
Darmasa
2
5.5500
Jm45/7
2
5.5600
Jm25/10
2
5.6400
Jm48/18
2
5.9500
Sig.
1.000
.085
.354
.060
.055
.135
1.000
1.000
.471
1.000

Note: Means for groups in homogeneous subsets are displayed a. Uses Harmonic Mean Sample Size = 2.000.

4.3. Days to 50% Flowering

Analysis of variance reflected significant difference (P=0.000) between the 20 genotypes. The shortest period for 50% panicle exertion was 31 days recorded fro RP2C2 and the longest period for 50% panicle exertion was 45 days recorded for: Darmasa, SP2C2, JM21/12 and JM48/18. Duncan’s test for means grouped the 20 genotypes into 8 groups according to days to 50% inflorescence Table 5. Most of the genotypes took a period of 33.0-40.5 days.

4.4. Days to Maturation

Analysis of variance reflected significant difference (P=0.000) for the 20 genotypes Table 2. The Shortest period for days to maturation was recorded for the genotypes JM49/17 (73.5 days) and RP1C2 (74.5 days); while the longest period was found to be 85.5 days recorded for the genotype Dramasa. Duncan’s test of means grouped the different genotypes into 9 groups Table 6.

4.5. Leaf Area (cm2)

Single leaf area exhibited significant difference for the 20 pearl millet genotypes Table 2. According to Duncan’s test of means, the 20 genotypes were grouped into 13 subsets Table 7. The smallest leaf area was found to be 59.9 cm2 recorded for the genotype SBV (Shambat bulk variety), while the largest leaf area was 237.28 cm2 recorded for JM 25/10, i.e. JM25/10 was almost 4 times Sbv. in leaf area.

4.6. Number of Leaves per Plant

Analysis of variance reflected significant difference (P = 0.000) for the 20 genotypes Table 2 in number of leaves per plant, which ranged between 6.3 (for Sbv.) and 10.4 (JM45/7). Ten homogenous subsets were displayed for the 20 genotypes using Duncan’s test of means Table 8.

Table-5. Duncan’s Test displayment for days to 50% inflorescence.

Genotype
N
Subset for alpha = .05
1
2
3
4
5
6
7
8
Rp2c2
2
31.0000
Ugandi
2
33.0000
33.0000
Madlakawia
2
34.0000
34.0000
Jm44/9
2
34.5000
34.5000
34.5000
Jm45/7
2
35.0000
35.0000
35.0000
35.0000
Bauda
2
36.0000
36.0000
36.0000
36.0000
Sbv
2
36.5000
36.5000
36.5000
36.5000
Jm25/10
2
37.0000
37.0000
37.0000
37.0000
Sp/c2
2
37.5000
37.5000
37.5000
37.5000
Jm36/11
2
37.5000
37.5000
37.5000
37.5000
Rp/c2
2
39.0000
39.0000
39.0000
39.0000
Jm3/16
2
39.5000
39.5000
39.5000
Jm30/13
2
39.5000
39.5000
39.5000
Jm23/13
2
40.5000
40.5000
40.5000
Jm24/15
2
42.5000
42.5000
42.5000
Jm49/17
2
42.5000
42.5000
42.5000
Darmasa
2
45.0000
45.0000
Sp2c2
2
45.0000
45.0000
Jm48/18
2
45.0000
45.0000
Jm21/2
2
45.0000
Sig.
.084
.063
.061
.063
.063
.133
.057
.195

Note:Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000.

Table-6. Days to maturation subsets as displayed by Duncan’s test.

Genotype
N
Subset for alpha = .05
1
2
3
4
5
6
7
8
9
Jm49/17
2
73.5000
Rp/c2
2
74.5000
74.5000
Bauda
2
76.0000
76.0000
Jm23/13
2
76.0000
76.0000
Jm21/2
2
76.5000
Jm3/16
2
77.0000
77.0000
Jm36/11
2
77.5000
77.5000
Sp2c2
2
78.5000
78.5000
Jm30/13
2
80.0000
80.0000
Rp2c2
2
80.5000
Sp/c2
2
81.5000
81.5000
Madlakawia
2
83.0000
83.0000
Jm25/10
2
83.0000
83.0000
Jm48/18
2
83.0000
83.0000
Jm45/7
2
83.0000
83.0000
Sbv
2
84.5000
84.5000
Jm44/9
2
84.5000
84.5000
Ugandi
2
84.5000
84.5000
Jm24/15
2
84.5000
84.5000
Darmasa
2
85.5000
Sig.
.211
.080
.096
.080
.067
.080
.096
.106
.260

Note:Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000

Table-7. Duncan’s Test output for leaf area (cm2).

Note: Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000

Table-8. Duncan’s Test output for average number of leaves per plant.

Genotype
N
Subset for alpha = .05
1
2
3
4
5
6
7
8
9
10
Sbv
2
6.3000
Bauda
2
6.6000
6.6000
Jm24/15
2
6.7000
6.7000
Ugandi
2
6.8000
6.8000
Rp2c2
2
6.9000
6.9000
Rp/c2
2
7.0000
7.0000
7.0000
Jm36/11
2
7.0000
7.0000
7.0000
Jm21/2
2
7.2000
7.2000
Jm30/13
2
7.4000
7.4000
Jm23/13
2
7.7000
7.7000
Sp/c2
2
7.7000
7.7000
Sp2c2
2
8.1000
Madlakawia
2
8.6000
Jm3/16
2
8.7000
Jm49/17
2
9.0000
9.0000
Jm44/9
2
9.3000
Jm48/18
2
9.9000
Jm25/10
2
10.0000
10.0000
Darmasa
2
10.3000
10.3000
Jm45/7
2
10.4000
Sig.
.052
.067
.064
.059
.138
.052
.052
.119
.052
.052

Note: Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000

4.7. Total Leaf Area (cm2)

Similar to single leaf area and number of leaves per plant, the total leaf area exhibited significant difference among the 20 genotypes Table 2. The highest total leaf area was almost 6 times the smallest one Table 9. The different genotypes were grouped into 11 subsets according to Duncan’s test of means.

4.8. Number of Reproductive Tillers per Plant:

Analysis of variance reflected significant difference between the 20 pearl millet genotypes (P = 0.000) in the number of reproductive tillers Table 2. According to the Duncan’s Test of means, the 20 genotypes were grouped into 7 subsets Table 10. Eleven genotypes had the lowest number of reproductive tillers (2.2 – 3.5). Two genotypes only recorded the highest number of reproductive tillers (7.5 – 8.0), these were JM45/7 and JM48/18, while the rest had the range of (4.0 - 7.0).

4.9. Number of Reproductive Branches

The 20 pearl millet genotypes exhibited significant difference in the number of reproductive branches, where (P=0.000) Table 2. The different genotypes were grouped into 7 groups Table 11. They varied greatly in the number of reproductive branches which ranged between 1.2 and 12.75.

4.10. Main Panicle Length (cm)

The analysis of variance reflected significant difference (P = 0.000) between the 20 genotypes Table 2.

The main panicle length ranged between 18.7 cm for Ugandi up to 32.0 cm for JM44/9. Seven subsets resulted Table 12 according to Duncan’s test of means.

Table-9. Duncan’s Test output for total leaf area (cm2).

Note: Means for groups in homogeneous subsets are displayed.
  a. Uses Harmonic Mean Sample Size = 2.000

Table-10. Average number of reproductive tillers per plant as displayed by Duncan’s.

Genotype
N
Subset for alpha = .05
1
2
3
4
5
6
7
Jm49/17
2
2.2000
Sbv
2
2.5000
Darmasa
2
2.2500
Jm21/2
2
2.7500
2.7500
Jm36/11
2
2.7500
2.7500
Jm23/13
2
2.8500
2.8500
Jm24/15
2
3.0500
3.0500
3.0500
Rp2c2
2
3.1000
3.1000
3.1000
Jm3/16
2
3.2500
3.2500
3.2500
3.2500
Jm30/13
2
3.3500
3.3500
3.3500
3.3500
Sp2c2
2
3.5000
3.5000
3.5000
3.5000
Sp/c2
2
4.0000
4.0000
4.0000
4.0000
Rp/c2
2
4.4500
4.4500
4.4500
Ugandi
2
4.5000
4.5000
4.5000
Bauda
2
4.7500
4.7500
Madlakawia
2
5.2500
Jm25/10
2
6.8500
Jm44/9
2
7.0500
Jm45/7
2
7.5000
7.5000
Jm48/18
2
8.5000
Sig.
.094
.104
.060
.051
.092
.347
.132

Note: Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000

Table-11. Average number of reproductive branches per plant as displayed by Duncan’s.

Genotype
N
Subset for alpha = .05
1
2
3
4
5
6
7
Sp2c2
2
1.2000
Jm36/11
2
1.2000
Sbv
2
2.0500
2.0500
Jm23/13
2
2.1500
2.1500
Bauda
2
2.1500
2.1500
Ugandi
2
2.2500
2.2500
Rp/c2
2
2.3000
2.3000
Jm24/15
2
3.7500
3.7500
Rp2c2
2
3.7500
3.7500
Madlakawia
2
4.2000
4.2000
Jm3/16
2
5.2500
5.2500
Darmasa
2
5.5000
5.5000
5.5000
Sp/c2
2
5.5000
5.5000
5.5000
Jm49/17
2
6.1000
6.1000
6.1000
Jm25/10
2
7.2500
7.2500
7.2500
Jm21/2
2
7.7500
7.7500
Jm45/7
2
7.9000
7.9000
Jm44/9
2
8.7500
Jm30/13
2
9.2500
Jm48/18
2
12.7500
Sig.
.372
.092
.065
.104
.057
.104
1.000

Note: Means for groups in homogeneous subsets are displayed.
 a. Uses Harmonic Mean Sample Size = 2.000

Table-12. Main panicle length (cm) as displayed by Duncan’s.

Genotype
N
Subset for alpha = .05
1
2
3
4
5
6
7
Ugandi
2
18.7000
Bauda
2
21.1500
21.1500
Jm49/17
2
21.3000
21.3000
Sp/c2
2
22.5000
22.5000
22.5000
Jm36/11
2
22.9000
22.9000
22.9000
Rp/c2
2
23.5000
23.5000
23.5000
23.5000
Sp2c2
2
24.2500
24.2500
24.2500
Jm48/18
2
24.2500
24.2500
24.2500
Jm24/15
2
24.6000
24.6000
24.6000
Jm23/13
2
24.7500
24.7500
24.7500
24.7500
Darmasa
2
25.5000
25.5000
25.5000
25.5000
25.5000
Madlakawia
2
25.5000
25.5000
25.5000
25.5000
25.5000
Rp2c2
2
26.4500
26.4500
26.4500
26.4500
Jm3/16
2
27.0000
27.0000
27.0000
27.0000
Jm30/13
2
27.1500
27.1500
27.1500
27.1500
27.1500
Jm45/7
2
28.1000
28.1000
28.1000
28.1000
Sbv
2
28.6000
28.6000
28.6000
28.6000
Jm25/10
2
29.7500
29.7500
29.7500
Jm21/2
2
30.0000
30.0000
Jm44/9
2
32.0000
Sig.
.056
.091
.073
.051
.053
.079
.054

Note: Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000

4.11. Main Panicle Width (cm)

Analysis of variance reflected significant difference (P=0.000) between the different pearl millet genotypes Table 2. Duncan’s test for means grouped the 20 genotypes into 7 subsets Table 13 with the range of 7.2 cm up to 10.6 cm.

It is worth mentioning that the genotype: JM24/15, JM30/13 and JM36/11 had bifurcated main stem panicles

4.12. Main Panicle Yield (g)

Analysis variance reflected significant difference (P = 0.000) between the 20 genotypes in the yield per main stem panicle Table 2. The main panicle yield ranged between 21.75g for RP2C2 up to 55.00g for JM44/9 Table 14. The 20 genotypes were grouped into 6 subsets according to the yield of the main panicle, using Duncan’s test for means.

4.13. Average Tiller Panicle Yield (g)

The average yield per tiller exhibited no significant difference (P = 0.110) among the 20 genotypes Table 2. The yield per a single tiller panicle ranged between 16.95g up to 26.75g Table 15. According to Duncan’s mean of test the 20 genotypes were divided into 3 groups only Table 15.

4.14. Total Tillers Panicle Yield (g)

This was computed by the multiplication of average number of reproductive tillers per plant by the average yield per tiller panicle (g). Although the average tiller panicle yield exhibited no significant difference between the 20 genotypes, the total yield per tillers exhibited significant difference (P = 0.000). Table 2 and Table 16, reflects the variation of the tillers yield, which ranged between 53.75g up to 207.00g. The different genotypes were grouped into 5 groups.

Table-13. Main Panicle width (cm) as displayed by Duncan’s

Genotype
N
Subset for alpha = .05
1
2
3
4
5
6
7
Jm3/16
2
7.2000
Jm25/10
2
7.5500
7.5500
Jm24/15
2
7.8000
7.8000
7.8000
Rp2c2
2
7.8000
7.8000
7.8000
Ugandi
2
8.1000
8.1000
8.1000
8.1000
Jm23/13
2
8.5000
8.5000
8.5000
8.5000
8.5000
Rp/c2
2
8.6000
8.6000
8.6000
8.6000
Jm49/17
2
8.9000
8.9000
8.9000
8.9000
8.9000
Jm36/11
2
8.9000
8.9000
8.9000
8.9000
8.9000
Jm45/7
2
9.1500
9.1500
9.1500
9.1500
Jm21/2
2
9.2000
9.2000
9.2000
Jm44/9
2
9.2500
9.2500
9.2500
Madlakawia
2
9.3500
9.3500
9.3500
9.3500
Sp/c2
2
9.4500
9.4500
9.4500
9.4500
Sp2c2
2
9.5000
9.5000
9.5000
Jm48/18
2
9.5000
9.5000
9.5000
Bauda
2
9.7500
9.7500
9.7500
Darmasa
2
10.0000
10.0000
Sbv.
2
10.5000
10.5000
Jm30/13
2
10.6500
Sig.
.056
.052
.052
.054
.075
.099
.060

Note: Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000

Table-14. Main panicle yield (g) as displayed by Duncan’s.

Genotype
N
Subset for alpha = .05
1
2
3
4
5
6
Rp2c2
2
21.7500
Jm3/16
2
23.6500
Ugandi
2
28.1500
28.1500
Jm45/7
2
30.3500
30.3500
30.3500
Bauda
2
31.0000
31.0000
31.0000
Rp/c2
2
31.0000
31.0000
31.0000
Jm49/17
2
34.8500
34.8500
34.8500
Jm23/13
2
35.8000
35.8000
35.8000
Jm24/15
2
36.4500
36.4500
36.4500
Jm21/2
2
38.3500
38.3500
38.3500
Madlakawia
2
39.1500
39.1500
39.1500
Sp2c2
2
40.1000
40.1000
40.1000
Jm36/11
2
40.9500
40.9500
Jm25/10
2
41.3000
41.3000
Darmasa
2
42.1000
42.1000
Sbv
2
42.7000
42.7000
Sp/c2
2
43.8500
43.8500
Jm48/18
2
47.1000
47.1000
Jm30/13
2
53.7000
Jm44/9
2
55.0000
Sig.
.058
.090
.052
.073
.078
.082

Note: Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000

Table-15. Duncan’s Test output for average yield (g) per a tiller panicle.

Genotype
N
Subset for alpha = .05
1
2
3
Sp/c2
2
16.9500
Ugandi
2
19.0000
19.0000
Rp2c2
2
19.3500
19.3500
19.3500
Jm3/16
2
20.3500
20.3500
20.3500
Bauda
2
20.7000
20.7000
20.7000
Madlakawia
2
21.3500
21.3500
21.3500
Jm23/13
2
21.5500
21.5500
21.5500
Rp/c2
2
21.7000
21.7000
21.7000
Jm25/10
2
22.6500
22.6500
22.6500
Jm44/9
2
23.4000
23.4000
23.4000
Jm21/2
2
24.1000
24.1000
24.1000
Jm49/17
2
24.5000
24.5000
Jm48/18
2
24.5000
24.5000
Darmasa
2
24.8500
24.8500
Sbv
2
25.2500
25.2500
Jm24/15
2
25.9500
25.9500
Sp2c2
2
26.5000
26.5000
Jm45/7
2
26.5000
26.5000
Jm36/13
2
26.7500
Jm36/11
2
26.7500
Sig.
.061
.053
.056

Note: Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000

Table-16. Total Yield per tillers panicle (g) as displayed by Duncan’s.

Genotype
N
Subset for alpha = .05
1
2
3
4
5
Jm49/17
2
53.7500
Darmasa
2
56.5250
Sbv
2
57.3000
Rp2c2
2
60.0100
Jm23/13
2
61.7850
Jm3/16
2
66.9750
66.9750
Sp/c2
2
67.3500
67.3500
Jm21/2
2
67.5000
67.5000
Jm36/11
2
73.4500
73.4500
Jm24/15
2
79.6200
79.6200
Ugandi
2
86.0000
86.0000
Jm30/13
2
90.2950
90.2950
Sp2c2
2
92.7500
92.7500
Rp/c2
2
96.4500
96.4500
Bauda
2
98.1500
98.1500
Madlakawia
2
111.9750
111.9750
Jm25/10
2
152.0500
152.0500
Jm44/9
2
166.5000
166.5000
Jm45/7
2
198.8500
Jm48/18
2
207.0000
Sig.
.071
.065
.057
.476
.067

Note: Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000

4.15. Yield per Single Branch Panicle (g)

The average yield per single branch tiller exhibited significant difference (P = 0.001) between the 20 millet genotypes Table 2. Duncan’s test for mean grouped the different genotypes into 4 groups (table 4.16). The mean branch tiller yield ranged between 8.5g up to 90.7g. Almost 15 genotypes were grouped into one subset with mean ranged between 10.5g – 14.7g Table 17.

4.16. Total Branch Tillers Yield (g)

Similar to the yield per single branch tiller, total branches tillers exhibited significant difference (P = 0.000) between the 20 pearl millet genotypes Table 2 this ranged between 17.47g up to 231.0g Table 18. Again Duncan’s test for means grouped the 20 genotypes into 8 groups. The genotypes JM48/18 had the highest record (231.9g) exhibited alone a single subset.

4.17. Total Yield per Plant (g)

This was calculated by the summation of the: Yield per main panicle, total tillers panicle yield and total branches panicle yield. The analysis of variance reflected no significant difference (P = 1.32) among the 20 pearl millet genotypes Table 2 although significant difference was exhibited for main panicle yield and total branch yield.

Total yield per plant Table 19 ranged between 630.59g and 3621.22g. Duncan’s test for means grouped the 20 genotypes into 3 subsets Table 19.

Table-17. Yield per a single branch tiller (g) as displayed by Duncan’s.

Genotype
N
Sunset for alpha = .05
1
2
3
4
Ugandi
2
8.5000
 
Jm25/10
2
8.5500
 
Jm3/16
2
10.5000
10.5000
 
Darmasa
2
10.6000
10.6000
 
Sp/c2
2
10.7500
10.7500
 
Jm30/13
2
10.8000
10.8000
 
Jm49/17
2
11.2500
11.2500
 
Rp/c2
2
11.3500
11.3500
 
Jm23/13
2
11.5000
11.5000
 
Rp2c2
2
11.9500
11.9500
 
Bauda
2
12.0000
12.0000
 
Jm44/9
2
12.4000
12.4000
 
Jm24/15
2
13.0000
13.0000
 
Jm21/2
2
13.7500
13.7500
13.7500
 
Madlakawia
2
14.4000
14.4000
 
Sbv
2
14.6000
14.6000
 
Sp2c2
2
14.7500
14.7500
 
Jm48/18
2
18.2500
18.2500
Jm36/11
2
18.8000
18.8000
Jm45/7
2
20.7000
Sig.
 
.056
.115
.055
.309

Note: Means for groups in homogeneous subsets are displayed.
 a. Uses Harmonic Mean Sample Size = 2.000

4.18. 1000 Grain Weight (g)

Analysis of variance reflected significant difference (P = 0.000) among the 20 genotypes in 1000 grain weight Table 2, with value ranged between 6.15g to 9.05g Table 20. Duncan’s test grouped the 20 genotypes into 8 subsets.

4.19. Number of Grains per Plant

This was calculated by dividing the total yield per plant by 1000 grain weight for each accession. Significant difference (P = 0.000) was exhibited between the 20 genotypes Table 2. The number of grains per plant ranged between 3.48x103  for Rp2c2 and 7.43*10 for Jm30/13 Table 21. According to Duncan’s test of means, they were grouped into 6 subsets.

4.20. Qualitative Characters Results

The 20 pearl millet genotypes were compared in: Anther color/colors, panicle shape, grain color/colors, glume color and seed exposure. The results for these qualitative characters are represented in Table 22.

Table-18. Total Yield of branch panicles (g) as displayed by Duncan’s

.

Genotype
N
Subset for alpha = .05
1
2
3
4
5
6
7
8
Sp2c2
2
17.4750
Ugandi
2
19.0000
Jm36/11
2
22.4700
22.4700
Jm23/13
2
24.7400
24.7400
24.7400
Bauda
2
25.9000
25.9000
25.9000
Rp/c2
2
26.2700
26.2700
26.2700
Sbv
2
29.5700
29.5700
29.5700
Rp2c2
2
47.0250
47.0250
47.0250
47.0250
Jm24/15
2
48.5000
48.5000
48.5000
48.5000
Jm3/16
2
55.0000
55.0000
55.0000
55.0000
Sp/c2
2
57.8500
57.8500
57.8500
Darmasa
2
58.3000
58.3000
58.3000
Madlakawia
2
59.4800
59.4800
59.4800
Jm25/10
2
62.2250
62.2250
Jm49/17
2
69.2550
69.2550
Jm30/13
2
100.5750
100.5750
Jm44/9
2
104.3000
104.3000
Jm21/2
2
105.8500
Jm45/7
2
163.5000
Jm48/18
2
231.0000
Sig.
.053
.060
.057
.239
.050
.759
1.000
1.000

Note: Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000

Table-19. Total Yield (g) as displayed by Duncan’s.

Genotype
N
Subset for alpha = .05
1
2
3
Sbv
2
630.5950
Ugandi
2
918.0550
918.0550
Jm21/2
2
1240.2300
1240.2300
1240.2300
Jm24/15
2
1259.4550
1259.4550
1259.4550
Jm36/11
2
1263.0450
1263.0450
1263.0450
Rp2c2
2
1280.8200
1280.8200
1280.8200
Bauda
2
1330.6300
1330.6300
1330.6300
Jm23/13
2
1511.9550
1511.9550
1511.9550
Sp/c2
2
1575.1650
1575.1650
1575.1650
Rp/c2
2
1594.6200
1594.6200
1594.6200
Jm49/17
2
1956.1900
1956.1900
1956.1900
Jm3/16
2
2042.7600
2042.7600
2042.7600
Sp2c2
2
2045.100
2045.100
2045.100
Jm30/13
2
2051.2600
2051.2600
2051.2600
Madlakawia
2
2303.6500
2303.6500
2303.6500
Jm44/9
2
2865.7500
2865.7500
2865.7500
Darmasa
2
3199.6150
3199.6150
Jm48/18
2
3333.4450
Jm45/7
2
3419.2800
Jm25/10
2
3621.2250
Sig.
.064
.059
.050

Note: Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000

Table-20. 1000 grain weight (g) output for the 20 genotypes.

Genotype
N
Subset for alpha = .05
1
2
3
4
5
6
7
8
Jm3/16
2
6.1500
Rp2c2
2
6.2500
6.2500
Jm21/2
2
6.2500
6.2500
Jm49/17
2
6.2500
6.2500
Jm23/13
2
6.7500
6.7500
6.7500
Ugandi
2
6.9000
6.9000
6.9000
6.9000
Madlakawia
2
7.0000
7.0000
7.0000
7.0000
Jm36/11
2
7.0000
7.0000
7.0000
7.0000
Sbv
2
7.1000
7.1000
7.1000
Bauda
2
7.1000
7.1000
7.1000
Jm25/10
2
7.2500
7.2500
7.2500
Jm30/13
2
7.2500
7.2500
7.2500
Sp2c2
2
7.5000
7.5000
7.5000
7.5000
Jm24/15
2
7.5000
7.5000
7.5000
7.5000
Jm48/18
2
7.7500
7.7500
7.7500
Sp/c2
2
8.0000
8.0000
8.0000
Jm45/7
2
8.2500
8.2500
8.2500
Jm44/9
2
8.7500
8.7500
Rp/c2
2
8.7500
8.7500
Darmasa
2
9.0500
Sig.
.052
.054
.087
.055
.075
.069
.054

Note: Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000

Table-21. Total number of grains per plant as grouped according to Duncan’s analysis. The values are multiple of 103.

Genotype
N
Subset for alpha = .05
1
2
3
4
5
6
Rp2c2
2
3.4800
Rp/c2
2
3.5450
Jm45/7
2
3.6850
Jm3/16
2
3.8500
3.8500
Ugandi
2
4.0800
4.0800
4.0800
Bauda
2
4.3650
4.3650
4.3650
4.3650
Darmasa
2
4.6550
4.6550
4.6550
4.6550
4.6550
Jm24/15
2
4.9100
4.9100
4.9100
4.9100
4.9100
Jm23/13
2
5.2900
5.2900
5.2900
5.2900
Sp2c2
2
5.4250
5.4250
5.4250
5.4250
Sp/c2
2
5.4800
5.4800
5.4800
Jm49/17
2
5.5700
5.5700
5.5700
Madlakawia
2
5.5900
5.5900
5.5900
Jm25/10
2
5.7150
5.7150
Jm36/11
2
5.8500
5.8500
Sbv
2
6.0150
6.0150
Jm48/18
2
6.1100
6.1100
Jm21/2
2
6.1550
6.1550
Jm44/9
2
6.2750
6.2750
Jm30/13
2
7.4300
Sig.
.079
.053
.066
.072
.053
.073

Note: Means for groups in homogeneous subsets are displayed.
a. Uses Harmonic Mean Sample Size = 2.000

5. DISCUSSION

Selection of favorable genotypes for certain characters depends mainly on the amount of variation present in the material under consideration. In this regards, the assessment of morphological and agronomic characters variation among populations is of great importance for successful application of selection procedures for the improvement of populations. The phenotypic variance is attributed to genotypic as well as environmental factors.

Table-22. Different qualitative characters results recorded for the 20 pearl millet genotypes.

Genotype
*Anther colour
**Panicle shape
***Grain colour
Glume colour
Seed exposure
Sbv
Org
Cyl, Csh
Gry
Light
Exposed
Bauda
Yel
Cyl
Wh
Light
Exposed
JM44/9
Yel.
Cyl,Csh
Gry, yel
Light
Exposed
Darmasa
Org, yel, vio
Cyl, Cl
Gry, yel
Dark
Exposed
SP2C2
Org, yel
Cyl
Gry, l. yel
Dark
Exposed
Ugandi
Wh, yel
Cyl
Gry, d.gry
Light/dark
Exposed
JM24/15
Org, yel
Cyl
Gry, d. gry
Light
Enclosed
RP2C2
Yel
Cyl
Gry, d. gry
Dark
Enclosed
JM23/13
Wh, yel, org
Cyl
Gry
Light/dark
Enclosed
RP1C2
Org, yel, vio
Cyl
Gry, yel rn
Light/dark
Intermediate
JM3/16
Og, yel
Cyl
Gry, d. gry, yel
Light
Exposed
Madekawia
Wh, yel, org
La, gos
Yel, gry, d. gry
Light
Exposed
JM21/2
Wh, yel, org, vio
Cyl
Gry, d. gry
Light/dark
Exposed
JM25/2
Wh, org, yel
Cyl
Yel, d. gry
Light
Exposed
JM30/13
Yel
Csh
Yel, l. gry
Dark
Exposed
JM49/17
Yel, d. yel
Cyl
Yel, gry, l. gry
Light
Exposed
SP1C2
Org, yel, vio
Cyl
Yel, gry, d. yel
Light
Enclosed
JM36/11
Org, yel
Cyl
Yel, gry
Light
Intermediate
JM48/18
Org, yel
Cyl, Csh
Yel, gry, d. gry
Light
Intermediate
JM45/7
Yel, vio
Cyl
Yel, gry
Light
Exposed

Note: * Anther colour/colours: The abbreviations: org= orange, yel = yellow, wh = white, vio = violet and d.yel = dark yellow.
** Panicles shapes: Cyl = cylindrical, Csh = candle shape, Cl = Club, La = Lanceolate, gos = goosy.
*** Grain colour: gry = grey, wh = white, d. yel = dark yellow, yel = yellow, l. yel = light yellow, d. gry = dark grey, l. gry = light grey and brn = brown.

In the present study, the twenty pearl millet genotypes exhibited significant differences in most of the characters studied. Similar findings were obtained by many workers in pearl millet and different cereal crops and under different environments (Abuelgasim, 1999; Fadlalla, 2002). Considering the twenty genotypes studied, there is a noticeable variation, not only among the different genotypes of the same species but also within the same genotype or cultivar. Such variation may be attributed to the open pollination system of this crop (Jauhur, 1981; Khairwal et al., 1990) and to independent domestication and emigrational events (Harlan, 1976). The genotypes Bauda, Darmasa, Madlkawia and JM45/7 had unnoticeable variation within the population itself, i.e. these genotypes have higher degree of homogeneity and stability.

In the 19 parameters studied, the different genotypes exhibited significant difference for most of the parameters except for: yield per tiller panicle and total yield per plant.

This variation between the different genotypes can be of a special value for adaptation to the local agro climatic conditions. For example the unpredictable rainfall in the western dry parts of the country can cause loss of the main stem and hence loss of the main stem panicle. The high tillering ( ability to produce tillers) capacity genotypes can compensate for the loss of the main panicle succeeding in preventing a famine and starvation. The same value can be related to the variation in days to 50% inflorescence and days to maturation. Earlier blooming genotypes can be very successful in areas of unpredictable rainy season duration. There are some genotypes with special and unique characters, which might be of importance to the plant protection. Example the long bristles possessed by the spike of the genotype Madlkawia which is thought to be useful against birds attack. It seems from this study that the bristles may be a good defence only during the end of the season, i.e. after the seed ripening since these bristles are soft during the milky stage, they become only pointed and stiff after seed ripening. Thus the high variability regarding the plant morphological features and head characteristics (Jain & El Ahmadi, 1981) helped in the high adaptation to the local agro climatic conditions especially to the dry western parts of the Sudan with its unpredictable amount and duration of the rainy seasons.

Funding: This study received no specific financial support.  

Competing Interests: The authors declare that they have no competing interests.

Acknowledgement: All authors contributed equally to the conception and design of the study.

REFERENCES

Abuelgasim, E. H. (1999). Pearl millet breeding and improvement programme in Sudan (Arabic). Khartoum: Ministry of Agriculture and Forestry.

Bono, M. (1971). Contribution to the morphosystem-atic of Pennisetum, annual cultivated for their grain in accidental Africa Francophore. Memoir for the Obtaining the Degree of Doctor Ingenieur.  

Brunken, J. N. (1977). A systematic study of Pennisetum sect. Pennisetum (Graminea). American Journal of Botany, 64, 161-176.Available at: https://doi.org/10.1002/j.1537-2197.1977.tb15715.x.

Chandra, M., & Matta, N. K. (1990). Characterization of millet protein fractions. Phytochemistry, 29(11), 3395-3399.

Fadlalla, H. A. (2002). Selection for drought tolerant in two random mating populations of pearl millet (Pennisetum glaucum). A Thesis Submitted in Fulfillment for the Requirement of Ph.D. Degree  to the Faculty of Agriculture, University of Khartoum.  

FAO. (1999). FAO quartery bulletin of statistics. 12(1), 31-32.

Gill, K. S. (1991). Pearl millet and its improvements. New Dalhi: Indian Council of Agricultural Research.

Gomez, K. A., & Gomez, A. A. (1984). Statistical procedures for agricultural research (2nd ed.). New York: John Wiley and Sons, Inc.

Harlan, J. R. (1976). Genetic resources in wild relation of crops. Crop Science, 16(3), 329-333.

Hazza, M. M. (1994). Line X Tester Analysis for the study of Heterosis and Combining Ability in some Genotypes of pearl millet (Pennisetum glaucum). A Thesis Submitted for the Degree of Master of Science (Agric.), University of Khartoum.  

Jain, R. P., & El Ahmadi, B. (1981). Pearl millet production and Research in the Sudan. Agricultural Research Division: Annual Reports.

Jauhur, P. P. (1981). Cytogenetic and breeding of pearl millet and related species (Vol. 1, pp. 2-3). New York: Alan R. Liss. Inc.

Khairwal, I. S., Ram, C., & Chhabra, A. K. (1990). Pearl millet seed production and technology. India: Manohar.

Kumar, K. A. (1993). Pearl millet in West Africa. Patancherw, India: Icrisat Centre.

Kumari, J., Bag, M. K., Pandey, S., Jha, S., Chauhan, S., Jha, G. K., & Dutta, M. (2016). Assessment of phenotypic diversity in pearl millet [Pennisetum glaucum (L.) R. Br.] germplasm of Indian origin and identification of trait-specific germplasm. Crop and Pasture Science, 67(12), 1223-1234.Available at: https://doi.org/10.1071/cp16300.

Murty, B. R., Upadhyay, M. K., & Marchonda, P. L. (1967). Classification and cataloguing of world collection of genetic stocks of Pennisetum. Indian Journal of Genetics and Plant Breeding, 27(spI), 313-394.

Salih, M. A. M. (1997). Cytogenetic studies in two pennisetum spp. A Thesis Submitted for the Degree of Master of Science (Botany) in the Dep of Botany, Faculty of Science, U. of Khartoum.  

Salih., M. A. M., El-Tigani, S., & Abdalla, A. H. (1999). C-banding in the Karyotype of two Pennisetum species. U. of K. Journal of Agriculture Science, 7(1), 34-47.

Serba, D. D., Muleta, K. T., St. Amand, P., Bernardo, A., Bai, G., Perumal, R., & Bashir, E. (2019). Genetic diversity, population structure, and linkage disequilibrium of pearl millet. The Plant Genome, 12(3), 1-12.Available at: https://doi.org/10.3835/plantgenome2018.11.0091.

Views and opinions expressed in this article are the views and opinions of the author(s), International Journal of Sustainable Agricultural Research shall not be responsible or answerable for any loss, damage or liability etc. caused in relation to/arising out of the use of the content.