|
African Crop Science Journal
African Crop Science Society
ISSN: 1021-9730 EISSN: 2072-6589
Vol. 8, Num. 2, 2000, pp. 145-152
|
African Crop Science Journal, Vol. 8. No. 2, pp. 145-152
African Crop Science Journal, Vol. 8. No. 2, pp. 145-152
VARIABILITY BETWEEN LOCAL AND EXOTIC BAMBARA GROUNDNUT LANDRACES IN BOTSWANA
S. K. KARIKARI Department of Crop Science and Production, Botswana College of Agriculture, Private Bag 0027, Gaborone, Botswana
Received 3 March, 1999; accepted January, 2000)
Code Number: CS00015
INTRODUCTION
Bambara groundnut (Vigna subterranea (L) Verdc.) is one of the most important leguminous crops grown by smallholder farmers in Botswana. Available estimates reveal that the crop is grown on 1,500 hectares of land from which a total yield of 40 tonnes of seed is obtained annually. Because the country is arid, dryland crop farming is limited to short season legumes which have inherent drought tolerance.
Previous Bambara groundnut germplasm collections in Botswana had a few landraces in two regions (north East and Kgatleng) based on pericarp colour (Ministry of Agriculture, Botswana, 1947; Mazhani and Appa-Rao, 1985; Appa Rao et al., 1986). The genetic base was quite narrow as the area covered was limited. Therefore, it is important that the genetic diversity be expanded with new sources of germplasm, including wild and exotic species. Previous research (Doku and Karikari, 1971) has shown that wild accessions of Bambara groundnut have useful genetic variability for maturity period, yield and seed characteristics. Observations on 27 genotypes in Ghana led to the conclusion
that the cultivated species evolved from Vigna subterranea var spontenea.
A useful tool in plant breeding for selecting useful traits is the analysis
of correlations among measurable characters. Highly significant correlations
between pairs of traits can be exploited. In a study of 27 genotypes under optimum
agronomic conditions in Ghana (Karikari, 1972), a simple correlation analysis
indicated that number of stems per plant and 100 seed weight were positively
correlated with grain yield and these characters were used for selection. The
characterisation and evaluation of Bambara groundnut at the International Institute
of Tropical Agriculture (IITA) revealed enormous agro-morphological diversity
which can be used for improvement. In correlation matrix study of the IITA collection,
Goli et al. (1997) found that characters most strongly correlated with
grain yield were number of leaves and pods, shell thickness and 100 seed weight.
Tanimu and Aliyu (1990) observed that shelling percentage and 100 seed weight
were important characters correlated with grain yield, and therefore used these
to select for grain yield within the germplasm collection at the Ahmadu Bello
University in Nigeria.
The objective of this study was to determine the variability between the landraces of Bambara groundnut within the new Botswana collection and also correlations between yield and agronomic traits to aid breeders in utilising these traits in breeding programmes.
MATERIALS AND METHODS
Field experiments were carried out at the Botswana College of Agriculture, Sebele, over a three year period (1994/95 - 1996/97). Sebele is located 24o33', 25o54E at an elevation of 994 m above sea level. Land preparation involved uniform cultivation to make a fine seedbed and application of basal fertiliser at 26 kg N and 140 kg P2O5 ha-1. The experimental design was a randomised complete block replicated four times, in which the Bambara groundnut landraces were assigned at random to the plots. Individual plot sizes were 1.5 x 10 m, with 0.5 m guard row planted in between each experimental plot. The landraces used were Botswana Red (BOTR), Gaborone Cream (GABC), Diphiri Cream (DIPC), Tshesebe Cream (TSHC), OMotswasele 1, 2 and 6 (OM1, OM2 and OM6),
National Tested Seed Red (NTSR) and Dodoma Red (DOR). All the landraces originated
from Botswana except NTSR and DOR which originated from Zimbabwe and Tanzania,
respectively. These have been characterised using the IBPGR Descriptors List
(IBPGR/IITA/GTZ, 1987), and their features described by Karikari (1996). Two
seeds were sown per station at 5 cm depth at a spacing of 30 x 30 cm, and seedlings
were thinned to one seven days after emergence. Irrigation was supplied to all
plots to ensure that the soil was at field capacity from sowing until crop establishment
(21 days after emergence). Thereafter, all plants were left to grow on moisture
stored in the soil profile plus any rain that fell during the season. The field
was kept weed-free throughout the period of the experiment.
Parameters measured were days from sowing to emergence and 50% flowering, canopy spread at pod initiation, petiole/internode ratio, 100 seed weight, shelling percentage, grain yield per hectare, total dry matter and days from sowing to 50% maturity. Techniques for these measurements have been described in detail by Karikari et al. (1996). Data of the quantitative characters for the three seasons were combined after Bartletts test (Sokal and Rohlf, 1969) showed homogeneity of error variances and were analysed using SAS/STAT (SAS Inst., 1990). Where significant effects were found, mean separation were accomplished using the Fishers LSD values (Steel and Torrie, 1986). Correlation matrix of agronomic characters and grain yield were also
established. Broad sense heritability (h2) were computed according to the method
of Moll et al. (1960) as:
h2 = |
d2g |
with SE = |
SEd2g |
|
|
d2g |
SEd2ph |
where |
d2g = genotypic variance SEd2ph = phenotypic variance SE = standard error |
RESULTS AND DISCUSSION
The variability of means between the landraces is shown in
Table 1, and the results of analysis of variance in Table 2. Number of days
from sowing to emergence varied from 14 to 23.4. Four landraces having red pericarp
(BOTR, OM6, NTSR and DOR) took longer (over 21 days) to emerge. They had small
seeds. The cream coloured landraces (DIPC, GABC and TSHC) with bigger seed sizes
emerged faster, within 14 to 17 days. Under conditions of low rainfall, rapid
emergence would be advantageous as this reduces the period over which seedlings
are susceptible to stress, and the quicker the roots develop the more likely
the seedling is able to withstand drought.
TABLE 1. Average quantitative characters of nine Bambara groundnut
landraces at Sebele, Botswana |
Landrace |
Grain yield (kg ha-1) |
100 seed weight (g) |
Shelling percentage |
Dry matter (g) |
Days to maturity |
Dayst to emergence |
Days to flowering |
Canopy spread (cm) |
Petiole/internode ratio |
BOTR |
505 |
54 |
40 |
30.01 |
112 |
21.4 |
50 |
44 |
5.71 |
GABC |
1208 |
86 |
44 |
41.17 |
88 |
16.5 |
40 |
22 |
8.83 |
DIPC |
1477 |
75 |
41 |
39.02 |
116 |
14.0 |
37 |
26 |
8.88 |
TSHC |
944 |
72 |
51 |
38.16 |
88 |
16.6 |
38 |
28 |
9.06 |
OM1 |
726 |
64 |
59 |
30.90 |
120 |
20.2 |
51 |
35 |
7.81 |
OM2 |
702 |
66 |
62 |
33.45 |
127 |
20.7 |
42 |
40 |
7.94 |
OM6 |
668 |
61 |
57 |
31.07 |
133 |
21.4 |
48 |
33 |
8.00 |
NTSR |
455 |
60 |
66 |
33.08 |
128 |
22.3 |
56 |
42 |
6.02 |
DOR |
380 |
50 |
62 |
35.21 |
165 |
23.4 |
52 |
46 |
6.66 |
Mean |
785 |
65 |
54 |
34.67 |
120 |
19.6 |
46 |
35 |
7.66 |
SD+ |
342.94 |
10.42 |
9.30 |
3.75 |
22.20 |
2.98 |
6.48 |
8.02 |
1.18 |
CV % |
43.69 |
15.96 |
17.36 |
10.81 |
18.55 |
15.18 |
14.09 |
22.84 |
15.4 |
F test |
** |
** |
** |
* |
* |
ns |
ns |
ns |
ns |
Landraces: BOTR = Botswana Red; GABC = Gaborone Cream;
DIPC = Diphiri Cream, TSHC = Tshesebe Cream; OM 1, 2 and 3 = OMotswasele
1, 2 and 3; NTSR = National Tested Seed Red; DOR = Dodoma Red. Significant
at the 0.05 level (*), 0.01 (**) or nonsignificant (ns)
|
TABLE 2. Analysis of variance table (sums of squares) for different
characters assessed in Bambara groundnut landraces in Botswana |
Sources of variation |
d.f. |
Grain yield |
100 seed weight |
Shelling percentage |
Total dry weight |
Days to maturity |
Days to emergence |
Days to flowering |
Canopy spread |
Petiole/internode ratio |
Replications |
3 |
61310.44 |
30.47 |
19.83 |
27.04 |
11.44 |
0.05 |
4.03 |
0.25 |
0.22 |
Landraces |
8 |
54097.58** |
2460.62** |
144.45** |
71.30* |
36.08* |
31.6 ns |
2.77 ns |
1.34 ns |
0.46 ns |
Residual |
24 |
60205.32 |
77.78 |
34.86 |
12.02 |
13.30 |
1.09 |
0.75 |
1.26 |
0.84 |
Days to flowering was correlated to the emergence pattern. Generally, within 21 to 24 days after emergence, flowering would occur among the early maturing landraces. The exotic landraces (NTSR and DOR) and a Botswana landrace (BOTR) flowered 28 - 40 days after emergence. These landraces are known to be photosensitive (Wigglesworth, 1996). Other researchers (Begemann, 1988; Nishitani et al., 1988; Linemann, 1991, Linemann and Azam-Ali, 1991) have also observed photoperiodic responses in local and exotic germplasm which also flowered in much longer time.
DOR, NTSR and BOTR had wider canopies whose spread were over 40 cm. These wider canopies were due to the low petiole/internode ratios (less than 7.00). The petiole/internode ratios in the cream coloured landraces were high, greater than 8.00, while the ratios in OM1, OM2 and OM6 were intermediate and lay between 7.00 and 8.00. The earlier classification based on growth habit was found to be useful for this autogamous crop species (Doku and Karikari, 1971; Doku and Opoku-Asiama, 1978). DOR, NTSR and BOTR fit into the Doku and Karikaris (1971) classification of the spreading varieties
while OM1, OM2 and OM6 fit a semi-spreading description. GABC, DIPC and THSC
belong to the bunch types. The growth habit of the landraces have strategic
use; the spreading ones could be used in intercropping situations where they
would form a more rapid ground cover and help suppress weed growth while the
bunch and semi-bunch types would be used in monocultures where optimum yield
would be achieved at higher plant populations. Seed sizes were highly significantly
bigger in the landraces with cream pericarp. Since the people of Botswana prefer
the cream coloured seeds for consumption, as these are claimed to be tastier
(Brink et al., 1996), selection for size has been in favour of these,
and therefore, their seed sizes have improved over time.
Shelling percentage was found to be significantly related to seed size; the low shelling percentage in the large seeded landraces of short season growth was because a high proportion of the pod mass was contributed to by the seed. Grain yield was significantly greater in GABC, DIPC and TSHC than DOR, NTSR and BOTR. GABC and DIPC produced over one tonne of seed while DOR, NTSR and BOTR produced less. The high seed yield was contributed to by the high total dry matter. The maturity periods indicated earliness (mean of 88 days) for GABC and TSHC and lateness for OM2 (127 days), NTSR(128 days) and OM6 (133 days). DOR was very late, maturing in 165 days.
Table 3 shows the correlation matrix of eight agronomic characters
and yield. The values of the coefficients are greater than 0.20, indicating
that they were statistically significant (Dowdy and Wearden, 1991). Since improved
grain yield is the major parameter being sought, its correlation with other
characters would be most desirable. Yield related parameters demonstrating very
highly significant and positive correlation with grain yields are seed size
(100 seed weight) (r = 0.88), shelling percentage (r = 0.82) and dry matter
productivity (r = 0.75). These characters are useful to plant breeders who may
include them in selection programme. The petiole/internode ratio was another
important growth related parameter which correlated positively with grain yield.
Plants with high petiole/internode ratio tended to be bunchy and upright and
so were more exposed to sunlight than those with low ratio which tended to creep.
Bunch habit is also an important consideration for ease of harvesting (Doku
and Karikari, 1971). The negative correlations between grain yield and canopy
spread and grain yield and days to maturity provide evidence that spreading
landraces take longer to mature and suffer yield decline due to reduced moisture
and cold weather.
TABLE 3. Pearsons correlation coefficients of agronomic parameters
and yield in Bambara groundnut landraces |
|
Days to emergence |
Days to flowering |
Canopy spread (cm) |
Petiole/ internode ratio |
100 seed weight (g) |
Shelling percentage |
Dry matter at harvest (g) |
Days to maturity |
Grain yield (kg ha-1) |
Days to emergence |
- |
0.89 |
0.89 |
-0.77 |
-0.88 |
-0.86 |
-0.69 |
0.74 |
0.96 |
Days to flowering |
0.89 |
- |
0.92 |
-0.88 |
-0.79 |
-0.84 |
-0.58 |
0.64 |
0.84 |
Canopy spread (cm) |
0.89 |
0.92 |
- |
-0.89 |
-0.86 |
-0.95 |
-0.51 |
0.71 |
-0.85 |
Petiole/ internode ratio |
-0.77 |
-0.88 |
-0.89 |
- |
0.75 |
0.92 |
0.58 |
-0.44 |
0.76 |
100 seed weight (g) |
-0.88 |
-0.97 |
-0.86 |
0.75 |
- |
0.87 |
0.77 |
-0.80 |
0.88 |
Shelling percentage |
-0.87 |
-0.84 |
-0.95 |
0.92 |
0.88 |
- |
0.61 |
-0.68 |
0.82 |
Dry matter at harvest (g) |
-0.69 |
0.58 |
-0.51 |
0.58 |
0.77 |
0.61 |
- |
-0.50 |
0.75 |
Days to maturity |
0.74 |
0.64 |
0.71 |
-0.44 |
-0.80 |
-0.68 |
-0.50 |
- |
-0.63 |
Broadsense heritability (h2) estimates for grain yield, 100
seed weight and shelling percentage, which had the highest r values, are presented
in Table 4. The highest h2 estimate was recorded for 100 seed weight. The h2
estimate for grain yield was low and had a large standard error. Yield as a
quantitative trait has been found to have low heritability (Bordia et al.,
1973).
TABLE 4. Estimates of heritability (h2) with their respective
standard errors (SE+) in nine Bambara groundnut landraces in Botswana |
Character |
h2 |
Grain yield |
71.63+14.31 |
100 seed weight |
25.14+8.83 |
Shelling percentage |
37.87+9.71 |
Dry matter plant-1 |
36.33+7.04 |
CONCLUSIONS AND RECOMMENDATIONS
The high heritability of 100 seed weight indicates that significant increases in yield can be obtained by selection for this trait within the Botswana landraces. Farmers are aware of this and a limited amount of farmer selection for seed size has occurred in Botswana. A farm survey (Brink et al., 1996) indicated that selection of large seeds for planting takes place. There is a need to explore in greater detail the genetic relationship between grain yield and other agronomic characters. Traits such as 100 seed weight may serve as a criterion for indirect selection for grain yield, although a more exhaustive and critical analysis (path coefficient analysis) of these correlations is needed.
This study shows that 100 seed weight and shelling percentage of the nine Bambara groundnut landraces in Botswana demonstrate substantial phenotypic variability and these characters could be used in breeding. However, given the relationship of the various characteristics with grain yield (Table 3), it is obvious that no single character is absolute in view of the high correlations of some of these characters with grain yield. Yield is a complex terminal outcome of growth and development to which there are diverse and interrelated developmental tracks. Bordia et al. (1973) and Uguru (1995) observed in cowpea (Vigna unguiculata L.), a legume in the same genus as Bambara groundnut, that characters have not only additive effects but also multiplicative effects. Therefore in selecting for high grain yield, several characters should be taken into consideration simultaneously. Although many characters were correlated with yield, it was most likely that they had more complex relationship with
each other which could not be explained in a linear relationship. In this study
100 seed weight was found to be the most important character to be considered
during selection and breeding of Bambara groundnut in Botswana and areas with
similar climate.
ACKNOWLEDGEMENT
The assistance given by Dr R. M. Sakia in the data analysis is duly acknowledged.
REFERENCES
- Appa Rao, S., Mazhani, L. M. and Attere, A. F. 1986. Collecting Bambara groundnut in Botswana. Plant Genetic Resources Newsletter 68:27-28.
- Begemann, F. 1988. Ecogeographic differentiation of Bambara groundnut in the collection of the International Institute of Tropical Agriculture (IITA). Wissenschaftlicher Fachverlag Dr Fleck, Niederkleen, Germany. 153 pp.
- Bordia, P. C., Yadavendra, J. P. and Kumar, S. 1973. Genetic variability and correlation studies in cowpea (Vigna unguiculata L.). Journal of Agricultural Science 1:39-44.
- Brink, M., Collinson, S. T. and Wigglesworth, D. J. 1996. Characteristics of Bambara groundnut cultivation in Botswana. In: Proceedings of the International Bambara Groundnut Symposium held at the University of Nottingham, U.K., 23 - 25 July, 1996. Azam-Ali, S., N., Sesay, A. and Collinson, S. T. (Eds), pp. 133-142.
- Doku, E. V. and Karikari, S. K. 1971. Operational selection in Bambara groundnut (Voandzeia subterranea) under cultivation. Ghana Journal of Science 11:45- 56.
- Doku, E. V. and Opoku-Asiama, Y. 1978. A classification of Bambara groundnut cultivars based on variation in morphological characteristics. In: Proceedings of the First National Symposium on Crop Improvement in Ghana, held at the University of Ghana, Legon, 22-24 August, 1978.Doku, E.V. (Ed.), pp 19-27. University of Ghana/CSIR, Accra, Ghana.
- Dowdy, S. and Wearden, S. 1991. Statistics for Research. John Wiley and Sons, Chichester 452 pp.
- Goli, A. E., Begemann, F. and Ng, N. Q. 1997. Characterisation and evaluation of IITAs Bambara groundnut. In: Bambara Groundnut (Vigna subterranea (L.) Verdc.). Promoting the Conservation and Use of Underutilised and Neglected Crops. No. 9. Heller, J., Begemann, F. and Mushonga, J. (Eds), pp. 101-118. IPGRI, Rome, Italy.
- IBPGR/IITA/GTZ, 1987. Descriptors for Bambara Groundnut. International Board for Plant Genetic Resources, Rome. 23 pp.
- Karikari, S. K. 1972. Correlation studies between yield and some agronomic characters in Bambara groundnut. Ghana Journal of Agricultural Science 5:79-83.
- Karikari, S. K. 1996. The status of Bambara groundnut genetic resources in Botswana. SACCAR Newsletter 34:8-16.
- Karikari, S. K., Sebolai, B. and Munthali, D. C. 1996. Field studies of Bambara groundnut in Botswana. In: Proceedings of the International Bambara Groundnut Symposium, held at the University of Nottingham, UK, 23-25 July, 1998, Azam-Ali, S. N., Sesay, A. and Collinson, S. T. (Eds.), pp. 72-84.
- Linnemann, A.R.1991. Preliminary observations on photoperiod regulation of phenological development in Bambara groundnut (Vigna subterranea). Field Crops Research 26: 295-304.
- Linnemann, A. R. and Azam-Ali, S. N. 1991 Bambara groundnut (Vigna subterranea (L) Verdc). In: Underutilised Crop Series. No. 2. Vegetables and Pulses. Williams, J. T. (Ed.), pp. 1-58. Chapman and Hall, London.
- Mazhani, L. M. and Appa-Rao, S. 1985. Groundnut, Bambara groundnut and finger millet: Germplasm collection in Botswana. Crops Research Bulletin No. 3, Department of Agricultural Research, Government of Botswana. 6 pp.
- Ministry of Agriculture, Government of Botswana, 1947. Jugo Beans. In: Review of Crop Experiments in the Bechunaland Protectorate. Department of Agricultural Research, Botswana. pp. 127-128.
- Moll, R. H., Robinson, F. H. and Cockerham, C. C. 1960. Genetic variability in advanced generation of a cross of two open-pollinated varieties of corn. Agronomy Journal 52: 171-173.
- Nishitani, T., Muraki, K. and Inoue, J. 1988. Effect of daylength on the flowering and fruiting in Bambara groundnut (Vigna subterranea (L.) Verdc). Japanese Journal of Tropical Agriculture 32:80-84.
- SAS Institute 1990. SAS/STAT Users Guide. 1990 Edition, SAS Institute, Inc. Cary NC
- Sokal, R. R. and Rohlf, F. J. 1969. Biometry. Principles and Practices of Statistics in Biological Research. W. H. Freeman and Co., San Francisco. 320 pp.
- Steel, R. G. D. and Torrie, J. 1986. Principles and Procedures of Statistics. A Biometrical Approach. McGraw-Hill Book Co., Auckland, 633 pp.
- Tanimu, B. S. and Aliyu, L. 1990. Genotypic variability in Bambara groundnut cultivars at Samaru, Nigeria. In: Genetics, Man and the Environment. Proceedings of the 17th Annual Conference of the Genetics Society of Nigeria, held at Obafemi Awolowo University, Nigeria, 11-16 March, 1990. Pages 54-60, Genetics Society of Nigeria.
- Uguru, M. I. 1995. Heritable relationships and variability of yield and yield related components in cowpea. African Crop Science Journal 3:23-28.
- Wigglesworth, D. J. 1996. The potential for genetic improvement of Bambara groundnut (Vigna subterranea (L.) Verdc) in Botswana. In: Proceedings of the International Bambara Groundnut Symposium held at the University of Nottingham, UK, 23-25 July, 1998, Azam-Ali, S. N., Sesay, A. and Collinson, S. T. (Eds.), pp. 182-191.
|