Published online 1 March 2007
Published in Crop Sci 47:737-748 (2007)
© 2007 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
CROP ECOLOGY, MANAGEMENT & QUALITY
Agronomic Performance of Striga Resistant Early-Maturing Maize Varieties and Inbred Lines in the Savannas of West and Central Africa
Baffour Badu-Apraku* and
A. Fontem Lum
International Institute of Tropical Agriculture (IITA), c/o Lambourn (UK), Limited, Carolyn House, 26 Dingwall Rd., Croydon CR9 3EE, UK
* Corresponding author (b.badu-apraku{at}cgiar.org).
 |
ABSTRACT
|
|---|
The potential maize (Zea mays L.) yield in the savannas of West and Central Africa (WCA) is constrained by Striga hermonthica (Del.) Benth. parasitism. Field trials at Ferkessedougou, Côte d'Ivoire; Abuja, Nigeria; and Mokwa, Nigeria, in 2002, and at Mokwa and Abuja in 2004, evaluated the performance of 11 early maturing maize varieties under artificial Striga infestation and Striga-free conditions. Another trial at Mokwa and Abuja in 2004 evaluated 100 early maturing inbred lines under Striga infestation and Striga-free conditions. The varieties differed significantly in grain yield under both conditions. Acr 94 TZE Comp 5-W, Acr TZE Comp 5-Y, and TZE-W Pop x 1368 STR C1 were identified as promising varieties, based on grain yield, Striga damage ratings, and Striga emergence counts. The most promising variety, Acr 94 TZE Comp. 5-W in terms of high grain yield, reduced Striga damage and low Striga emergence, outyielded the reference entry by 2% under Striga-free conditions and 45% under Striga infestation. Ward cluster analysis of the varieties and inbred lines produced four major clusters each, under both Striga-infested and Striga-free conditions. In each case, the varieties and inbred lines assigned to each cluster under Striga infestation differed from those in the Striga-free conditions. Several inbred lines with high grain yield, low Striga emergence and reduced Striga damage were identified as sources of Striga resistance for maize breeding programs in WCA.
Abbreviations: ASI, anthesis-silking interval • EPP, number of ears per plant • GEI, genotype x environment interaction • IITA, International Institute of Tropical Agriculture • WAP, weeks after planting • WCA, West and Central Africa.
|
INTRODUCTION
|
|---|
THE SAVANNA ecology of sub-Saharan Africa has a high potential for maize production and productivity due to high solar radiation and low night temperatures (Kassam et al., 1975). As a result, maize has been adopted extensively in the zone, replacing the traditional crops, sorghum [Sorghum bicolor (L.) Moench] and millet [Pennisetum glaucum (L.) R. Br.] (Fakorede et al., 2003).
A major constraint to attaining the potential maize yield in the savannas is S. hermonthica parasitism. The levels of Striga infestation are often so high that maize can suffer total yield loss and farmers may be compelled to abandon their fields. In an effort to combat this threat, several high yielding, early maturing Striga resistant/tolerant populations, over 100 advanced early maturing Striga resistant/tolerant inbred lines, and several Striga resistant/tolerant early maturing varieties of various grain types and texture have been developed through the collaborative efforts of the International Institute of Tropical Agriculture (IITA)-West and Central Africa Collaborative Maize Research Network (WECAMAN) and the national agricultural systems (NARS) of WCA. In this paper, tolerance is used to denote the ability of the host plant to withstand the effects of the parasitic plants that are already attached. Resistance refers to the ability of the host plant to prevent the parasite from attaching itself to its roots (Kim, 1994).
Phenotypic classification of inbred lines is invaluable in the selection of parental inbreds for the development of heterotic populations and the introgression of desirable genes from diverse germplasm sources into the available genetic base (Thompson et al., 1998). The most commonly used approach for studying genetic differences between different germplasm and for placing inbred lines in heterotic groups is the pedigree method. However, several other approaches including molecular markers and multivariate methods are available. Among the available multivariate methods, cluster analysis, principal component analyses (PCAs), principal coordinate analyses (PCoA), and multidimensional scaling (MDS) are particularly useful and very common (Melchinger, 1993; Johns et al., 1997; Thompson et al., 1998; Brown-Guedira et al., 2000). Badu-Apraku et al. (2005) used the PCA and Ward's minimum variance cluster analysis to study the genetic diversity in 65 extra-early maturing maize inbred lines under Striga-infested and Striga-free conditions. Based on the preliminary grouping of the inbreds, it was reported that the selection of specific combinations of inbred lines for the development of hybrids, or synthetics, or for introgression into breeding populations may best be done by selecting parents in each cluster that combined high grain yield with reduced Striga damage symptoms and numbers of emerged Striga plants, as well as high numbers of ears per plant. The clustering technique is also very useful for the study of effects of pedigree and the origin of genotypes on their phenotypic behavior in various environments (Shorter et al., 1977).
In spite of the good progress that has been made in the identification and formation of new heterotic groups for intermediate and late maturing maize in the tropics (Vasal et al., 1992a, 1992b; Kim et al., 1999; Menkir et al., 2003), information is completely lacking on the heterotic patterns and the extent of diversity in the Striga resistant early maturing germplasm. Also, studies have shown that the heterotic patterns of inbred lines and populations are influenced by the environment under which they are evaluated (Gutierrez-Gaitan et al., 1986; Kim and Ajala, 1996; Vasal et al., 1993; Menkir et al., 2003). Therefore, information on the heterotic patterns of inbred lines in stressed and nonstressed environments could be very useful for the development of an efficient breeding strategy for the growing environments in WCA. Furthermore, knowledge and understanding of genetic relationships among the early maturing inbred lines would be very useful in identifying parental lines for making crosses, designing mating schemes for the field testing of heterotic patterns of the inbred lines, assigning them to specific heterotic groups, and making precise identification with respect to plant varietal protection (Hallauer and Miranda, 1988).
The objectives of this study were to: (i) evaluate the performance of early maturing Striga resistant maize varieties and inbred lines under artificial S. hermonthica infestation and noninfested conditions, (ii) determine the usefulness of the inbred lines as parents of open-pollinated synthetic varieties, as well as sources of resistance for the national maize breeding programs, (iii) use Ward's minimum variance cluster analysis for preliminary grouping of the lines for use in designing mating schemes to facilitate field testing of their heterotic patterns, and (iv) compare groups of the early inbred lines identified through cluster analysis under Striga-infested and Striga-free conditions.
|
MATERIALS AND METHODS
|
|---|
Experimental Protocol
Two field studies were conducted between 2002 and 2004 to evaluate the performance of early maturing Striga resistant inbred lines and varieties under Striga-infested and noninfested conditions. In the first study, 11 early maturing elite Striga resistant varieties (Table 1) were evaluated in the Regional Striga Trial–early (RSVT-early) at Ferkessedougou (9°30' N, 5°10' W; 325 m altitude), Côte d'Ivoire, in 2002, and at Mokwa (9°18' N, 5°04' E and elevation 457 m) and Abuja (9°16' N, 7°20' E and elevation 300 m), Nigeria, in 2002 and 2004. A randomized complete block design with four replications was used in each trial. Planting distance was 0.75 m between rows 5 m long and 0.4 m between hills, giving a plant population density of 66000 plants ha–1. Each plot consisted of two rows. Two of the four replications of the trial were artificially infested with Striga using a metal scoop which delivered about 5000 germinable S. hermonthica seeds per hole at planting.
View this table:
[in this window]
[in a new window]
|
Table 1. Characteristics of early-maturing maize varieties evaluated under Striga-infested and Striga-free conditions at Ferkessedougou, Côte d'Ivoire, in 2002 and at Mokwa and Abuja, Nigeria, in 2002 and 2004.
|
|
In the second study, 100 early maturing Striga resistant inbred lines developed from five diverse germplasm sources with tolerance or resistance to Striga and Maize streak virus, and/or tolerance to drought (TZE-W Pop DT STR C0, WEC STR, TZE-Y Pop DT STR C0, TZE Comp 5-Y C6, and TZE-W Pop x 1368 STR C0), were evaluated under Striga-infested and Striga-free conditions at Mokwa and Abuja in 2004 to determine the usefulness of the inbred lines as parents of open-pollinated synthetic varieties, as well as sources of resistance for the national maize breeding programs. A 10 by 10 lattice design with two replications was used for the trial. There were two rows per plot and a row of each plot was infested artificially with about 5000 germinable S. hermonthica seeds per hole at planting.
Artificial Striga Infestation and Field Management
The Striga infestation method developed by IITA Maize Program that ensures uniform Striga infestation with no escapes (Kim, 1991; Kim and Winslow, 1991) was used for the two studies. Striga hermonthica seed collected from maize and sorghum fields and stored for at least 6 mo were used for the artificial field infestations. To stimulate suicidal germination of existing Striga seeds in the field, ethylene gas was applied to the soil before artificial Striga infestation in Mokwa and Abuja. Apart from the Striga seed infestation, management practices were the same for both Striga-infested and noninfested plots. Three maize seeds were planted per hill. The plants were thinned to two per stand about 2 wk after emergence to give a final population density of 66000 plants ha–1. Fertilization of the artificially infested maize field was delayed until about 30 d after planting. At this stage of plant growth, about 30 kg N ha–1, 26 kg P ha–1, and 50 kg K ha–1 were applied as 15–15–15 N–P–K. Weeds other than Striga were manually controlled.
Collection of Agronomic Data
Data were collected on grain yield, number of ears and plants harvested, plant height, ear aspect, anthesis-silking interval (ASI), stalk lodging, and days to 50% anthesis and silking in both infested and noninfested plots. In addition, host plant damage syndrome rating (Kim, 1991) and emerged Striga counts were made at 8 and 10 wk after planting (WAP) in the Striga-infested rows. Striga damage syndrome rating was recorded on a scale of 1 to 9 where 1 = no damage, indicating normal plant growth and high resistance, and 9 = complete collapse or death of maize, indicating high level of susceptibility. Anthesis-silking interval was determined as the difference between 50% silking and anthesis. Number of ears per plant (EPP) was determined by dividing the total number of ears per plot by the number of plants harvested. Grain yield was calculated from the shelled kernel dry weight and adjusted to 15% moisture. Even though data were collected on several traits, only those on the most important traits in the studies are presented in the results.
Statistical Analysis
Combined analyses of variance were conducted for grain yield and other genotypic variables using the General Linear Model Procedure (GLM) of the statistical analysis systems (SAS) package (SAS, 1990). The variance of Striga counts has been found to increase with the mean, therefore a log transformation [log (counts + 1)] was used to remove the heterogeneity of variance. Data for seven of the inbred lines were not usable and were therefore discarded.
Principal component analyses were performed on the average values for each trait to identify the group of traits that accounted for most of the variance in the data set and could therefore be used to rank the Striga resistant varieties and inbred lines for their performance under Striga-infested and Striga-free conditions. This was achieved using PRINCOMP procedures of the SAS package (SAS, 1985). A simple phenotypic correlation analysis between the principal component (PC) scores and each trait was performed to determine the contribution of each trait to the principal component axis. Traits that did not have a significant correlation with PC scores were discarded. Clustering of the varieties and the inbred lines was performed separately by subjecting standardized genotypic variables to the Ward's minimum variance method (Ward, 1963) available in the CLUSTER procedure of SAS statistical package (SAS, 1985). Ward's method is particularly noted for its effectiveness in maximizing the variance among groups while minimizing the variance within groups.
|
RESULTS
|
|---|
Performance of Early Maturing Varieties under Striga-Infested and Striga-Free Conditions
Analyses of variance of data combined across locations and years for the RSVT-early revealed significant differences among the genotypes for all traits measured under Striga-infested and Striga-free conditions, except plant height and EPP under Striga-free conditions (Table 2). The genotype x environment interactions (GEI) were not significant for the traits measured except Striga emergence count at 8 and 10 WAP under Striga infestation and EPP under Striga-free conditions. The other components of variation (location, year, and location x year) had significant mean squares for most traits under Striga-infested and Striga-free conditions. The results indicated that the outstanding varieties in terms of grain yield, Striga emergence counts, and Striga damage ratings were Acr 94 TZE Comp. 5-W, Acr 94 TZE Comp. 5-Y, and TZE W-Pop x 1368 STR C1. The most promising variety in terms of grain yield and with the lowest Striga damage rating and emergence count, Acr 94 TZE Comp. 5-W, outyielded the reference entry, Kamboinse 88 Pool 16DT, by 45% under Striga infestation and 2% under Striga-free conditions. The Striga emergence counts were generally high, ranging from 172 to 332 plants plot–1 at 8 WAP and 203 to 366 plants plot–1 at 10 WAP. The high Striga emergence counts indicate that the varieties are tolerant to Striga.
View this table:
[in this window]
[in a new window]
|
Table 2. Performance of early-maturing varieties under Striga-infested and Striga-free conditions at Ferkessedougou, Mokwa, and Abuja in 2002 and 2004.
|
|
View this table:
[in this window]
[in a new window]
|
Table 3. Grain yield and other agronomic traits of 93 early-maturing inbred lines evaluated under Striga-infested and Striga-free conditions at Abuja and Mokwa in 2004.
|
|
The large genotypic variation in the phenotypic traits of the Striga resistant varieties under Striga-infested and noninfested conditions prompted the use of multivariate analysis to identify the major genotypic patterns. The first three PCs with eigenvectors 
0.3 summarized 86% of the multivariate variation in variety means under Striga infestation and 75% under Striga-free conditions (data not shown) and were the most important. The traits loaded on the three axes were used for the grouping of the early varieties. The early varieties evaluated under Striga infestation were clustered using grain yield, days to 50% silking, days to anthesis, ASI, plant height, EPP, ear aspect, Striga emergence counts, and Striga damage rating. On the other hand, the early varieties evaluated under Striga-free conditions were grouped based on grain yield, days to 50% silking, days to anthesis, ASI, plant height, stalk lodging, EPP, and ear aspect. At a normalized distance of 0.1, dendrograms with four major clusters were produced for the early varieties under Striga infestation and Striga-free conditions using Ward's minimum variance cluster analysis (Fig. 1
and 2
). Under Striga infestation, Group I contained two varieties (AC2 and AC3); Group II, one variety (EW2); Group III, five varieties (WC1, E97, EVY, WC2, and TZE), and Group IV, three varieties (EW1, AC1, and KAMB). The only variety placed in Group II, EW2, was the lowest yielding variety under Striga-free conditions. It was rated moderately susceptible under Striga infestation but surprisingly, it had the lowest Striga emergence counts (172 plants plot–1 at 8 WAP and 203 plants plot–1 at 10 WAP). Group III contained the second highest yielding variety, TZE, and the lowest yielding variety, under Striga infestation, WC1. It is surprising to note that even though WC1 was rated tolerant (Striga damage ratings of 5 and Striga emergence counts of 322 plants plot–1 at 8 WAP and 362 plants plot–1 at 10 WAP), this was not translated into high grain yield. The three varieties in Group IV were low yielding with the group's average grain yield far below the trial mean. The varieties also had high Striga damage rating (score of 5 at 8 WAP and 6 at 10 WAP). The number and type of varieties within the four clusters under Striga-free conditions were different from those in the corresponding clusters generated under Striga infestation (Fig. 1 and 2). Under Striga-free conditions, Group I comprised TZE, AC3, and EW1, which had moderate to high grain yield. As observed under Striga infestation, Group II contained only one variety, EW2, which was the lowest yielding under Striga-free conditions. The four varieties in Group III had moderate to high grain yield. The highest yielding variety under Striga-free conditions, WC2, was found in this group. Group IV consisted of E97, WC1, and AC1. The average grain yield for the varieties in this group was higher than the mean yield of the trial.
View larger version (8K):
[in this window]
[in a new window]
|
Figure 1. Dendrogram of 11 early-maturing maize varieties derived from quantitative traits using the first four principal components and the Ward's minimum variance cluster analysis under Striga infestation at Abuja and Mokwa in 2004. See Table 2 for codes for the varieties.
|
|
View larger version (9K):
[in this window]
[in a new window]
|
Figure 2. Dendrogram of 11 early maturing maize varieties derived from quantitative traits using the first four principal components and the Ward's minimum variance cluster analysis under Striga-free conditions at Abuja and Mokwa in 2004. See Table 2 for codes for the varieties.
|
|
Performance of Early Maturing Inbred Lines under Striga-Infested and Striga- Free Conditions
Significant differences were detected among the inbred lines for all traits except ASI and stalk lodging under Striga-free conditions (Table 2). The location mean squares were also significant for most traits measured under Striga-infested and Striga-free conditions. Significant location x inbred interactions were detected for only grain yield, days to anthesis, days to silking, plant height, Striga emergence count at 8 and 10 WAP, Striga damage rating at 8 WAP, and ASI under Striga infestation. On the other hand, significant location x inbred interactions occurred only for grain yield, days to anthesis, and days to silking under Striga-free conditions. Grain yield ranged from 217 Mg ha–1 for TWX12 to 1799 Mg ha–1 for TYC01 under Striga-infested conditions while under Striga-free conditions, it ranged from 463 Mg ha–1 for LD01 to 3322 Mg ha–1 for TW06. The outstanding inbred lines based on high grain yield, low Striga damage ratings, and reduced Striga emergence at 8 and 10 WAP were TWX02, TWX04, TWX07, TY16, TYC01, TYC03, TYC04, TYC05, and TYC06. These inbred lines may be classified as S. hermonthica resistant. On the other hand, the lines TW02, TW10, TW12, TW14, TW16, TWX03, TWX18, TW20,TY14, TY21, TYC07, and TY27 had high grain yield, low Striga damage scores, but high Striga emergence, and may be described as tolerant to S. hermonthica.
Significant positive phenotypic correlations were obtained between grain yield and plant height (ra = 0.31), and EPP (ra = 0.51), while significant negative phenotypic correlations existed between grain yield and days to anthesis (rp = –0.25), days to silking (rp = –0.28), Striga damage ratings at 8 WAP (rp = –0.56) and 10 WAP (rp = –0.54), Striga emergence counts at 8 WAP (rp = –0.26) and 10 WAP (rp = –0.24), and ear aspect (rp = –0.66) under Striga infestation (Table 4). The phenotypic correlations between grain yield and stalk lodging and ASI were not significant under Striga infestation. Apart from days to anthesis, EPP was significantly correlated with all other traits under Striga-infested conditions. Strong phenotypic correlations existed between Striga damage ratings at 8 and 10 WAP (rp = 0.74); Striga emergence counts at 8 and 10 WAP (rp = 0.97). However, there was low correlation between Striga damage ratings at 8 WAP and Striga emergence counts at 8 WAP (rp = 0.24) and 10 WAP (rp = 0.18); Striga damage ratings at 10 WAP and Striga emergence counts at 8 WAP (rp = 0.19) and 10 WAP (rp = 0.14). Under Striga-free conditions, strong positive correlations were found between grain yield and plant height (rp = 0.36) and EPP (rp = 0.35), while significant negative correlations were detected between grain yield and days to anthesis (rp = –0.23), days to silking (rp = –0.35), ASI (rp = –0.21), stalk lodging (rp = –0.20), and ear aspect (rp = –0.66) (Table 4).
View this table:
[in this window]
[in a new window]
|
Table 4. Correlation coefficients between grain yield and agronomic traits of early-maturing inbred lines evaluated under Striga-infested (upper diagonal) and Striga-free (below diagonal) conditions at Mokwa and Abuja in 2004.
|
|
Grouping of Early Maturing Inbred Lines under Striga-Infested and Striga- Free Conditions
Following the restriction that only eigenvectors equal to or greater than 0.3 made a significant contribution to the PC axis, the variables were grouped into four PCs each, under Striga-infested and Striga-free conditions (Tables 5 and 6). The PCs accounted for 77% each of the total genetic variation among the 93 inbreds evaluated under Striga infestation (PC1 = 0.32, PC2 = 0.24, PC3 = 0.13, and PC4 = 0.09) and under Striga-free conditions (PC1 = 0.33, PC2 = 0.19, PC3 = 0.14, and PC4 = 0.12). Under Striga-infested conditions, grain yield, Striga damage ratings at 8 and 10 WAP, EPP, and ear aspect were loaded on PC1. Striga emergence counts at 8 and 10 WAP had their highest weightings assigned to PC2 and were considered interrelated. Days to silking was loaded on PC3 while days to anthesis, ASI, and plant height, were loaded on PC4. Percentage of stalk lodging was not loaded on any PC under Striga infestation. Under Striga-free conditions, grain yield and ear aspect were loaded on PC1 and were considered interrelated. Days to silking and ASI were loaded on PC2. Days to anthesis and plant height had their highest weightings on PC3. EPP and percentage of stalk lodging were assigned the highest weighting on PC4. The traits loaded on the four axes were employed to group the inbred lines into closely related clusters using the Ward's minimum variance cluster method. Four major clusters were identified under both Striga-infested (Fig. 3
) and Striga-free conditions (Fig. 4
). Under Striga infestation, cluster I was represented by 28 inbred lines of white and yellow endosperm color derived from three germplasm sources (TZE-W Pop, WEC STR, and TZE-Y Pop). The grain yield of the inbred lines in this cluster ranged from 432 Mg ha–1 for TW01 to 1598 Mg ha–1 for TW12 and they had a mean grain yield (844 Mg ha–1) about the same as the trial mean. The Striga emergence counts and Striga damage scores of the inbred lines in the cluster generally had wide ranges and large variances and a cluster mean greater than the trial mean (842 Mg ha–1). Two of the inbred lines, TW12 and TW20, were outstanding in terms of grain yield but had high Striga emergence and sustained moderately high Striga damage suggesting that they are tolerant to Striga. Cluster II contained 12 inbred lines of white and yellow endosperm color from four genetic backgrounds, TZE-W Pop, TZE-W Pop x 1368 STR, WEC STR, and TZE-Y Pop. The inbred lines were low yielding, generally sustained moderately high Striga damage, and had moderately high Striga emergence. Cluster III comprised 21 inbred lines of white and yellow grain color and were extracted from the four germplasm sources, TZE-W Pop C0, TZE-W Pop x 1368 STR, TZE-Y Pop C0, and TZE-Comp5-Y C6. The inbred lines in this cluster had high mean grain yield, a large cluster variance, and a wide range. The outstanding inbred lines in the study in terms of high grain yield, low Striga emergence and reduced Striga damage were classified into this cluster. They were TWO2, TWX02, TWX04, TWX18,TY14, TY16, TY21, TYC01, TYC03, TYC04, TYC05, TYC06, and TY27. Cluster IV was composed of 32 inbred lines extracted from TZE-W Pop C0, TZE-W Pop x 1368 STR, WEC STR, TZE-W Pop x LD, TZE-Y Pop C0, and TZE-Comp5-Y C6. These inbred lines had moderately high grain yield with a wide range and a large variance. The lines TWX07 and TWX08 were outstanding in terms of grain yield, Striga emergence, and Striga damage rating. Several lines, such as TWX13, TWX16, TY05, and TY13, had low Striga emergence and sustained low Striga damage but this did not result in high grain yield.
View this table:
[in this window]
[in a new window]
|
Table 5. Eigenvectors of the first four principal component axes (PC1, PC2, PC3, and PC4) of 93 early-maturing maize inbred lines evaluated under Striga-infested conditions at Mokwa and Abuja in 2004. Only eigenvectors with values equal to or higher than 0.3 are shown.
|
|
View this table:
[in this window]
[in a new window]
|
Table 6. Eigenvectors of the first four principal component axes (PC1, PC2, PC3, and PC4) of 93 early-maturing maize inbred lines evaluated under Striga-free conditions at Mokwa and Abuja in 2004. Only eigenvectors with values equal to or higher than 0.3 are shown.
|
|
View larger version (15K):
[in this window]
[in a new window]
|
Figure 3. Dendrogram of 93 early maturing maize inbred lines derived from quantitative traits using the first four principal components and the Ward's minimum variance cluster analysis under Striga infestation at Abuja and Mokwa in 2004. See Table 3 for codes for the lines.
|
|
View larger version (14K):
[in this window]
[in a new window]
|
Figure 4. Dendrogram of 93 early-maturing maize inbred lines derived from quantitative traits using the first four principal components and the Ward's minimum variance cluster analysis under Striga-free conditions at Abuja and Mokwa in 2004. See Table 3 for codes for the lines.
|
|
Under Striga-free conditions, 48 lines were grouped in cluster I. Grain yield of the lines in this cluster ranged from 1293 Mg ha–1 for TY15 to 3322 Mg ha–1 for TW06 with a mean of 2040 Mg ha–1 and a large variance (208 209.82). The lines in the cluster were developed from five germplasm sources (TZE-W Pop C0, TZE-W Pop x 1368 STR, WEC STR, TZE-Y Pop C0, and TZE-Comp5-Y C6) and had white and yellow endosperm colors. The outstanding lines in terms of grain yield in the study came from this cluster and were TY20,WC04, TWX23, TW12, TWO6, and TWO2. On the other hand, the cluster also contained lines such as TWX05, TWX14, TWX22, TW23, WC11, WC14, TY15, and TY16 with grain yield below the trial mean. Cluster II contained only one inbred, LD01 with low grain yield (463 Mg ha–1) and long days to anthesis. The 34 inbred lines grouped in cluster III were derived from five source populations, TZE-W Pop C0, TZE-W Pop x 1368 STR, WEC STR, TZE-Y Pop C0 and TZE-Comp5-Y C6. The inbred lines in this cluster had mean grain yield of 1422 Mg ha–1 with a range of 829 Mg ha–1 for TY11 and 2267 Mg ha–1 for TYC05. The 10 inbred lines in cluster IV had low to moderate grain yield, low percentage stalk lodging, desirable plant height, and fitted well in the early maturity group. The variance in grain yield for the inbreds within this cluster was quite large (84006).
Examination of the corresponding clusters of the lines under Striga-infested and Striga-free conditions revealed that only 23 of the lines in cluster I were common, while five were common in cluster II and six in cluster IV (Table 7). No lines were common in cluster IV under Striga-infested and Striga-free conditions. Furthermore, even though the lines were assigned to four major clusters under infested and Striga-free conditions, the inbred lines within each cluster under each environment were different.
View this table:
[in this window]
[in a new window]
|
Table 7. Composition of different clusters, means, ranges, and variances of clusters derived from quantitative characters of 93 early maturing maize inbred lines evaluated under Striga-infested and Striga-free conditions at Mokwa and Abuja, 2004.
|
|
|
DISCUSSION
|
|---|
One of the primary objectives of this study was to identify high yielding Striga resistant or tolerant varieties for production in the Striga endemic savannas of WCA. The results showed that the outstanding varieties in terms of grain yield, Striga emergence counts and Striga damage ratings were Acr 94 TZE Comp. 5-W, Acr 94 TZE Comp. 5-Y, and TZE W-Pop x1368 STR C1. The superior performance of the three Striga tolerant varieties has been confirmed in on-farm trials conducted under artificial Striga infestation in the savanna ecologies of Nigeria, Côte d'Ivoire, and Mali. For example, 10 on-farm trials were conducted in 2005 in the northern Guinea savanna of Nigeria to evaluate the reaction of the two Striga tolerant varieties, Acr 94 TZE Comp. 5-W, EV DT 99 STR, and the farmers's variety to S. hermonthica infestation. Results showed that the farmer's variety produced fewer plants per hectare, and lower grain yield than the Striga tolerant varieties. Acr 94 TZE Comp. 5-W and EV DT 99 STR supported fewer emerged Striga plants and were less damaged than the farmer's variety. The grain yield of Acr 94 TZE Comp. 5-W was 14% higher than the farmers' variety, and EV DT 99 STR was 17% higher (WECAMAN, unpublished data, 2006). Based on the results of these studies, Acr 94 TZE Comp. 5-W, 2000 Syn WEC, and EV DT 99 STR are being promoted through farmer participatory on-farm trials for adoption by NARS of WCA.
The high Striga emergence counts recorded for Acr 94 TZE Comp. 5-W, Acr 94 TZE Comp. 5-Y, TZE W-Pop x 1368 STR C1, and the other varieties in this study indicates that they are tolerant to S. hermonthica (Kim, 1994). This suggests the need to adopt integrated approach involving rotation of the available Striga tolerant varieties with soybean [Glycine max (L.) Merr.], cowpea [Vigna unguiculata (L.) Walp.], and groundnut (Arachis hypogaea L.) varieties that can cause suicidal germination of Striga seed for the control of Striga in WCA (Kling et al., 1997; Badu-Apraku et al., 2005).
During the early stages of the IITA-WECAMAN breeding program, the emphasis was on the selection for tolerance, which allowed the pathogen to reproduce, thus increasing the buildup of the Striga seedbank in the soil at the end of each season. Since 1999, the emphasis of the breeding program has been on the selection for reduced Striga emergence, resulting in the development of the inbred lines such as TWX02, TWX04, TWX07, TY16, TYC01, TYC03, TYC04, TYC05, and TYC06 identified in this study which have reduced Striga emergence counts, sustain low Striga damage, and have high grain yield under Striga infestation. These inbred lines could serve as invaluable sources of Striga resistance for the development of synthetic varieties and for use by NARS in their breeding programs. Similar findings have been reported by Badu-Apraku et al. (2005). The use of these sources of high Striga resistance will lead to drastic reduction in the reproduction of the parasite thereby depleting or lowering the levels of the Striga seed bank in soils of the Striga endemic areas of the savanna ecology.
To ensure that Striga tolerant materials from the IITA-WECAMAN maize program have resistance to potentially different biotypes of S. hermonthica, Ferkessedougou in Côte d'Ivoire and Mokwa and Abuja in Nigeria have been used in our screening and evaluation programs. The lack of significant GEI for grain yield and most other traits of the early maturing varieties under Striga-infested and Striga-free conditions suggests that there are no different biotypes of S. hermonthica at the different locations. The implication is that the tolerance of the available varieties will hold in all the environments used in the evaluations.
The strong positive phenotypic correlations between Striga emergence at 8 and 10 WAP suggests that either of this parameters could be effectively used as a selection parameter for the evaluation of the lines for Striga resistance or tolerance. The moderate negative phenotypic correlations between grain yield and Striga damage scores at 8 and 10 WAP confirms that the the traits are fairly reliable for selection for improved grain yield and Striga resistance. This finding is in agreement with the results of Kim (1991, 1994) and Badu-Apraku et al. (2004). On the other hand, the low and nonsignificant correlations between grain yield and Striga emergence count at 8 and 10 WAP corroborates the findings of several other workers (Kim, 1991; Akanvou et al., 1997; Badu-Apraku et al., 2004) that the Striga emergence counts is not an effective selection criterion for improvement of Striga resistance.
Ward's minimum variance cluster method was employed for the preliminary classification of the Striga resistant early maturing inbred lines into heterotic groups. It is interesting to note that inbred lines derived from different genetic backgrounds and of different grain color were grouped in the same cluster. For example, 21 inbred lines of white and yellow grain color extracted from the different four germplasm sources, TZE-W Pop C0, TZE-W Pop x 1368 STR, TZE-Y Pop C0, and TZE-Comp5-Y C6, were placed in cluster III under Striga infestation. Similarly, under Striga-free conditions the 34 inbred lines grouped in cluster III were derived from five different germplasm sources, TZE-W Pop C0, TZE-W Pop x 1368 STR, WEC STR, TZE-Y Pop C0, and TZE-Comp5-Y C6. The implication is that the clustering of the inbred lines was independent of the genetic backgrounds and grain color, and that they may represent several heterotic groups that could be exploited for the development of high yielding Striga resistant populations and synthetic varieties. Similar findings were reported by Badu-Apraku et al. (2005).
The results of this study have shown that the most appropriate traits for grouping the early lines under Striga infestation are grain yield, days to anthesis, days to silking, plant height, Striga emergence count at 8 and 10 WAP, Striga damage rating at 8 and 10 WAP, ASI, EPP, and ear aspect. Under Striga-free conditions, the most important variables were grain yield, ear aspect, days to silking, ASI, days to anthesis, plant height, EPP, and percentage of stalk lodging.
An important objective of this study was to use the results of the cluster analysis of the inbred lines to develop guidelines for the management of diversity in the inbred lines in the IITA-WECAMAN breeding program. Based on the results of this study, preliminary grouping of the lines has been done and mating schemes have been designed for field testing of the heterotic patterns to allow efficient exploitation of the diverse early germplasm in our program. So far, diallel crosses have been made between representative samples of lines from each of the four clusters. The crosses derived from the clusters are presently being evaluated under artificial Striga infestation in the field to determine the general and specific combining ability of the lines and thus confirm the heterotic groupings.
Received for publication April 17, 2006.
|
REFERENCES
|
|---|
- Akanvou, L., E.V. Doku, and J. Kling. 1997. Estimates of genetic variances and interrelationships of traits associated with Striga resistance in maize. Afri. Crop Sci. J. 5:1–8.
- Badu-Apraku, B., M.A.B. Fakorede, A. Menkir, A.Y. Kamara, L. Akanvou, and Y. Chabi. 2004. Response of early maturing maize to multiple-stresses in the Guinea savanna of West and Central Africa. J. Genet. Breed. 58:119–130.
- Badu-Apraku, B., A. Menkir, and A.F. Lum. 2005. Assessment of genetic diversity in extra-early Striga resistant tropical inbred lines using multivariate analyses of agronomic data. J. Genet. Breed. 59:67–80.
- Brown-Guedira, G.L., J.A. Thompson, R.L. Nelson, and M.L. Warburton. 2000. Evaluation of genetic diversity of soybean introductions and North American ancestors using RAPD and SSR markers. Crop Sci. 40:815–823.[Abstract/Free Full Text]
- Fakorede, M.A.B., B. Badu-Apraku, A.Y. Kamara, A. Menkir, and S.O. Ajala. 2003. Maize revolution in West and Central Africa: An overview. p. 3–15. In B. Badu-Apraku et al. (ed.) Proc. of a regional maize workshop IITA Cotonou, Benin Republic. 14–18 May 2001. WECAMAN/IITA.
- Gutierrez-Gaitan, M.A., H. Cortez-Mendoza, E.N. Wathika, C.O. Gardner, M. Oyervides-Garcia, and A.R. Hallauer. 1986. Testcross evaluation of Mexican maize populations. Crop Sci. 26:99–104.[ISI]
- Hallauer, A.R., and J.B. Miranda. 1988. Quantitative genetics in maize breeding. 2nd ed. Iowa State Univ. Press, Ames, IA.
- Johns, M.A., P.W. Skrotch, J. Neinhuis, P. Hinrichsen, G. Bascur, and C. Munoz-Schick. 1997. Gene pool classification of common bean landraces from Chile based on RAPD and morphological data. Crop Sci. 37:605–613.[Abstract/Free Full Text]
- Kassam, A.H., J.M. Kowal, M. Dagg, and M.N. Harrison. 1975. Maize in West Africa: Its potential in the savanna area. World Crops 17:75–78.
- Kim, S.K. 1991. Breeding maize for Striga tolerance and the development of a field infestation technique. p. 96–108. In S.K. Kim (ed.) Combating Striga in Africa. Proc. Int. Workshop IITA, ICRISAT, and IDRC. 22–24 Aug. 1988. IITA, Ibadan, Nigeria.
- Kim, S.K. 1994. Genetics of maize tolerance of Striga hermonthica. Crop Sci. 34:900–907.[Abstract/Free Full Text]
- Kim, S.K., and S.O. Ajala. 1996. Combining ability of tropical maize germplasm in West Africa II. Tropical vs temperate x tropical origins. Maydica 41:135–141.[ISI]
- Kim, S.K., S.O. Ajala, and J.L. Brewbaker. 1999. Combining ability of tropical maize germplasm in West Africa III. Tropical maize inbreds. Maydica 44:285–291.[ISI]
- Kim, S.K., and M.D. Winslow. 1991. Progress in breeding maize for Striga tolerance/resistance at IITA. p. 494–499. In J.K. Ransom et al. (ed.) Proc. 5th Int. Symp. on Parasitic Weeds, Nairobi. 24–30 June 1991. IITA, Ibadan, Nigeria.
- Kling, J.G., D.K. Berner, and O.A. Ibikunle. 1997. Sources of resistance to Striga in tropical maize and teosinte. p. 71. Agronomy Abstracts. ASA, Madison, WI.
- Melchinger, A.E. 1993. Use of RFLP markers for analyses of genetic relationships among breeding materials and prediction of hybrid performance. p. 621–628. In D.R. Buxton (ed.) Proc. Int. Crop Sci. Congress. 14–22 July 1992. CSSA, Madison, WI.
- Menkir, A., B. Badu-Apraku, C. The, and A. Adepoju. 2003. Evaluation of heterotic patterns of IITA's lowland white maize inbred lines. Maydica 48:161–170.[ISI]
- SAS. 1985. SAS user's guide, statistics. 5th ed. SAS Institute Inc., Cary, NC.
- SAS. 1990. SAS/STAT user's guide. Version 6, 4th ed. SAS Institute Inc., Cary, NC.
- Shorter, R.D., D.E. Byth, and V.E. Muntgomery. 1977. Genotype x environment interactions and environmental adaptation. I. Pattern analysis–application to soybean populations. Aust. J. Agric. Res. 25:59–72.[Medline]
- Thompson, J.A., R.L. Nelson, and L.O. Vodkin. 1998. Identification of diverse soybean germplasm using RAPD markers. Crop Sci. 38:1348–1355.[Abstract/Free Full Text]
- Vasal, S.K., G. Srinivasan, F.C. Gonzalez, D.L. Beck, and J. Crossa. 1993. Heterosis and combining ability of CIMMYT's quality protein maize germplasm: II. Subtropical. Crop Sci. 33:51–57.[Abstract/Free Full Text]
- Vasal, S.K., G. Srinivasan, G.C. Han, and F.C. Gonzalez. 1992a. Heterotic patterns of eighty-eight white subtropical CIMMYT maize lines. Maydica 37:319–327.[ISI]
- Vasal, S.K., G. Srinivasan, S. Pandey, H.S. Cordova, G.C. Han, and F.C. Gonzalez. 1992b. Heterotic patterns of ninety-two white tropical CIMMYT maize lines. Maydica 37:259–270.[ISI]
- Ward, J.H., Jr. 1963. Hierarchical grouping to optimize an objective function. J. Am. Stat. Assoc. 58:236–244.[CrossRef][ISI]
TOP
ABSTRACT
INTRODUCTION
