Crop Science
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 1 March 2007
Published in Crop Sci 47:674-682 (2007)
© 2007 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Menkir, A.
Right arrow Articles by Kling, J. G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Menkir, A.
Right arrow Articles by Kling, J. G.
Agricola
Right arrow Articles by Menkir, A.
Right arrow Articles by Kling, J. G.
Related Collections
Right arrow Crop Genetics

CROP BREEDING & GENETICS

Response to Recurrent Selection for Resistance to Striga hermonthica (Del.) Benth in a Tropical Maize Population

A. Menkira,* and J. G. Klingb

a International Institute of Tropical Agriculture (IITA), c/o L.W. Lambourn Ltd., Carolyn House, 26 Dingwall Rd., Croydon CR9 3EE, UK
b Oregon State Univ., Crop Science Building 107, Corvallis, OR 97331-3002

* Corresponding author (a.menkir{at}cgiar.org).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Striga hermonthica (Del.) Benth is an obligate root parasite infecting maize (Zea mays L.) and causing considerable yield losses in Africa. Few studies have been conducted to determine the effectiveness of recurrent selection for improving resistance to S. hermonthica in maize. The objective of this study was to evaluate genetic gain achieved in a composite subjected to six cycles of selection under S. hermonthica infestation. The selection cycles and checks were evaluated with and without S. hermonthica infestation at two locations in Nigeria for 2 yr. Selection for improved performance under S. hermonthica infestation significantly increased grain yield by 24% cycle-1 and ears per plant by 9% cycle-1. At the same time, the gain per cycle was –7% for relative yield loss, –5% for host damage rating, –9% for emerged S. hermonthica plants, –4% for anthesis–silking interval, and –5% for ear aspect. Selection under S. hermonthica infestation was accompanied by a concomitant increase in grain yield and improvement in plant aspect and ear aspect without S. hermonthica infestation. The observed progress in performance under S. hermonthica infestation demonstrates the effectiveness of recurrent selection for increasing polygenetic resistance against the parasite in tropical populations.

Abbreviations: IITA, International Institute of Tropical Agriculture.


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
THE MOIST savannas of West and Central Africa are considered to be high potential zones for maize production due to favorable growing conditions and low incidence of pests and diseases. Striga is an endemic obligate root hemiparasite in the savannas and poses a serious threat to maize (Zea mays L.) production in this zone (Doggett, 1984; Carson, 1986; Lagoke et al., 1991). This parasite has devastating effects on grain yield of maize and other cereals by robbing its host of C, N, and inorganic salts (Gurney et al., 1999), while at the same time diminishing the growth and photosynthetic capacity of the host plant (Frost et al., 1997). According to Gressel et al. (2004), 64% of the total land area devoted to cereal production in West Africa is severely infested with Striga. Among the four economically important species of Striga that are endemic to the savannas of Africa, S. hermonthica (Del.) Benth is the most widespread and destructive species affecting cereals, with maize being the most susceptible (Lagoke et al., 1991; Ramaiah, 1991; Berner et al., 1995). Estimated yield losses due to Striga infestation vary widely depending on crop variety, level of infestation, rainfall pattern, and soil degradation (Elemo et al., 1995). Losses in grain yield of up to 100% have been recorded in maize trials artificially infested with S. hermonthica (Ransom et al., 1990; Kim et al., 2002).

Several options, including the use of hand pulling, appropriate rate of fertilizer, herbicides, rotation of legume trap crops with cereals, intercropping, biocontrol, and resistant maize varieties have been recommended for controlling Striga in farmers' fields (Lagoke et al., 1991; Berner et al., 1995). However, the diversity of the farming systems in Africa and that of the parasite have rendered the use of a single control method ineffective against Striga (Berner et al., 1995). A combination of some of these options would thus be more effective in controlling Striga on subsistence farmers' fields. Breeding cultivars that can withstand parasite infection is considered to be central to any integrated approach to control S. hermonthica in maize. Host resistance to S. hermonthica is simple and economical for subsistence farmers and can be applied over a broad range of environmental and socioeconomic conditions.

Maize cultivars that withstand parasite infection can be resistant to the parasite by diminishing its growth, development, and survival or tolerant to the effects of a large number of attached parasites to their roots (Ejeta et al., 1991; Kim, 1994; Lane et al., 1997). The earliest search for Striga resistant maize germplasm was not successful (Saunders, 1933) and led to a view that resistance genes were very rare in the maize genome (Ramaiah, 1987). The advent of effective and uniform infestation techniques for screening maize germplasm in the field and screenhouse (Kim, 1991) and extensive screening of a large number of maize genotypes have facilitated the identification of resistance to Striga among elite germplasm and landrace accessions (Ransom et al., 1990; Kim, 1991; Mumera and Below, 1996; Kling et al., 2000) as well as from wild species of maize (Lane et al., 1997; Kling et al., 2000; Gurney et al., 2003).

Some studies have been conducted to examine the mode of inheritance of traits associated with resistance to Striga in maize (Kim, 1994; Akanvou et al., 1997; Gethi and Smith, 2004). Field resistance to Striga in maize is considered a quantitatively inherited trait, with additive gene effects being more important in regulating host plant damage symptom rating and grain yield under infestation. On the other hand, nonadditive gene action played a prominent role in controlling the number of emerged S. hermonthica plants in diallel crosses among 10 maize inbred lines (Kim, 1994). In other studies conducted under artificial infestation with S. hermonthica and S. asiatica (L.) Kuntze, both additive and nonadditive gene effects were significant for host plant damage symptom rating, number of emerged Striga plants, and grain yield under infestation, with the former making the largest contribution to the inheritance of these traits (Gethi and Smith, 2004). Estimates of narrow-sense heritability in a tropical population were 0.33 for host plant damage rating, 0.32 for grain yield under infestation, and 0.14 for number of emerged Striga plants (Akanvou et al., 1997). It appears that genes that impart a reduced level of parasite infection are present at low initial frequencies in tropical populations (Kling et al., 2000).

Recurrent selection is effective for improving quantitative traits with low and intermediate heritability. Recurrent selection involves systematic testing and selection of desirable progeny derived from a population followed by recombination of the selected progeny to form an improved population. This type of selection method has been used successfully for increasing resistance in maize to diseases (Jenkins et al., 1954; Miles et al., 1980, 1981; Ceballos et al., 1991; Lambert and White, 1997) and European corn borers (Ostrinia nubilalis Hübner) (Tseng et al., 1984; Klenke et al., 1986). So far, the use of recurrent selection for improving resistance to parasitic plants in maize and other cereals has been limited. The International Institute of Tropical Agriculture (IITA) has conducted recurrent selection in several maize populations with diverse genetic backgrounds, maturities, and grain colors under artificial infestation with S. hermonthica (Kling et al., 2000). Selfed progeny and full-sib family selection schemes have been used to gradually and continually accumulate and increase the frequency of desirable alleles for traits associated with resistance to S. hermonthica as well as alleles for desirable agronomic traits in these populations. One of these populations, TZL COMP1-W, was subjected to six cycles of recurrent selection for resistance to S. hermonthica. This is the first known maize composite that has been subjected to a long-term recurrent selection to improve resistance to S. hermonthica. The composite, therefore, provides a unique opportunity to examine the effectiveness of recurrent selection for improving resistance to S. hermonthica in maize. This study was, therefore, conducted (i) to examine the response to six cycles of recurrent selection for traits associated with resistance to S. hermonthica and (ii) to determine if any concomitant changes in agronomic traits occurred in Striga noninfested environment as a result of the selection in the composite.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
A late-maturing (120 d from planting to physiological maturity) composite, TZL COMP1-W, was developed at IITA in 1988 by crossing a Maize streak virus–resistant population, TZB-SR, with seven Striga-resistant inbred lines (Tzi 1, Tzi 9, Tzi 12, Tzi 13, Tzi 17, Tzi 24, and Tzi 30) identified from the registered IITA inbred lines (Kim et al., 1987). These lines were selected for resistance to S. hermonthica after repeated evaluation in the field and screenhouse. The selected lines had moderate level of resistance to S. hermonthica. TZB-SR was a Nigerian composite (B) formed by intercrossing 43 maize cultivars from West Africa and the Caribbean. TZB-SR was resistant to the Maize streak virus and was commonly grown by farmers in the savannas of Nigeria. TZL COMP1-W was improved using selfed progeny (S1 and S2 line) or full-sib family selection schemes under artificial infestation with S. hermonthica at Abuja (9°16' N, 7°20' E, altitude 490 m) and Mokwa (9°18' N, 5°04' E, altitude 210 m) in Nigeria. The soil type at Abuja was a ferric luvisol Plinthustalf, with 81% sand, 12% silt, and 7% clay, whereas the soil type at Mokwa was Tropeptic Haplustox, which is fine and kaolinitic in nature. Abuja received an annual rainfall of 1292 mm, while Mokwa received an annual rainfall of 1069 mm.

Independent field trials consisting of 271 S1 lines extracted from the base population (C0), 225 S1 lines from the first cycle (C1), 180 full-sib families from the second cycle (C2), 151 S2 lines from the third cycle (C3), 464 full-sib families from the fourth cycle (C4), and 267 S1 lines from the fifth cycle (C5) were evaluated under artificial Striga infestation between 1989 and 1999. Progeny evaluation trials for the first two cycles were conducted only under artificial infestation with S. hermonthica, while those for subsequent cycles (C2–C5) were conducted with and without S. hermonthica infestation at Abuja and Mokwa. The lines or families were divided into sets, with genotypes included in each set arranged in a lattice design with two to four replications. Each family or line was grown in a single row plot of 5 m in length spaced 0.75 m apart with 0.5 m between plants within a row. To ensure uniform infection with the parasite, seeds of S. hermonthica were used to infest single plants in each S1 line or full-sib family (Kling et al., 2000).

The S. hermonthica seeds were collected from sorghum [Sorghum bicolor (L.) Moench] fields at different localities around Abuja and Mokwa in Nigeria at the end of the previous season. Fifty S. hermonthica seeds randomly drawn from the collected seed lot and disinfected with 1% sodium hypochlorite for 5 min were placed on moist filter paper in each of 10 sterile petri dishes. The petri dishes containing the Striga seeds were incubated in complete darkness at 25°C for 14 d. The filter paper in the petri dish was moistened daily with sterile deionized water. Two drops (10 µL) of 10 mg of a synthetic germination stimulant (GR24) were added to each petri dish, which was then placed in a dark incubator at 30°C for 24 to 48 h. Germinated Striga seeds with emerging radicles were observed under a dissecting microscope. Germination percentage was calculated from the ratio of seeds with emerging radicles to the total number of seeds placed on each petri dish. The germination percentage was thus used to determine the required amount of S. hermonthica seed mixed with fine sand at 1:99 ratios to deliver approximately 3000 to 5000 germinable Striga seeds per hill in field trials.

Three maize kernels of each line or family were planted in a 6-cm-deep hole infested with approximately 3000 to 5000 germinable Striga seeds by placing 8.5 g of sand-mixed S. hermonthica seed inoculum. Also three maize kernels of each line or family were planted in Striga noninfested rows, which were treated with ethylene 2 wk before planting to eliminate any potential Striga seeds present in the soil. A 3-yr rotation of cotton–soybean–maize and a 2-yr rotation of soybean–maize were used at Abuja and Mokwa, respectively, to reduce the incidence of S. hermonthica in Striga noninfested plots and to maintain the uniformity and productivity of the experimental fields. The two nonhost crops, cotton (Gossypium hirsutum L.) and soybean [Glycine max (L.) Merr.], stimulated suicidal germination of Striga seeds in previously infested maize fields, thereby reducing the reservoir of Striga seeds in the soil. Two weeks after planting, all the maize plants were thinned to two plants per hill. Fertilizer was applied at the rate of 30 kg ha–1 each of N, P, and K at planting and an additional 30 kg ha–1 N was applied 7 wk after planting. The N rate, which was only half the recommended rate for maize in the savannas of Nigeria, was used to ensure optimal development of S. hermonthica that allowed differentiation among lines or families for Striga damage symptom rating and ensured a minimum of 50% yield reduction under Striga infestation. Weeds other than Striga were removed manually with hoes on a regular basis. All management practices for both infested and Striga noninfested plots were the same.

All traits recorded in progeny trials of each selection cycle were obtained from Striga infested and Striga noninfested plots, except host plant damage symptom rating and numbers of emerged Striga plants that were recorded only in Striga infested plots. The number of emerged Striga plants was recorded on a plot basis. Ear height, husk cover, and plant aspect were recorded only in Striga noninfested plots. There was no Striga emergence in ethylene-treated Striga noninfested control plots. Host plant damage symptom was visually rated at each location on the two rows at 8 and 10 wk after planting using a scale of 1 to 9, where 1 = no visible host plant damage symptom and 9 = all leaves completely scorched resulting in premature death (Kim, 1994). Numbers of emerged S. hermonthica plants were also counted at 8 and 10 wk after planting. Days to anthesis and days to silking were recorded as the number of days from planting to when 50% of the plants shed pollen and had emerged silks, respectively. Plant and ear heights were measured in centimeters as the distance from the base of the plant to the height of the first tassel branch and the node bearing the upper ear, respectively. Plant aspect was rated on a scale of 1 to 9, where 1 = uniform plants with minimal reduction in height and ear size and low ear placement and resistance to foliar diseases and lodging and 9 = plants with severely stunted growth and small ears and susceptible to foliar diseases and lodging. Ear aspect was scored on a 1 to 9 scale, where 1 = clean, uniform, and large ears and 9 = rotten, variable, and small ears. Husk cover was rated on a scale of 1 to 9, where 1 = husks tightly arranged and extended beyond the ear tip and 9 = ear tips exposed. Ears per plant were calculated as the number of ears counted at harvest divided by the number of plants counted at harvest. All ears harvested from each plot were weighed and representative samples of ears were shelled to determine percent moisture. Grain yield adjusted to 15% moisture was, computed from ear weight and grain moisture assuming a shelling percentage of 80%.

Selection of lines or families from each replicated trial for recombination was based on a base index that involved selected traits using their lattice-adjusted means expressed in standard deviation units. Index scores were calculated as I = (2.0YLI + 1.0YLN + 1.0EHV – 1.0SDR8 – 1.0SDR10 – 1.0ESP8 – 1.0ESP10 – 1.0EROT), where YLI was yield in Striga infested plots, YLN was yield in Striga noninfested plots, EHV was the number of ears at harvest in Striga infested plots, SDR8 and SDR10 were Striga damage ratings at 8 and 10 wk after planting, respectively, ESP8 and ESP10 were the number of emerged Striga plants at 8 and 10 wk after planting, respectively, and EROT represented ear rot rating in Striga noninfested plots. The best 11 to 28% of the lines or families that combined higher yield and ear number under infestation with lower host plant damage symptom rating, fewer emerged Striga plants as well as acceptable yield and lower ear rot scores in Striga noninfested plots were selected using index scores. Remnant seeds of the selected lines or families were intercrossed to form each new selection cycle. A balanced composite of seed from crosses involving all selected parents formed the new cycle.

Seeds for this study were produced at Ibadan (7°26' N, 3°54' E, altitude 150 m) in Nigeria during the dry season of 2002 by making plant-to-plant crosses of randomly chosen 500 plants within each cycle of the population. At harvest 200 representative ears were selected, shelled, and bulked to obtain enough seed of each recurrent selection cycle for the field trial. The six recurrent selection cycles along with a Striga resistant open-pollinated variety (ACR97 TZL COMP1-W), a Striga tolerant hybrid (9022-13), and Striga susceptible hybrid (8338-1) checks were evaluated in a field trial at Abuja and Mokwa in 2003 and 2004 with and without artificial S. hermonthica infestation, which were hereafter referred to as Striga infested and Striga noninfested plots, respectively. ACR 97 TZL COMP1-W was developed by intercrossing the top 14 full-sib families derived from the fifth cycle of selection in TZL COMP1. The tolerant and susceptible hybrid checks were single crosses developed at IITA in the 1980s and have been used in screening trials as standard indicators of the level of Striga infestation attained during the season. The recurrent selection cycles and the checks were arranged in a randomized complete block design with four replications. Each genotype was planted in two Striga infested and two Striga noninfested rows of 5 m length spaced 0.75 m apart with 0.25-m spacing between plants within each row. The genotypes randomly assigned to each block were planted in adjacent Striga infested and Striga noninfested strips separated by a 1.5-m alley to get a precise estimate of yield loss from Striga (Kling et al., 2000). Plots were laid out in a serpentine fashion so that Striga infested rows occurred back-to-back in strips across the field, alternating with Striga noninfested strips. This arrangement minimized movement of Striga seeds into Striga noninfested areas, which were treated with ethylene 2 wk before planting to eliminate any potential Striga seeds present in the soil. A 3-yr rotation of cotton–soybean–maize and a 2-yr rotation of soybean–maize were used at Abuja and Mokwa, respectively, to reduce the incidence of S. hermonthica in Striga noninfested plots and to maintain the uniformity and productivity of the experimental fields.

All traits, including grain yield, days to silking, days to anthesis, plant height, ear height, number of ears, host plant damage symptom rating, and number of emerged Striga plants, described for progeny trials of the selection cycles were recorded in Striga infested and Striga noninfested plots in this study. In addition, the total number of plants was counted in each plot 2 wk after emergence. The number of Striga per plant was then calculated as the proportion of the total number of emerged Striga plants in a plot divided by the total number of plants. Anthesis–silking interval was computed as interval in days between dates of silking and anthesis. Relative yield loss was computed as a proportion of the difference in yield between a Striga noninfested plot and a Striga infested plot to grain yield in a Striga noninfested plots multiplied by 100.

Analysis of variance (ANOVA) combined over environments was conducted with PROC GLM in SAS using a RANDOM statement with the TEST option (SAS Institute, 2001). Independent analyses of variance were conducted for data collected from Striga infested and Striga noninfested strips. The location–year combinations were referred to as environments in the combined ANOVA for each trait. Environments and replications within environments were considered random effects, while recurrent selection cycles and checks were considered fixed effects. The significance of the mean squares for the main and interaction effects were tested using the appropriate mean squares, obtained from the RANDOM option in SAS (SAS Institute, 2001). Checks were excluded from the data set and regression coefficients (b) were computed for each trait to estimate the response to selection with cycles (0–6) being independent and cycle means dependent variables with PROC REG in SAS (SAS Institute, 2001). The average gain per cycle was calculated as the regression coefficient divided by the corresponding cycle zero mean.


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Response to Selection under Striga Infestation
Grain yield under Striga infestation, host plant damage symptom rating, and number of emerged Striga plants are considered to be important traits associated with resistance to S. hermonthica. In the combined analyses of variance, environments and genotypes were significant sources of variation for these and other traits recorded under Striga infestation (Table 1). The environmental effect represented 27 to 69% of the total variation in eight traits and less than 20% of the total variation in the remaining four traits. The genotype x environment interaction was significant only for host plant damage symptom ratings and number of emerged Striga plants. The variation among genotypes for the two traits was, however, higher than the corresponding variation due to genotype x environment interaction (Table 1). Further assessment of the genotype x environment interaction using rank correlation analyses between each pairs of the four environments revealed significant (P < 0.05) and positive correlation coefficients for host plant damage symptom ratings (r = 0.62 to 0.88) and number of emerged Striga plants (r = 0.61 to 0.98) at 8 and 10 wk after planting. Only 2 of the 24 rank correlation coefficients between pairs of environments were not significant for the two traits recorded at 8 and 10 wk after planting.


View this table:
[in this window]
[in a new window]

  		Table 1. Mean squares from the combined analyses of variance for traits of recurrent selection cycles and checks evaluated under Striga infestation at Abuja and Mokwa in Nigeria in 2003 and 2004.

 
In this trial, the recurrent selection cycles and checks sustained a 54.8% average grain yield loss due to S. hermonthica infection. The susceptible hybrid (8338-1) had the largest yield loss (81%) under Striga infestation, sustained the highest host plant damage symptom and supported the largest number of emerged Striga plants (Table 2). The base population (C0) undergoing improvement produced significantly more grain yield under Striga infestation than the susceptible hybrid check (8338–1), suggesting that the composite had some level of tolerance to S. hermonthica at the initial stage. The advanced cycle (C6) of TZL COMP1-W significantly out-yielded the base population (C0) by 1628 kg ha–1 and sustained a yield loss of 36%, which was nearly half of the yield loss (70%) recorded in C0 (Table 2). Mean grain yield of C6 did not differ significantly from those of the resistant variety (ACR97 TZLCOMP1-W) and the tolerant hybrid check (9022-13). In general, the observed increase in grain yield with advances in selection was accompanied by a concomitant reduction in yield losses due to S. hermonthica (Table 2 and Table 3). In the meantime, host plant damage symptom rating and number of emerged Striga plants recorded at 8 and 10 wk after planting significantly decreased from C0 to C6 (Table 2). The decrease in the host plant damage symptom rating from C0 to C6 was 48% at 8 wk and 33% at 10 wk after planting. The advanced cycle of selection (C6) supported significantly fewer S. hermonthica plants at 8 and 10 wk after planting in comparison to C0. Also, six cycles of selection significantly increased days to anthesis and ears per plant, reduced anthesis–silking interval, and improved ear aspect score. The changes in plant height and days to silking with selection were not significant (Table 2).


View this table:
[in this window]
[in a new window]

  		Table 2. Means of recurrent selection cycles and checks included in a trial evaluated under infestation with Striga hermonthica at Abuja and Mokwa in Nigeria in 2003 and 2004.

 

View this table:
[in this window]
[in a new window]

  		Table 3. Mean squares from the combined analyses of variance for traits of recurrent selection cycles and checks evaluated under Striga noninfested conditions at Abuja and Mokwa in Nigeria in 2003 and 2004.

 
After six cycles of a combination of selfed progeny and full-sib family selections in TZL COMP1-W, there were significant linear increases in grain yield, ears per plant, and days to anthesis under Striga infestation. The realized gain from selection was 24.3% per cycle for grain yield, 9.2% per cycle for ears per plant, and 0.4% per cycle for days to anthesis (Table 4 and Fig. 1 ). On the other hand, the linear changes in relative yield loss, host plant damage symptom ratings, number of emerged Striga plants, anthesis–silking interval, and ear aspect with selection were significant and negative. The selection gain per cycle was –7.2% for relative yield loss, –5.3% for host plant damage symptom rating at 10 wk, –8.5% for numbers of emerged Striga plants at 10 wk, –3.5% for anthesis–silking interval, and –4.8% for ear aspect (Table 4 and Fig. 1). The coefficient of determination for grain yield and other traits varied from 75 to 90% indicating a high proportion of the cycle variance was associated with the variance for each trait. The changes in plant height and days to anthesis and silking with selection were not significant.


View this table:
[in this window]
[in a new window]

  		Table 4. Linear response per cycle for traits of recurrent selection cycles of TZLCOMP1 evaluated at Abuja and Mokwa under Striga infestation in 2003 and 2004.

 

Figure 1
View larger version (13K):
[in this window]
[in a new window]

  		Figure 1. Response to selection for grain yield with and without Striga infestation, Striga damage rating, and number of emerged Striga per plant in TZL COMP1.

 
Simple correlations among traits were calculated for selection cycles and the checks (Table 5). Grain yield was negatively and significantly correlated with Striga damage rating, emerged Striga plants, anthesis–silking interval, ear aspect, and days to silking and positively correlated with number of ears per plant (Table 5). The correlation of Striga damage rating with emerged Striga plants, anthesis–silking interval and ear aspect was significant and positive, while its correlation with plant height and ears per plant were significant and negative. The relationship of emerged Striga plants with plant height, ears per plant, and days to anthesis was significant and negative, while its correlation with ear aspect was significant and positive (Table 5). Plant height was significantly and positively correlated only with ear aspect. The correlation of ears per plant with anthesis–silking interval, ear aspect, and days to silking was significant and negative. Anthesis–silking interval was significantly and positively correlated with ear aspect and days to silking. The correlations between the remaining pairs of traits were not significant (Table 5).


View this table:
[in this window]
[in a new window]

  		Table 5. Simple correlation coefficients between pairs of traits recorded in recurrent selection cycles of TZLCOMP1 evaluated at Abuja and Mokwa under Striga infestation in 2003 and 2004.

 
Response to Selection in Striga Noninfested Environment
In the combined analyses of variance, environmental effects were significant for all the traits recorded in Striga noninfested plots (Table 3). Genotypic differences were significant for grain yield, days to anthesis, plant height, plant aspect, and husk cover. The genotype x environment interaction was significant only for grain yield, days to silking, and plant aspect in Striga noninfested plots (Table 3). As shown in Table 6, grain yield of C0 of TZL COMP1-W did not differ significantly from those of the three checks in Striga noninfested plots, indicating that the yield potential of the composite was an appropriate baseline for initiating recurrent selection. The advanced cycle (C6) produced significantly more grain yield than the base population (C0). Linear regression analysis found significant improvement in grain yield, plant aspect, and ear aspect (analysis not shown). The liner response for these traits accounted for 74 to 83% of the total variation among cycles. The realized gain from selection per cycle was 4.4% for grain yield, –2.1% for plant aspect, and –1.8% for ear aspect. The linear regressions did not find significant negative trends for other agronomic traits recorded under Striga noninfested environment.


View this table:
[in this window]
[in a new window]

  		Table 6. Means of recurrent selection cycles and checks included in a trial evaluated in Striga noninfested environment at Abuja and Mokwa in Nigeria in 2003 and 2004.

 

   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Expression of genetic variability for traits associated with resistance to S. hermonthica in maize, including grain yield under Striga infestation, host plant damage symptom rating, and number of emerged Striga plants, is largely dependent on the presence of severe infection with the parasite. Improvement in infestation techniques and use of low to moderate level of N (30–60 kg N ha–1) have been effective in eliciting differential reactions of maize genotypes to S. hermonthica (Kim, 1991; Efron, 1993; Kling et al., 2000). The observed large loss in grain yield (81%), high host plant damage symptom rating, and large number of emerged S. hermonthica plants recorded in the susceptible hybrid check (8338-1) were clear indicators of the severe parasite pressure attained during evaluation of the recurrent selection cycles at the two locations. In spite of the differences in soil type, temperature, and amount of rainfall of the two test locations that are 400 km apart, the genotype x environment interaction for grain yield under Striga infestation were not significant. The significant rank correlations also suggested that the observed significant genotype x environment interaction for host plant damage symptom rating and number of emerged Striga plants represented a noncrossover type of interaction. These results indicate that the relative ranking of the recurrent selection cycles and the checks for these traits was consistent across environments. Evaluating lines or families derived from each cycle of selection of TZL COMP1-W under severe artificial infestation with S. hermonthica at the two test locations to form new cycles of selection could contribute to the observed stable expression of the three Striga resistance parameters across environments.

A combined use of selfed progeny and full-sib family selection schemes significantly increased grain yield by 24% per cycle under Striga infestation in TZL COMP1-W. This grain compares favorably to the yield gain of 19 to 22% per cycle reported for early maturing maize pools (Ceballos et al., 1991), but was higher than the yield gain of 6 to 8% reported for other pools and populations subjected to recurrent selection for resistance to diseases (Ceballos et al., 1991; Carson and Wicks, 1993). A significant increase in grain yield was attained at C3 and the subsequent cycles of selection. A slight decrease in grain yield, however, was observed from C0 to C4 possibly due to the low selection intensity (28%) used to identify promising lines for recombination to form C4. The heavy weight assigned to grain yield under Striga infestation in the selection index, the high selection intensity, and the large number of progeny evaluated under infestation seem to be effective for rapid improvement in grain yield and resistance to the parasite. The continuous increase in grain yield with advances in selection suggests that adequate genetic variability still exists in TZL COMP1-W to make further improvement in this trait with recurrent selection under Striga infestation.

Variance component estimates in TZL COMP1-W showed that additive genetic variance played a prominent role in conditioning Striga damage rating and grain yield under Striga infestation, while the dominance variance was of great importance for number of emerged Striga plants and grain yield under Striga noninfested condition (Akanvou et al., 1997). When additive genetic variance is more important than the dominance variance in a population, S1 progeny selection would be superior to full-sib family selection in most instances (Weyhrich et al., 1998). Full-sib family selection would use both additive and dominance genetic variances available in the composite for improvement of grain yield under Striga infestation and other traits. A combination of selfed progeny and full-sib family selection schemes were thus used to take full advantage of the additive and dominance genetic variances present in the composite. Although selfed progeny selection is expected to be superior to full-sib family selection for improvement of grain yield under Striga infestation in TZL COMP1-W, significant yield gains under Striga infestation were recorded when the later was used. Holthaus and Lamkey (1995) also found a lack of response to inbred progeny selection compared to reciprocal recurrent and half-sib selection in Iowa Stiff Stalk Synthetic (BSSS). It appears that full-sib family selection utilized both additive and dominance genetic variance for grain yield and other traits recorded under Striga infestation, which were included in the selection index, more effectively than the inbred progeny selection schemes.

Selection for simultaneous reduction in host plant damage symptom and number of emerged Striga plants has been recommended for increasing the level of field resistance to S. hermonthica in maize (Kim, 1994; Kling et al., 2000) and sorghum (Haussmann et al., 2001b). Six cycles of recurrent selection reduced host plant damage symptom rating by 33% and number of emerged Striga plants by 59% at 10 wk after planting in TZL COMP1-W. In spite of the reported low heritability estimates for number of emerged S. hermonthica plants (Berner et al., 1995; Akanvou et al., 1997), the observed gain from selection for this trait was higher than that of host plant damage symptom rating. Recurrent selection can thus further reduce Striga infection, which may contribute to significant reduction in parasite reproduction and result in reduced parasite infestation in the long term. Some studies have demonstrated that the number of emerged Striga plants recorded aboveground is significantly correlated with the number of Striga plants attached to the roots in maize (Kim et al., 1999; Menkir, 2006) and sorghum (Rodenburg et al., 2005). Several mechanisms, including low germination stimulant production (Weerasuriya et al., 1993; Reda et al., 1994; Vogler et al., 1996; Haussmann et al., 2001a), reduction in successful establishment of parasitic plants on roots (Lane et al., 1997), restricted parasite penetration of roots due to lignified cells with silica deposits (Maiti et al., 1984), and reduced capacity to elicit haustorial induction of Striga (Rich et al., 2004), have been implicated in lowering the number of emerged Striga plants. Ejeta et al. (2000) had also summarized other potential postgermination mechanisms of resistance that impede attachment and emergence of Striga in crops. Whether the observed reduction in the number of emerged Striga plants with selection in TZL COMP1-W are due to these or other mechanisms is yet to be determined.

Host plant damage symptom rating represents a visual assessment of the extent of leaf chlorosis and scorching and reduction in plant height as well as ear and tassel size of the host plant caused by Striga. The rating is obtained by visually integrating the overall appearance of all the plants in Striga infested plots (Kim, 1994). In our study, marked reduction in host plant damage symptom rating was recorded from C0 to C6 suggesting that a significant improvement in the overall growth and health of the host plant had occurred in TZL COMP1-W after six cycles of recurrent selection. Since plant size sets a limit to photosynthetic area (Troyer and Brown, 1976; Troyer, 1990), the observed reduction in host plant damage symptom may improve the capacity of the composite to maintain high rate of photosynthesis under Striga infestation. Gurney et al. (2002) found that a maize cultivar that sustained less damage from S. asiatica maintained higher rates of photosynthesis than the cultivars that sustained greater damage from S. asiatica. Also the results of our study showed that recurrent selection for resistance to Striga significantly shortened anthesis–silking interval by nearly 4% per cycle and ears per plant by 9% per cycle. In studies with tropical maize populations, a short anthesis–silking interval was related to an improvement in partitioning of assimilates to developing ears, resulting in higher ear growth rates, reduced embryo abortion after fertilization, and fewer barren plants under drought stress and low soil N (Lafitte and Edmeades, 1994; Bänziger et al., 1999).

Breeding for improved performance under S. hermonthica infestation can be attained through selection for resistance, to reduce parasite infection level, or tolerance, to minimize the impact of parasite infection on host performance (Rodenburg et al., 2005). Combining these two forms of resistance has been regarded as a promising strategy to breed crops for durable resistance to Striga (Kim, 1994, 1996; Haussmann et al., 2001a). Relative yield loss due to Striga results from the combined effect of resistance and tolerance and can thus be used as a selection criterion in crops (Rodenburg et al., 2005). However, this parameter was not used for selecting promising lines or families that formed new cycles of selection of TZL COMP1-W because the trait was often associated with lower mean grain yields (Kling et al., 2000). In our study, the observed changes in relative yield loss with selection did not follow consistent trend from C0 to C4, but decreased with selection thereafter. This inconsistency could arise from the fluctuations in the relative importance of resistance and tolerance to the overall defense against Striga, resulting from year-to-year variation in the level of parasite infestation at the time of evaluation of selfed progenies and full-sib families derived from the composite. According to Rodenburg et al. (2005), resistance assumes greater prominence against Striga with increasing level of parasite infestation in sorghum. Seasonal changes in environmental factors such as soil moisture, fertility, and temperature contribute to fluctuations in the level of Striga infestation in the field (Kim, 1996). Consequently, selection for low relative yield loss should be combined with selection for reduced number of emerged Striga plants and high yield in Striga infested and Striga noninfested environments to develop maize genotypes with superior resistance to S. hermonthica in the field.

The second objective of recurrent selection in TZL COMP1-W was to improve performance under Striga infestation while at the same time increasing grain yield and maintaining other agronomic traits in Striga noninfested environment. Selection for resistance to Striga using an index resulted in significant progress in the desired direction for traits recorded not only in Striga infested plots but also for grain yield, plant aspect, and ear aspect recorded in Striga noninfested plots. Furthermore, the changes in other agronomic traits recorded in Striga noninfested plots were not significant. Contrary to our findings, Ceballos et al. (1991) did not find significant yield increases in disease-free environments for tropical pools, which were subjected to recurrent selection for increasing levels of resistance to northern corn leaf blight [Exserohilum turcicum (Pass.) K.J. Leonard and E.G. Suggs]. These results indicate that recurrent selection for improved performance under Striga infestation may result in concomitant improvement in yield potential and acceptable performance in other agronomic traits in Striga noninfested environment.

The combined use of selfed progeny and full-sib family selection schemes has been effective in increasing alleles imparting reduced level of parasite infection. Continual screening of lines or families derived from the broad-based composite under artificial Striga infestation may facilitate the combination of multiple-resistance alleles into single lines through recombination. Such lines with multiple-resistance alleles can be used for the development of maize cultivars with resistance that are durable and broadly effective against a variable and rapidly changing parasite (Berner et al., 1995). In fact, some inbred lines derived from TZL COMP1-W evaluated under artificial infestation with populations of S. hermonthica and S. asiatica in Kenya exhibited higher level of resistance to the two species of Striga than locally adapted maize inbred lines (Gethi and Smith, 2004). Thus, the composite can be a good source of inbred lines with reduced level of parasite infection that can be used as parents in a breeding program for developing Striga resistant germplasm. Our study provides evidence that selecting for increased resistance to S. hermonthica may not adversely affect performance in a Striga noninfested environment. The observed continuous progress in performance under Striga infestation demonstrates the potential effectiveness of recurrent selection for increasing polygenetic resistance against the parasite in tropical populations.


   ACKNOWLEDGMENTS
 
This research was conducted at the IITA and financed by IITA. The authors express their appreciation to all staff members involved in carrying out field experiments.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Received for publication July 26, 2006.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 




This article has been cited by other articles:


Home page
 Agron. J.Home page
B. Badu-Apraku, M. A. B. Fakorede, A. F. Lum, and R. Akinwale
Improvement of Yield and other Traits of Extra-Early Maize under Stress and Nonstress Environments
Agron. J., March 4, 2009; 101(2): 381 - 389.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Menkir, A.
Right arrow Articles by Kling, J. G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Menkir, A.
Right arrow Articles by Kling, J. G.
Agricola
Right arrow Articles by Menkir, A.
Right arrow Articles by Kling, J. G.
Related Collections
Right arrow Crop Genetics

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Vadose Zone Journal
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome