| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
a Misión Biológica de Galicia, Spanish Council for Scientific Research (CSIC), Apartado 28, 36080 Pontevedra, Spain
b Estación Experimental de Aula Dei, Spanish Council for Scientific Research (CSIC), Apartado 202, 50080 Zaragoza, Spain
* Corresponding author (gsandoya{at}mbg.cesga.es).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Abbreviations: ECB, European corn borer • GxE, genotype x environment • PSB, pink stem borer
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Received for publication February 14, 2007.
a Misión Biológica de Galicia, Spanish Council for Scientific Research (CSIC), Apartado 28, 36080 Pontevedra, Spain
b Estación Experimental de Aula Dei, Spanish Council for Scientific Research (CSIC), Apartado 202, 50080 Zaragoza, Spain
* Corresponding author (gsandoya{at}mbg.cesga.es).
Selection against pink stem borer (Sesamia nonagrioides Lef) (PSB) attack was performed in the maize (Zea mays L.) synthetic EPS12. The direct response of the EPS12 population to three cycles of selection to reduce tunnel length damage by PSB while maintaining yield was evaluated. S0 (cycles of selection) and S1 (selfed cycles of selection) generations and testcrosses to three testers were evaluated under artificial infestation with two corn borers, PSB and European corn borer (Ostrinia nubilalis Hubner) (ECB). Genotypes and the genotype x environment interaction were significant for tunnel length and yield. Differences among cycles of selection were significant for tunnel length, and the linear decrease for this trait (–1.80 cm cycle–1) achieved during selection accounted for most of these differences. Yield was not significantly reduced with selection. Inbreeding for yield significantly increased due to selection. EPS12 crossed to EP42 showed a significant increase in yield with selection, while crosses to other testers showed a nonsignificant reduction in yield. Although crosses to EP42 were not significantly different for tunnel length, the high heterosis between EPS12 and EP42 and the increased yield of EPS12 x EP42 during selection suggest that inbred lines developed from advanced cycles of EPS12 could be crossed to EP42 to generate promising hybrids. In general, resistance to PSB and ECB was improved, while yield was maintained, inbreeding was increased, and yield of the cross EP42 x EPS12 was improved.
Abbreviations: ECB, European corn borer • GxE, genotype x environment • PSB, pink stem borer
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Resistance to ECB is polygenic, the additive genetic effects being the most important (Chiang and Hudon, 1973; Jennings et al., 1974; Guthrie et al., 1989). Similarly, resistance against PSB was described as additive with some dominant effects and no significant genotype x environment (GxE) interaction (Butrón et al., 1998a,b, 1999b; Cartea et al., 1999, 2001; Velasco et al., 2002). Intrapopulational recurrent selection has been proposed as an adequate breeding method when additive effects are the most important. Recurrent selection increases the frequency of favorable alleles. Some examples of successful use of these methods for improving maize resistance against stem borers have been reported (Penny et al., 1967; Barry et al., 1983). Klenke et al. (1986b) performed four cycles of selection to improve resistance to the first and second generations of ECB in the synthetic BS9, reducing damage by both generations. No selection programs for resistance to PSB have been reported so far. S1 recurrent selection to improve maize resistance to insect pests has been associated with unfavorable responses in other agronomic traits, such as yield decreases (Russell et al., 1979; Klenke et al., 1986a; Nyhus et al., 1989; Butrón et al., 2002).
An intrapopulational recurrent selection program for improving PSB resistance in the maize synthetic population EPS12 was initiated in the 1990s. EPS12 was obtained after three cycles of recurrent selection for yield in EPS7 (Vales et al., 2001). EPS7 was the result of intercrossing four landraces from the Ebro Valley and eastern Spain (Ordás, 1991). As EPS7 and some of its original landraces had shown resistance against PSB attack (Malvar et al., 1993; Velasco et al., 2002), EPS12 was chosen as the base material to carry out a selection program for improvement of PSB resistance. The intensity of selection applied was 10%, selecting 10 families, out of 100, with shortest stem tunnel length and yield above the 100-family mean. At the moment, three cycles of S1–progeny selection have been completed in EPS12. Molecular changes during selection (Butrón et al., 2005b) and the changes in leaf and sheath antibiosis (Butrón et al., 2005a) have been monitored; however, the direct response to the three cycles of selection has not been evaluated.
The objectives of this work were (i) to evaluate the direct response of the synthetic EPS12 to the selection program for resistance to PSB and (ii) to determine if the selection process significantly modified inbreeding and the performance of the population in crosses to testers.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
In 1996, progenies from the first cycle of selection were obtained by selfing EPS12(S)C1-syn1 plants. Simultaneously, EPS12(S)C1-syn1 was recombined: 300 seeds were planted, thinned to 150 plants, and crosses were made using each plant only once as male or female. At harvest, at least 50 ears were obtained and an equilibrated bulk of 1000 seeds was collected establishing the synthetic EPS12(S)C1. In a similar way, EPS12(S)C2 and EPS12(S)C3 were obtained in 1999 and 2002, respectively.
EPS12(S)C1 seeds were accidentally and irreparably mixed with seeds from another maize synthetic; therefore they were not included in the present study. However, the selection process was not affected because S1 families were obtained from EPS12(S)C1-syn1 (Butrón et al., 2005b).
Evaluation of the Selection Program
In 2002, seeds from the synthetics EPS12(S)C0, EPS12(S)C2 and EPS12(S)C3 were multiplied, selfed, and testcrossed to inbred lines A639, B93, and EP42. Inbred lines A639, B93, and EP42 were used because they represent different heterotic groups (Reid, Lancaster, and humid Spain heterotic groups, respectively). The inbred lines A639 and B93 were reported as resistant to PSB (Butrón et al., 1999a; Butrón et al., 2006), and EP42 as susceptible (Butrón et al., 1999a). To multiply the seed from each synthetic, plant-to-plant crosses among 150 plants were made using each plant as male or female. For each synthetic, at least 50 plants were selfed to obtain the synthetics EPS12(S)C0S1, EPS12(S)C2S1, and EPS12(S)C3S1. Finally, for the testcrosses, each cycle contributed with approximately 100 male plants. Pollen from those plants was bulked and pollinated around 30 plants per inbred.
The three cycles of selection per se, EPS12(S)C0, EPS12(S)C2, and EPS12(S)C3 (S0 generation), testcrosses to A639, B93, and EP42, and selfed cycles (S1 generation) were evaluated jointly in a split-plot design. The two levels of inbreeding were assigned to the main plots (inbred and noninbred) and genotypes within each inbreeding level to subplots.
The research was performed at two different locations of Spain: Pontevedra (42°24' N, 8°38' W, 20 m above sea level) and Zaragoza (41°44' N, 0°47' W, 230 m above sea level), in 2003 and 2004. Genotypes were evaluated under artificial infestation with two different corn borers (PSB and ECB) in separate experiments. Trials infested with PSB and ECB were adjacent in each year and location.
In Pontevedra, each experimental plot was hand planted and consisted of two rows spaced 0.80 m apart with 25 two-plant hills spaced 0.21 m apart. Plots were overplanted and thinned, obtaining a final density of 
60,000 plants ha–1. In Zaragoza, plots were machine planted and consisted of two rows spaced 0.75 m apart with 27 two-plant hills spaced 0.18 m apart. Plots were overplanted and thinned, with a final density of 74,000 plants ha–1.
At silking, 10 plants per plot were infested with a mass of about 40 eggs of PSB or ECB, depending on the trial. The infestation was performed according to Butrón et al. (1999b). The insect-rearing method used to obtain eggs of PSB was described by Eizaguirre (1989). The eggs of ECB were supplied by the Centre de Recherches de Poitou-Charentes in the Institute National de la Recherche Agronomique (France).
We evaluated the direct response to selection by recording stem tunnel length and yield because these traits were used as selection criteria. At harvest, 10 infested plants per plot were dissected and tunnel lengths (cm) made by borers were measured. Grain yield of the plot was adjusted to kernel moisture of 140 g H2O kg–1 and expressed as Mg ha–1.
Statistical Analysis
Combined analyses of variance across years, locations, and infested species of the cycles per se and testcrosses were performed. Locations, years, and replications were considered random effects, while infested species and genotypes were fixed effects.
Since the infested species and the infested species x year interaction were not significant, a new analysis of variance was performed, considering each combination of year, location, and infested species as a different environment. Environments and replications were considered random effects, and genotypes were considered fixed effect. The sums of squares due to genotypes were orthogonally partitioned into cycles per se, testcrosses to A639, testcrosses to B93, testcrosses to EP42, and between groups. The sums of squares of the GxE interaction were similarly partitioned. Moreover, the sums of squares due to each genotype group were partitioned into sums of squares due to linear regression, using the orthogonal coefficients derived from the Carmer and Seif formula (Carmer and Seif, 1963) and residual sums of squares. The linear regression coefficient was considered the average response per cycle of selection. Means were compared by Fisher's protected LSD method (Steel and Torrie, 1980).
Phenotypic correlations were calculated according to Johnson et al. (1955). For each cycle of selection, inbreeding depression in absolute units was calculated as the difference between the cycle per se (average of S0 plants) and the cycle selfed (average of S1 plants). Percentage of inbreeding depression was calculated by dividing the inbreeding depression in absolute units by the S0 mean and multiplying by 100. Standard errors for inbreeding depression in absolute units were calculated as the square root of the sum of the variances of noninbred and inbred cycle means (Lamkey and Smith, 1987). Significance of inbreeding depression was tested by a t test according to Satterthwaite (1946). Data analyses were performed using the SAS software package (SAS Institute, 2005).
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Significant differences among genotypes for stem tunnel length and yield were found. The GxE interaction was also significant (P < 0.01) for those traits (Table 1 ). The significant GxE interaction for stem tunnel length could be the result of the unusual weather conditions in Pontevedra in 2003 and Zaragoza in 2004 that favored exceptional PSB development and increased the infestation rate compared to standard conditions. No significant GxE interactions for tunnel length under more moderate infestation rates have been detected previously (Butrón et al., 1999b; Cartea et al., 1999).
|
|
|
EPS12(S)C0 showed the highest yield when crossed to B93, but the cross EPS12 x EP42 significantly increased its yield with selection, almost achieving the yield performance of EPS12(S)C0 x B93. The yield of the testcross to A639 and testcross to B93 showed a nonsignificant trend toward a decrease during selection, while the yield of EPS12 x EP42 increased significantly with selection, showing a linear increase of 0.16 Mg ha–1 per cycle (Table 3). If inbred lines were developed from the third or later cycles of selection in EPS12, they would be crossed to EP42. S1–progeny selection was effective in improving resistance to PSB and ECB. Several authors have reported improved resistance to pests by using this selection method. Penny et al. (1967) and Tseng et al. (1984) improved the resistance of maize against the first generation of ECB, Barry et al. (1983) developed a breeding population resistant to the second generation of ECB, and Klenke et al. (1986b) produced a synthetic resistant to both generations. In contrast, Williams and Davis (1983) were unsuccessful with a recurrent selection program used to improve resistance to Diatraea grandiosella. In the present study, EPS12 resistance was improved and yield was maintained. We avoided the negative results in yield obtained by other authors because they only selected for reduced tunnel length (Russell et al., 1979; Klenke et al., 1986a, 1988; Nyhus et al., 1989; Butrón et al., 2002). Anglade et al. (1996) improved the tolerance rating to ECB and the yield of the FS12 synthetic by two cycles of recurrent multitrait selection, but tolerance rating was not directly related to tunnel length.
In agronomic terms, the gains observed would be of limited importance because the 2-cm reduction in tunnel length achieved (per cycle) via selection, relative to the average length, 50 cm, would not be expected to significantly reduce lodging. However, the real response achieved in this synthetic population could have been underestimated because infestation rates were exceptionally high. In 2003, temperatures were higher than usual in Pontevedra, favoring pest development, especially PSB, which has an African origin. An investigation performed to study population dynamics and the distribution of corn borers in the Atlantic coast in northwestern Spain has pointed out that the development of these pests was higher than usual (Velasco et al., 2007). These exceptional conditions could be responsible for underestimating the real resistance because, with high levels of infestation, few or no genotypes could be resistant (Granados and Paliwal, 2001). In fact, after removing the trial performed at Pontevedra in 2003 under PSB infestation, the response to selection increased, achieving a tunnel length reduction of 2.3 cm cycle–1.
Genetic drift could also limit the progress achieved by selection. Higher selection pressure could lead to a greater expected progress with selection; however, high selection pressure in small populations could render drastic changes in the population structure due to the small effective population size causing some inbreeding effects (Hallauer and Miranda, 1988). In each selection cycle, 10 families from a total of 100 were recombined, and that could generate genetic drift. There was not any clear trend of change for inbreeding depression for tunnel length (Table 4 ). These results are expected because the inheritance of this trait is basically additive (Butrón et al., 1999b; Cartea et al., 1999); therefore, allele fixation should not entail inbreeding depression. The trend of inbreeding depression for yield was to decrease with selection. Under the assumption of genes of equal effects and no epistasis, the decrease of inbreeding depression due selection would entail an increase of inbreeding respect to the initial population. The increase of inbreeding could be a consequence of genetic drift or favorable allele fixation (Benson and Hallauer, 1994). Since the selection scheme used increased the inbreeding and could limit gains for yield, we recommend in the future increasing the number of families in the evaluation to diminish the effect of genetic drift when selecting the best 10% families.
|
In conclusion, crop yield was maintained and resistance against PSB and ECB was improved, while inbreeding was increased by the selection process. The inbreeding increase would limit yield gains because it would be a consequence of genetic drift rather than of allele fixation. Limited success could be a result of many factors; particularly important is that selecting for two traits diminishes the selection pressure in each one, especially when they are negatively correlated. Yield changes of the testcrosses to Reid and Lancaster testers during selection were similar to yield changes of the population per se. However, the yield of the cross EPS12 x EP42 increased with the selection process.
G. Sandoya acknowledges three consecutive grants from the C. Iturriaga and Maria de Dañobeitia Foundation, the Mediterranean Agronomic Institute of Zaragoza (CIHEAM), and the Ministry of Education and Science (Spain). The authors thank Emma Muiños and Raquel Díaz for rearing insects. This research was also supported by the Plan Nacional I+D+I (AGL2003-00961 and AGL2006-1314) and the "Excma. Deputación Provincial de Pontevedra."
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 February 14, 2007.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
In this step, EPS12C1 was mixed with another maize population, but selection was not affected.