| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a ENSA, 2 Place Viala, 34060 MONTPELLIER Cedex, France
b CIMMYT, P.O. Box 6-641, 06600 Mexico D.F., Mexico
c 52A Williams Street, Cambridge, New Zealand
* Corresponding author (pmonneveux{at}yahoo.fr)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: ASI, anthesis-silking interval • C, carbon • CIMMYT, Centro Internacional de Mejoramiento de Maíz y Trigo • DA, days to anthesis • DS, days to silking • DTP-W, drought tolerant population-white • DTP-Y, drought tolerant population yellow • DTP1, drought tolerant population 1 • DTP2, drought tolerant population 2 • N, nitrogen • PAR, photosynthetically active radiation • PH, plant height.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Maize is particularly sensitive to water stress in the period 1 wk before to 2 wk after flowering (Grant et al., 1989). Drought during this period results in an easily measured increase in the anthesis-silking interval (ASI) as silk emergence is delayed (Edmeades et al., 2000) and in grain abortion (Boyle et al., 1991). Grain abortion commonly occurs during the first 2 to 3 wk after silking (Westgate and Boyer, 1986; Schussler and Westgate, 1991b). It is exacerbated by any stress that reduces canopy photosynthesis and the flux of assimilates to the developing ear so that it falls below a threshold level necessary to sustain grain formation and growth (Edmeades and Daynard, 1979; Tollenaar et al., 1992). The reduction in photosynthesis can be due to a decrease in radiation interception, associated with reduced leaf expansion, leaf rolling (Bolaños et al., 1993) and foliar senescence (Wolfe et al., 1988), and to a reduction in C fixation per unit leaf area because of stomatal closure or a decline in carboxylation capacity (Bruce et al., 2002). On the basis of considerations of heritability and correlation with yield under stress, barrenness, ASI, leaf senescence, and leaf rolling were proposed by Bänziger et al. (2000) as secondary traits useful for improving maize yields in drought-prone environments.
Nitrogen stress also reduces final grain number by increasing kernel abortion (Lemcoff and Loomis, 1986; Pearson and Jacobs, 1987; Uhart and Andrade, 1995b). Around 85% of the abortion occurs during the first 20 d after female flowering (Monneveux et al., 2005). This increase in grain abortion is closely related to a lack of post-flowering nitrogen uptake by the crop (Below, 1997). Nitrogen deprivation reduces leaf area index and hence radiation interception. It also accelerates senescence of lower leaves (Wolfe et al., 1988; Moll et al., 1994), decreases radiation use efficiency (Uhart and Andrade, 1995a), and increases ASI (Jacobs and Pearson, 1991; Edmeades et al., 2000). Bänziger and Lafitte (1997) and Bänziger et al. (2000) proposed ASI and foliar senescence as secondary traits for improving maize for low-nitrogen target environments.
CIMMYT initiated breeding for drought tolerance in maize in 1975. Since most elite germplasm has a low frequency of alleles conferring drought tolerance, CIMMYT’s initial approach was to utilize recurrent selection techniques in elite populations. Evaluation for drought tolerance was based on replicated trials at one or two water stress levels during a rain-free period using controlled irrigation. Methods of selection and water stress management have been described by Fischer et al. (1989) and Bolaños and Edmeades (1993a). Severe water stress was induced during flowering and grain filling such that average grain yield in these trials was reduced 15 to 30% relative to unstressed yields. The same germplasm was also grown under well-watered conditions. Selection was based on an index involving grain yield under drought and well-watered conditions and ASI, barrenness, leaf senescence, and leaf rolling under drought (Bolaños and Edmeades, 1993a; Edmeades et al., 1999). This methodology was used to develop drought tolerant versions of several elite lowland tropical populations (Edmeades et al., 1999). Several publications document selection gains under a range of environmental conditions using this approach (Bolaños and Edmeades, 1993a, 1993b; Bolaños et al., 1993; Chapman and Edmeades, 1999; Edmeades et al., 1999). For example, Edmeades et al. (1999) observed an increase in grain yield under drought of 0.26 Mg ha–1 (12.6%) cycle–1 following S1 recurrent selection using a selection intensity of 5 to 10%, and a lower rate of improvement (0.08 Mg ha–1 or 3.8% cycle–1) following full-sib recurrent selection with a selection intensity of 26 to 32%. These gains were associated with increases in ears per plant and in harvest index, a slight reduction in leaf senescence, and a decline in ASI and stem biomass. They also reported a small but significant increase in grain yield, ears per plant, kernel number per fertile ear, and individual kernel weight under well-watered conditions. In the same study, Chapman and Edmeades (1999) reported reductions in plant height, time to anthesis, and tassel primary branch number. In an earlier study using similar methods, recurrent selection produced no measurable effect on plant water status (Bolaños et al., 1993). Bänziger et al. (1999) evaluated contrasting selection cycles of several of these populations under low N and observed similar yield gains under low N as well. They suggested that common mechanisms were responsible for increased partitioning of assimilates to the developing ear and for increased yields under both types of stress. This perception is reinforced by findings of Andrade et al. (2002) who found that a common curve described the response of kernel number to crop growth rate around flowering whether the crop was stressed by inadequate water or by nitrogen deficiency.
More recently, two tropical maize populations DTP1 and DTP2 have been developed at CIMMYT using known sources of drought tolerance followed by recurrent selection (Edmeades et al., 1997). These populations were formed to capture most of the known sources of drought tolerance adapted to the lowland and mid-altitude tropics in two populations of mixed color and grain type. During the development of DTP1 and DTP2, diverse components were recombined first by a half-sib recurrent selection for two to three cycles under no direct selection for drought tolerance. This was followed by recurrent S1 selection under managed drought stress in Mexico, accompanied by two cycles of international progeny testing in the case of DTP1. DTP2, formed later than DTP1, included large elite fraction of DTP1. To make these source populations more useful for national programs, elite fractions of DTP1 and DTP2 were first merged and then split into yellow- and white-grained populations (Edmeades et al., 1997). These populations differ in important ways from the lowland tropical populations La Posta Sequía, Pool 26 Sequía, and Tuxpeño Sequía studied by Edmeades et al. (1999). First, they were assembled from germplasm sources ranging from Corn Belt experimental hybrids to relatively unimproved landraces, selected on the basis of the criteria of proven superior performance under drought stress and adaptation to altitudes of 1800 m elevation or less. Second, they are not well adapted to any specific tropical environment, are of mixed grain texture, and have no well-defined heterotic response. Finally, because they were assembled from components with a putative high frequency of drought-adaptive alleles, the focus on selection initially was less on drought tolerance and more on uniformity and adaptation. Both populations were intended to be used as sources of drought tolerance and never for direct release by national programs. In early studies of performance, selections from these populations showed acceptable levels of drought tolerance, and good performance under optimal conditions (Edmeades et al., 1999).
The objectives of the present study were to assess the level of tolerance of these populations to drought stress and to low N and to evaluate changes due to selection in traits included in the selection procedure (i.e., grain yield, ears per plant, anthesis-silking interval and leaf senescence) as well as correlated responses in other traits when cycles of selection were grown under drought, low nitrogen, and optimal conditions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
DTP2
While DTP1 was being improved as a closed population, additional source materials were systematically screened under well-watered conditions and under moderately severe mid-season drought stress. Selection was based on delayed foliar senescence, superior grain yield, a reduced level of barrenness under stress, a short ASI, and lodging resistance under stress. Sources evaluated in this manner were 300 landrace collections held by CIMMYT’s Germplasm Bank from collection sites <1000 m elevation and annual rainfall <600 mm, 104 elite selections with reputed drought tolerance from Mexico, southern Africa, Thailand, USA, and CIMMYT’s conventional breeding program, and 52 elite selections with known drought tolerance from CIMMYT’s drought breeding program (Edmeades et al., 1999). Sources were evaluated per se vs. DTP1 in the winter season at Tlaltizapán, then the best 50% were crossed with an elite fraction of DTP1 C5 and evaluated in a second season before final selection of components. The population DTP2 was formed by the introgression of 25 new drought tolerant sources into DTP1 C5. Thus DTP1 formed 58% of DTP2, and the balance consisted of contributions from elite lowland topical drought tolerant sources (20%), elite Corn Belt sources (10%), landraces (5%), and miscellaneous tropical hybrids and lines (7%). Adaptation of components of DTP2 was estimated as lowland tropical (65%), subtropical (15%), and temperate (20%). Initially, components were recombined during three cycles of half-sib recombination under mild selection pressure, followed by one cycle of S1 recurrent selection under drought and heat. Entry numbers and screening and selection procedures in Ciudad Obregón and Tlaltizapán were as described for DTP1. Recombination of the best 40 families from C4 was by grain color to form DTP2-Y C5 and DTP2-W C5.
DTP-Y and DTP-W
The best 200 S1 families from each color fraction of DTP1 C7 were evaluated under two levels of drought stress in a combined trial with 200 DTP2 C5 S1 families of matching color in Tlaltizapán. The superior 60 families in each color class were recombined to form DTP-Y C8 and DTP-W C8, and 200 S1 progenies were generated and tested under three water regimes (well-watered, stressed at flowering, and stressed during grainfilling). Selection was for increased grain yield and ears per plant and for reduced leaf senescence and ASI. Recombination was of the superior 40 S1 families from each population, and a new set of S1 progenies of each was formed. Cycle balanced bulks of C9 were created from a similar number of seeds from each S1 family. Both DTP-Y and DTP-W are intermediate in maturity, with a final germplasm composition of approximately 60 to 65% lowland tropical, 15 to 20% subtropical, and 20% temperate.
Generation of Cycle Bulks
Bulk seed generated from progenies of each cycle was advanced to at least F2 to eliminate residual heterosis among crosses. Bulks of each cycle were sown in a block of 300 plants and each plant was crossed with one other to create F1 ears. Each of the harvested ears was shelled and a bulk created from two seeds per ear. The process was repeated to create F2 seed, and harvested ears were shelled and bulked for experimentation. Evaluations were performed under drought, low N, and optimal (well-watered and well-fertilized) conditions and involved four cycles of DTP1 (C0, C3, C6W, and C6Y), four cycles of DTP2 (C0, C3, C5W, C5Y) and DTP-W C9, and DTP-Y C9. In subsequent calculations, the performances of DTP-W C9 and DTP-Y C9 were included in gain estimates for DTP2. The term cultivar is used hereafter to denote cycles of selection in these populations.
Experimental Conditions
The current study was conducted under drought, low N, and optimal conditions at CIMMYT’s experimental station at Tlaltizapán, Morelos, Mexico, during the dry (winter) season of 2002–2003. The soil is a calcareous vertisol (Isothermic Udic Pellustert) 1.3 to 1.8 m in depth, with a pH of 7.6. Total water holding capacity in the top 1 m of soil is 265 mm, of which about 100 mm is available to the crop (Bolaños and Edmeades, 1993a). Air vapor pressure varied between 1.0 and 1.2 kPa. Daily minimum and maximum temperatures averaged 12 and 32°C, respectively. The average daily photosynthetically active radiation (PAR) increased from 9.0 to 13.0 MJ m–2 d–1 and Penman ETo increased from 4.0 to 9.0 mm d–1 during the growing season.
Trials under drought, low N, and optimal conditions were established separately using an 
(0,1) lattice design with three replications. Experiment plots were sown in four rows 5 m in length and 0.75 m apart. Plots were overplanted and thinned to a distance between plants in the row of 25 cm for an established plant density of 5.3 plants m–2. All trials were irrigated by sprinkler before soil preparation to destroy volunteer seeds and after sowing to ensure uniform emergence. All trials except those allocated to the drought treatments were irrigated by furrow irrigation to field capacity throughout their life cycle at intervals of approximately 2 wk, to represent the optimal treatment. Plots receiving the drought treatment were also irrigated every 2 wk until 21 d before anthesis, when water was withdrawn. One additional final irrigation was applied 17 to 19 d after 50% anthesis (Bänziger et al., 2000). The low N treatment was established according to Bänziger et al. (1999). No nitrogen was applied to the experimental field for a 5-yr period, during which maize was grown and all aboveground biomass removed from the field. In all other plots, a total of 150 kg N ha–1 was applied two times (before sowing and at V6 stage) as ammonium sulfate (NH4)2 SO4 in droughted and well-watered trials. In the low N treatment, N-NO3 concentration in the 0- to 30-, 30- to 60-, 60- to 90-, and 90- to 120-cm layers was 6.36, 2.09, 2.06, and 2.27 mg kg–1. In the other two treatments, N-NO3 concentration was 49.5, 24.12, 6.73, and 6.54 mg kg–1, respectively. All trials received 60 kg P2O5 (triple superphosphate with 46% P2O5) ha–1 applied before sowing, and no potassium since previous tests has shown no response to this element on these soils. The experiments were kept free from weeds, insects, and diseases. Seeds were treated before sowing with a mixture of one insecticide and two fungicides. The insecticide used was thiodicarb (dimethyl N,N'-[thiobis[(methylimino)carbonyloxy]]bis[ethanimidothioate]) and the fungicides were fludioxonil (4-(2,2-difluoro-1,3-benzodioxol-4-yl)-1H-pyrrole-3-carbonitrile), and metalaxyl (methyl N-(2,6-dimethylphenyl)-N-(methoxyacetyl)-DL-alaninate). A herbicide treatment was applied during pre-emergence at the rate of 2.2 kg ha–1 atrazine (6-chloro-N-ethyl-N'-(1-methylethyl)-1,3,5-triazine-2,4-diamine) and 1.7 kg ha–1 s-metolachlor (2-chloro-N-(2- ethyl-6-methylphenyl)-N-[(1S)-2-methoxy-1-methylethyl] acetamide). Plants were also treated to control fall armyworm [Spodoptera frugiperda (JE Smith)] with permethrin [(3-phenoxyphenyl)methyl 3-(2,2-dichloroethenyl)-2,2-dimethylcyclopropanecarboxylate] granules.
Measurements
Beginning when the sixth leaf tip was visible, six plants per plot, sampled as thinnings from unthinned border rows, were uprooted at 2- to 3-d intervals. Apical meristem length was measured under a dissecting microscope fitted with an eyepiece graticule whose divisions were calibrated in millimeters. The meristem was considered initiated if its length exceeded 0.5 mm (Siemer et al., 1969). Examination ceased when tassels were initiated in all six plants examined. The date when 50% of the apical meristems were >0.5 mm in length was recorded as tassel initiation for each plot.
Days to anthesis (DA) and silking (DS) were recorded from a well-bordered group of 40 plants in each plot. A plant was considered as having reached anthesis or silking if at least one extruded anther or one silk was visible. A plot was considered as having reached anthesis or silking when at least 50% of the plants reached these stages. Anthesis-silking interval (ASI) was calculated as DS–DA. After completion of male flowering, plant height (PH) was recorded as the distance between the ground surface and the node bearing the flag leaf. PH values were recorded on ten plants per plot and averaged. In all trials, in vivo relative chlorophyll concentration of the ear leaf of 10 plants per plot was assessed initially 2 wk after male flowering, and subsequently at 2 wk intervals, using a portable chlorophyll meter (SPAD-502, Minolta, Tokyo, Japan), and was expressed in arbitrary absorbance (or SPAD) values (Dwyer et al., 1991). Senescence and leaf rolling were assessed 2 wk after male flowering and then at 2-wk intervals, according to Bänziger et al. (2000). Senescence was scored on a scale from 0 to 10, derived from the percentage estimated total leaf area that was dead divided by 10. Leaf rolling was scored on a scale from 1 (unrolled) to 5 (completely rolled) in droughted plots.
One day after DA six well-bordered plants per plot (from a 1.125-m2 area) were cut at the soil surface, divided into tassels, ears, stems (including leaf sheaths), leaves and husks, and oven-dried at 80°C to constant weight in a forced-air circulation oven. The procedure was repeated on eight well-bordered plants per plot at maturity, when grain was separated from the cob. The weights of individual plant parts were combined to obtain the total aboveground biomass dry weight. At maturity, harvest index was calculated as the ratio of grain weight to total aboveground biomass dry weight.
In the two border rows of each plot, five areas of six plants each were marked for observation of ovule number per ear and kernel abortion. The number of ovules per upper ear was determined at silking on sampled ears as the product of ovule number per row (averaged from two ear rows) and number of rows on each ear. Ears were harvested 20, 30, 40, and 50 d from DS for the whole plot without removing the plants, so that effective bordering was maintained. Grain number per ear was determined by the same procedure as for ovules. Abortion rate was estimated in each population as the slope of the regression of ovule or grain number on days after silking after verifying the significance of the linear regression. Since most kernel abortion occurred during the first 30 d after anthesis, the regression was calculated with the ovule or grain number at 0, 20, and 30 d after anthesis. At physiological maturity, grain yield from 18 bordered plants from a 3.4-m2 area within each plot was measured by hand-harvesting. Ears were counted (an ear was defined as having one or more grains on it) and air dried, shelled, and weighed. Ear number per plant and grain yield were calculated, weight of 1000 kernels was determined, and grains per ear were calculated from grain yield and 1000-kernel weight. All grain weights were expressed on a dry weight basis.
Statistical Analysis
Data were subjected to analysis of variance (ANOVA), considering entries as fixed and replicates, plots, and incomplete blocks within replicates as random factors. Analyses were made by SAS, version 8.1. (SAS Institute, 1987). Effects of genotype, environment, and their interaction were determined for yield and yield components. Duncan’s t test was used to establish differences between cycles in each population. Selection progress per cycle for yield and yield components was estimated in each population by regressing genotype means on cycle of selection and by testing the significance of the linear regression (Edmeades et al., 1999). Percentage change per cycle was computed as the ratio of the regression slope to the mean performance of C0. Phenotypic correlations were used to determine the relationships among traits, within each environment, and across environments.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Since recurrent selection and evaluation were conducted in the same environment, yield progress was probably overestimated because increasing adaptation to the selection site was confounded with increased tolerance to drought. Byrne et al. (1995) evaluated cycles of selection in Tuxpeño Sequía in the normal growing season at sites that differed from the selection location and reported gains that were around 80% of those observed in the selection environment. Gains similar to those observed at the selection site have also been observed for tropical populations tested in the subtropics and southern Corn Belt (G.O. Edmeades, unpublished data), suggesting that improvements were relatively consistent even in locations where the populations were poorly adapted. However, even assuming 80% efficiency in transferring these gains to other environments, an increase in yield of at least 67% in DTP1 and 29% in DTP2 can be expected by evaluating this material under drought conditions that reduce yield to one third of its potential. The efficiency of being able to select for drought tolerance under managed irrigation conditions in the dry season where the stress can be managed for intensity, timing, and frequency normally greatly exceeds losses due to specific adaptation to a dry season screening environment (Barker et al., 2005; Edmeades et al., 2005).
In DTP1, yield gain was associated with a strong increase in ear number per plant under stress, consistent with that found by Bolaños and Edmeades (1993a) and Chapman and Edmeades (1999). If stress is of sufficient magnitude, complete abortion of the ear can occur. Decreasing barrenness under stress at flowering remains a highly effective way to increase yield under this condition, especially in relatively unimproved germplasm. Progress for this trait obtained in DTP1 across six cycles represented a gain of more than 18 000 ears ha–1 when sown at standard plant density (53 000 plants ha–1). Changes in DTP2 for this character were influenced by a significant increase in prolificacy under optimal conditions that accompanied selection (Table 1). Two of the sources of DTP2 had been previously selected specifically for prolificacy (Edmeades et al., 1997) and were the likely sources of this trait.
Recurrent selection for improved tolerance to stresses imposed at flowering led to a strong decrease in ASI, similar to that reported by others (Fischer et al., 1989; Bolaños and Edmeades, 1993b; Chapman and Edmeades, 1999). This trait is an easily observed external indicator of ear growth rate and hence partitioning and is a reasonably reliable predictor of grain yield under stress (Edmeades et al., 2000). It was highly negatively correlated with ear weight at DA (r = –0.522, P < 0.01) and final grain yield (r = –0.529, P < 0.01) across stress levels. On the other hand, there were no clear changes in chlorophyll concentration, leaf senescence, and leaf rolling with selection in the present study (data not shown), consistent with Bolaños et al. (1993) in the population Tuxpeño Sequía. Chapman and Edmeades (1999) pointed out that the apparent lack of progress in leaf senescence rate, a reasonably heritable trait, may reflect an increased level of internal competition for nitrogen caused by the greater demands of a larger ear in advanced selection cycles when grown in dry or N-deficient soils.
Correlated Responses in Unselected Traits
Changes in unselected traits can arise because direct selection affects plant processes that form part of a chain of yield-determining events. Recurrent selection resulted in significant increases in number of grains per fertile ear in both populations, as previously observed by Bolaños and Edmeades (1993a) in Tuxpeño Sequía. In agreement with Chapman and Edmeades (1999), we found no selection response in 1000-kernel weight under drought stress, but the reduction in weight per kernel attained significance in DTP2 where prolificacy has increased, leading to increased intraplant competition for assimilates.
Further evidence of changes in dry matter partitioning in favor of ear formation was observed. There was a significant increase in harvest index, as reported by others (Bolaños and Edmeades, 1993a; Edmeades et al., 1999) but no change in aboveground biomass. Thus the improvements in grain yield, reduced ASI, and reduced barrenness all point to important changes in dry matter partitioning within the shoot that increase the chances of grain formation under stress at flowering. Our results agree with Bolaños et al. (1993) who postulated that improved drought tolerance in Tuxpeño Sequía was due to increased partitioning of dry matter toward the developing ear rather to a change in water status. In temperate maize, Echarte et al. (2004) also found that genetic yield improvement was attributable to increased partitioning of dry matter to the ear during the critical period bracketing silking. Furthermore, in DTP1, there was a significant increase of ear dry weight and a decrease of tassel and stem dry weight, both at anthesis and maturity. The decline in tassel dry weight was closely associated with a decrease in stem dry weight but not with changes in plant height as reported in other similar studies (Bolaños and Edmeades, 1993a; Chapman and Edmeades, 1999). Ear and tassel dry weights were negatively correlated across cultivars, suggesting intraplant competition for assimilates between these reproductive organs. An association between reduced tassel size and improved partitioning toward the ear has been suggested by Fischer et al. (1989) and confirmed by the effects of plant detasseling on yield under drought (Sangoi and Salvador, 1998). Selection for reduced tassel size may also increase canopy photosynthesis through reduced shading (Duncan et al., 1967). Reduced tassel size appears to be a relevant breeding objective under stress, especially in tropical germplasm. This trait can be easily altered by selection (Duvick, 1997), and is highly heritable (Bolaños and Edmeades, 1996). Reduction of tassel weight was much greater than the reduction in tassel primary branch number reported by others (Bolaños et al., 1993: –2.6% cycle–1; Chapman and Edmeades, 1999: –5.9% cycle–1), suggesting that selection for reduced tassel weight may be performed without decreasing tassel ramification and pollen production at the same rate. Intraplant competition also may be found between the ear and the husk, but in the current study, husk dry weights increased in step with ear dry weights in response to selection.
We found a strong reduction in grain abortion with selection in both DTP1 and DTP2, largely explaining the differences observed across cycles in the number of grains per fertile ear (r = –0.926, P < 0.001). There was also a significant decrease of the number of ovules with selection. These results are similar to that reported by Edmeades et al. (1993) who observed a 2.5% cycle–1 reduction in ovule number in Tuxpeño Sequía. This change could be associated with early and more vigorous silking (Edmeades et al., 1993). It has been also postulated that fewer ovules can grow more quickly, thereby more readily achieving some threshold weight needed to prevent abortion at the onset of the linear growth phase (Edmeades et al., 2000). During the lag phase of their growth, the growing grains apparently have little capacity to attract preanthesis photosynthates (Schussler and Westgate, 1991a), and their development depends strongly on current photosynthesis. When grains enter the linear phase of filling, their sink strength increases, and stored assimilates are mobilized from the stem and husk. These hypotheses are consistent with the significant positive correlation between abortion rate and ASI (a reflection of ear growth and assimilate availability) and the strong negative correlation observed between ovule number at flowering and abortion rate. The regression of abortion rate on ASI in the present study predicted that for every day increase in ASI, there was a loss of 40 ovules over the first 30 d of grain filling.
Responses under Low N
In DTP1, selection under drought also led to yield gains under low N conditions. Beck et al. (1996) and Bänziger et al. (1999) reported that selection for drought tolerance increased grain yield under a range of N levels. These results suggest a common mechanism of tolerance to stress at flowering (Andrade et al., 2002; Bruce et al., 2002). According to Bänziger et al. (1999), yield gain under low N can be attributed to delayed leaf senescence, larger aboveground biomass, or more extensive root systems. In the present study, however, yield gain under low N was not associated with significant changes in senescence or aboveground biomass. Moreover, increased drought tolerance was found by Bolaños et al. (1993) to be associated with reduced root biomass in the upper 50 cm of rooting depth. Recent observations also confirmed that drought tolerant lines and hybrids had less developed root systems, compared with susceptible cultivars (P. Monneveux, unpublished data), though rooting depth may well be increased. As observed under drought stress, stem and/or tassel weights were reduced by selection, when observed under low N conditions. A significant negative correlation was noted between tassel weight and yield in different types of germplasm (lines, open pollinated varieties, hybrids), and an association between reduced tassel size and improved partitioning toward the ear in tropical germplasm has been suggested by Monneveux et al. (2005). The increase in yield and ears per plant in DTP1 under low N appeared to be driven more by changes in dry matter partitioning than by changes in biomass production or senescence.
Effects under Optimal Conditions
The effects of selection under water stress on yield under optimal conditions remain controversial. Selection under drought for traits that are related to survival under drought may carry a yield penalty under optimal environments (Ludlow and Muchow, 1990). However, Bolaños and Edmeades (1993a), Byrne et al. (1995), and Chapman and Edmeades (1999) have observed significant yield gains in maize in optimal environments because of selection under a range of conditions that deliberately included a well-watered control. The rate of gain observed arising from this type of selection has been consistent across yield levels (Bolaños and Edmeades, 1993a), though in other studies it has sometimes showed a decrease as yield levels increased (Barker et al., 2005). A significant increase in yield under optimal conditions was not found in the current study. This may not appear surprising, given that the type of germplasm used to form these two source populations was chosen on the basis of its observed or putative drought tolerance rather than its yield potential. Moreover, in the recurrent selection procedures used in these populations, a much greater emphasis was placed on performance under stress than under well-watered conditions, compared with elite tropical populations previously evaluated by Edmeades and coworkers. The increase in prolificacy observed in DTP2 under optimal conditions does appear to be associated with a significant increase in harvest index.
Implications for Maize Breeding
Our study confirms the efficiency of recurrent selection using grain yield and selected secondary traits to improve drought tolerance in these two tropical maize source populations. Even though these populations have a high putative level of drought tolerance because they were assembled from drought tolerant sources, further gains have been made in drought tolerance. Furthermore, gains obtained under water deficits were at no cost to yield under unstressed environments, suggesting that increased yield and yield stability across environments varying for water availability is a reasonable breeding goal. Efficiency of this selection strategy, however, requires access to adequate genetic variation and to managed stress environments, precise field phenotyping conditions, and the use of advanced designs and analysis techniques to remove spatial trends in data that occur under water or N stress (Barker et al., 2005). Lines extracted from the cycle C9 of DTP2-W and DTP2-Y have been screened for drought and low N tolerance and additionally for their adaptation to lowland tropical conditions. The best of these are available to breeders as unique sources of drought tolerance and have also been used to create synthetics and single and three-ways hybrids that are currently being evaluated under drought and low N conditions.
Improved tolerance over cycles of selection was found to be mainly due to increased partitioning of biomass toward the developing ear, as suggested by Bolaños and Edmeades (1993a), rather than to changes in water status or senescence. Echarte et al. (2004) also found that yield progress in Argentinean hybrids was mainly attributable to increased partitioning of dry matter to the ear. Most gains in ear growth were attributable to decreases in tassel and stem weight, i.e., to successful competition between the ear at flowering and other organs for available C products. There was also a reduction in grain abortion in advanced cycles of selection for drought tolerance, largely explained by a reduction of ovule number at silking, leading to less competition among developing grains. This sink reduction, also observed under optimal conditions, if a continued outcome of selection, could eventually limit potential yield. However, this can easily be offset by increasing the level of prolificacy in the populations, a process that appears to have already started in DTP2.
Most physiological studies dealing with drought tolerance in maize have focused on mechanisms involved in light interception (senescence and reduction of leaf expansion) and CO2 assimilation (stomatal conductance, photosynthetic capacity) (Bruce et al., 2002) or of water capture. Mechanisms of intraplant competition for assimilates allowing sink adjustment to available C levels, and favoring ear and grain growth have been, in general, poorly analyzed. Further research on drought tolerance in tropical maize should consequently give more attention to reduction of competition between developing grains and other organs that are growing at a time that coincides with kernel set. Reducing plant height may well be one way in which this could be implemented (Edmeades et al., 2000), and reductions in husk and cob biomass could also be of use. In this manner, it seems likely that yield gains under stress will translate into increases in harvest index and grain yield under all growing conditions, including those where water stress is not present.
Received for publication April 26, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. F. Stanger and J. G. Lauer Corn Stalk Response to Plant Population and the Bt-European Corn Borer Trait Agron. J., April 4, 2007; 99(3): 657 - 664. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
