| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Dipartimento di Agronomia, Universitá di Bologna, Via Filippo Re, 6, 40126 Bologna, Italy
Corresponding author (plandi{at}agrsci.unibo.it)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: ABA, abscisic acid • ABA-3, leaf ABA concentration measured at mid-period of stem elongation • ABA-4, leaf ABA concentration measured at tassel appearance • ABA-5, leaf ABA concentration measured at mid-end of silking • DSI, drought stress index • gs, stomatal conductance • H-P, population selected for high-leaf ABA concentration • L-P, population selected for low-leaf ABA concentration • RWC, relative water content of the leaf • V0, V1, and V2, irrigation volumes corresponding to 0, 60, and 120% of actual evapotranspiration after accounting for rainfall, respectively • 
, leaf water potential
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
One trait universally known to be affected by drought is the concentration of abscisic acid (ABA; Zeevaart and Creelman, 1988; Davies and Zhang, 1991; Quarrie, 1991; Jackson, 1993). A widely recognized effect of an increase in ABA concentration is a reduction in stomatal conductance (Tardieu et al., 1993; Trejo et al., 1995), an adaptive feature that allows the plant to maintain a favorable water balance. Other important adaptive changes related to increased ABA concentration are an increase in the ratio between root and shoot biomass (Saab et al., 1990; Sharp, 1996) and, possibly, a reduction in leaf expansion rate (Bacon et al., 1997). These adaptive changes improve survival under natural conditions of limited water availability, but may reduce biomass production and yield in agricultural systems. Selection for high ABA might prove beneficial in agricultural systems when the dynamics and/or the intensity of the drought stress are such that genotypes characterized by protective mechanisms related to an ABA increase and favoring plant survival are better adapted than genotypes with higher water requirements. These conditions are more likely to occur under tropical environments in which crop establishment plays an important role in determining yield potential of a crop. Larque-Saavedra and Wain (1974) pointed out that the tropical line Latente, which shows enhanced ability to withstand drought condition, is also characterized by high ABA concentration.
To clarify the role of ABA accumulation, different experimental approaches can be followed, relying either on the direct manipulation of ABA levels by means of exogenous treatments (Quarrie, 1991) or, preferably, on the evaluation of genetic stocks differing in their capacity to endogenously accumulate ABA (Innes et al., 1984; Read et al., 1991; Pekic et al., 1995; Sanguineti et al., 1996). Mutants impaired in accumulating ABA (Schwartz et al., 1997) can also provide useful information on the role of this phytohormone. It is difficult, however, to properly utilize this information in the context of breeding materials because of the complex genetic basis of ABA concentration in leaf tissue of crops grown under field conditions (Ivanovic et al., 1992; Lebreton et al., 1995; Tuberosa et al., 1998) and the possible effects of ABA on traits which in turn influence yield.
To analyze properly the effects associated with genetic variation of a quantitative trait, such as leaf ABA concentration, divergent selection can be used to obtain populations differing for the trait in question but sharing a common genetic background. The correlated responses to divergent selections for leaf ABA concentration, which were successfully conducted in wheat (Triticum aestivum L.) and in maize by Quarrie and collaborators, were to some extent controversial (Innes et al., 1984; Read et al., 1991; Quarrie, 1993). Innes et al. (1984) and Read et al. (1991) tested the same wheat lines obtained by divergent selection for ABA and showed that in both cases higher ABA phenotypic values were associated with higher yields under water-limited conditions. However, Read et al. (1991) observed that the lines selected for high ABA concentration had an average ABA concentration lower than that of the lines selected for low ABA concentration. Maize populations obtained following selection for high ABA concentration, when tested under drought conditions, confirmed the effectiveness of selection and showed significantly lower grain yield than populations selected for low ABA concentration from the same population (Quarrie, 1993). The different responses to divergent selection for ABA observed in wheat and maize clearly suggested that further studies were needed to better understand the effects of genetic variation in ABA concentration. Furthermore, in both species, only one cross was used as source population, thus limiting our understanding of the genetic basis (linkage and/or pleiotropy) of correlated responses to selection.
On the basis of these premises, we undertook a divergent selection for leaf ABA concentration in maize through subsequent generations of selfing. To reduce the possible biases due to the fixation of specific gene combinations of the parents and to gain information on the selection effects in different genetic backgrounds, two F2 populations were used as sources. Selection was conducted in the field under conditions of moderate drought stress, and preliminary results of the selection work obtained by testing the selected F3 and F4 lines were reported in Sanguineti et al. (1996). However, inbreeding depression could have partially biased the assessment of the selection responses for traits, such as grain yield, that are affected by important non-additive gene action.
The goal of the present study was to evaluate the direct and correlated responses to selection unbiased by inbreeding effects, by testing the populations developed through intermating of the selected materials. A secondary objective was to evaluate responses to selection in populations derived from one F2 source when tested under different irrigation regimes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Selection was conducted in Cadriano (44° 33' N lat., 11° 24' E long. Po valley, northern Italy) on a Udic Ustochrept (fine silty, mixed, and mesic) soil. All selection trials were sown almost 2 months later than usual, to increase the chance of drought stress occurring in the period from stem elongation to flowering; irrigation volumes applied corresponded to about 50% of actual evapotranspiration, after accounting for rainfall. Selection began in 1992 and for each population, 480 F2 plants were selfed and examined for leaf ABA concentration. Leaf samples were collected at the mid-period of stem elongation and at tassel appearance, corresponding to Growth Stages 3 and 4, respectively, according to Hanway (1963). Leaf samples were immediately frozen and stored at -30°C until ABA analysis was carried out. The two leaf samples of each plant were bulked and an average leaf ABA concentration was determined by a radioimmunoassay with an ABA-specific monoclonal antibody (Quarrie et al., 1988). The 30 plants with the highest (H) and the 30 plants with the lowest (L) leaf ABA concentration were selected. In 1993, the 60 F3 families (30 H and 30 L) selected within each source population were compared in adjacent field trials; and ABA concentration of each F3 family was determined on a bulk of leaves collected at Growth Stages 3 and 5 (mid-end of silking, i.e., when about 75% of plants had silks visible). In addition, the 60 F3 families of each population were grown in the nursery and six plants of each family were selfed and analyzed for leaf ABA concentration at Growth Stages 4 and 5.
With respect to the Os420 x IABO78 material, the eight F3 families with the highest and the eight with the lowest leaf ABA concentration determined on the basis of analysis of bulked leaves were identified in the field trial within the H and the L groups, respectively. Then, utilizing the data obtained in the nursery, the plant with the highest leaf ABA concentration was selected within each of the eight H-F3 families, while the plant with the lowest leaf ABA concentration was selected within each of the eight L-F3 families. In 1994, the eight F4 families of each ABA group were crossed according to the diallel scheme. Four to five crosses were made for each of the 28 H and 28 L combinations and within each group about 130 plants were used as female parents and about 130 plants as male parents. An equal amount of seed of each combination was bulked to constitute the H- and L-populations (hereafter identified as H-P and L-P, respectively). In the same year, the source F2 was reproduced by intermating about 100 F2 random plants (as female parents) with about 100 F2 random plants (as male).
With respect to the Mo17 x B88 material, the divergent selection among families was conducted not only in F3 but also in F4, because of the limited, though highly significant, response to divergent selection observed in F3. Sixty F3 families were evaluated in the 1993 trial and 12 F3 families were selected in each directions. In the nursery, where F3 plants were selfed, the plant with the highest and the plant with the lowest leaf ABA concentration were selected within each of the 12 F3 families of the H and L group, respectively. In 1994, the corresponding 24 F4 families were tested in a field trial and leaf ABA concentration was analyzed at Growth Stages 3 and 5. The eight F4 families with the highest leaf ABA concentration and the eight F4 families with the lowest leaf ABA concentration based on analysis of bulked leaves were selected within the H and L groups, respectively. In 1995, the H-P, L-P, and the source F2 were produced following the same procedures described for the Os420 x IABO78 material.
Responses to Selection in Os420 x Iabo78
The three populations (F2, H-P, and L-P) were tested at Cadriano, at three irrigation volumes in both 1995 and 1996. The field layout was a split-plot design with four replications. Irrigation volumes were assigned to main plots and populations to subplots. The main plots were separated by four border rows. Each subplot consisted of two rows 4.32 m long and spaced 0.70 m apart; rows of adjacent subplots were spaced 1.30 m to favor the measurements of some physiological traits, such as leaf temperature and stomatal conductance. The three irrigation volumes were V0 (non irrigated) V1, and V2 (corresponding to 60 and 120% of the actual evapotranspiration after accounting for rainfall, respectively). Details of the procedure followed for calculating the crop evapotranspiration were provided by Landi et al. (1995).
Trials were hand-sown on 25 May 1995 and on 22 May 1996, i.e., 40 to 50 d later than usual. At thinning, 44 plants were left per subplot (5.1 plants m-2). Fertilizers were applied before sowing at rates of 150 kg ha-1 for N and 44 kg ha-1 for P; at thinning a further application of N (100 kg ha-1) was made. K fertilizer was not applied because of its high availability in this type of soil and because previous studies indicated that its application could have negative effects. Weeds were removed by mechanical cultivation and by hand. Irrigation was applied from mid-stem elongation to till the end of silking. Total irrigation volumes provided in V2 corresponded to 177 and 158 mm of water in 1995 and 1996, respectively; in V1 the volumes were halved. Total rainfall from sowing to harvest was 424 mm in 1995 and 421 mm in 1996. Trials were hand-harvested on 3 Oct. 1995 and 11 Oct. 1996.
Leaf samples to measure ABA concentration were collected at Growth Stage 4 (ABA-4) and at Growth Stage 5 (ABA-5) on 32 plants per subplot, after excluding three plants on each end of the row. The collection dates were 18 July (Growth Stage 4) and 26 July (Growth Stage 5) in 1995, and 17 July (Growth Stage 4) and 24 July (Growth Stage 5) in 1996. The measurement of ABA concentration of each plot was carried out on duplicate samples following the procedures described in Sanguineti et al. (1996). Samples were collected in the morning (from 1030 to 1200 h, local time) from the central part of the lamina of the fifth and third leaf from the top at Growth Stage 4 and Growth Stage 5, respectively. Leaf samples were immediately frozen and stored at –30°C. The concentration of physiologically active, unconjugated ABA [2-cis(+)-ABA] was determined with a radioimmunoassay as it was during the selection process; the results are reported as ng ABA g-1 dry weight (DW).
The following traits were recorded on the same 32 plants per subplot used to determine ABA.
1. Drought sensitivity index (DSI), as a visual score from 1 (plants with fully turgid leaves) to 5 (plants with leaves severely wilted), in the afternoon (from 1330–14:30 h, local time) of the same day in which leaf samples were collected for measuring ABA-4. Data were taken only in V0 because of negligible symptoms of drought sensitivity in V1 and lack of symptoms in V2.
2. Leaf relative water content (RWC), assessed according to the procedure described by Sanguineti et al. (1999) on the same leaves used for determining ABA-4 and ABA-5. RWC (%) was computed as 100 x (fresh weight - dry weight)/(turgid weight - dry weight).
3. Leaf temperature, measured with an infrared thermometer (Model 112, Everest Interscience, Inc. Tucson, AZ) in each of the 3 d following the two leaf samplings for ABA analysis. Data were taken under clear weather from 1200 to 1300 h, local time. Two measurements were taken for each plot by holding the thermometer at a distance of 1.2 m from the plants with an oblique angle so as to measure only the central portion of the canopy above the ear.
4. Interval from sowing to pollen shedding date, assessed when about 50% of the plants had extruded anthers.
5. Silk delay, calculated as difference between silking date (assessed when about 50% of plants had extruded silks) and pollen shedding date.
6. Root lodging, counting as lodged those plants leaning more than 30° from the vertical.
7. Grain yield, adjusted to 155 g kg-1 grain moisture.
8. Number of ears per plant, calculated as ratio between the number of ears (carrying at least 10 kernels) per subplot and the number of plants per subplot.
9. Mean kernel weight, adjusted to 155 g kg-1 grain moisture on a sample of 300 kernels.
10. Number of kernels per plant, calculated as ratio between grain yield per plant and kernel weight.
The following traits were recorded on 12 competitive plants per subplot.
1. Stomatal conductance to water vapor (gs), measured by two operators, each one using a steady state porometer (Model LI-1600, LI-COR Inc., Lincoln, NE). Measurements were made on leaf blades fully exposed and unshaded just prior to measurements. The central portion of each sampled leaf blade was measured on both the abaxial and adaxial surfaces. Each operator measured one complete replicate from 0930 to 1330 h, local time, on the same day in which leaf samples for measuring ABA-4 and ABA-5 were collected; the other two replicates were analyzed the following day, in the same way and during the same hours. The fourth and second uppermost leaves per plant were considered at Growth Stages 4 and 5, respectively. Further details on the measurements of gs are reported in Tuberosa et al. (1994).
2. Leaf water potential (), measured at Growth Stages 4 and 5 at the same time and on the same leaf blades used for measuring gs. To measure , the distal portion of the leaf blade (about 20 cm long) was cut after placing it in a humidified clear plastic bag. Within 10 to 15 s, about 3 cm of the proximal portions of the leaf blade flanking the midrib were removed and the leaf blade was placed in a pressure chamber (PSI, Corvallis, OR) lined with moist filter paper to reduce water loss due to transpiration. The elapsed time between excision and pressurization with compressed N2 was about 50 s. Balancing pressure was determined by observing with a magnifying lens when xylem sap appeared at the cut surface of the middle vein.
3. Plant height, measured to the flag leaf collar after silking.
The leaf samples for measuring ABA concentration and all other physiological traits (i.e., DSI, RWC, gs, , and leaf temperature) were taken during clear weather, with no wind and with incident photosynthetic photon flux density varying from 800 to 1600 µmol m-2 s-1. All traits were measured in both years, with the exception of gs and (only in 1995).
The analysis of variance (ANOVA) was conducted on subplot mean values and combined across samplings (two for leaf ABA concentration, RWC, gs, and ), measurements (six for leaf temperature), and years (two for all traits except gs and ). The ANOVA for root lodging percentage was conducted on data subjected to angular transformation. The differences among the three irrigation volumes and the corresponding interactions were partitioned into linear and quadratic components. The differences among populations and the corresponding interactions were partitioned into H-P vs. L-P (to evaluate the selection effect) and into (H-P and L-P) vs. F2 (to evaluate the asymmetry of the selection responses). For the F-test, the effects of years were considered as random and the effects of samplings, measurements, irrigation volumes, and populations as fixed.
Responses to Selection in Mo17 x B88
The source F2, H-P, and L-P were tested in three environments, i.e., Cadriano in both 1996 and 1997 and Rovigo (45° 40' N lat., 11° 48' E long. Po valley; sandy-loam soil not yet characterized according to the U.S. soil taxonomy system) in 1997. Only one irrigation volume was applied corresponding to about 60% of the evapotranspiration after accounting for rainfall. The field layout was a randomized complete block design with 12 replications. Each plot consisted of one row 5.2 m long and spaced 0.75 m from adjacent rows.
Trials were sown on 22 May 1996 and on 12 May 1997 at Cadriano, and on 14 May 1997 at Rovigo. The field practices were similar to those already described. After thinning, each plot included 20 plants (5.1 plants m-2). Irrigations provided a total volume corresponding to 79 mm (Cadriano, 1996), 66 mm (Cadriano, 1997), and 64 mm (Rovigo, 1997) m3 ha-1. Trials were hand-harvested on 11 Oct. 1996 and on 24 Sept. 1997 at Cadriano, and on 30 Sept. 1997 at Rovigo.
The following traits were investigated on 14 plants per plot after excluding three plants on each end of the row: ABA concentration and RWC (at both Growth Stages 4 and 5), DSI, pollen shedding date, silk delay, plant height, grain yield, number of ears per plant, number of kernels per plant, and kernel weight. The procedures followed for measuring or calculating these traits were the same as those described above. Traits were investigated in all three environments with the exception of DSI (Cadriano, 1996 and 1997), RWC (Cadriano, 1996), and number of ears per plant, number of kernels per ear, and kernel weight (Cadriano, 1997 and Rovigo, 1997). In the trial of Cadriano 1997, plants showed symptoms of a virus attack which that was mainly due to maize dwarf mosaic virus. The level of plant susceptibility was rated 1 wk before the onset of pollen shedding by a visual score from 0 (plants with no visible symptoms) to 9 (plants with severe mosaic and stunted growth).
The ANOVA was conducted on plot mean values and combined across samplings (leaf ABA concentration, and RWC) and environments (all traits except RWC and virus tolerance). The differences among populations and the population x environment interaction were partitioned as indicated above. For the F-test, the effects of environments were considered as random and the effects due to both samplings and populations as fixed.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
and no significant interaction with measurements (date) was detected for leaf temperature; therefore, only mean values across samplings or measurements are considered for these traits. The higher order interaction involving years, irrigation volumes, and populations was not significant for any trait. For the sake of conciseness, only data concerning irrigation volumes, populations, and their interaction are presented and discussed.
The mean trait values for the three irrigation volumes across the two years and the three populations are reported in Table 1. Irrigation led to a decline in ABA-4, ABA-5, leaf temperature, silk delay, and kernel weight, while it led to an increase in RWC, gs, , plant height, root lodging percentage, grain yield, number of ears per plant, and number of kernels per plant. Most of these findings were consistent with those obtained in studies conducted in maize grown at various levels of water supply (e.g., for RWC, gs, and see Bolaños et al., 1993; Gutierrez-Rodriguez et al., 1998; for leaf temperature and agronomic traits see Landi et al., 1995; Bolaños and Edmeades, 1996).
|
|
Table 2 shows the mean trait values for the three populations across the 2 yr and the three irrigation volumes. With respect to leaf ABA concentration at both Growth Stages 4 and 5, the significance of the difference among populations was due to the comparison between H-P and L-P, while the comparison between their mean value and that of the source F2 was not significant. In particular, H-P exceeded L-P by 199 ng ABA g-1 DW (90.5%) at Growth Stage 4 and by 208 ng ABA g-1 DW (74.0%) at Growth Stage 5. These results indicate that divergent selection was effective in changing the selected trait in both directions and that responses were symmetrical. These symmetrical responses could be due, at least partly, to the fact that, in this source population, the allelic frequencies at the segregating loci are equal to 0.5 and that leaf ABA concentration is controlled largely by additive effects (Sanguineti et al., 1996). In such conditions, it is expected that the additive genetic variance in the source population is at the highest level and that it declines (as a result of the divergent selection) following similar trends in both directions, thus favoring symmetrical responses (Falconer, 1981).
|
) and gs, as the differences between H-P and L-P were not significant. This lack of significance should be mainly ascribed to the negligible size of the differences, rather than to high variability of RWC, , and gs data, as indicated by their coefficients of variation (CV) that were smaller than that of DSI. Significant and negative correlations between leaf ABA concentration and both RWC and gs were observed by Sanguineti et al. (1999), who tested 80 random F4 families derived from Os420 x IABO78. The same authors also found a partial overlap between the quantitative trait loci (QTLs) affecting leaf ABA concentration and those affecting RWC or gs. The inconsistency between the results of Sanguineti et al. (1999) and those herein presented could be ascribed to the different level of inbreeding of the tested materials and, consequently, to the higher intensity of drought stress experienced by the F4 families compared with the more vigorous populations (equivalent to an F2) tested in this study. This inconsistency could be also due to the genotype x environment interaction because the different materials were grown in different environments (years and irrigation volumes).
It is worth mentioning that Pekic et al. (1995), comparing maize lines obtained by divergent selection for leaf ABA concentration, were unable to show significant differences for between the high-ABA and the low-ABA groups. Furthermore, Bolaños et al. (1993), following a recurrent selection for drought tolerance on a tropical maize population, did not detect any correlated response for predawn , when the source and the selected populations were examined at various irrigation regimes. They hypothesized that the improved drought tolerance of the selected populations was mainly due to changes in biomass partitioning toward the ear rather than to changes in traits affecting the plant water status. These findings could be related to the fact that maize is an isohydric species that tends to maintain a quite constant day-time leaf water status by varying stomatal conductance (Tardieu, 1996). Thus, plants experiencing different levels of drought stress could show fairly similar water potential values while differing for gs. If this was the case, H-P (according to its higher leaf ABA concentration) should have shown a lower gs than L-P. However, this was not observed here, possibly because of differences between the two selected populations in one or more of the following traits (not considered in our study) which can influence the water status and the water balance of the plant: leaf area index, root size, osmotic potential, hydraulic conductance in the pathway from roots to leaves, and/or stomatal sensitivity to ABA. In rice (Oryza sativa L.), selection for high leaf ABA concentration was associated with a decrease in stomatal sensitivity to ABA (Austin et al., 1982). Genetic variation in sensitivity to ABA has seldom been investigated on a large scale. Preliminary work carried out by Conti et al. (1994a) indicated that significant differences were present among maize inbred lines. In addition, studies of Zhang and Davies (1990), Tardieu et al. (1991)(1992), and Tuberosa et al. (1994) pointed out that gs in drought-stressed maize is more tightly regulated by ABA concentration in the xylem sap rather than in the leaf.
Table 2 also shows that other traits were affected by divergent selection for ABA and that their correlated responses to selection were always generally symmetrical, with the exception of sowing-pollen shedding interval and kernel weight. H-P showed lower grain yield, number of fertile ears per plant, number of kernels per plant, plant height, and root lodging percentage, while it showed higher leaf temperature and longer sowing-pollen shedding interval and silk delay. Most of these results were consistent with those obtained during selection by Sanguineti et al. (1996), who evaluated the H- and L-F3 families at only one irrigation volume (similar to V1). Furthermore, our results, in contrast to those for RWC and gs, were in good agreement with the results of a correlation analysis conducted on 80 random F4 families derived from the same single cross (Sanguineti et al., 1999), indicating that leaf ABA concentration was positively associated with DSI, leaf temperature, and silk delay and negatively associated with grain yield. Quarrie (1993) conducted a divergent selection for leaf ABA concentration induced by an artificial laboratory test (detached leaf test) and also found that the population selected for high ABA flowered later and was less productive than that selected for low ABA.
Because we used as the source population an F2 not subjected to intermating prior to selection, the correlated responses for some traits could be due at least partly to linkage effects. This could be the case for DSI and grain yield, because the high-ABA parent (i.e., Os420) was characterized by higher DSI, in comparison with the low-ABA parent (IABO78) (Conti et al., 1994b; Sanguineti et al., 1996) and by lower grain yield potential both per se (Sanguineti et al., 1999) and in hybrid combination (Landi et al., 1995). As to grain yield, H-P showed a mean value (3.61 Mg ha-1) which can be considered quite low in our location, even under conditions of water shortage. This finding could be due to the recovery of the low-yield alleles supplied by Os420 and tightly linked to high-ABA alleles because of the limited recombination that occurred during selection. However, the lack of a large difference between L-P and H-P for grain yield (1.53 Mg ha-1) suggests that this possible recovery did not play an important role. More likely, the low yield level of H-P should be related to the poor performance that, on the whole, was exhibited by the three Os420 x IABO78 populations. Such a poor performance should not be surprising if one considers that Os420 is an old inbred line representative of the first decade of maize breeding in the USA (Lamkey and Smith, 1987).
Differences for leaf temperature between H-P and L-P could be related to their different leaf angle, the leaves of L-P being more erect than those of H-P (data not shown). Differences in leaf angle were already noted during selection. Because IABO78 (i.e., the low-ABA parent) is characterized by leaves more erect than those of Os420 (Sanguineti et al., 1996), it is likely that also the correlated response for leaf habit was due to linkage effects.
The longer silk delay of H-P was consistent with its lower yield performance across environments. In fact, a negative association between silk delay and grain yield is often found in maize, particularly under drought conditions (e.g., Bolaños and Edmeades, 1996). The lower plant height of H-P can be related at least partially to the effects of ABA on this trait in cereals (Quarrie, 1991).
In addition, H-P showed a much lower percentage of root lodging, which could be ascribed to its lower height but not to its lower productivity. In fact, H-P showed less lodging than L-P not only in 1995, when root lodging occurred near physiological maturity, but also in 1996, when root lodging occurred at the end of flowering (i.e., when grain yield was not yet expressed). These findings thus suggest that H-P has a stronger soil anchorage than L-P. Lebreton et al. (1995), investigating an F2 maize population grown under drought conditions, found that leaf ABA concentration was positively associated with root pulling resistance and with the number of nodal roots, i.e., traits that can be associated with root lodging resistance (e.g., Fincher et al., 1985; Varlet-Grancher et al., 1987). Lebreton et al. (1995) also identified a region on chromosome 2 containing a QTL for both leaf ABA concentration and root pulling resistance. The same chromosome region was shown to contain a major QTL for leaf ABA also in a mapping population derived from the same cross herein studied (Tuberosa et al., 1998; Sanguineti et al., 1999). We are presently preparing segmental isolines for the QTL in question to elucidate its role in controlling leaf ABA concentration as well as root traits.
Responses to Selection in Mo17 x B88
The ANOVA for leaf ABA concentration conducted across the two samplings (at Growth Stages 4 and 5) revealed the significance of the population x sampling interaction (not shown); therefore, the results of ABA-4 and ABA-5 are presented separately. Significant differences among environments were found for ABA-4, ABA-5, grain yield, and its components. The population x environment interaction was significant only for ABA-5 and grain yield, and was due only to the component (H-P vs. L-P) x environment. For ABA-5, H-P always exceeded L-P, whereas for grain yield L-P always exceeded H-P (Fig. 2)
, indicating that the significance of the interaction was due to magnitude effects. The difference between H-P and L-P followed different trends for the two traits from one environment to the other; in particular, the largest difference for ABA-5 and the smallest difference for grain yield were noted in the same environment (Cadriano 1996).
|
|
The consistency between the correlated responses for DSI observed in Os420 x IABO78 and in Mo17 x B88 is noteworthy because the two pairs of parents were characterized by different combinations of values for leaf ABA concentration and DSI (see Materials and Methods for details). Therefore, these correlated responses are probably not merely due to linkage effects but also to pleiotropy, i.e., to the fact that high leaf ABA concentration and high sensitivity to drought are determined to a certain extent by a common set of genes. A consistency between correlated responses of the two groups of populations was also found for grain yield, despite the fact that the two pairs of parents were characterized by different combinations of ABA and grain yield values, similarly to the situation for ABA and DSI. In fact, the high ABA parent was characterized by lower yield potential than the low ABA parent in the first cross, while the reverse was true in the second cross (Sanguineti et al., 1996). Therefore, the correlated responses for grain yield may also be at least partly due to pleiotropy, i.e., to the fact that high leaf ABA concentration and low grain yield were affected by a common set of genes.
For grain yield, the difference between H-P and L-P of Mo17 x B88 was similar to the difference between H-P and L-P of Os420 x IABO78 at the same irrigation volume (i.e., V1), as can be deduced from Fig. 1 (-2.24 and -2.11 Mg ha-1, respectively). In contrast, the differences between H-P and L-P for leaf ABA concentration at both Growth Stages 4 and 5 were much smaller in Mo17 x B88 than in Os420 x IABO78. These findings could be related to differences in the linkage phase (coupling vs. repulsion) of genes for yield linked to genes influencing ABA or to different pleiotropic actions of genes affected by selection. In this respect, Sanguineti et al. (1999) detected a fairly broad overlap of QTLs for grain yield and QTLs for leaf ABA concentration and provided evidence that an increase in ABA was associated with either an increase or a decrease in grain yield, according to the chromosomal region involved. Differences in sensitivity to ABA in the two source populations could also account for the similar responses for grain yield, as compared with the different responses for leaf ABA concentration.
The correlated response to selection for the level of virus susceptibility suggests that in Mo17 x B88 the selection for low leaf ABA concentration favored those genotypes with greater tolerance, while the reverse occurred when selecting for high leaf ABA. It is difficult to interpret this correlated response because studies on the relationship between virus susceptibility and ABA concentration have not been reported in maize. However, it is noteworthy that Whenham and Fraser (1990) found that in tobacco (Nicotiana tabacum L.), infection by tobacco mosaic virus markedly increased ABA concentration.
In accordance with its lower drought tolerance, H-P also showed a significantly lower RWC than L-P, though the difference between the two mean values was small.
The comparison between the F2 and the mean of the two selected populations was significant for DSI, silk delay, and grain yield, and its components (number of ears per plant and number of kernels per plant), thus indicating asymmetrical correlated responses. At least to some extent, this could be due to random drift because each selected population was obtained by intermating only eight F4 families. For grain yield, the asymmetrical response to selection could also be due to the fact that the F2 source performed quite poorly, though there are not plausible explanations for this finding.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Received for publication January 19, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  P Landi, M. Sanguineti, C Liu, Y Li, T. Wang, S Giuliani, M Bellotti, S Salvi, and R Tuberosa Root-ABA1 QTL affects root lodging, grain yield, and other agronomic traits in maize grown under well-watered and water-stressed conditions J. Exp. Bot., January 1, 2007; 58(2): 319 - 326. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Giuliani, M. C. Sanguineti, R. Tuberosa, M. Bellotti, S. Salvi, and P. Landi Root-ABA1, a major constitutive QTL, affects maize root architecture and leaf ABA concentration at different water regimes J. Exp. Bot., December 1, 2005; 56(422): 3061 - 3070. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. TUBEROSA, S. SALVI, M. C. SANGUINETI, P. LANDI, M. MACCAFERRI, and S. CONTI Mapping QTLs Regulating Morpho-physiological Traits and Yield: Case Studies, Shortcomings and Perspectives in Drought-stressed Maize Ann. Bot., June 15, 2002; 89(7): 941 - 963. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) across two years (1995 and 1996) and three Os420 x IABO78 maize populations (H-P, L-P, and F2)
