| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dpto. de Producción Vegetal, Facultad de Agronomía, Universidad de Buenos Aires. Av. San Martín 4453 (C1417DSE), Ciudad de Buenos Aires, Argentina
b Instituto Nacional de Tecnología Agropecuaria (INTA), Estación Experimental Agropecuaria Pergamino, Buenos Aires, Pergamino
* Corresponding author (kdandrea{at}agro.uba.ar)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: ASI, anthesis– silking interval • Exp. 1, experiment 2000–2001 • Exp. 2, experiment 2001–2002 • HI, harvest index • k, light attenuation coefficient of the canopy • LAI, leaf area index • ln, leaf number n • PAR, photosynthetically active radiation • PGR, plant growth rate • R2, onset of active grain filling • RUE, radiation use efficiency • Vn, leaf stage
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Today, the largest investments in maize breeding are made by the private sector, where the whole selection process (i.e., from early inbred line development to commercial hybrids) takes place in the absence of N restriction (R. Mella, Monsanto Argentina, pers. comm.). The industry does not find it profitable to develop genotypes for areas or markets with low economic return, such as with low soil fertility. This may have resulted in the loss of some adaptive traits to these environments, such as the lack of genetic variability for N absorption found in studies performed with CIMMYT germplasm (Lafitte et al., 1997).
Research on phenotypic and physiological traits related to N metabolism in maize has been reported mostly for hybrids, but there has been no account of the genotypic variability of these traits because only one or a few genotypes were included in the experiments (Lemcoff and Loomis, 1986; Sinclair and Horie, 1989; Uhart and Andrade, 1995). Information on this topic is scarce for inbred lines, the basic genetic material for hybrid seed development. Until present, the few published studies on the response to N availability of this type of germplasm (Betrán et al., 2003b; Lafitte and Edmeades, 1995) did not report on genotypic differences in important traits related to biomass production and final grain yield, such as leaf area growth and senescence, light interception capacity, radiation use efficiency, and biomass partitioning (e.g., response of kernel number per plant to plant growth rate around silking). A better understanding of these attributes in inbreds is particularly important because, in spite of the poor correlation for grain yield between inbreds and their hybrids (Hallauer and Miranda, 1988), predictors of hybrid performance could increase the efficiency of hybrid breeding programs (Betrán et al., 2003a).
The objective of this study was to analyze, in a set of maize inbred lines, the genotypic variability in the response to N availability of several morphological and physiological traits known to be related to N use by crops. Because this variability may be limited among inbred lines in current maize breeding programs (Lafitte et al., 1997), inbreds with different origins and from different breeding eras were included.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Sowing took place on 6 December (Exp. 1) and on 12 November (Exp. 2). Treatments were a factorial combination of 11 (Exp. 1) or 10 (Exp. 2) inbred lines (Table 1) and two N levels. Seed of lines B098 and B101 was only available for Exp. 1 and that of line LP2541 only for Exp. 2. Nitrogen levels were low N (control), with no N added, and high N, with 400 kg N ha–1 added as urea in four applications between planting and the 12-leaf (ligulated leaves) stage (V12, Ritchie and Hanway, 1982).
A split-plot design with three replicates was used, with N availability as the main plots and inbred lines as the subplots. Plant population was 7 plants m–2, and each subplot had three rows, 0.7 m apart and 6 m long. Plots were hand-planted at three seeds per hill and thinned to the desired plant population at V3. Water stress was prevented by means of sprinkler irrigation, with the uppermost 1 m of soil held near field capacity throughout the growing season. The experiments were not under water deficit or flood. Crops were kept free of weeds, pests and diseases. Daily values of incident global solar radiation, mean air temperature, and rainfall were obtained from a LI 1200 (LI-COR, Inc., Lincoln, NE) weather station installed at the experimental field. Daily incident photosynthetically active radiation (PAR) was estimated by multiplying daily incident solar radiation by 0.45 (Monteith, 1965), and accumulated thermal time (TT, base temperature 8°C) was computed from mean daily air temperatures from sowing onward (Ritchie and Ne Smith, 1991).
Measurements
Leaf Area and Light Interception
Five plants were tagged in the central row of each plot at V3 to follow leaf appearance and senescence dynamics. Tags were placed at identified positions along the stem (e.g., between leaf 3 and leaf 4), which allowed the identification of individual leaves (ln) and the determination of total leaf number. The numbers of ligulated and senesced (more than half of leaf yellow) leaves per plant were registered weekly between seedling emergence and physiological maturity on all tagged plants. Individual leaf area was computed as lamina length x maximum width x 0.75 (Montgomery, 1911). Before silking, leaf area per plant was calculated as the sum of the areas of green ligulated leaves plus the final area of the following two leaves (Muchow and Carberry, 1989). After silking, it was measured as the sum of the area of all green leaves. A sigmoid function of the type indicated in Eq. [1] was fitted to the presilking evolution of green leaf area per plant as a function of the number of ligulated leaves.
 | [1] |
Leaf area index (LAI) was calculated as the product of leaf area per plant and number of plants per unit land area. The effect of N availability on maximum LAI was estimated (Eq. [2]) as the relative reduction in maximum LAI at low N as compared to maximum LAI at high N.
 | [2] |
Postflowering senescence was estimated for each N treatment (Eq. [3]) as the reduction in LAI at maturity relative to maximum LAI
 | [3] |
The fraction of incident PAR intercepted by the canopy was measured every 2 wk during a period around solar noon using a line quantum-sensor (LI-191SA, LI-COR, Lincoln, NE). Four determinations per plot were taken at midday, between 1130 and 1430 h, on clear days, with 1 m of the sensor placed diagonally across the rows immediately below the lowermost green leaves of the canopy (Gallo and Daughtry, 1986). Daily fractional interception between observation dates was estimated by linear interpolation. Daily incident PAR intercepted by the canopy was estimated as the product between daily values of incident PAR and fractional interception. The light attenuation coefficient of the canopy (k) was estimated from the exponential function in Eq. [4].
 | [4] |
Flowering Data
The dates of anthesis (i.e., at least one extruded anther visible) and silking (i.e., at least one extruded silk visible) were recorded on all tagged plants. The anthesis-silking interval (ASI) was calculated for each tagged plant as the difference in days between silking and anthesis dates (Uribelarrea et al., 2002), and averaged for each plot. The duration of the silking period was computed as the time in days between silking of the first and the last tagged plant in each subplot.
Biomass Production and Partitioning
Shoot biomass was estimated for each tagged plant at silking (Exp. 1 and Exp. 2), and at the onset of active grain filling (R2, Ritchie and Hanway, 1982) on 10 to 12 d after silking (Exp. 2 only), using nondestructive allometric models based on the relationship between plant biomass and morphometric variables (Borrás and Otegui, 2001; Maddonni and Otegui, 2004; Vega et al., 2000; Vega et al., 2001). Model parameter values were obtained from destructive harvests of at least eight plants from each genotype x N combination (two or three plants per subplot). Most of the relationships, 24 out of 31, were significant at P < 0.001 and all the relationships were significant at P < 0.05. The N levels produced a wide range of the morphometric variables (e.g., the volume ranged between 134 and 1430 cm3 and the biomass between 47 and 202 g at flowering in Exp. 2). There was at least one meter border plants separating tagged plants from the sample area at flowering or at R2 (border or inner rows) to leave tagged plants with adequate border area until physiological maturity. The morphometric variables used were plant height from ground level to the uppermost visible ligule, stalk diameter at the base of the plant, and maximum apical ear diameter at R2. Shoot biomass estimates for each tagged plant were used to calculate plant growth rate at flowering (PGR; in g plant–1 d–1) (Eq. [5]), and averaged for each plot.
 | [5] |
All plants tagged in the central row at V3 were sampled at physiological maturity, determined as black layer observed in grains of the mid portion of the ear (Daynard and Duncan, 1969) from 40 d from silking in two ears for each plot in border areas. Each plant was separated into leaves, stem plus sheaths and tassel, husks, and cob plus kernels. All plant material was oven dried at 60°C for 7 d and weighed.
Grain Yield and Its Determinants
Potential kernel number was estimated by counting the number of spikelets present on the apical ear of the eight plants sampled at silking for each plot. Grain yield per plant and its components (ears per plant, kernels per ear, and kernel weight) were determined for each tagged plant. Ears harvested at physiological maturity were individually hand-shelled, and grains were weighed to determine plant grain yield. Kernel number was counted for each ear, and kernel weight was calculated as the quotient between plant grain yield and kernel number per plant. Prolificacy was expressed as the number of grained ears per plant, and harvest index (HI) was calculated as the ratio between plant grain yield and aboveground plant biomass at physiological maturity.
Statistical Analysis
ANOVA was used to evaluate the effects of treatments and their interactions, and a Tukey test was applied to determine significant differences (P < 0.05) between the means. The relationships among attributes were evaluated by regression analysis. Models in Eq. [1] and [4] were fitted by TBLCURVE (Jandel Scientific, 1992), and the confidence intervals (P < 0.05) of the parameters of each function were used to determine the significance of the differences between models.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
|
0.95) at the stand density used in the experiments, and maximum fraction of incident PAR seldom reached values of 0.9 (Table 2). Inbreds differed (P < 0.05) in their light attenuation capacity (coefficient k in Eq. [4]), which was greater (k = 0.47–0.52) for B100, ZN6, P578, LP2, LP2541, and LP561 and smallest (k = 0.35–0.40) for the others inbreds. Nitrogen availability never affected k. In general, reduced N availability promoted greater postflowering leaf senescence than at high N, but inbreds differed in the magnitude of the response and a significant (P < 0.05) genotype x N interaction was detected for LAI at physiological maturity (Table 2). Some genotypes (e.g., LP2) exhibited strong relative leaf senescence (Eq. [3]) after flowering regardless of the year or N availability (between 80 and 98%), whereas others held a high proportion of maximum LAI at physiological maturity in most conditions (e.g., LP561) or only under no N stress (e.g., ZN6 and LP611).
Crop Phenology and Flowering Dynamics
Environmental conditions of each experimental year (Fig. 1) did not affect thermal time requirements to anthesis, silking, or physiological maturity (data not shown), but higher mean air temperatures experienced by the crop during Exp. 1 than Exp. 2 reduced the number of days to reach these events in the former (Table 3). Inbreds differed significantly (P < 0.05) in thermal time requirements up to anthesis and silking. Genotypes B100 and LP662 were among the earliest (range between 905 and 986°Cd for anthesis and between 910 and 1010°C for silking), LP611 and LP561 were the most delayed (range between 1007 and 1069°Cd for anthesis and between 1078 and 1120°Cd for silking). A difference of 7 to 10 d was registered in time to flowering between these extreme groups. For most inbreds, reduced N availability resulted in an increase in thermal time requirements up to anthesis and silking, especially during Exp. 2, but this trend was significant (P < 0.05) only for time to silking of inbred LP2541. For this inbred, many tagged plants never reached silking at low N, and a value equivalent to the mean date of physiological maturity was assigned to the silking date of these plants to include them in ASI computations. For all inbreds during Exp. 2, the delay in silking because of low N was longer than the delay in anthesis and resulted in an increased ASI (Table 3). The increase in ASI, however, was not always matched by a longer silking period of the crop as were expected (Table 3). A tendency to simultaneous increase in both parameters (ASI and duration of silking period) was only observed in inbreds ZN6 and LP2. For most other inbreds (B100, LP521, LP662, LP153, LP611, and LP561), there was a strong year effect on the response of the duration of silking period to N availability.
|
|
Grain Yield and Yield Determinants
The year effect was also evident on grain yield and most of its components (Table 3), except for the potential number of kernels represented by the number of spikelets per ear (data not shown). This difference between years could be attributed to the extremely high temperatures registered around silking in Exp. 1, which may have promoted reduced prolificacy and kernel set for most treatments in this experiment and diminished the effect of contrasting nitrogen availabilities.
Potential kernel number did not respond significantly to N availability, with the exception of inbred LP2541. This genotype exhibited a 75% reduction in the average number of spikelets per ear when cropped at low N, mostly because of the early arrest of ear growth in many plants under this N treatment. This effect of N stress was evident in the lack of silking already described, which resulted in the greatest reduction of prolificacy (47%) and kernel number per plant (65%) registered in this study.
Grain yield always responded significantly (P < 0.01) to the variation in kernel number (r2 = 0.79 for Exp. 1, and r2 = 0.75 for Exp. 2). This variation was promoted by differences among genotypes in Exp. 1, and among genotypes and N treatments in Exp. 2 (Table 3). Reduced N availability in Exp. 2 resulted in a reduction of 43 (LP521) and 74% (LP2541) in plant grain yield, and of 33 (LP521) and 65% (LP2541) in kernel number per plant.
Grain yield was not related to the other grain yield components (i.e., prolificacy and kernel weight) in Exp. 1, but 59% and 24.8% of the variation in grain yield of Exp. 2 was explained by prolificacy (Fig. 3a ) and kernel weight, respectively. Kernel weight, like HI, were less affected by N availability than kernel number (Table 3). Differences in HI were strongly related to kernel number per plant (r2 = 0.79; n = 42; P < 0.001). The HI increased at a rate of 0.0012 per each additional kernel set up to a threshold of 272 kernels plant–1, above which it tended to stabilize at a constant value of 0.368 (Fig. 3b). Data from Exp. 1 were all below this threshold.
|
The number of kernels per plant did respond to variations in PGR around silking, but inbreds were classified in two groups on the basis of differences in the response pattern. A curvilinear response of kernel number per plant to PGR was established for the group represented by inbred P578 (Fig. 4a ), and the only effect of N availability on kernel number was through changes in the independent variable. Inbreds B100, LP521, and LP611 shared this response pattern. The second group, represented by LP2, was characterized by a linear relationship for the explored data range (Fig. 4b). The distinctive aspect of this group, however, was that the effect of N on final kernel number was not exclusively explained by its effects on PGR. In this group, shared with genotypes ZN6, LP2541, and LP662, kernel set per unit PGR was lower when plants were cropped at low N than when cropped at high N.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
An important result of the present work was the assessment of the start of N effects on leaf area production in inbreds (V10–V17), which matched information obtained for hybrids by many authors (Cox et al., 1993; Lemcoff and Loomis, 1986; Muchow, 1988; Novoa and Loomis, 1981) but appeared to be delayed with respect to other reported data (Uhart and Andrade, 1995). This apparent controversy among results from different sources disappears when the start of N effects on LAI is referred to as emerged leaf area per plant instead of a leaf stage (i.e., Vn). Even under the severe N restrictions represented by initial soil levels of 10 to 40 kg N-NO3 per ha, significant differences between N treatments never occurred before the expansion of approximately 3000 cm2 of green leaf area per plant (i.e., LAI = 2.1 in the present experiments), as it can also be observed in data reported by Uhart and Andrade (1995) or by Muchow (1988).
In spite of the agreement between inbreds and results from hybrids in the initial response to N of leaf area production, no treatment reached the critical LAI in our experiments, with the concomitant disadvantage for light capture. Maximum light interception values registered in this study were low as compared with records obtained for hybrids cropped in a similar environment (Maddonni et al., 2001) but represented an expected level for inbred lines because of their usually reduced leaf growth (Rasse et al., 2000). The negative effects of low N availability on leaf senescence additionally conditioned light interception efficiency, particularly after flowering. Relative reductions in postflowering leaf senescence, however, decreased under poor kernel set (i.e., Exp. 1), evidence of a control exerted by the postflowering source–sink ratio on leaf area duration (Borrás et al., 2003). This control should alert breeders to the importance of defining comparable source–sink ratios when selecting for valuable attributes like leaf area duration or the stay-green phenotype because of the effect of this physiological aspect on their expression.
Reduced N availability resulted in greater reductions in biomass production and PGR around silking than data available for hybrids (Muchow and Davis, 1988; Uhart and Andrade, 1995). These effects were more related to the above-mentioned reductions in leaf expansion and light interception efficiency than to the decline in RUE. Radiation use efficiency levels detected for maize inbreds were within the range of records available for hybrids (Sinclair and Muchow, 1999), and a similar comment applies to differences observed in this trait along the cycle (Otegui et al., 1995b) and in response to N deficiency (Uhart and Andrade, 1995). Our results support the idea that reduced biomass production usually observed in inbred lines is more related to the negative effects of inbreeding on leaf expansion and light interception than to photosynthesis levels (Tollenaar et al., 2004). Abiotic stresses, such as low N availability, intensify these trends.
Grain Yield Determination
Plant grain yield was severely reduced by N deficiencies, and the response was mainly related to variations in kernel number per plant. Grain yield was less responsive to variations in kernel weight and HI promoted by N supply, but there was a strong year and inbred effect on these relationships, as observed by Lafitte and Edmeades (1994) for inbreds developed for low N soils in tropical environments. Variations in HI were mostly due to variations in kernel number per plant, particularly at low kernel number values like those registered in Exp. 1. The slight decrease in HI under reduced N availability registered for most inbreds (range between 0 and 15%) may be indicative of a primary effect of N supply on shoot biomass accumulation rather than on biomass partitioning to the ear. This general trend, however, did not apply to inbreds ZN6, LP2, LP2541, and LP662, which exhibited harvest index reductions between 16 and 22%. This difference in the HI response to N supply between groups of inbreds may be the final result of the particular response of kernel number per plant to PGR around silking. Inbreds of the first group (i.e., slight reduction of the index) were those for which the effect of N on kernel number was exerted exclusively through changes in PGR (Fig. 4a). For inbreds in the second group mentioned above, an additional effect of N availability was detected as reduced kernel set at a given PGR (i.e., reduced reproductive efficiency evidenced in the ordinates of Fig. 4b). Thus, biomass allocation to the reproductive sink in this second group did not depend exclusively on the inherent biomass partitioning pattern of the genotype around silking, but appeared to be additionally controlled by N availability. This finding has not been reported previously for maize crops grown in the field in response to an abiotic stress, and supports data obtained in vitro by Below et al. (2000), who mentioned direct effects of N deficiencies on spikelet growth and kernel set. Differences in kernel set per unit PGR around silking have been reported for hybrids (Echarte et al., 2004; Luque et al., 2006; Tollenaar et al., 1992), but no differences from the general response model were identified for N and water deficit data obtained with one hybrid studied by Andrade et al. (2002). On the other hand, Abbate et al. (1995) detected a reduced efficiency for converting ear dry matter at anthesis to reproductive sinks (i.e., kernels) under N deficient conditions in wheat field crops, which supports our findings for the observed effect of N on the relationship between kernel number and PGR around silking in some maize inbreds.
For all inbreds, final kernel number was always smaller than the potential number identified as spikelet number per apical ear, independently of reduced N. availability. This potential number, however, was slightly reduced (range between 10 and 25%) at the N deficiency levels tested in this work. This trend suggests that delayed silk appearance rate from the husks (Cárcova et al., 2000) and increased kernel abortion (Otegui et al., 1995a) may be the main determinants of final kernel set under N stress. The only exception to this general pattern corresponded to inbred LP2541, which exhibited a dramatic reduction (75%) in potential kernel number (i.e., in ear morphogenetic processes) linked to the early arrest in ear development under N stress.
Finally, inbreds reported here had long ASI values because of the strong genetic control of this trait, which partially masked the negative effects of reduced N. Lack of response of kernel number per plant to the ASI in Exp. 1 was probably due to the unusually high temperatures registered between silking and R2 in this experiment, which may have promoted very poor kernel set (Schoper et al., 1986) independently of the ASI duration and the effects of N on this attribute. These facts determined a weaker relationship between these variables than previously reported (Bolaños and Edmeades, 1993). A similar analysis is valid for prolificacy, which is also considered a valuable secondary trait in maize breeding because of its strong relationship with final kernel number and grain yield (Bolaños and Edmeades, 1993). In summary, the ASI will be a good predictor of variations in kernel number per plant or prolificacy provided factors not determinant of ear growth (e.g., heat shock like in Exp. 1) do not impair ovary fertilization and kernel set.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Received for publication July 11, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. A. Lee and M. Tollenaar Physiological Basis of Successful Breeding Strategies for Maize Grain Yield Crop Sci., December 18, 2007; 47(Supplement_3): S-202 - S-215. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Carcova and M. E. Otegui Ovary Growth and Maize Kernel Set Crop Sci., May 31, 2007; 47(3): 1104 - 1110. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
