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


     

Published online 19 March 2008
Published in Crop Sci 48:666-677 (2008)
© 2008 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tanaka, W.
Right arrow Articles by Maddonni, G. A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Tanaka, W.
Right arrow Articles by Maddonni, G. A.
Agricola
Right arrow Articles by Tanaka, W.
Right arrow Articles by Maddonni, G. A.
Related Collections
Right arrow Crop Physiology & Metabolism
Right arrow Seed Quality

Pollen Source and Post-Flowering Source/Sink Ratio Effects on Maize Kernel Weight and Oil concentration

Walter Tanaka* and Gustavo Angel Maddonni

Dep. de Producción Vegetal, Fac. de Agronomía, Univ. de Buenos Aires, Av. San Martín 4453, Ciudad de Buenos Aires (C1417DSE), Argentina. Financial support from the National Council for Research (CONICET. PIP 5540). G.A. Maddonni is a member of and W. Tanaka has a scholarship from CONICET

* Corresponding author (wtanaka{at}agro.uba.ar).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Previous studies have documented pollen effect on maize (Zea mays L.) kernel oil concentration and the stability of this trait for a wide range of post-flowering source/sink ratios. Few efforts, however, have been devoted to establishing the functional relations among pollen source, source/sink ratio, and kernel oil concentration. Kernels of a normal oil content hybrid (DK752) self-pollinated and those of the same hybrid but pollinated with a high oil pollinator genotype (5MG) were sampled during the grain-filling period to evaluate the effect of different post-flowering source/sink ratios on kernel and embryo growth dynamics and oil deposition in the embryos. Final weight of kernels and embryos were related to post-flowering source/sink ratio, but embryo oil concentration was not modified. Pollen source affected both embryo weight (~31 and 41 mg for DK752xDK752 and DK752x5MG, respectively) and embryo oil concentration (~330 and 380 g kg–1 for DK752xDK752 and DK752x5MG, respectively). Final weight of kernels and embryos were closely related to variations in their growth rates (R2 = 0.79–0.82). The robust relationship between embryo growth rate and kernel growth rate determined the steady embryo/kernel ratio (~12.6 and 16.1% for DK752xDK752 and DK752x5MG, respectively) and kernel oil concentration of each cross (~68 and 93 g kg–1 for DK752xDK752 and DK752x5MG, respectively).

Abbreviations: EL, leaf subtending the ear • V3, three-ligulated leaf stage


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
CURRENT MAIZE (Zea mays L.) production in Argentina is approximately 16 million tons, and more than 70% of the annual marketable maize production is exported as grain, while less than 30% is locally industrialized. Hence, during the last decades, breeding efforts in Argentina were mainly focused on grain yield increase (Echarte et al., 2000; Luque et al., 2006), and less attention has been given to kernel composition (i.e., starch, oil, and protein concentration). Maize kernels, however, are used in several industrial processes that depend on kernel characteristics (e.g., endosperm vitreousness, kernel friability, and milling characteristics) and kernel composition. Thus, in feed rations of livestock and poultry, maize kernels with a high oil concentration are preferred because of their energy value and as a substitute for animal fats (Thomison et al., 2003).

Traditional maize hybrids, commonly cultivated in Argentina, yielded kernels with an oil concentration of about 50 g kg–1 (Maddonni and Otegui, 2006). In 2001 Dekalb-Monsanto Argentina S.A. introduced a high oil maize production system based on pollen-source effect on kernel composition (Xenia effect; Bulant and Gallais, 1998). The pollen parent affected kernel oil concentration without modifying kernel weight (Letchworth and Lambert, 1998). The high oil maize production system involved a blend of two types of genotypes, a high proportion (~90%) of male-sterile high-grain yield traditional hybrid (oil concentration 45–55 g kg–1) and a low proportion (~10%) of a male-fertile, high-oil hybrid (120–150 g kg–1) used as pollinator. As a result, cropped kernels exhibited an oil concentration in the range of 60 to 70 g kg–1 (Lambert et al., 1998).

In an initial effort to understand the Xenia effect on the increased kernel oil concentration, Lambert et al. (1998) attributed this fact to both a small increase of embryo weight and embryo oil concentration, and to a lower proportion of endosperm in the kernel. These results suggested an additive or dominant genetic effect on both embryo growth and oil deposition in the embryo, traits not characterized in the mentioned work.

Lower yields of the high oil maize production system were associated with failures in pollination, an inferior breeding effort applied to high oil pollinators, and to the high energy required for oil synthesis (Lambert et al., 1998). The stability of paternal pollen effect on kernel oil concentration, however, was confirmed for a wide range of environments (combination of years, locations, and sowing dates). Averaged across environments, oil levels in kernels were 31 g kg–1 higher in the blend than for the control hybrids (Thomison et al., 2003). To our knowledge, few efforts have been devoted to establishing functional relations among genotypes, environmental conditions, and kernel oil concentration.

In sunflower, seed oil concentration was modified by post-flowering environmental conditions (Connor and Hall, 1997). During this period brief intervals of high temperatures (Rondanini et al., 2003) and low irradiance values reduced seed oil concentration (Aguirrezábal et al., 2003). In maize, high temperatures ( >35°C) during the grain-filling period also affected kernel composition (Monjardino et al., 2005). In the main maize production area of Argentina (from 32 to 36°S, 58 to 64°W), however, the occurrence of high-temperature stress is not usual (Hall et al., 1992), but a shortage of assimilate availability per grain (post-flowering source/sink ratio), such as that promoted by low irradiance values or defoliation, may be reflected on both kernel weight and kernel composition (Borrás and Otegui, 2001; Borrás et al., 2002). Protein and starch contents (mg per kernel) were maximized at the same level of post-flowering source/sink ratio as kernel weight. A decrease in the source/sink ratio beyond this threshold promoted a decrease in the protein concentration and an increase in starch concentration. Kernel oil concentration, however, was not related to the post-flowering source/sink ratio suggesting a constant embryo/endosperm ratio (Paddick and Sprague, 1939) and embryo oil concentration (Ingle et al., 1965) for a wide range of kernel weights. We hypothesize that a dramatic reduction of the post-flowering source/sink ratio could lead to a similar kernel oil concentration of traditional and high oil hybrids due to the extra energy requirement of the latter for the synthesis of a high-energy compound (Penning de Vries, 1974). To test this hypothesis, kernels of (i) a traditional maize hybrid with normal oil content and (ii) those of the same hybrid but pollinated with a high oil pollinator, were sampled along the entire grain-filling period to evaluate the effect of different post-flowering source/sink ratios on the dynamics of both kernel and embryo growth and oil deposition in the embryo.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Experimental Design
Field experiments were conducted in Argentina during the growing seasons of 2004–2005 and 2005–2006 in Pergamino (34°56' S 60°34' W) on a silty clay loam soil (Typic Argiudoll). Seeds of the single-cross maize hybrid DK752 were hand-planted at 9 pl m–2 on 18 October 2004 and 28 October 2005. The experiments comprised 36 to 40 rows, 0.7m apart, and 36 m long. Seeds of a high oil concentration pollinator (5MG, Dekalb-Monsanto Argentina S.A.) were also sown on three successive planting dates with a 5 to 7 d interval to provide fresh pollen for the hand-pollinations.

The experiments were over-sown and thinned to the desired plant population at the three-ligulated leaf stage (V3; Ritchie et al., 1993). Weeds, insects, and diseases were adequately controlled. To minimize nitrogen restrictions, urea was applied at V4 (250 kg N ha–1). Water stress was prevented by means of sprinkler irrigation, with the soil near field capacity throughout the growing season.

The effect of two factors on kernel weight and kernel composition of DK752 was studied: the pollen source, and the post-flowering source/sink ratio. Both factors were arranged in a split-plot design with three replicates. The pollen source (DK752 and 5MG) was assigned to the main plot and the post-flowering source/sink ratio (three and five levels during 2004–2005 and 2005–2006, respectively) to the subplot. Each subplot was made up of five rows 0.70 m apart and 10 m long.

Pollination and Post-Flowering Source/Sink Ratios
Different post-flowering source/sink ratios were obtained by modifying the number of kernels per ear (the sinks) and the source size. The sink size was varied by different pollination timings and the source size was altered by defoliation and shading.

During 2004–2005 and 2005–2006, at least 50 plants per subplot were tagged approximately 15 d before silking and apical ears were bagged before silk emergence. Silking date (at least one silk visible) of the apical ear of each tagged plant was registered. Hand pollinations were performed by adding fresh pollen of each parental genotype to tagged plants of DK752. For pollen collection, tassels of both parental genotypes with visible anthers only in the main branch were bagged late in the afternoon, and sampled for pollen next morning. Three pollination treatments were applied: restricted, synchronized, and pseudo-natural pollination. Pollinations were performed 2 d and 4 d after the silking date of each plant of the restricted- and synchronized-pollination treatment, respectively. These treatments were performed to reduce (restricted pollination) or increase (synchronized pollination) kernel set. For the pseudo-natural pollination treatment, fresh pollen was daily added from 2 d to 6 d after silking date of each plant, mimicking open pollination conditions. Ears remained bagged after pollinations were performed to prevent pollen contamination. During 2005–2006, plants of 5MG were self-pollinated as was described above for the pseudo-natural pollination treatment.

During 2005–2006 the post-flowering source size was modified by shading and defoliation applied at the onset of the effective grain-filling period of plants from the pseudo-natural pollination treatment. Three subplots of each cross were shaded from approximately 12 d after silking to physiological maturity with black synthetic cloths 10 m long by 3.5 m wide stretched approximately 0.20 m above the crop on cane and wire structures. Hence, each structure comprised five rows, 10 m long. Measurements of incident solar radiation above the canopy with and without the black synthetic cloths indicated a reduction of incident solar radiation of about 50%. Defoliation consisted of ~50% leaf area reduction and was applied to all plants of the three subplots. About 12 d after silking, the leaf subtending the ear (EL) of each plant was identified and the lamina of the alternate leaves (from EL+1 to the topmost leaf and EL–1 to the lowermost leaf) were cut.

Measurements
During 2004–2005 and 2005–2006 five plants per subplot were used to characterize the post-flowering source/sink ratio of each treatment. The source size was quantified by the post-flowering plant leaf area duration and the aboveground plant biomass increase during the effective grain-filling period. While the former can be used as an estimator of the capacity of plants to intercept radiation, the latter quantifies the capacity of plants to both intercept radiation and transform the intercepted radiation into biomass. Plant leaf area was measured at R3 (~12 d after silking) using a nondestructive method. Individual leaf area was computed as lamina length x maximum lamina width x 0.75 (Montgomery, 1911). Plant leaf area was calculated as the sum of the area of all green leaves. Area of senesced leaves (i.e., when half or more of leaves had yellowed) was discounted from plant leaf area at R3, to obtain weekly values of plant leaf area until maturity. Post-flowering plant leaf area duration was the integral of the evolution of plant leaf area values on a thermal time basis (°Cd degree day units, with a base temperature of 8 °C; Ritchie and NeSmith, 1991). Plant-biomass increase during the effective grain-filling period of the same five plants per subplot was quantified as the difference between plant biomass at physiological maturity and plant biomass at R3, estimated by morphometric variables (Maddonni and Otegui, 2004). At physiological maturity each plant was individually sampled and kernel number per ear (i.e., the sink size) was counted. Post-flowering source/sink ratio was calculated as the quotient of plant-leaf area duration or above-ground biomass increase and kernel number per plant. These plants were also used to determine grain yield, kernel weight, and kernel composition (Near-infrared transmittance; Infratec 1227, Tecator, Sweden).

During 2004–2005 and 2005–2006, kernel and embryo growth dynamics and oil deposition during the effective grain-filling period were determined on kernels sampled from similar spikelet positions to avoid the variations of these traits along the ear (Lambert et al., 1967; Tollenaar and Daynard, 1978). The apical ears of three plants per subplot were harvested weekly from 15 d after silking to physiological maturity. At least 40 kernels per ear were sampled from spikelet positions 10 to 20 from the bottom of the ear. From this bulk of kernels, 10 were randomly selected and used to measure individual kernel weight and the others were dissected to extract the embryos. Samples were dried at 80°C until constant weight. Embryo/kernel ratio was calculated for each sampling date. Oil concentration was measured on both the entire kernels and the embryos by an exhaustive lipid extraction technique (Rondanini et al., 2003), using an initial solvent extraction (Hara and Radin, 1978) followed by a supercritical CO2 extraction (Am 3-96, AOCS, 1996) of the residual oil in the sample. The evolution of kernel weight, embryo weight, embryo/kernel ratio, and oil content along the grain-filling period was expressed on a thermal time basis.

Statistical Analyses
Results were subjected to analysis of variance (ANOVA) to evaluate the effects of treatments and their interaction on grain yield per plant, grain-yield components, kernel composition, final embryo/kernel ratio and the post-flowering source/sink ratio.

Bilinear (Eq. [1] and [2]) and tri-linear models (Eq. [1], [3], and [4]) with plateau were fitted by using the nonlinear routine of Table Curve V 3.0 (Jandel T.B.L.C.U.R.V.E., 1992).

Formula 1[1]

Formula 2[2]

Formula 3[3]

Formula 4[4]
Where a is the intercept, b and d the slopes, and c and e the breakpoints of the functions.

For each cross, pollination timing, and experiment, a bilinear model with plateau was fitted between kernel weight and thermal time after silking, where b and c parameters estimated grain-filling rate and the duration of the grain-filling period, respectively. The duration of the lag phase was calculated as a/b. The bilinear model with plateau was also used to describe the evolution of embryo growth, kernel oil weight, and embryo oil weight. A tri-linear model was fitted between embryo/kernel ratio and thermal time after silking. The b parameter estimated the rate of the first phase, c and e the breakpoint between both phases, and d the rate of the second phase. For DK752xDK752 and DK752x5MG bi- and tri-linear models with plateau were fitted to the evolution of kernel oil concentration and embryo oil concentration, respectively. For 5MGx5MG bi- and tri-linear models with plateau were fitted to the evolution of embryo oil concentration and kernel oil concentration, respectively.

Treatment effects on model parameters were tested by ANOVA. When model parameters were not affected by treatments, models were re-fitted to the whole data set.


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Grain Yield, Grain-Yield Components, and Kernel Composition
Grain yield per plant and grain-yield components of DK752 were not affected by pollen source (Table 1 ); in contrast, grain yield per plant decreased (P < 0.06–0.001) by both reductions of the sinks (i.e., restricted-pollination treatment) and of the post-flowering source size (i.e., shading and defoliation) (Table 1). Reductions of grain yield per plant were promoted by pollination timing effect on kernel number per plant (P < 0.001). The lowest kernel number was recorded on plants of the restricted-pollination treatment (~25% and 50% kernel number reduction, in comparison with those of the pseudo-natural pollination treatment, during 2004–2005 and 2005–2006, respectively). During 2005–2006, shading and defoliation imposed 12 d after silking not only reduced the post-flowering source size (~50%), but also kernel number per plant (~23 and 13% of kernel number reduction, in comparison with those of the pseudo-natural pollination treatment, for shading and defoliation, respectively). Consequently, kernels of the restricted-pollination treatment were growing under higher (P < 0.1–0.01) post-flowering source/sink ratios (Table 1). Their larger kernel weight (~19 and 23% heavier than those of the pseudo-natural pollination treatment during 2004–2005 and 2005–2006, respectively) was not fully compensated by their lower kernel number. During 2005–2006, kernels of the restricted pollination treatment exhibited the highest (P < 0.001) protein concentration (~110 vs. 94 g kg–1) but the lowest (P < 0.05) oil concentration (~75 vs. 83 g kg–1).


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

  		Table 1. Post-flowering source/sink ratio, grain yield per plant, grain-yield components (kernel number per plant and kernel weight) and kernel composition (oil, protein, and starch concentration) of maize hybrid DK752 hand-pollinated by two pollinators (DK752 and 5MG) at different pollination timings during 2004–2005 and 2005–2006. In the second experiment some plants from the pseudo-natural pollination treatment were defoliated (Pseudo-naturalD) and shaded (Pseudo-naturalS) after silking.

 
In contrast to that mentioned above for grain-yield components, pollen source only affected (P < 0.01) kernel composition (Table 1). For all pollination treatments, pollen from 5MG increased kernel oil concentration (~93 and 68 g kg–1 for DK752x5MG and DK752xDK752, respectively), without modifying protein concentration. Starch concentration was only reduced by pollen of 5MG (~573 and 544 g kg–1 for DK752xDK752 and DK752x5MG, respectively) during 2005–2006.

Pollen-source x pollination-timing interactions on grain yield, grain yield components, and kernel composition were not recorded.

Kernel and Embryo Weight, Embryo Oil Concentration, and Post-Flowering Source/Sink Ratio
Modifications in the number of reproductive sinks per plant and in the source of assimilates during the effective grain-filling period established a wide range of post-flowering source/sink ratios (Table 1). Independently of the nature of the pollen, source/sink ratio explained 77 to 80% of the variation in the mean weight of all kernels of the ear (i.e., mean kernel weight). Bilinear models with plateau were fitted to mean kernel weight and post-flowering source/sink ratio, the latter variable quantified as plant leaf area duration per kernel (Fig. 1A ), or as plant weight gain per kernel (Fig. 1B). Values of model parameters were significant (P < 0.001) and differed from zero (P < 0.01). The ordinate of both fitted functions, i.e., the extrapolated mean kernel weight for a post-flowering source/sink ratio close to zero, was approximately 175 mg k–1. Almost all data points, with the exception of some points of the restricted pollination treatment, coincided with the responsive part of these relationships. The slope of the function fitted to mean kernel weight and plant weight gain per kernel was lower (P < 0.01) than 1, indicating an increase of about 42% mean kernel weight per each mg of post-flowering plant weight gain per kernel. Maximum mean kernel weights (~302 mg k–1) were recorded for post-flowering source/sink ratios ≥ 312 mg k–1, or ≥ 1.54 m2 °Cd k–1.


Figure 1
View larger version (19K):
[in this window]
[in a new window]

  		Figure 1. Mean kernel weight (A and B) and weight of kernels located at the middle of the ear (C and D) as a function of post-flowering source/sink ratio quantified as plant leaf area duration per kernel (A and C) or plant weight gain per kernel (B and D). Symbols represent mean values of each pollination treatment of both pollinators (DK752 and 5MG). Lines indicate the models fitted to the data. For A: Kernel weight = 186.4 + 79.9 source/sink ratio (r2 = 0.80, n = 16, P < 0.001), for source/sink ratio < 1.54 m2 °Cd k–1, Kernel weight = 309.5 mg, for source/sink ratio ≥ 1.54 m2 °Cd k–1. For B: Kernel weight = 163.6 + 0.42 source/sink ratio (r2 = 0.77, n = 16, P < 0.001), for source/sink ratio < 312 mg k–1, kernel weight = 295.6 mg, for source/sink ratio ≥ 312 mg k–1. For C: Kernel weight = 210.1 + 67.8 source/sink ratio (r2 = 0.56, n = 16, P < 0.001), for source/sink ratio < 1.37 m2 °Cd k–1, kernel weight = 302.9 mg, for source/sink ratio ≥ 1.37 m2 °Cd k–1. For D: Kernel weight = 201.5 + 0.29 source/sink ratio (r2 = 0.62, n = 16, P < 0.001), for source/sink ratio < 348.6 mg k–1, kernel weight = 302.9 mg, for source/sink ratio ≥ 348.6 mg k–1.

 
Pollination timing also affected (P < 0.001) the weight of kernels located at the middle of the ear; pollen source did not modify the weight of these kernels (Table 2 ). During 2004–2005 only, pollen from 5MG reduced (P < 0.05) kernel weight of synchronized pollination plants.


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

  		Table 2. Kernel and embryo growth dynamics (lag phase duration, rate and duration of the effective growth period, and final weight), and final embryo/kernel ratio of kernels located at the middle of the ear of maize hybrid DK752 hand-pollinated by two pollinators (DK752 and 5MG) at different pollination timings during 2004–2005 and 2005–2006. In the second experiment, some plants from the pseudo-natural pollination treatment were defoliated (Pseudo-naturalD) and shaded (Pseudo-naturalS) after silking.

 
Post-flowering source/sink ratio accounted for 56 to 62% of the variation in the weight of kernels located at the middle of the ear of the whole data set (Fig. 1C and 1D). Parameter values of the bilinear models fitted to kernel weight and post-flowering source/sink ratio were similar to those obtained from the functions fitted to the mean kernel weight and post-flowering source/sink ratio data set (Fig. 1A and 1B). These results collectively suggest that, independently of the pollen source, variations of the post-flowering source/sink ratio were reflected in the weight of kernels at any position of the ear.

Both the pollen source (P < 0.05–0.01) and the pollination timings (P < 0.001) affected the weight of embryos located at the middle of the ear (Table 2). During 2004–2005 only, embryo weight of the synchronized pollination plants was not affected by pollen source.

Oil concentration of embryos was only modified (P < 0.01) by the pollen source (~330 and 380 g kg–1 for DK752xDK752 and DK752x5MG, respectively; Table 3 ). When embryo weight values were expressed as relative to the maximum embryo weight of each cross (38 mg and 48.5 mg for DK752xDK752 and DK752x5MG, respectively), embryo weight variability of the whole data set was significantly (P < 0.001) related to post-flowering source/sink ratios (Fig. 2A and 2B ). Bilinear models with plateau were fitted to the mentioned variables. Relative embryo weight attained its maximum value (~1) above post-flowering source/sink ratios (1.32 ± 0.35 m2°Cd k–1 and 247 ± 32 mg k–1 for plant leaf area duration per kernel and plant weight gain per kernel, respectively), similar to those (1.37 ± 0.32 m2°Cd k–1 and 348.6 ± 67 mg k–1 for plant leaf area duration per kernel and plant weight gain per kernel, respectively) which maximized kernel weight. As pollination timing did not affect embryo oil concentration (Table 4 ), variations of the post-flowering source/sink ratio were not reflected on embryo oil concentration of each cross (Fig. 2C and 2D).


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

  		Table 3. Embryo oil accumulation dynamics (lag phase duration, rate, duration of the effective oil accumulation period, and final embryo oil weight), and final oil concentration in embryos and kernels located at the middle of the ear of maize hybrid DK752 hand-pollinated by two pollinators (DK752 and 5MG) at different pollination timings during 2004–2005 and 2005–2006. In the second experiment, some plants from the pseudo-natural pollination treatment were defoliated (Pseudo-naturalD) and shaded (Pseudo-naturalS) after silking.

 

Figure 2
View larger version (18K):
[in this window]
[in a new window]

  		Figure 2. Relative embryo weight (A and B) and relative embryo oil concentration (C and D) of kernels located at the middle of the ear as a function of post-flowering source/sink ratio quantified as plant leaf area duration per kernel (A and C) or plant weight gain per kernel (B and D). Symbols as in Fig. 1. Lines indicate the models fitted to the data. For A: Relative embryo weight = 0.70 + 0.23 source/sink ratio (r2 = 0.49, n = 16, P < 0.01), for source/sink ratio < 1.32 m2 °Cd k–1, relative embryo weight = 1, for source/sink ratio ≥ 1.32 m2 °Cd k–1. For B: Relative embryo weight = 0.55 + 0.0017 source/sink ratio (r2 = 0.62, n = 16, P < 0.001), for source/sink ratio < 247.7 mg k–1, relative embryo weight = 0.97, for source/sink ratio ≥ 247 mg k–1.

 

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

  		Table 4. Dynamics of embryo/kernel ratio, embryo oil concentration, and kernel oil concentration during the post-flowering period of kernels located at the middle of the ear of DK752 hand-pollinated by two pollinators (DK752 and 5MG) during 2004–2005 and 2005–2006. During 2005–2006 the dynamics of the 5MG self-pollinated were also included.

 
Kernel and Embryo Growth Dynamics
The evolution of the weight of kernels located at the middle of the ear was well described (R2 > 0.96, P < 0.001) by bilinear models with plateau (Table 2). Pollination timing affected the grain-filling rate (P < 0.01–0.001) without modifying the duration of both the lag phase (~202°Cd) and the effective grain-filling period (~641°Cd). In contrast, pollen source did not modify kernel growth dynamics. Hence, grain-filling rate accounted for 79% of the variation in kernel weight of the whole data set A).

Embryo growth dynamics were also significantly (R2 > 0.98, P < 0.001) described by bilinear models with plateau (Table 2). Both pollen source (P < 0.05–0.01) and pollination timing (P < 0.001) affected final embryo weight through modifications of the embryo growth rate. Embryos started the effective growth period at approximately 284°Cd after silking, and lasted approximately 603°Cd, when maximum embryo weight was attained. Embryos obtained from ovaries of DK752, pollinated with pollen of 5MG, exhibited a greater (P < 0.001) growth rate (~0.067mg °Cd–1) than those pollinated with DK752 (~0.057mg °Cd–1). Consequently, for the whole data set, modifications in final embryo weight were only related to (R2 = 0.86) changes in embryo growth rate during the effective embryo growth period (Fig. 3B ).


Figure 3
View larger version (15K):
[in this window]
[in a new window]

  		Figure 3. Kernel weight as a function of grain-filling rate (A), embryo weight as a function of embryo growth rate (B) and embryo oil weight as a function of embryo oil accumulation rate (C) of kernels located at the middle of the ear. Symbols as in Fig. 1. Lines indicate the models fitted to the data.

 
As variations of the post-flowering source/sink ratio had a similarly quantitative effect on both kernel and embryo growth rate (Fig. 4 ), final embryo/kernel ratio of each cross was stable, independently of final kernel weight (Table 2). In contrast, as pollen source only affected embryo growth (Table 2), kernels from DK752x5MG exhibited a greater (P < 0.05) embryo/kernel ratio (~16%) than those from DK752xDK752 (~12%). For each cross, embryo/kernel ratio increased along the entire grain-filling period (R2 = 0.97–0.99), but with two distinctive rates (Fig. 5A and Table 4), a higher rate from 230°Cd to 450°Cd and a lower one from 450°C to about 870°Cd (i.e., physiological maturity). Pollen source affected (P < 0.1–0.05) both rates without modifying the duration of the periods (Table 4). For 5MGx5MG (Fig. 5A), kernels exhibited the highest rates and highest final embryo/kernel ratio (~20.4%).


Figure 4
View larger version (13K):
[in this window]
[in a new window]

  		Figure 4. Variations of embryo growth rate as a function of variations of grain-filling rate. For both variables, values represent the variations (%) of each treatment from those of the pseudo-natural pollination treatment at each growing season. Symbols as in Fig. 1. Line indicates the model fitted to the data.

 

Figure 5
View larger version (16K):
[in this window]
[in a new window]

  		Figure 5. Evolution of embryo/kernel ratio (A), embryo oil concentration (B) and kernel oil concentration (C) of kernels located at the middle of the ear, from DK752xDK752, DK752x5MG, and 5MGx5MG crosses. Lines indicate the models fitted to the data. For embryo/kernel ratio: Embryo/kernel ratio = a + b TT, for TT ≤ c; embryo/kernel ratio = a+ bc +d (TT – c), for TT > c and TT ≤ e; Embryo/kernel ratio = a +bc+ d (e-c), for TT > e. For DK752xDK752: a = –9, b = 0.039, c = 461, d = 0.0084, and e = 868 (R2 = 0.98, n = 77); for DK752x5MG: a = –11.5, b = 0.050, c = 441, d = 0.012, and e = 901 (R2 = 0.98, n = 77); for 5MGx5MG: a = –12.9, b = 0.058, c = 461, d = 0.017, and e = 846 (R2 = 0.99, n = 10). For embryo oil concentration of Dk752xDK752 and DK752x5MG: Embryo oil concentration = a + b TT, for TT ≤ c; embryo oil concentration = a + bc +d (TT – c), for TT > c and TT ≤ e; embryo oil concentration = a + bc + d (e – c), for TT > e; For DK752xDK752: a = 206.5, b = 0.38, c = 546, d = –0.24, and e = 912 (R2 = 0.75, n = 76); for DK752x5MG: a = 232.3, b = 0.41, c = 536, d = –0.23, and e = 878 (R2 = 0.67, n = 76). For embryo oil concentration of 5MGx5MG: Embryo oil concentration = 175.5 + 0.61 TT, for TT < 448°Cd; embryo oil concentration = 448 mg g–1, for TT ≥ 448°Cd. For kernel oil concentration of K752xDK752 and DK752x5MG: Kernel oil concentration = a + b TT, for TT < c. Kernel oil concentration = a + bc, for TT < c. For DK752xDK752: a = 20.4, b = 0.057, and c = 531 (R2 = 0.67, n = 77); for DK752x5MG: a = 7, b = 0.118, and c = 548 (R2 = 0.87, n = 77). For kernel oil concentration of 5MGx5MG: Kernel oil concentration = a + b TT, for TT ≤ c. Kernel oil concentration = a + bc + d (TT – c), for TT > c and TT ≤ e. Kernel oil concentration = a + bc + d (e – c), for TT > e where a = 6.8, b = 0.159, c = 542, d = 0.042, and e = 923 (R2 = 0.96, n = 10).

 
Oil-Accumulation Dynamics
The evolution of oil accumulation in the embryos was well described (R2 > 0.94, P < 0.001) by bilinear models with plateau functions (Table 3). Both pollen source and pollination timing affected (P < 0.01–0.001) parameters of the dynamics, and variations of final embryo oil weight were closely (R2 = 0.70) related to the rate of embryo oil accumulation (Fig. 3C). Interestingly, the effective period of oil accumulation in the embryo (~545°Cd, Table 3) was shorter than that of kernel growth (~641°Cd, Table 2). In contrast, the lag phase of kernel growth (~201°Cd, Table 2) was shorter than that of embryo oil accumulation (~280°Cd, Table 3). Final embryo oil concentration was only modified (P < 0.05) by pollen source (~329 and 374 g kg–1 for DK752xDK752 and DK752x5MG, respectively, Table 3). Pollen source effect on this trait, however, was of a lower magnitude (~18 and 12% increase during 2004–2005 and 2005–2006, respectively) than that promoted on embryo/kernel ratio (~26 and 29% increase during 2004–2005 and 2005–2006, respectively). For each cross, embryo oil concentration increased during 550°Cd after silking (Table 4), when maximum embryo oil concentrations were attained (Fig. 5B). At this stage, embryos of DK752xDK752 had the lowest (P < 0.10) oil concentration (~410 g kg–1). From 550°Cd to physiological maturity, embryo oil concentration of DK752xDK752 and DK752x5MG decreased with a similar rate (Table 4). In contrast, embryo oil concentration of 5MGx5MG remained close to the maximum value attained at 550°Cd.

The different dynamics of embryo/kernel ratio and embryo oil concentration among crosses determined the differential evolution of kernel oil concentration during the grain-filling period (Fig. 5C and Table 4). At early stages of kernel growth, kernel oil concentration was increased by an increase of both embryo/kernel ratio and embryo oil concentration. At 520 to 540°Cd from silking, kernels of DK752xDK752 (~52 g kg–1) and DK752x5MG (~76 g kg–1) attained their maximum oil concentration. In contrast, kernels of 5MGx5MG also increased their oil concentration from 540°Cd to physiological maturity but with a lower rate (0.048 g kg–1 °Cd–1).


   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Our experimental conditions allowed us to explore pollen source and post-flowering source/sink ratio effects on maize kernel weight and kernel oil concentration. Manipulations of both the source (shading, defoliations) and the sinks (pollination timings) established a wide range of post-flowering assimilate availability per kernel, which varied from 30 to 235% compared to that of the control (i.e., pseudo-natural pollination treatment). As was previously reported (Borrás and Otegui, 2001; Borrás et al., 2002; Borrás et al., 2004), (i) variations in the post-flowering source/sink ratio accounted for changes in kernel weight without modifying kernel oil concentration, (ii) kernel weight decreased (~9%) whenever source/sink ratio was reduced, (iii) kernel weight increased (~22%) in response to enhanced source/sink ratio up to a saturation value (~310 mg k–1) above which kernels attained the maximum weight (~300 mg), and (iv) kernel weight was similarly affected at any positions along the ear. These results collectively indicate that under most tested conditions, kernel weight was source limited, but kernel oil concentration was unaffected.

Differences among genotypes in kernel weight response to source/sink ratio were previously established by comparisons of parameter values (i.e., slope, source/sink ratio threshold, and maximum kernel weight) of the mentioned relationship (Borrás and Otegui, 2001). To our knowledge the effect of pollen source on this relationship was not analyzed. In our work, pollen source affected kernel oil concentration (~93 and 68 g kg–1, for DK752x5MG and DK752xDK752, respectively) without modifying kernel weight. Hence, the DK752xDK752 and DK752x5MG crosses exhibited similar kernel weights at any source/sink ratio, indicating that the relationship between kernel size and source/sink ratio was not canceled out by the different caloric content of kernels. From Penning de Vries (1974) we have estimated how much glucose was required theoretically for the synthesis of 1 g of DK752xDK752 (~68, 86, and 600 g kg–1 of oil, protein, and starch concentration, respectively) and DK752x5MG (~98, 89, and 573 g kg–1 of oil, protein, and starch concentration, respectively) kernel. The small difference in the glucose requirement (~1.37 and 1.44 g glucose per 1 g DK752xDK752 and DK752x5MG kernel, respectively) could be supported by the stored non-structural carbohydrates (Kiniry et al., 1992), or by a higher post-flowering assimilate production promoted by the increased sink demand (Reynolds et al., 2000). In our work, pollen source did not modify either plant leaf area duration (~338 and 335 m2 °Cd pl–1 for DK752xDK752 and DK752x5MG, respectively) or post-flowering biomass production per plant (~95 and 93 g pl–1 for DK752xDK752 and DK752x5MG, respectively), suggesting a differential contribution of pre-flowering stored non-structural carbohydrates to sustain kernel growth. Hence, for both crosses, the post-flowering source/sink ratio estimators similarly quantified the assimilate availability per kernel imposed by pollination treatments.

Since kernel oil concentration is mainly located within the embryos (Watson, 1987), we have also analyzed embryo growth and oil deposition in the embryos under contrasting post-flowering source/sink ratios. Despite the pollen effect on embryo growth, variations of the post-flowering source/sink ratio accounted for changes in final embryo weight of both crosses. Interestingly, embryo weight was maximized at similar source/sink ratio values to those above which kernels attained the maximum weight. Moreover, for each cross, post-flowering source/sink ratio had a similar quantitative effect on both grain-filling rate and embryo growth rate, and did not affect kernel or embryo growth duration. In a recent work (Gambín et al., 2006), differences in maize kernel weight due to genotypes or environments were related to the source/sink ratio established during the very early stages of grain-filling, and these differences were associated with changes in kernel growth rate during the effective grain-filling period. In contrast, environmental conditions did not affect grain-filling duration. Based on these results, post-flowering source/sink ratio effects on grain-filling rate were probably related to the different source/sink ratio established around silking by both manipulations of the sinks (pollination timings), and also the source (shading and defoliation). For shading and defoliation treatments, we had tried to modify the post-flowering source without altering sink number, to produce a drastic reduction of the post-flowering assimilate availability per kernel. Under severe source limitations we expected to modify both grain-filling rate and grain-filling duration (Echarte et al., 2006; Sala et al., 2007), and also to promote a differential hierarchy between the major kernel tissues (i.e., endosperm and embryo) for assimilate supply. Unfortunately, shading and defoliation were imposed at about 200°Cd, i.e., within the lag phase (246.5°Cd) of the tested genotype (Borrás and Otegui, 2001). Any reduction of assimilate availability within this early post-silking period affected kernel set (Otegui and Bonhomme, 1998), counterbalancing the reduction of the source.

The robust relationship between embryo growth rate and grain-filling rate helped us to understand the steady embryo/kernel ratio of each cross (~12.6 and 16.1% for DK752xDK752 and DK752x5MG, respectively) for the wide range of kernel weights, which supports previous evidences of the constant embryo/endosperm ratio of maize kernels (Paddick and Sprague, 1939). This result shows the importance of the period around flowering not only in the definition of kernel number per plant (Andrade et al., 1999; Echarte et al., 2000; 2004; Luque et al., 2006, Pagano and Maddonni, 2007), and kernel weight (Gambín et al., 2006), but also in the steady embryo/kernel ratio. Under an ecological framework, we speculate that maize plants would have the capacity to adjust the fitness of the next generation (i.e., the number of individuals, the balance-sheet between seed reserves and embryo size) from environmental conditions around silking (Sadras, 2007). A drastic post-flowering source reduction, however, could cancel out the stability of the embryo/kernel ratio, if an interruption of the grain filing period occurs during the 450°Cd from silking.

Pollen-source effect on both embryo growth pattern and oil deposition in the embryos, support the relevant evidence of the additive or dominant gene action of the male gametophyte on these traits (Lambert et al., 1998) and indicates a differential partition of assimilates among kernel tissues (i.e., pericarp, endosperm, and embryo). Interestingly, the embryo oil concentration dynamics revealed the existence of late (from 500°Cd to physiological maturity) deposition compounds different from oil, which reduced embryo oil concentration. A detailed analysis of starch, protein, and oil deposition within the embryos, together with histological studies of kernels, would assist in understanding differences between crosses in embryo size and embryo oil concentration dynamics.

Finally, the stability of both embryo/kernel ratio and embryo oil concentration under different post-flowering source/sink ratios helped us to understand previous evidence of the high homeostasis of kernel oil concentration (Borrás et al., 2002; Thomison et al., 2003). In contrast to previous studies (Alexander and Lambert, 1968; Lambert et al., 1998), our findings suggest a possible increase of kernel oil concentration without reductions of kernel weight and grain yield.


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Kernel oil concentration was determined by the embryo/kernel ratio and the embryo oil concentration. Pollen source affected the temporal evolution of both traits and this effect was not suppressed by the restrictive post-flowering conditions tested in our experiments. Hence, a reason other than a source limitation for a high energy compound would produce the documented grain-yield reductions of the high oil maize production system.


   ACKNOWLEDGMENTS
 
The authors wish to thank to F. Vartorelli, D. Rondanini, D. Ravetta, M.E. Otegui, L. Blanco, E. Pagano, and A. Cirilo for their valuable help and E. Whitechurch for the revision of English style. This work was supported by the National Council for Research (CONICET, PIP 5440) and Dekalb-Monsanto Argentina. W. Tanaka has a scholarship from and G.A. Maddonni is a member of CONICET.


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

Received for publication August 14, 2007.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Crop Sci.Home page
W. Tanaka and G. A. Maddonni
Maize Kernel Oil and Episodes of Shading during the Grain-Filling Period
Crop Sci., October 22, 2009; 49(6): 2187 - 2197.
[Abstract] [Full Text] [PDF]

Home page
 ANN BOT (LOND)Home page
W. Tanaka, A. I. Mantese, and G. A. Maddonni
Pollen source effects on growth of kernel structures and embryo chemical compounds in maize
Ann. Bot., August 1, 2009; 104(2): 325 - 334.
[Abstract] [Full Text] [PDF]

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

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