Crop Science Grow Your Career with CSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

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 Related articles in Crop Science
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
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 ISI Web of Science (10)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by de la Vega, A. J.
Right arrow Articles by Hall, A. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by de la Vega, A. J.
Right arrow Articles by Hall, A. J.
Agricola
Right arrow Articles by de la Vega, A. J.
Right arrow Articles by Hall, A. J.
Related Collections
Right arrow Crop Cytology
Crop Science 42:1202-1210 (2002)
© 2002 Crop Science Society of America

CROP ECOLOGY, MANAGEMENT & QUALITY

Effects of Planting Date, Genotype, and Their Interactions on Sunflower Yield

II. Components of Oil Yield

Abelardo J. de la Vega* and Antonio J. Hall

IFEVA, Facultad de Agronoma, Universidad de Buenos Aires/CONICET, Av. San Martn 4453, (1417) Buenos Aires, Argentina

* Corresponding author (avega{at}waycom.com.ar)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Sunflower (Helianthus annuus L.) yields are strongly reduced by late sowing dates. The objectives of this study were to investigate the physiological bases of the sowing date (S), genotype (G), and genotype x sowing date (G x S) interaction effects on sunflower yield and contribute to development of ideotype-based selection strategies for improving yield at late plantings. Nine differentially adapted hybrids were evaluated during two seasons at normal and late planting dates at Venado Tuerto, Argentina, by means of the framework that defines oil yield as the product of grain number and weight and oil concentration. Sowing date was the major source of variation for oil yield. The G x S interaction, in turn, accounted for a portion of the total variability four times higher than the contribution of G. The S and G x S interaction effects on grain number were mostly determined by grain set in the central portion of the capitulum. Duration of grain filling was the main determinant of the S and G x S interaction effects on grain weight and grain oil concentration. Variation in grain oil concentration between sowing dates was largely due to changes in kernel oil proportion, rather than changes in kernel percentage. Relative changes in rate of grain filling and kernel percentage also contributed to the observed G x S interactions for grain weight and oil concentration, respectively. Although some degree of compensation among oil yield components underlie the observed G x S interactions, it was found that seed set in the central portion of the head can be used as an indicator of adaptation to late planting dates.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
SEVERAL STUDIES HAVE demonstrated that oil yield in sunflower is strongly reduced when normal sowing dates are delayed (Robinson, 1970; Johnson and Jellum, 1972; Unger, 1980; Beard and Geng, 1982; Andrade, 1995; Bange et al., 1997). As sunflower is frequently planted after the optimal period in several situations and growing regions, the formulation of breeding strategies to improve yield under these conditions is needed. Developing a physiological understanding of the yield reductions linked to late plantings and the associated G x S interactions (Robinson, 1970; Beard and Geng, 1982; Miller et al., 1984) could serve to identify traits useful for breeding programs targeted at overcoming this constraint to production. In a companion paper (de la Vega and Hall, 2002), we used the framework that defines yield as the product of total oil-corrected biomass and harvest index to investigate S and G x S interaction effects on crop performance for a reference set of differentially adapted sunflower hybrids. We found that both S and G x S effects on oil-corrected grain yield were mostly explained by variations in attributes and processes expressed postanthesis. The marked differences between planting dates for mean oil-corrected biomass at physiological maturity were the dominant determinants of the S effect on oil-corrected grain yield, and the genotype-specific responses for harvest index were the dominant determinant of the observed G x S interaction. We also found that the hybrids that improve their relative performance at late planting dates are characterized by canopy stay green, indicating that this trait can be used as putative indicator of adaptation to late planting dates.

In the present paper we analyze the results of the same experiments using the framework which views crop oil yield as the product of its primary components (i.e., grain number, grain weight, and grain oil content). Other authors (Robinson, 1970; Johnson and Jellum, 1972; Unger, 1980; Beard and Geng, 1982; Miller et al., 1984; Andrade, 1995; d'Andria et al., 1995; Bange et al., 1997) have reported the effects of late sowings on these yield components and have examined some S and some G x S interaction effects. However, these papers have neither examined the physiological bases of the observed changes in yield components, e.g., dynamics of grain growth, hull to kernel ratio, etc. (Robinson, 1970; Johnson and Jellum, 1972; Unger, 1980; Miller et al., 1984; Andrade, 1995; d'Andria et al., 1995) nor used a set of genotypes in which G x S interaction effects were not expressed (Bange et al., 1997); nor have they tried to link their results to crop improvement.

Here we examined the performance of nine differentially adapted sunflower hybrids, which exhibited significant G x S interactions under conditions of normal (October) and late (December) planting dates during 2 yr at Venado Tuerto, central Argentina (de la Vega and Hall, 2002). Our analyses of S, G, and G x S interaction effects encompass not only the primary components of oil yield, but also extend to the dynamics of grain growth, the role of floret differentiation and grain set in grain number determination, and the contributions of kernel percentage and kernel oil content to grain oil proportion. Thus, our present analyses extend the previous ones (de la Vega and Hall, 2002) and seek to explore issues not dealt with in other papers on this topic. The overall objective was to enhance the current understanding of variations observed for sunflower oil yield components associated with planting date and G x S interaction. We were also interested in searching for secondary traits that might enhance the formulation of ideotype-based selection strategies for improving yield at late plantings. In this context, our results suggest that the diameter of the empty center of the head may be a useful indication of adaptation to late planting.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Cultural Details
Supplementary irrigated crops of a reference set (Fox and Rosielle, 1982) of nine sunflower hybrids (Table 1) were grown on a deep coarse loam soil (Typic Hapludoll) at the Advanta Semillas Research Center, Venado Tuerto, Argentina (33° 41' S, 61° 57' W). Crops were sown in October (S1) and December (S2) during 1996-1997 (E1) and 1998-1999 (E2) seasons. These hybrids exhibit differential adaptation to northern and central regions of Argentina and to normal (S1) and late (S2) plantings at Venado Tuerto (de la Vega et al., 2001), as well a strong G x S interaction for oil-corrected grain yield (de la Vega and Hall, 2002). Further details of the experimental material, test environments and experimental design are given in de la Vega and Hall (2002).


View this table:
[in this window]
[in a new window]
  		Table 1. Agronomic characters of the sunflower hybrids evaluated at Venado Tuerto during 1996–1997 and 1998–1999 seasons, using October (S1) and December (S2) planting dates (mean values for two trials per sowing date). All genotypes are single-cross hybrids, developed by the sunflower breeding program of Advanta Argentina, except Morgan 734, provided by Dow-Morgan Argentina.

 
Yield and Its Components
All grain (achene, includes kernel and hull) yield data are presented at 110 g kg-1 moisture. Yield and yield components were determined by hand harvesting of 3.99 m2 (one central row, discarding the border plants; plot size of at least 8 rows x 6 m and interrow spacing of 0.70 m). Grain weight was determined by weighing and averaging three random 100-achene samples per plot, and grain number per square meter was calculated from the grain yield. Grain oil concentration was determined on a 10-g, oven-dried achene sample by nuclear magnetic resonance (Granlund and Zimmerman, 1975) with a Newport Analyzer (Newport-Oxford Instruments Ltd, Newport Pagnell, Buckinghamshire, England). Oil yield was calculated as a function of oil content and grain yield.

The number of disc flowers was estimated from four heads per plot sampled at physiological maturity, counting on 1/8 of the head surface. The same heads were used to measure the diameter of the empty center (lack of seed set). The dynamics of individual grain weight was followed during the entire grain-filling period. Samples were harvested every 3 to 4 d in two different portions of the head: upper periphery and intermediate portion. In E1 S1, E2 S2 and E2 S1, the center of the floral disk was also sampled. At each harvest, five randomly selected plants per plot were subsampled and three grains were harvested or three plants were subsampled and five grains were harvested from each position on the head. The grains were dried at 70°C for at least 48 h before weighing. Kernel percentage (i.e., kernel weight as a percent of total achene weight) was measured on the grains of the last harvest.

Statistical Analyses
The nonlinear routine of TBLCURVE (Jandel TBLCURVE, 1992) was used to fit piecewise bilinear regression models to individual grain weight/time relationships to estimate the rate and duration of grain filling, and final grain weight. A conditional model (Ploschuk and Hall, 1995) was used with a first stage where

[1]
for DAA < C, and a second stage where:

[2]
where a and b are the intercept and the slope, respectively, of the linear regression corresponding to the first stage, DAA is days after anthesis, and the constant C is the unknown breakpoint of the function indicating the end of grain filling. Functions were fitted to data from all replications and all sampled portions of the head per treatment.

Pooled analysis of variance over 2 yr (Y) and two planting dates (S), based on a fixed effects linear model, was performed to separate Y, S, G, and their interaction effects for all characters recorded. The statistical significance of Y, S, and Y x S was tested against the pooled error (a) (i.e., Y x S x rep), and that of G, G x S, G x Y, and G x S x Y, against the pooled error (b) (i.e., residual error). The partitioning of the treatment sums of squares was used as a measure of the relative contribution of each source of variation to the total variability for each attribute. Correlations among traits were computed using hybrid means for S1 and S2 across 1996-1997 and 1998-1999 seasons.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Oil Yield
Combined analysis of variance showed that sowing date affected oil yield (more in S1 than in S2, Tables 1, 2, and 3), accounting for almost 75% of total variability. The relative magnitude of the difference observed for oil yield between S1 and S2 varied between experiments (Table 2), resulting in a significant Y x S interaction (Table 3). Significant G and G x S effects for oil yield were also found, the sum of which were equivalent to about 67% of the variability due to the total of unpredictable effects (i.e., Y, Y x S, G x Y, and G x S x Y). The effect of G on oil yield was significant (Table 3), but it accounted for a portion of the total variability almost four times lower than the contribution of the G x S interaction (Table 3). This is consistent with the strong crossover interaction found between planting dates (Table 1). As observed for oil-corrected grain yield (de la Vega and Hall, 2002), G x S interaction for oil yield accounted for a larger portion of the treatment sums of squares than the unpredictable interactions involving G (i.e., G x Y and G x S x Y). The relative importance of the predictable interactions suggests that traits related to specific adaptation could be used in formulating breeding strategies to improve yield at late planting dates.


View this table:
[in this window]
[in a new window]
  		Table 2. Mean values for oil yield and its primary and secondary components in October (S1) and December (S2) planting dates at Venado Tuerto, during 1996–1997 and 1998–1999 seasons (mean values for nine hybrids).

 

View this table:
[in this window]
[in a new window]
  		Table 3. Partitioning of the treatment sums of squares and significances derived from the pooled analysis of variance for oil yield and its primary and secondary components of nine sunflower hybrids grown in four environments (Venado Tuerto, two seasons and two planting times). Values in the body of the table are in percent of total SS after removal of residuals.

 
Multiple regression analysis showed that variations in grain oil concentration accounted for 77% of the oil yield variation, while grain number and grain weight explained roughly equivalent proportions of the remaining variability (data not shown). The relative importance of the S, G, and G x S interaction effects on the primary components of yield (grain number, weight, and oil concentration) varied with the component (Table 3, Fig. 1) . Thus, G accounted for large proportions of total variability in grain number and weight, while S was predominant in its effects on grain oil concentration. Significant G x S interaction effects were found for all three primary components of oil yield (Table 3). The dashed arrows in Fig. 1 highlight the strength of these interactions, which suggest that it would be very difficult to find broad adaptation to both environment types via selection for any single component of yield in a single environment. In what follows, we examine S, G, and G x S effects on the primary components of oil yield and their secondary determinants.



View larger version (16K):
[in this window]
[in a new window]
  		Fig. 1. Relationships between oil yield and (A) grain number per square meter, (B) grain weight, and (C) grain oil concentration, for nine sunflower hybrids grown in October (S1) and December (S2) sowing dates at Venado Tuerto during 1996-1997 and 1998-1999. Values are means of the 2 yr. Dashed arrows highlight some contrasting genotypic responses to change from S1 to S2, and which underlie G x S interactions (Treatment 1, triangles; Treatment 9, circles). Genotype groups derived from cluster analysis of a multienvironment trial (de la Vega et al., 2001), i.e., G1: Northern adapted, G2: Central adapted, G3: broadly adapted. See Table 1 in de la Vega and Hall (2002) for genotype group membership.

 
Grain Number
Grain number in sunflower is a function of floret number and grain set (i.e., the ratio between filled grains and florets). Floret number is determined in a brief period (about 8 d under usual temperature conditions for sunflower cultivation), most of which occurs before the crop reaches the bud-visible stage (Cantagallo and Hall, 2000). Grain set is strongly influenced by temperature and radiation conditions prevailing around anthesis, but can also be affected by levels of these factors during and immediately after floral differentiation (Cantagallo et al., 1997; Cantagallo and Hall, 2000; Chimenti et al., 2001).

In the present experiments, the slight S effects on floret number was not significant (Tables 2 and 3). This result is consistent with known effects of temperature and radiation on floret differentiation (Cantagallo et al., 1997). Radiation and temperature conditions were similar across sowings within each year (data not shown), as was preanthesis leaf area development (de la Vega and Hall, 2002). Year effect was important (Tables 2 and 3) and may reflect differences between years in conditions prevailing during floret differentiation (data not shown). As is to be expected (Cantagallo et al., 1997), G effects for floret number contributed strongly to total variability, but G x S interaction for this variable was small and not significant (Table 3).

In addition to the responses to Y and G noted for floret number, grain number also showed significant S and G x S interaction effects (Table 3), and oil yield was correlated with grain number across sowings (Fig. 1 and Table 4), in contrast to the lack of correlation of oil yield with floret number (Table 4). The strong G effect on the number of florets per head (Tables 1 and 3) accounted for much of the positive association between this trait and the number of filled grain, accounting for more than 50% of the total variation for the latter variable (Tables 1 and 4). The significant S effect on grain number (greater in S1 than in S2, Tables 1, 2, and 3), combined with a lack of effect on floret number (Tables 2 and 3), suggests that variation in grain number derives from failures in grain set. This is consistent with the significant increase in the diameter of the empty center of the capitulum in S2 (Tables 2 and 3), the known sensitivity of grain set to conditions around anthesis (Cantagallo and Hall, 2000), and the previously noted (de la Vega and Hall, 2002) observation that most of the planting date effects on yield in these experiments were linked to processes and attributes expressed postanthesis. The observed failures in seed set in S2, because of either lack of fertilization or embryo abortion in the central portion of the floral disc, could be due to the reduction observed in the daily incident radiation from anthesis to physiological maturity in that environment [see Table 2 in de la Vega and Hall (2002)].


View this table:
[in this window]
[in a new window]
  		Table 4. Correlations, based on hybrid means, among sunflower traits measured in October and December planting dates across two seasons (1996–1997 and 1998–1999).

 
The G x S interaction for grain number contributed to the observed G x S interaction for grain yield (Tables 1 and 3) and was linked, in turn, to the G x S interaction for diameter of the empty center of the head (Table 3). Because there was no G x S interaction for floret number, we conclude that the G x S interaction for grain number involves genotype-specific responses for traits expressed postanthesis.

Grain Weight
Oil yield was correlated with grain weight across sowings (Fig. 1B; Table 4), and within S2 (Fig. 1). In sunflower, as with other crops, grain weight may be regarded as the outcome of the product of its secondary determinants, rate and duration of grain filling. In the present experiments, most of the sources of variation considered in the analysis of variance affected grain weight, with G and S contributing a large portion of total variation (Table 3). The G effects across sowing dates on grain weight were largely associated with the usual (Cantagallo et al., 1997; López Pereira et al., 1999) inverse correlation between grain number and weight across genotypes (P < 0.01; data not shown), with large-seeded hybrids having lower grain numbers. When plotting the hybrid data within sowing dates, grain weight was correlated negatively to floret number (Table 4), but the negative association between grain weight and grain number was not significant (Table 4). This, together with the negative correlation found between floret number and rate of grain filling (Table 4), suggests that the restriction of seed size imposed by the number of florets differentiated at anthesis cannot be counteracted by the seed capacity to respond to improved spatial conditions derived from failures in seed set. Consistent with the findings of López Pereira et al. (1999), contributions of G effects to total variation were larger when rate of grain filling was considered (Fig. 2A ; Table 3), while S effects explained the dominant proportion of variations in duration of grain filling (Fig. 2B; Table 3). The proportion of the treatment sums of squares accounted for by the G effects for grain weight and grain filling rate decreased from the periphery to the intermediate portion of the capitulum (Table 3), while the opposite trend was observed for S effects. The higher heritability expressed by these traits in the head periphery is consistent with the fact that the effects of stresses are most clearly seen in the central portion of the floral disc (Connor and Hall, 1997).



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 2. Relationships between grain weight and (A) rate of grain filling, and (B) duration of grain filling, for nine sunflower hybrids grown in October (S1) and December (S2) sowing dates at Venado Tuerto during 1996-1997 and 1998-1999. For genotype groups (i.e., G1, G2, G3) see legend of Fig. 1.

 
A significant G x S interaction was found for grain weight (Tables 1 and 3), which contributed to the observed G x S interaction for oil yield. The portion of the total variation accounted for by the G x S interaction for grain weight and its components increased from the periphery to the inner portion of the capitulum (Table 3). Even when the significant genotype-specific responses to S2 in terms of both rate and duration of grain filling underlie the G x S interaction for grain weight (Table 3), some degree of compensation between these attributes among the genotypes adapted to S2 was observed. Figure 3 shows the grain weight vs. time relationships for three genotypes of contrasting responses for rate and duration of grain filling in S1 and S2. Genotype 1, which improved its relative performance in S1, showed a strong reduction in both rate and duration of grain filling in S2. Genotypes 2 and 9, which showed a contrasting relative performance across environments in comparison to Genotype 1, showed different types of response to S2 in terms of the responses of the grain weight determinants. Genotype 2 increased grain filling rate in S2, reaching a higher grain weight than in S1, in spite of the reduction observed in the duration of grain filling. Genotype 9, in turn, showed a reduction in the rate of grain filling in S2, but maintained the duration of grain filling better than the other hybrids.



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 3. Bilineal relationships between grain weight and time from full anthesis for Genotypes 1, 2, and 9 in October (S1) and December (S2) planting dates at Venado Tuerto, 1996-1997. Final grain weight and grain oil concentration values reached by those hybrids in those environments are superimposed. Error bars indicate standard deviations. See Table 1 for the genotype codes.

 
Grain Oil Concentration
Oil yield was correlated to grain oil concentration across sowings (Fig. 1C; Table 4) and within S1 (Fig. 1C). Planting date accounted for almost 65% of the treatment sums of squares for grain oil concentration (Table 3), which was higher in S1 than in S2 (Tables 2 and 3).

Sunflower grain comprises pericarp (hull), derived from the ovary wall, and the kernel, which is mostly embryo (Knowles, 1978). As most of the grain oil is deposited in the kernel, grain oil concentration is affected by kernel percentage and kernel oil concentration. The patterns of growth of the whole grain, hull and kernel, and deposition of oil (see Villalobos et al., 1996; Connor and Hall, 1997) provide a useful framework for analyzing the influence of the rate and duration of grain filling on grain oil concentration. Hull growth is completed early during the grain filling period, while kernel mass is still increasing. Deposition of reserve lipids commences after the start of rapid increase in kernel mass (Villalobos et al., 1996). Consequently, terminal stresses are likely to affect final grain oil concentration simply through changes in the rate and duration of embryo and oil increases (Hall et al., 1985), i.e., a shortening of the grain filling period would result in a greater proportion of pericarp.

The strong reduction in the duration of grain filling observed in S2 determined a significant reduction in the mean grain oil concentration (Fig. 4B) . The partitioning of the treatment sums of squares (Table 3) revealed that S accounted for most of the association between both attributes (Table 4). According to the framework described above, it was expected that the observed reduction in grain filling duration would affect negatively grain oil concentration through changes in kernel percentage. However, the mean changes in final grain oil proportion associated to planting date were largely due to changes in kernel oil proportion (Tables 2 and 3). The effects on kernel percentage, although fairly consistent and predominantly in the direction that would reduce oil concentration, contributed little to the observed behavior (Tables 2 and 3). Diagonal isolines in Fig. 4A represent response to kernel percentage expected for the indicated constant kernel oil concentrations. Arrows for Genotypes 1 and 9 exemplify the change in grain oil concentration between S1 and S2. Although the responses of these hybrids in terms of grain oil concentration and kernel percentage were strongly contrasting, both suggest that changes in kernel percentage are not the main determinant of grain oil concentration responses to late plantings.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 4. Relationships between grain oil concentration and (A) kernel percentage, and (B) duration of grain filling, for nine sunflower hybrids grown in October (S1) and December (S2) sowing dates at Venado Tuerto during 1996-1997 and 1998-1999. Dashed arrows highlight some contrasting genotypic responses to S2, which underlie G x S interactions (Treatment 1, Triangles; Treatment 9, circles). For genotype groups (i.e., G1, G2, G3) see legend of Fig. 1.

 
The observed shortening of the grain filling period in S2, which affected both grain weight and grain oil concentration, is noteworthy since low temperatures within the suboptimum range tend to prolong the duration of grain filling (Ploschuk and Hall, 1995). Earlier reports (Johnson and Jellum, 1972; Unger, 1980) of crops maturing late in the season at low temperatures show that the low oil content of the seed was associated with lighter seed, possibly because of problems with grain growth. Bange et al. (1997) also found a reduction in the thermal duration of grain filling and in grain oil concentration in crops planted late, which filled grain at lower temperatures than crops planted early. Dosio et al. (2000) have shown that reductions in radiation intercepted per plant during the grain filling phase can decrease grain oil proportion because of variations in kernel oil proportion rather than kernel percentage. The effects of late planting dates on grain oil concentration found in this study may therefore be allied to the responses to lower intercepted radiation reported by Dosio et al. (2000) [see Table 2 in de la Vega and Hall (2002)]. Other factors, such as photoperiod (de la Vega et al., 2002), may also play a part. Future studies involving manipulation of management, environmental, and genetic factors are needed to dissect out and quantify the relationships and the interactions between factors.

Genotypic variation was also found for grain oil concentration (Tables 1 and 3). The genotypic variability within planting dates related more to variations in kernel percentage than to changes in the duration of grain filling (Fig. 4A). The differential performance for grain oil concentration showed by the hybrids with white-pigmented achene hypodermis (kernel less than 68% of achene) and those of uncolored hypodermis (kernel more than 73% of achene) accounted for most of the significant correlation between kernel percentage and grain oil concentration found in S1 (r = 0.90, P < 0.01). The dominant factor Hyp determining the presence of white pigmented hypodermis is located in the same map interval as one quantitative trait locus with major effects on seed oil percentage (Leon et al., 1996). In S2, the positive association between grain oil concentration and kernel percentage was not significant (r = 0.46), in part because of the improved relative performance of the hybrids with white-pigmented hypodermis in terms of both grain oil concentration and kernel percentage (see dashed arrow corresponding to Genotype 9 in Fig. 4A).

A significant G x S interaction was found for grain oil concentration (Tables 1 and 3), which partially determined the G x S interaction found for oil yield (see dashed arrows in Fig. 1C). As observed for the other yield components, the portion of variation explained by the predictable interactions (i.e., G x S) was higher than that explained by the unpredictable interactions involving G (i.e., G x Y and G x S x Y; Table 3). The genotype-specific responses to S2 in terms of both kernel percentage (Table 3) and duration of grain filling (Table 3) underlie the G x S interaction for grain oil concentration (see dashed arrows in Fig. 4). Genotype 9, for example, which improved its relative performance for grain oil concentration in S2, maintained the duration of grain filling better than the other hybrids (Fig. 3 and 4B) and increased kernel percentage (Fig. 4A) in S2.

Grain Set in the Head Center as a Secondary Trait Associated with Adaptation to Late Sowing
As observed for the determinants of oil-corrected grain yield (de la Vega and Hall, 2002), some degree of compensation among the attributes determining oil yield was observed among the genotypes adapted to S2. Figure 3, for example, shows the contrasting responses of Genotypes 2 and 9, both improving their relative performance in S2 in comparison to Genotype 1, in terms of rate and duration of grain filling. An increase in grain filling rate in S2 allowed Genotype 2 to reach a higher grain weight than in S1. A better maintenance of the duration of grain filling allowed Genotype 9 to maintain better grain oil concentration in S2 than the other hybrids. This kind of compensation among the attributes underlying the genotype-specific responses for oil yield within the same genotype groups represents a difficulty for the identification of common indicators for an ideotype-based selection strategy. However, it was found that the hybrids that improve their relative performance in S2 showed a better seed set in the central portion of the floral disc than the genotypes that improve their relative performance in S1. Excluding hybrid 3, which improved its relative performance in S2 in spite of increasing the diameter of the empty center, we found a significant correlation between both attributes (r = 0.69; P < 0.05). We conclude that the grain set in the central portion of the capitulum is an easy-to-assess trait that could complement canopy stay green (de la Vega and Hall, 2002) as putative indicator of adaptation to late planting dates.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
The results of experiments and analyses described here are consistent with earlier reports showing that late planting dates affect negatively sunflower yields through reductions in all its components. However, in contrast to the results of Beard and Geng (1982) and Bange et al. (1997), these results show that G x S interactions can account for a portion of total variability sufficiently high to severely complicate selection for broad adaptation to normal and late planting times. Further, these results suggest that a large portion of the observed G x S interactions is the result of differences in partially identifiable adaptive trait combinations that can be exploited by selection for specific adaptation. Although yield components are obvious candidates for physiological and genetic analyses aimed at understanding the biophysical bases of G x E interactions, the results of such studies are often of limited value as guides to plant breeders (Austin, 1993). In the present case, this is partly because of compensation between components (Austin, 1993), i.e., the patterns of genotype relative performance for yield become blurred when the focus moves towards the primary components of yield, and partly because of the strong G effect showed by the components of oil yield. In this study, for example, Genotype 1 produced more grains, lighter in weight and of greater oil concentration than Genotype 3 in both S1 and S2 (Table 1), but the significant noncrossover interactions shown by the three components determined a significant crossover interaction for oil yield (Tables 1 and 3). Nonetheless, the partial understanding of the underlying physiology of the S, G, and G x S interaction effects shown by the primary yield components allowed the identification of indicators of improved relative performance for yield at late planting dates. A selection strategy aimed at exploiting both specific and broad adaptation would use identifiable variation attributed to G x S interaction and G effects, respectively. Variations in grain set were the dominant determinant of the G x S interaction for grain number, and duration of grain filling was the main determinant of the G x S interaction effects on grain weight and grain oil concentration. The diameter of the empty center and canopy stay green (de la Vega and Hall, 2002) proved to be efficient putative secondary indicators related to seed set and duration of grain filling, respectively, that can be used to improve the efficiency of selection for specific adaptation to late plantings. While kernel oil proportion largely explained the S effect for grain oil proportion, kernel percentage mostly determined the significant G effect for this yield component, showing a significant positive association with it across planting times (r = 0.75; P < 0.05). This is an example of how the study of both G x S interaction and G effects can be used to identify putative traits, i.e., canopy stay green and kernel percentage, respectively, relevant for selection strategies aimed at exploiting specific and broad adaptation.


   ACKNOWLEDGMENTS
 
We thank Advanta Semillas (Steven Quast, Alan Scott, Ricardo Siciliano) for making this research possible. We also thank Carlos Ghanem, Aldo Martínez, Sergio Solián, Daniel Kennedy, Ney Flores, and César Sánchez for collaborating in the field experiments, and Drs. Martín Grondona and Alberto León for help with the statistical analyses.

Received for publication March 21, 2001.


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


Related articles in Crop Science:

This issue in Crop science

Crop Science 2002 42: 1069-1070. [Full Text]  



This article has been cited by other articles:


Home page
 J Exp BotHome page
T. Hewezi, M. Leger, W. El Kayal, and L. Gentzbittel
Transcriptional profiling of sunflower plants growing under low temperatures reveals an extensive down-regulation of gene expression associated with chilling sensitivity
J. Exp. Bot., September 1, 2006; 57(12): 3109 - 3122.
[Abstract] [Full Text] [PDF]

Home page
 ANN BOT (LOND)Home page
A. I. MANTESE, D. MEDAN, and A. J. HALL
Achene Structure, Development and Lipid Accumulation in Sunflower Cultivars Differing in Oil Content at Maturity
Ann. Bot., June 1, 2006; 97(6): 999 - 1010.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
R. A. Ruiz and G. A. Maddonni
Sunflower Seed Weight and Oil Concentration under Different Post-Flowering Source-Sink Ratios
Crop Sci., February 1, 2006; 46(2): 671 - 680.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
A. J. Leon, F. H. Andrade, and M. Lee
Genetic Analysis of Seed-Oil Concentration across Generations and Environments in Sunflower
Crop Sci., January 1, 2003; 43(1): 135 - 140.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
A. J. de la Vega and A. J. Hall
Effects of Planting Date, Genotype, and Their Interactions on Sunflower Yield: I. Determinants of Oil-Corrected Grain Yield
Crop Sci., July 1, 2002; 42(4): 1191 - 1201.
[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 Related articles in Crop Science
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
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 ISI Web of Science (10)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by de la Vega, A. J.
Right arrow Articles by Hall, A. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by de la Vega, A. J.
Right arrow Articles by Hall, A. J.
Agricola
Right arrow Articles by de la Vega, A. J.
Right arrow Articles by Hall, A. J.
Related Collections
Right arrow Crop Cytology

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