Crop Science 41:1390-1395 (2001)
© 2001 Crop Science Society of America
CROP BREEDING, GENETICS & CYTOLOGY
Grain Fill Duration in Twelve Hard Red Spring Wheat Crosses
Genetic Variation and Association with Other Agronomic Traits
L. E. Talbert*,
S. P. Lanning,
R. L. Murphy and
J. M. Martin
Dep. of Plant Sciences and Plant Pathology, Montana State University, Bozeman MT 59717
* Corresponding author (usslt{at}montana.edu)
 |
ABSTRACT
|
|---|
The rate and duration of grain fill period has been shown to be associated with increased yield in cereals when water deficits occur during maturation. Grain fill duration can be approximated as the time between heading date and physiological maturity, allowing measurements to be obtained for large breeding populations. Our objectives were to determine the genetic variation for grain fill duration, and to determine its association with agronomic traits in a set of spring wheat (Triticum aestivum L.) crosses. Fifty F3.5 progenies from 12 spring wheat crosses plus parents were grown in replicated trials in three Montana environments. Heritabilities were consistently high across crosses for grain protein (mean = 0.92) heading date (mean = 0.89) and test weight (mean = 0.79), intermediate for physiological maturity (mean = 0.64) and grain yield (mean = 0.59) and lowest for grain fill duration (mean = 0.4). Earlier heading and later physiological maturity were associated with longer grain fill across crosses. Earlier heading also was often significantly correlated with higher grain protein concentration and higher test weight, but usually was not associated with grain yield. Longer grain fill duration was usually associated with higher grain protein in two environments, but with lower grain protein for the cool, wet environment. Grain fill duration showed no significant association with grain yield in most instances. Selection for early heading in early generations followed by selection for grain yield and grain fill duration in later generations using multi-location trials may circumvent the negative association between grain yield and grain protein.
|
INTRODUCTION
|
|---|
MAJOR OBJECTIVES for hard red spring wheat (Triticum aestivum L.) breeding programs include high yield potential and high grain protein concentration. A challenge is that these traits are generally negatively correlated, and selection for one tends to decrease performance for the other. The necessity for high grain protein concentration has no doubt limited yield gains in hard red spring wheat. There would be value in developing selection criteria that circumvent this negative relationship.
Evaluation of grain protein and grain yield are challenging when dealing with large numbers of genotypes and small quantities of seed in early generations. As a result, there has been a long-standing interest in identifying plant characteristics associated with these traits that might be used as selection criteria on a single row or even single plant basis. For plant characteristics to be useful on the scale employed in early generation line or plant selection, ease and rapidity of measurement and high heritability are critical.
One trait of interest in cereal grains is grain fill duration. Grain fill is the period between heading and physiological maturity. Physiological maturity has been approximated visually in spring wheat at the time when heads lose green color (Hanft and Wych, 1982; Metzger et al., 1984; Mou and Kronstad, 1994). Grain fill duration generally has been found to be positively correlated with grain yield in corn (Zea mays L.) (Daynard et al., 1971; Daynard and Kannenberg, 1976; Ottaviano and Camussi, 1981), although Hartung et al. (1989) were unable to show any relationship. Borrell et al. (2000) investigated the impact of green leaf area duration on grain yield in sorghum using a set of nine hybrids placed under three different water regimes. They found a high correlation (r = 0.75) between these traits under conditions of terminal water deficit; i.e., where the crop depended primarily on stored soil moisture for growth and maturation. There was no association between green leaf area duration and grain yield under well-watered conditions.
Associations between grain fill duration and grain yield in wheat and other small grains have varied across several different experiments. Housley et al. (1982) found only slight differences in grain fill duration among a set of winter wheat cultivars, and no association with final grain yield. Darroch and Baker (1990) found that a set of three spring wheat cultivars did not show consistent differences for grain fill duration over environments. Bruckner and Frohberg (1987) found significant differences for grain fill duration in a set of 20 spring wheat lines, but no association with grain yield, kernel weight, or test weight. Nass and Reiser (1975) also found no association between grain fill duration and grain yield in a set of ten spring wheat lines. Similarly, Metzger et al. (1984) found that a set of barley (Hordeum vulgare L.) lines selected for divergent grain fill duration did not differ significantly in grain yield. On the other hand, Gebeyehou et al. (1982a)(b) found a set of eleven durum wheat (Tricitum turgidum L. var. durum) cultivars differed significantly for grain fill duration, which was significantly correlated with final grain yield (r = 0.39). Most previous reports with grain fill duration in wheat have not examined associations with grain protein concentration. Mou et al. (1994) found a significant positive correlation (r = 0.44) between grain fill duration and protein concentration in a winter wheat population. Knott and Gebeyehou (1987) found a positive correlation between kernel protein concentration and grain fill duration (r = 0.22) for three durum wheat crosses in one of two years.
Spring wheat production in the northern Great Plains generally is characterized by water deficits, where plants are largely dependent on stored soil moisture for growth and maturity. In general, breeders have tended to select plants with short life cycles that could reach maturity before severe water deficits occurred. However, given results obtained with other cereals there may be merit in incorporating long grain fill duration into cultivars developed for the region. This is also suggested by a trend in dryland regions of Montana for higher yielding new cultivars to have longer grain fill duration (Talbert and Lanning, unpublished data, 2000). Our objectives were to determine the genetic variation for grain fill duration, and its association with agronomic traits important to hard red spring wheat variety development in a set of spring wheat crosses.
|
MATERIALS AND METHODS
|
|---|
Twelve crosses were made among 10 elite hard red spring wheat cultivars (Burkhamer et al., 1998). The crosses were ‘Amidon’/‘Newana’ (AN), ‘Fortuna’/‘Hi-Line’ (FH), Fortuna/‘Lew’ (FL), ‘Glenman’/Amidon (GA), Glenman/Lew (GL), Glenman/‘Marberg’ (GM), ‘Grandin’/‘Pondera’ (GP), Hi-Line/Newana (HN), Hi-Line/Pondera (HP), Lew/Amidon (LA), ‘Len’/Glenman (LG), and Len/Newana (LN). Seed of fifty random F3-derived F5 lines per cross was obtained via a generation of single seed descent followed by a generation of seed increase using seed from a single F3 plant.
Fifty randomly chosen F3-derived F5 lines per cross plus parents were planted in a randomized complete block split plot design. The 12 crosses were whole plots, and the 50 random lines plus parents were subplots. Each subplot was a single 3 m row with 30 cm between rows. The experiment was grown at the Arthur H. Post Field Research Farm near Bozeman, MT in 1995. The experiment was grown at the same location in two separate adjacent trials in 1996 with one being an irrigated environment and the other receiving no irrigation. Seed for the 1996 trials was obtained from a single replication of the 1995 trial. Thus, F3-derived F6 lines were used in 1996. The irrigated trial received 8.9 cm additional moisture on both 24 to 25 June and 9 to 10 July. There were three replications in 1995 and two replications for each environment in 1996. The trials were planted 19 May 1995 and 2 May 1996. Plots were harvested 13 to 15 Sept. 1995. The non-irrigated trial was harvested 20 to 21 Aug. 1996 and the irrigated trial was harvested 29 to 30 Aug. 1996. There was 72.6 kg/ha of stored available N in 1995 and 27.2 kg/ha was added in the form of urea, and in 1996, there was 79.4 kg/ha of stored available N and 22.6 kg/ha was added in the form of urea.
Heading date was the number of days after planting to when 50% of the heads in a plot were completely emerged from the flag leaf sheath. Physiological maturity was the number of days after planting to when complete loss of green color from the glumes was observed in 75% of the plot (Hanft and Wych, 1982). Grain fill duration was determined as the difference between physiological maturity and heading date. Test weight was measured from a sample of grain on a Seedburo (Chicago, IL) test weight scale. Grain protein concentration was obtained on whole grain samples using an Infratec (Tecator, Höganäs, Sweden) whole kernel analyzer.
We computed an analysis of variance for all traits from separate analyses for each cross using PROC MIXED in SAS (SAS Institute, 1997). All factors except environments were considered random. Analyses then were combined over crosses where environments and crosses were considered fixed and the remaining factors considered random. Narrow sense heritability values were computed for each trait-cross combination on an entry mean basis using variance components from each cross as:
where 
2g is genetic variance component, 2ge is genotype x environment variance component, 2 is random error, e is number of environments and r is harmonic mean for number of replications in an environment. Correlations among traits were computed using the entry means for each environment and from entry means over environments. Equality of correlation matrices over crosses within years and from means combined over years was tested using a multivariate generalization of Bartlett's (1937) test for homogeneity of variances (Rencher, 1995). Homogeneity of correlations over years for individual crosses was performed in similar fashion. The tests were performed in SAS (SAS Institute, 1988) using the DISCRIM procedure with METHOD=NORMAL and POOL=TEST options.
|
RESULTS
|
|---|
Analysis of variance (data not shown) showed that crosses and their interaction with environments were significant (P < 0.01) sources of variation for all traits. Environments also elicited different responses for traits, with the smallest variation in mean performance observed for test weight (Table 1). Long term average growing season temperature and precipitation for this site are 15.4°C and 19.44 cm, respectively. The 1995 average growing season temperature was 12.6°C, with 35.92 cm of precipitation. The 1996 weather was more typical for this site, where average growing season temperature was 14.6°C, with 17.0 cm of precipitation. Average grain fill duration was 48.5 d in 1995 as opposed to 36.1 d for the 1996 non-irrigated environment and 45.6 d for the 1996 irrigated environment (Table 1). The diverse environmental conditions created variable combinations of vegetative and reproductive periods among the three environments. Lack of moisture in the non-irrigated trial hastened maturity compared to the irrigated trial in 1996; consequently physiological maturity occurred 11 d earlier. Heading date was more than 6 d earlier in 1995 than in 1996. Total growing season moisture was about equal for the 1995 and 1996 irrigated environments, yet grain yield was 1350 kg ha-1 greater in the 1996 irrigated environment.
View this table:
[in this window]
[in a new window]
|
Table 1. Environment means for agronomic traits averaged over fifty random lines from twelve spring wheat crosses.
|
|
Significant genetic variation (P < 0.01) was observed in all crosses for heading date, protein concentration, and test weight (Table 2). Heritability values exceeded 0.7 in all cases except the FL cross for test weight (h2 = 0.46). All crosses also showed significant genetic variation for physiological maturity, where heritability values ranged from 0.35 to 0.87. Ten of 12 crosses displayed significant (P < 0.05) genetic variation for grain yield, and heritability values ranged from 0.22 to 0.78. Grain fill duration was least heritable among the traits with values ranging from 0.05 to 0.69, and eight of 12 crosses showed significant (P < 0.05) genetic variation.
View this table:
[in this window]
[in a new window]
|
Table 2. Heritability values for grain fill duration, heading date, physiological maturity, grain yield, test weight, and grain protein for twelve spring wheat crosses.
|
|
Since correlation structure was determined heterogeneous over the three environments (P < 0.01) and over the 12 crosses within an environment (P < 0.01), correlations were calculated for environment/cross combinations in addition to mean data combined over environments. Several trait correlations were as expected. First, heading date was positively correlated with physiological maturity for all crosses, as correlations exceeded 0.51 based on data combined over years (P < 0.01) (data not shown). Grain fill duration was always negatively correlated with heading date, significantly so in eight of 12 crosses based on data combined over environments. Grain fill duration was always positively correlated with physiological maturity, significantly so for eight of 12 crosses based on data combined over environments (P < 0.05). Correlations between grain yield and grain protein concentration were significant and negative when combined over environments except for the HN cross (r = -0.09). Within environments, grain yield-grain protein concentration associations were always negative, but were higher in 1995 than for the 1996 environments. Nine of 24 correlations were not significant at the 5% level in 1996.
Table 3 shows correlations of heading date with yield, test weight and grain protein. Heading date and yield were positively correlated (P < 0.05) for four of 12 crosses based on means combined over environments, and for nine of 36 cross/environment combinations. Within environments correlations ranged from -0.19 to 0.48. Heading date was negatively correlated with test weight (P < 0.05) for 10 of 12 crosses based on means, and for 25 of 36 cross/environment combinations. Heading date was also negatively correlated (P < 0.05) with grain protein concentration for seven of 12 crosses based on means, and for 17 of 36 cross/environment combinations.
Physiological maturity was positively associated (P < 0.05) with grain yield for three of 12 crosses based on means over environment with values ranging from -0.14 to 0.54 (Table 4). Within environments 11 of 36 cross/environment combinations were significant (P < 0.05) and positive, and correlations ranged between -0.18 and 0.70. Physiological maturity was negatively correlated (P < 0.05) with test weight for seven of 12 crosses with range -0.6 to 0.06 based on means, and for 13 of 36 cross/environment combinations where the range was -0.56 to 0.39. Most of the negative correlations occurred in the cold wet growing season of 1995. Correlations of physiological maturity and grain protein concentration varied with environments, with this value being significantly negative (P < 0.05) for nine of 12 crosses in 1995, and positive for five of 12 crosses in the 1996 non-irrigated environment. One significant negative correlation and one significant positive correlation were observed in the 1996 irrigated environment (Table 4).
Grain fill duration and grain yield were independent based on means combined over environments (Table 5). Within environments, significant (P < 0.05) positive correlations were observed in five of 36 cross/environment combinations, and one significant negative correlation was observed. Correlations ranged between -0.21 and 0.31 in 1995, between -0.39 and 0.56 in the non-irrigated 1996 environment, and between -0.18 and 0.37 in the 1996 irrigated environment.
Correlations between grain fill duration and test weight ranged between -0.28 and 0.34 with three being significantly positive and one significantly negative (P < 0.05) (Table 5) based on mean performance over environments. In 1995 correlations ranged between -0.28 and 0.27 with one significant negative correlation. Nine of 12 correlations were significantly positive (P < 0.05) for the 1996 non-irrigated environment and correlations ranged from -0.20 to 0.51. Similarly, three correlations were significantly positive for the 1996 irrigated environment with one being significantly negative while the range between them was -0.32 to 0.52.
Grain fill duration and grain protein concentration were positively related in 10 of 12 crosses with five of 12 being significant (P < 0.05) with a range between -0.13 and 0.43 based on mean performance over environments. Correlation between grain fill duration and grain protein varied markedly depending upon the environment. These correlations were usually negative with seven of 12 being significant (P < 0.05) in 1995, and always positive with eight of 12 being significant (P < 0.05) in both non-irrigated and irrigated 1996 environments (Table 5).
|
DISCUSSION
|
|---|
Environments and crosses were a significant source of variation for grain fill duration and other traits. Heritabilities were highest for grain protein concentration (mean = 0.92) heading date (mean = 0.89) and test weight (mean = 0.79), intermediate for physiological maturity (mean = 0.64) and grain yield (mean = 0.59) and lowest for grain fill duration (mean = 0.4) (Table 2). This suggests less genetic variation for grain fill duration than for other traits measured.
Associations between traits varied with both crosses and environments. However, some general trends were discerned. First, earlier heading date was consistently associated with both higher test weight and grain protein concentration, but it was statistically correlated with lower grain yield in about one-fourth of the cross/environment combinations. Second, physiological maturity was usually positively correlated with grain yield, but negatively correlated with test weight. Correlations between physiological maturity and grain protein concentration varied considerably with environment and cross.
Grain fill duration was independent of grain yield, as the correlation ranged between -0.11 and 0.15 for the twelve crosses using combined means, and was not significant in 30 of 36 cross/environment combinations. Thus, for most crosses in these environments, selection for long grain fill duration would show minimal correlated response in grain yield. Longer grain fill duration was often correlated with higher test weight in the 1996 non-irrigated environment, but that association was not observed in the other two environments where moisture was not limited. This result suggests that selection for long grain fill duration may lead to increased test weight for many genetic backgrounds in environments where moisture is limited during grain filling as was the case in the 1996 non-irrigated trial.
The correlation between grain fill duration and grain protein concentration showed remarkable environmental variation for individual crosses. Longer grain fill duration was associated with lower grain protein concentration in 1995 but with higher grain protein concentration in the 1996 environments.
Farmers are paid for hard red spring wheat based on two primary criteria, grain yield and grain protein concentration. Test weight must also be above a minimal value to avoid a discount. Measurement of traits such as grain yield is difficult in early generations where limited seed supply will not allow multi-environment replicated trials. Breeders would like selection criteria that show desirable correlations with other traits such as grain yield or grain protein concentration to use in early generations. Our results point to the possibility that selection for early heading and long grain full duration may help to circumvent the negative association between grain protein concentration and grain yield observed in many selection experiments (McNeal et al., 1972; Delzer et al., 1994). In particular selection for earlier heading and longer grain fill duration may lead to higher grain protein concentration without negatively impacting grain yield for many genetic backgrounds in typical spring wheat growing environments such as those encountered during our 1996 trials.
There are concerns related to selection for long grain fill duration. Heritability for grain fill duration varied across genetic backgrounds, and on average was lower than for other traits measured. This implies genetic gain in grain fill duration would require selection based on multi-location replicated trials. Additionally, our results in 1995 suggest that longer grain fill duration could lead to lower grain protein concentration in certain environments. However, it is worth noting that the 1995 environment was atypical for hard red spring wheat growing regions, with a relatively late seeding date and a cool, wet growing season. A possible strategy would be to select for early heading in early generations and then select for grain fill duration and grain yield in later generations using replicated multi-location trials. Heading date was as highly correlated with grain protein concentration as grain fill duration, the association was consistent across environments, and heading date was highly heritable across all genetic backgrounds.
A problem often encountered by spring wheat in the Great Plains is extreme high temperatures during grain fill, which has variable deleterious effects across genotypes (Al-Khatib and Paulsen, 1990; Gibson and Paulsen, 1999). Truncation of maturation due to high temperature may be especially harmful to genotypes requiring long grain fill periods (Bruckner and Frohberg, 1987). This may be of particular concern if grain fill duration and grain fill rate are negatively correlated, as reported by Mou et al. (1994) for a winter wheat cross. However, other authors have found no relationship between grain fill duration and grain fill rate (Bruckner and Frohberg, 1987). A pragmatic concern for farmers in areas with short growing seasons may be that long grain fill duration, if associated with late physiological maturity, will delay harvest date. Thus, while results of the present study suggest that selection for increased grain fill duration may have an associated positive impact on grain protein concentration for spring wheat in typical Northern Great Plains environments, there are complicating factors that also should be considered.
|
NOTES
|
|---|
Contribution no. J-2000-72 Montana Agric. Exp. Stn.
Received for publication October 6, 2000.
|
REFERENCES
|
|---|
- Al-Khatib, K., and G.M. Paulsen. 1990. Photosynthesis and productivity during high temperature stress of wheat genotypes from major world regions. Crop Sci. 30:1127–1132.[Abstract/Free Full Text]
- Bartlett. M.S. 1937. Some examples of statistical methods of research in agriculture and applied biology. J. Roy. Statist. Soc. Suppl. 4:137–183.
- Borrell, A.K., G.L. Hammer, and R.G. Henzel. 2000. Does maintaining green leaf area in sorghum improve yield under drought? II. Dry matter production and yield. Crop Sci. 40:1037–1048.[Abstract/Free Full Text]
- Bruckner, P.L., and R.C. Frohberg. 1987. Rate and duration of grain fill in spring wheat. Crop Sci. 27:451–455.[Abstract/Free Full Text]
- Burkhamer, R.L., S.P. Lanning, R.J. Martens, J.M. Martin, and L.E. Talbert. 1998. Predicting progeny variance from parental divergence in hard red spring wheat. Crop Sci. 38:243–248.[Abstract/Free Full Text]
- Darroch, B.A., and R.J. Baker. 1990. Grain filling in three spring wheat genotypes: Statistical analysis. Crop Sci. 30:525–529.[Abstract/Free Full Text]
- Daynard, T.B., and L.W. Kannenberg. 1976. Relationships between the length of actual and effective grain filling periods and the grain yield of corn. Can. J. Plant Sci. 56:237–242.
- Daynard, T.B., J.W. Tanner, and W.G. Duncan. 1971. Duration of grain filling period and its relationship to grain yield in corn, Zea mays L. Crop Sci. 11:45–48.
- Delzer, B.W., R.H. Busch, and G.A. Hareland. 1994. Recurrent selection for grain protein in hard red spring wheat. Crop Sci. 35:730–735.
- Gebeyehou, G., D.R. Knott, and R.J. Baker. 1982a. Rate and duration of grain filling in durum wheat cultivars. Crop Sci. 22:337–340.
- Gebeyehou, G., D.R. Knott, and R.J. Baker. 1982b. Relationships among durations of vegetative and grain filling stages, yield components, and grain yield in durum wheat cultivars. Crop Sci. 22:287–290.[Abstract/Free Full Text]
- Gibson, L.R., and G.M. Paulsen. 1999. Yield components of wheat grown under high temperature stress during reproductive growth. Crop Sci. 39:1841–1846.[Abstract/Free Full Text]
- Hanft, J.M., and R.D. Wych. 1982. Visual indicators of physiological maturity of hard red spring wheat. Crop Sci. 22:584–588.[Abstract/Free Full Text]
- Hartung, R.C., C.G. Poneleit, and P.L. Cornelius. 1989. Direct and correlated responses to selection for rate and duration of grain fill in maize. Crop Sci. 29:740–745.[Abstract/Free Full Text]
- Housley, T.L., A.W. Kirleis, H.W. Ohm, and F.L. Patterson. 1982. Dry matter accumulation in soft red winter wheat seeds. Crop Sci. 22:290–294.[Abstract/Free Full Text]
- Knott, D.R., and G. Gebeyehou. 1987. Relationships between the lengths of the vegetative and grain filling periods and agronomic characters in three durum wheat crosses. Crop Sci. 27:857–860.[Abstract/Free Full Text]
- McNeal, F.H., C.F. McGuire, and M.A. Berg. 1978. Recurrent selection for grain protein content in hard red spring wheat. Crop Sci. 18:779–782.[Abstract/Free Full Text]
- Metzger, D.D., S.J. Czaplewski, and D.C. Rasmusson. 1984. Grain filling duration and yield in spring barley. Crop Sci. 24:1101–1105.[Abstract/Free Full Text]
- Mou, B., and W.E. Kronstad. 1994. Duration and rate of grain filling in selected winter wheat populations: I. Inheritance. Crop Sci. 34: 833–837.[Abstract/Free Full Text]
- Mou, B., W.E. Kronstad, and N.N. Saulescu. 1994. Grain filling parameters and protein content in selected winter wheat populations: II. Associations. Crop Sci. 34:838–841.[Abstract/Free Full Text]
- Nass, H.G., and B. Reiser. 1975. Grain filling period and grain yield relationships in spring wheat. Can. J. Plant. Sci. 55:673–678.
- Ottaviano, E., and A. Camusi. 1981. Phenotypic and genetic relationships between yield components in maize. Euphytica 30:601–609.[Web of Science]
- Rencher, A.C. 1995. Methods of multivariate analysis. John Wiley & Sons, New York.
- SAS Institute Inc. 1988. SAS/STAT Users guide, Release 6.03 ed., Cary, NC.
- SAS Institute Inc. 1997. SAS/STAT Software: Changes and enhancements through release 6.12. SAS Institute Inc., Cary, NC.
This article has been cited by other articles:
|
|
|
|
 
N. K. Blake, S. P. Lanning, J. M. Martin, J. D. Sherman, and L. E. Talbert
Relationship of Flag Leaf Characteristics to Economically Important Traits in Two Spring Wheat Crosses
Crop Sci.,
March 1, 2007;
47(2):
491 - 494.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
