| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Agron., Iowa State Univ., 2104 Agronomy Hall, Ames, IA 50011
b Dep. of Agron., Moore Hall, Univ. of Wisconsin, 1575 Linden Dr., Madison, WI 53706
* Corresponding author (palle{at}iastate.edu)
Received for publication September 1, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
Abbreviations: DAE, days after emergence
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
Seed yield is determined by the number of seeds per unit area and seed mass. However, most, but not all, environmentally induced yield differences are due to difference in seed number. Seed mass is often inversely correlated with seeds per unit area (Hanson, 1986). Seed mass is determined by the rate of seed growth and the duration of seed fill, both of which are genetically controlled (Egli et al., 1981, 1984; Guldan and Brun, 1985) although there are environmental influences as well (Egli et al., 1985; Egli and Wardlaw, 1980; Meckel et al., 1984). Attempts to increase yield through increases in seed number or seed mass have been somewhat unsuccessful due to the compensation that occurs between these components (Swank et al., 1987). The compensation that occurs between seed mass and number led Hanson (1986) to conclude that soybean seeds are receptacles for assimilate and that yield-limiting factors occur somewhere outside the seed.
Yield decreases resulting from drought stress depend both on the phenological timing of the stress and on the degree of yield component compensation. Schou et al. (1978) reported that yield is more influenced by changes from flowering to physiological maturity compared with the emergence to flowering period. Numerous studies (Egli and Yu, 1991; Johnston et al., 1969; Schou et al., 1978) have indicated that seed number (per unit ground area) was responsive to altered environmental conditions during flowering and pod set. Cultivar differences in yield response to irrigation regimes (Kadhem et al., 1985a), however, depend not only on differences in individual yield component responses, but also on differences in yield component compensation (Kadhem et al., 1985b).
The negative effects of stress are particularly important during flowering, seed set, and seed filling where stress can reduce yield by reducing number of pods, number of seeds, and seed mass (Ashley and Ethridge, 1978; Doss and Thurlow, 1974; Sionit and Kramer, 1977). Specht and Williams (1984) noted that genotype x environment interactions often involve a "specific adaptation" component (i.e., a consistent superiority of some genotypes over others in specific environments but an inverse performance rank in other environments).
Cultivar adaptability to a region and its influence on soybean yield and yield components can be affected by growth habit and planting date. Since the hectares of soybean have increased in the northern USA, it is important to evaluate the magnitude of the genotype x management system interaction on soybean seed yield components. Information is lacking on the impact of management systems on soybean yield components under cooler temperature in the upper Midwest. The objectives of this research were to (i) determine the yield component response of soybean cultivar to management system and planting date and (ii) describe the soybean yield component development process throughout the growing season in the upper Midwest.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
The experimental design for each management system was a randomized complete block in a split-plot arrangement with four replications. Main plot was planting date (early May vs. late May). The subplots were three soybean cultivars: Hardin (released in 1980; MG 2.0), DeKalb CX232 (1995; MG 2.3), and Spansoy 250 (1995; MG 2.5). Plot size was 3 by 15.2 m, and plots were further divided into two subplots of 3 by 7.6 m where one of the subplots was used for harvest that previously has been published (Pedersen and Lauer, 2003) and the other subplot was used for monitoring yield components. Seeding rate was 432000 seed ha–1. Sections of 0.76 m2 were hand-harvested from each cultivar and planting date plot and were used to determine dry matter on 21-d intervals starting 21 d after emergence (DAE). There were six individual sampling dates (sections) throughout the growing season (21, 42, 63, 84, 105, and 126 DAE). Each section was randomly selected from the center rows of each cultivar and thinned to approximately 350000 plants ha–1 before the first sampling date. Development, growth stage, and plant height information were taken based on a sample of three plants randomly collected from the hand-harvested section. Plant growth stages were determined according to the methods by Fehr and Caviness (1977). The three plants were separated into leaves, stems, pods, and seeds. The yield components measured were as follows: harvest index, which is the ratio of seed dry weight to total aboveground dry weight at time of measurement; seed number per square meter, a seed was counted when the diameter was larger than 3 mm; pod number per square meter, which was determined as a pod when larger than 1 cm; seeds per pod; and seed mass per 100 seeds, which was determined by a random sample of 100 seeds from the harvested seed from each plot. All dry weight samples were oven-dried at 60°C to a constant weight to determine yield on a dry weight basis. Seed mass per 100 seeds was adjusted to 130 g kg–1 moisture content.
Gravimetric soil moisture content was measured in samples taken every 3 wk from the time of the first hand-harvest sampling. Gravimetric soil moisture content was determined in each replication and in each management system by collecting two random soil samples with a 2.5-cm (inner diam.) soil probe from depths of 0 to 15, 15 to 30, 30 to 60, and 60 to 90 cm. Data from 30 to 60 and 60 to 90 cm were collected for another research project and will not be presented here. Samples of field-moist soil were composited for each sampling depth, and a subsample of approximately 0.4 kg was oven-dried at 105°C for 3 d to calculate gravimetric moisture content.
Data were subjected to an analysis of variance using the PROC MIXED procedure (Littell et al., 1996) of SAS (SAS Inst., 1995) with the six sampling dates analyzed as sub-subplots (Gomez and Gomez, 1984). Individual analysis by year using the restricted maximum likelihood method for variance component estimation indicated that error variances were heterogeneous. Block was treated as a random effect in the individual analysis by year. Management system, cultivar, and planting date were treated as fixed effects in determining expected mean squares and appropriate F-tests in the analysis of variance. Management system was treated as fixed effect rather than random to determine interactions involving management system. Homogeneity of error variances was found for data collected during 1998 and 1999, and a combined analysis of variance was performed. For ease of illustration, most emphasis will be focused on the combined analysis; however, data will be discussed for each individual year if they deviate from the combined analysis. Analysis across years (1998 and 1999) treated year as a fixed effect to determine interactions involving year in PROC MIXED. Mean comparisons were made using Fisher's protected LSD test (P 
0.05). Phenotypic correlation coefficients between grain yield from the machine harvest plots and the grain yield components were computed using PROC CORR of SAS.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
|
|
|
|
|
Yield Components
Seed Mass
No interactions were observed for seed mass in the combined analysis. However, a management system x cultivar interaction was detected for seed mass in 2000. Hardin had a 37% higher seed mass in the management system at Hancock compared with the management systems at Arlington in 2000. Seed mass of CX232 and Spansoy was not influenced by any of the management systems during 2000. Additionally, a management system x planting date interaction was detected for seed mass in 2000. No differences in seed mass were observed among management systems for the early planting date. For the late planting date, however, seed mass was 31 and 19% lower in the no-tillage system at Arlington compared with the management system at Hancock and the conventional tillage management system at Arlington, respectively.
Soybean grown in the management system at Hancock produced on average 28% higher seed mass than all management systems at Arlington (Table 1). Seed mass at Hancock was highly correlated with yield (r = 0.65; P < 0.001) whereas no significant correlations were observed at Arlington (Table 2). It is well known that the soybean plant adjusts its sink size in response to environmental stress by aborting flowers, pods, or seeds (Shibles et al., 1975). It is speculated that the higher seed mass at Hancock may have resulted from a more uniform flowering pattern resulting in higher seed mass. At Arlington, the two no-tillage management systems produced 16% greater seed mass than the conventional tillage management systems. Irrigation did not affect seed mass in any tillage system at Arlington, perhaps because of the plentiful and evenly distributed precipitation throughout the growing seasons, which may have favored relatively long seed-filling periods, high seed mass, and equalized planting date and cultivar factors. Ashley and Ethridge (1978) showed that water deficit during seed filling reduces seed size and yield due to shorter seed-filling period and earlier maturity. This is in agreement with our observations since we never observed drought conditions during the study. However, Woodward and Begg (1976) showed a reduced seed mass for irrigated soybean.
|
Hardin and Spansoy 250 produced 15% larger seed mass than CX232 (Table 1). No difference was found in seed mass among cultivars in 1997 and 2000 (data not shown). Gay et al. (1980) and Woodward and Begg (1976) showed that yield advantages between cultivars were correlated with seed mass, partially as a result of the number of seeds available for filling, the duration of the filling period, and total photosynthate production. Seed mass was overall significantly correlated with yield (r = 0.33; P < 0.001), but the r value was relatively small (Table 3). Board (1987) and Carter and Boerma (1979) also reported a weak relationship between seed mass and yield.
|
Seed number was influenced by management system (Table 1). The four management systems at Arlington produced 19% more seeds per square meter than the management systems at Hancock. Irrigation did not influence seed number at Arlington since insufficient moisture during seed set and seed filling was not observed. However, tillage system influenced seed number, with the conventional tillage systems producing 12% more seeds per square meter than the no-tillage systems. These results support previous work that tillage practice has a positive effect on seed number per square meter (Frederick et al., 1998). Management system did not influence seed number in 1997.
Early planting date produced 3607 seeds m–2, or 10% more seeds than the late planting date (3238 seeds m–2; Table 1). This result is consistent with Beatty et al. (1982), who showed that early planted soybean would take advantage of favorable soil moisture conditions and seed number decreases consistently with later planting. Planting date did not influence seed number in 1997 and 2000.
Cultivars differed in seed number with Hardin and Spansoy 250 having 17% more seed m–2 than CX232. No difference was found among cultivars in 1997 and 2000.
Egli et al. (1978) suggested that the number of seeds produced by a soybean community is a function of the amount of photosynthate available for seed growth since soybean seed number is associated with crop growth rate during flowering and pod set (Egli, 1993; Egli and Yu, 1991; Herbert and Litchfield, 1984; Ramseur et al., 1985). Differences in seeds per square meter could therefore be derived from a more efficient utilization of assimilate in seed. Seed number was inversely correlated with seed mass at Arlington averaging –0.79 across the four management systems (Table 2). This is in agreement with observations by Hanson (1986). No correlation was observed between seed mass and seed number at Hancock (Table 2).
Pod Number
No interactions were observed for the combined analysis (Table 1). However, a management system x planting date interaction was found for pod number in 2000 where early planted soybean had 21% higher pod number per square meter than the late-planted soybean in the conventional tillage system with irrigation at Arlington. A management system x cultivar interaction was detected for pod number in 2000 with pod number per square meter being 39% higher for Hardin than for Spansoy 250 in the conventional tillage system with irrigation at Arlington, and no difference was observed between Hardin and CX232 or CX232 and Spansoy 250.
Pod number was influenced by management system in 1998 and 1999, with the management systems at Arlington producing 20% more pods per square meter than the management system at Hancock (Table 1). Irrigation did not influence pod number at Arlington since insufficient moisture during pod set was not observed. However, tillage system influenced pod number with the conventional tillage system producing 12% more pods per square meter than the no-tillage system. No difference was observed among pods per square meter and management systems in 1997.
Planting date affected pod number with the early planted soybean having 12% more pods per square meter than the late-planted soybean, which is in agreement with Beatty et al. (1982), who reported that delayed planting reduced pod number. No difference was observed among pods per square meter and planting dates for 1997 and 2000.
Cultivars differed in pod number with Hardin and Spansoy 250 both having 19% more pods per square meter than CX232. No difference was found among cultivars in 1997 and 2000. Our data correspond well with those of Woodward and Begg (1976), who showed that reduced seed mass of soybean resulted in greater pod and seed number (Table 3). However, no correlation between seed mass and pod number was observed in the nonirrigated, no-tillage system at Arlington (Table 2).
Seed Number per Pod
A cultivar x management system interaction was found in the combined analysis with the two newer cultivars, CX232 and Spansoy 250, having 8% more seeds per pod than Hardin at Arlington (Table 1). No difference was found among the three cultivars at Hancock. A management system x planting date interaction was found for seeds per pod in 1998. No difference was found between planting dates for the irrigated no-tillage management system at Arlington and the management system at Hancock. However, the remaining three management systems averaged 5% more seeds per pod for the late planting date.
Management systems influenced seeds per pod, but inconsistently (Table 1). The management system at Hancock and the no-tillage management system without irrigation and the irrigated conventional tillage management at Arlington had the highest number of seeds per pod, averaging 2.46 seeds pod–1 compared with the remaining two management systems that averaged 2.38 seeds pod–1. Irrigation had a positive effect on seeds per pod in the conventional tillage system and a negative effect in the no-tillage system at Arlington (Table 2). No differences were found among seeds per pod and management systems in 1997 and 2000.
Delaying planting increased the number of seeds per pod from 2.40 to 2.46 seeds pod–1. No differences were observed among planting dates and seeds per pod in 1997 and 2000.
Seeds per pod differed between cultivars with Hardin having 5 and 9% less seeds per pod than CX232 and Spansoy 250, respectively.
Differences between seasons for seeds per pod were small as expected (Dominguez and Hume, 1978) and did not correlate with seed yield (Table 3). However, pod number was inversely correlated with seeds per pod (r = –0.37; P < 0.001).
Harvest Index
A management system x cultivar interaction was detected for harvest index in the combined analysis (Table 1). Hardin had 8% higher harvest index in all management systems at Arlington compared with CX232 and Spansoy 250. No differences were observed among cultivars for harvest index at Hancock. A management system x cultivar interaction was detected for harvest index in 1997. Hardin and CX232 had on average 7% higher harvest index in all management systems at Arlington compared with Spansoy 250. No differences were observed among cultivars and harvest index at Hancock. A management system x planting date x cultivar interaction was detected for harvest index in the combined analysis. At Hancock, Hardin had 16% lower harvest index at the early planting and 5% higher harvest index at the late planting compared with CX232 and Spansoy 250. Few and inconsistent differences were observed between cultivars and planting date at the management systems at Arlington (data not shown). A planting date x cultivar interaction was detected in 1997. Hardin had on average a 9% higher harvest index for the late planting date, and no differences were observed between planting dates for CX232 and Spansoy 250.
Soybean grown in the management system at Hancock had 2% higher harvest index than the management systems at Arlington. Tillage system did not influence harvest index at Arlington, whereas irrigation lowered the harvest index by 2% on average. No difference was found among management systems and harvest index in 1997.
Early planting date had 2% higher harvest index than late planting date. No difference was found among management systems and harvest index in 1997 and 2000.
Highest and lowest harvest index was observed for Hardin and Spansoy 250, respectively. Hardin had 9% higher harvest index than Spansoy 250, and no difference was found between Hardin and CX232 or CX232 and Spansoy 250 (Table 1). No difference was found among cultivars and harvest index in 1997. Based on these results, it is assumed that the old cultivar Hardin had a more efficient utilization of assimilates during seed set than CX232 and Spansoy 250. Spaeth et al. (1984) showed harvest index to be a stable characteristic of cultivar with respect to variations in water availability and photoperiod, which corresponds well with our data. However, the association between harvest index and seed yield has been contradictory (Frederick et al., 1991; Schapaugh and Wilcox, 1980). Schapaugh and Wilcox (1980) found no correlation between harvest index and yield, whereas Frederick et al. (1991) found a relationship between increased harvest index and improved yield potential. Harvest index was in this study significantly correlated with yield and with all other yield components, with an inverse relationship between harvest index and seeds per pod (Table 3).
|
SUMMARY |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
Seed yield was highly correlated with seed mass and harvest index. In general, the two locations (Arlington and Hancock) were different for development of yield components. The management systems at Arlington had higher pod and seed number but lower seed mass compared with the management system at Hancock. Except for seeds per pod, small differences were found between irrigated and nonirrigated management systems at Arlington because of adequate precipitation during all growing seasons. Tillage system had an effect on soybean yield components at Arlington with the no-tillage systems having greater seed mass, seed number, and pod number than the conventional tillage systems. Planting date had an effect on soybean yield components with the early plating date having greater seed number, pod number, and harvest index than the late planting date. There was a difference in the development of yield components associated with planting date, but no treatment differences were detected by 105 DAE for most yield components. Hardin had a higher harvest index than the two new cultivars CX232 and Spansoy 250, indicating a greater efficiency of dry matter use in Hardin.
It was concluded that, despite cultivar differences in yield components and their development, the ability of cultivars to compensate among yield components was more affected by year variability than by management system and planting date.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. L. De Bruin and P. Pedersen Soybean Cultivar and Planting Date Response to Soil Fumigation Agron. J., June 16, 2008; 100(4): 965 - 970. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. De Bruin and P. Pedersen Soybean Seed Yield Response to Planting Date and Seeding Rate in the Upper Midwest Agron. J., May 7, 2008; 100(3): 696 - 703. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Bastidas, T. D. Setiyono, A. Dobermann, K. G. Cassman, R. W. Elmore, G. L. Graef, and J. E. Specht Soybean Sowing Date: The Vegetative, Reproductive, and Agronomic Impacts Crop Sci., March 19, 2008; 48(2): 727 - 740. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Crop Science | 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 | |||
= conventional tillage at Arlington; 
= irrigated, no-tillage management system at Arlington; and 
= no-tillage management system at Arlington. Vertical bars represent the LSD (P 
= irrigated, conventional tillage management system at Hancock. Vertical bars represent the LSD (P 
= late May. Vertical bars represent the LSD (P 