Crop Science 40:1285-1294 (2000)
© 2000 Crop Science Society of America
CROP PHYSIOLOGY & METABOLISM
Light Interception Efficiency and Light Quality Affect Yield Compensation of Soybean at Low Plant Populations
Jim Board
Dep. of Agronomy, Louisiana Agric. Exp. Stn., LSU Agric. Ctr., Baton Rouge, LA 70803 USA
jboard{at}agctr.lsu.edu
 |
ABSTRACT
|
|---|
Greater understanding of how soybean [Glycine max (L.) Merr.] yield compensation occurs across plant populations would aid research aimed at reducing optimal plant population. Objectives were to determine how net assimilation rate (NAR) and leaf area index (LAI) contribute to crop growth rate (CGR) equilibration across low, medium, and high plant populations during the vegetative (emergence–R1) and early reproductive periods (R1–R5). Determinate cultivar Delta Pine 3606 (Maturity Group VI) was planted at an optimal planting date during 1995 and 1996 at low (80000 plants ha-1), medium (145000 plants ha-1), and high (390000 plants ha-1) plant populations on a Commerce silt loam near Baton Rouge, LA (30°N). Yield was unaffected by plant population. Equilibration of CGR for low vs. higher plant populations near R1 was achieved through greater NAR for the low plant population during the vegetative period, created by greater light interception efficiency (LIE, light interception per unit leaf area). Although NAR equilibrated to minimal levels across plant populations near R1, low population maintained CGR parity with higher populations until R5 through greater relative leaf area expansion rate (RLAER) during the late vegetative and early reproductive periods. Higher relative leaf area expansion rate for low vs. higher plant populations resulted from increased partitioning of dry matter into branches, probably induced by greater red/far red light ratios within the canopy. In conclusion, a possible genetic characteristic conducive to low optimal plant population is greater partitioning of dry matter into branches.
Abbreviations: CGR, crop growth rate [g m-2 (land area) d-1] • LAI, leaf area index • LER, leaf expansion rate [cm2 m-2 (land area) d-1] • LI, light interception (%) • LIE, light interception efficiency (%) • NAR, net assimilation rate [g m-2 (leaf area) d-1] • RGR, relative growth rate (g g-1 d-1) • RLAER, relative leaf area expansion rate [cm2 m-2 (leaf area) d-1] • TDM, total dry matter m-2 [g m-2 (land area)]
|
INTRODUCTION
|
|---|
PLANTING SOYBEAN at the minimal population for best yield (optimal plant population) reduces seeding costs, avoids some diseases, and minimizes lodging (Boquet and Walker, 1980). In previous research, optimal plant population varied from 30000 to 500000 plants ha-1 (Lehman and Lambert, 1960; Leffel and Barber, 1961; Lueschen and Hicks, 1977; Costa et al., 1980; Parks et al., 1982; Egli, 1988; Wells, 1991). Optimal plant population can vary by 100% across years, even when the same cultivar is grown in the same location (Moore and Longer, 1987; Wells, 1991). Much of this variability can be explained by environmental conditions, with optimal plant population increasing under adverse growing conditions (Wells, 1991).
Understanding how growth dynamic factors in low vs. medium or high plant populations result in similar yield will aid in identifying genetic and environmental strategies for reducing optimal plant population. Similar yield across plant populations results from equilibration of CGR by the early reproductive period, which causes an equivalent number of pods per square meter (Carpenter and Board, 1997a, 1997b). Crop growth rate equilibration across plant populations occurs through adjustments of LAI and/or NAR (Hunt, 1978). Greater NAR [and also relative growth rate (RGR)] in low compared with normal plant populations during the late vegetative and early reproductive periods is sometimes a contributing factor to CGR equilibration (Carpenter and Board, 1997b; Wells, 1993). Greater light interception per unit LAI (LIE) was associated with this NAR advantage for low plant populations (Carpenter and Board, 1997b). However, sampling in this previous study did not include the early and middle parts of the vegetative period, making it difficult to assess how important NAR is for CGR equilibration across plant populations.
Relative leaf area expansion rate (RLAER) has been reported to be greater in sparse vs. dense stands (Wells, 1993), suggesting that more rapid LAI development per unit LAI may also contribute to CGR equilibration across plant populations. Since most LAI development during the early reproductive period [R1–R5, stages according to Fehr and Caviness (1977)] occurs on branches (main stem vegetative growth stops shortly after R1 in determinate soybean), greater partitioning of total dry matter (TDM) into branches by low populations (Carpenter and Board, 1997b) could result in relatively faster LAI expansion.
Differences in red/far red light ratios within the canopy may help explain why soybean in low plant populations partitions a greater percentage of TDM into branches compared with normal populations. Kasperbauer (1987) demonstrated under field conditions that an increased ratio of red/far red light resulted in greater branch development in soybean. Growth chamber studies confirmed phytochrome involvement. During vegetative development (prior to canopy closure), plants in sparse stands usually experience greater red/far red light ratios within the canopy than those in dense stands (Sanchez et al., 1993). This occurs because of greater interplant reflection of far red light within dense stands. Thus, greater RLAER development in sparse vs. dense stands may ultimately be related to higher ratios of red/far red light within the canopy of the sparse stands. Objectives of this study were to increase understanding of how NAR and LAI contribute to CGR equilibration across plant populations by (i) assessing the importance of LIE during the vegetative period as a factor contributing to greater NAR in sparse vs. dense stands and (ii) determining the role of red/far red light ratios in affecting branch and LAI development in sparse vs. dense stands.
|
Materials and methods
|
|---|
Determinate cultivar Deltapine 3606 (Maturity Group VI) was planted at the Ben Hur Research Farm near Baton Rouge, LA (30°N Lat), on a Commerce silt loam soil (fine-silty, mixed, nonacid, thermic Aeric, Fluvaquent). Seed were machine planted 23 May 1995 and 17 May 1996 on a 75-cm row width at high plant populations. Experimental plots were 10 contiguous rows with a 6.1-m row length. Within each plot, two adjacent rows with border rows were used for plot yield determination and three separated rows (each bordered) were used for sampling. Prior to V3, plots were thinned to three plant populations: low (80000 plants ha-1) , medium (145000 plants ha-1), and high (390000 plants ha-1). Plant populations were the averages of seven stand counts starting at 2 wk after emergence and ending near R2. Stand counts were randomly taken by counting the number of plants occupying a 66-cm length of bordered row from interior portions of the plot and then multiplying by two to obtain plants per square meter. As with other soybean growth dynamic studies, use of multiple cultivars and locations was precluded by the amount of data obtained (Schonbeck et al., 1986; Egli, 1988; Wells, 1993). At R1, plants were supported with nylon netting (12.7 by 12.7 cm mesh) to prevent lodging. Weeds, diseases, and insects were controlled with recommended practices. Fertilizer was applied before planting at the rate of 0-0-67 kg ha-1 (N-P-K) according to soil test recommendations.
Experimental design for growth dynamic data was a randomized complete block in a split plot arrangement with four replications. Main plots were the low, medium, and high plant populations described above, and split plots were sampling dates at 14, 21, 28, 35, 42, 49, 56, and 80 (85 in 1996) d after emergence. Data obtained were light interception [LI (%)], LAI, light interception efficiency (LIE = LI/LAI), TDM (g m-2), and partitioning of TDM into plant parts [leaves, petioles, main stems, branches (% TDM)], number of plants (no. m-2), number of leaves (no. m-2), and area per leaf (cm2 leaf-1).
Light interception was determined with a 1-m-long LI-COR Line Quantum Sensor (LI-COR, Lincoln, NE) connected to a LI-1000 data logger. Light interception was determined by first measuring light intensity at soil level (average of several measurements made parallel to the row at 13-cm intervals between rows). Ambient light intensity at the top of the canopy was then obtained and percentage interception calculated. Plant samples were obtained by harvest of 66 cm from interior portions of bordered rows. The number of plants and leaves were counted. Samples were separated into leaves, petioles, branches, and main stems. Leaf area index was determined by placing the leaf blades through a LI-COR 3000 portable leaf area meter. Plant parts were dried in a forced-air dryer at 60°C to a constant weight. Total dry matter was the sum of all plant parts. Dry matter partitioning for a plant part was the weight of the plant part divided by TDM. Area per leaf was calculated by dividing LAI by leaf number and LIE was determined as LI/LAI. Analysis of variance was according to the SAS General Linear Model (SAS Institute, Cary, NC) with mean separation according to individual t tests. Individual t tests were used in preference to a general LSD value because of changes in plant size occurring during the sampling period (Dr. Lynn LaMotte, Dep. of Experimental Statistics, LSU, 1998, personal communication).
Total dry matter and LAI were regressed against time (Hunt and Parsons, 1981) to obtain CGR [g m-2 (land area) d-1], CGR rate per plant (g plant-1 d-1), RLAER [cm2 m-2 (leaf area) d-1], RGR (g g-1 d-1), and NAR [g m-2 (leaf area) d-1]. Linear, quadratic, and cubic components of each regression equation were successively tested for significance and included in the equation if they significantly reduced the residual sum of squares. Significant differences were determined by t tests using standard errors calculated by the regression program. Crop growth rate averaged during R1 to R5 was calculated by subtracting TDM(R1) from TDM(R5) and dividing by the number of days of the R1 to R5 period. Leaf expansion rate (LER) during R1 to R5 [cm2 m-2 (land area) d-1] was calculated by subtracting LAI(R1) from LAI(R5) and dividing by the number of days from R1 to R5.
Experimental design for yield, yield components obtained at maturity, CGR(R1–R5), and LER(R1–R5) was a randomized complete block with one factor (plant population), four replications, and 2 yr. Yield (kg ha-1) was determined by combine harvest of two interior rows (6.5 m2) of each plot that had been end-trimmed to 4.3 m and corrected to 130 g kg-1 moisture. Shortly after R7, 8, 16, and 36 plants were randomly selected from the low, medium, and high plant population plots, respectively, for yield component analysis. Because plot yields were based on 6.5 times as many plants as sample yield, the two parameters did not always coincide (compare Tables 1 and 2)
. Yields based on sampling usually are greater than plot yield because during sampling, malformed and/or barren plants are rejected. Such plants do not contribute to plot yield and are rejected since yield formation is the process being studied. Thus, sample yields from medium and high populations were greater relative to plot yields, because malformed and barren plants were not included in the sample. Sample and plot yields were similar in the low population, because of fewer malformed or barren plants. Thus, in all treatments, plants representative of those making the contribution to yield were randomly selected. The following yield components were determined for branches, main stem, and whole plant: sample yield (g m-2), seed size (g 100 seed-1), seed per pod (no.), number of nodes (no. m-2), number of reproductive nodes (no. m-2) (reproductive node is a node having at least one pod containing at least one seed), percentage reproductive nodes (% of nodes becoming reproductive), and pods per reproductive node (no.). Plant height data were also obtained from these samples. Seed number for whole plants (no. m-2) was determined by converting plot yield into yield per square meter (g m-2) on a dry weight basis and dividing by weight per seed (g seed-1). Number of pods for whole plants (no. m-2) was calculated as number of seed (no. m-2) divided by seed per pod. Harvest index (sample seed yield/sample TDM) was also determined. Analysis of variance was according to the SAS General Linear Model with mean separation according to DMRT.
View this table:
[in this window]
[in a new window]
|
Table 1 Yield, agronomic data, and whole plant yield components for soybean grown at low, medium, and high plant populations, averaged across 1995 and 1996, Baton Rouge, LA
|
|
View this table:
[in this window]
[in a new window]
|
Table 2 Branch and main stem yield, and branch yield components for low, medium, and high plant populations, averaged across 1995 and 1996, for soybean planted near Baton Rouge, LA. Branch dry matter, number of seed, and number of pod per meter, and seed per pod were not significantly different between populations (P < 0.05)
|
|
Experimental design for red/far red light ratios was a randomized complete block in a split plot arrangement with four replications. Red (645 nm) and far-red (735 nm) light irradiances (Kasperbauer, 1987) were determined on all experimental plots at 14, 28, 42, and 56 d after emergence with a LI-COR portable spectroradiometer (band width of 1 nm) equipped with a fiber optic probe. Based on these recordings, the red/far-red light ratio was calculated. Measurements were taken at the midpoint of plant height according to a method described in Kaul and Kasperbauer (1988). A single plant was removed and replaced with a wooden stake to support the fiber optic probe. The probe was oriented to measure incoming light parallel to the soil surface at north, south, east, and west positions. Ambient red/far-red light ratio was also determined by placing the probe toward the sky at a position just above and parallel to the top of the canopy. Analysis of variance was according to SAS GLM with mean separation by LSD.
|
Results
|
|---|
Yield Compensation Between Populations
Although year and year x population effects were not significant at P < 0.05 (data not shown), plant population trends for both years are shown in Fig. 1
. Plant populations throughout the growing season were stable for the low and medium populations, indicating that little plant attrition occurred (Fig. 1). In contrast, the high plant population showed a steady decline between emergence and R5 in both years. From an initial stand of 45 plants m-2 at 2 wk after emergence, plant population fell to 28 plants m-2 (38% decline) by R5. Results are consistent with other studies demonstrating high rates of plant attrition when heavy seeding rates are used (Ethredge et al., 1989). Plant population and year x plant population had no significant effect (P < 0.05) on yield, with all treatments producing 
4000 kg ha-1 (Table 1).
Significant (P < 0.05) year and year x population effects also did not occur for plant height, days to R7, harvest index, number of seed, seed size, seed per pod, number of pods, branch yield, main stem yield, number of branch nodes, percentage branch nodes becoming reproductive, number of branch reproductive nodes, and branch pod per reproductive node. Similar yield occurred because CGR, initially smaller in low vs. medium and high plant populations, equilibrated across populations by R1 or shortly thereafter in both years (Fig. 2 and 3)
. Crop growth rate equilibration was manifested in similar or greater branch development in low compared with medium and high plant populations (Fig. 4) . Thus, the numbers of seed and pods were similar across populations (Table 1), because low plant populations had a greater number of branch pod per reproductive node than other populations, while also producing a similar number of branch reproductive nodes (Table 2). A similar mechanism of yield compensation occurred in previous studies (Carpenter and Board, 1997a). Plant population and the plant population x year interaction had no significant effect on seed size, seed per pod, or days to R7 (Table 1). Harvest index, although numerically greater in low compared with medium and high plant populations, was statistically similar (P < 0.05) across populations (Table 1). Plants were significantly shorter (P < 0.05) with each reduction in plant population (Table 1). Similar results have been reported (Hoggard et al., 1978).
Initiation of Yield Compensation in Low vs. Higher Plant Populations During the Vegetative Period
Crop growth rate equilibration across plant populations by R1 resulted in similar TDM(R5) of 700 to 800 g m-2 for all treatments (Fig. 2 and 3). This TDM level is considered optimal for pod and seed production (Egli et al., 1987; Board and Harville, 1994). Associated with CGR equilibration, LI(R1) for all populations was near the 95% optimum level (Shibles and Weber, 1966) (Table 3) . Data in Table 3 are presented within years because of a significant (P < 0.05) year x population x days after emergence interaction for LI. Observation of CGR per plant indicated that equilibration of low vs. higher plant populations commenced between 14 to 21 d after emergence (Fig. 5)
. At 14 d after emergence, CGR per plant was similar or nearly similar across populations in both years, indicating that little interplant competition occurred during the first 2 wk of growth for any treatment (Loomis and Connor, 1992). However, by 21 d after emergence, CGR per plant was almost twice as great in low compared with the high population in 1995 (P < 0.05) and 35% greater for the same comparison in 1996 (P < 0.05). Thus, with greater dry matter production per plant commencing at 14 to 21 d after emergence, CGR on an area basis for the low population could accelerate faster than in the higher plant populations, resulting in eventual CGR equilibration near R1. Differences in CGR per plant in low vs. higher plant populations increased during the remainder of the vegetative period and into the reproductive period (Fig. 5). Throughout this time, CGR per plant was consistently greater in low vs. medium populations and in medium vs. high populations.
View this table:
[in this window]
[in a new window]
|
Table 3 Leaf area index (LAI) and light interception (LI) at R1 and R5, and average crop growth rate (CGR) and average leaf expansion rate (LER) during R1 to R5, for soybean planted at low, medium, and high populations near Baton Rouge, LA, in 1995 and 1996
|
|
Greater CGR per plant and the start of CGR equilibration across plant populations was created by greater LIE in low vs. higher populations starting near 21 d after emergence and extending to near R1 (Fig. 6 and 7)
. Throughout the last half of the vegetative period, LIE was usually significantly (P < 0.05) greater in low vs. high plant population, with the medium plant population having an intermediate level. By R1, canopy closure had been reached by most treatments and LIE was at similar minimal levels for all populations. Greater light interception per unit LAI resulted in an early NAR advantage for low compared with higher plant populations during most of the vegetative period (Fig. 6 and 7). For both years, NAR was significantly (P < 0.05) greater in low vs. high plant populations during the entire period (except at 14 d after emergence). Relative growth rate patterns between plant populations followed a pattern similar to that of NAR (Fig. 6 and 7).
In contrast to NAR, LAI (the other factor affecting CGR) during the vegetative period was usually significantly (P < 0.05) less in low vs. higher plant populations (Fig. 2 and 3). Leaf area index differences between plant populations reflected TDM equilibration that occurred during the growing season. By R5, LAI was similar across plant populations, as was the case for TDM. Thus, CGR equilibration during the vegetative period was probably more influenced by greater NAR in low vs. higher plant populations than in any LAI differences. Branch development also played little role in CGR equilibration during the vegetative period, as only 15 to 20% of branch dry matter production occurred during this time (Fig. 4).
Yield Compensation Between Low and Higher Plant Populations During the Early Reproductive Period
At R1, LAI for all populations for both years was near or above 3.00 and LI was close to 95% (Table 3), the optimal level for maximizing CGR (Shibles and Weber, 1966). With near canopy closure, LIE declined to the minimal level [near 20 units of % LI m-2 (leaf area)] across plant populations, resulting in equilibration of NAR and RGR (Fig. 6 and 7). Final equilibration of LIE, NAR, and RGR to minimal levels across plant populations probably occurred shortly after R2 (56 d after emergence) in both years, since 95% LI had been obtained for all populations (data not shown) by that time. Similar relationships between 95% LI, NAR, and RGR have been reported in previous studies (Loomis and Connor, 1992; Carpenter and Board, 1997b). Despite similar NAR across plant populations during the R1 to R5 period and significantly less (P < 0.05) LAI in low vs. higher plant populations at R1 (Table 3), low populations were able to maintain CGR(R1–R5) at levels similar to or better than higher populations (Table 3) in both years. Consequently, TDM, which had been significantly (P < 0.05) less in low vs. higher plant populations at R1 (Fig. 2 and 3), equilibrated across plant populations by R5. Since the NAR advantage of low plant populations was no longer present after canopy closure (50–56 d after emergence), it could not explain how the low plant population maintained CGR(R1–R5) equivalent to higher plant populations despite having lower LAI at the start of reproductive growth.
Low plant populations in both years maintained CGR(R1–R5) relative to higher plant populations because of greater leaf area expansion during the R1 to R5 period (Table 3). In both years, LAI(R1) was significantly less (P < 0.05) in low vs. medium or high plant populations. By R5, LAI was similar across plant populations, ranging from 5.26 to 6.01, indicating that LER had been greater (numerically but not significantly greater) for the low vs. medium and high plant populations during R1 to R5. Greater potential for leaf area expansion in low vs. higher plant populations can be observed better by comparing relative leaf area expansion rates (RLAER) during the vegetative and early reproductive periods (Fig. 8)
. During the first 35 d after emergence, RLAER was similar or only slightly greater in low vs. medium and high plant populations (Fig. 8). Relative leaf area expansion rates began diverging after this time, with significantly (P < 0.05) greater rates in low vs. medium or high plant populations. These differences increased with time, and by 56 d after emergence (R2), RLAER was more than twice as great for low vs. high plant populations in 1995 and four times greater for the same comparison in 1996. Thus, similar CGR(R1–R5) was maintained across plant populations largely because of greater RLAER in low vs. higher plant populations.
Greater RLAER in low vs. higher plant populations during the early reproductive period was due to a greater number of leaves rather than greater area per leaf (Table 4)
. Data in Table 4 are presented within years to be compatible with the RLAER data. For any year–developmental stage combination, area per leaf was never significantly affected by plant population (P < 0.05). The number of leaves at R1 or R2 was less for low vs. higher plant populations, but then accelerated faster for the low plant population such that by R5 it was similar across plant populations. Since most leaf development during R1 to R5 occurs on branches (main stem growth stops shortly after R1 in determinate soybean), greater branch development in low vs. higher plant populations (Fig. 4) explains the greater rate of leaf number production (R1–R5) for the low plant population. Branch development commenced 2 wk prior to R1 for all plant populations (Fig. 4). Despite less TDM for the low vs. higher plant populations between this time and R5 (Fig. 2 and 3), branch dry matter was always equal to or significantly (P < 0.05) greater in low vs. higher plant populations. This occurred because of greater TDM partitioning into branches by the low plant population throughout the branch development period (from 2 wk before R1 to R5) (Fig. 9 and 10)
. Thus, low plant population maintained a CGR(R1–R5) similar or greater than higher plant populations because of greater partitioning of TDM into branch dry matter, which accelerated RLAER and LAI relative to these higher populations. When data were averaged across years [year x population x days after emergence interactions were significant (P < 0.05) only for petioles], population had a large effect on altering partitioning into main stems compared with branches (Fig. 11)
. Partitioning into leaves and petioles was fairly constant throughout the vegetative and into the reproductive period. In contrast, starting at 28 d after emergence, the low population began partitioning relatively more of its TDM into branches vs. main stems compared with the medium and high populations. By 56 d after emergence, the low population partitioned 17.5% of TDM into branches vs. only 6.4% for the high population. Partitioning to branches for the medium population (12.6%) was intermediate. In contrast, partitioning into main stems at this time was only 22.1% for the low population compared with 36.2% for the high population. The medium population again showed an intermediate value of 26.8%.
View this table:
[in this window]
[in a new window]
|
Table 4 Area per leaf and leaf number per unit land area for soybean planted at low, medium, and high plant populations near Baton Rouge, LA, in 1995 and 1996
|
|
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 11 Partitioning of total dry matter (TDM) into branches, main stems, petioles, and leaves as affected by plant population, averaged across 1995 and 1996, for soybean planted near Baton Rouge, LA. Means not followed by the same letter are significantly different according to LSD (P < 0.05). Low population = 80000 plants ha-1; medium population = 145000 plants ha-1; high population = 390000 plants ha-1
|
|
Near initiation of branch development and during the rest of the vegetative period, red/far red light ratios received at the main stem were consistently greater for low vs. higher plant populations (Fig. 9 and 10). Initial recordings made 14 d after emergence indicated that low and medium plant populations had ratios of 1.0 to 1.12 in both years, whereas the high plant population had a significantly lower ratios of 0.6 to 0.8 (Fig. 9 and 10). As canopy development proceeded during the vegetative period, red/far red ratios declined for all plant populations, but maintained the same general relationship of greater in low vs. medium and medium vs. high plant populations. By 42 d after emergence, red/far red light ratios were still about twice as great in low vs. higher plant populations.
|
Discussion
|
|---|
Optimum yield of 4000 kg ha-1 was achieved at a plant population of only 80000 plants ha-1, a level less than one-half the recommended plant population for optimal planting dates in Louisiana (192000 plants ha-1). Yield compensation by the low vs. higher plant populations occurred through the same mechanism of CGR equilibration and yield component adjustments (branch reproductive nodes and pods per reproductive node) described previously (Carpenter and Board, 1997a, 1997b). Low optimal plant population reflected lack of any major environmental stresses occurring during the growing season for both years that could have reduced CGR below the level for optimal yield. Planting at the highest rate (390000 plants ha-1) resulted in considerable plant death (Fig. 1), indicating that many initial plants made no contribution to final yield. Planting at the highest rate also increased plant height (Table 1), which is conducive to lodging (Cooper, 1971).
Although previous research showed NAR and RLAER to be involved in CGR and yield compensation in low vs. higher plant populations (Wells, 1993; Carpenter and Board, 1997a, 1997b), the current study demonstrates how and when they influence CGR and identifies the factors affecting them. Crop growth rate compensation in low vs. higher plant populations commenced as early as 21 d after emergence (Fig. 5), eventually resulting in CGR equilibration near R1 in both years and TDM equilibration by R5 (Fig. 2 and 3). During the vegetative period, greater NAR in low vs. higher plant populations was the dominant factor affecting CGR compensation across plant populations (Fig. 6 and 7). Greater NAR for the low plant population resulted from greater LIE. Apparently, interplant shading and competition for light were less severe in low vs. higher plant populations during much of the vegetative period in both years of the study. Thus, low plant populations intercepted a greater amount of light per unit leaf area and therefore had greater NAR and RGR compared with higher plant populations. Consequently, the early CGR advantage of the high and medium plant populations relative to the low plant population (caused by denser stands) was virtually erased by R1 (Fig. 2 and 3).
Although many studies have dismissed the importance of reduced source strength (i.e., CGR and its components LAI and NAR) during the vegetative period in yield formation (Brun, 1978; Christy and Porter, 1982; Jiang and Egli, 1993), the current study indicates that source strength during the vegetative period greatly influences the yield compensatory ability of low plant populations. Any environmental factor that adversely affected the NAR advantage of a low plant population would impair the ability of sparse stands to achieve the same yield as normal stands. Common environmental stresses in Louisiana and other parts of the southeastern USA are drought and waterlogging. For example, drought stress during the vegetative period could restrict CO2 uptake to such an extent that light intensity was no longer limiting NAR. Also, soil waterlogging could reduce stomatal conductance (Oosterhuis et al., 1990) and lower leaf N content below 2.4 g m-2 (leaf area) (Sinclair and Horie, 1989). Consequently, CO2 or leaf N content could replace photosynthetic irradiance as the factor limiting NAR. Under these conditions, greater LIE for low vs. higher plant population could not stimulate NAR.
As canopy closure was achieved near R1 to R2 (Table 3), LIE, NAR, and RGR for all plant populations were near similar minimal levels (Fig. 6 and 7). Thus, after this time these growth dynamic factors played no further role in yield compensation across plant populations. However, leaf area index was lower at R1 (Table 3) in low vs. higher populations. This would be expected since most leaf development during emergence to R1 occurs on the main stems. The question arises as to how the low population was able to maintain similar or greater CGR(R1–R5) (Table 3) compared with higher populations, despite its lower LAI and lack of the compensatory advantage of greater NAR that it had during the vegetative period. The answer lies in greater RLAER for the low population commencing at 35 d after emergence (Fig. 8). This allowed for greater LER (Table 3) for the low plant population during R1 to R5, resulting in similar LAI across populations at R5 (Fig. 2 and 3). Thus, more rapid production of LAI for low vs. higher populations during R1 to R5 allowed the low population to maintain CGR parity with the other populations.
Since area per leaf was similar across populations at R1, R2, and R5, greater RLAER for low vs. higher populations during the late vegetative and early reproductive periods resulted from a greater rate of leaf number production per square meter (Table 4). Since most leaf production during this time occurs from branch nodes (main stem growth stops shortly after R1 for determinate soybean), and branch dry matter per area and branch nodes per area are highly correlated (Carpenter and Board, 1997a), the most likely explanation for greater RLAER in the low population was increased branch dry matter per area for low vs. higher populations (Fig. 4). The low populations achieved greater levels of branch dry matter per unit land area compared with higher populations throughout the late vegetative and early reproductive periods because of greater TDM partitioning into branches (Fig. 9 and 10). Commencing at 35 to 40 d after emergence, the same time that the low plant population started showing greater RLAER compared with medium and high populations (Fig. 8), this greater partitioning became obvious and was maintained at least until 60 d after emergence. Greater partitioning of TDM into branches was compensated for by less partitioning into main stems (Fig. 11). In other words, as population declines, partitioning of TDM increases to branches and decreases to main stems. Because most LAI development during R1 to R5 occurs on branches in determinate soybean and most yield is from branches (Table 2), the low compared with medium and high plant populations was partitioning more TDM into processes that enhanced yield formation.
Greater TDM partitioning into branch dry matter for low vs. higher populations could be related to altered ratios of red/far light within the canopy. Kasperbauer (1987) and Sanchez et al. (1993) recognized red/far red light as affecting branching and tillering in crop plants. They showed that plants grown at dense populations received lower ratios of red/far red light within the canopy than sparse populations. Although normal sunlight has a red/far red light ratio of 1.2 (Sanchez et al., 1993), this ratio declines as sunlight penetrates the canopy, because red light is absorbed by the vegetation, whereas far red light is reflected. Thus, denser canopies associated with high populations would have lower red/far red light ratios compared with sparse populations. This indeed is what occurred in the current study, where the low population had consistently greater red/far red light ratios than the higher populations (Fig. 9 and 10).
Were differences in red/far red light ratios in the current study sufficient to cause the altered partitioning into branches that helped the low population attain the same yield as higher populations? Based on studies conducted by Kasperbauer (1987), the answer appears to be yes. In his studies, different red/far red light ratios were applied within soybean canopies using altered plant population–row spacing combinations. By 6 wk after emergence, red/far red light ratio (converted from Kasperbauer's far red/red light ratios) received in the highest population (narrowest row spacing) was 0.30, whereas that received in the lowest population (widest row spacing) was 0.71. Concomitant with these changes, the number of branches per square meter (converted from Kasperbauer's number of branches per plant) was only 22 branches m-2 in the highest population compared with 72 branches m-2 in the lowest population. Although Kasperbauer did not measure the number of branch nodes, field studies provide ample evidence that branch dry matter per unit land area, number of branches per area, and number of branch nodes per area are highly correlated (r2 = 0.91–0.93, Carpenter and Board, 1997a). The number of branches for intermediate populations in Kasperbauer's study were between these extremes. He also conducted controlled-environment studies demonstrating that these morphological responses were indeed regulated by red/far red light ratios acting through the plant's phytochrome pigment. In the current study, red/far red light ratios within the canopy prior to the initiation of greater TDM partitioning into branches were comparable with the data presented by Kasperbauer (1987); that is, low populations had ratios of 0.87 to 1.00, whereas medium and high populations had ratios of 0.62 to 0.72 and 0.38 to 0.42, respectively. Thus, the most likely cause of greater TDM partitioning into branch dry matter was altered red/far red light ratios within the canopies.
|
Conclusions
|
|---|
Low population (80000 plants ha-1) in the current study achieved the same yield as medium (145000 plants ha-1) and high (390000 plants ha-1) populations through greater NAR (caused by greater LIE) during the vegetative period and greater RLAER (caused by increased TDM partitioning to branches induced by greater red/far red light ratios) during the late vegetative and early reproductive periods. Because of these two factors, low populations achieved and maintained the same CGR as higher populations during the emergence to R5 period. Consequently, the number of pods per area and yield were similar across populations. Results indicate that growth dynamics during the vegetative period are important for yield compensation between low and higher populations. Thus, farmers wishing to reduce seed costs by planting at lower rates must avoid plant stresses during the vegetative period that would hinder the capacity of low populations to reach CGR comparable with those in higher populations. Results also have implications for genetic approaches to identify and develop cultivars able to achieve optimal yield at low populations. Such cultivars would have greater partitioning of TDM into branch dry matter, as this enhances the RLAER during the late vegetative and early reproductive periods.
|
NOTES
|
|---|
Approved for publication by the Director of the Louisiana Agric. Exp. Stn. as manuscript no. 99-09-0559.
Received for publication October 13, 1999.
|
REFERENCES
|
|---|
- Board J.E., Harville B.G. A criterion for acceptance of narrow-row culture in soybean. Agron. J. 1994;86:1103-1106.[Abstract/Free Full Text]
- Boquet D.J., Walker D.M. Seeding rates for soybeans in various planting patterns. Louisiana Agric. 1980;23:22-23.
- Brun W.A. Assimilation. In: Normal A.G., ed. Soybean physiology, agronomy, and utilization. New York: Academic Press, 1978:45-76.
- Carpenter A.C., Board J.E. Branch yield components controlling soybean yield stability across plant populations. Crop Sci. 1997;37:885-891 a.[Abstract/Free Full Text]
- Carpenter A.C., Board J.E. Growth dynamic factors controlling soybean yield stability across plant populations. Crop Sci. 1997;37:1520-1526 b.[Abstract/Free Full Text]
- Christy A.L., Porter C.A. Development, carbon metabolism and plant productivity. In: Govindjee, ed. Photosynthesis. Vol. II. New York: Academic Press, 1982:499-511.
- Cooper R.L. Influence of soybean production practices on lodging and seed yield in highly productive environments. Agron. J. 1971;63:490-493.[Abstract/Free Full Text]
- Costa J.A., Oplinger E.S., Pendleton J.W. Response of soybean cultivars to planting patterns. Agron. J. 1980;72:153-156.[Abstract/Free Full Text]
- Egli D.B. Plant density and soybean yield. Crop Sci. 1988;28:977-981.[Abstract/Free Full Text]
- Egli D.B., Guffy R.D., Heitholt J.J. Factors associated with reduced yields of delayed plantings of soybean. J. Agron. Crop Sci. 1987;159:176-185.
- Ethredge W.J., Ashley D.A., Woodruff J.M. Row spacing and plant population effects on yield components of soybean. Agron. J. 1989;81:947-951.[Abstract/Free Full Text]
- Fehr W.R., Caviness C.E. Stages of soybean development. Ames, IA: Iowa Agric. Exp. Stn. Spec Rep. 80, 1977.
- Hoggard A.L., Shannon J.G., Johnson D.R. Effect of plant population on yield and height characteristics in determinate soybeans. Agron. J. 1978;70:1070-1072.[Abstract/Free Full Text]
- Hunt R. Plant growth analysis. Southampton, UK: Camelot Press, 1978 Inst. Biol. Stud. Biol. 96..
- Hunt R., Parsons E.T. Plant growth analysis. Users instructions for the stepwise and spline programs. Unit of comparative ecology. Sheffield, UK: Univ. of Sheffield, 1981.
- Jiang H., Egli D.B. Shade induced changes in flower and pod number and flower and fruit abscission in soybean. Agron. J. 1993;85:221-225.[Abstract/Free Full Text]
- Kasperbauer M.J. Far-red light reflection from green leaves and effects on phytochrome-mediated assimilate partitioning under field conditions. Plant Physiol. 1987;85:350-354.[Abstract/Free Full Text]
- Kaul K., Kasperbauer M.J. Row orientation effects on FR/R light ratio, growth and development of field-grown bush bean. Physiol. Plant. 1988;74:415-417.
- Leffel, R.C., and G.W. Barber. 1961. Row widths and seeding rates in soybeans. Univ. of Maryland Agric. Exp. Stn. Bull. 470. College Park, MD.
- Lehman W.F., Lambert J.W. Effects of spacing on soybean plants between and within rows on yield and its components. Agron. J. 1960;52:84-86.[Abstract/Free Full Text]
- Loomis, R.S., and D.J. Connor. 1992. Community concepts. p. 32–59. In Crop ecology: Productivity and management in agricultural systems. Cambridge Univ. Press, Cambridge, UK.
- Lueschen W.E., Hicks D.R. Influence of plant population on field performance of three soybean cultivars. Agron. J. 1977;69:390-393.[Abstract/Free Full Text]
- Moore, S.H., and D.E. Longer. 1987. Optimum plant populations for maximum yield in soybean. p. 11. Arkansas Farm Res. July–August.
- Oosterhuis D.M., Scott H.D., Hampton R.E., Wullschleger S.D. Physiological responses of two soybean [Glycine max (L.) Merr.] cultivars to short-term flooding. Environ. Exp. Bot. 1990;30:85-92.
- Parks, W.L., J. Davis, R. Evans, M. Smith, T. McCutchen, L. Sofley, and W. Sanders. 1982. Soybean yields as affected by row spacing and within row plant density. Univ. of Tenn. Agric. Exp. Stn. Bull. 615. Knoxville, TN.
- Sanchez R.A., Casal J.J., Ballare C.L., Scopel A.L. Plant response to canopy density mediated by photomorphogenic processes. In: Buxton D.R., et al. , ed. International Crop Science I. Madison, WI: CSSA, 1993:779-786.
- Shibles R.M., Weber C.R. Interception of solar radiation and dry matter production by various soybean planting patterns. Crop Sci. 1966;6:55-59.
- Schonbeck M.W., Hsu F.C., Carlsen T.M. Effect of pod number on dry matter and nitrogen accumulation and distribution in soybean. Crop Sci. 1986;26:783-788.[Abstract/Free Full Text]
- Sinclair T.R., Horie T. Leaf nitrogen, photosynthesis, and crop radiation use efficiency: a review. Crop Sci. 1989;29:90-98.
- Wells R. Soybean growth response to plant density: relationships among canopy photosynthesis, leaf area, and light interception. Crop Sci. 1991;31:755-761.[Abstract/Free Full Text]
- Wells R. Dynamics of soybean growth in variable planting patterns. Agron. J. 1993;85:44-48.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
A. Murillo-Williams and P. Pedersen
Early Incidence of Soybean Seedling Pathogens in Iowa
Agron. J.,
September 8, 2008;
100(5):
1481 - 1487.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
A. Murillo-Williams and P. Pedersen
Arbuscular Mycorrhizal Colonization Response to Three Seed-Applied Fungicides
Agron. J.,
May 7, 2008;
100(3):
795 - 800.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. T. Edwards, L. C. Purcell, and D. E. Karcher
Soybean Yield and Biomass Responses to Increasing Plant Population among Diverse Maturity Groups: II. Light Interception and Utilization
Crop Sci.,
August 1, 2005;
45(5):
1778 - 1785.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. E. Board and H. Modali
Dry Matter Accumulation Predictors for Optimal Yield in Soybean
Crop Sci.,
August 1, 2005;
45(5):
1790 - 1799.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
P. Pedersen and J. G. Lauer
Soybean Growth and Development in Various Management Systems and Planting Dates
Crop Sci.,
March 1, 2004;
44(2):
508 - 515.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
B. Rigsby and J. E. Board
Identification of Soybean Cultivars That Yield Well at Low Plant Populations
Crop Sci.,
January 1, 2003;
43(1):
234 - 239.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
Medium and high plant populations are significantly different according to LSD (P < 0.05). 
