Printer Friendly

Dry matter accumulation predictors for optimal yield in soybean.

IDENTIFICATION OF TDM predictors for optimal soybean yield would aid farmers in identifying anticipated yields and determination of environmental factors limiting crop yield. In the current study, optimal yield for the area of study (Louisiana) is considered to be 4500 kg [ha.sup.-1]. This level is based on field studies conducted across a 10-yr period (Board et al., 1990; Board et al., 1992; Linkemer et al., 1998; Board and Harville, 1996; Carpenter and Board, 1997; Board, 2000) and reports from extension personnel (personal communication, D. Lanclos, soybean specialist). Optimal yield levels in other environments and/or with different genotypes may be different from those stated here. Because most cultural and environmental factors influence yield through differences in seed number per area (Jiang and Egli, 1995; Egli, 1998), and seed number per area is largely determined between emergence and R5 (Pigeaire et al., 1986; Board and Tan, 1995), a logical strategy is to develop TDM predictors for the easily recognizable R1 and R5 stages (stages according to Fehr and Caviness, 1977). In this study, a TDM predictor for optimal yield is the TDM required by a certain stage to allow for attainment of optimal yield. Attainment of this level would not necessarily guarantee optimal yield, as stress conditions occurring after the specific stage may result in suboptimal yield. However, establishment of such predictors indicates a producer is on-track to achieve optimal yield, assuming no subsequent stress events, and demonstrates the status of yield formation during certain developmental periods.

Failure to achieve a certain predictor by either R1 or R5, combined with knowledge of timing of stress events, would indicate to the farmer what cultural practices (row spacing, plant population, tillage, irrigation, pesticide application, etc.) to use to remedy the situation. In a study involving row spacing and planting dates, Egli et al. (1987) suggested that a TDM(R5) of 500 g [m.sup.-2] was required to optimize seed number per area and yield. A more recent study involving 14 cultivars and two plant populations supported this finding (Rigsby and Board, 2003). Verification of TDM(R5) as a predictor for optimal yield across a wide range of cultural and environmental factors has not been established. No research has reported on the use of TDM(R1) as a predictor for optimal yield. However, previous studies involving late-planted soybean have linked TDM(R1) with yield (Board et al., 1992; Starling et al., 1998).

Dry matter accumulation predictors have greater practicality than other possible predictors for optimal yield such as yield components (e.g., seed size, seed number per area, node number per area, etc.), leaf area index, and light interception. Although useful for small plot trials, yield components are not promising predictors because of their high variation and the difficulty for assessing them in large-production environments (Board et al., 1990). Another problem with using yield components as predictors of optimal yield is that some of them (e.g., seed size) are formed late in the growing season. Although achievement of 95% light interception by R5 has been proposed as a predictor for optimal yield (Shibles and Weber, 1966), subsequent research demonstrated that, in some cases, 95% light interception before R5 was required to optimize yield (Egli, 1988; Board et al., 1992). Thus, timing for achievement of 95% light interception to obtain optimal yield is not consistent enough to be a valid criterion. Achievement of optimal leaf area index (leaf area index required for 95% light interception) has also been suggested as a predictor for optimal yield but suffers from the same problems as optimal light interception.

Use of TDM at R1 and R5 as predictors for optimal yield potential has appeal because these developmental stages are predictable occurrences for specific cultivar and environment situations (Board and Boethel, 2001). In addition, TDM can be easily determined through spectral analysis (Ma et al., 2001; Pinter et al., 1981). The relative amounts of red and infrared light reflected by the canopy can be used to calculate vegetation indices [e.g., the normalized difference vegetation index and the simple ratio (Aparicio et al., 2000; Penuelas et al., 1997)] which indicate the level of leaf area index and TDM of the crop. Because TDM(R1) and TDM(RS) can be efficient and accurate predictors for optimal yield and because little research has been done on this subject, our primary objectives were to determine (i) If TDM(R1) and TDM(R5) can be used as predictors for optimal yield as influenced by cultural and environmental factors and (ii) to identify yield components that explain the observed relationships between TDM and yield. This aids in explaining how achievement of a certain dry matter predictor results in optimal yield, thus providing a sounder basis for accepting the predictor.


Data Set Construction

Treatment means from six previous studies were combined into a single data set to study relationships between TDM(R1), TDM(R5), yield, and yield components. These studies and the specific treatment combinations that were entered into the data set are listed in Table 1.

Data were analyzed across these studies, rather than within, to have the widest possible range for testing the hypotheses that TDM(R1) and TDM(R5) can be used as predictors for optimal yield. This method of combining data across experiments to determine the efficacy of agronomic relationships has been used by previous researchers (Loomis and Connor, 1992; Robert and Andrew, 1989; Purcell and Specht, 2004). The purposes in these cases, as in the current study, were not to find similar trends within experiments, but to use data from across several studies to identify significant relationships. The studies selected for the data set contained a variety of cultural treatments that altered environmental growing conditions (planting dates, row spacings, plant populations, and water-logging stress) but had limited cultivar variability. Thus, the data set provided a wide range of environmentally influenced yield and TDM levels (with little genetic variation) for achieving our goals. In all cases, field studies were closely monitored to avoid and/or ameliorate any abiotic and/or biotic stresses other than treatment effects.

Data used for the current study were yield, TDM(R1), TDM(R5), and the following yield components: seed size (g/100 seed), seed per pod (no.), seed number per area (no. [m.sup.-2]), pod number per area (no. [m.sup.-2]), pod per reproductive node (no.), reproductive node number per area (no. [m.sup.-2]) (node bearing at least one pod having at least one seed), node number per area (no. [m.sup.-2]), and percent reproductive nodes (%) (percentage of all nodes that become reproductive). Treatment means (four replications) for these parameters were combined into one set and subjected to regression analyses.

Yield at maturity was determined by combine harvest of interior rows (6.5 [m.sup.2]) of each plot that had been end-trimmed to 4.3 m and corrected to 130 g [kg.sup.-1] moisture. A 10-plant sample was taken at maturity to determine seed per pod, pod per reproductive node, and harvest index. Samples were dried to constant weight in a forced-air dryer at 60[degrees]C and then separated into seed and stem portions. Harvest index was calculated as (seed weight/total sample weight) x 100. Yield components were determined as follows.

1. Seed size (g per 100 seed) was determined by counting 300 seed from each yield sample with an automatic seed counter, drying the seed for 3 d to constant weight at 60[degrees]C in a forced-air dryer, weighing the sample, and then dividing the weight by three.

2. Seed number per area (number [m.sup.-2]) was determined by first converting yield in kg [ha.sup.-1] at 130 g [kg.sup.-1] moisture to dry yield in g [m.sup.-2] at 30 g [kg.sup.-1] (the same moisture content for seed size), and then dividing dry yield by seed size (as g per seed). Thus, g [m.sup.-2] (dry yield)/g [seed.sup.-1] (individual seed size) calculates seed number per area.

3. Seed per pod (number) was determined from the same 10-plant sample used to determine harvest index. The number of bulging locules in a subsample of 100 randomly selected pods were counted to determine seed per pod.

4. Pod number per area (number [m.sup.-2]) was calculated by dividing seed [m.sup.-2] by seed per pod (seed [m.sup.-2]/seed per pod = pods [m.sup.-2]).

5. Pod per reproductive node (number) was determined from the same sample used for determination of harvest index. All reproductive nodes (a reproductive node is defined as a node bearing at least one pod having at least one seed) and pods in the samples were counted and pod per reproductive node determined by dividing pod number by reproductive node number.

6. Reproductive node number per area (number [m.sup.-2]) was determined by dividing pod [m.sup.-2] by pod per reproductive node (pod [m.sup.-2]/pod per reproductive node = reproductive node [m.sup.-2]).

7. Percentage of reproductive nodes (%) was determined from the same sample used to determine harvest index. Reproductive and total node numbers were determined and the fraction of nodes becoming reproductive determined as (reproductive nodes/total nodes) x 100.

8. Node number per area (number [m.sup.-2]) was calculated as reproductive node [m.sup.-2]/fraction of nodes becoming reproductive.

9. Harvest index was calculated from the 10-plant sample as (seed weight/total sample weight) x 100.

The aforementioned yield components were classified into primary traits affecting yield (seed number per area and seed size), secondary traits affecting seed number per area (seed per pod and pod number per area), tertiary traits affecting pod number per area (pod per reproductive node and reproductive node number per area), and quaternary traits affecting reproductive node number per area (percentage reproductive nodes and node number per area).

Appropriate methods for regression analyses of the data were selected based on consultations with Dr. Jay Geaghan, Department of Experimental Statistics at Louisiana State University (personal communication). Regression analyses were done with the PROC NLIN, PROC UNIVARIATE, and SAS regression (PROC GLM) procedures of the SAS system. Relationships between TDM(R5) or TDM(R1) with yield, seed number per area, pod number per area, reproductive node number per area, and node number per area were analyzed by negative exponential regression analyses provided by PROC NLIN and PROC UNIVARIATE. This model is described as Y = Ymax (1 - [e.sup.-cx]). Ymax is the asymptotic Y value and c is a constant. All other analyses were done using SAS regression (PROC GLM) in which linear, quadratic, and cubic components were successively tested for significance and included if the residual sum of squares was significantly reduced (p < 0.05). No procedure was employed to identify and remove outlier data points.


Yield was inversely related to harvest index (Fig. 1) in a linear relationship ([r.sup.2] = 0.40). In contrast, Yield was closely positively related with TDM(R1) and TDM(R5) but not in a linear manner (Fig. 1). The relationship between yield and TDM(R5) was described by a close negative exponential model ([r.sup.2] = 0.89). Yield increased steeply with TDM(R5) at low dry matter levels (<300 g [m.sup.-2]). Yield responses to increased TDM(R5) progressively declined as dry matter rose above this level. Yield did not respond to TDM(R5) above a level of about 600 g [m.sup.-2]. Yield responded to TDM(R1) in a similar fashion ([r.sup.2] = 0.79) and reached a plateau at about 200 g [m.sup.-2].


Among the primary yield components, seed number per area was more closely related to both dry matter parameters and yield than was seed size (Fig. 2, Table 2). Seed number per area and yield demonstrated a close linear relationship ([r.sup.2] = 0.83, Fig. 2, Table 2). Seed number per area responded to TDM(R5) in a manner similar to that shown for yield (negative exponential relationship, [r.sup.2] = 0.82, Fig. 2). Large increases in seed number per area with increased TDM(R5) occurred at low dry matter levels (<300 g [m.sup.-2]) and then progressively declined as TDM(R5) increased to 500 g [m.sup.-2], a level somewhat below that shown for yield. Seed number per area was also closely linked to TDM(R1) (negative exponential relationship, [r.sup.2] = 0.70, Fig. 2) and plateaued at a TDM(R1) of about 150 g [m.sup.-2]. In contrast to seed number per area, seed size was not closely related to either TDM(R1) or TDM(R5) ([r.sup.2] = 0.20 and 0.26, linear and quadratic relationships, respectively; Table 2). Similarly, seed size and yield were not highly correlated ([r.sup.2] = 0.27, linear relationship, Table 2).


For secondary yield components affecting seed number per area, pod number per area was more closely linked to yield, TDM(R1), and TDM(R5) than was seed per pod (Fig. 3, Table 2). Pod number per area was closely linearly correlated with yield ([r.sup.2] = 0.91, Fig. 3). Pod number per area responded to TDM(R1) and TDM(R5) in a fashion similar to that shown for seed number per area and yield (Fig. 3), demonstrating close negative exponential relationships ([r.sup.2] = 0.75 and 0.88 for TDM(R1) and TDM(R5), respectively). Plateau levels of TDM(R5) and TDM(R1) for maximal pod number per area were 600 and 200 g [m.sup.-2], respectively, the same levels shown for yield. In contrast, seed per pod was not closely correlated with yield ([r.sup.2] = 0.18, linear relationship, Table 2) and with both dry matter parameters ([r.sup.2] = 0.20, linear relationships, Table 2).


Of the two tertiary yield components affecting pod number per area, reproductive node number per area had a closer relationship with yield, TDM(R1), and TDM(R5) than did pod per reproductive node (Fig. 4, Table 2). Yield responded to reproductive node number per area in a quadratic fashion ([r.sup.2] = 0.72, Fig. 4). Yield increased steeply as reproductive node number per area increased to 400 nodes [m.sup.-2]. Above this level, yield responses became smaller and appeared to reach a plateau at about 800 nodes [m.sup.-2]. In contrast, pod per reproductive node demonstrated only a slight linear correlation with yield ([r.sup.2] = 0.20, Table 2). Reproductive node number per area was related to TDM(R1) in a moderately close negative exponential relationship ([r.sup.2] = 0.55, Fig. 4) and to TDM(R5) in a similar fashion ([r.sup.2] = 0.61, Fig. 4). Levels of TDM(R1) and TDM(R5) associated with maximal reproductive node number per area were 150 and 400 g [m.sup.-2], respectively, levels lower than those for pod number per area, seed number per area, and yield. Pod per reproductive node was not closely linked to both dry matter parameters [[r.sup.2] = 0.20 and 0.27 for TDM(R1) and TDM(R5), respectively, linear relationships] (Table 2).


For the quaternary yield components affecting reproductive node number per area, node number per area was more closely linked with yield, TDM(R1), and TDM(R5) than was percent reproductive nodes (Fig. 5, Table 2). Yield responded to node number per area in a quadratic response similar to the yield response to reproductive node number per area (Fig. 5). Yield appeared to reach a plateau at about 1000 node [m.sup.-2]. The relationships between node number per area with TDM(R1) ([r.sup.2] = 0.58, negative exponential relationship) and TDM(R5) ([r.sup.2] = 0.63, negative exponential relationship) paralleled the patterns shown for reproductive node number per area, pod number per area, seed number per area, and yield (Fig. 5). Levels of TDM(R1) and TDM(R5) that maximized node number per area were the same as those for reproductive node number per area. In contrast, percent reproductive nodes was not significantly related to yield, TDM(R1), or TDM(R5) (Table 2).



TDM(R1) and TDM(R5) as Predictors for Optimal Yield

The results of the current study clearly indicate that TDM(R1) and TDM(R5) can be used as predictors for optimal yield in soybean. The existence of a negative relationship between yield and HI (Fig. 1) demonstrated that environmental and cultural factors did not cause greater yield through increased dry matter partitioning into the seed. Instead, these factors affected yield predominantly through dry matter accumulation. A similar result, where yield was found to be more related to crop growth rate than to partitioning of assimilate, was reported by Egli (1988) and Egli and Yu (1991).

The relationships of yield with TDM(R1) and TDM(R5) were well described by negative exponential relationships. In essence, dry matter accumulation affected yield up to a point, but not beyond. A TDM(R1) of 200 g [m.sup.-2] can be used as a predictor for optimal yield (Fig. 1). At R5, a TDM of 600 g [m.sup.-2] accurately predicted optimal yield. Greater [r.sup.2] for yield vs. TDM(R5) compared with yield vs. TDM(R1) probably resulted because TDM(R5) is accrued over a longer period of dry matter accumulation during the period for which seed number per area is determined (Board and Tan, 1995). Another factor may be that R5 occurs closer to harvest. However, the close relationship between yield and TDM(R1) indicates that the environmental and cultural changes affecting yield had an important influence during the vegetative period, and not just in the reproductive period as claimed by others (Egli, 1997). Results indicate that producers may need to be watchful of environmental stresses not only after but also before R1. Our results support previous findings (Egli et al., 1987; Rigsby and Board, 2003), demonstrating the robustness of TDM(R5) as a predictor for optimal yield in soybean. Duncan (1986) also postulated that greater TDM(R5) resulted in higher seed yields of soybean.

Identification of TDM levels at R1 and R5 as predictors for optimal yield potential aids farmers in identifying stresses adversely affecting yield. Most environmental stresses reduce soybean yield through reducing crop growth rate (between emergence and R5), TDM(R5), and seed number per area (Egli and Yu, 1991; Board et al., 1992). Exceptions do occur when certain stresses affect yield formation events without affecting TDM (e.g., temperature effects on pollen sterility); however, such factors have less of an impact on yield (Egli, 1998). Monitoring seasonal TDM accumulation pattern helps identify environmental stresses affecting yield (Loomis and Connor, 1992). In cases where suboptimal TDM accumulation (TDM below the predictor level for optimal yield) is apparent early in the growing season (e.g., at R1), the factor limiting crop growth would be something acting soon after crop emergence such as mineral toxicities or deficiencies, soil pH extremes, inadequate inoculation, suboptimal plant population, or row spacing that is too wide. However, if TDM is optimal early in the growing season but becomes suboptimal by R5 [as evidenced by achievement of optimal TDM(R1), but suboptimal TDM(R5)], then the environmental stress is something that was initially available in adequate amounts, but became limiting as growth progressed. Examples of this are water stress, mineral deficiencies and inadequate prevailing light. Suboptimal yield, even with the achievement of TDM(R1) and TDM(R5) at levels equal to or above that required for optimal yield, indicates that the stress factor did not affect the crop growth rate between emergence and R5 but was acting during the seed filling period. In this case, possible stress factors include late-season biotic stresses (diseases, insects, or weeds) or a severe late-season drought stress. Thus, identification of TDM(R1) and TDM(R5) predictors for optimal yield combined with a knowledge of seasonal environmental events will aid farmers in identifying the stresses reducing their yields.

Suitability of TDM(R1) and TDM(R5) as predictors for optimal yield is enhanced by other factors. The R1 and R5 stages are easily recognizable (Fehr and Caviness, 1977) and are highly predictable for a given cultivar and environment situation (Board and Boethel, 2001). Another advantage is that TDM can cheaply, quickly, and easily be determined from vegetation indices determined by remote sensing (Tucker, 1979; Ma et al., 2001).

Yield Component Analysis

Analyses of relationships between yield and yield components with TDM(R1) and TDM(R5) explain why yield responded to TDM(R1) and TDM(R5) in the manner presented in this study. Regression analyses indicated that yield components which responded to TDM accumulation [as measured by either TDM(R1) or TDM(R5)] were also the most important in yield formation. These were seed number per area, pod number per area, reproductive node number per area, and node number per area. Seed size, seed per pod, pod per reproductive node, and percent reproductive nodes were not closely linked with yield or dry matter accumulation. The pattern of yield response to TDM(R1) and TDM(R5) reflected dry matter accumulation effects on the four aforementioned yield components that played important roles in yield formation. However, these four yield components did not have similar optimal TDM(R1) and TDM(R5) levels. Yield, seed number per area, and pod number per area showed plateau responses to TDM(R1) at 150 to 200 g [m.sup.-2] and to TDM(R5) at 500 to 600 g [m.sup.-2] (Fig. 1, 2, and 3). Reproductive node number per area and node number per area showed lower levels of TDM(R1) (150 g [m.sup.-2]) and TDM(R5) (400 g [m.sup.-2]) for optimal production (Fig. 4 and 5). Apparently, dry matter levels that saturate production for these yield components are less than those required for pod and seed production, and yield.

Data suggest that dry matter accumulation is important to yield formation up to a point, but not beyond. Production of yield components significant for yield formation strongly responds to increased TDM when TDM levels are low. At greater TDM levels, production of significant yield components (per unit increased TDM) declines and eventually becomes zero. Thus, supraoptimal TDM levels may be unnecessary and even harmful for increased yield [e.g., increased lodging (Cooper, 1971)]. Results of this study indicate to farmers what levels are necessary for optimal yield. Thus, cultural (and possibly genetic) factors should be adjusted to achieve these levels.

Although seed number per area, pod number per area, reproductive node number per area, and node number per area responded to TDM in a similar nonlinear fashion (negative exponential relationships), yield responses to these yield components differed. Yield was related to seed number and pod number per area in strong linear relationships (Fig. 2 and 3), whereas yield responded to reproductive node and node number per area in quadratic fashions (Fig. 4 and 5), suggesting that as node numbers increased, each node became less efficient at producing pods and seeds. Results indicated that although yield continued to increase with pod and seed production, it was saturated at 800 reproductive node [m.sup.-2] or 1000 node [m.sup.-2]. This suggests that seed number per area and yield are controlled by node production up to these levels, but that once optimal node number is achieved, some other factor controls how environmental and cultural factors affect pod and seed production, as well as final yield. Interactions between yield components (to be discussed in a companion publication) may help elucidate this issue.

Because of the limited genetic variability across experiments, data from the current study cannot be extrapolated from the environmental to the genotypic level. However, previous studies involving broad ranges of cultivars demonstrated that genotypic yield differences were controlled by seed number per area and pod number per area (Board et al., 2003), a mechanism similar to that shown in the current study. However, linkages between TDM with yield and yield components were not studied in Board et al. (2003), and so the role of TDM in explaining genotypic yield differences remains unclear. Further evidence from Wisconsin suggests that genotypic yield formation may reflect the environmental yield formation pattern described in the current paper (Pedersen and Lauer, 2004). Modern cultivars Spansoy and CX232 had greater dry matter accumulation than the old cultivar Hardin, suggesting that genetic yield improvement was linked with greater dry matter accumulation. Perhaps yield component responses to TDM (on the genotypic level) may help explain why certain genotypes or cultivars yield better than others in a specific environment.

Extrapolation of our results to the midwestern USA is, of course, a speculative matter. Yield levels in the midwestern USA are much greater than those in the southeastern USA. For example, average yield in Iowa during 2000 to 2002 (2999 kg [ha.sup.-1]) was 50% greater than that in Louisiana (2020 kg [ha.sup.-1]) (Wilcox, 2004). However, it is our belief that, although maximal yield may be greater, yield-TDM-yield component relationships outlined in the current manuscript would be similar to those found in the midwestern USA. Recently, research from Wisconsin demonstrated that consistent yields across a range of environmental factors were associated with TDM(R5) above 500 g [m.sup.-2] (Pedersen and Lauer, 2004), a finding that supports our contention of what dry matter level is required for predicting optimal yield. It is intuitive that a certain amount of TDM is necessary for yield formation and that, at least to some extent, TDM and yield are linked. At the same time, it has long been recognized that greater TDM was not always associated with greater yield (Shibles and Weber, 1966; Weber et al., 1965). Our results demonstrate that yield and TDM are tightly linked at low TDM levels but that this relationship is less direct as TDM increases until a certain TDM plateau is reached, beyond which yield does not increase. Although maximal yields and TDM plateaus for maximal yield may differ in other environments compared with ours, the general response of yield and yield components to TDM probably will follow patterns similar to those described here.


Predictors for optimal yield were established at a TDM(R1) of 200 g [m.sup.-2] and a TDM(R5) of 600 g [m.sup.-2]. These levels for optimal dry matter accumulation reflected yield responses to certain yield components and how these yield components, in turn, were influenced by TDM. Yield components important in yield formation were seed number per area, pod number per area, reproductive node number per area, and node number per area. These yield components responded to TDM(R1) and TDM(R5) in negative exponential relationships that were similar to those shown by yield. Predictors for optimal yield based on TDM levels can be useful tools to farmers for predicting yield potential and diagnosing stress factors that limit yield.

Abbreviations: R1, first flowering; R5, start of seed filling; TDM, total dry matter.


We wish to thank Dr. Jay Geaghan of the Department of Experimental Statistics at Louisiana State University for advice on regression analyses of the data. Gratitude is also extended to Vinay Desai, Dinesh Maricherla, and Rajesh Bodapaty for assistance with computer data analyses and construction of figures.


Aparicio, N., D. Villegas, J. Casadesus, J.L. Araus, and C. Royo. 2000. Spectral vegetation indices as nondestructive tools for determining durum wheat yield. Agron. J. 92:83-91.

Board, J.E. 2000. Light interception efficiency and light quality affect yield compensation of soybean at low plant populations. Crop Sci. 40:1285-1294.

Board, J.E., and D.J. Boethel. 2001. Light interception: A way for soybean farmers to determine when to spray for defoliating insects. Louisiana Agric. 4:8-10.

Board, J.E., and B.G. Harville. 1996. Growth dynamics during the vegetative period affects yield of narrow-row, late-planted soybean. Agron. J. 88:567-572.

Board, J.E., B.G. Harville, and A.M. Saxton. 1990. Narrow-row seed yield enhancement in determinate soybean. Agron. J. 82:64-68.

Board, J.E., M. Kamal, and B.G. Harville. 1992. Temporal importance of greater light interception to increased yield in narrow-row soybean. Agron. J. 84:575-579.

Board, J.E., M.S. Kang, and M.L. Bodrero. 2003. Yield components as indirect selection criteria for late-planted soybean cultivars. Agron. J. 95:420-429.

Board, J.E., and Q. Tan. 1995. Assimilatory capacity effects on soy- bean yield components and pod number. Crop Sci. 35:846-851. Carpenter, A.C., and J.E. Board. 1997. Branch yield components controlling soybean yield stability across plant populations. Crop Sci. 37:885-891.

Cooper, R.L. 1971. Influence of early lodging on yield of soybean. Agron. J. 63:449-450.

Duncan, W.G. 1986. Planting patterns and soybean yield. Crop Sci. 26:584-588.

Egli, D.B. 1988. Alterations in plant growth and dry matter distribution in soybean. Agron. J. 80:86-90.

Egli, D.B. 1997. Cultivar maturity and response of soybean to shade stress during seed filling. Field Crops Res. 52:1-8.

Egli, D.B. 1998. Yield components-Regulation by the seed. p. 113-153. In Seed biology and the yield of grain crops. CAB International, New York.

Egli, D.B., R.D. Guffy, and J.J. Heitholt. 1987. Factors associated with reduced yields of delayed plantings of soybean. J. Agron. Crop Sci. 159:176-185.

Egli, D.B., and Z. Yu. 1991. Crop growth rate and seeds per unit area in soybean. Crop Sci. 31:439-442.

Fehr, W.R., and C.E. Caviness. 1977. Stages of soybean development. Spec. Rep. 80. Iowa Agric. Exp. Stn., Ames, IA.

Jiang, H., and D.B. Egli. 1995. Soybean seed number and crop growth rate during flowering. Agron. J. 87:264-267.

Linkemer, G., J.E. Board, and M.E. Musgrave. 1998. Waterlogging effects on growth and yield components in late-planted soybean. Crop Sci. 38:1576-1584.

Loomis, R.S., and D.J. Connor. 1992. Community concepts, p. 32-39. In Crop ecology: Productivity and management in agricultural systems. Cambridge Univ. Press, Cambridge, England.

Ma, B.L., L.M. Dwyer, C. Costa, E.R. Cober, and M.J. Morrison. 2001. Early prediction of soybean yield from canopy reflectance measurements. Agron. J. 93:1227-1234.

Pedersen, P., and J.G. Lauer. 2004. Soybean growth and development in various management systems and planting dates. Crop Sei. 44: 508-515.

Penuelas, J., R. Isla, I. Filella, and J.L. Araus. 1997. Visible and near infrared reflectance assessment of salinity effects on barley. Crop Sci. 37:198-202.

Pigeaire, A.C., C. Duthion, and O. Turc. 1986. Characterization of the final stage in seed abortion in indeterminate soybean, white lupin and pea. Agronomie 6:371-378.

Pinter, P.J., Jr., R.D. Jackson, S.B. Idso, and J.R. Reginato. 1981. Multidate spectral reflectance as predictors of yield in water stressed wheat and barley. Int. J. Remote Sens. 2:43-48.

Purcell, L.C., and J.E. Specht. 2004. Physiological traits for ameliorating drought Stress. In H.R. Boerma and J.E. Specht (ed.) Soybeans: Improvement, production, and uses, 3rd ed. Agron. Monogr. 16. ASA, CSSA, and SSSA, Madison, WI.

Rigsby, B., and J.E. Board. 2003. Identification of soybean cultivars that yield well at low plant populations. Crop Sci. 42:234-239.

Robert, K.M.H., and J.W. Andrew. 1989. Interception of solar radiation by the crop canopy, p. 28-36. In An introduction to the physiology of crop yield. John Wiley & Sons, New York.

Shibles, R.M., and C.R. Weber. 1966. Interception of solar radiation and dry matter production by various soybean planting patterns. Crop Sci. 6:55-59.

Starling, M.E., C.W. Wood, and D.B. Weaver. 1998. Starter nitrogen and growth habit effects on late-planted soybean. Agron. J. 90: 658-662.

Tucker, C.J. 1979. Red and photographic infrared linear combinations for monitoring vegetation. Remote Sens. Environ. 8:127-150.

Weber, C.R., R.M. Shibles, and D.E. Byth. 1965. Effect of plant population and row spacing on soybean development and production. Agron. J. 58:99-102.

Wilcox, J.R. 2004. World distribution and trade of soybean. In H.R. Boerma and J.E. Specht (ed.) Soybeans: Improvement, production, and uses, 3rd ed. Agron. Monogr. 16. ASA, CSSA, and SSSA, Madison, WI.

James E. Board * and Harikrishna Modali

Dep. of Agronomy and Environmental Management, Louisiana Agric. Exp. Stn., Louisiana State Univ. Agric. Ctr., Baton Rouge, LA 70803. Research support provided by the Louisiana Soybean Promotion Board. Received 13 Oct. 2004. * Corresponding author (jboard@agctr.
Table 1. A listing of journal articles from which data were obtained
for the current study with various cultural, environmental, and
genotypic data for individual studies conducted for soybean grown
near Baton Rouge, LA, 1987 through 1996.

Reference                   Years   Soil                 Planting dates

Board et al., 1990          1987    Mhoon silty clay     22 May 1987
                            1988       clay ([dagger])   12 July 1987
                                                         24 May 1988
                                                          6 July 1988
Board et al., 1992          1989    Mhoon silty clay     26 July 1989
                            1990                         10 July 1990
Board and Harville, 1996    1992    Commerce silt loam   15 July 1992
                            1993      loam (double       12 July 1993
Carpenter and Board, 1997   1994    Commerce silt loam   25 May 1994
                            1995                         23 May 1995
Linkemer et al., 1998       1993    Commerce silt loam   12 July 1993
                            1994                         16 Aug. 1994
Board, 2000                 1995    Commerce silt loam   23 May 1995
                            1996                         17 May 1996

Reference                   Years   Cultivars (MG)        Row spacing


Board et al., 1990          1987    'Centennial' (VI)         100
                            1988    'Forrest' (V)              50
Board et al., 1992          1989    Centennial (VI)           100
                            1990                               75
Board and Harville, 1996    1992    Centennial (VI)           100
                            1993                               50
Carpenter and Board, 1997   1994    'DP415' (V)                75
Linkemer et al., 1998       1993    Centennial (VI)           100
                            1994                               25
Board, 2000                 1995    'DP3606' (VI)              75

Reference                   Years    Plant population    Irrigation

                                    plants [ha.sup.-1]

Board et al., 1990          1987          220 000        No irrigation
                            1988          325 000
Board et al., 1992          1989          325 000        Irrigation
                            1990                           applied
Board and Harville, 1996    1992          220 000        No irrigation
Carpenter and Board, 1997   1994          234 000        No irrigation
                            1995          189 000
                                          164 000
                                           70 000
Linkemer et al., 1998       1993          220 000        No irrigation
Board, 2000                 1995          390 000        No irrigation
                            1996          145 000
                                           80 000

([dagger]) Fine-silty, mixed, nonacid, thermic Typic Fluvaquent.

([double dagger]) Fine-silty, mixed, nonacid, thermic Aeric Fluvaquent.

Table 2. [R.sup.2] values and F values for yield and total dry matter
(TDM) at R1 and R5 as related to harvest index and various yield
components for soybean grown across a range of environ-mental and
cultural conditions near Baton Rouge, LA, 1987 through 1996. Linear
and quadratic components are designated by [B.sub.1] and [B.sub.2],

variable    Independent            [R.sup.2]            F values
(Y)         variable (X)            values

Yield       harvest index            0.40      [B.sub.1] = 34.6 ****
Yield       seed number per area     0.83      [B.sub.1] = 16.01 ****
Yield       seed size                0.27      [B.sub.1] = 19.3 ****
Yield       seed per pod             0.18      [B.sub.1] = 11.6 **
Yield       pod number per area      0.91      [B.sub.1] = 23.47****
Yield       pod per rep. nd          0.2       [B.sub.1] = 12.9 ***
Yield       rep. node per area       0.72      [B.sub.1] = 122.04****,
                                               [B.sub.1] = 10.17 **
Yield       node number per area     0.71      [B.sub.1] = 116.64 ****,
                                               [B.sub.2] = 9.7 **
Yield       per rep. nodes           0.03      NS
TDM (R5)    seed size                0.26      [B.sub.1] = 13.9 ***,
                                               [B.sub.2] = 4.4 *
TDM (R5)    seed per pod             0.2       [B.sub.1] = 13.5 ***
TDM (R5)    pod per rep. nd          0.27      [B.sub.1] = 20 ****
TDM (R5)    per rep. nodes           0.007      NS
TDM (R5)    seed size                0.2       [B.sub.1] = 13.3 ****
TDM (R5)    seed per pod             0.2       [B.sub.1] = 10.9 **
TDM (R5)    pod per rep. nd          0.2       [B.sub.1] = 13.5 ***
TDM (R5)    per rep. nodes           0.02      NS

* F value is statistically significant at the 0.05 probability level.

** F value is statistically significant at the 0.01 probability level.

*** F value is statistically significant at the 0.001 probability

**** F value is statistically significant a at the 0.0001 probability
COPYRIGHT 2005 Crop Science Society of America
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2005 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Board, James E.; Modali, Harikrishna
Publication:Crop Science
Geographic Code:1USA
Date:Sep 1, 2005
Previous Article:Nondestructive measurement of carotenoids in plant tissues by fluorescence quenching.
Next Article:Planting configuration x cultivar effects on soybean production in low-yield environments.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters