Soybean Yield and Yield Component Response to Plant Density in Narrow Row Systems

Search PMN

Michael Epler and Scott Staggenborg, Department of Agronomy, Kansas State University, Manhattan, KS 66506

Abstract

Soybean (Glycine max) yield is responsive to changes in plant population and row spacing under dryland conditions. Soybean seeding rate recommendations in many states were developed prior to the release of herbicide tolerant varieties when soybean seed costs were lower. As a result, many recommendations are high, and soybean performance at low plant densities in narrow rows needs to be reevaluated. This study was conducted to evaluate soybean yield response to plant density in narrow row, drilled systems (7.5 and 15 inches) and evaluate the effects of plant population on yield components. The study was conducted in six environments with optimum plant populations ranging from 80,000 to 140,000 plants/acre. The two yield components responsible for the yield optimization were pods per plant and pods per acre. As plant population increased, pods per plant decreased steadily; however, yield was not reduced either year of the study by the loss of pods per plant because pods per acre increased as plant population increased. The optimum seeding rates for each year of the study were below the 150,000 to 160,000 plants/acre seeding rate normally recommended in the region. The decreased optimum plant populations in both years did not adversely affect yield and might be useful in reducing input costs in dryland soybean production.

Introduction

Soybean yield responses to plant density and row spacing have been studied extensively during the past two decades. Many studies examined physiological responses to narrow row spacing (7.5-inch or less) as this management system evolved (5,6,9,13,14,16,26). Early seeding rate recommendations for soybean grown in narrow rows were quite high (~ 200,000 seed/acre), but seed costs were quite low prior to the release of herbicide tolerant varieties (27).

Early narrow row soybean research was conducted under conventionally tilled systems with grain drills designed for wheat seed metering and placement. The combination of this tillage system and planting equipment often resulted in low emergence rates. This, combined with low seed costs is likely the genesis of high seeding rate recommendations.

The ability of soybean to compensate for low population has resulted in little or no response to changes in plant population from 70,000 up to more than 400,000 seed/acre (1,2,4,11,19,20,25). Devlin et al. (12) found that yields in Kansas are normally affected by water availability, planting date, and soil type. This illustrates that environmental conditions can contribute to the non-response of soybean yields to increasing seeding rates.

Soybean responds to row spacing and population shifts through yield component adjustments. Carpenter and Board (10) attributed soybean yield compensation at low densities to increased branching, while others reported that pods per plant and seeds per plant were the source of compensation (3,6,20,22,28). Board et al. (6) and Bullock et al. (9) reported that early season light interception was the reason for narrow row soybean having higher yields than those grown in wide rows. In wider rows, decreased light interception led to decreased branching, pods per plant, and seeds per plant which provided the yield advantage in narrow row soybean. These researchers also reported decreased pods per plant, and Bullock et al. (9) reported decreased branching as row width increased, which suggests that added inter-row space does not always translate into increased branches and pods per plant.

Because soybean has the ability to adjust yield components at low plant densities to maintain yield levels, it is possible that soybean seeding rates in narrow row systems can be reduced without reducing grain yields. Therefore, our objectives were to: (i) evaluate yield, yield component, and plant characteristic responses to plant population and row spacing; and (ii) determine the minimum soybean plant population that can be achieved in Kansas without decreasing yields.

Field Studies

This study was conducted in 2005 and 2006. Plots were planted at Manhattan, KS, Ottawa, KS, and Rossville, KS in 2005 and 2006. Table 1 shows the soil series and taxonomic description for each of the locations and years included in the study.

Table 1. Soil series and taxonomic descriptions for all 2005 and 2006
plot locations.

Year	Location	Soil series	Taxonomic description
2005	Manhattan	Reading silt loam	Fine-silty, mixed, superactive, mesic Pachic Argiudolls
2005	Ottawa	Woodson silt loam	Fine, smectitic, thermic Abruptic Argiaquolls
2005	Rossville	Eudora silt loam	Coarse-silty, mixed, superactive, mesic Fluventic Hapludolls
2006	Manhattan	Smolan silt loam	Fine, smectitic, mesic Pachic Argiustolls
2006	Ottawa	Woodson silt loam	Fine, smectitic, thermic Abruptic Argiaquolls
2006	Rossville	Eudora silt loam	Coarse-silty, mixed, superactive, mesic Fluventic Hapludolls

All small plots were sowed with a no-till drill (Model 3P605NT, Great Plains Mfg., Salina, KS) with 7.5-inch row spacing. In 2005 and 2006 treatments consisted of five seeding rates: 60, 100, 140, 180, and 220,000 seeds/acre. In 2005, plots had 7.5-inch row spacings and in 2006, a 15-inch row spacing treatment was added at all locations. At all three locations in 2005, the experimental design was a randomized complete block with four replications. In 2006, a split plot within a randomized complete block design with row spacing as the main plot and seeding rate as the sub-plots was used at all locations. Plot size was 6 by 30 ft at all locations. The tillage system at the 2005 and 2006 Manhattan and Ottawa locations was no-till, while the 2005 and 2006 Rossville locations were planted into a conventional-till system. Table 2 shows the planting date, row spacing, previous crop, and variety of seed used for all locations in 2005 and 2006.

Table 2. Planting dates, row spacing, previous crop, and soybean variety for 2005 and 2006 soybean plots.

Year	Location	Planting date	Row spacing (inches)	Previous crop	Soybean variety
2005	Manhattan	17 May	7.5	soybean	Asgrow brand 3701 RR
2005	Ottawa	27 May	7.5	sorghum	Asgrow brand 3701 RR
2005	Rossville	20 May	7.5	corn	Asgrow brand 3701 RR
2006	Manhattan	16 May	7.5 & 15	sorghum	Garst brand 3824 RR
2006	Ottawa	19 May	7.5 & 15	sorghum	Garst brand 3824 RR
2006	Rossville	8 June	7.5 & 15	corn	Garst brand 3824 RR

At all locations and both years, final plant population was determined approximately 30 days after planting. To determine plant population, plants were counted on both sides of a 10 ft pipe that was randomly placed between two non-border rows. Pods per plant, with respect to origin on plant (branches or mainstem), plant height, and height of the lowest pod were determined at approximately growth stage R7 by randomly selecting two plants from each plot for measurement. The yield components seeds per pod and pods per acre were calculated using the plant populations and the pods per plant, and 200 seeds were counted from every plot to determine seed weight. All 2005 locations were combined for analysis using GLM (SAS v 9.1, SAS Institute Inc., Cary, NC) after variance testing indicated the variances were equal from each location. In 2006, the Manhattan, Ottawa, and Rossville locations were combined for analysis using GLM (SAS v 9.1).

Yield Responses to Narrow Rows

In this study, the soybean planting dates and seeding rates used included the normal ranges used in Kansas. Across both years, yields ranged from 24 to 64 bu/acre which was attributed to the wide variation in temperature, soil quality, and rainfall throughout the two years of the study.

In 2005, adequate rainfall was received for much of the growing season, which contributed to the moderately high yields found at all three locations (Fig. 1). The adequate rainfall across the region resulted in yield variances that were equal for each of the three locations, so the data were combined for analysis and only plant population affected yields (Table 3). Across all three locations, yields increased as plant population increased until reaching a plateau at 80,000 plants/acre (Fig. 2A).

Fig. 1. Cumulative normal precipitation and precipitation measured in 2005 and 2006 for Manhattan, Ottawa, and Rossville, KS.

Fig. 2. Soybean yield response to plant density averaged across Manhattan, Ottawa, and Rossville, KS in 2005 (A) and for individual locations in 2006 (B).

Table 3. Significance levels for soybean plots in 2005 and 2006 at Manhattan, Ottawa, Rossville, and Parsons KS.

	Seed yield	Plants /acre	Pods/ plant	Pods /acre	Branch pod %	Plant height	First pod height	Seeds /pod	Seed weight
	2005
Rep	*	NS	*	NS	NS	*	*	NS	NS
location	NS	*	*	*	NS	NS	NS	NS	*
population	*	*	*	*	*	*	*	NS	NS
loc × pop	NS	NS	NS	NS	NS	NS	NS	NS	NS
	2006
Rep	*	*	NS	NS	NS	*	*	NS	*
location	*	*	*	*	NS	*	*	NS	*
population	*	*	*	*	*	*	*	*	NS
row spacing	NS	NS	NS	NS	NS	*	NS	NS	NS
rs*pop	NS	NS	NS	NS	NS	NS	NS	NS	NS
loc*pop	*	*	NS	NS	NS	*	NS	NS	NS
Loc*rs	*	*	NS	NS	NS	*	NS	NS	NS
locrspop	NS	NS	NS	NS	NS	NS	NS	NS	NS

* Significant at α = 0.05 level.

In 2006, rainfall was not as abundant in the region, and yields were lower than 2005. Locations, location × population and location × row spacing affected yields (Table 3). The 7.5-inch row spacing produced higher yields at all locations in 2006, and Manhattan and Ottawa both yielded higher than Rossville in 2006 (Table 4). As plant population increased, yield increased in a quadratic manner at Manhattan and Ottawa with optimum yields occurring at plant populations 136,000 and 140,000 plants/acre, respectively (Fig. 2B). A severe hailstorm at Rossville in 2006 at approximately the V2 growth stage decreased the final plant population and delayed the crop (Table 4). The lower plant densities at Rossville contributed to the significant location × row spacing interaction in 2006 (Table 3). The lower plant densities and delayed development at Rossville in 2006 resulted in a similar yield response to increasing plant populations as the other locations. Based on the regression analysis and extrapolation beyond the data, the optimum plant population at Rossville in 2006 was approximately 150,000 plants/acre. This is not an unreasonable conclusion, as the physical hail damage and delayed development might require more plants to develop adequate pods and grain to reach the yield potential compared with the other two locations where hail was not a factor.

Optimum plant densities of 80,000 and 140,000 plants/acre (2005 and 2006, respectively) are below the recommended seeding rates for Kansas (17) and Nebraska (15) who recommend 180,000 and 150,000, respectively and in agreement with seeding rates for 30-inch rows in Missouri (18). However, our results are reported in plants per acre and seeding recommendations are made in seed per acre meaning that emergence rates must be considered. This creates the greatest challenge associated with the seeding rate decision as seeding equipment (drill vs. planter), weather (rain at planting) and seed quality can play an important role in emergence rates.

Pods Per Plant and Pods Per Acre

Yield components responded similarly across locations and years. In 2005 and 2006 plant population affected pods per plant (Table 3). Pods per plant were different at each location in both 2005 and 2006 (Table 5). Throughout the study, as population increased, pods per plant decreased in a polynomial manner, reaching minimums from 150,000 to 190,000 plants/acre (Fig. 3A). The negative response in pods per plant to increasing plant population agree with the findings of several other researchers (3,7,20,22,28).

Fig. 3. Soybean pods per plant as affected by plant population at Manhattan, Ottawa, and Rossville, KS in 2005 and 2006 (A); Soybean pods per acre averaged across locations in 2005 and 2006 (B); Branch contribution to the total pods per plant across locations in 2005 and 2006 (C).

In 2005 and 2006, pods per acre were affected by location and plant population (Table 3). Pods per acre were different across locations both years (Table 5). It appears that the lower plant densities and damage to the plants from the hail at Rossville in 2006 had less impact on the plants’ ability to compensate than might be expected. Pods per plant at Rossville were equal to those at Ottawa and greater than at Manhattan. Pods per acre at Rossville were lower than at the other two locations. As population increased, pods per acre increased in a linear manner (Fig. 3B). Pods per plant was not correlated to yield (r = 0.02), but pods per acre was correlated to yield (r = 0.72). These results agree with the findings of Parvez et al. (23), but conflict with those of Ethredge et al. (16).

Branch Contribution to Total Pods Per Plant

In 2005 and 2006, pods per plant on branches were affected by plant density (Table 3). As population increased in both years, the percent of branch pods on each plant decreased (Fig. 3C). It was apparent that increased competition within the rows caused the plants to branch less, and therefore receive less of the total yield from branches. Branch pod contribution had a low correlation to yield (r = 0.25) but had a larger influence on pods per acre (r = 0.67). These findings agree with work done by Probst (24) and Carpenter and Board (10), who showed the compensatory ability of soybeans as they react to increases and decreases in available space.

Plant Height

In 2005, plant height was affected by plant density at all locations (Table 3). As plant density increased, plant height increased in a quadratic manner, reaching a maximum at plant population of approximately 195,000 plant/acre (Fig. 4). In 2006, plant height was affected by location, plant density, and row spacing as indicated by significant location × population and location × row spacing interactions for the three combined locations (Table 3). Plant heights were not different when comparing the two row spacings at Manhattan and Rossville in 2006 (Table 6). At Ottawa in 2006, plants in the 7.5-inch rows were shorter than plants in the 15-inch rows.

Fig. 4. Soybean plant height response to plant density averaged across Manhattan, Ottawa, and Rossville, KS in 2005 (A) and for individual locations in 2006 (B).

Plant height was not affected by plant population to the same degree at Manhattan and Ottawa in 2006 as in 2005. In 2005, plant heights across all locations had a range of 4 inches, whereas at these two locations in 2006, the range was less than 2 inches. In 2006, only at Rossville did plant height increase as rapidly as plant density increased. Plant height in narrow row soybean is often a concern to producers, as taller plants are believed to have greater risk of lodging.

Height of First Pod

Of greater interest to producers than plant height is the location of the first pod from the soil surface. Across all locations each year, height of first pod was affected by plant population (Table 3). The location main effect was also significant in 2006. In 2005, height to the first pod increased linearly and reached a plateau at approximately 110,000 plants/acre (Fig. 5). In 2006, average height of the lowest pod increased as plant population increased and reached a maximum at approximately 140,000 plants/acre (Fig. 5). These observations agree with the findings of Lueschen and Hicks (21), who reported similar responses of first pod height to plant population.

Fig. 5. Height of first pod averaged across Manhattan, Ottawa, and Rossville, KS in 2005 and 2006.

Seeds Per Pod and Seed Weight

Population affected seeds per pod in 2006 (Table 3) but seeds per pod were not affected by any treatments in 2005. As population increased, seeds per pod decreased in a linear manner (seeds/pod = 2.71 – 3.21 × 10^-6 * plant population) with the data ranging from 2.3 to 1.5 seeds/pod (data not shown). In both 2005 and 2006, growing conditions at each location affected seed weights differently (Table 3). In 2005, seed weights were different at each location (Table 5). In 2006, the delayed development because of the hail damage at Rossville reduced the time available for grain fill and seed weights at Rossville were lower than those measured at Manhattan and Ottawa. This illustrates that the yield component most affected by the hail damage was seed weight compared with pod per plant and pods per acre. As discussed earlier, pod per acre were approximately 10% lower at Rossville than Manhattan, whereas seed weights are approximately 20% lower at Rossville compared with the other two locations in 2006.

Seed weight influenced grain yield (r = 0.79), but seeds per pod did not influence yield (r = -0.06). Seed weight and seeds per pod yield responses were mainly affected by rainfall and high temperatures late in the growing season after growth stage R5. These findings agree with those of Board et al. (6) and Lehman and Lambert (20), who both reported that seeds per pod and seed weight typically do not display as large of a response to plant density as do the yield components pods per plant and pods per acre.

Summary

Row spacing had little effect on soybean yield and yield components, but population and environment played major roles in dictating yield responses. Yields were maintained across changing plant densities due to the compensatory abilities of soybean to fluctuate yield components in response to intra- and inter-row competition. The 2005 plots failed to increase in yield above 80,000 plants/acre and the 2006 plots had optimum plant populations below the currently recommended seeding rates for Kansas and Nebraska and were in agreement with those from Missouri. Therefore, it is likely that the recommended soybean seeding rates for dryland environments can be safely lowered by at least 20 to 25%.

Literature Cited

1. Adams, P. D., and Weaver, D. B. 1998. Brachytic stem trait, row spacing, and plant population effects on soybean yield. Crop Sci. 38:750-755.

2. Ablett, G. R., Beversdorf, W. D., and Dirks, V. A. 1991. Row width and seeding rate performance of indeterminate, semideterminate, and determinate soybean. J. Prod. Agric. 4:391-395.

3. Ball, R. A., Purcell, L. C., and Vories, E. D. 2000. Optimizing soybean plant population for a short-season production system in the Southern U.S. Crop Sci. 40:757-764.

4. Beuerlein, J. E. 1988. Yield of indeterminate and determinate semidwarf soybean for several planting dates, row spacings, and seeding rates. J. Prod. Agric. 1:300-303.

5. Board, J. E., and Harville, B. G. 1992. Explanations for greater light interception in narrow rows vs. wide-row soybean. Crop Sci. 32:198-202.

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

7. Boquet, D. J. 1990. Plant population density and row spacing effects on soybean at post-optimal planting dates. Agron. J. 82:59-64.

8. Buhler D. D., Philbrook, B. D., Oplinger, E. S. 1990. Velvetleaf and giant foxtail control for solid-seeded soybean production in three tillage intensities. J. Prod. Agric. 3:302-308.

9. Bullock, D., Khan, S., and Rayburn, A. 1998. Soybean yield response to narrow rows is largely due to enhanced early growth. Crop. Sci. 38:1011-1016.

10. Carpenter, A. C., and Board, J. E. 1997. Growth dynamic factors controlling soybean yield stability across plant populations. Crop Sci. 37:1520-1526.

11. Costa, J. A., Oplinger, E. S., and Pendleton, J. W. 1980. Response of soybean cultivars to planting patterns. Agron. J. 72:153-156.

12. Devlin, D. L., Fjell, D. L., Shroyer, J. P., Gordon, W. B., Marsh, B. H., Maddux, L. D., Martin, V. L., and Duncan, S. R. 1995. Row spacing and seeding rates for soybean in low and high yielding environments. J. Prod. Agric. 8:215-222.

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

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

15. Elmore, R. W., and Specht, J. E. 2000. Soybean seeding rates. Coop. Ext. Pub no. G1395. Univ. of Nebraska, Lincoln, NE.

16. Ethredge, W. J. Jr., Ashley, D. A., and Woodruff, J. M. 1989. Row spacing and plant population effects on yield components of soybean. Agron. J. 81:947-951.

17. Fjell, D. L. 1997. Soybean production handbook. Kansas St. Res. and Ext. Pub. No. C-449. Manhattan, KS.

18. Hensel, Z. R., and Minor, H. C. 1993. Soybean production in Missouri. Coop. Ext. Pub. no. G4410. Univ. of Missouri, Columbia, MO.

19. Kratochvil, R. J., Pearce, J. T., and Harrison, M. R. Jr. 2004. Row spacing and seeding rate effects on glyphosate-resistant soybean for Mid-Atlantic production systems. Agron. J. 96:1029-1038.

20. Lehman, W. F., and Lambert, J. W. 1960. Effects of spacing of soybean plants between and within rows on yield and its components. Agron. J. 52:84-86.

21. Lueschen, W. E., and Hicks, D. R. 1977. Influence of plant population on field performance of three soybean cultivars. Agron. J. 69:390-393.

22. Norsworthy, J. K., and Shipe, E. R. 2005. Effect of row spacing and soybean genotype on mainstem and branch yield. Agron. J. 97:919-923.

23. Parvez, A. Q., Gardner, F. P., and Boote, K. J. 1989. Determinate- and indeterminate-type soybean cultivar responses to pattern, density, and planting date. Crop Sci. 29:150-157.

24. Probst, A. H. 1945. Influence of spacing on yield and other characters in soybeans. J. Amer. Soc. Agron. 37:549-554.

25. Riess, W. D., and Sherwood, L. V. 1965. Effects of row-spacing, seeding rate, and potassium and calcium hydroxide additions on soybean yields on soils in southern Illinois. Agron. J. 57:431-433.

26. Safo-Kantanka, O., and Lawson, N. C. 1980. The effect of different row spacings and plant arrangements on soybeans. Can. J. Plant Sci. 60:227-231.

27. Staggenborg, S. A., Devlin, D. L., Fjell, D. L., Shroyer, J. P., Gordon, W. B., Marsh, B. H., and Maddux, L. D. 1996. Soybean response to row spacing and seeding rates in Northeast Kansas. Pub. No. SRL 112. Kansas Agric. Exp. Stn., Manhattan, KS

28. Weber, C. R., Shibles, R. M., and Byth, D. E. 1966. Effects of plant population and row spacing on soybean development and production. Agron. J. 58:99-102.