Index | Search | Home | Table of Contents

Henderson, T.L., A.A. Schneiter, and N. Riveland. 1993. Row spacing and population effects on yield of grain amaranth in North Dakota. p. 219-221. In: J. Janick and J.E. Simon (eds.), New crops. Wiley, New York.

Row Spacing and Population Effects on Yield of Grain Amaranth in North Dakota

T.L. Henderson, A.A. Schneiter, and N. Riveland

  5. Table 1
  6. Fig. 1
  7. Fig. 2

Grain amaranth (Amaranthus spp.), a high protein pseudo-cereal which originated in Central and South America, was a staple crop of ancient Aztec and Inca civilizations. Since the mid-1970s, amaranth has received attention as a new crop for North America. Trials conducted in North Dakota since 1981 indicate that certain cultivars of grain amaranth are well adapted in the eastern part of this state, producing more than 2000 kg/ha (Henderson et al. 1991).

Results of research to determine optimal row spacing have been inconclusive (Robinson 1986). Grain yield response to plant density has been shown to be influenced by environments, species, and cultivars (Putnam 1990). Robinson (1986) reported a decline in yield at populations greater than 210,000 plants/ha. Yet, Haas (1983) identified a much higher range of optimal plant density, between 323,000 and 360,000 plants/ha. A positive yield response to increasing plant density was observed for lower populations (20,000 to 60,000 plants/ha) (Edwards and Volak 1980). At excessively high populations, competition for moisture and nutrients reduces grain yield (Weber 1990). The relationship between plant density and grain yield may be influenced by cultivar (Edwards and Volak 1980; Haas 1983).

This study was conducted as part of research at North Dakota State University to establish production guidelines for grain amaranth in North Dakota. Our objective was to determine the most suitable row spacing and plant population for grain amaranth production in North Dakota.


Four grain amaranth cultivars, 'K283' and 'MT-3' (both A. cruentus), and 'K343' and 'K432' (both A. hypochondriacus x A. hybridus), were evaluated at Prosper and Williston, North Dakota, in 1989 and 1990. Populations of 74, 173, and 272 thousand plants/ha were established for each cultivar at narrow (30.5 cm) and wide (76.2 cm) row spacing. Stands were oversown and thinned by hand to establish the desired populations. Experimental design was a randomized complete block with a split-split plot arrangement. Row spacing, established plant population, and cultivars constituted the main, sub, and sub-sub plots, respectively. All plots were hand-harvested. Grain yield was determined for each plot. Precipitation was recorded daily at each environment.


Main effects of row spacing, established population, and cultivar on grain yield for individual environments are presented in Table 1. In the combined analysis, a significant environment x cultivar interaction, which can be attributed to precipitation differences, was obtained for grain yield. All cultivars exhibited a decrease in grain yield at the driest environment, Williston 1990, which received only 9.4 cm of precipitation during the growing season, compared to 21.7 cm at Prosper 1989 and 20.0 cm at Prosper 1990. In the combined analysis, highest yields were produced by 'MT-3' and 'K283'.

In the combined analysis, established population had a significant effect on grain yield, with highest yields achieved at the lowest population. The row spacing x established population interaction was significant for grain yield as a yield advantage for wider rows at the higher populations (Fig. 1). This response occurred presumably due to increased competition within the rows which decreased plant population after establishment, resulting in more water and nutrients for surviving individual plants in the wider rows.

A similar row spacing x population interaction effect was observed at Prosper during 1990. To examine the influence of row width and established population on plant competition, final plant population at harvest was determined for each plot at the Prosper 1990 environment. A significant row spacing x established population interaction was observed for final population. Intrarow competition was greater at the 76.2 cm row width, resulting in a substantial loss of plant population by harvest time, especially at the higher established populations. Significant losses in plant population were not observed with the narrower row spacing. Grain yield increased as final population decreased (Fig. 2). Results of the combined analysis suggest that at the lowest established population, row width was not critical. At the higher established populations, a yield advantage was obtained with wider row spacing, due to lowered plant density resulting from competition within the rows.


Highest amaranth grain yields were obtained at the Prosper location in eastern North Dakota, which received more precipitation during the growing season. 'MT-3' and 'K283' (A. cruentus) were the highest yielding cultivars in the combined experiment. The lowest established population, 74,000 plants/ha, consistently produced the highest grain yield. Row spacing had no effect at the lowest population, while at the higher populations, more grain was produced with the wider (76.2 cm) row spacing. With the wider rows, plants within each row were spaced closer together, leading to increased competition at high established populations. This intrarow competition caused substantial losses in plant numbers after establishment. Grain yield of the surviving plants was higher as a result of the lowered plant population in the wider rows.


Table 1. Mean amaranth grain yield values by environment for main effects of row spacing, established population, and cultivar.

Mean grain yield (kg/ha)
Environment Prosper 1989 Prosper 1990 Williston 1990
Row width (cm) * NS **
30.5 867 1414 344
76.2 1228 1452 477
LSD 0.05 217 --- 31
Established population
NS * **
74000 1114 1513 479
173000 993 1450 410
272000 1035 1337 344
LSD 0.05 --- 132 32
Cultivar ** ** **
K283 1193 1467 571
K343 989 1566 291
K432 661 976 294
MT-3 1348 1724 487
LSD 0.05 180 131 47
NSIndicates not significant at the 0.05 level.
*Indicates significant at the 0.05 level.
**Indicates significant at the 0.01 level.

Fig. 1. Mean amaranth grain yield values for established population x row spacing interaction, based on combined analysis of three environments in North Dakota. (Least significant difference 0.05 = 93.6 kg/ha.)

Fig. 2. Mean amaranth grain yield values as a function of row spacing and final plant population at Prosper, ND in 1990. (Least significant difference 0.05 = 185.9 kg/ha and 7781 plants/ha for grain yield and final population, respectively.)
Last update September 10, 1997 aw