Index | Search | Home | Table of Contents
Apaza-Gutierrez, V., A. Romero-Saravia, F.R. Guillén-Portal, and D.D. Baltensperger. 2002. Response of grain amaranth production to density and fertilization in Tarija, Bolivia. p. 107109. In: J. Janick and A. Whipkey (eds.), Trends in new crops and new uses. ASHS Press, Alexandria, VA.
V. Apaza-Gutierrez, A. Romero-Saravia, F.R. Guillén-Portal, and D.D. Baltensperger
Grain amaranth has shown potential for the conditions of the Central Valley of Tarija, Bolivia (Guillén-Portal 1992a). In this region, Amaranthus caudatus L., Amaranthaceae, locally known as coime, has been traditionally produced in temperate mountain areas through an amaranth-maize (Zea mays L.) intercropping system, which is adequate for subsistence agriculture (Guillén-Portal 1992b). Recent advancements on the potential of grain amaranth as a cash crop or as raw matter for industrial purposes has led to consider the expansion of this crop throughout the area. However, for large-scale amaranth production, which would make the crop profitable, development of solo-crop production technology is required. When considering solo-crop production of grain amaranth, the type of fertilizers for local soil conditions and plant density to which the crop must be accommodated in the field are important factors to consider. Studies conducted in Peru on A. caudatus indicated that optimum grain yields can be obtained at plant densities of about 450,000 plants ha-1 and fertilization levels of 100 N138 P180 K (Sumar-Kalinowsky et al. 1992). The objective of this study was to determine, at a preliminary level, the optimum fertilization formula and plant density for solo-crop grain amaranth production in Tarija, Bolivia.
Two separate studies were conducted in representative areas of the Central Valley of Tarija (Apaza-Gutierrez 1994; Romero-Saravia 1999). In one study, carried out in the village of Yesera Norte (21° 10' S, 64° 26' W, 2,200 m) during the summer of 19911992, two grain amaranth genotypes, 87f-51 (Amaranthus hypochondriacus, introduced) and 01-0023-1 (A. caudatus, local) were evaluated for their response to eight levels of fertilizer: control, chemical (404020 and 808040 NPK), organic (7.5 tonnes ha-1, 15 t ha-1 dried ovine manure), and mixed (202010 + 3.75 t ha-1, 404020 + 7.5 t ha-1; and 606030 + 11.25 t ha-1). The trial was planted on Dec. 10, 1991, on a recent-alluvial-terrace soil, slightly alkaline. One hundred percent of the organic fertilizer and 100% of P and K was applied at the planting time on each plot at the corresponding levels, whereas 50% of N was applied at planting time and the rest at the cultivation time. Plots were thinned by hand to an equivalent plant density of 60,000 plants ha-1 and hand weeded. Irrigation was applied only occasionally to prevent extreme drought.
In another study, carried out in the village of Sella Cercado (21°23' S, 64°42' W, 2,080 m) at the same time as the above study, four grain amaranth genotypes, 87f-51 (A. hypochondriacus, introduced), 886s-445 (A. cruentus, introduced), 1023 (Amaranthus spp., introduced), and 01-0014-0 (A. caudatus, local) were evaluated for their response to four plant densities: 55,000, 110,000, 166,000, and 222,000 plants ha-1. This trial was planted on Dec. 15, 1991 on a fluvio-lacustre terrace soil, 30 cm depth and slightly alkaline. Plants were thinned by hand to the corresponding plant density levels in each plot. A fertilization formula of 606060 NPK was applied to each plot, with 50% N, 100% P and 100% K applied at planting time and the remaining N at cultivation time. Other cultural practices were the same as the experiment above. In both studies, a factorial arrangement of treatments (genotype × fertilizer, genotype × plant density) using a randomized block design with four replications was used. The experimental unit consisted of four rows 5 m long and 0.7 m apart. Response variables were plant height (m), stem diameter (cm), inflorescence length (cm), grain yield (kg ha-1), for both studies, and additionally grain yield plant-1 (g plant-1) and 1000 seed weight (1) for the plant-density study.
Soils at the site of the trials were characterized as having moderate levels of organic matter and low levels of NPK (data not shown). Considering the magnitude of the studies, the analyses of the results will focus mainly on grain yield. Across fertilizers, the introduced cultivar yielded 1.59 t ha-1, 17% higher than the local type (Table 1). However, the latter was more responsive to the levels of fertilizer applied (1.5, 3, and 1.7 times more responsive under chemical, organic, and mixed fertilizers, respectively) (Fig. 13). Grain yield showed a linear response to chemical and organic fertilizers, and a quadratic response to the mixed fertilizer. Highest response corresponded to chemical and mixed fertilizers (both showing a yield of 1.65 t ha-1, 80% and 29% higher than control and organic, respectively) (Fig. 13). The fertilization levels delayed the phenology of the crop, although not uniformly between cultivars (data not shown).
Table 1. Comparison of two grain amaranth cultivars across fertilization levels.
|01-0023-1 (A. caudatus)||102.5a z||1.9a||47.5||1.36a|
|87f-51 (A. hypochondriacus)||63.6b||2.2b||45.6||1.59b|
zValues in rows followed by the same letter are not statistically significant (p < 0.05).
|Fig. 1. Regression based on mean values of chemical fertilization on grain yield across two grain amaranth cultivars, Tarija, Bolivia, 1991.||Fig. 2. Regression based on mean values of organic fertilization on grain yield across two grain amaranth cultivars, Tarija, Bolivia, 1991.|
Fig. 3. Regression based on mean values of mixed (chemical + organic) fertilization on grain yield across two grain amaranth cultivars, Tarija, Bolivia, 1991.
Cultivars responded similarly in grain yield to variations in plant density (avg. 2.14 t ha-1) (Table 2. Fig. 4). Grain yield increased linearly within the range of densities (approx. 2 kg of grain per 1,000 plants). At the highest density (220,000 plants ha-1), grain yield reached a yield of 2.30 t ha-1. Based on this, the materials studied can still respond positively to increased plant density. An analysis of the effect of density on grain yield per plant and other characteristics pointed to stem diameter as the character with the highest effect on grain yield per plant. Grain yield per plant and stem diameter decreased quadratically with increased plant density (Fig. 5). It is likely that this decrease is the result of interplant competition. Thus, grain yield per unit area might be directly related to the ability of the plant to store nutrients on the stem. This is an aspect that should be studied in more detail.
Table 2. Comparison of four grain amaranth cultivars across plant densities, Tarija, Bolivia, 1991.
|01-0014-0 (A. caudatus)||1.6a z||1.5a||46.1a||0.079b||25.3||2.11|
|886s-445 (A. cruentus)||1.1bc||1.3b||39.9c||0.079b||23.5||2.15|
|87f-51 (A. hypochondriacus)||1.1c||1.3b||42.8b||0.087a||25.0||2.15|
|1023 (A. spp.)||1.2b||1.4b||47.4a||0.080b||24.3||2.15|
zValues in columns followed by the same letter are not statistically significant (p < 0.05).
|Fig. 4. Regression based on mean values of plant density
on grain yield and grain yield/plant across four grain amaranth cultivars.
Tarija, Bolivia, 1991.
||Fig. 5. Regression based on mean values of plant density on grain yield/plant and on stem diameter across four grain amaranth cultivars. Tarija, Bolivia, 1991.|
We conclude that solo-crop amaranth production in the Central Valley of Tarija should be carried out under a fertilizer formula of at least 808040 NPK or 606030 NPK + 11 t ha-1 dried ovine manure, and a plant density of at least 220,000 plants ha-1. Further studies throughout the region and replicated in years including the economic feasibility of the application of chemical and organic fertilizers are required before making recommendations for the farmer.