Grain amaranths (Amaranthus spp.) are high protein content and quality pseudo-cereal crops originally domesticated in Latin America whose favorable nutritional profile belies their potential to alleviate nutrition and food insecurity in developing countries. Grain amaranth was introduced as a nutrient dense food into the Kamuli District, eastern Uganda, in 2006 through Iowa State University's Center for Sustainable Rural Livelihoods (CSRL) in conjunction with Ugandan partners Volunteer Efforts for a Development Concerns (VEDCO), a local non-governmental organization (NGO), and Makerere University. Initial analysis of protein content of amaranth grain pooled from farms in the Kamuli District indicated protein levels from 11.7% to 12.5%, lower than the average value of 15% found in the literature.

A study was therefore designed to determine: 1) common production practices for grain amaranth in Uganda; 2) variability of amaranth grain yields and protein content between farms; 3) variability of amaranth grain yields and protein content between varieties; and 4) the effects of soil physical and chemical properties, and the use of organic soil amendments, on amaranth grain yields and protein content.

A non-random snowball survey of amaranth-growing farms in the Kamuli District was conducted in June 2009 to determine common agronomic practices and use of soil amendments. Soil samples to a depth of 15 cm were also taken to determine soil physical and chemical properties of amaranth fields in the area. Survey results indicated that grain amaranth is grown over variable demographic and geographic ranges within the Kamuli district. Farmers reported growing amaranth on small plots, which had previously supported a wide range of crops. Amaranth was most frequently grown in pure stands, with "golden" as the most common variety. Pest pressure was generally moderate to low in severity. Most respondents did not use organic soil amendments to improve crop yields, but the majority of those who applied amendments used cow manure. Soil amendments were variable in residue age and management practices.

On-station grain amaranth trials were conducted in the Wakiso District, 18 km north of Kampala in south-central Uganda, to test effects of poultry manure and composted manure applied at 0, 1, 1.5, and 3 ton ha-1 on amaranth yield and protein content. Three identical experiments were conducted in this regard (Experiments 1-3), with two reported here, due to accidental destruction of the amaranth crop in Experiment 1.

Experiment 2 was conducted during the dry season beginning on 17 July 2009 and harvested on 9 October 2009 (91 DAP), with data collected on amendment nutrient composition, seedling establishment, plant height, leaf number, insect and weed populations, grain yield and protein content. Mean grain yields for Experiment 2 were 1110 kg ha-1 (SD: 66 kg ha-1) with no significant differences between treatments, but yields for "cream" (1013 kg ha-1) and "golden" (1208 kg ha-1) were significantly greater than those for `Plainsman' (191 kg ha-1) (P=0.0083). Amaranth grain was pooled across all treatments and varieties in Experiment 2 to determine a mean protein content of 14.5% (SD: 0.2%).

Experiment 3 was conducted during the rainy season and planted on 3 October 2009, with harvest occurring on three different dates in late December 2009 and early January 2010 (80 - 98 DAP). Data was taken for all parameters listed above for Experiment 2. Mean amaranth grain yields for Experiment 3 were 1886 kg ha-1 (SD: 921.1 kg ha-1), with no significant differences between treatments, but yields for "cream" (2750 kg ha-1) were significantly greater than those for "golden" (1742 kg ha-1) and `Plainsman' (1166 kg ha-1) (P=0.0043). Grain yields were also significantly greater for Column 3 (2724 kg ha-1) than for Column 1 (1709 kg ha-1). Pooled amaranth grain across all treatments and varieties in Experiment 3 contained a mean protein content of 15.1% (SD: 0.2%)

On-farm experiments were conducted during the short rainy season of 2009 in the Kamuli District to determine the effects of cattle manure applied at 27 ton ha-1 on "cream" and "golden" amaranth varieties compared to a control treatment (0 ton ha-1). Ten farms participated in the experiment; data on soil physical and chemical properties, planting and harvest date, plant height, plant density, plant health and disease parameters, qualitative weather assessments, and grain yield were collected. Yields were significantly different between farms, with a maximum reported yield of 1230 kg ha-1 and a minimum of 239

kg ha-1. Application of cattle manure provided a significant yield increase, where treatment averages were 667 kg ha-1 where soil amendments were utilized, compared to 420 kg ha-1 in control plots, representing a yield increase of 58.6% with the addition of cattle manure.

There were no significant differences between varieties, with a mean of 619 kg ha-1 for "cream" and 468 kg ha-1 for "golden" varieties (SD: 487 kg ha-1 and 261 kg ha-1, respectively). Amaranth grain protein content was significantly different between farms, ranging from 12.1% to 15.1%. Average protein content of 14% across treatments and varieties revealed a greater amaranth protein content than previously reported for the Kamuli District. Linkages with organic associations, such as NOGAMU (National Organic Agriculture Movement of Uganda, Kampala, Uganda), can help facilitate marketing strategies for amaranth farmers whose production exceeds family needs and who desire to enter commercial markets.