EFFECTS OF FERTILIZER APPLICATION AND PINCHING ON THE YIELD OF GRAIN AMARANTH (Amaranthus hypochondriacus)

Inadequate food and imbalanced nutrients are major causes of human malnutrition \nfor the poor, sick, children, elderly and People Living with HIV/AIDS (PLWAS) \nwho are the most vulnerable. These groups have high protein requirements but are \nunable to access adequate amounts from animal sources due to high cost. \nTherefore, there is need to explore cheap plant foods with comparable protein \nquality and quantity as animal protein foods such as grain amaranth (Amaranthus \nhypochondriacus). The current level of production is quite low (0.25 – 1 t ha-1) as \ncompared to the potential of 3 t ha-1. This study investigated the effects of \ncombined organic and inorganic fertilizer, fertilizer pelleting and pinching on the \ngrowth and yield of grain amaranth. The study was carried out at the Maseno \nUniversity demonstration farm during the short rains in 2010 and the long rains in \n2011 in two experiments which were laid out as Randomized complete block \ndesign (RCBD) with split plot arrangement and replicated three times. For \nfertilizer and pinching experiment study, the main plots were pinching at different \ntimes and the subplots were the different inorganic and organic fertilizer \ncombinations, while in pelleting study, the main plots were pelleting and the \nsubplots were three different levels of manure and inorganic fertilizer \ncombinations. The data collected was on days to 50 % germination, days to 50 % \nflowering, days to 50 % maturity, average plant height, stem width, number of \nleaves, height of flower head, canopy, plants dry matter weight, grain yield and \n1000 seed weight. The data were subjected to Analysis of Variance and \nsignificant means separated by the least significant difference at P<5%. A \nregression of yield and growth parameters was done to determine the factor with \nthe greatest influence on yield. To assess the effect of organic and inorganic \nfertilizer combination and fertilizer pelleting on the soil and plant tissues nutrient \nstatus during growth and development of amaranth, soil and leaf samples were \ntaken at seedling, flowering and at harvesting growth stages and analyzed for \nnitrogen (N), phosphorous (P) and potassium (K) .The soil and leaf nutrient data \nwas subjected to regression analysis. Regression analysis was used to observe the \ninfluence of the nutrient levels in the soil and leaves on the crop yields. Pinching \nat 28 days after planting was the best practice for increased stem width, number of \nflowering stems, canopy, number of leaves and yield. Fertilizer combination of \n75 % organic N and 25 % inorganic N had the highest grain yield of 1.185 t \nha-1while fertilizer combination of 25 % organic and 75 % inorganic recorded the \nlowest grain yield of 0.665 t ha-1both in 2010 and 2011. Stepwise multiple \nregressions of the factors with yield showed plant height to have the highest \ninfluence on yields. All the pellet fertilizer treatments had better growth \nparameters, higher dry matter weight, 1000 seed weight and grain yield than the \ncontrol. In 2011, pellet fertilizer treatment with 75 % organic N and 25 % \ninorganic N had mean grain yield of 743 kg ha-1, while non- pellet fertilizer of the \nsame treatment had mean grain yield of 533 kg ha-1. In the same season pellet \nfertilizer treatment with 25 % organic N and 75 % inorganic N had mean grain \nxii \nyield of 413 kg ha-1 while the non-pellet fertilizer of the same treatment had mean \ngrain yield of 231 kg ha-1 (Least Significant Difference = 141.3). The regression \nmodels were statistically significant at P<5% in estimating the linear dependency \nof; % N in leaves on % N in the soil at both seedling and flowering stages, yield \non the level of potassium in the soil and in the leaves at both seedling and \nflowering stages. The regression models using fertilizer pellets showed positive \ncorrelation between % N levels in the leaves and % N levels in the soil at \nflowering stage representing a normal curve. The regression models positively \nrelated P levels in the soil and yield at seedling stage which is normally the case. \nThe study recommends the use of a mixture of 6.75 kg ha-1 organic manure and \n22 kg ha-1 inorganic N to ensure prolonged source of nutrients to the plant and to \nenhance yield. Coating of fertilizers containing high leachable nutrients or \npelleting with manure is also recommended