| dc.description.abstract | Declining soil fertility in general and nitrogen (N) in particular limit crop production on many smallholder farms in western Kenya. Soil N deficiency is usually ameliorated with inorganic N fertilizers but their high costs preclude their use by resource poor smallholder farmers. In addition there are concerns that increased use of inorganic fertilizers could lead to environmental degradation. Organic inputs offer sustainable alternatives to expensive inorganic N fertilizers but the common organic materials on smallholder farms are not available in adequate amounts. The use of none traditional organic materials such as water hyacinth has been reported to supply plant nutrients but has not been adequately tested. Integration of legumes such as the common bean in cropping systems is another option that has the potential to contribute soil N through symbiotic nitrogen fixation. However there is paucity of information on the abundance, symbiotic effectiveness and identity of native bacterial strains from soils in western Kenya thus making it difficult to make informed decisions on the inoculation need. This study was conducted to compare the response of common bean to water hyacinth compost, inorganic N fertilizer and determine the abundance, symbiotic efficiency (SE) and identity of the native bacteria nodulating legumes. Soil sampling was carried out in three farms that had been used for growing common bean for at least two seasons and one fallow land with no known history of growing common bean or inoculation in Kisumu and Kakamega. The abundance of soil bacteria and symbiotic efficiency (SE) were determined in a greenhouse experiment. Deoxyribonucleic acid (DNA) was extracted from the bacteria and the 16S rRNA gene sequenced to establish their identity. Field experiments were laid out in a randomized complete block design (RCBD) replicated four times at Masinde Muliro University of Science and Technology and Kisumu for two consecutive seasons during the short rains of October to December, 2013 and the long rains of April to August, 2014. The treatments consisted of two bean varieties, rhizobium inoculation and non inoculation, with and without phosphorus (P), urea and water hyacinth compost. The population of bacteria ranged from 3.2 × 101 to 3.5 × 104 cells per gram of soil. Cluster analysis of the 16S rRNA gene grouped the native bacterial strains into five genera: Rhizobium, Bacillus, Pantoea, Enterobacter and Klebsiella. The bacterial strains had high SE compared to the first commercial the inoculant, CIAT 899 and ranged from Rhizobium sp (74%) to Klebsiella variicola (170%). The two species of Klebsiella variicola (MUST005 and KSM005) and Enterobacter hormaechei recorded higher SE than strain 446, a second inoculant. These native species should be exploited for the development of inoculation programmes to enhance N fixation in the region. Low levels of P were recorded in fallow soils at the two sites while Aluminium levels were high in Kakamega. Urea significantly improved yields of common bean in the short rain season compared to the other treatments at the two sites. However, in the long rains season water hyacinth compost gave higher yields than urea and control. This study reveals that the use of water hyacinth compost can serve as a substitute for inorganic N source. | en_US |
