Maize, the main dietary staple in Kenya, is one of the crops most susceptible to contamination by aflatoxin. To understand sources of aflatoxin contamination for home grown maize, we collected 789 maize samples from smallholder farmers’ fields in Eastern and South Western, two regions in Kenya representing high and low aflatoxin risk areas, respectively, and determined aflatoxin B1 (AFB1) using ELISA with specific polyclonal antibodies. AFB1 was detected in 274 of the 416 samples from Eastern Kenya at levels between 0.01 and 9091.8 μg kg−1 (mean 67.8 μg kg−1). In South Western, AFB1 was detected in 233 of the 373 samples at levels between 0.98 and 722.2 μg kg−1 (mean 22.3 μg kg−1). Of the samples containing AFB1, 153 (55.8%) from Eastern and 102 (43.8%) from South Western exceeded the maximum allowable limit of AFB1 (5 μg kg−1) in maize for human consumption in Kenya. The probable daily intake (PDI) of AFB1 in Eastern Kenya ranged from 0.07 to 60612 ng kg−1 bw day−1 (mean 451.8 ng kg−1 bw day−1), while for South Western, PDI ranged from 6.53 to 4814.7 ng kg−1 bw day−1 (mean 148.4 ng kg−1 bw day−1). The average PDI for both regions exceeded the estimated provisional maximum tolerable daily intake of AFB1, which is a health concern for the population in these regions. These results revealed significant levels of preharvest aflatoxin contamination of maize in both regions. Prevention of preharvest infection of maize by toxigenic A. flavus strains should be a critical focal point to prevent aflatoxin contamination and exposure.

Striga hermonthica (witchweed) is a parasitic weed that attacks and significantly reduces the yields of maize, sorghum, millet, and sugarcane throughout sub-Saharan Africa. Low cost management methods such as hand weeding, short crop rotations, trap cropping, or conventional biocontrol have not been effective. Likewise, Striga-tolerant or herbicide-resistant maize cultivars are higher yielding, but are often beyond the economic means of sustenance farmers. The fungal pathogen, Fusarium oxysporum f.sp. strigae, has been the object of numerous studies to develop Striga biocontrol. Under experimental conditions this pathogen can reduce the incidence of Striga infestation but field use is not extensive, perhaps because it has not been sufficiently effective in restoring crop yield and reducing the soil Striga seed bank. Here we brought together Kenyan and US crop scientists with smallholder farmers to develop and validate an effective biocontrol strategy for management of Striga on smallholder farms. Key components of this research project were the following: (1) Development of a two-step method of fungal delivery, including laboratory coating of primary inoculum on toothpicks, followed by on-farm production of secondary field inoculum in boiled rice enabling delivery of vigorous, fresh inoculum directly to the seedbed; (2) Training of smallholder farmers (85% women), to produce the biocontrol agent and incorporate it into their maize plantings in Striga-infested soils and collect agronomic data. The field tests expanded from 30 smallholder farmers to a two-season, 500-farmer plot trial including paired plus and minus biocontrol plots with fertilizer and hybrid seed in both plots and; (3) Concerted selection of variants of the pathogen identified for enhanced virulence, as has been demonstrated in other host parasite systems were employed here on Striga via pathogen excretion of the amino acids L-leucine and L-tyrosine that are toxic to Striga but innocuous to maize. This overall strategy resulted in an average of >50% increased maize yield in the March to June rains season and >40% in the September to December rains season. Integration of this enhanced plant pathogen to Striga management in maize can significantly increase the maize yield of smallholder farmers in Kenya.

No abstract

Smut disease of sugarcane, caused by the fungus Ustilago scitaminea Sydow, can cause considerable yield losses and reductions in cane quality. To investigate the reactions of the seedlings to smut, three different inoculation methods were employed. Data on number of smut whips per stool, disease incidents per population and number of tillers per plot were recorded and analysed. Results showed that screening for smut resistance at the first stage of selection to assess seedling reaction to smut is possible. The wound paste method of inoculation gave better results than the other techniques evaluated.

Beans (Phaseolus vulgaris L.) are important legumes in the semi-arid areas of Eastern Kenya. But their production is constrained by water stress within the season and between seasons. A field trial was conducted in the short rains in the year 2011 (October to December) and long rains of 2012 (March to June) at KARI-Katumani (10 35'S and 37014'E, and 1560 metres above mean sea level) to determine the effect of on soil moisture and soil nutrients on grain yield of three market -preferred bean genotypes; namely: Katumani bean 1, NUA 1 and NUA 4. The experiment was conducted using a randomized block design with three replications with treatments in split plot arrangement. Results showed that bean grain yields of Kat B1 and NUA1 grown on tied ridges with manure at 5 tons ha-1 produced significantly (P<0.05) higher yields than furrow or flat tillage beans with and without manure in the short rains of 2011. In the long rains of 2012, Kat B1 and NAU1 beans grown on tied ridges with manure out yielded all the other treatments. In the same season, grain yield of NUA 4 grown on tied ridges with manure, was higher than beans grown on open furrows, flat tillage with and without manure. In both seasons, Kat B1 produced significantly higher grain yields than all the other genotypes except NUA 4 during the short rains of 2011.

Drylands which cover one third of the earth’s land surface and almost 80% of Kenya’s land surface are being used to grow dryland crops such as maize, beans, sorghum, millets and livestock. Studies show that refined farming systems can be used in enhancing ecosystem sustainability, through the promotion of species and crop diversity. For example, cropping patterns involving intercropping legumes and cereals have demonstrated varying success in maintenance of crop diversity in the Kenyan drylands showing land equivalent ratios (LER) > 1.0, although such benefits are often lost during low rainfall seasons. Research show that some genotypes can be used to reduce soil erosion, enhance nutrient availability, soil moisture retention, microbial earthworm activities and land use efficiency. Thus critical examination of farming systems for dryland areas suggests that long term multiple effects of the ecosystem, rather than the short term benefits not only increases yields but sustains the life of ecosystems. In this reveiw we submit that monocropping systems should be modified to include varieties that are suitable for different plots in the same site to enhance efficient utilization of underground diversity. In developing farming systems modelling approaches utilizing plant genotypic and epigenetic variations, ecological, edaphic and microbial cycles should be evaluated for dryland ecosystems.