Weed Diversity And Their Response To Glyphosate Applications In Coffee Farms In Kiambu County, Kenya

Abstract

The problem of weed control in the east and west of rift in Kenya differs quite remarkably between the coffee growing areas. This has been verified by past studies conducted with the aim of improving coffee production by use of various means among them being the use of glyphosate in weed management. Various weed species have been identified to be associated with coffee. These programs have been applied routinely in the past decade in the coffee farms that are more than ten hectares. Production costs are a major factor in estimating coffee profitability. The annual average cost of chemical control per hectare is 10,000 Kenya shillings representing about 7% of the direct cost of weeding. Common broadleaved weeds such as black jack and gallant soldier and common soft grasses compete with the coffee crop. Their presence makes spray operations, fertilizer application and harvesting difficult and costly. Various herbicides rates have been used as influenced by factors such as weed population, resources available to the farmers, farmer practice models such as the tank mixing of two or more different herbicides among others. Particular suspect weed species have developed dominance in such farms and tend to be found in the highlands. Research studies were thus carried out to investigate the weed species diversity and their responses upon exposure to varied doses of the glyphosate herbicide (Glyphogan 480SL) in coffee farms in Kiambu County, Kenya. The weeds diversity survey was carried out quantitatively, data analyzed and presented in summary tables. These were their frequency, uniformity, relative equivalents, density and relative abundance. A total of 47 different weed species (39 broadleaved weeds 7 grasses and 1 sedge were identified. Bidens pilosa L. was the most frequent weed species at 89%. Thirteen treatments were applied in a completely randomized design investigating effective mortality, necrosis and chlorosis on treated plants. The plant mortality means against the test doses were established and separated in the data analysis. The efficient lethal concentrations established were 5.0Lt/Ha and 4.5 Lt/Ha for the 2-4 & 6-8 leaves growth stages respectively. A field evaluation was carried out pursuant to the dose tests. A completely randomized block design with eight treatments was used to establish the efficient doses. This was replicated three times and tested at the same growth stages whose efficient doses established were 5.5 and 4L/Ha respectively. It was established that such high rates are likely not to be economical if used as the only available Bidens pilosa L. weed management tool in coffee.