The acute toxicity of three formulations of commercial detergent (ROMA®, FOCA® and BLANCA NIEVES®) was evaluated using the polychaete Capitella sp. C in static bioassays over a 48-h exposure period. The probit method was used to determine the median lethal concentration (LC50) of each formulation as a whole as well as the LC50 of the active ingredient, linear alkylbenzene sulfonate (LAS), using 95 % confidence intervals. The formulations and LAS showed LC50 values of between 70.79 and 147.91 ppm and 15.48 and 22.38 ppm, respectively, at 48 h. FOCA® was the most toxic detergent, followed by BLANCA NIEVES® and finally ROMA®. The variation in the toxicity of the three detergents could have been caused both by differences in the relative concentrations of the anionic surfactant LAS contained in each formulation and the presence of other ingredients (enzymes, sodium silicate, sodium tripolyphosphate, bleachers and perfumes) which can also increase formulation toxicity. Correlation analysis revealed that percent mortality of Capitella sp. C increased with increase in the concentrations of the detergent over the 48-h exposure period. The risk quotient was greater than one for all three evaluated detergents, indicating that there is a high risk that they adversely affect the aquatic biota, particularly sediment-dwelling organisms such as the test species.
Larvicidal activity was carried out as described by the WHO  with minor modifications as described by Rahuman et al.,  using third instar larvae. Four different concentrations; 0.125, 0.25, 0.5 and 1 mg/mL were tested. The extracts were prepared in 0.1 mL of DMSO and 99.9 mL of distilled water contained in a 250 mL beaker. This solvent system in a similar ratio served as the control. Three batches of twenty larvae were used for all the experiments and the number of dead larvae in each test was counted and removed after 24, 48 and 72 h of exposure. The mortality data were subjected to Probit analysis to calculate lethal concentration values (LC50 and LC90) and lower and upper 95% fiducial limits. LC50, LC90 and Chi-square values were calculated using the EPA (U.S. Environmental Protection Agency) computer Probit analysis program (version 1.5).
We are grateful to the staff at the International Centre of Insect Physiology and Ecology (ICIPE), Duduville, Nairobi, who provided support without which our research would not have been possible and Komivi Akutse for guidance on data analysis. Special thanks are extended to David Tchouassi and Vincent Nyasembe for helpful comments on the manuscript and Richard Ochieng and Milcah Gitau for their technical support. We thank Simon Mathenge (formerly of the Botany Department, University of Nairobi) for help in identification of plants. The work was supported by Training Health Researchers into Vocational Excellence (THRiVE) in East Africa, Grant number 087540, funded by Wellcome Trust. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the supporting offices. 2b1af7f3a8