Table 3.

Mean trap catches of Anopheles gambiae in a semi-field set-up to compounds identified in bacterial headspace samples.

			BB	0.03		0.10		0.20
Compound	L	T	Mean ± SE	Mean ± SE	E	Mean ± SE	E	Mean ± SE	E
1-butanol	<0.001	0.10	38.4 ± 6.7	47.0 ± 9.5		37.4 ± 6.7		28.6 ± 4.9
2,3-butane-dione	<0.001	0.02	29.0 ± 4.4	29.5 ± 6.3		55.2 ± 14.4	+	30.0 ± 7.2
2-methyl-1-butanol	0.54	0.20	24.9 ± 4.3	23.2 ± 2.8		22.6 ± 3.5		33.4 ± 6.3
2-methyl-butanal	0.04	0.07	29.6 ± 4.7	18.4 ± 3.3 -	-	30.1 ± 5.2		23.3 ± 3.9
2-methyl-butanoic acid	0.001	0.93	19.6 ± 3.1	19.0 ± 4.6		18.0 ± 3.9		20.1 ± 4.3
3-hydroxy-2-butanone	<0.001	<0.001	44.0 ± 9.3	52.8 ± 6.4		33.3 ± 5.8	-	27.8 ± 4.3	-
3-methyl-1-butanol	<0.001	<0.001	19.1 ± 3.3	29.1 ± 6.2		61.3 ± 9.6	+	29.8 ± 6.4	+
3-methyl-butanal	0.007	0.670	25.8 ± 7.2	22.3 ± 3.6		24.4 ± 7.1		18.9 ± 2.3
3-methyl-butanoic acid	<0.001	<0.001	32.4 ± 4.1	11.9 ± 0.8	-	58.0 ± 5.8	+	13.0 ± 1.9	-

All traps were provided with CO₂ and the basic blend (BB). To three traps a LDPE sachet of different thickness (0.03 mm, 0.10 mm or 0.20 mm) was added containing one of the 10 test compounds. The effect of trap location (L) and treatment (T) on the mean number of mosquitoes caught per night (± standard error) is given (GLM, y = location*x₁ + treatment*x₂, P-values given). The effect of the compound tested on the 'attractiveness' of the BB is indicated: + = significant increase of mosquito catches compared to the BB, - = significant reduction compared to the BB (GLM, P < 0.05).