BACKGROUND : African trypanosomosis, primarily transmitted by tsetse flies, remains a serious public health and economic challenge in sub-Saharan Africa. Interventions employing natural repellents from non-preferred hosts of tsetse flies represent a promising management approach. Although zebras have been identified as non-preferred hosts of tsetse flies, the basis for this repellency is poorly understood. We hypothesized that zebra skin odors contribute to their avoidance by tsetse flies.
METHODOLOGY/PRINCIPAL FINDINGS : We evaluated the effect of crude zebra skin odors on catches of wild savannah tsetse flies (Glossina pallidipes Austen, 1903) using unbaited Ngu traps compared to the traps baited with two known tsetse fly management chemicals; a repellent blend derived from waterbuck odor, WRC (comprising geranylacetone, guaiacol, pentanoic acid and δ-octalactone), and an attractant comprising cow urine and acetone, in a series of Latin square-designed experiments. Coupled gas chromatography-electroantennographic detection (GC/EAD) and GC-mass spectrometry (GC/MS) analyses of zebra skin odors identified seven electrophysiologically-active components; 6-methyl-5-hepten-2-one, acetophenone, geranylacetone, heptanal, octanal, nonanal and decanal, which were tested in blends and singly for repellency to tsetse flies when combined with Ngu traps baited with cow urine and acetone in field trials. The crude zebra skin odors and a seven-component blend of the EAD-active components, formulated in their natural ratio of occurrence in zebra skin odor, significantly reduced catches of G. pallidipesby 66.7% and 48.9% respectively, and compared favorably with the repellency of WRC (58.1%– 59.2%). Repellency of the seven-component blend was attributed to the presence of the three ketones 6-methyl-5-hepten-2-one, acetophenone and geranylacetone, which when in a blend caused a 62.7% reduction in trap catch of G. pallidipes.
CONCLUSIONS/SIGNIFICANCE : Our findings reveal fundamental insights into tsetse fly ecology and the allomonal effect of zebra skin odor, and potential integration of the three-component ketone blend into the management toolkit for tsetse and African trypanosomosis control.
S1 Table. Amount (in grams) of each compound in 4.5 g of each blend reflecting their natural ratios of occurrence in zebra skin odor.
S2 Table. Mean catches ± SEM in the initial trials to determine the optimum repellent doses for each compound and blend in three replicates trials.
Blend A, K, and Z indicate 4-component blend of aldehydes, 3-component blend of ketones and 7-component blend of all EAD-active compounds, respectively, in their natural ratios of occurrence in zebra skin odor. Values underlined represent catches at optimum repellent dose (i.e. dose with the least catch or not significantly different from the dose that did).
S3 Table. Release rates ± SEM of individual compounds and blends from polyethylene sachets in three replicates.
Blend A, K, Z and 2C Blend K indicate 4-component blend of aldehydes, 3-component blend of ketones, 7-component blend of all EAD-active compounds and 2-component blend of ketones respectively, in their natural ratios of occurrence in zebra skin odor.
S1 Fig. GC/EAD profiles showing antennal responses of G. pallidipes to commercial standards of the physiologically-active compounds in zebra skin odor extract.
A aldehydes and B ketones. Upper trace is the GC/FID and lower traces represent EAD responses. 1, heptanal; 2, 6-methyl-5-hepten-2-one; 3, octanal; 4, acetophenone; 5, nonanal; 6, decanal; and 7, geranylacetone. The EAD runs were scaled to 10mV/div.