Main content area

Ecological characteristics of Simulium breeding sites in West Africa

Cheke, Robert A., Young, Stephen, Garms, Rolf
Acta tropica 2017 v.167 pp. 148-156
Simulium damnosum, air temperature, altitude, analysis of variance, breeding sites, cluster analysis, correspondence analysis, data collection, habitats, hydrology, insects, larvae, principal component analysis, pupae, relative humidity, rivers, stream channels, variance, vegetation, water temperature, Benin, Ghana, Liberia
Twenty-nine taxa of Simulium were identified amongst 527 collections of larvae and pupae from untreated rivers and streams in Liberia (362 collections in 1967–71 & 1989), Togo (125 in 1979–81), Benin (35 in 1979–81) and Ghana (5 in 1980–81). Presence or absence of associations between different taxa were used to group them into six clusters using Ward agglomerative hierarchical cluster analysis. Environmental data associated with the pre-imaginal habitats were then analysed in relation to the six clusters by one way ANOVA. The results revealed significant effects in determining the clusters of maximum river width (all P<0.001 unless stated otherwise), water temperature, dry bulb air temperature, relative humidity, altitude, type of water (on a range from trickle to large river), water level, slope, current, vegetation, light conditions, discharge, length of breeding area, environs, terrain, river bed type (P<0.01), and the supports to which the insects were attached (P<0.01). When four non-significant contributors (wet bulb temperature, river features, height of waterfall and depth) were excluded and the reduced data-set analysed by principal components analysis (PCA), the first two principal components (PCs) accounted for 87% of the variance, with geographical features dominant in PC1 and hydrological characteristics in PC2. The analyses also revealed the ecological characteristics of each taxon’s pre-imaginal habitats, which are discussed with particular reference to members of the Simulium damnosum species complex, whose breeding site distributions were further analysed by canonical correspondence analysis (CCA), a method also applied to the data on non-vector species.