Aim This study was conducted to investigate the potential of predicting alpha diversity and turnover rates of a highly diverse herbivorous insect family (Geometridae) based on vascular plant species richness and vegetation structure.
Location The study was carried out on the south-western slopes of Mount Kilimanjaro within a wide range of habitats between 1200 and 3150 m elevation.
Methods The floristic and structural composition of the vegetation was recorded at 48 plots of 400 m2. Geometrid moths were sampled manually at light sources located at the plot centres. Principal components analysis, redundancy analysis and multiple linear regression were used to explore how alpha diversity and species turnover of geometrid moths are related to vegetation structure and plant species richness.
Results Alpha diversity of geometrid moths was significantly correlated with species diversity patterns in the most common vascular plant families (R2 = 0.49) and with plant structural parameters (R2 = 0.22), but not with overall floristic diversity. Species turnover of geometrid moths was strongly linked to diversity changes in a range of plant families (40% explained variance), less strongly to changes in vegetation physiognomy (25%), and only weakly to overall floristic diversity (5%). Changes in elevation were a better predictor of both alpha diversity and species turnover of geometrid moths than any principal component extracted from the vegetation data.
Main conclusions Vegetation composition, diversity and structure all showed significant correlations with the diversity and species composition of geometrid moth assemblages. Nevertheless, in most cases relationships were indirect, via environmental parameters such as temperature and humidity, which influenced both vegetation and moth fauna. Possible direct links between geometrid diversity and potential food plants were much weaker. The lack of a significant correlation between overall plant species richness and geometrid diversity indicates that tropical geometrid moths may not be very selective in their food plant choice. Accordingly, a clear correlation between floral diversity and herbivore species richness must be regarded as overly simplistic, and the diversity of vascular plants cannot universally be used as a suitable biodiversity indicator for diverse insect taxa at higher trophic levels.