Do flower mixtures with high functional diversity enhance aphid predators in wildflower strips?

Abstract

Among the semi-natural elements in agricultural landscapes, wildflower strips sown at field margins or within fields are potential habitats for the natural enemies of insect pests. As insects are sensitive to a variety of flower traits, we hypothesized that mixtures with high functional diversity attract and support a higher abundance and species richness of aphid flower visiting predators than mixtures with low functional diversity. During a field experiment, repeated over two years (2014 and 2015) in Gembloux (Belgium), aphid predators (i.e., lacewings, ladybeetles and hoverflies) were pan-trapped in five sown flower mixtures (including a control mixture, with three replicates of each mixture) of low to high functional diversity based on seven traits (i.e., flower colour, ultra-violet reflectance and pattern, start and duration of flowering, height and flower class, primarily based on corolla morphology). In both years, the species of flowering plants in the sown mixtures (i.e., sown and spontaneous flowers) were listed, and the realized functional diversity in each plot calculated. Over the two years, a high functional diversity was not associated with high abundance and richness of aphid predators. Moreover, ladybeetles, which made up the majority of the predators trapped, were more abundant in mixtures with very low or intermediate functional diversity at sowing, especially in 2014. We hypothesize that certain flowers, which were abundant in certain mixtures (and not in those exhibiting the highest functional diversity), attracted predators and were sufficiently abundant to support them. Our results present novel information that could be used for developing flower mixtures that provide effective ecosystem services, such as pest control.

Keywords

• syrphidae
• coccinellidae
• chrysopidae
• agri-environmental measure
• conservation biological control
• wildflower strips
• functional diversity
• rao's index