Flower Strips in Wheat Intercropping System: Effect on Pollinator Abundance and Diversity in Belgium

Abstract

The decline of pollinators in agricultural areas has been observed for some decades, this being partly due to landscape simplification in intensive agrosystems. Diversifying agricultural landscapes by sowing flower strips within fields could reduce these adverse effects on biodiversity. In this context, the study presented here aimed at assessing and comparing the abundance and diversity of bees (Hymenoptera: Anthophila) and hoverflies (Diptera: Syrphidae) found and visiting flowers in three types of flower strips in Belgium: (i) a mixture of 11 wild flowers, (ii) a monofloral strip of Dimorphoteca pluvialis (Asteraceae) and (iii) a monofloral strip of Camelina sativa (Brassicaceae), where the last two are considered to be intercrops since they are valuable on the market, all sown within a field of winter wheat (Triticum aestivum L.). Pollinators were captured with pan traps and by netting in standardised transects from May to July 2017. One-thousand one-hundred and eighty-four individuals belonging to 43 bee species and 18 hoverfly species were collected. Significant differences in hoverfly diversity were found between the different flower strips. The multifloral treatment supported a greater diversity of syrphid species. Various pollinator species visited the different flowers composing the mixture and also D. pluvialis. The pollinator community proved to be predominantly generalist, with the exception of an oligolectic species in Belgium, Andrena nitidiuscula. Moreover, the three tested flower strips were effective in attracting hoverflies, among them natural enemies of insect pests. This study opens new perspectives in the design of intercropping systems with flower strips towards the design of sustainable agro-ecosystems. Improving economic profitability of sowing flower strips could encourage farmers to diversify their agricultural systems and foster conservation biology strategies.

Keywords: Andrena nitidiuscula; Apoideae; Camelina sativa; Coriandrum sativum; Dimorphoteca pluvialis; Fagopyrum esculentum; Syrphidae; ecosystem services; sustainable agriculture.

Figure 1

Experimental setup.

Figure 2

Bar plots of mean values of both Simpson and Shannon indexes for the different treatments. The different letters represent a significant difference calculated from the post-hoc Dunn’s test comparison (p-value < 0.05).

Figure 3

Comparison between pollinator communities from the three floral treatments (denoted by colours and solid dots) by sample-size-based rarefaction (solid lines) and extrapolation (dashed curves) curves based on abundance data of hoverflies and bees together (A), bees alone (B) and hoverflies alone (C). Each panel displays Hill numbers of order N = 0 (left panel), N = 1 (middle panel) and N = 2 (right panel). The 95% confidence intervals (coloured-shaded regions) were obtained by a bootstrap method based on 100 replications.

Figure 4

Principal coordinate analysis (PCoA) ordination of the three treatments (red circle: D. pluvialis; green circle: Multifloral; blue circle: C. sativa) based on the data collected with (a) pan traps and (b) a net through transects. Ellipses show the 80% confidence interval of the locations grouped by flower strip. Species scores are represented with numbers (Table S2).

Figure 5

Principal coordinate analysis (PCoA) ordination of the four flower species with data collected with a net (red circle: D. pluvialis; yellow circle: C. sativum; green circle: F. esculentum; blue circle: C. sativa). Ellipses show the 80% confidence interval of the locations grouped by flower species. Species scores are represented with numbers (Table S2).