In this post Paul van Rijn discusses his recent paper ‘Nectar accessibility determines fitness, flower choice and abundance of hoverflies that provide natural pest control’

The industrialization of agriculture has strongly impoverished our countryside. The amount and quality of non-crop habitats have declined, and the biodiversity within and among crops has dropped even more. In recent years it has become clear that this not only impacts nature. The loss of biodiversity also strikes back to our economic activities, including agriculture itself. Our attempts to make agriculture more efficient have, unknowingly, removed parts of the agroecosystem that contributed to crop production.

One example is the loss of flowers from our rural landscapes. The removal of ‘non-productive’ habitats and the generous use of fertilizers and herbicides have strongly reduced the number and diversity of herbs and scrubs bearing insect-pollinated flowers. Due to a better understanding of the life cycles of beneficial insects and mites we now know that for many of them floral resources are essential for their existence. Many pollinators of our crops and predators of our pests require nectar and/or pollen during part of their life cycle; also when the crop is not in bloom.

Field margin strips are now increasingly being promoted to offset the loss of these resources. One of the projects aiming at this was the Functional Agrobiodiversity (FAB) project performed in the Dutch agricultural region of the Hoeksche Waard. For this project I was asked to design flower strips that could support natural enemies of pests, especially of aphids, in arable crops. A literature search revealed that recommendations to use specific plants are mainly based on the observed visitation of these plants by a limited number of natural enemies or on a few experiments with a very limited set-up. Clearly, not a very sound base for a project that aims at demonstrating the benefits of field margin strips. Therefore, I set out to study the fitness consequences that a range of different plant species have for important predators. The recent publication in Journal of Applied Ecology is the result of this extended study. The focus is on probably the most important natural enemies of these pests: zoophagous hoverflies. Like many other insect predators, these hoverflies feed on aphids and other prey when juvenile, but on pollen and nectar when adult.

Many flowers produce nectar to attract pollinators, but also have adaptations to limit access to a specific group of pollinators, e.g. by the length of the corolla tube. The study showed that zoophagous hoverflies can use only a very limited range of flowers. The flowers on which they survive can be characterized as open or very shallow. The maximum flower depth of 1.6 mm is much less than for other hoverflies and most bees. Until now plants of the daisy family (Asteraceae) are generally recommended as resources for hoverflies and other natural enemies, but this study showed that most species from this family are actually unsuitable. The problem probably lies in the narrow florets that, in combination with the anthers and stamen, limit the access to the nectar. Exceptions are those plants that produce extrafloral nectar, such as cornflower. Preliminary results with green lacewings (Chrysoperla sp.), species that also preys on aphids when juvenile and on flowers when adult, indicate a very similar pattern of flower suitability.

These studies have important consequences for the design and management of flower strips and other habitats when the aim is to restore natural pest control services. Many seed mixtures recommended for field margin strips may be suitable to support bees, but appear largely unsuitable for supporting important predators. Even the seed mixtures that I originally designed for the FAB project, which are still used in various field margin projects in Europe, do, in hindsight, contain not only species that support hoverflies. When comparing field margin strips that differed in flower composition the number of zoophagous hoverflies was directly related to the amount of accessible flowers. Deeper flowers (> 2mm), successfully exploited by bees and non-zoophagous hoverflies, did not have any effect on the number of zoophagous hoverflies.

Creating functional field margin strips on nutrient rich arable land proved to be an art: providing support for the right group of insects during a sufficient part of the season, avoiding pest and weed support in annual strips, avoiding dominance by grasses in perennial strips, etc. But it can be done, as farmers in the Hoeksche Waard show by the many hundreds of kilometers of field margin strips they maintain for many years now. Even when field margin strips are also created to assist farmland birds and bees, the narrow range of flowers suitable for the natural enemies of our pests should be the starting point, to which species supporting other ecosystem services can easily be added.