In this post Associate Editor Yann Clough discusses a paper he recently handled by Mattias Jonsson and colleagues ‘Experimental evidence that the effectiveness of conservation biological control depends on landscape complexity

Brassica crops worldwide are attacked by a range of herbivorous insects, and frequent insecticide use is common. New research shows that the natural biological control of pests can be achieved by establishing strips of buckwheat, which increases yields in kale crops. The effectiveness of this measure is greatest in landscapes with a significant share of annual crops.

The creation of habitat supporting natural enemies in agricultural landscapes is a cornerstone of conservation biological control and ecological intensification of agricultural production. Flower strips and grassy beetle banks have been shown to support natural enemies and pest control in many crops across the world. Buckwheat has been shown to increase both predators and parasitoids of brassica pests (Orre et al. 2013).

A buckwheat strip. Photo credit: Mattias Jonsson
A buckwheat strip. Photo credit: Mattias Jonsson.

It is important to realize that the effectiveness of these habitats can vary considerably and systematically with the type of landscape. In simplified landscapes, where disturbed habitats such as annual crop fields dominate, and undisturbed habitat is scarce, adding habitat provides refuges and food resources that are otherwise lacking. In highly complex landscapes with much undisturbed habitat, enough resources are already available, and natural biological control may be difficult to improve by adding habitat.

This principle, known as the ‘intermediate landscape complexity hypothesis’ (Tscharntke et al. 2005) can in principle help prioritize the distribution of habitats across a region, and may help growers locally to decide whether or not to incur the costs for habitat establishment. Empirical data to inform conservation biological control is still scarce, however. This is a problem for translating theory into practice because (1) while effects on biodiversity have often been shown, effects on yield are more elusive, and (2) a complex or simple landscape depends on the context.

Natural enemies visiting buckwheat flower strips. D. semiclausum (left) and ladybird (right). Photo credit: Mattias Jonsson.
Natural enemies visiting buckwheat flower strips. D. semiclausum (left) and ladybird (right). Photo credit: Mattias Jonsson.

Mattias Jonsson and colleagues established buckwheat strips in landscapes differing in annual crop cover and assessed the pests, natural enemies and crop growth in an adjacent, untreated kale strip. Aphids and the diamondback moth were among the important pests in the crop, with parasitoids being primarily responsible for their control. Jonsson et al. use structural equation models to disentangle the cascading effects of landscape context, presence or absence of buckwheat strips and fertilization on parasitism, pest densities, and crop yield.

The authors find that landscapes with up to 75% non-crop habitat – a high value – are impoverished enough for buckwheat strips to have a significant impact on the biological control and plant size. This is quite different from other numbers put forward in other parts of the world (e.g. Tscharntke et al. 2005). Jonsson and colleagues suggest that most non-crop habitats in New Zealand were dominated by a low diversity of exotic species, and are therefore potentially less valuable for natural enemies.


Orre, G.U.S., Wratten, S. D., Jonsson, M., Simpson, M., & Hale, R. (2013). ‘Attract and reward’: Combining a herbivore-induced plant volatile with floral resource supplementation–Multi-trophic level effects. Biological Control, 64(2), 106-115.

Tscharntke, T., Klein, A. M., Kruess, A., Steffan‐Dewenter, I., & Thies, C. (2005). Landscape perspectives on agricultural intensification and biodiversity–ecosystem service management. Ecology letters, 8(8), 857-874.