Spatial Diversification Of Agroecosystems Towards Biological Control Of Insect Pests:Focus On Intercropping And Wildflower Strips

Facing the limits of input-intensive agriculture,agroecology aims at thinking ways to design a sustainable agriculture that is economically viable and socially relevant.It notably invites to mobilize ecological processes within agroecosystems in order to enhance the delivery of ecosystem services towards reducing the use of external inputs – among others insecticides.For enhancing biological control of insect pests,a strategy is to spatially diversify agroecosystems at the field scale.Whereas increasing plant diversity could directly negatively affect pest development on the one hand(i.e.bottom-up effect),providing flowering features could allow the enhancement of natural enemies and their direct effect on pest populations on the other hand(i.e.top-down effect).The present thesis focuses on intercropping(i.e.the cultivation of at least two crop species simultaneously in a same field)as a way to diversify crop habitat,and the sowing of wildflower strips as a non-crop feature.A systematic analysis of the literature revealed that,in most of studies,wheat(Triticum aestivum L.)-based intercropping allows a reduction of insect pests on crops,without necessarily favouring their natural enemies,compared to pure-stands.Besides,the provision of flowering resources,by for instance sowing wildflower strips,can attract and support flower visiting predators and parasitoids.Hence,in a first set of field experiments,combining the two tactics of increasing crop diversity and providing flowering resources was tested.First in China,wheat crop was associated with oilseed rape(Brassica napus L.),but it neither allowed reducing aphid(Hemiptera: Aphididae)abundance nor enhancing their natural enemies compared to pure stands.Instead,aphid density – independently from the treatments – affected natural enemy abundance,and interspecific relations between aphids and their natural enemies were observed.Second in Belgium,wildflower strips were sown within a wheat field,which led to a reduction of aphid density in wheat plots in between flowering features and an increase of aphidophagous hoverflies(Diptera: Syrphidae)compared to pure-stand wheat.Nevertheless,the presence of flowering strips did not affect the other natural enemies,i.e.lacewings(Neuroptera: Chrysopidae),ladybeetles(Coleoptera: Coccinellidae)and parasitoid wasps(Hymenoptera: Braconidae).Therefore,a second set of field experiments focussed on ways to compose mixtures of wild flowers attractive to a diversity of natural enemies.Flower functional traits were considered due to their effect on insect behaviour.First,the hypothesis that mixtures with high functional diversity attract and support a high abundance and diversity of aphid flower visiting predators was tested.This hypothesis was not verified.Instead,the high density in the plots of some flower species(especially the Asteraceae Leucanthemum vulgare Lam.)known to be attractive to flower visitors was supposed to have overwhelmed the effect of functional diversity.Second,a methodology was developed to identify which flower traits significantlyaffect natural enemy abundance – in this experiment parasitoids of oilseed rape beetle pests(i.e.Meligethes spp.[Coleoptera: Nitidulidae] and Ceutorhynchus spp.[Coleoptera:Curculionidae])– in flower mixtures.Among seven traits,visual traits(i.e.colour,ultra-violet reflectance)and the one related to food availability(i.e.corolla morphology)were found to significantly affect parasitoid abundance.These results highlight that(i)increasing plant diversity at the field scale can – but not systematically – favour a reduction of insect pests,(ii)including flowering features can enhance some – but not all – of their natural enemies,and(iii)in order to compose flower mixes attractive to natural enemies,specific flower traits – rather than functional diversity at the mixture level – can be considered.These results are discussed in a broader perspective.Indeed,strategies to spatially diversify crop and non-crop habitats in agroecosystems are various,as well as the ways to compose,manage and design such habitats.Also,processes at larger scales than the field may be determinant.Moreover,insects are not the only pests,and pests are not the only biotic or abiotic elements that need to be regulated in agroecosystems.Controlling multiple pests simultaneously but also enhancing the provision of multiple regulating services represent challenges for future research in agriculture.