|dc.description.abstracteng||Land-use change is a major cause of biodiversity loss. In agricultural landscapes, which cover 40% of European land, changes in land-use are spatially and temporally very dynamic, thereby changing the organisms’ habitat availability. In the last decades agriculture has been intensified through (1) shortening of crop rotations and smaller numbers of crops grown, (2) enlargement of cultivated fields and (3) loss of semi-natural habitats such as hedgerows or grassy strips. This resulted in loss of spatial and temporal heterogeneity of cropland. The spatial heterogeneity of crops can be described by two components: the diversity of crop types (compositional heterogeneity) and the spatial arrangement of cultivated fields (configurational heterogeneity) in the landscape. The temporal heterogeneity represents changes in crop composition due to annual crop succession. In intensified and dynamic landscapes, biodiversity plays a crucial role in sustaining environmental sound food production. Indeed, farmland biodiversity fulfils important ecosystem services such as biological pest control. For example, predatory arthropods such as carabid beetles and spiders can regulate pest outbreaks. Loss of species and functional diversity can have negative consequences on ecosystem services. While it is well known that semi-natural habitats interspersed within cropland enhance beneficial arthropods and associated services, there is no evidence that the heterogeneity of crops in space and time supports them.
The first and second part of this thesis present results from a pan-European and Canadian study analysing species diversity response to compositional and configurational heterogeneity. We selected 435 landscapes within five countries (Spain, France, U.K., Germany and Canada) along orthogonal gradients of increasing crop diversity (compositional heterogeneity) and increasing field border density (configurational heterogeneity). In each landscape, we sampled bees, hoverflies, carabid beetles, spiders and we monitored butterflies and birds in crops commonly grown in the regions. We showed for the first time that higher crop configurational and compositional heterogeneity can enhance carabid beetles, spiders, hoverflies, birds and plants diversity, while wild bees remained unaffected and configurational heterogeneity negatively affected butterflies diversity. Further, in the second part we showed that generalist predators (carabid beetles and spiders) and flower-visitors (butterflies and hoverflies) species traits responded to cropland heterogeneity in cereal fields across the four European countries. We found that landscapes with more crop diversity favoured spider species that live both in soil and vegetation strata and hoverflies with low reproductive potential. Landscapes with more field borders (smaller field on average) selected for specialised feeders of carabid beetles and butterflies. This is likely to have consequences on ecosystem services such as pest biological control and pollination provided by generalist predators and flower-visitors. Such cascading effects are further investigated in the next part of the thesis.
In the third part, we explored the cascading effect of landscape crop heterogeneity and generalist predators’ functional diversity of communities (carabid beetles and spiders) on biological pest control. We estimated biological control potential as the predation rates of aphids added into the same cereal fields as described above. The biological control potential was positively affected by crop composition and carabid abundance. Landscape crop diversity had a positive effect on the biological control potential, but the effect became negative at higher crop diversity levels. Furthermore, carabids communities dominated by omnivorous species enhanced biological control. However, this diet trait was not affected by landscape crop heterogeneity. Spider abundance was positively affected by higher field borders in the landscape (smaller field on average), while communities shifted to smaller species with smaller field sizes. These results show that landscape crop heterogeneity and ground-dwelling arthropods’ community traits affect biological pest control, though landscape effects are not mediated by communities’ traits.
In the fourth part, I examined crop heterogeneity effects on cereal aphid pests, predators and pest control with a focus on spatial and temporal changes in the landscape. On a subset of cereal fields in the Göttingen region in Germany, we monitored live aphids, parasitized aphids and their vegetation-dwelling enemies (e.g. hoverfly larvae). Only aphids and their parasitism were affected by both spatial and temporal crop heterogeneity. Aphid infestation decreased in landscapes with higher crop diversity when land cover of aphid resource habitats (cereal, maize and grasslands) had decreased compared to the year before. Aphid control through parasitism decreased with the inter-annual expansion in aphid host habitat, but only in landscape with small field size. These results show that pest infestation can be reduced by higher crop heterogeneity in the landscape in space and time.
Focusing on the Göttingen region in Germany, I further explored in the fifth part the effect of crop heterogeneity on carabid beetle inter-specific (at community level) and intra-specific body size traits (within three species). Landscape crop diversity decreased the community’s average body size in oilseed rape crops, while smaller-field landscapes decreased community’s average body size but increased body size of the omnivorous beetle Poecilus cupreus in cereal crops. These results suggest that not only communities’ structure can be affected by the spatial crop heterogeneity, but within-species trait variation as well.
Overall, I conclude that landscape-wide crop heterogeneity should be considered if we want to maintain and enhance agro-biodiversity and associated ecosystem services. We show for the first time that loss of crop heterogeneity or landscape homogenisation may (1) have negative consequences on species diversity, (2) favour generalist species over species with more specialized requirements and (3) reduce biological pest control. Likewise, inter-annual increase in land cover of one particular crop hosting pests can promote pest densities. This is a further argument to halt the expansion of monoculture or short crop rotations. Thus, we recommend that future landscape management strategies in Europe encourage both crop diversification and small-scale farming to sustain important ecosystem services.||de