The loss and fragmentation of semi-natural habitats are a major threat to biodiversity in agricultural landscapes, with potential negative consequences for the biological control of pest insects and pollination of crop plants. As habitat loss and fragmentation usually occur together, most existing studies have examined only their combined effect. While habitat loss usually has large, consistently negative effects on biodiversity, the effect of fragmentation per se is little known. Habitat isolation, expressed as the distance to neighbouring habitat patches, is one aspect of habitat fragmentation that is expected to reduce biodiversity. Habitat loss can be expressed as the change of landscape composition, i.e. the reduction of the share of a landscape covered by suitable habitat. The proposed study differentiates between the effect of habitat isolation and landscape composition on biodiversity, pollination and pest control. To this end we planted groups of young cherry trees as standardised habitat patches in 2008, which will be the study sites for 2010 - 2012.Thirty groups of seven trees were established along a gradient of landscape composition. The percentage of woody habitats ranges from 3.6 to 74.2% within 500 m radius around the sites. Independent of this gradient in landscape composition, the study sites have three levels of habitat isolation: Ten of the sites are located at the edge of dense and tall-growing forest to represent no isolation from native habitat. Another 10 sites are connected to small-sized woody habitats such as hedgerows or single trees. The remaining 10 sites are isolated from any woody habitat by at least 100 m distance. Diversity, density and parasitism of arthropods are recorded and analysed with respect to habitat isolation and landscape composition. In addition, experiments are conducted in the field to quantify the influence of predators and parasitoids on black cherry aphids. Further, we study flower visitation and pollination success of the cherry trees. The two ecosystem services aphid control and pollination are analysed with respect to habitat isolation and landscape composition. As the planted cherry trees will be followed over five years, we will be able to test the influence of habitat age on effects of habitat isolation and landscape composition.