Plant-herbivore-predator communities and grassland management intensity - Implications for biodiversity conservation practices on local and landscape scales

The rapid growth of the human population and the concomitant demands for fuel, food and agricultural products put increasing pressures on ecosystems around the world. This has become a major driver of global environmental change, resulting in the destruction or modification of natural terrestrial ecosystems, often followed up by dramatic losses in biological diversity. Effective management strategies are needed, balancing biodiversity conservation and agricultural production. <br/> In this thesis, we analyse the direct and indirect influences of local land use intensification on diversity and abundance patterns of herbivore and predator insect communities in natural grassland ecosystems across three different regions in Germany. We assess landscape composition and configuration effects on pest insects spilling over from crop fields to grasslands, affecting fruit set of wild plants. Finally, we conclude with suggestions for biodiversity conservation practices on local and landscape scales. <br/> Chapter 1 gives an overview of the topics covered in this thesis. In Chapter 2, we use the framework of a large-scale and multi-site experiment, located in three regions of Germany along a 600-km north-south gradient, to show that local land use intensification reduces diversity and abundance of the herbivore community consistently across all study regions. Mowing frequency, affecting vegetation height and local plant species richness, excelled as key predictor of declines in herbivore diversity. Predators were bottom-up controlled via herbivore diversity and not directly influenced by management intensification. Experimentally established short-term and small-scale succession enhanced herbivore diversity, which in turn increased predator diversity via bottom-up effects. Interestingly, efficiency of this conservation practice increased with land use intensity, in particular with mowing frequency. Temporarily allowing successional subplots on intensively used grasslands appeared to be a low cost but high benefit conservation measure introducing and sustaining habitat heterogeneity and insect diversity despite of ever changing land use practices and environmental conditions. <br/>Chapter 3 relates management-induced reduction in host plant height to colonization success of specialized grass-shoot miners. In this study, changes in host grass heights resulted from experimentally excluding subplots from management (1-year set aside) or of an a priori selection of continuously managed and already abandoned (≥2-year set-aside) grassland patches. We show that abundances of herbivores are negatively affected by a management–induced reduction in host shoot length over time. Especially on continuously managed grasslands, overlaps in attack heights between species increased, potentially narrowing spatial niche width for co-occurring species. Mean grass shoot length above 1 m in the abandoned areas was a threshold, boosting colonization success for two out of three species. We conclude that implementation of small grassy strips e.g. located at grassland edges, excluded from grazing and mowing for at least two years should be highly effective in supporting a diverse and abundant insect community and may provide suitable refuges from which stem-borers can re-colonize sward islets on managed grasslands. <br/> Chapter 4 presents a grassland phytometer experiment that quantifies scale dependent crop-noncrop spillover of rape pollen beetles (Brassicogethes aeneus) and fruit set of insecticide-treated and untreated wild mustard plants (Sinapis arvensis) as a consequence of herbivore damage. We used a set of 20 grassland plots along a gradient of increasing proportion of oilseed rape (OSR) within eight circles (with 250 to 2000 m radius) around the centre of each plot. Our results show that pest insect spillover from crop fields is triggered by pulsing availability of mass-flowering crops at different spatial scales. Mustard in the direct neighbourhood of cropland (250 m) suffered first from increased herbivore spillover since damage increased with decreasing distance from crop fields after mass-flowering peak of crops. Thus, a minimum distance of 250 m between large crop monocultures and wild plants (of the same family) in grasslands may reduce potentially negative herbivore spillover effects. <br/> Taken together, the results presented in this thesis provide evidence that land use intensification reduces diversity and abundance of insect communities in grasslands via direct and indirect effects. Mowing frequency excelled as a key driver of herbivore diversity decline in all three regions. Pest insect spillover across the crop-noncrop interface increases with higher proportions of crop fields at small spatial scales and can reduce fruit set in wild plants, suggesting to keep a minimum distance between crop monocultures and habitats when high biodiversity (e.g. on calcareous grasslands) needs to be protected. We further show that even small and easily set up successional islets on meadows and pastures can serve as important refuges for the diversity of herbivores and associated predators. Since the efficiency of this conservation practice increases with (i) land use intensity and (ii) duration, we expect the highest benefit of long-term grassland set-asides on intensively managed grasslands.