As our primary research systems, we use insect communities associated with the English oak (Quercus robur) and with cattle dung, and insect food webs of the high Arctic. Our projects span a broad range of questions, yet are united by fundamental themes. Most importantly, we use findings from individual study systems to develop general models, and to generate hypotheses testable across target taxa. In all cases, we combine long-term observational studies with targeted experiments aimed at resolving causal relationships.