Statistical models to evaluate invertebrate-plant trophic interactions in arable systems.

Over the past 40 years there have been marked shifts in arable farmland management that are widely believed to have had a considerable impact on flowering plants and invertebrates and the small mammals and birds that rely upon them. It is not yet possible to predict the dynamics of plants and invertebrates either with past or future changes in farmland management. This study investigates whether a basic invertebrate classification, formed of broad trophic groups, can be used to describe interactions between invertebrates and their resource plants and evaluate management impacts for genetically modified, herbicide-tolerant (GMHT) and conventional herbicide management in both spring- and winter-sown oilseed rape. It is argued that the analyses validate trophic-based approaches for describing the dynamics of invertebrates in farmland and that linear models might be used to describe the changes in invertebrate trophic group abundance in farmland when driven by primary producer abundance or biomass and interactions between invertebrates themselves. The analyses indicate that invertebrate dynamics under GMHT management are not unique, but similar to conventional management occurring over different resource ranges, and that dynamics differed considerably between spring- and winter-sown oilseed rape. Thus, herbicide management was of much lower impact on trophic relationships than sowing date. Results indicate that invertebrate dynamics in oilseed rape are regulated by a combination of top-down and bottom-up trophic processes.