BeskrivningThe ecological intensification of agriculture requires models that predict responses of natural pest control to landscape-scale land use, but responses to landscape characteristics are very variable between agroecosystems. Reliable predictions should stem from mechanistic approaches that are general enough to represent the diversity of agroecological systems. Here, we review existing modeling approaches and present a new framework that uses traits of crop, pest and enemy organisms to reduce system complexity, while preserving the organisms’ functional characteristics. The majority of the existing models of natural control of invertebrate crop pests at the landscape scale are statistical, while the majority of the mechanistic models are specific to one system. Among general mechanistic models, the vast majority employ systems of equations, which can generate precise predictions, without necessarily being realistic. Only a few general frameworks, described as conceptual models, sacrifice precision in favor of a realistic representation. We argue for a combined application of these two general approaches to model natural pest control, which should allow tackling uncertainty in our knowledge of the system. System representations based on different simplifying assumptions result in a robust approach to truth as the intersection of independent lies. We employed stochastic differential equations and qualitative mathematical models to represent the same agroecological systems, as these are summarized by key traits of the involved organisms. Both models predict similar, counter-intuitive system responses to land use change, which also agree with system observations. The main challenge of this approach consists in developing intuitive decision support tools for agricultural management.
|Period||2019 feb 6|
|Evenemangstitel||Swedish Oikos meeting 2019|
|Typ av evenemang||Konferens|
Dokument och länkar
Aktivitet: Deltagit i eller arrangerat evenemang › Arrangerat workshop/ seminarium/ kurs
Forskningsoutput: Tidskriftsbidrag › Översiktsartikel › Peer review