Understanding the relative importance of neutral and non-neutral processes for adaptive evolution is essential to understand the origin of biodiversity [25]. While the vast majority of theory on ecological specialisation has been developed for situations under population equilibrium or single species dynamics [5,26,27], there is mounting evidence that eco-evolutionary dynamics [28] largely depend on the population dynamics and metacommunity context [10,29]. I will establish a large, replicated experimental study using the mite Tetranychus urticae as a model (e.g., [23,30,31,32,33]). Spider mites adapt very swiftly to novel host plants [34]. The species’ genome is small, sequenced and annotated [35], and enables us to obtain allelic frequency information at a genome wide scale [36]. I specifically aim to test