Species living in human-dominated landscapes experience frequent and severe alteration of their habitats, which is identified as the single greatest threat to species conservation. The amount of habitat loss that a species can tolerate depends on the spatial pattern of habitable and unsuitable habitats. Populations occupying highly fragmented networks of heterogeneous patches are often connected by migration and consistently tend to occupy larger or more favorable habitat. Migration rate in turn is influenced by species traits and the connectivity between habitat patches. Altered migration rates affect species-specific demographic parameters and thus local dynamics and persistence in patches.

In the laboratory, we attempted to gain an enhanced understanding of the impact of spatial heterogeneity and temporal dynamics of the landscape on extinction thresholds and population dynamics using a recently developed analytical metapopulation model. Specifically, we tested in experimental labscapes the effect of habitat loss and patch dynamics by manipulating the amount and availability of resources as well as the patch turnover in space and time, employing the model insect Tribolium castaneum (Red flour beetle). We also manipulated the connectivity between patches, thereby modifying the ease with which species can disperse between resource patches. Our laboratory experiments are the first empirical application of the analytical model linking spatial realism with a stochastic landscape.