METAPOPULATION DYNAMICS: EFFECTS OF HABITAT QUALITY AND LANDSCAPE STRUCTURE

Metapopulation dynamics have received much attention in conservation and population biology, but the standard approach has also been criticized for being too restrictive, as it is based on the effects of habitat patch area and isolation only. Here we demonstrate how the effects of habitat quality (extra environmental factors) and detailed landscape structure (described with GIS [Geographical Information System]) can be included in a spatially realistic metapopulation model, the incidence function model. Expanded models are tested with a large data set on the Glanville fritillary butterfly ( Melitaea cinxia ). The incidence function model supplemented with additional environmental factors revealed some new and confirmed some previously known interactions between M. cinxia and its environment. However, the ability of the additional environmental factors to explain the error in the fit of the basic model was generally low (≤15%). In the second variant of the basic model, landscape structure was used to modify effective patch isolations. This approach, though biologically appealing, failed to improve significantly the fit of the incidence function model. There are several possible reasons for this failure, including inaccurate satellite data, problems with habitat classification, and most importantly, generic problems in the modeling of migration. Our results demonstrate that additional complexity beyond the effects of habitat patch area and isolation does not necessarily improve the predictive power of a metapopulation model.