Abstract
We hypothesize that the ability of an organism to move through a landscape is determined by the interaction between its movement behaviour and the landscape structure. In contrast, models predicting spatial distribution, local population stability, or metapopulation stability typically assume that movement ability is independent of landscape structure. These model predictions will be invalid if the assumption of constant movement ability is incorrect. To assess the influence of landscape structure on movement behaviour (and therefore movement ability), we tracked individual goldenrod beetles (Trirhabda borealis) through microlandscapes composed of three patch types (goldenrod, cut vegetation, and cut vegetation containing camouflage netting to a height of 50 cm) that differed in terms of available food resources and structural complexity. In goldenrod patches, beetles moved infrequently in brief bursts of slow meandering movements. In cut patches, beetles moved frequently in sustained bursts of slow directed movements. In netting patches, beetles moved frequently in brief bursts of fast meandering movements. Using mark-release experiments, we determined that T. borealis did not detect goldenrod from afar or respond to edge type. Since T. borealis movement behaviour differed between patch types, its movement ability must depend on landscape structure. If this general result applies to other species, it implies that predictions of local population and metapopulation responses to landscape alteration could be erroneous. Effects of landscape alteration on movement behaviour should be incorporated into models of population response to landscape alteration.