Faster movement in non-habitat matrix 1 promotes range shifts in heterogeneous landscapes 2 3 4

Ecologists often assume that range expansion will be fastest in landscapes composed 23 entirely of the highest quality habitat. Theoretical models, however, show that range expansion 24 depends on both habitat quality and habitat-specific movement rates. Using data from 78 species 25 in 70 studies, we find that animals typically have faster movement through lower-quality 26 environments (73% of published cases). Therefore, if we want to manage landscapes for range 27 expansion, there is a tradeoff between promoting movement with non-hostile matrix, and 28 promoting population growth with high-quality habitat. We illustrate how this tradeoff plays 29 out using an exemplar species, the Baltimore checkerspot butterfly. For this species, we 30 calculate that the expected rate of range expansion is fastest in landscapes with ~15% high31 quality habitat. Behavioral responses to non-habitat matrix have often been documented in 32 animal populations, but rarely included in empirical predictions of range expansion. Considering 33 movement behavior could change land planning priorities from focus on high-quality habitat 34 only to integrating highand low-quality land cover types, and evaluating the costs and benefits 35 of different matrix land covers for range expansion. 36