Landscape context affects genetic diversity at a much larger spatial extent than population abundance.

Regional landscape context influences the fate of local populations, yet the spatial extent of this influence (called the "scale of effect") is difficult to predict. Thus, a major problem for conservation management is to understand the factors governing the scale of effect such that landscape structure surrounding a focal area is measured and managed at the biologically relevant spatial scale. One unresolved question is whether and how scale of effect may depend on the population response measured (e.g., abundance vs. presence/absence). If scales of effect differ across population outcomes of a given species, management based on one outcome may compromise another, further complicating conservation decision making. Here we used an individual-based simulation model to investigate how scales of effect of landscapes that vary in the amount and fragmentation of habitat differ among three population responses (local abundance, presence/absence, and genetic diversity). We also explored how the population response measured affects the relative importance of habitat amount and fragmentation in shaping local populations, and how dispersal distance mediates the magnitude and spatial scale of these effects. We found that the spatial scale most strongly influencing local populations depended on the outcome measured and was predicted to be small for abundance, medium-sized for presence/absence, and large for genetic diversity. Increasing spatial scales likely resulted from increasing temporal scales over which outcomes were regulated (with local genetic diversity being regulated over the largest number of generations). Thus, multiple generations of dispersal and gene flow linked local population patterns to regional population size. The effects of habitat amount dominated the effects of fragmentation for all three outcomes. Increased dispersal distance strongly reduced abundance, but not presence/absence or genetic diversity. Our results suggest that managing protected species at spatial scales based on population abundance data may ignore broader landscape effects on population genetic diversity and persistence, lending support to the importance of managing large buffers surrounding areas of conservation concern.