Habitat loss and fragmentation are the leading threats to biodiversity worldwide. In fragmented landscapes, movement of organisms among patches of suitable habitat is critical to ecological and evolutionary processes. Movement of organisms allows maintenance of genetic diversity, rescue of declining populations, and recolonization following local extinctions. Connectivity, the degree of movement among patches, is dependent on both the spatial arrangement of habitats, or other landscape elements, and behavioral responses of individuals to physical structure of a landscape. Hence, connectivity for a species is best understood by examining movement at multiple scales. Here, I examine movement of Poliocitellus franklinii (Franklin???s ground squirrels), a grassland species of conservation concern, at multiple scales in agricultural landscapes in the Midwestern United States. First, I investigated the relative influence of habitat quality, area, and connectivity on patterns of site occupancy and extinction for P. franklinii at the landscape scale. While many methods are used to measure connectivity, debate about which metrics are preferred for applied conservation issues is ongoing. Thus, while investigating occupancy dynamics for P. franklinii using 3 years of presence-absence data collected at 55 sites, I also compared predictive performance of Nearest neighbor (NN) and Incidence Function Model (IFM) metrics based on both Euclidean and cost-distances. I found habitat quality, area, and connectivity were all positive predictors for occupancy by P. franklinii. Only isolation was a positive predictor of extinction, suggesting connectivity was especially important in allowing dispersers to rescue populations from local extinction. A simple NN metric measuring Euclidean distance between a target patch and nearest occupied source (NS) outperformed IFM metrics (Euclidean and cost-distance) in predicting occupancy and extinction for P. franklinii, indicating simple NN metrics may be adequate when measuring connectivity for rare species with few occupied habitat patches within dispersal distance. Low densities, secretive behavior, and a tendency to burrow underground in thick grass make it difficult to survey for P. franklinii. Therefore, while conducting occupancy surveys for P. franklinii, I collaborated with Working Dogs for Conservation to develop use of detection dogs in occupancy surveys for cryptic small mammals. I livetrapped at 62 sites for P. franklinii and surveyed 40 of those sites using detection dogs. Independent surveys of a site by 2 dog-handler teams took <1 hr and resulted in detection rates comparable to 2 daily livetrapping surveys (detection rate = 83-84%). Surveys by 2 dog-handler teams can cover more and larger sites in a shorter time than 2 daily livetrapping surveys, with only a moderate increase in cost. I conclude a two-stage strategy could be used effectively in large-scale surveys for a variety of rare and cryptic species, whereby livetrapping is conducted only at sites where detection dog surveys indicate presence. Next, I used a field experiement to examine movement decisions of adult P. franklinii in agricultural landscapes during mating season in the spring. By translocating radiocollared adult squirrels across fallow crop fields and tracking routes home, I tested if individuals based gap-crossing decisions on lengths of alternate movement routes or on individual energetic constraints. I used giving-up densities to determine that P. franklinii perceived a higher risk of predation in crop fields than in grass, but travel speed, calculated using telemetry locations, was not strongly adjusted to counteract risk when traveling through crop fields. Squirrels did not appear to base gap-crossing decisions on lengths of alternate movement routes. Instead, body mass was the only predictor of gap crossing; lighter squirrels were more likely than were heavier squirrels to cross crop fields. The importance of body mass in gap crossing decisions suggests sciurid movements through heterogeneous landscapes are partly explained by trade-offs between predation risk and energetic constraints. Additionally, all squirrels translocated <400 m homed within 24 hrs of release, but likelihood of homing decreased with increasing crossing distance. High homing success suggests P. franklinii can be proficient at moving across an agricultural landscape, but the low frequency of field crossing implies fallow fields may be areas of high predation risk that inhibit adult movements and gene flow during the mating season in the spring. Finally, I examined how variation in individual personality might influence movement and habitat use for P. franklinii. I used hole-board tests to characterize personalities of individuals and then related personality to space use by squirrels. P. franklinii individuals demonstrated repeatable personalities characterized by vigilance and activity. For radiocollared, juvenile males, individual home range size and length of movements related negatively to vigilance and positively to activity demonstrated during hole-board testing. Findings suggest variation in individual personalities characterized by vigilance and activity can affect space use, and given a positive relationship between home range size and dispersal distance, may potentially influence connectivity and metapopulation dynamics at a larger scale.