Search

Habitat modeling across a landscape that has gradients of habitat conditions requires potential predictor data that can be quantified at biologically relevant scales. We used remotely sensed data to develop a multi‐scale density model in 2018 for the golden‐cheeked warbler (Setophaga chrysoparia; warbler), a species that breeds in Ashe juniper (Juniperus ashei)‐oak (Quercus spp.) woodlands in central Texas, USA. We first classified Ashe juniper and broadleaf tree cover at a 1‐m resolution and used this to map potential habitat across the warbler's >67,000‐km2 breeding range. We then designed a survey for estimating warbler density based on hierarchical distance sampling. We used stratified random sampling to survey for male warblers at 1,804 points across the continuum of tree canopy cover and composition and detected 810 warblers during our surveys. We developed a suite of potential predictor variables for modeling warbler density that reflected vegetation, topography, climate, and anthropogenic land use conditions across the breeding range and developed these at 3 scales representing the territory, site, and landscape. We modeled warbler density and used the best fit model to produce a spatially explicit estimate. Predicted warbler density was influenced by tree canopy cover and canopy height at the territory scale (100‐m radius); tree canopy cover, percent of the canopy comprised of juniper, and an interaction between canopy cover and compound topographic index at the site scale (1‐km radius); and annual temperature range at the landscape scale (5‐km radius). We estimated a population size of 217,444 male warblers (95% CI = 153,917–311,965) and >3,000 males in each recovery unit. After controlling for the duration of point count surveys, our estimate of population size was similar to that reported from the only previous breeding range survey conducted in 2008–2009. Our model results indicated that management activities to increase warbler density should promote woodlands with high tree canopy cover, approximately 60–80% Ashe juniper composition, and tree heights >3 m. In contrast to a patch‐based approach, our treatment of habitat variables as continuous helped to credibly map the warbler distribution across areas with broad transitions from woodlands to shrublands. By measuring these predictor variables at biologically relevant scales, we allowed the warbler survey data to define habitat relationships instead of using anthropogenically defined habitat patches. Outcomes from our study show the benefits of developing spatial products tailored to individual species of interest for conservation and management decisions. [ABSTRACT FROM AUTHOR]

City greenbelts or greenways are usually created to achieve multiple goals, often including conservation of agricultural or range land, preservation of unique environments and heritage, recreation, and creation of transportation and utility corridors. Tradeoffs may be required in meeting these multiple goals. Analysis of potential future trajectories for urban development with such greenbelts is important to characterize the drivers and spatially explicit patterns of change that may ensue. Spatially explicit multi-criteria analysis provides one approach for examination of driver and patterns. In this paper we use a spatial multi-criteria analysis shell (MCAS-S) to explore two case studies: (1) a 1250 ha subset of the Toronto Greenbelt in the Durham region of Ontario (Canada); and (2) the Balcones Canyon lands Conservation Plan area for endangered species protection on the western side of Austin, Texas (USA). In both cases, urbanization and encroachment are major pressures. The study areas were described by GIS data layers grouped into themes that capture major competing factors-e.g., residential preference, physical constraints to development, agricultural and environmental assets, tourism, and infrastructure development. Rules were constructed that combined preference for access to green space with limits to development, and a range of scenario layers were created to indicate likely future development areas. Development in the Toronto Greenbelt was mostly associated with areas of low likelihood of development, probably because areas that met all requirements-limited impact on agriculture and environment but helpful for tourism and infrastructure-were difficult to find. By contrast, development in the Austin BCCP was associated with urban infilI modified by proximity to preserves. Although the results are only associative and not causative, they illustrate the difficulties of satisfying multiple competing goals in greenbelts, and the tendency for preserved patches to be incrementally surrounded by development. [ABSTRACT FROM AUTHOR]