The networks of protected areas in the United States represent our nation’s best opportunity for responding to pressures such as climate change and intensifying land use that threaten biological resources. This assembly of public and privately-owned lands allows for natural adaptive responses associated with gene flow, migration, range expansion, and other important ecological processes. A simple yet robust analysis of our protected area networks is needed to inform management decisions and to identify additional priority areas for conservation. We present a series of analyses that illustrate how the basic tools of network theory can be used to assess this broad-scale conservation challenge. We used the Conservation Biology Institute (CBI) database on protected lands in the United States to identify protected areas and grouped these according to World Wildlife Fund (WWF) biomes and ecoregions. Protected lands were further classified according to their ownership, level of protection, and degree of human disturbance. Analyses were conducted for two distinct taxa: large carnivores and songbirds. Dispersal capabilities and minimum-area requirements of songbirds were considered nearly two orders lower then those of the large carnivores. The carnivore analysis explored the suitability of protected area networks in each biome for species which have been extirpated from many parts of the country and for which connectivity could be crucial to reintroduction efforts (e.g., mountain lions, bears, wolves). The songbird analysis examined the connectivity of protected area networks in three ecoregions identified by the WWF as being of global concern.

Results/Conclusions

Public protected lands were generally well connected for large carnivores, although a few biomes were only sparsely connected. The worst of these, the temperate grasslands, had a well-defined break in the network between the western and the eastern portions of the biome. Private lands played an important role in filling this gap. For the songbird analysis, the ecoregions of concern were completely disconnected without private lands. Even when including private lands in the network, connectivity was extremely sparse. Adding extractive-use lands (e.g., national and state forests) improved potential connectivity for songbirds but still resulted in a poorly connected network. We conclude that additional lands should be set-aside for long-term persistence of songbirds and other species with similar life-history characteristics. Network theory can be used to identify locations for additional protected areas that will maximize landscape connectivity and uphold the ecological processes that rely on connectivity.