Connectivity Restoration in Large Landscapes: Modeling Landscape Condition and Ecological Flows

Without intervening areas of natural habitat, relatively intact protected areas become isolated over time, limiting evolutionary processes and reducing species and genetic diversity. As a result, maintaining large-landscape habitat connectivity has been recognized as an important goal of conservation since at least the 1980s (Noss 1983, Merriam 1984). During that time, robust methods have arisen to conceptualize connectivity and implement on-the-ground actions with the objective of maintaining well-connected landscapes and facilitating movement of organisms and ecological processes (Noss 1983, Crooks and Sanjayan 2006, Beier et al. 2011). Restoration ecology has traditionally focused on local or site-scale projects, and while these projects may have cumulative positive large-landscape effects over time, only recently has the problem of restoring habitat connectivity at broader scales become recognized (Noss 1991, Carroll et al. 2003). As many important connective areas occur in a matrix of public and private lands under diverse management regimes, it seems logical that enhancing habitat connectivity in large landscapes will require restoration projects aimed at increasing the overall permeability of matrix lands (Donald and Evans 2006). In many cases, restoration may provide a viable alternative to more traditional conservation actions. For example, removing a barrier to movement may be more expedient and cost-effective than acquiring land or development rights for conservation, especially in landscapes with a high demand for conversion to agricultural, commercial, or residential land uses. Moreover, habitat restoration frequently focuses on restoring an ecological process to an area (e.g., fire, aquatic flows; Noss et al. 2006). By considering habitat connectivity, practitioners can ensure that other processes, such as gene flow through fragmented landscapes or species range shifts under climate change scenarios, will also be restored (Dobson et al. 1999, Hannah et al. 2002, McRae and Beier 2007, Heller and Zavaleta 2009).

Understanding and quantifying habitat connectivity requires grappling with variation in matrix quality, which influences the ability of organisms to move among patches or the capacity for ecological processes to occur naturally (Merriam 1984, Hanski 1997, Rouget et al. 2006). Habitat quality and resistance to movement are not perfectly correlated, as organisms often require more restricted habitat conditions for some activities, such as breeding, foraging, or overwintering, than they do for others, such as dispersal or migration (Harrison 1992, Haddad and Tewksbury 2006). Restoration of large-landscape habitat connectivity can therefore afford to focus on broader measures of habitat condition and is more akin to restoration of other ecological processes, in that if connectivity for one process or organism is restored, then other processes and organisms will likely benefit (Noss et al. 2006). At the same time, a discrete barrier to movement (e.g., a busy road) may have a disproportionate effect on landscape connectivity relative to its effect on habitat quality (Fahrig and Rytwinski 2009). Movements of individuals and genes and connectivity of landforms and ecological systems have been modeled at large-landscape scales using GIS (Singleton et al. 2002; Cushman et al. 2006; Tracey 2006, WHCWG 2010), and such models can provide guidance for monitoring and prioritizing restoration efforts.

Background: Connectivity Modeling

Connectivity models can be placed into 2 main categories: 1) those that map areas important for movement among discrete landscape locations representing habitat patches, intact areas of high naturalness, or existing protected areas (Baldwin et al. 2010, Beier et al. 2011); and 2) those that map connectivity as a continuous landscape property, mapping pathways from each point on the landscape to all other points on the landscape, and showing areas important for maintaining flows more generally (Carroll et al. 2011, Landguth et al. 2011, Theobald et al. 2012). Both approaches can employ centrality analyses, in which the value or importance of linkages or sites is assessed relative to the overall network being considered (Carroll et al. 2011, Theobald et al. 2012). Mapping linkages among core polygons is intuitively appealing and may help define project areas better than continuous methods. When a project has specific, local connectivity goals (e.g., connecting 2 protected areas...