Creating Effective Urban Greenways and Stepping-stones: Four Critical Gaps in Habitat Connectivity Planning Research

Urban Connectivity Strategies: Corridors and Stepping-stones

Strong empirical evidence suggests that functional habitat connectivity impacts urban species richness. Less clear is what functional connectivity looks like in developed areas. Creating sizable patches of habitat in urban centers is both infeasible and undesirable. Adding large amounts of green space would reduce the density of cities and its associated efficiencies. A more practical strategy is to create a configuration of green space that connects small habitats to each other and—more importantly—to core natural areas. There are two broad, conceptual strategies for creating these connections in fragmented landscapes: corridors and stepping-stones.

Corridors create structural connectivity. They are direct, linear connections between habitats. There has been significant debate over the years on whether organisms successfully use corridors or whether they lead to increased mortality (Beier and Noss 1998). The most recent consensus for urban areas is that they do create functional connectivity in otherwise fragmented areas (Vergnes, Kerbiriou, and Clergeau 2013; Munshi-South 2012; Lee and Peres 2008; Beier and Noss 1998). The other option, stepping-stones, is small islands of habitat that serve as refuges between larger habitat patches (Carbo-Ramirez and Zuria 2011; Fernandez-Juricic and Jokimaki 2001). To support animal dispersal, stepping-stones must be functionally connected, which means that organisms must be able to move through the landscape between the refuges (Baum et al. 2004). A recent and extensive meta-analysis indicates that corridors—structural connections—are more effective at supporting urban biodiversity than stepping-stone habitats (Beninde, Veith, and Hochkirch 2015). But evidence does support the benefits of stepping-stones, particularly where the landscape between patches is relatively benign. If corridors are not possible, stepping-stones could provide habitat that allows urban organisms to move across the landscape (Leidner and Haddad 2011).

A variety of urban green spaces can serve as corridors or stepping-stones. Munshi-South (2012), for example, found that small animals use environments as diverse as cemeteries, vegetated road medians, and forested residential areas to move between city parks. The list of candidates for corridors and stepping-stones in cities is long: greenways, cemeteries, public parks and gardens, road and rail corridors, spontaneous green space (vacant lots, brownfields), green roofs, residential gardens, and residential and commercial open spaces (Ignatieva, Stewart, and Meurk 2011). The following sections review research on the state of our understanding of how they can be planned and designed to foster the movement of urban wildlife.

Do Urban Greenways Create Connectivity?

Over the past twenty-five years, biologists and ecologists have examined the connectivity functions of several types of urban greenways. Publicly owned linear parks and riparian corridors are among the most studied. As previously discussed, the ability of a natural feature to support wildlife movement will depend upon the characteristics of focal species as well as the qualities of the green space. Consequently, a useful way to discuss the impact of greenways on animals is to divide studies into taxonomic groups: arthropods, mammals, and birds. These divisions are broad, but the groups have diverging dispersal characteristics, so they provide a general idea of greenway success in supporting animal movement.

For small species such as arthropods, the quality of habitat within greenways may be as important as any movement they facilitate (Angold et al. 2006; Chovanec et al. 2002). One study found that habitat patches that were connected to railway and river corridors had butterfly and beetle communities similar to those that were disconnected. The authors concluded that “for many species of invertebrates in the urban environment…good quality habitat is the key to their continued survival rather than the more difficult task of increasing habitat connectivity” (Angold et al. 2006, 203). But other research suggests that linear landscape elements do facilitate movement by arthropods and may be crucial in maintaining urban biodiversity (Van Rossum and Triest 2012; Vergnes, Le Viol, and Clergeau 2012). The results suggest that stepping-stone habitats impact arthropod movement, and corridors probably do, but the two could work best synergistically.

Evidence that linear landscape elements help small mammals move throughout an urban landscape is less equivocal. Several studies suggest that connectivity provided by urban greenways could be beneficial for small- to medium-sized mammals—particularly those with limited habitat availability and dispersal capabilities (Angold et al. 2006; Schiller and Horn 1997; Vergnes, Kerbiriou, and Clergeau 2013). A study of suburban Paris, for example, concluded that without vegetated corridors, the movement of shrews between habitats would be limited and shrew populations genetically compromised (Vergnes, Kerbiriou, and Clergeau 2013). Evidence for larger mammals depends upon the species in question (Vergnes, Kerbiriou, and Clergeau 2013). Some suggest that large mammals move between habitats with ease and do not benefit from greenway corridors (Angold et al. 2006). A study of thirty-eight greenways in the southeastern United States concluded that a primary factor in the presence of white-tailed deer within the corridor is connectivity to large-scale natural areas beyond the greenway. The authors conclude that some greenway segments are valuable for large mammals because they provide access to adjacent lands with high habitat value (Schiller and Horn 1997).

Greenways are also supportive of bird populations—both local and migrating. A study of vulnerable bird species in suburban North Carolina found that landscape connectivity was an important predictor of bird richness and abundance in greenways but not as significant as forest size and characteristics (Mason et al. 2007). A study in Tokyo compared three urban greenways with different vegetation profiles and found that the corridors were effective at facilitating the movement of most birds. They concluded that corridors with supportive vegetation structure, even if they were originally intended for recreation, are important factors in the persistence of urban bird species (Matsuba, Nishijima, and Katoh 2016). Greenways also serve as stopover locations for migrating birds. In another study of North Carolina greenways, migrating birds used greenways of all characteristics but were more abundant in wider corridors with greater vegetation complexity and interior habitat (Kohut, Hess, and Moorman 2009).

Research on the relationship between urban greenways and wildlife movement is oriented toward river, rail, and public trail corridors. This is unsurprising because urban greenways are most common in areas where green space is limited, and creating corridors along existing linear features or recreational amenities is a practical strategy. Studies on roadside and residential subdivision greenways also exist but are largely conceptual (Viles and Rosier 2001; Arendt 2004). Few studies examine the ecological differences between the types of greenways and whether one type may have advantages over others. This research could be useful for planners charged with deciding between multiple corridor options. For greenways with transportation features, such as railways, roads, or trails, the width of the feature could make a difference. Greenways built around narrow transportation pathways could support species movement more effectively than alignments with wider features. In one study of urban barriers, for example, railways allowed for the greatest passage by forest birds because the transportation component was the narrowest (Tremblay and St. Clair 2009). Other studies conclude that riparian areas are particularly associated with native bird presence (Palmer et al. 2008) and retain their high-quality native habitat in otherwise disturbed landscapes (Jim and Chen 2003; Parker et al. 2008). However, some evidence suggests that the quality or diversity of habitat in a greenway is more important than the type of corridor. In their study of a riparian corridor in England, Dallimer and colleagues (2012) suggest that the dynamics of urban riparian zones are no different from nonriparian zones, and that other forms of urban green space such as parks or “other linear features such as abandoned railway lines, could be equally able to enhance connectivity in urban settings” (p. 751). Given the spatial constraints of greenway routes in cities, ecological greenway planners may find this reassuring.

Linear Transportation Infrastructure: A Fourth Type of Greenway

One of the more pervasive modes of disturbance and fragmentation in urban areas is transportation, particularly roads. Yet transportation corridors are linear elements and often edged with trees or other managed vegetation. Several authors note the potential for roadsides to serve as greenways (Ignatieva, Stewart, and Meurk 2011; Jim and Chen 2003; Viles and Rosier 2001) Others do not use the term “greenway” but note similar potential connectivity benefits (Carbo-Ramirez and Zuria 2011; Davies, Frandsen, and Hockridge 2014; Fernandez-Juricic and Jokimaki 2001; Munshi-South 2012; Oprea et al. 2009; O’Sullivan et al. 2017). Linear transportation infrastructure such as roads, railways, and pipelines consists of a transportation lane and a “verge,” a vegetated strip around or on top of the alignment. Some also include power lines, canals, and navigable rivers in this category (Jeusset et al. 2016). Where verges are vegetated, they could create corridors that support species richness in adjacent areas, particularly if they are designed and managed to do so (Ikin et al. 2013; Saumel, Weber, and Kowarik 2016; Viles and Rosier 2001) or if they are no longer active (Foster 2014). A case study of an abandoned railway corridor in Paris, for example, notes that the “vacant” space has biodiversity comparable to the city’s two large woodlands and facilitates ecological connectivity along its length (Foster 2014). But active road corridors are much more common, particularly in urban areas, and the topic of most studies of transportation verges.

As with many facets of connectivity, the success of road verges in facilitating wildlife movement depends upon the vegetation and focal species. Studies of bats (Oprea et al. 2009), birds (Carbo-Ramirez and Zuria 2011; Fernandez-Juricic and Jokimaki 2001), and mammals (Munshi-South 2012) have concluded that road corridors could increase connectivity in urban areas, although the effect is not always strong (Oprea et al. 2009). But public roads also have safety and utility management standards that constrain management for wildlife (Carbo-Ramirez and Zuria 2011; Viles and Rosier 2001), particularly since native street vegetation is more supportive of some animals than exotic vegetation (Ikin et al. 2013). Two recent reviews (Davies, Frandsen, and Hockridge 2014; O’Sullivan et al. 2017) discuss the environmental benefits of transportation verges in the United Kingdom and identify management strategies such as tree retention and reduced mowing frequency, which could enhance their ability to support biodiversity. These strategies would also augment the verges’ ability to support species movement. While the body of evidence is still developing, existing information suggests that the relationship between transportation verges and connectivity is sufficient to create a fourth category of greenway, a transportation greenway.

Greenway Characteristics that Enhance Connectivity

Planners can support biodiversity by creating greenways that allow species to traverse the landscape and to access large, core habitat patches. The design of a conservation corridor determines many of its ecological benefits. As Noss and Harris noted in 1986, “In almost any landscape are linear or sinuous structures…that can potentially connect or sever landscape patches and nodes. Whether a given feature serves as a barrier or as a dispersal corridor depends on habitat structure within the corridor, corridor width, [and] the position of the corridor relative to habitat patches in the landscape” (p. 304). Based on a review of existing evidence, recommendations for greenways that support animal movement are identified below and discussed in the following section. These could be described as best practices of greenway design. Some of the actions (e.g., wide greenways with connections to core habitat areas) are well supported by empirical observations, and others (e.g., encouraging private landowners to manage their land as habitat) are frequently recommended but rarely examined.

Greenway planning recommendations:

Create wide greenways with narrow trails that are located near the edge of the corridor.

Many studies highlight the importance of width in the ability of greenways to support animal movement. A corridor should be wide enough to provide space for resources such as food and shelter, minimize disturbance from surrounding land uses, and facilitate travel between habitat patches (Parker et al. 2008). Yet studies on the minimum width standards for any type or location of corridor are rare (Harris and Atkins 1991; Lee and Peres 2008; Parker et al. 2008). The primary reason for this is probably the practical considerations of green space delineation in cities. Open space availability rather than a designer often dictates urban greenway widths (Parker et al. 2008). A second reason is that greenways should be designed to provide functional connectivity for a specific focal species, so the appropriate dimensions vary. Wider may not even be better for some species (Andreassen, Halle, and Ims 1996). Harris and Atkins (1991) note that, “the width of corridors should be gauged in units of the landscape, not simple, human-imposed numbers…a vegetation strip along a creek may be extremely great for small creatures [but]…totally useless for other species…” (p. 124). They go on to cautiously propose a scale running from a width of 30–300 feet (9–90 meters) for a few well-known species and short-term function to 3,000+ feet (900+ meters) for greenways that will support the movement of a broad number of species over the long term (Harris and Atkins 1991). All studies that discuss urban greenway width note that wider is better, and several do suggest minimum thresholds, particularly for bird populations. Austin (2012) suggests 10–30 meters for moderately human-adapted species, while Schiller and Horn (1997) classify greenways of less than 30 meters as narrow and greater than 30 meters as wide (Austin 2012).

Species that are more sensitive to disturbance will require wider corridors. One study found that interior bird species were not present in urban greenways narrower than 50 meters, and more sensitive species were only present in greenways wider than 300 meters. The authors suggest greenways wider than 100 meters (Mason et al. 2007). Another avian diversity study suggests that 150 meters or wider is most desirable (Kohut, Hess, and Moorman 2009). One way to add width without investing in additional greenway space is to encourage adjacent property owners to retain vegetation (Kohut, Hess, and Moorman 2009; Parker et al. 2008). Not only would this widen the effective greenway, but evidence suggests that adjacent land uses are a factor in the ability of greenways to support birds and mammals (Kohut, Hess, and Moorman 2009; Schiller and Horn 1997). Several authors recommend the strategy, but no peer-reviewed literature directly examines its implementation or outcomes.

Another factor in greenway planning is disturbance. Greenways are often designed to support recreation as well as wildlife, so many have landscaped, paved, and mowed areas (Turner 2006; Hay 1991). Evidence suggests that arthropods (Angold et al. 2006), birds (Mason et al. 2007), and mammals (Schiller and Horn 1997) can be impacted by disturbance from recreation uses of trails and paths. As Schiller and Horn (1997) note, the creation “of greenways can potentially reduce habitat value…if remnant urban wildlands are converted to highly managed greenways” (p. 114). More generalist species are present, and overall species abundance and richness is lower in greenway segments with more managed surface (Mason et al. 2007). The result is concerning but unsurprising. Avian and insect species are known to decline as impervious surfaces increase (Dallimer et al. 2012; Schutz and Schulze 2015). Mason and colleagues (2007) argue that greenways with small amounts of managed area may support up to twice as many development-sensitive bird species as those with two to four meter wide trails surrounded by a mowed area (p. 159). Narrow forest greenways with wide paved trails are also associated with high risks of avian nest predation (Sinclair et al. 2005). Planners can support development-sensitive species by minimizing the width of greenway trails and adjacent mowed, paved, and bare surfaces (Mason et al. 2007; Kohut, Hess, and Moorman 2009; Mortberg and Wallentinus 2000). One simple action would be to locate trails near the edge of a greenway instead of in the center (Mason et al. 2007). The action would maximize the width of the undisturbed area without impacting recreation uses. It is unclear whether this strategy has been adopted. It has not been examined empirically.

To support species movement, land within a greenway should be managed as habitat. Matsuba, Nishijima, and Katoh (2016) found that corridor vegetation could have a greater impact on corridor effectiveness than width. The risk of organism mortality can also be greater when a corridor does not provide important resources such as food and shelter (Parker et al. 2008). Ideally, a greenway network would protect a representative array of regional ecosystems (Flores et al. 1998). Forest cover is a prominent feature in many greenway ecology studies. Tree cover is often associated with species movement, abundance, and diversity (Harris and Atkins 1991; Kohut, Hess, and Moorman 2009; Mason et al. 2007; Matsuba, Nishijima, and Katoh 2016; Mortberg and Wallentinus 2000; Schiller and Horn 1997; Tremblay and St. Clair 2009). Many of the studies on corridor vegetation focus on sensitive bird species. Urban green space corridors that contain mature and decaying trees are supportive of vulnerable bird species (Mortberg and Wallentinus 2000). Another study suggests that tree cover and understory vegetation support the movement of sensitive, urban-avoiding birds. The latter could be problematic for greenway planners, as a thick understory can appear unmanaged or dangerous and undermine aesthetic and recreation values. The authors suggest a practical strategy of managing greenway vegetation based on the goal of individual segments. Corridors near large undisturbed green spaces could be managed for urban averse birds, and corridors near more developed land use for recreation and urban-adapted species (Matsuba, Nishijima, and Katoh 2016). The quality of green spaces that are adjacent to greenway corridors is also important. Corridors surrounded by natural land cover are more likely to support wildlife (Kohut, Hess, and Moorman 2009; Schiller and Horn 1997). One way to maximize the amount and quality of natural land cover surrounding a greenway is to ensure that it connects to large protected core habitat patches. Ideally, these core areas would serve as nodes and be located throughout a greenway network (Kohut, Hess, and Moorman 2009). Connections to larger habitat areas are associated with wildlife presence in a greenway (Schiller and Horn 1997) and critical in supporting species that are not adapted to disturbance (Austin 2012; Matsuba, Nishijima, and Katoh 2016).

Planning Greenways for Animal Movement

There are a variety of strategies for identifying, connecting, and preserving a network of greenway corridors (Ndubisi, DeMeo, and Ditto 1995; Arendt 2004). Use of a network concept in green space planning is positively associated with on-the-ground connectivity (Lynch 2016). Research also suggests that delineating corridors and patches in advance of development yields greater landscape connectivity and network attributes that are more supportive of goals such as biodiversity (Benedict and McMahon 2006). In growing suburban areas, preservation can occur along with development, for example, through conservation subdivisions and open space requirements, but the scale of the resulting greenways is still relatively small. Truly landscape-scale strategies are only feasible in rural and fringe areas where significant natural land is still available. Several techniques, however, can be applied or adapted to more urban areas. This section describes the state of knowledge of greenway planning for connectivity. The median publication date of the thirty-eight included studies (2006) is notable as the oldest in the review.

Greenway planning typically includes three steps: (1) identify important natural or cultural areas, (2) identify a network of greenways that includes or connects them, and (3) identify planning strategies to create and protect the network. Several studies also recommend public involvement as a standard greenway planning component, which enhances local support and could impact ecological outcomes (Mugavin 2004; Taylor, Paine, and FitzGibbon 1995; Toccolini, Fumagalli, and Senes 2006; Yokohari, Amemiya, and Amati 2006). Greenway planning process studies (Table 2) also highlight the importance of identifying avenues for implementation (Floress et al. 2009; Mugavin 2004; Taylor, Paine, and FitzGibbon 1995), balancing natural with cultural or recreation goals (Ahern 1995; Mugavin 2004; Taylor, Paine, and FitzGibbon 1995), and taking advantage of historic corridors such as rivers (Erickson 2004) and abandoned railways (Toccolini, Fumagalli, and Senes 2006).

OPEN IN VIEWER

Table 2.

Studies that Examine the Greenway Planning Process or Have Clear Planning Process Implications.

Study (Author)	Date	Case Location(s) (n)	Greenway Planning Recommendations
Greenbelt to Greenways: Four Canadian Case Studies (Taylor, Paine, and FitzGibbon)	1995	Ottawa, Calgary, Saskatoon, and Toronto, Canada (2)	Delineate greenways with an emphasis on connectivity and natural and cultural features Maintain biodiversity while accommodating appropriate human use Gather public support by involving local residents Create efficient and affordable implementation and management strategies
Greenways as a Planning Strategy (Ahern)	1995	North Brabant, the Netherlands; Massachusetts, USA (3)	Efforts should be made to protect natural elements that are not part of the linear greenway network To be truly sustainable, greenways must integrate historic, cultural, aesthetic, and recreational goals as well as the fundamental goal of biodiversity
Adopting a Modern Ecological View of the Metropolitan Landscape: The Case of a Green Space System for the New York City Region (Flores et al.)	1998	New York City Metropolitan Area, USA (1)	Acknowledge that wildlife is harbored in land units that are not formally classified as “green space” Discourage thinking about green spaces as isolated land units that have discrete, easily mapped, boundaries and static, homogeneous ecological characteristics Connect ecological characteristics with a wide variety of environmental services, despite the frequent lack of information on natural systems and functions of interest
Adelaide’s Greenway: River Torrens Linear Park (Mugavin)	2004	Adelaide, Australia (1)	Stage or phase implementation and engage with the public and their attitudes and ideas Community processes may be as important to greenway planning as biophysical processes
The Relationship of Historic City Form and Contemporary Greenway Implementation: A Comparison and Milwaukee, Wisconsin (USA) and Ottawa, Ontario (Canada) (Erickson)	2004	Milwaukee, USA; Ottawa, Canada (2)	Regional greenways programs need a strong coordinated approach Historic green space frameworks, especially river corridors, are an advantage Successful metropolitan greenway networks benefit from visionary thinking and the leadership and cooperative structure to carry it out
Greenways Planning in Italy: The Lambro River Valley Greenways System (Toccolini, Fumagalli, and Senes)	2006	Municipalities along the Lambro River, Italy (35)	Numerous public meetings are key in ensuring a plan addresses the priorities of local residents Historic routes (e.g., old rural roads, abandoned railways) can help to create a greenway network that is cost-effective and finds a new function for old infrastructure
The History and Future Directions of Greenways in Japanese New Towns (Yokohari, Amemiya, and Amati)	2006	Tsukuba Science City and Kohoku New Town, Japan (2)	Public involvement in greenway planning helps residents to view the risks of greenway use more realistically and allow for the retention of more robust vegetation Citizen involvement in greenway planting and maintenance, and actions that indicate ownership of public spaces can help to strike a balance between ecology and safety
The Quality of Greenways Planning in Northwest Indiana: A Focus on Sustainability Principles (Floress et al.)	2009	Jurisdictions in Northwest Indiana, USA (32)	High-quality greenway plans tend to include discussions of implementation, funding, possible greenway routes, and potential partners to assist with greenway implementation Incorporate principles of sustainability when planning for greenways

The first two steps in the greenway planning process can best be described as delineation. While there are many ways to accomplish the tasks, they are straightforward. The process involves data collection, mapping, analysis, and often modeling. The third step in the greenway planning process presents a greater challenge. Environmentally significant land is often privately owned. A Georgia study found that 75 percent of environmentally sensitive features that researchers identified for inclusion in a greenway network were located on private lands (Ndubisi, DeMeo, and Ditto 1995). As a consequence, supporting greenways in the long term is a complex issue. Communities can rarely acquire an entire greenway network from private landowners and must often prioritize the preservation of key segments (Ndubisi, DeMeo, and Ditto 1995). Furthermore, common land conservation tools such as conservation easements are more complicated if the land is planned for public use and are expensive to use near urban areas where land value is high. A larger issue is that studies oriented toward mapping greenways often stop short of identifying concrete strategies for retaining them, much less monitoring the strategies’ effectiveness. Perhaps as a consequence of this, most greenway planning studies emphasize delineation, few examine implementation tools, and even fewer directly assess the connectivity outcomes of greenway projects. Seven of the thirty-eight greenway planning studies included in this review are broad theoretical or historical texts. For ease of summary, I divide the remaining thirty-one into three categories: delineation, implementation, and outcomes.

A delineation study is one that proposes or demonstrates a process for identifying or demarcating a greenway. Of the thirty-one studies relating to urban greenway planning and connectivity, the largest number (fifteen) fall into this category (Appendix Table A1). These publications are well distributed over time and geographically. The features that guide greenway delineation depend upon the goals of the future corridors. Common ecological selections include riparian areas, wetlands, prime farmland, and wildlife habitat (Arendt 2004; Ndubisi, DeMeo, and Ditto 1995) but may also consider attributes such as size, current condition, stormwater needs, development pressure, and other locally important features (Conine et al. 2004; Oh, Lee, and Park 2011; Rudd, Vala, and Schaefer 2002). Orsini and colleagues (2014), for example, delineated greenway corridors that connect roofs that have the potential to support rooftop agriculture. For greenway projects with stronger recreation or cultural goals, characteristics could include potential pedestrian destinations such as residential areas, commercial centers, or recreation facilities (Conine et al. 2004; Rudd, Vala, and Schaefer 2002). Several planning studies emphasize the importance of balancing natural and cultural features and not overlooking the importance of culture and recreation in ecologically successful greenway planning (Ahern 1995; Mugavin 2004; Taylor, Paine, and FitzGibbon 1995).

Strategies for connecting greenway corridors also depend upon the overall goals of the network, but generally assume either the “core strategy” of connecting a few large habitat patches or the “node strategy” of connecting many small habitat patches, either with each other or with the core habitat patches (Orsini et al. 2014; Parker et al. 2008; Rudd, Vala, and Schaefer 2002; Zhang and Wang 2006). Many of these methods are technologically demanding and not frequently applied in practice. Least-cost path analysis is a core strategy and is often used for networks intended to facilitate wildlife movement (Hong et al. 2013; Oh, Lee, and Park 2011; Teng et al. 2011). Least-cost path analysis, as applied to species mobility, is a Geographic Information System (GIS) analysis that identifies the most likely (i.e., “least cost”) path for an organism between designated source and destination green spaces. The technique identifies a network of greenways that are likely to facilitate species movement between habitat patches. Integrating field survey data on indicator species could improve upon the strategy for urban areas (Hong et al. 2013). Other greenway mapping strategies, typically of the node variety, are network analysis and graph theory. Graph theory conceptualizes the environment as a series of nodes (e.g., habitat patches) and the links (e.g., greenway corridors) that facilitate flow between them. Network analysis describes these elements and their interrelationships and helps users to identify a network that maximizes desired characteristics such as interconnectivity (Hellmund and Smith 1996; Zhang and Wang 2006). The strategy can, for example, identify a network that creates the greatest connectivity between nodes such as public parks with the smallest number of preserved greenways (Zhang and Wang 2006) or between nodes that are the most ecologically significant (Rudd, Vala, and Schaefer 2002). A benefit of the node strategy is that connections to other natural areas buoy the ecological function of small ecological features (Jennings and Lev 1986). More multipurpose greenway models must integrate recreation, ecology, and cultural values (Conine et al. 2004; Teng et al. 2011). Models with recreation goals may incorporate data on greenway use or demand rather than the ability of wildlife to traverse the landscape.

Greenway projects in highly urbanized areas are likely to have a combination of recreation, cultural, and ecological goals and to be space constrained. In these environments, the delineation process described above will be challenging. Planners can accomplish multiple greenway goals in such environments, but they must take advantage of existing linear features such as roads, railways, and rivers (Erickson 2004; Jim and Chen 2003; Teng et al. 2011; Toccolini, Fumagalli, and Senes 2006). The width of available green spaces and linear features in such environments often dictates the size and location of greenways rather than the landscape itself (Parker et al. 2008). This practical and opportunistic strategy helps to address urban land cost and availability challenges (Teng et al. 2011). A greenway study in urban Australia found, for example, that the most continuous corridor scenarios included riparian areas that were set aside as drainage when the area was originally developed (Parker et al. 2008). Erickson’s study of Milwaukee and Ottawa further supports this strategy. She finds that a preexisting connected open space pattern can be retrofitted for social and ecological needs and “is an enormous asset on which to build contemporary greenway corridors” (Erickson 2004, 220).

Similarly, a study on the potential for an ecological network in a highly developed city in China identified a road corridor, city wall corridor, and a river corridor as possible greenways. The authors suggest that these greenways could serve both recreation and urban wildlife dispersal functions and connect to other scales of green space (Jim and Chen 2003). An additional benefit of including road verges is that they are within the public right-of-way and have one landowner: the government. Public ownership simplifies the acquisition, management, and maintenance processes. Public land can also foster connectivity in areas where private property owners are not participating in actions to support urban wildlife (Viles and Rosier 2001). Roads are pervasive in urban and suburban areas and thus could create an extensive corridor network if a community can create a wide enough verge and strike a successful balance between the disturbances of traffic and management for habitat (Viles and Rosier 2001).

Studies that delineate greenways, almost half of the greenway planning research described in this study, rarely identify specific strategies for preserving the corridors. Some suggest incorporating greenways into the comprehensive plan (Conine et al. 2004; Ndubisi, DeMeo, and Ditto 1995), which does not necessarily ensure protection, particularly when the information on what should be added is vague. Local plans, such as comprehensive plans, that guide greenway development tend to lack specific implementation recommendations (Floress et al. 2009). As the authors of a greenway plan quality study note, “Because implementation is the ultimate goal of planning, this aspect is one of the greatest strengths of excellent plans” (Floress et al. 2009, 70). The majority of plans they examined (70 percent) did not meet their standards for “excellent” and provided little to no discussion of actual implementation or funding. Even plans rated as “good” presented greenways as “more of a ‘good idea’ and less of a feasible option” (Floress et al. 2009).

Prominent options for protecting delineated greenway corridors are land acquisition, land use regulation (Erickson 2004; Taylor, Paine, and FitzGibbon 1995), development review (Erickson 2004; Maruani and Amit-Cohen 2009), incentives, and education (Maruani and Amit-Cohen 2009; Ndubisi, DeMeo, and Ditto 1995). Milwaukee, Wisconsin, limits development by preventing the extension of urban services into greenway corridors (Erickson 2004). Adelaide, Australia, found success with a phased, section-by-section approach (Mugavin 2004). Ndubisi, DeMeo, and Ditto (1995) surveyed local public- and private-sector stakeholders in a greenway planning project and found that most preferred regulatory measures. The rules typically require development permit applicants to satisfy a series of regulatory requirements (Ndubisi, DeMeo, and Ditto 1995). But Ndubisi also suggests that the large percentage of potential greenway land that is privately owned indicates that incentives and education would be more promising strategies.

More greenway protection strategies exist than are examined in peer-reviewed literature. As with several of the aforementioned greenway design best practices, implementation measures are quite often recommended but rarely examined for connectivity outcomes. As a result, the outcomes of many greenway implementation measures are unknown or unclear. It is difficult to tell where and how well they work, much less whether they are successful in creating greenways with the design characteristics that best support connectivity.

An implementation study is one that describes a specific strategy or tool for creating, protecting, or maintaining a greenway. Thirteen studies fall into this category (Table 3). In four of them, implementation is a relatively minor component. They identify or suggest implementation tools but do not examine them in any depth. Three are historic case studies that identify implementation tools and their outcomes indirectly. Viles and Rosier (2001) do directly discuss the benefits and challenges of roadways as an urban corridor strategy, but they suggest rather than assess the tool. The remaining eight studies are designed specifically to examine the effectiveness or potential of a tool that could implement greenways. Seven of those examine just one tool: conservation subdivision regulations. The eighth examines the success of riparian corridor protection policies in Israel. But while eight studies examine implementation tools, only half of those use ecological measures to understand connectivity outcomes.

OPEN IN VIEWER

Table 3.

Greenway Studies with Implementation Conclusions or Implications.

Study (Author)	Date	Location or Scale	Implementation Tool(s)
Environmentally Sensitive Areas: A Template for Developing Greenway Corridors (Ndubisi, DeMeo, and Ditto)	1995	Walton County, Georgia, USA	Regulatory measures, tax benefits, education, and land acquisition
Greenbelt to Greenways: Four Canadian Case Studies (Taylor, Paine, and FitzGibbon)	1995	Ottawa, Ontario; Calgary, Alberta; Saskatoon, Saskatchewan; and Toronto, Ontario, Canada	Land purchase, development freeze, partnerships with land owners, local zoning, and land use controls
How to Use Roads in the Creation of Greenways: Case Studies in Three New Zealand Landscapes (Viles and Rosier)	2001	North Island, New Zealand (three sites)	Road verges
Linked Landscapes: Creating Greenway Corridors through Conservation Subdivision Design Strategies in the Northeastern and Central United States (Arendt)	2004	Neighborhood scale, USA	Conservation subdivision
Adelaide’s Greenway: River Torrens Linear Park (Mugavin)	2004	Adelaide, Australia	Sector-by-sector (segment-by-segment) incremental approach
The Relationship of Historic City Form and Contemporary Greenway Implementation: A Comparison and Milwaukee, Wisconsin (USA) and Ottawa, Ontario (Canada) (Erickson)	2004	Milwaukee, USA; Ottawa, Canada	Environmentally oriented development review, land acquisition, intergovernmental cooperation, leadership, and advocacy
A Framework for Understanding Conservation Development and Its Ecological Implications (Milder)	2007	Neighborhood scale, USA	Conservation subdivision
Conserving Biodiversity and Ecosystem Function through Limited Development: An Empirical Evaluation (Milder, Lassoie, and Bedford)	2008	Neighborhood scale, Eastern USA	Conservation subdivision
The Effectiveness of the Protection of Riparian Landscapes in Israel (Maruani and Amit-Cohen)	2009	National scale, Israel	Riparian protection policies
Conservation versus Cluster Subdivisions and Implications for Habitat Connectivity (Freeman and Bell)	2011	Falmouth, Maine, USA	Conservation subdivision
Barriers to Successful Implementation of Conservation Subdivision Design: A Closer Look at Land Use Regulations and Subdivision Permitting Process (Goçmen)	2013	Waukesha County, Wisconsin, USA	Conservation subdivision
Assessing the Environmental Merits of Conservation Subdivision Design (Goçmen)	2014	Waukesha County, Wisconsin, USA	Conservation subdivision
Balancing Housing Growth and Land Conservation: Conservation Development Preserves Private Lands Near Protected Areas (Mockrin et al.)	2017	Colorado, USA (thirteen counties)	Conservation subdivision

Note: Conservation subdivision studies are in bold.

Conservation subdivision strategies are only useful in undeveloped fringe areas but are prominent in the greenway planning literature. The seven studies included in this review suggest that the strategy has the potential to support connectivity (Arendt 2004; Ndubisi, DeMeo, and Ditto 1995). Furthermore, this literature has two remarkable qualities. First, four of the studies directly measure on-the-ground connectivity outcomes, a rarity in urban habitat connectivity research (Table 4). Second, these assessments comprise one of the few areas of strong overlap between urban planning and ecology literature. Studies of conservation subdivision developments are, for example, published in Conservation Biology and BioScience. It is unclear why research on this particular regulatory strategy has been able to bridge natural and social sciences research while others have not, but it can serve as a model. These studies suggest ecologists and conservation biologists recognize the impact and importance of land use planning tools.

OPEN IN VIEWER

Table 4.

Conservation Subdivision (Implementation) Studies that Directly Measure Connectivity Outcomes.

Study (Author)	Date	Location	Sample Size	On-the-ground Connectivity and Network Outcomes Measured
Conserving Biodiversity and Ecosystem Function through Limited Development: An Empirical Evaluation (Milder, Lassoie, and Bedford)	2008	Eastern United States	Ten conservation and limited development subdivisions, three conservation subdivisions	Edge effect, perforation, fragmentation, off-site connectivity, and site conservation targets
Barriers to Successful Implementation of Conservation Subdivision Design: A Closer Look at Land Use Regulations and Subdivision Permitting Process (Göçmen)	2013	Waukesha County, Wisconsin, USA	Fifty-four conservation subdivisions	Percent of subdivision as open space, percent of open space within environmental corridors, percent of open space adjacent to other open spaces, and percent natural or wooded
Assessing the Environmental Merits of Conservation Subdivision Design (Göçmen)	2014	Waukesha County, Wisconsin, USA	Twenty-six conservation subdivisions and twenty-six conventional subdivisions	Area of open space, percent open space adjacent to vegetated environmental corridors, wildlife habitat, woodland wildlife habitat, neighboring open space, naturally vegetated neighboring open space, and neighboring woodland open space
Balancing Housing Growth and Land Conservation: Conservation Development Preserves Private Lands Near Protected Areas (Mockrin et al.)	2017	Thirteen counties, Colorado, USA	343 conservation subdivisions	Area of open space, percentage of perimeter length shared with other protected areas, nearest neighbor

A conservation subdivision is a type of cluster development that delineates residential open space based upon the location of environmentally sensitive and important features (Arendt 2004; Göçmen 2014). If designed to be continuous, these ecological spaces could form effective corridors. The strategy has the most promise in fringe suburban areas with high development pressure. As the region develops, each new residential area would add a segment to a growing greenway network (Arendt 2004), effectively preserving important landscape features that would otherwise be developed. Evidence suggests that conservation subdivision design provides more functional landscape connectivity than conventional development and cluster development that is not based on natural features (Freeman and Bell 2011; Milder, Lassoie, and Bedford 2008). However, it is usually not the function of conservation development to implement a greenway plan. The strategy adds green space in an ad hoc, piecemeal manner.

Evidence supports the ability of conservation subdivisions to create and preserve connections between protected open spaces. Preserved open space within the subdivisions is often located adjacent to existing wildlife habitats, environmental corridors, and other conservation subdivision open spaces (Göçmen 2014; Mockrin et al. 2017). Mockrin and colleagues examined 343 conservation developments in Colorado and found that 76 percent were adjacent to at least one protected area. Variations of conservation subdivisions such as limited development projects appear to result in less green space fragmentation, more connectivity, and create refuges for wildlife (Milder 2007; Milder, Lassoie, and Bedford 2008). But the strategy does not necessarily create functional habitat (Hostetler and Drake 2009) or corridors. With natural features such as farmland, recreation trails, and steep slopes competing for consideration as preserved green space, retaining a wide corridor with diverse, high-quality habitat that supports organism movement could be a challenge. Connectivity with large core open spaces could also be an issue, particularly as development does not always occur with order and adjacency. Short greenway segments that do not connect to larger protected areas would have little value for species movement. In addition, as Hostetler and Drake point out in their 2009 review, planners and developers often focus on the design phase of conservation subdivisions and deemphasize the impact of construction and postconstruction management. As previously discussed, the management actions of private landowners and the quality and characteristics of green space have a significant impact on the ability of all types of wildlife to use and traverse a landscape. Furthermore, while conservation subdivision research is the most advanced of the implementation tools examined in this review, Milder (2007) finds that more “empirical work is needed to evaluate…the conservation outcomes of all four conservation development techniques” (p. 766).

While literature on the success of other regulatory implementation options is thin, two possibilities are zoning and riparian or floodplain regulations. Overlay zones that follow designated linear features, for example, could protect greenway corridors in the subdivision and land development process. Riparian buffers with vegetation and management requirements could assist in creating stream corridors that facilitate wildlife movement (Jim and Chen 2003; Parker et al. 2008). A study of riparian corridor protection policies in Israel suggests that protecting river corridors through local planning and permitting has limitations (Maruani and Amit-Cohen 2009). More research is needed on the effectiveness of these strategies for creating corridors and the width, habitat quality, and core green space connectedness that results.

Several additional strategies with significant potential are not traditional regulatory or public ownership-based protection but collaborative approaches where private landowners or groups of landowners such as horticulture clubs or community groups manage sections of a corridor system (Colding, Lundberg, and Folke 2006). Conventional public land preservation and regulation reinforces a dichotomy between conservation land and exploitable or private land that is not helpful in supporting biodiversity (Colding, Lundberg, and Folke 2006). As several delineation studies note, effective corridors are likely to include both public and private lands (Parker et al. 2008; Rudd, Vala, and Schaefer 2002). One option is co-production, a cooperative process between a government entity like a city and local citizens. Through co-production, local residents can help to fill the gap between what the city can provide and what is needed. These initiatives can work on large scales. A watershed stewardship program in Portland, Oregon, for example, resulted in 800 contributing organizations and 130 community projects including installations such as green roofs and bioswales, many of them on private property (Shandas and Messer 2008). Supported by a strong public planning process and municipal support, collaborative management could help to create more functional and biodiverse greenways. Given the visibility, popularity, and community benefits of greenways, they may be ideal candidates for co-production.

The final category of greenway planning studies is outcomes, encompassing both monitoring and assessment. Monitoring is often underemphasized in the planning process but can help to “close the loop” on planning and ensure that planning processes and outcomes continue to improve. Few studies examine the outcomes of greenway or urban corridor planning efforts. The deficiency is in line with a general lack of monitoring and empirical evaluation of urban green space plans (Ahern 2013). Gippoliti and Battisti (2017) note, for example, that “monitoring program[s] testing the effectiveness of ecological network plans using a project cycle logic are totally lacking, at least in a large part of Europe” (p. 687). Ten articles and chapters identified in this review examine the connectivity or habitat results of existing, on-the-ground greenways or green space networks and can be classified as outcome studies. Four of them examine conservation subdivisions and have already been discussed (Table 4). They are recent and based in the United States. The remaining six studies (Table 5) are more geographically diverse; only one examines a greenway in the United States. But all are more than ten years old.

OPEN IN VIEWER

Table 5.

Greenway Planning Outcome Studies.

Study (Author)	Date	Location	Green Spaces Included	On-the-ground Connectivity and Habitat Quality Outcomes Measured
Greenbelt to Greenways: Four Canadian Case Studies (Taylor, Paine, and FitzGibbon)	1995	Ottawa, Calgary, Saskatoon, and Toronto, Canada	Four greenway systems	Qualitative: success in supporting greenway connectivity and in creating a network of public green spaces
Sustainability and Urban Greenways: Indicators in Indianapolis (Lindsey)	2003	Indianapolis, USA	Seven greenways	Canopy cover, frequency, and dominance of riparian tree species within the greenways
The Relationship of Historic City Form and Contemporary Greenway Implementation: A Comparison and Milwaukee, Wisconsin (USA) and Ottawa, Ontario (Canada) (Erickson)	2004	Milwaukee, USA; Ottawa, Canada	Two greenway systems	Qualitative: connections between greenways and other green spaces and recreation areas and amenities, the quality of the corridors
Adelaide’s Greenway: River Torrens Linear Park (Mugavin)	2004	Adelaide, Australia	One greenway	Qualitative: adjacent land uses and connecting tributary green spaces
Planning an Ecological Network of Xiamen Island (China) using Landscape Metrics and Network Analysis (Zhang and Wang)	2006	Xiamen Island, China	One greenway system	Edge density, mean patch size, landscape shape index, Euclidian nearest neighbor, connectivity, Shannon’s diversity index, and Shannon’s evenness index of the greenway system
The History and Future Directions of Greenways in Japanese New Towns (Yokohari, Amemiya, and Amati)	2006	Tsukuba Science City and Kohoku New Town, Japan	Two greenway systems	Qualitative: role of the greenways in a citywide ecological network

Note: Includes only studies that directly measure on-the-ground connectivity outcomes. Excludes conservation subdivision studies, which are shown in Table 4.

Of the six studies shown in Table 5, four are longitudinal, historical case studies that describe connectivity outcomes qualitatively and not as a major focus. Only two studies assess the habitat and connectivity outcomes of an existing, on-the-ground, greenway network. The more recent of the two, a 2006 article by Zhang and Wang, examines the landscape ecology of an existing and a planned greenway system. The goal of their study was not to assess the existing system but to understand the potential impact of the new plan and the utility of landscape metrics and network analysis. While this does not detract from the strength of their analysis, it does impact their recommendations, which look forward to the benefits of a new plan rather than closing the loop on the existing system.

A 2003 study more directly examines the results of a greenway system and is more similar to the types of outcome assessments that a community might conduct. Lindsey (2003) compares the goals and outcomes of Indianapolis’s seven greenways. A major goal of the greenways was to “Protect important wildlife and promote the conservation of open space, forest, and wetland areas” (p. 171). But the city did not systematically collect data on ecological indicators, so information on attributes such as canopy cover or vegetation diversity within greenways was unavailable. Indianapolis is not alone; few cities routinely collect ecological data (Berke 2007). In their study of the New York City region’s green space plan, Flores et al. (1998) note that their success in incorporating ecological knowledge into connectivity planning was “often limited by the availability of ecological information on the systems and functions of interest” (p. 307). A study of three New Zealand road corridors encountered similar challenges (Viles and Rosier 2001). The Flores study, however, goes on to note the existence of a “rich history of biological scholarship in some North American urban areas that is often glossed-over or underutilized in planning efforts” (p. 307).

Lindsey did make use of biological data collected through other studies. It was limited and covered only a few greenway segments but indicated that greenway ecosystems were degraded. Together, this and the general lack of environmental data suggest two things. First, greenways were not meeting ecological goals, and second, ecological conditions were not a particular emphasis of the Indianapolis Greenway Division, despite their stated conservation goal. Lindsey (2003) confirms that the data (or lack thereof) reflected the priorities of the Greenway Division. He notes that the Division sought to build user support for expanding the system by emphasizing recreation corridors over conservation corridors. While identified here as a monitoring and assessment deficiency, this is also an implementation issue. In scenarios where the multiple goals of urban greenways are competing, recreation objectives, the more visible of the two, can overtake conservation objectives. Erickson reports a similar shift with Milwaukee’s greenway system. The initial greenway corridors were protected to meet recreation and environmental objectives, but by 2004, recreation and active transportation were more important. This was despite the regional environmental quality goals that drive the broader corridor program (Erickson 2004). These types of shifts contribute to the erroneous view that greenways are merely trails.

Hundreds of communities are engaged in greenway projects, but no recent literature describes their ecological results, much less longer-term connectivity outcomes. None of the greenway studies that examine on-the-ground features use connectivity or habitat metrics comprehensively and all are over ten years old. With monitoring technologies constantly changing and new greenways being added and connected, literature from a decade ago has waning relevance to modern greenway planning. Research on conservation subdivision outcomes, however, does exist and draws upon a variety of connectivity and habitat metrics. Greenways have even more connectivity potential and would benefit from similar assessments.

Informal Green Spaces

An informal green space is a vacant site with unmanaged or spontaneous vegetation. Also known as wasteland (Bonthoux et al. 2014) or liminal green space (Rupprecht and Byrne 2014), the most common type is a brownfield, but wastelands also include vacant land and interstitial spaces such as railway and canal sidings. Due to their vegetation diversity and lack of management, wastelands support a variety of wildlife (Angold et al. 2006; Bonthoux et al. 2014). Research on the role that wastelands play in cities is still coalescing (see reviews by Bonthoux et al. 2014; Rupprecht and Byrne 2014), but current evidence suggests that wastelands support more species than managed urban green spaces such as parks and lawns (Bonthoux et al. 2014 and Rupprecht and Byrne 2014) and are sometimes biologically comparable to a city’s core natural areas (Foster 2014).

Large vacant lots and brownfields have characteristics that could make them effective stepping-stones, but this role of informal green spaces is understudied outside of transportation verges and local residents may find unmanaged areas undesirable or even unsafe (Burkholder 2012). Research does examine the role of abandoned lots and how shrinking cities can reclaim them for open space (Burkholder 2012; Frazier and Bagchi-Sen 2015). Frazier and Bagchi-Sen, for example, use landscape ecology metrics to inform a residential demolition strategy that increases green space connectivity. As the authors note, urban residential properties are typically too small (>0.1 acres) to be considered as part of a green space network (Frazier and Bagchi-Sen 2015). By aggregating them, the properties could provide more significant refuge to species passing through the landscape.

Green Roofs

Strong evidence suggests that green roofs also support urban habitat connectivity. Green roofs can be categorized as intensive or extensive. Intensive roofs have deep growing media and could be better described as rooftop gardens. They can support shrubs, trees, and even food production. Extensive roofs are lighter, less expensive, and more common. They have shallow growing media and most often a low-growing succulent called sedum. But extensive “biodiverse” green roofs are possible and can support grasses, flowers, or other small plants that can withstand the harsh roof environment. Because they are more common, most research addresses extensive green roofs. Studies of bats and arthropods conclude that green roofs are functionally connected to nearby ground-level green spaces (Braaker et al. 2014; Pearce and Walters 2012) and to each other (Braaker et al. 2014). As one would expect, connections between green roofs and surrounding features are most pronounced for high-mobility species such as bees and bats (Braaker et al. 2014; Pearce and Walters 2012), but species such as spiders are also functionally connected to ground-level vegetation (Braaker et al. 2014). Results suggest that green roofs function as nodes in urban ecological networks.

The green roof attributes that most contribute to their support of wildlife are site connectedness (Braaker et al. 2017) and vegetation diversity (Braaker et al. 2017; Pearce and Walters 2012). These characteristics matter more than the age, size, or design of the roof (Braaker et al. 2017). Furthermore, the connectivity role of green roofs can be augmented if green roofs are planned as part of an ecological network (Braaker et al. 2014). While field-based studies of intensive green roofs are absent, models suggest that they too can serve as stepping-stones or nodes in a green infrastructure network (Orsini et al. 2014). Orsini and colleagues (2014) argue that rooftop gardens designed with appropriate shelter and vegetation can support beneficial species such as pollinators. They go on to map out a corridor system that connects rooftops with garden potential (i.e., flat) to each other and to core green spaces at the urban periphery using ground-level features such as street trees.

Stepping-stone Characteristics that Enhance Connectivity

Evidence suggests that planners can support biodiversity by creating a series of stepping-stone habitats that help organisms to traverse the landscape and to access core habitat patches. Actions and characteristics that will augment the function of stepping-stones are identified below and discussed in the section to follow. Several of the characteristics are similar to those that support species movement through greenways (e.g., diverse, native, structurally complex vegetation), but others (e.g., taking advantage of informal green spaces) are distinct.

Stepping-stone planning recommendations:

In more developed areas, create larger stepping-stones at closer intervals.

Stepping-stones best facilitate the dispersal of species that are able to traverse some distance through an urban area (Leidner and Haddad 2011; McFrederick and LeBuhn 2006). Corridors are continuous, but stepping-stones are surrounded by other land uses. The ability of an organism to migrate from one small habitat patch to another therefore depends upon its ability to move through the intervening landscape. For flying organisms, the distance may not be limiting. Urban birds forage in tree patches as far away as 85 kilometers (Hostetler and Holling 2000). But terrestrial organisms move more slowly and must pass directly through nonhabitat areas. A reasonable dispersal distance for small mammals from patch to patch, through a corridor, or in the open, for example, is just one kilometer (Bennett 1990). Some put that value far lower, at 100 or 200 meters (Hüse et al. 2016). Baum and colleagues (2004) suggest that both stepping-stones and corridors support high levels of connectivity when the intervening land—the matrix—does not provide significant resistance to species dispersal. In environments where the matrix is hostile to the species of concern, corridors are a better choice. But if an animal can successfully traverse the spaces between patches, stepping-stones may be sufficient. The result suggests that stepping-stones should be closer in environments where the matrix is difficult to traverse, such as an urban center, but farther in suburban areas, which are greener and more generally supportive of animal movement. A second implication is that stepping-stones are likely to be unsuccessful at facilitating movement in environments that are hostile to wildlife. In these areas, the structural connectivity afforded by corridors is necessary. Planners should consider the matrix and mobility characteristics of local species of concern and strike a careful balance with refuge locations.

Area of green space is also important and has a strong positive association with bird species richness (Carbo-Ramirez and Zuria 2011; Fernandez-Juricic and Jokimaki 2001; Schutz and Schulze 2015; Strohback, Lerman, and Warren 2013). Minimum thresholds for a successful stepping-stone are commonly set between 1,000 and 2,000 square meters (Palmer et al. 2008; Rudd, Vala, and Schaefer 2002), although habitats smaller than 2,000 square meters may support mainly disturbance-tolerant species (Palmer et al. 2008). Very small green spaces are predominately edge habitat, which even in optimal vegetation conditions facilitates predation and disturbance (Fernandez-Juricic and Jokimaki 2001; Schutz and Schulze 2015). So while sensitive species may be attracted to small urban parks and gardens, use of the areas could lead to increased mortality (Carbo-Ramirez and Zuria 2011). This result indicates that a patch of habitat can be “too small” to serve as a refuge and that every effort should be made to create and retain larger green spaces in the development process and to increase the size of very small green space patches (Palmer et al. 2008). This also impacts stepping-stone placement. Small patches must be placed where organisms from larger core green spaces can continually repopulate them (Rudd, Vala, and Schaefer 2002). More ecological studies are needed to understand how successfully small green spaces support wildlife (Carbo-Ramirez and Zuria 2011) and the appropriate size for functional urban stepping-stones in different environments and for various species of concern.

Another important characteristic of stepping-stones is vegetation diversity. Evidence suggests that increasing vegetation diversity and complexity enhances species diversity in small green spaces (Belaire, Whelan, and Minor 2014; Fernandez-Juricic and Jokimaki 2001; Palmer et al. 2008). Bird communities appear particularly affected (White et al. 2005). Vegetation diversity is also an important factor in the ability of green roofs (Braaker et al. 2017) and informal green spaces to support of wildlife (Angold et al. 2006). Angold and colleagues (2006) argue, for example, that planners could best support urban connectivity and biodiversity “by not hurrying to tidy up and redevelop brownfield sites” (p. 203). These results suggest that maximizing the plant diversity of stepping-stones can make them more effective at providing refuge services and consequently at facilitating animal dispersal. But while informal green spaces are appealing to animals, humans may not find them attractive. Local residents may see unmanaged land that could attract crime or pests and reduce property values. Some research exists on integrating informal green spaces into formal environments in a way that legitimizes them (Burkholder 2012), particularly in the field of landscape architecture. More is needed to connect this work to planning.

Native vegetation is also important. As White and colleagues (2005) put it, “remnants of native vegetation act as vital refugia of indigenous fauna in the urban landscape” (p. 133). Evidence indicates that native vegetation facilitates the movement of species through an urban landscape to a greater extent than exotics (White et al. 2005). A variety of studies suggest that public and private individuals can support stepping-stones by planting native vegetation in refuge areas (Beumer and Martens 2016; Hüse et al. 2016; Ikin et al. 2013; Leidner and Haddad 2011; Mason et al. 2007; Parker et al. 2008; White et al. 2005).

Planning Stepping-stones for Animal Movement

Stepping-stones are smaller and simpler to create, retain, and manage than larger linear features. Most cities already have a variety of them, in the form of small public parks, residential yards or gardens, and other vegetated areas. Literature on urban stepping-stones is still developing. Most (79 percent) of the forty-two stepping-stone publications included in this review are from the past decade. The median publication date of the twenty-seven planning-oriented studies is 2014, the most recent of the five categories. Many of these examine ecological value or potential. The most common research question is some variation of “does (or how does) this type of small green space support local biodiversity?” Ten studies go further to identify the role of the stepping-stones in supporting habitat connectivity (Table 6), but connectivity is rarely the subject of the analysis. Just two European studies (Braaker et al. 2014; Ioja et al. 2014) directly examine the role of the focal green space in connecting urban habitats.

OPEN IN VIEWER

Table 6.

Stepping-stone Studies with Connectivity Conclusions or Implications.

Study (Author)	Date	Location	Type of Stepping-stone	n	Connectivity Conclusions
Retrofitting the Suburban Garden: Morphologies and Some Elements of Sustainability Potential of Two Australian Residential Suburbs Compared (Ghosh and Head)	2009	Illawarra Region, Australia	Residential Yards (Gardens)	Two suburban neighbor-hoods (thirty-one traditional parcels and forty-one modern)	The rear garden (yard) areas are larger in the traditional suburban neighborhood, forming a green space corridor. This makes them capable of supporting environmental and ecological functions through connectivity of open spaces.
Scaling Up from Gardens: Biodiversity Conservation in Urban Environments (Goddard, Dougill, and Benton)	2009	N/A (literature review)	Residential Yards (gardens)	N/A (literature review)	Gardens (yards) and adjacent habitats form interconnected networks. A landscape ecology framework is necessary to understand the relationship between the spatial configuration of gardens and their biodiversity.
Are Small Greening Areas Enhancing Bird Diversity? Insights from Community-driven Greening Projects in Boston (Strohbach et al.)	2013	Boston, USA	Pocket parks, gardens, and street trees	Twelve greening projects, twelve urban sites, and six large parks	Small greening projects appear to be most valuable for urban biodiversity if they target preserving, increasing, and connecting existing green space.
The Potential of School Green Areas to Improve Urban Green Connectivity and Multi-functionality (Ioja et al.)	2014	Bucharest, Romania	Campus (public school)	461 public schools	School green spaces are an important element of the urban environment and serve as stepping-stones to support species movement. The role of these small, specialized green areas and their positive impact on connectivity suggests that they can be used to address urban green space deficits.
Potential Contributions of Green Spaces at Business Sites to the Ecological Network in an Urban Agglomeration: The Case of the Ile-de-France region, France (Serret et al.)	2014	Ile-de-France (Paris) region, France	Business Site Green Spaces	8,700 hectares of commercial site green space	Business site green spaces can support ecological connectivity within the region. Because of their number and strategic locations at the interface between rural and urban landscapes, these spaces could serve as stepping-stones that allow wildlife to enter urban areas.
Habitat Connectivity Shapes Urban Arthropod Communities: The Key Role of Green Roofs (Braaker et al.)	2014	Zurich, Switzerland	Green Roofs	Forty green roofs and forty ground sites	The green roof community composition indicates frequent exchange of individuals among roofs and a high level of connectivity between roofs. The integration of green roofs into urban spatial strategies has the potential to enable higher connectivity among green spaces.
Having Our Yards and Sharing them too: The Collective Effects of Yards on Native Bird Species in an Urban Landscape(Belaire, Whelan, and Minor )	2014	Chicago, USA	Residential Yards	Twenty-five transects	Groups of neighboring yards, in the aggregate, were more important for native bird species richness than environmental characteristics at the neighborhood or landscape scale. Migratory birds were observed on transects with more wildlife-friendly features.
Exploring the Production Capacity of Rooftop Gardens (RTGs) in Urban Agriculture: The Potential Impact on Food and Nutrition Security, Biodiversity and Other Ecosystem Services in the city of Bologna (Orsini et al)	2014	Bologna, Italy	Rooftop Gardens	One city with 3,500 flat rooftops	Flat rooftops in urbanized areas are numerous enough to support significant food production and to form a network of flying routes that would support urban biodiversity.
The Effects of Habitat Area, Vegetation Structure and Insect Richness on Breeding Bird Populations in Beijing Urban Parks (Huang et al.)	2015	Beijing, China	Urban Parks	Twenty-six urban parks	Connecting park vegetation with that of surrounding green spaces allows species to avoid impervious surfaces and enhances the effective size of green spaces. But even small urban parks support breeding birds and provide essential links within the urban green system.
BIMBY’s First Steps: A Pilot Study on the Contribution of Residential Front-Yards in Phoenix and Maastricht to Biodiversity, Ecosystem Services, and Urban Sustainability(Beumer and Martens)	2016	Phoenix, USA; Maastricht, Nether-lands	Residential Yards	512 front yards	Yard contribution to sustainable living is highly variable. Yet management decisions can foster small-scale habitats that function as ecological stepping-stones to the wider environment.

Note: N/A = not available.

One method for distributing small green spaces throughout an urban landscape is to use existing well-distributed public features such as school yards. A study of Bucharest, Romania, found that public school green spaces served as stepping-stones that increased the connectivity of the city’s green infrastructure (Ioja et al. 2014). Other studies suggest that college campuses (Alexio et al. 2014) and commercial green spaces (Serret et al. 2014) can also serve as refuges for urban wildlife and facilitate movement to more supportive habitats.

Another strategy is to take advantage of residential gardens. A large percentage of land in suburban communities is yard space, so it is unsurprising that conservation network models often include residential green spaces (Rudd, Vala, and Schaefer 2002; Parker et al. 2008). Studies of residential subdivision configurations find that many traditional gridded developments have mature trees and backyards that connect to form blocks of vegetated open space (Ghosh and Head 2009). But management is still an issue. Many yards are lawns, which are monocultures, and do not support urban wildlife. Intensive management can also stress and disturb animals (Beumer and Martens 2016). But if managed to support wildlife, gardens can serve as effective stepping-stones and form nodes in a broader conservation network (Goddard, Dougill, and Benton 2009; Rudd, Vala, and Schaefer 2002; White et al. 2005).

Gardens are commonly clustered in residential neighborhoods, particularly among adjacent properties (Hunter and Brown 2012). Due to their size, groupings form more ecologically significant stepping-stones than isolated gardens. A study of the Chicago area by Belaire, Whelan, and Minor (2014) found that groups of yards were more important for native bird species richness than neighborhood- or even landscape-scale environmental characteristics. Planners can support biodiversity by encouraging landowners to cluster gardens and to create gardens near adjacent habitat areas such as parks (Goddard, Dougill, and Benton 2009). Landowners can further facilitate habitat connectivity by creating gardens with supportive characteristics such as native vegetation in locations where a need for stepping-stones is identified (Belaire, Whelan, and Minor 2014; Beumer and Martens 2016; Ikin et al. 2013; Leidner and Haddad 2011; Mason et al. 2007; Parker et al. 2008).

Given the number of stepping-stones needed to provide connectivity in large urban areas, some level of co-production of urban refuges will be necessary. As with corridors, programs that encourage landowners in key areas to manage their land for habitat effectively expand conservation networks without the need for public land acquisition and management. In one of the few stepping-stone studies to examine an implementation strategy, Hunter and Brown (2012) suggest that social contagion is at play in the clustering of residential gardens and could be used as a mechanism to spread sustainable practices and create the groupings needed to support biodiversity. A variety of initiatives by nonprofit organizations support biodiversity on private property through programs that acknowledge homeowner efforts. The National Wildlife Federation, for example, consults with property owners on characteristics that support wildlife and certifies private landscapes that provide necessities such as food, water, and cover as “Certified Wildlife Habitat” (2018). However, an examination of the program by Widows and Drake (2014) noted “little to no connectivity between certified yards or between certified yards and larger areas of protected habitat” (p. 37). Connectivity was not the subject of their study, but the finding suggests that more purposeful planning will be necessary to create functional clusters of private habitat.

While no literature directly links them to connectivity, development regulations can also support stepping-stones. Local ordinances impact the type and extent of vegetation in new developments and could be used to retain trees and encourage native plantings. To manage stormwater, cities also increasingly incentivize or require decentralized stormwater facilities such as rain gardens and green roofs. Like stepping-stones, the effectiveness of these natural stormwater solutions may be dependent upon their critical mass and distribution throughout the landscape (Roy et al. 2008). The same network that manages stormwater could support wildlife. Research on green roofs confirms their role in urban connectivity (Braaker et al. 2014) and rain gardens, while smaller, typically include dense and diverse vegetation and could also play a role.

Stepping-stones are typically added opportunistically rather than through an ecological network planning process. But, like corridors, stepping-stones are most effective at supporting species when planned as part of a broader network (Braaker et al. 2014; Goddard, Dougill, and Benton 2009; Strohback, Lerman, and Warren 2013). Goddard, for example, highlights the role of a structurally disconnected, but functionally connected, network of residential gardens in supporting urban bird populations (Goddard, Dougill, and Benton 2009; White et al. 2005). The challenge for planners is that very few studies directly examine the role of stepping-stones in a network. This review identifies only four.

Two studies (Rudd, Vala, and Schaefer 2002 and Parker et al. 2008) examine stepping-stones as a step in the process of delineating habitat corridors. In Parker et al.’s (2008) study of urban Australia, the authors use rapid assessment methodologies to identify “conservationally significant stepping-stone sites” between two core habitat patches and the corridors that could connect them (p. 49). Rudd and colleagues (2002) use network analysis to examine the patterns of corridors necessary to link urban green space nodes in greater Vancouver, Canada. They concluded that smaller nodes provide important peripheral habitat and consider their most appropriate position in the habitat network. Other urban network analyses use similar strategies to highlight the role of nodes such as green roofs and public parks in providing functional connectivity in urban areas (Huang et al. 2015; Orsini et al. 2014), but they are limited by a focus on just one type of green space.

A 2016 study of Debrecen, Hungary, examined the role of a variety of discrete public and private green spaces in supporting urban biodiversity. The authors found that 65 percent of structurally disconnected urban habitat patches are within a reasonable dispersal distance and consequently close enough to be functionally connected. This creates the possibility for species to disperse between them and to connect to broader ecological networks at the urban periphery. They suggest that including ecologists in the process of urban development can facilitate these connections (Hüse et al. 2016). An earlier study by van Langevelde, Claassen, and Schotman (2002) modeled two conservation strategies for human-dominated areas: enlarging existing core green spaces and allocating stepping-stones between the sites. They found that both models increase the percentage of occupied habitat, but stepping-stones are more important where the amount of habitat available for animals is relatively low. In this situation, small green spaces increase functional connectivity and consequently the population of empty patches. Stepping-stones are less useful when good habitat is widely available (van Langevelde, Claassen, and Schotman 2002). To this, Rudd, Vala, and Schaefer (2002) add that smaller habitat nodes have higher rates of extinction and require close proximity to core habitat areas to replenish them. As fragmentation increases, smaller nodes become farther away, making dispersal difficult. Together, Rudd and van Langevelde suggest that stepping-stones are most useful in areas with a moderate level of fragmentation. When small nodes are too dispersed, they cannot be replenished and when good habitat is plentiful they are unnecessary. Research that could provide further clarification on the level of fragmentation at which for stepping-stones would be most helpful would benefit planners. Where, for example, are these spaces typically located along the urban–rural continuum? An answer to this question could help prioritize future stepping-stone research and green space funding and programs.

The Implementation and Impacts of Best Practices

Research in the fields of ecology and conservation biology suggest a variety of design strategies for creating functional habitat connectivity in urban and suburban areas. This review summarized the findings into two sets of best practices that could enhance the connectivity outcomes of greenways and stepping-stones, respectively. Among them, connecting features should be managed as habitat, rather than paths between habitats, and include vegetation that is as diverse, native, and structurally complex as possible. One way to accomplish this is to allow informal green spaces such as vacant lots to remain minimally managed. When complex vegetation is not feasible along the entire length of a greenway, patches of high-quality habitat can support species movement. Communities should also encourage private landowners to manage land along greenways and within their own yards and gardens as habitat to create functionally larger stepping-stones and wider corridors. Ecology research further highlights the importance of connecting greenways and stepping-stones to larger core habitat patches. To facilitate access to these spaces and among patches of green space, stepping-stones should be larger and closer in more developed areas, and greenways should be as wide as possible with narrow active transportation lanes that are located near the edge of the corridor rather than the center.

But while ecological research suggests that connecting features incorporate these characteristics, planning studies rarely examine their implementation and impact. Planning literature, for example, does recognize the need for connections between greenways and core habitat areas, but the emphasis results mostly in conceptual delineation studies. Research rarely describes communities’ success in creating actual, functional connections to larger patches. Vegetation management is also a frequent recommendation in both planning and ecology literature. But few studies examine associated planning strategies or the degree to which resulting green space networks incorporate diverse, native, and structurally complex spaces. Many greenways, for example, are mown for practical and safety reasons. For stepping-stones, spacing is also an unknown. Generally, closer is better. And to an extent, the spacing of stepping-stones depends upon the species of concern. But research by van Langevelde, Claassen, and Schotman (2002) and Rudd, Vala, and Schaefer (2002) suggests that there is a moderate level of development in which they could be particularly effective. Studies of communities that have adopted a stepping-stone approach could help to illuminate the effectiveness of the strategy in various conditions and create models that would encourage implementation.

Best practices that depend upon the actions of private landowners have great potential but are also understudied. When landowners vegetate the spaces adjacent to greenways, parks, and private gardens, they increase the effective size of conservation features. Programs that encourage property owners to create habitat, such as the National Wildlife Federation’s certified habitat program, enhance human contact with nature and add to wildlife space without the need for additional public land acquisition. But additional incentives may be needed to ensure that individual landowner decisions add up to an ecologically sound whole. For greenways, private management actions could work well because public ownership is most necessary for recreation access. Communities could feasibly acquire narrow greenway trail corridors and create wide vegetated buffers around them through private land management. Trail users would enjoy visual access to the space, and animals could make use of additional private greenway width with minimal human disturbance. These types of strategies are promising, but more research is needed to understand their feasibility, implementation, and level of success in supporting connectivity goals.

Corridor and Stepping-stone Implementation Strategies

A significant deficiency in urban connectivity planning research is implementation. It is particularly notable for greenways. The purchase of land in-fee is the simplest and most expensive way to create corridors, but there are a variety of regulatory, partnership, and private land management possibilities. Riparian regulations and greenway co-management, for example, have potential but are understudied. One tool that does appear regularly in urban connectivity planning literature is conservation subdivision design. Ecologists, not planners, have conducted about half of existing research on the strategy. While the strategy itself has significant limitations, the shared interest in the outcomes of a land use tool is promising. More research is needed on how land use and land management tools can create effective corridors and stepping-stones, from both a planning and an ecological perspective. Studies that examine the effectiveness of implementation strategies at creating connective features with best practice characteristics are particularly needed.

The Characteristics and Outcomes of On-the-ground Connectivity Features

The most significant gap in urban connectivity planning research is the lack of information on outcomes. Stepping-stones are a relatively new addition to planning literature, so the lack of outcome assessments is unsurprising. There are few, if any, planned stepping-stone-based networks to examine. But the absence of recent greenway outcome assessments is more remarkable. The problem is not a lack of subjects. If urban plans and designs have sought structural connectivity since the eighteenth century, there is much to study. Revisiting former research sites, for example, would be illuminating. Even short-term studies that examine whether greenway plans lead to landscapes that support wildlife or whether riparian regulations create ecologically sound corridors would add significantly to the field.

This review could identify no study of the ecological results of an on-the-ground greenway that has been published in the past ten years. Only one, from 2003, attempts to close the loop on greenway planning by comparing a project’s conservation goals to its outcomes (Lindsey 2003). That study did not examine connectivity and relied upon incomplete secondary data. There are two possible explanations for the lack of greenway outcomes in the literature. Planners—academics or practitioners—are not performing these studies or they are conducting the assessments for internal use but not publishing them. Either is possible, but a brief review of the presentations at greenway and trail conferences suggests that the former is more likely. As previously discussed, very few presentations address habitat connectivity. In addition, both planning and ecology literature do assess the configuration of conservation subdivisions, so connectivity outcome studies are achievable and can produce meaningful results.

Shifting Greenway Goals

Some evidence suggests that the goals of greenway planning have changed over time from conservation to recreation and active transportation (Erickson 2004; Lindsey 2003). But many greenway planning studies are over ten years old or are historical longitudinal studies, which makes it difficult to track shifts in goals. The lack of connectivity outcome studies suggests there has been at least some movement away from ecological objectives. If greenway goals have shifted from conservation to recreation or active transportation, the impacts on urban habitat connectivity could be significant. Recreation greenways have a different configuration than ecological greenways, with more mowed and paved area and less green space. The “green” aspects of trail or bike path networks will derive from improvements in areas such as energy consumption, air quality, and public health rather than natural green space. More research is needed on current greenway planning goals and how recreation can be balanced with ecology. A 2016 study, for example, suggests that different segments of the same greenway can be managed to meet different goals (e.g., biodiversity, recreation; Matsuba, Nishijima, and Katoh 2016). As greenways become more popular, and consequently more extensive and interconnected, they will have a greater capacity to support goals such as economic development and public health but also to connect urban habitat patches and core green spaces. Currently, there is no evidence of communities taking advantage of this opportunity. More research on both greenway and stepping-stone strategies could help communities to revalue urban habitat connectivity and adopt planning goals and strategies that support biodiversity and balance the needs of wildlife and humans.