The results of this case study illustrate the wildlife conservation potential and limitations of urban green spaces as they exist in a major metropolitan area. Both of our comparison green spaces supported substantial native mammalian diversity, including diverse native mesopredators: coyote, bobcat, raccoon, and long-tailed weasel. However, neither green space appeared to support the full array of native mammal species of the region (Ingles 1965), and we recorded no occurrences of any of the regionally widespread native apex predators—black bear (Ursus americana) and mountain lion (Felis concolor)—or largest herbivores—e.g., elk (Cervus elaphus). Additional widespread native mammals—e.g., striped skunk (Mephitis mephitis)—also appeared to be absent. Thus, while urban green spaces can support substantial diversity, even a large, relatively unfragmented urban green space such as SESP is likely to support only a subset of the regional species pool (Aronson et al. 2016; Lerman et al. 2021).
These results also demonstrate the differences in wildlife value of relatively large urban green spaces subject to different levels of urbanization impact. Our cameras detected a greater diversity of native mammal species, including a more diverse and robust native predator guild and significantly higher levels of occupancy of several native species in SESP, which is larger, less fragmented, and situated amid more wildlife-friendly land uses than JWP. These differences were observed in both upland and riparian forest habitats. As these patterns did not obtain for the largest and most conspicuous species--coyote and deer—the wildlife impact of urbanization may not be apparent to the public and government officials charged with land-use planning and green space management.
Unlike native species, non-native species occupancy levels were largely similar in the two parks (Fig. 3). The sole exception was Virginia opossum, which showed higher riparian habitat occupancy at SESP, although its upland habitat occupancy did not differ significantly between parks. Opossum, eastern gray squirrel and cottontail rabbit activity levels were high in both parks. These results suggest that the range in urbanization of our two study sites had little influence on non-native species, and that some non-native mammal species are likely to be prominent faunal elements of even the largest and most ecologically intact urban green spaces.
Some native species were more sensitive to the impacts of urbanization than others. All native rodent species were more frequently detected at SESP, with Douglas squirrel and mountain beaver relatively common in SESP, but rarely present in JWP, and Townsend’s chipmunk and flying squirrel, detected at low levels in SESP and never detected in JWP. This pattern contrasts with the non-native eastern gray squirrel’s high occupancy levels in both habitat types in both parks. Eastern gray squirrel has proven to be highly invasive in a variety of locations where it has been introduced (e.g., British Columbia, United Kingdom, Italy), occupying diverse habitats, including all but the most intensely urbanized sites (Bruemmer et al. 2000; Bonnington et al. 2014).
Among the native mesopredators, bobcat occupancy was much greater in SESP, while coyote and raccoon occupancy patterns were mixed, with no consistent differences between the parks. These results parallel findings elsewhere that bobcats avoid more urbanized sites (Ordeñana et al. 2010; Lesmeister et al. 2015) while coyotes (Gehrt et al. 2010) and raccoons (Prange et al. 2004) are more tolerant of urban conditions (Bateman and Fleming 2012; Rodriguez et al. 2021). Differential sensitivity to urbanization across species resulted in substantially different predator communities in the two parks, with JWP characterized by greater presence of coyote and domestic cat, and lower presence of bobcat. This suggests that beyond the absence of apex predators, the predator communities and predator-prey dynamics may differ substantially in more and less urbanized green spaces (Crooks 2002; Fisher et al. 2012; Smith et al. 2018; Rodriguez et al. 2021), with possible ecosystem impacts, e.g., through trophic cascades (Jones et al. 2016; Wilson et al. 2020).
This study did not seek to identify the specific aspects of urbanization that were most important in producing the wildlife differences we observed. However, our results suggest the importance of urban features that are barriers to animal movement, such as roads and other hostile land uses around green spaces (matrix). The two largest and most mobile native species we observed, deer and coyote, were prominent in both parks, and in some cases were more active in JWP, while smaller native species—all of the native rodent species—were less active in JWP. Studies elsewhere (Rico et al. 2007; McGregor et al. 2008) indicate that small animals may be particularly vulnerable to the barrier effects caused by roads and other features of urbanization. The smallest non-native predator species we recorded—the opossum—also had lower occupancy in JWP, perhaps reflecting its vulnerability to barrier effects and road mortality (Glista and DeVault 2008; Smith-Patten and Patten 2008). The reduced presence of bobcat in JWP also may be, at least in part, a result of barriers, since this species was present in the riparian habitat of JWP where it can travel along stream corridors, but was absent in uplands where access requires road crossings. Other studies have found that bobcats favor riparian habitats for movement (Hilty and Merenlender 2004), avoid roads (Poessel et al. 2014), and are susceptible to road-related mortality (Tigas et al. 2002).
These results have implications for urban green space planning, management, and conservation. First, they provide clear and readily understandable evidence of the value of urban green spaces as wildlife habitat, and they illustrate the greater conservation value of larger, less fragmented, and more favorably situated green spaces as they exist in an urban area. By providing a measure of the wildlife impact of greater and lesser levels of urbanization, the comparative results of studies such as this can be used by local officials and planners to present accessible, site-specific information to the public, allowing informed decision making that includes biodiversity values, which are often not available and therefore are only included in decision making processes in general terms or left out altogether (Miller et al. 2009; Wilkinson et al. 2013; Apfelbeck et al. 2020). This is particularly important given that widespread occurrence of a small number of conspicuous urban adapter species—e.g., coyote and deer—which may obscure habitat value differences of green spaces for the public. Greater public awareness of less visible wildlife, and the charismatic nature of some of those species—e.g., bobcat and flying squirrel—may foster greater community interest and support for biodiversity conservation in land-use decisions (Stokes et al. 2010).
As one of the largest and most ecologically intact green spaces in the Seattle area (Stokes et al. 2014), SESP may approximate a best-case situation for urban wildlife, and our results offer a possible standard against which the potential and performance of other urban green spaces in the Pacific Northwest region can be assessed. As this study shows, even the most favorable urban habitats (e.g., SESP) do not support all of a region’s native species (Lerman et al. 2021). The faunal diversity of SESP, as opposed to the complete native diversity of the region, may be an appropriate standard of evaluation of the wildlife conservation value of other urban green spaces. Thus, SESP may serve as a regional urban reference site (Klaus and Kiehl 2021) for forested urban green spaces. A similar urban reference site function may be served by particularly high quality urban green spaces in other regions.
Clearly, even large green spaces such as SESP and JWP are too small, by themselves, to support viable populations of many of the mammal species that occur there. For example, the average territory size of a single female bobcat in western Washington is more than 300 ha (Link 2007), an area larger than SESP. Even small animals such as Douglas squirrel (Steele 1999) and mountain beaver (Arjo et al. 2007) are unlikely to occur at densities high enough to permit long-term viability of populations solely within the boundaries of SESP. This implies that the future persistence of these and similar species in green spaces depends on their continued access to and use of habitats beyond green space boundaries. To maintain wildlife presence in urban parks, land-use planning must address wildlife habitat value outside those parks. Given the negative impact of barriers on wildlife diversity suggested by our results, planning, design, and management of roads and other potential barriers to animal movement should receive particular attention.
The prominence of non-native mammals in a very large urban green space such as SESP highlights the importance of gaining a better understanding of the impact of those nonnatives on native species in urban green spaces. While the negative effects of domestic cats and dogs are well documented (e.g., Loss et al. 2013; Doherty et al. 2017), the impacts of eastern gray squirrel, eastern cottontail rabbit, and Virginia opossum on native fauna are less investigated (but see Bruemmer et al. 2000; Bonnington et al. 2014) and merit study, with the goal of informing appropriate management of these nonnatives.
Finally, our study identifies several species that may serve as practical indicators of native wildlife habitat quality of forested green spaces in urban areas of the Pacific Northwest. Possible indicator species are Douglas squirrel, bobcat, and mountain beaver, which are both sensitive to urbanization impacts and relatively easily monitored with camera traps. An additional potential indicator is the flying squirrel, a species primarily associated with mature forests in the Pacific Northwest (Holloway and Smith 2011; Smith 2012). We detected few flying squirrels; however, our camera deployments were not designed to maximize detections of this species. With cameras positioned appropriately—aimed at tree trunks and downed logs—(Boulerice and Van Fleet 2016; DS unpubl. data]), or using acoustic methods (Diggins et al. 2016), this species may be sufficiently detectable to serve as a useful indicator. Analogous indicator species (e.g., McKenzie et al. 2018) can be used to evaluate habitat value of urban green spaces in other regions with different species pools.