Introduction
The global urban population is expected to increase by 2.5 billion people over the next 30 years (Seto, Güneralp and Hutyra 2012), following decades of continued urban growth (Seto, Fragkias, Güneralp and Reilly 2011). Consequently, urban landscapes have doubled in the last few decades leading to direct and indirect forest loss (van Vliet 2019). By 2030, global urban land cover is projected to increase between 430,000 km2 and 12,568,000 km2(Seto, Güneralp and Hutyra 2012). With urbanization comes a loss of natural habitats – especially tree cover – and increase in impervious surfaces, of low value to biodiversity (Nowak and Greenfield 2020). Thus urbanization generally correlates to losses in biodiversity (McDonald, Güneralp, Huang, Seto and You 2018, van Vliet 2019). However, many anthrophilic species continue to coexist with humans in urban environments (Magle, Hunt, Vernon and Crooks 2012, Møller 2012) and there is sustained advocacy, research, and planning for urban areas that promote wildlife-human coexistence (Apfelbeck, Snep, Hauck, Ferguson, Holy, Jakoby, Scott MacIvor, Schär, Taylor and Weisser 2020, Larson, Lerman, Nelson, Narango, Wheeler, Groffman, Hall and Grove 2022).
Although urban wildlife ecology as a discipline was advocated by Aldo Leopold in the 1930s it remains markedly understudied in academia (Adams 2005). The biodiversity outcomes of urban development have been varied, from local extirpation of undesirable species such as predators (McCance, Decker, Colturi, Baydack, Siemer, Curtis and Eason 2017) to multi-taxic rapid phenotypic changes in urban centers, implicating urbanization as a mode of evolutionary change (Alberti, Correa, Marzluff, Hendry, Palkovacs, Gotanda, Hunt, Apgar and Zhou 2017). In summary species’ responses vary within and among cities and contextualizing the mechanisms behind responses remains a key endeavor (Magle, Hunt, Vernon and Crooks 2012, McDonald, Mansur, Ascensão, Crossman, Elmqvist, Gonzalez, Güneralp, Haase, Hamann and Hillel 2020, Seto, Güneralp and Hutyra 2012).
We know that globally, large carnivores are one of the first groups extirpated, as we seek to “make safe” urban places for humans. One of the outcomes of extirpating large carnivores from urban environments is providing prey species with refugia from predation (Gallo, Fidino, Lehrer and Magle 2019, Møller 2012), often combined with substantial foraging subsidies for browsing and grazing herbivores (DeStefano and DeGraaf 2003). These anthropogenic changes to landscapes and wildlife communities have led to the perception of an “urban deer (Odocoileus spp .) problem” (Bowman 2011, Conover 1995, Rondeau and Conrad 2003) in wildlife management. Fifty-four percent of the world’s population lives in urban areas and is expected to increase to 66% by 2050 (Soulsbury & White 2019). Within cities, low-medium density housing areas the highest likelihood of urban wildlife interactions due to high species richness and low species extinction rates (Magle et al. 2016) and the greatest areas of greenspace and diversity of landcover (Loram et al. 2007). Yet people living within these low-medium density housing tend to react most negatively to human-wildlife conflict (Wine et al. 2015).
More than a problem however, urban wildlife is an unplanned scientific experiment that allows us to examine the roles of different forms of landscape heterogeneity on species’ space-use and resource selection. Urban areas are complex mosaics of impervious surfaces (buildings, roads, parking lots), natural or semi-natural greenspaces (parks), and heavily modified greenspaces (yards, gardens, golf courses), each offering different resources and risks for different species. Those resources are the outcome of social and economic drivers within the human population (Belaire, Westphal and Minor 2016).
One interesting outcome observed in urban ecology is the “luxury effect” wherein differences in affluence among neighborhoods generates differences in biodiversity. Evidence for a luxury effect dates back thousands of years, arising from Egyptian archaeological records, and continue through the Anthropocene (Leong, Dunn and Trautwein 2018). The luxury effect spans spatial scales, occurring both within and among cities (Magle, Lehrer and Fidino 2016), albeit inconsistently. Among 20 North American cities studied (Magle, Fidino, Lehrer, Gallo, Mulligan, Ríos, Ahlers, Angstmann, Belaire and Dugelby 2019) per capita income played a role in explaining vertebrate diversity in half; instead species richness was highly (negatively) correlated with urban intensity (Magle, Fidino, Sander, Rohnke, Larson, Gallo, Kay, Lehrer, Murray and Adalsteinsson 2021). Affluence is thus a proxy measure for biological properties associated with rich neighborhoods (Magle, Fidino, Sander, Rohnke, Larson, Gallo, Kay, Lehrer, Murray and Adalsteinsson 2021): low human density, energy subsidy, and especially greenness. Indeed the luxury effect is generally amplified in arid environments (Leong, Dunn and Trautwein 2018).
Most research on luxury effect uses species richness of plant or animal assemblages as the metric. For large mammals, individual behavior is a key mechanism explaining response to urban development (Honda, Iijima, Tsuboi and Uchida 2018), so we examine luxury effect from this different angle. We examine resource selection by highly abundant urban black-tailed deer (Odocoileus hemionus columbianus ; deer), a native to the western Nearctic including the Canadian province of British Columbia (BC). They are important prey for BC’s diverse carnivore population (Ballard, Lutz, Keegan, Carpenter and deVos Jr 2001) but the changing landscape has led to abundant urban deer populations. Predator persecution is an obvious culprit, but we suspect landscape change is an important driver. Deer select high-energy and high-nutrient plants as forage (Weckerly 1994) and are very sensitive to factors affecting the recruitment of young deer into the breeding population (Forrester and Wittmer 2013, Gilbert and Raedeke 2004). The abundant backyard gardens of urban and suburban areas in affluent neighborhoods (Larson, Lerman, Nelson, Narango, Wheeler, Groffman, Hall and Grove 2022) provide ample deer food, potentially allowing deer to breed more often and more successfully than in ‘natural’ (non-urban) landscapes.
However, even in natural systems the trade-off between security from predation and food resources is not well understood (Bowyer, Kie and Van Ballenberghe 1998), so how deer perceive risk in urban areas – and how they capitalize upon potential resource subsidies – remains unknown. Urban environments have been shown to impact wildlife behaviour, resulting in unique adaptations that differ from their non-urban counterparts (Schell, Stanton, Young, Angeloni, Lambert, Breck and Murray 2021, Wright, Adams, Stent and Ford 2020). Similarly animal behaviour and personalities influence the efficacy of behavioural tools for urban wildlife management such as hazing deterrents and culls (Honda, Iijima, Tsuboi and Uchida 2018). A better understanding of urban deer resource selection, avoidance, and spatial landscape use would help determine if the luxury effect is impacting individual deer behavioural, and if so, what are some of the proximal mechanisms for this effect. This information is also an important tool for suburban deer management, both in terms of minimizing the impacts on wildlife population processes as well as negative human-wildlife interactions (Klees van Bommel et al. 2020).
We quantified deer locations via satellite telemetry collars and estimated home-range sizes to better understand urban deer habitat selection. To evaluate our hypotheses­­­ about the luxury effect, we used resource selection function (RSF) analyses to make inferences about how deer use different landcover features (Boyce and McDonald 1999, Seidel, Dougherty, Carlson and Getz 2018). RSFs have been used extensively to assess animal movement patterns, response to novel anthropogenic features, and identify movement pathways (Abrahms, Sawyer, Jordan, McNutt, Wilson and Brashares 2017, Anderson, Turner, Forester, Zhu, Boyce, Beyer and Stowell 2005, Chetkiewicz and Boyce 2009, Darlington, Ladle, Burton, Volpe and Fisher 2022, Laforge, Brook, van Beest, Bayne and McLoughlin 2016, Stewart, Darlington, Volpe, McAdie and Fisher 2019). We examined the role of (1) vegetation productivity and tree cover, (2) residential lot size, (3) road density, (4) golf courses and public green spaces, and combinations thereof. We included variables measuring these features in a generalized linear model and examined effect sizes (β coefficients). We hypothesized that if the luxury effect was apparent, then residential lot size would show the significantly positive effect size. We also predicted that road density was a risk deer avoided, and that native (parks) and non-native (golf courses) forage sources would be selected, but with smaller effect sizes.