Achieving the Urban Tree Trifecta: Scenario Modelling for Salubrious, Resilient, and Diverse Urban Forests in Densifying Cities

Introduction

The critical role natural ecosystems play within urban environments has been increasingly acknowledged in both research and policy over the past 4 decades. Human dependence on forests and biodiversity is highlighted in the United Nations’ Sustainable Development Goal 15 (United Nations 2022), recognized as a pillar of health-promoting environments by the World Health Organization (World Health Organization 2017), and embedded into municipal sustainability plans such as the “One New York” plan in New York City (City of New York 2015), Vancouver’s “Urban Forest Strategy” (Vancouver Board of Parks and Recreation 2018), and the “First Step” urban forest management plan in Los Angeles (City of Los Angeles 2018). Indeed, natural landscapes are now being touted for their potential to provide nature-based solutions (NBS) to the complex set of challenges arising from climate change, whether social, economic, or ecological (European Commission 2015).

Urban forests are increasingly recognized as being salubrious (health-giving) for urban residents. Looking specifically at the urban forest as a form of nature, the range of benefits for both physical and mental health is robust. With respect to physical health outcomes, the presence of 10 street trees per block in Toronto, Canada had the same impact on overall health improvements as an additional $10,000 in yearly household income (Kardan et al. 2015). Likewise, individuals living in areas with the highest tree density within 1 kilometer of their homes in New York City were 23% more likely to report very good or excellent health in comparison to residents of areas with the lowest tree density (Reid et al. 2018). In terms of psychological health, the evidence is even more robust. Greater street tree density was linked to significant reductions in depressive symptoms among residents of socioeconomically deprived neighbourhoods across the Netherlands (Gubbels et al. 2016) and to reduced rates of antidepressant prescriptions among lower-socioeconomic status individuals residing in Leipzig, Germany (Marselle et al. 2020). More complex assessments of urban forests also demonstrate benefits. A recent evaluation of the 3-30-300 rule for urban forestry carried out in Barcelona, Spain found that individuals who could see at least 3 trees from their windows, had 30% canopy cover surrounding their homes (as assessed via a measure of residential surrounding greenness), and were within 300 meters of a park or other greenspace were 69% less likely to have visited a psychiatrist or psychologist over the prior year (Tonne et al. 2021).

Healthy urban forests also indirectly benefit human health by serving as a solution in climate-change resilience efforts. Resilient urban forests seek to reduce the harms to humans associated with high greenhouse gas (GHG) emissions (European Commission 2015) and to increase capacity to help cities recover from disturbances related to climate change (Huff et al. 2020). A systematic analysis of existing reviews that have examined the potential of various urban natural environments to serve as NBS found robust evidence for the substantial cooling effect of urban greenery, offsetting the increased risks of heat stress, chronic -disease exacerbation, and excess mortality in increasingly hot cities (van den Bosch and Sang 2017). Additional insights come from modelling efforts: a comprehensive urban greening scenario that integrated a 30% increase in street trees along with green roofs and vegetated impervious surfaces (such as carparks) across greater Manchester in the United Kingdom linked this “deep green” development with a reduction of 3 to 4.9 °C in surface temperatures, even in the context a high-emissions future (Carter et al. 2015).

Mitigation strategies are just as crucial as adaptation solutions, because they aim to stave off the most drastic harms of climate change by reducing greenhouse gas emissions or increasing carbon storage (European Commission 2015). One such strategy is urban densification, which has been linked to improvements in building energy efficiency and population-level shifts to lower-emissions transportation methods—such as walking, cycling, and the use of public transit—both of which can reduce per capita greenhouse gas emissions (Tonne et al. 2021). Conversely, more dispersed community designs have been linked to significantly larger GHG emissions, due both to a greater reliance on private-vehicle travel over larger distances (Ewing and Hamidi 2015) and to a higher amount of per-capita infrastructure, such as single-family homes rather than multi-unit dwellings (Norman et al. 2006; Hamidi et al. 2015). How urban forests perform with increased density is not well understood (Haaland and van den Bosch 2015). In terms of urban forest health, urban densification can also lead to a lack of permeable surfaces, insufficient soil volume to grow healthy trees, and less space for large patches of urban greenspace.

At the same time, many urban-planning policies that seek to increase density may threaten existing urban forests. In already developed urban cores, these threats occur in 2 principal areas. First, as lots that once held single-family homes or low-rise multi-unit dwellings are redeveloped into larger multi-unit developments, existing private trees are often cleared for the larger required footprint. Numerous cities have tree protection ordinances (TPOs) in place that regulate the replacement of lost trees, governing issues such as the size and species of replacement trees, which trees are subject to replacement, and how these rules are enforced (Lavy and Hagelman 2019). However, delving into the TPO for Vancouver, Canada demonstrates a number of factors that may result in replacement trees that provide fewer ecosystem services than the original components of the urban forest. In Vancouver, these issues include the following: (1) the rule applies to trees of at least 20 cm in diameter at removal, but replacements need only be 6 cm at the time of planting; (2) tree species are required to be selected “for disease resistance and hardiness” but not to be chosen with a changing climate in mind; and (3) the requirement as a whole can be met in lieu by paying a relatively small fee (City of Vancouver 2022a). In addition, even carefully selected and well-cared-for trees face additional pressures as their surrounding environments densify, including soil compaction due to land development, solid impediments to root proliferation (such as building foundations and roads), and toxic levels of heavy metals deposited in soil from high levels of vehicular traffic (Day et al. 2010).

Finally, many of the direct, short-term effects of climate change pose perils to tree health as well as human health. One recognized way to help address future unknown disturbances is to carefully and strategically increase the diversity of the tree population in terms of age, size, and species (Morgenroth et al. 2016). In British Columbia (BC), Canada, which includes the site of the current case study, recent years have been marked by a number of climate-related disasters, including devastating wildfires, a record-breaking heat dome, and severe flooding (Crawford 2022). Across coastal areas of the region, such floods are projected to become more common in the future, with simulation models indicating that the atmospheric rivers responsible for extreme precipitation will become 50% more frequent under projected warming scenarios (Espinoza et al. 2018). In addition to this increase in precipitation in the wet season, models indicate significantly less precipitation during the dry season, with the length of dry spells expected to increase by 22% over the historical average by the 2050s (Metro Vancouver 2016). Each of these climate-related disasters can negatively impact the urban forest. Wildfires can result in increased air pollution that can reduce the rate of photosynthesis or increase susceptibility to microbial and entomological stressors (Depietri et al. 2012). Heatwaves lead to crown dieback and mortality among both native and exotic species (Marchin et al. 2022). Flooding reduces nutrient availability in the soil (Depietri et al. 2012).

In order to develop and manage urban forests that are resilient in the face of these threats while providing the full range of nature-based solutions to nearby residents, we must be able to project into the future. In this way, we can understand precisely what climatic conditions the urban forest will face over the coming decades, select the tree species that will be able to thrive under these conditions, and detail the full range of ecosystem services that will be provided as a result. The overarching objective of this study is to outline a framework for creating future urban-forest scenario models by clarifying the impact of each stage of the scenario-development process on a model’s outputs and applicability to policy decisions. We report the results of a case study in a neighbourhood in Vancouver, Canada with both low building density and low canopy cover in 2020, envisioning a future for the area in which both of these attributes would shift substantially by 2050. In addition to these overall criteria, the future scenario sought to model, and thereby investigate, the implications of 3 additional planning decisions that reflect policy priorities in the region: (1) designing the urban forest in a manner that would achieve greater public-health benefits via the integration of “blue-green streets” as a specific form ofNBS; (2) prioritizing the inclusion of tree species that would be maximally resilient under a changing climate; and (3) integrating planting programs into the existing network of trees on both public and private property to ensure the highest possible diversity across the urban forest. This study was designed as a complement to Czekajlo et al. (in press), which offers a comprehensive description of the underlying scenario-development methodology, by elucidating the effects of embedding specific policy priorities.

Materials and Methods

Scenario Development and Assessment

The final phase of the scenario modelling process was simulating and assessing the impact of the planting process in the year 2050. Figure 1 outlines the inputs, processes, and outputs used to create, run, and analyze the scenario models. Scenario data assignment and model development were conducted systematically using a rules-based approach via R scripts (R Core Development Team 2017), including the following packages: ggplot2 (Wickham 2016), tidyverse (Wickham et al. 2019), rgdal (Bivand et al. 2022), raster (Hijmans et al. 2015), and sf (Pebesma et al. 2022). Additionally, some minor manual editing was performed using ArcGIS Pro (Version 2.9.2). Inputs used to create the 2050 scenario models included information about which trees were directly sourced from the baseline 2020 model (i.e. “existing” trees). Trees that were not selected for the 2050 models due to pre-determined mortality rates, based on rates provided by Hilbert et al. (2019) and species prioritization rules, were removed and considered “aged out”. For replacement and additional trees that were new in the 2050 models, quantities were determined through analysis of available spaces in the neighbourhood. Representative species were selected across small, medium, and large categories from the climate-adapted tree species list (Metro Vancouver 2019a), ensuring that diverse families and genera were represented. In alignment with the blue-green streets concept, trees that are known to tolerate the conditions in stormwater swales were prioritized. Trees from each size category were then selected for a site based on the location of the planting area. For example, we “planted” new trees to reach a goal of 2 trees per parcel in private yards, selecting small and medium species in recognition of the fact that most yards are unable to accommodate large trees.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Outline of process used to develop future scenario models.

To efficiently run the scenarios, all trees were “planted” in the modelled environment during the year 2020 so they would be of exactly the same age 3 decades later. Data on species’ average 30-year growth curves were unavailable or inconsistent across the range of online plant guides and databases appropriate for the study area’s climatic region, which included those produced by Oregon State University (OSU), the University of British Columbia (UBC), and the University of Florida (Breen 2022; University of Florida 2022; UBC Botanic Garden 2022). The OSU data provided the greatest extent of coverage about mature tree height for the species included on our planting list, however, so those details were used to estimate each planted tree’s height following 30 years of growth (estimated as 80% of mature height). Data on trees not included in the OSU database were collected from Kwantlen Polytechnic University (Kwantlen Polytechnic University 2015), Missouri Botanical Gardens (Missouri Botanical Gardens [date unknown]), and Plants For A Future (Plants for a Future [date unknown]). Because only mature tree heights were consistently provided within the consulted online plant databases, 30-year canopy spread (i.e., diameter) was estimated as 45% of average mature tree height. Details about growing trees of each species to 30-year sizes for this project, and further information about scenario data and models, are provided in Czekajlo et al. (in press).

Results

The results of the virtual planting process carried out for our case study offer indications of the potential of the urban-forest scenario modelling process to inform the development of salubrious, resilient, and diverse urban forests in the context of a changing climate and densifying populations. Table 1 lists the top 15 species planted across public and private settings, demonstrating significant differences in the composition of these 2 components of the overall urban forest. Most notable is the absence of coniferous trees as common planting choices in the public realm, and their relative abundance in the private realm, with only Prunus found as a top planting choice in both.

This distinction in planted species between the public and private realms is particularly important in light of a separate finding from the case study: resilience and diversity require an increasing reliance on trees planted on private property in order to achieve a substantial increase in canopy cover. At our study’s baseline, the total canopy cover within the sandbox was 6.6%, with 74% of this coverage coming from publicly maintained trees and 26% from private trees. By 2050, the canopy cover had grown substantially to 15.6% (which was just over half of the scenario’s stated aim of 30%) and the tree ownership ratio had shifted dramatically: 57% of coverage was now provided by private trees.

This change in tree ownership was accompanied by a shift in the average size of trees, both with respect to height and crown radius. In the baseline tree inventory, tall trees (those greater than 15 m) comprised 25% of the canopy, but this fell to 15% of the canopy in 2050, due primarily to the replacement of fully mature trees (at least 50 years old) with the moderately mature trees planted in the scenario (Czekajlo et al. in press). When sizing was standardized, the difference in canopy size was found to be associated with the chosen species, with most climate-adapted species tending to be smaller overall. In addition, there was a wider range of tree heights at baseline than in 2050, due to the greater variation in mature height than that seen at 30 years following planting. Conversely, crown radii were grouped more distinctly at 30 years, with a skew toward the lower end of the range and very few trees found in the higher end. Figure 2 captures the changes in the composition of the urban forest between baseline and 2050, with statistics on the distribution of public versus private trees, as well as the relative size of trees with respect to height and crown radius.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Urban forest composition comparison between baseline and Scenario 4. Tree sizes are based on mature tree height and include small (< 10 m), medium (10 to 15 m), and large (> 15 m) groupings.

Looking next at the resilience of the future urban forest in our case study, we found that a relatively large proportion of the trees planted at baseline would not be resilient to the shifting climate, as indicated by the color coding in Figure 3. The extent of this issue is greater for the trees planted in the public realm than in the private realm, primarily due to the predominance of new plantings of private trees, for which the list of climate-resilient species developed by Metro Vancouver was used to guide the selection.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Resilience of commonly planted trees on public and private property.

Finally, with respect to diversity, our case study indicated that a high level of diversity could be achieved via a strategic planting program. Even staying within the confines of the much smaller set of trees known to be climate-resilient, the implementation of a rule precluding both Acer spp. and Prunus spp. resulted in a significant decrease in the predominance of the families Sapindaceae and Rosaceae respectively, allowing us to fulfill the 30-20-10 planting rule in 2050, as depicted in Figure 4.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Urban forest family and genus diversity at baseline and 2050. Genus totals do not add to 100% because not all genus are listed in the diagram for legibility reasons.

Discussion

Our results demonstrate that scenario modelling can play an important role in developing evidence-based plans for developing resilient and diverse urban forests in densifying cities such as the study setting of Vancouver, Canada. We purposively selected a neighbourhood that is anticipated to undergo rapid densification over the coming decades and then developed a case study to achieve 3 distinct aims: (1) maximizing public-health benefits; (2) selecting tree species best suited to thrive in a changing climate; and (3) creating a holistic urban forest that integrates both public and private plantings to promote diversity. Using this approach, this study clarified the path toward achieving the urban tree trifecta, with lessons that can be applied by urban foresters and their municipal partners to increase canopy cover while advancing equity.

Looking first at the public-health benefits of the scenario via the lens of the urban forest as a form of nature-based solutions (NBS), our model was successful at advancing this component by more than doubling canopy cover (from 6.6% at baseline to 15.6% in 2050) and by creating “blue-green streets” that may provide multiple benefits, including outdoor shading and cooling (Yu et al. 2020; Jiang et al. 2021; Qiu et al. 2023). A greater amount of canopy cover both reduces average temperatures and increases the availability of shade, which is particularly important in cities such as Vancouver that are predicted to experience more frequent heatwaves in a changing climate. The health benefits of an expansive urban forest during heatwaves were demonstrated in a study conducted by the British Columbia Centre for Disease Control in Vancouver, which reported that individuals who died during the city’s 2021 heatwave were more likely to live in areas of higher building density and lower surrounding greenness (Henderson et al. 2022). Ensuring that tree canopy cover is incorporated in a manner that creates an inviting environment for physically active modes of transportation such as walking and cycling, which is precisely what our model’s blue-green streets provide, is embedded in Vancouver’s regional growth strategy specifically to “to promote positive mental and physical health” (Metro Vancouver 2022).

These streets also respond to calls in the literature to fully integrate urban-forest planning into the broader socio-ecological system of cities, paying careful attention to the intersection between NBS and other forms of infrastructure, and ensuring that trees are integrated in a way that provides ecosystem services both within individual neighbourhoods and across cities (Hansen et al. 2022).

The selection of trees is as vital as their placement in order for the urban forest as a whole to be resilient to the impacts of climate change. Urban forests possess the potential to mitigate the worst outcomes of climate change if they can thrive under rapidly, and often dramatically, changing conditions (Depietri et al. 2012), but climate change can lead to a reduction in both the level and range of ecosystem services provided if resiliency is not taken into account (Runting et al. 2017). Recognizing the importance of street trees to mitigation, numerous cities have made tree-planting programs a cornerstone of their sustainability plans (City of Los Angeles 2018; City of New York 2015; Vancouver Board of Parks and Recreation 2018). However, less attention has been paid to the management costs that are required to address issues of longterm vitality among newly planted trees (Roman et al. 2021). Our case study identified a greater lack of resiliency among public trees than private trees due to the impact of historical planting decisions, making adequate planning and funding for the maintenance of the existing urban forest an essential complement to tree-planting programs.

Similarly, the lack of diversity in the baseline urban forest, which was primarily composed of public trees, calls attention to a challenge that is common to numerous locations beyond Vancouver: few species currently planted are on the “highly suitable” list for Metro Vancouver, as demonstrated by the findings captured in Figure 2. These findings make it clear that some species of trees should be put on a pause list for planting, both because of their unsuitability in projected future climates and because they have been planted in such large numbers over prior decades, resulting in a less diverse urban forest that is more vulnerable to the effects of urbanization and to climate-related disasters such as droughts, flooding, and pest outbreaks.

Promoting high biodiversity in urban ecosystems is a cornerstone of preserving global biodiversity (Alvey 2006), and there is significant evidence that urban landscapes can achieve high biodiversity (Ordóñez and Duinker 2012; Sushinsky et al. 2012). Tratalos et al. (2007) argue that higher-density nodes in large areas of greenspace create a complex form that improves biodiversity, if designed well. In a modelling study comparing compact and sprawling development patterns, researchers found that compact developments had comparatively low ecological impact, but that this “depends crucially on maintaining high-quality interstitial greenspaces between high density developments” (Sushinsky et al. 2012, p. 407).

This final point also draws attention to the importance of public-private partnerships, especially in light of the shift in trees planted on public property to those on private property that our results demonstrate is essential to achieving anything close to the City of Vancouver’s overall canopy coverage aims. Although in the private realm, it’s critical that these trees be selected from our list of identified climate-resilient species, which can only be accomplished via a public awareness campaign accompanied by the provision of saplings of these species to private property-holders. Such an effort is already underway in London, England, where the Mayor’s Trees for London program is working with a nationwide greenspace advocacy group to provide packs of 50 carefully selected 2-year-old saplings at no cost to organizations such as housing associations and schools, along with a detailed planting guide that covers issues of aftercare and long-term maintenance (The Conservation Volunteers 2022). As the inclusion of aftercare instructions acknowledges, supporting private trees immediately after planting is vitally important, particularly to offset the increased vulnerability of newly planted trees to both drought and flooding. Maintenance issues were also one of the largest concerns expressed by residents who were invited to participate in The Greening of Detroit’s street tree-planting program, along with frustrations about the limited opportunities to select trees they found aesthetically pleasing and a good fit for their properties (Carmichael and McDonough 2018). In light of the fact that achieving the diversity aims of our urban tree trifecta will necessarily constrict the types of trees available to plant on private property, this last finding further highlights the importance of both broad-based public education campaigns and more detailed guidance for participating planters.

As with any modelling study, this effort has a number of limitations. Principal among these limitations is the fact that any model is only as robust as its underlying inputs, and our inputs relied on a number of assumptions. These include the extrapolation of baseline species based on what could be visually assessed via Google Street View in a sample of 100 randomly chosen locations across the case study area, rather than an actual inventory of all trees on private property. For future efforts, this limitation could be overcome by the incorporation of a survey sent to neighbourhood residents asking them to count and identify the trees on their property or by the integration of a citizen-science approach in which individuals are trained to conduct ground-truthing of species data by taking photographs of private trees. Alternately, novel machine-learning approaches may successfully detect more than 70% of all public and private trees identified via ground-truthing (Lumnitz et al. 2021), allowing for less time-intensive assessments. Other assumptions included the estimated height and crown spread of planted trees following 30 years of growth, for which we integrated mature size data from a reliable database but necessitated a generalized formula to calculate 30-year sizing across all species due to limited and inconsistent existing data and literature. With the increase in open, online tree inventories, future work could incorporate tree allometry and growth measurements, per species and urban environment, for improved model robustness.

For example, incorporating climate impacts such as drought, increases in potential insect and disease infestations, and extreme weather conditions into tree species selection was done through the use of a regional climate-adapted tree species list. Relying on such dated information, even ifrecently produced, is another limitation. Increasingly, climate impact information is being updated at a rapid pace, and future models could incorporate more recent impact assessments. Robust models also require consideration of cooperation among departments at the municipal and regional levels (i.e., the development of blue-green streets can only be carried out successfully when urban foresters work hand in hand with transportation planners and civil engineers). Although included, this consideration was somewhat limited, focusing primarily on the potential for planted trees to provide sufficient shade along blue-green streets. Other factors to examine would be the potential risks to pedestrians and cyclists associated with falling leaves and branches; conflicts with below and aboveground infrastructure; and the impact of land-use types, such as single-family vs. multi-family housing, on the appropriate selection and maintenance of trees. The current effort was aimed at sparking a conversation with local governments about the potential benefits and limitations of a range of future urban forests with the notion that future iterations would incorporate local governance measures more concretely.

Conclusions

As the case study described here demonstrates, it is possible to achieve the tree trifecta of a salubrious, resilient, and diverse urban forest even alongside rapid densification and within the constraints imposed by historical planting choices. Reaching this goal will require the creation of strong collaborative relationships between urban-forest managers and a range of other municipal officials, as well as the development of novel public-private partnerships to plant new trees that prioritize both short-term community education and long-term maintenance. In addition, although much of the required effort will take place at the municipal or regional scale, factoring in variation that exists at the neighbourhood, block, or lot level is critical, as is laying out a clear path for reaching the time- and density-based goals common to urban sustainability plans. Different neighbourhoods will have varying extant urban forests, as well as distinct strengths and capabilities in terms of growing their local urban forests. If cities plan urban forests at the neighbourhood scale with the future climate in mind, then species diversity can be tailored appropriately and ecosystem services assessed more accurately. These characteristics must be accounted for to avoid exacerbating existing inequities in canopy coverage that result in a lack of environmental justice from the many ecosystem services the urban forest provides.

Conflicts of Interest

The authors reported no conflicts of interest.