Assessing climate risk to support urban forests in a changing climate

Globally, cities are planning for resilience through urban greening initiatives as governments understand the importance of forests in improving quality life and mitigating climate change. However, persistence ecosystem benefits they provide threatened by change, systematic assessments causes tree dieback mortality environments rare. Long-term monitoring studies adaptive management needed to identify prevent change-driven failures mortality. Research when coupled with forecasting will enable incorporate change into forestry planning. Future scenarios which resilient or decline depend on actions we make today. The is a key element across globe. Urban services well other nature-based solutions 4.2 billion people living cities. continue do so effectively, need be able thrive an increasingly changing climate. Trees vulnerable extreme heat drought events, predicted increase frequency severity under Knowledge species' vulnerability therefore, crucial ensure provision desired benefits, improve species selection, maintain growth reduce mortality, stress forests. Yet, We reviewed state knowledge globally, finding very few frameworks that detection impacts no long-term assessing direct driver effects remain poorly understood quantified, constraining ability A nivel mundial, las ciudades están expandiendo áreas verdes medida que los gobiernos comprenden la importancia de bosques urbanos para mitigar el cambio climático y mejorar calidad vida ciudadanos. Sin embargo, supervivencia servicios ecosistémicos brindan se ven amenazados por actualmente, son muy raros estudios sistemáticos sobre causas muerte árboles urbanos. Se necesitan monitoreo largo plazo gestión adaptativa identificar prevenir mortalidad en provocada climático. Dicha investigación monitoreo, combinados con predicciones clima, permitirán efectos adversos del través planificación forestal urbana. Los escenarios futuros sean resilientes o estén declive dependerán acciones realicemos hoy. More than live areas, represent ~3% Earth's land area (Liu et al., 2014), 2050, it this number 6.6 (~70% global population) (UN, 2018). Within cities, comprise trees, shrubs associated vegetation, soils fauna variety settings including streets, residential park woodlands green belt vegetation (Miller 2015). numerous such mitigation, reduced stormwater runoff, biodiversity conservation improvement human health (Keeler 2019), solutions, roofs walls temperature energy savings (Alexandri & Jones, 2008). forests, both public private spaces, can also help mitigate adverse absorbing greenhouse gases storing carbon (Bastin 2019; Cimburova Pont, 2021). Both mitigation storage contribute meeting target limiting rise 1.5°C above pre-industrial levels (IPCC, these require healthy, functioning Climate change—that is, any over time, whether due natural variability result activity 2014)—is potential stressor affecting performance (Brandt 2016; Esperon-Rodriguez, Baumgartner, 2021; Ordóñez Duinker, 2014). assessment showed more 50% all plant present exceeding their current climatic tolerance mean annual temperature, and, proportion increase, potentially further jeopardising ecosystems (Esperon-Rodriguez, increases weather heatwaves, severe droughts floods, threaten (Hilbert Meehl Tebaldi, 2004; Staudhammer 2011; Yan Yang, 2018; Zscheischler These events widespread increased (Escobedo Roman 2014; Smith 2019) (Figure 1). Therefore, reasonable assume ongoing anthropogenic play role determining survival future composition Management activities, providing supplemental irrigation, may some negative (Van der Veken however, complex ecosystems, not only experience but face harsh conditions, growing situations limited soil volume nutrients, compaction extremes moisture availability, exposure de-icing salt cold climates, cause damage vitality trees (Day Bassuk, 1994; Gregory 2006; Mullaney drivers challenging essential environmental socio-economic losses sustained (Cimburova In general, often from slow accumulation many stresses time interactions among multiple factors (Czaja 2020; Franklin 1987; Hauer, Hanou, Sivyer, Koeser, Hilbert 2019). decisions along way either exacerbate ameliorate risks 2). activities (i.e. management) biophysical contributing Inadequate include unsuitable site poor nursery planting stock, inappropriate technique, insufficient preparation maintenance during establishment period, construction (e.g. new development redevelopment) vandalism (Hauer, van Doorn McPherson, Biophysical climate, pests diseases, herbivory browsing Ultimately, strongly interrelated because involve alteration factors, structure water nutrient availability Systematic proximate rare, especially those aims study were (1) highlight mortality; (2) review targeting reporting drivers; (3) propose recommendations changes being others. Thus, incorporating practices policymakers forest managers economic losses. date, topic remains understudied. affected gradual parameters rising temperatures precipitation patterns), pulse rapid one-off storms). Gradual less studied, research mortality—that inventory dynamics—is contrast, seem easier research—measure before immediately after, allowing attribution response specific event. cases, recognised 120 China, example, storms (Yan Similarly, Sweden Norway, low heavy snow, have been identified Pedersen Brun, 2013; Sjöman Slagstedt, Indeed, linked conditions after planting, occur frequently, necessarily attributed Furthermore, lacked data required detect attribute responses Santiago, Chile, distinguished using 12-year-period (2002–2014) concluded was influenced improper choice) alone 2016). To there paucity (see details about our literature search Methods S1). gap argue effective possible its guises shifts patterns) crown dieback, visible injury, defoliation growth, secondary change-induced diseases (Linnakoski As changes, become difficult excessive offset deficits, particularly regions supply (Pataki cost-effective options available air trees. sustainability depends identification cultivars suitable given location (McPherson Identifying genotypes likely tolerant climates option expanding palette within different locales Brandt 2017; Burley Esperon-Rodriguez McPherson Sadeghabadi Steenberg 2009). Initiatives Citree database (Vogt 2017) Germany, Vermont Tree Selection tool (https://vtcommunityforestry.org/resources/tree-care/tree-selection) United States, Species Guide (Hirons Sjöman, British Isles Which Plant Where programme (www.whichplantwhere.com.au) Australia science-based evidence inform selection. Iran, recently developed failure model (TFMmlp tool) provides decision support system artificial intelligence at risk wind storm) (Jahani Saffariha, Published valuable information limits Brune, McBride Laćan, Such thresholds based known distributions realised niches). Metrics envelopes degree days) hardiness zones) used selection For metrics useful, must policymakers, growers informed relation mainly base field trials assign particular zones classes, approach past does account Additionally, tolerances choice educate consumers. lags between production nurseries, local lifespan individual tree. Given comparatively rates promoting longevity, selections planned years even decades advance. Information climate-sensitive embedded practice arboriculture clearly collated accessible. available, reports grey anecdotal, restricted scale frequently scope terms numbers sites, usefulness broad applicability. around globe, councils municipalities) keep inventories, let accurately consistently record plantings established (van 2020). Dynamic inventories costly, thus, financial limitations collecting extremely perpetuate lack (Ramage 2013). Remote sensing canopy resolution do, offer cover (Hanssen 2021), though missing. Here, tools found two components missing literature: caused Further, via assessments. needs arise three main reasons. First, researchers resource yet undertake sufficient Second, signal, present, overwhelmed noise drivers. Finally, acute dispersed space complicating antecedent observed disentangle recommend conducting incorporates detailed inventories. Also, taking demography 2018), would entail removals (incorporates recognises sometimes cut down still alive) recruitment. Importantly, aid identifying successes develop plans (Venter emphasise implementing easy, methods document every stage metric evaluate success programmes (Roman collection should longitudinal, tracking surveys undertaken annually 2 (during e.g. <2 years) 5 (for established, mature >5 assess threats pests. Regular size height (trunk crown) stem diameter assessments, symptoms assessed leaf damage; Power, This last important, affect without killing them; go largely undetected Measuring traits attributes, bud burst, flowering colour, useful evaluating climate-driven phenology, damage. experimental functional (Esperon-Rodriguez Hirons Unfortunately, uncertainty decision-making Establishing standardised physiological turgor loss point, critical photosynthesis proline content) developing national international protocols databases means relate widest geographic areas. species, relevant history traits. Global regional trait TRY, AusTraits BROT 2.0 Mediterranean Basin (Falster Kattge Tavşanoğlu Pausas, 2018) tolerances. Field quantitative guidelines successful existing standardise addition ground-based remote Monitoring capture allow regular intervals dynamic inventories) events. Incorporating trends meteorological winter temperatures, summer high season precipitation) altered regimes test links rates. US Long-Term Ecological Network example how type conducted address questions non-urban (Mirtl periods status health, arising factors. Data could integrate removal permits ‘living inventories’ instead ‘static inventories’. community capturing detected removal. ‘Become Citizen Forester’ City Melbourne, Australia, citizens create resilient, healthy diverse landscapes. web visualise Melbourne's has locate send emails directing concerns updates each (melbourneurbanforestvisual.com.au). California USA's ‘Climate ready trees’ multi-partner promising underused tolerate types insights participate selecting appropriate species. species-level indicators observable verifiable, qualitative, decision-making, measurable, relatively short periods) rely (Tyler, 1996). When compare most risk. indices reflect safety margin. thermal tolerance, margin defined difference function occurs (Sunday good indicator capacity cope (Clusella-Trullas Gallagher Species' unknown. niche informative. case, (or climate) calculate niche-based margins (Box grow outside native ranges (Kendal sensitivity inferred solely interpreted carefully supplemented observations where possible. approaches differences analysis: Using (S) calculated. indicates much warmer drier), city exceeded calculated limit variable planted 3). Vulnerability index. index components: individuals species; rate period; period. considered lower individuals, inventory. Integration integration An aggregated shown 4). Details calculation accompanying examples Tables S1–S3. Determining challenging. Nonetheless, minimised niche, requirements; high-quality stock free defects exhibits proper form; best techniques, applying rigorous protocols, adapting site-conditions (4) preserve related weather, controlled conduct rates, heatwaves susceptibility context (Table Develop plots stratified stressors performance. Identify origin decrease probability niches, half least one currently Developing Box 1) requirements failure. country, zone, Köppen classification) volume, nutrients) sites Document this, precipitation, maximum temperature) Observe standard distances structures rooting follow considering Assess visual inspection issues, vitality; issue detected, examination confirm nature; defect confirmed risk, measured, recorded made corresponding actions, Authorities nurseries collaborate standards material. Australian standard, AS2303 Stock Landscape Use, specifies above- below-ground criteria (AS 2303, implement collaboration researchers, industry stakeholders protocols. adapted (Keller Konijnendijk, 2012; 1999). Protocols adaptable flexible meet goals needs, practitioners, protocol process simple users 2013) phase. Growth collected post-establishment phases examine survivorship stewardship outcomes real benefits. threat data, difficulties discerning previously. necessary factor Whereas improved underpin management, secure climate-resilient application predictive analysis simulation basis informing needed. Effective Holling, 1978; Walters, 1986) combination forecasts scenarios. Currently, planning, use decadal century timescales. computational purpose-built forestry. sustainable considerations adequate effort efficient plantings, alongside consistent, time. acknowledge, nevertheless, differ locations access resources. big accounting Hanssen Laumer disciplines—including arborists, landscape architects owners, social practitioners—as central partners co-production (Campbell transdisciplinary take socio-ecological perspectives (Steenberg 2017). recording cases might fundamental expansion estimates indices, proposed here, implemented forward. call governments, scientists general work together management. understanding requires shorter timescales better adapt pace longevity landscape, thank Leslie (USDA Forest Service, USA), Jakub Kronenberg (University Lodz, Poland), Shawn Landry South Florida, USA) Per Anker (Faculty Society, Norwegian University Life Sciences) thoughts contributions. MER, PR, SP MGT Leigh Staas (Macquarie University) funding Hort Frontiers Green Cities Fund, part strategic partnership initiative Innovation, co-investment Macquarie University, Western Sydney NSW Department Planning, Industry Environment contributions Government. DNB acknowledges Council Norway ENABLE project BiodivERsA COFUND 2015-2016 proposals. BW FORMAS (dia.nr 2016-20098). anonymous reviewers thoughtful work. opinions findings expressed paper authors construed official USDA Government determination policy. declare conflict interest disclose. SP, conceived article. MER wrote All contributed discussion content edited manuscript submission. Authors, excluding listed alphabetically. Supporting Figure S1 Relationship breast (GRDBH; N = 66 species). Ribbons indicate 95% confidence interval predictions linear [“lm”] Table Comparison (GR; cm/year) (MR; dead individuals/year) five city, values (VI), (VN), (VGR) (VMR). (VI) allows ranking C ranked vulnerable, respectively S2 cm/year), SMAT) (VMR) (VSSM). S3 Number (trees), DBH, SMAT, °C), SMAT (VGR), allowed Please note: publisher responsible functionality supporting supplied authors. Any queries (other directed author