Implications for Urban Greening and Preliminary Functional Classification: A Review of BVOC Emissions and Tree Carbon Sequestration in China

Abstract

Urban trees have a vital role in climate mitigation by enhancing air quality, providing thermal comfort, promoting biodiversity, and improving human well-being, thereby aiding in the achievement of Sustainable Development Goals 3, 11, 13, and 15. Nevertheless, in particular environmental circumstances, their release of biogenic volatile organic compounds (BVOCs) may facilitate the generation of ozone (O₃) and secondary organic aerosols (SOA), resulting in a context-sensitive compromise. This analysis consolidates data regarding carbon sequestration and BVOC emissions to guide the choice of sustainable urban tree species in China. A literature review encompassing works from 2002 to 31 August 2026 resulted in 132 entries: 127 peer-reviewed papers, four dissertations, and one preprint. Experimental measurements were emphasized, as they constitute the foundation for remote sensing inversions and model simulations; these techniques offer supplementary validation. The gathered information encompassed carbon sequestration, photosynthetic carbon absorption, BVOC emission rates and their composition, measurement foundations, and spatiotemporal dimensions. The data predominantly depicted urban and forest settings in China, constraining comparisons among analogous indicators and measurement parameters. The research revealed variability in carbon sequestration evaluations and suggested that overall BVOC emission rate by itself cannot define atmospheric hazards. Consequently, we suggest a provisional, trade-off-oriented, synergistic, and provisionally low BVOC-emission category categorization that incorporates carbon-related data, compound-specific BVOC traits, and environmental context. This investigative synthesis bolsters evidence-driven urban greening by connecting climate mitigation, air quality, ecosystem services, and human welfare; yet, it does not represent a verified mechanistic framework or conclusive planting guidelines. The reactions noted under particular Chinese meteorological circumstances, pollution metrics, management strategies, and developmental phases necessitate independent verification prior to their use in other areas or commercial forestry. Broader execution requires regional evaluations of ozone and secondary organic aerosol production potential, local nitrogen oxides, meteorological conditions, human-induced emissions, species characteristics, and sustained monitoring.

1. Introduction

Biogenic volatile organic compounds (BVOCs) are volatile non-methane hydrocarbons produced by plants through secondary metabolism in their leaves, flowers, fruits, roots, and other tissues. They include isoprene, monoterpenes, sesquiterpenes, and oxygen-containing derivatives [1]. Global BVOC emissions are estimated at approximately 1 Pg (10¹⁵ g) per year, nearly nine times higher than VOC emissions from anthropogenic sources, with forests being the primary source [1]. BVOCs are highly reactive atmospheric compounds that promote the formation of tropospheric ozone (O₃) and secondary organic aerosols (SOAs), thereby affecting air quality. Conversely, many BVOCs possess antibacterial, anti-inflammatory, and mood-regulating properties and participate in the global carbon cycle, linking plant ecology, atmospheric processes, and human health [1,2,3,4,5,6].

As shown in Figure 1, urban trees function simultaneously as carbon sinks and sources of atmospheric chemical effects. Through photosynthesis, they absorb and store atmospheric CO₂, release O₂, and sequester carbon. However, VOCs emitted by trees can react with atmospheric pollutants such as NO_x, promoting the formation of O₃ and SOAs, which, in turn, affect urban air quality and human health. Consequently, the ecological effects of urban trees encompass interrelated processes such as the carbon cycle, pollution formation, environmental regulation, and urban well-being. Therefore, tree species selection should balance carbon sequestration potential, BVOC emission characteristics, and associated environmental impacts to optimize benefits in carbon sink enhancement, pollution control, and ecological health.

Figure 1. Urban greening trees simultaneously serve as carbon sinks and exert atmospheric chemical effects.

Current urban greening practices emphasize carbon sequestration capacity while often overlooking the atmospheric costs associated with BVOC emissions. Existing studies typically examine BVOC emission characteristics, driving factors, and spatial patterns—or carbon sequestration and storage—primarily from a one-dimensional perspective. Consequently, interactions among these functions, their combined ecological and environmental effects, and the potential applications of different species for pollution reduction, carbon reduction, and promoting ecological health remain underexplored. This study systematically reviews the carbon sequestration capacity and BVOC emission characteristics of common urban landscaping trees, examines their interactions, and proposes a preliminary species selection framework based on reported measurement results. This study further constructs a comprehensive “theory–effect–application” framework that bridges basic research and practice to support synergistic pollution and carbon reduction under China’s “Dual Carbon” goals [7].

2. Literature Review Methods

2.1. Article Searching and Screening

Article Search and Selection

This comprehensive review considered publications released between 1 January 2002 and 31 August 2026. The main search log, dated 28 September 2026, documented five searches of the Web of Science Core Collection, which returned 1241, 316, 8, 3194, and 43 records, respectively, as well as two searches of China National Knowledge Infrastructure (CNKI), which returned 748 and 362 records, respectively. These entries totaled 4802 Web of Science records and 1110 CNKI records, resulting in a total of 5912 entries retrieved before deduplication. These figures represent the search counts for each individual search, not the number of unique publications. The main search log documented the absence of language restrictions and other filtering criteria applied during the search phases. Search terms and available settings are provided in the supplemental literature screening materials.

Records were managed using EndNote and Zotero. The reported screening pathway removed 845 duplicate entries, leaving 5067 records. Title screening excluded 2368 records, leaving 2699; abstract screening excluded 1713, leaving 986. An initial full-text assessment excluded 588 publications; a further eligibility assessment of the remaining 398 excluded 290, leaving 108 publications. Supplementary searches of Google Scholar, SpringerLink, ScienceDirect, and Wiley Online Library contributed 24 additional eligible publications. Consequently, the final set comprised 132 distinct publications: 127 peer-reviewed articles, four theses, and one preprint. Publication counts were distinguished from extracted datasets, species records, and duplicate measurements (Figure 2).

Figure 2. Principles utilized in data selection and analysis in the present review.

Eligibility requirements called for evidence related to tree carbon storage or accumulation, photosynthetic CO₂ uptake, BVOC emissions or composition, or their relationships to urban greening, atmospheric chemistry, and ecosystem services. Publications that dealt solely with environmental pollutants without relevant evidence regarding trees or that lacked sufficient information for the intended synthesis were excluded.

The first author conducted the initial screening and full-text evaluation. Co-authors subsequently reviewed the preliminary screening results and discussed relevance and the adequacy of the evidence. This was not an independent, duplicate screening by two reviewers. Incomplete specifications regarding some search terms and fields, supplementary platform search records, duplicate matching procedures, and publication-level exclusion criteria limited the assessment of full historical reproducibility and selection bias.

Figure 3 shows the annual publication output from the Web of Science Core Collection for the period from 2006 to 2026. This bibliometric analysis was conducted independently of the evidence synthesis. The period from 2006 to 2026 was selected to provide a unified, comprehensive perspective on publication trends in these two academic fields. Since 2026 is not yet complete, the interpretation of annual publication trends is limited to the completed calendar years 2006–2025; no inferences are made regarding the total number of publications for the full year of 2026.

Figure 3. Annual publication frequency of research on BVOCs and the carbon sequestration potential of trees cataloged in the Web of Science Core Collection database from 2006 to 2026 (data for 2026 currently unavailable).

2.2. Data Extraction, Source Verification, and Standardization

For each publication, the following information was extracted where available: publication identifier, taxonomic unit, study location, measurement method, environmental conditions, developmental stage, temporal and spatial basis, indicator definition, reported values, original units, and calculation procedures. Missing information was treated as unreported rather than inferred from another species, study, or measurement scale.

For quantitative interpretation, each record must have a traceable link to its literature source and the original table, figure, or calculation. Source identifiers imported into the compilation are not assumed to automatically correspond to reference numbers in this manuscript. Records with unresolved citation mappings or taxonomic identities remain unverified and cannot independently establish species-level classification.

BVOC records are distinguished based on compound coverage, reporting basis, and measurement conditions. Total reported amounts, the sum of selected compounds, and compound-specific measurements are retained separately. The total reported amount represents the compounds covered by the original analytical method and does not necessarily include all emitted BVOCs. Carbon records are distinguished based on carbon pools, accounting boundaries, assessment periods, and spatial basis.

Unit conversions require documented conversion factors and reproducible calculations. Values for which this information is lacking shall retain their original units and be interpreted within their original measurement context. Conversion to common units alone is insufficient to establish comparability.

Distinguish between independent supporting studies and literature source entries. Technical replicates, repeated seasonal measurements, and repeated citations of the same underlying study through reviews do not constitute independent studies. Where source counts exist only at the compilation level, these are described as source entries rather than independent study counts [8,9].

2.3. Categorization and Integration of Evidence

2.3.1. Assessment of BVOC Emission Classifications

The comprehensive BVOC classification encompasses solely observations with analogous chemical representation and a standardized foundation. Quantifications of specific chemicals or compound categories (such as isoprene, monoterpenes, sesquiterpenes, and oxygenated BVOCs) are preserved as distinct evidence and are not regarded as synonymous with total BVOC emissions. Consequently, observations of isoprene solely contribute to the evaluation of isoprene emission potential but are omitted from the computation of the overall BVOC threshold.

For species $s$ within study $j$, analogous repeated measurements are consolidated utilizing the study-level median:

$${\mathrm{E}}_{\mathrm{s}\mathrm{j}}=\mathrm{m}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{a}\mathrm{n}\left({\mathrm{E}}_{\mathrm{s}\mathrm{j}1},\dots ,{\mathrm{E}}_{\mathrm{s}\mathrm{j}\mathrm{n}}\right)$$

(1)

Subsequently, the representative value at the species level is determined as the median of the accessible study-level medians:

$$E_s=median\left(E_{s1},\dots ,E_{sm}\right)$$

(2)

Every autonomous study yields a singular cumulative value. Technical replicates, repeated sampling intervals, and seasonal assessments within a single study are not regarded as separate investigations. The application of inverse variance weighting was omitted due to the inconsistency of variance estimations with the documented sampling strategies.

Species-level values exhibiting positivity were subjected to log₁₀ transformation. In each analogous dataset, the 20th, 40th, 60th, and 80th percentiles were determined utilizing Excel’s PERCENTILE.INC function. The thresholds were subsequently reverted to the initial emission units:

$$Q_p=10^{PERCENTLE\left({\log }_{10}{E}_{s,p}\right)}, p=0.20,0.40,0.60,0.80$$

(3)

The five-category BVOC emission scale and corresponding emission intervals are presented in Table 1 [3,10,11,12,13]:

Table 1. BVOC emission level categories and corresponding emission rate intervals (μg·gDW⁻¹·h⁻¹).

BVOC Emission Levels	Range of BVOC Emission Rates (μg·gDW−1·h−1)
Very high	E > 32.93
High	17.08 < E ≤ 32.93
Moderately high	7.21 < E ≤ 17.08
Medium	2.52 < E ≤ 7.21
Moderately low	E ≤ 2.52

“E” indicates reported BVOC emission rates based on dry mass. Each cutoff value is assigned to the lowest emission category. These categories describe relative emission levels as reported and do not represent validated atmospheric risk thresholds.

Assessments performed at a consistent temperature of 30 °C and a photosynthetic photon flux density of 1000 μmol m⁻² s⁻¹ are deemed the most analogous. Information provided based on fresh mass, leaf area, stand level, carbon mass, or compositional hierarchy is preserved for descriptive objectives but is not necessarily incorporated into the official rate categorization without robust conversions. For species with numerous observations, the median is initially computed inside each separate study and then an equally weighted median is derived across all studies, thus averting the dominance of studies with extensive measurements on species-level estimations. Only isoprene measurements are preserved as an indicator of isoprene emission potential; yet, due to their inability to represent the entire BVOC composition, they are omitted from the overall BVOC threshold calculations. Crucially, emission categories serve merely as relative indicators of BVOC emission potential rather than direct proxies for atmospheric ozone or secondary organic aerosol hazards, since these risks are also affected by compound-specific reactivity, ambient NO_x concentrations, oxidant availability, meteorological factors, background anthropogenic emissions, and atmospheric transport [12,14,15,16,17,18,19,20,21,22].

2.3.2. Assessment of Tree Carbon Sequestration Classifications

In this review, carbon fixation refers to the assimilation of inorganic carbon into organic compounds through photosynthesis. Measurements based on net gas exchange in leaves are described as net photosynthetic CO₂ uptake, with specifications regarding the integration period and area basis. Carbon storage refers to the amount of carbon held in a defined carbon pool at a specific point in time, while carbon accumulation refers to the increase in that carbon pool over a given period. Annual changes in tree biomass carbon stocks are reported as annual biomass carbon increments. Carbon sequestration is used to describe the broader function of atmospheric carbon uptake and storage, while quantitative results retain their specific metric definitions and accounting boundaries. Biomass carbon increments and net carbon balance over the life cycle are explained separately.

Carbon-related observations were classified into five distinct indicator categories: annual biomass carbon increment per tree, life-cycle annualized net carbon balance, short-term photosynthetic CO₂ absorption, annual carbon accumulation per unit of stand area, and existing carbon stocks. For every observation, we documented the species or vegetation classification, indicator definition, raw data and units, study region, evaluation timeframe, spatial extent, and calculation methodology. Quantitative classification was executed among datasets that shared analogous indicator definitions, units, temporal and geographical dimensions, and accounting parameters. The values delineated for each tree, each unit of leaf area, and each unit of land area were not combined [23].

Where applicable, the annual biomass carbon increment per tree served as the principal species-specific metric. For biomass-related research, this metric was determined by analyzing variations in biomass carbon reserves over the evaluation timeframe:

$$A_C={\displaystyle \frac{C_{end}-C_{start}}{\Delta t}}$$

(4)

In this framework, $A_C$ denotes the yearly biomass carbon accumulation (kg C·tree⁻¹·yr⁻¹), whereas $C_{start}$ and $C_{end}$ signify the biomass carbon stocks at the commencement and conclusion of the evaluation period, respectively, and $\mathrm{\Delta }\mathrm{t}$ indicates the duration in years. This metric quantifies the annual change in biomass carbon stocks within the specified pool, rather than the overall carbon equilibrium of a complete ecosystem or life cycle. Stock variations are characterized as life-cycle accumulation solely when the initial research comprehensively addresses the complete life cycle.

In the absence of important measures, a comprehensive life-cycle net carbon balance can serve as an alternative annualized measure:

$$A_{LC}={\displaystyle \frac{B_{LC}}{L}}$$

(5)

In this framework, $B_{LC}$ delineates the net carbon equilibrium across the life cycle, considering variations in carbon reserves alongside emissions during the stages of establishment, maintenance, removal, and disposal, while $L$ signifies the associated service duration. Carbon and CO₂ values are differentiated and converted consistently where applicable. Annualized life-cycle values are documented independently from biomass accumulation due to differing accounting parameters.

When suitable measurements are accessible, short-term photosynthetic CO₂ absorption is determined using trapezoidal integration of the net photosynthetic rate at the leaf scale:

$$U_{{\mathrm{C}\mathrm{O}}_2}=10^{-6}{\sum }_{i=1}^{n-1}{\displaystyle \frac{P_i+P_{i+1}}{2}}\left(t_{i+1}-t_i\right)3600$$

(6)

In this context, $P_i$ denotes the net photosynthetic rate (μmol CO₂ m⁻² leaf s⁻¹) at time i, measured in hours, while $U_{{\mathrm{CO}}_2}$ signifies the total absorption per unit leaf area (mol CO₂ m⁻² leaf). Multiplying by 44 or 12 transforms this amount into grams of CO₂ or grams of carbon, correspondingly. The conversion to a ground-area basis utilizing the leaf area index is conducted solely when corroborated by source studies. These estimations denote photosynthetic absorption over the observation duration or a specifically extrapolated timeframe and must not be construed as annual total-tree net carbon sequestration. Stand-level accumulation is likewise designated for comparisons at the stand or community level. Current carbon reserves are not included in the classification of sequestration rates [24,25,26,27,28,29].

For each valid dataset, we used Excel’s PERCENTILE.INC function to calculate the 20th, 40th, 60th, and 80th percentiles from the raw values. The resulting descriptive percentiles and input-pool characteristics are summarized in Table 2.

Table 2. Author-calculated descriptive percentiles of carbon-related indicators, with input-pool definitions and input counts.

Indicator and Pool	Temporal and Spatial Basis	Unit	Source	n	P20	P40	P60	P80
G1. Cumulative increase in tree biomass carbon over the modeled life cycle	Modelled life cycle; individual tree	kg C·tree−1	[25]	32	2.951	13.276	53.588	557.194
G2. Annual biomass carbon increment per tree	Annual; species-by-DBH-class means	kg C·tree−1·yr−1	[24]	18	3.950	6.222	8.910	16.482
G4. Annual increment-based CO2 fixation estimate by forest functional type	Annual; unit stand area	t CO2·ha−1·yr−1	[29]	4	1.268	2.120	4.280	5.888
G5. Annual increment-based CO2 fixation estimate by stand age class	Annual; unit stand area	t CO2·ha−1·yr−1	[29]	5	3.110	3.654	4.214	5.000
G6. Annual increment-based CO2 fixation estimate by forest composition or dominant-species group	Annual; unit stand area	t CO2·ha−1·yr−1	[29]	6	2.900	3.560	5.840	6.760
G3. Daily photosynthesis-derived fixation per unit leaf area, summarized by landscape application type	Daily; unit leaf area; application-type means	g·m−2·d−1a	[27]	8	6.240	8.276	8.636	9.726
G7. Annual photosynthesis-derived CO2 uptake per unit ground area	Annual total extrapolated from seasonal measurements; per unit ground area	g CO2·m−2 (annual total)	[26]	2	4079.294	4191.888	4304.483	4417.077

The superscript “a” following the unit “g·m⁻²·d^−1a” in row G3 is a footnote marker; it is not part of the unit, nor does it represent “year”.

The input pool identifier corresponds to the supplemental carbon workbook. n is the number of numerical inputs used per row; it is not an independent study count, nor is it necessarily the number of species. All four percentiles within the same row use the same input values. The percentiles were calculated by the authors based on compiled inputs; they are not thresholds directly reported in the cited publications. G1 describes changes in cumulative biomass carbon and does not represent the net carbon balance over the life cycle. G4–G6 are estimates based on biomass increments expressed as CO₂ mass, rather than direct gas exchange measurements or annual carbon increments per individual tree. These three pools categorize forest information from the same source in different ways and do not provide independent species-level verification. ^a G3 uses Table 1 from Reference [27], which reports the mean values for eight landscape application types. This pool is used solely as supplementary descriptive evidence. For G7, the four percentiles were obtained by interpolating between two numerical inputs and do not represent four independent observations or a fully characterized species-level distribution.

Range endpoints, deduced values, and consolidated vegetation classifications are not regarded as autonomous species-level observations. Species-level assessments presented in the first study are valid only when the definitions and measurement foundations of the indicators are thoroughly recorded.

Principal indicators delineate species-specific classifications; photosynthetic metrics and stand-level indicators offer additional elucidation. The uniformity of categories at a single level is deemed descriptive consistency only when the indicators are methodologically equivalent. Inconsistencies across many tiers are marked for additional scrutiny and categorized distinctly instead of being averaged. All thresholds indicate comparative locations within the examined dataset and may not represent general ecological benchmarks. Thresholds obtained from limited datasets are deemed tentative and necessitate corroboration through further independent data.

Appendix A contains Table A1 and Table A2. Table A1 summarizes carbon-related parameters from the literature, including reported values, units, source references, and calculation methods; Table A2 presents a decision matrix used to interpret these indicators based on data availability, measurement scales, and accounting boundaries. ChatGPT (5.6sol) and DeepSeek (Web Version V3 Base Model and R1 Deep Thinking Model) assisted in extracting and organizing numerical data from selected literature. Before performing the percentile calculations according to the procedure described in this section, the authors verified the extracted values against the original literature. The authors reviewed the calculation outputs and the final tables. Tool identifiers, assistance with English translation, and the limitations of the retained AI logs have been disclosed in the Acknowledgments.

3. Carbon Sequestration and BVOC Emission Patterns and Their Trade-Off Relationship

3.1. Ecological and Environmental Significance of BVOC Emissions

3.1.1. Physiological Significance and Environmental Impacts of BVOC Emissions

BVOCs connect plant physiology, ecosystem processes, and atmospheric chemistry. Their emissions are contingent upon temperature, radiation, leaf maturity, phenological stage, organ classification, water and nutrient accessibility, tissue impairment, and measurement parameters. Consequently, emission rates assessed under regulated circumstances of elevated light and temperature cannot be readily juxtaposed with field fluxes in fluctuating environments. Thermal stress could amplify the release of photochemically active BVOCs from tropical and subtropical flora, possibly aiding leaf functionality and adaptation [2]. Emission rates can differ by over an order of magnitude among various tree species and tissues, underscoring their significance in environmental reactions and biosphere-atmosphere dynamics, along with their relevance in vegetation evaluation and emission modeling [30,31,32].

BVOCs serve as significant precursors for ozone (O₃) and SOAs. Their oxidation facilitates the creation of SOAs and enhances ozone (O₃) generation when exposed to sunlight and nitrogen oxides (NO_x), potentially exerting a more significant influence in regions characterized by dense flora and elevated anthropogenic pollution. Some monoterpenes and sesquiterpenes demonstrate antibacterial properties and may affect psychophysiological reactions, establishing a biological foundation for forest therapy, restorative settings, and nature-oriented leisure pursuits [33,34]. Consequently, the choice of urban tree species must incorporate species-specific BVOC emissions and carbon sequestration potential alongside local environmental and pollution factors to harmonize ecosystem services, prospective health advantages, and atmospheric hazards (Figure 4) [1,14,15,16,17,18,19,20,24,25,34,35,36,37,38].

Figure 4. Urban trees simultaneously function as carbon sinks, emit BVOCs that influence atmospheric chemistry, and provide context-dependent air quality and health benefits.

Biogenic and anthropogenic VOCs may jointly influence urban O₃ and SOA formation. Field observations in Taiyuan, an industrial city in northern China, revealed that coking and coal/biomass combustion are the primary contributors to the potential for VOC-derived O₃ and SOA formation. Under high-temperature and high-NO_x conditions, tree-source isoprene and reactive monoterpenes may interact with the oxidation chemistry of anthropogenic VOCs, promoting radical cycling and the formation of condensable products. Their combined effects may be nonlinear and depend on compound composition, the VOC/NO_x ratio, radiation, atmospheric stagnation, and regional transport. Although the Taiyuan study did not quantify the contribution of tree-source BVOCs separately, it highlights the need to consider local anthropogenic pollution when assessing the impact of BVOCs. Source apportionment measurements and atmospheric chemistry modeling are needed to quantify their combined contribution to O₃ and SOA formation [39,40,41,42,43,44,45,46].

3.1.2. Comprehensive Evaluation of BVOC Emissions for Urban Greening Tree Species Selection

BVOCs influence environmental systems through atmospheric chemistry and plant–atmosphere interactions. Their high reactivity promotes the formation of ozone (O₃) and SOAs, thereby exacerbating regional air pollution [4,17,34,37,38,47]. Emissions vary with temperature, radiation, phenology, leaf age, physiological status, water availability, and stress; therefore, climate-driven phenological shifts may alter BVOC emissions, atmospheric composition, and climate processes [5,48]. Their effects also depend on NO_x concentrations, meteorological conditions, vegetation structure, pollution levels, and anthropogenic emissions [10,11,12,33,49,50,51,52,53]. Under low-pollution or low-NO_x conditions, the contribution of BVOCs to O₃ formation may be negligible, whereas transportation, industry, power generation, and residential activities may dominate urban pollution. Therefore, the impact of BVOCs should be assessed on regional and seasonal scales, rather than assuming they are universally harmful contributors to secondary pollution [54,55].

Bao et al. (2025) found that BVOC emissions from trees in 65 residential areas in Beijing were correlated with socioeconomic conditions and were particularly high in some newly developed communities [6]. Therefore, urban species selection should consider plant functional traits, BVOC composition, landscape value, and vegetation cover. For carbon sink forests, tree species with high carbon sequestration and low pollution risk should be prioritized, whereas health-oriented forests may favor species with higher monoterpene content and lower ecological risk. Caution is warranted when planting tree species with high isoprene emissions on a large scale in urban or polluted environments [56,57]. Consequently, BVOC characteristics can inform preliminary species comparisons, vegetation management, and ecological spatial planning [6,18,33,52,56].

Evidence from Taiyuan, a city in a northern Chinese basin, indicates that, against a backdrop of spatially heterogeneous anthropogenic emissions, high temperatures and high NO_x concentrations strongly influence summer VOC distribution and O₃ formation [58]. In photochemically active environments, tree-source BVOCs may act as additional reactive precursors, particularly in areas with dense plantings of isoprene-emitting species and limited ventilation. The magnitude of this contribution depends on compound composition, NO_x levels, atmospheric dilution, radiation, and regional transport. Although the Taiyuan study did not isolate the contribution of urban vegetation to O₃, it highlights the importance of evaluating tree-source BVOCs in the context of anthropogenic pollution. Therefore, source-resolved measurements of BVOCs and anthropogenic VOCs, combined with NO_x observations and atmospheric chemistry modeling, are needed to determine the impact of urban tree species in northern Chinese cities on O₃ [59,60,61,62,63,64].

3.1.3. Spatially Explicit Modeling of the Interaction Between Urban Greenery and Air Pollution

The atmospheric impact of urban vegetation cannot be deduced merely from species-specific BVOC emission rates, as it is also affected by planting arrangements, architectural designs, airflow, microclimatic conditions, and the spatial arrangement of human-made emission sources. Horizontal greening systems (such as street trees, parks, and green roofs) and vertical systems (including green facades and living walls) can influence airflow, solar radiation, temperature, pollutant dispersion, and public visibility. As a result, their ecological impacts can differ markedly among urban canyons, residential areas, industrial districts, and expansive suburban terrains.

Spatially explicit modeling provides a good approach for understanding these interactions. Akopov et al. created an object-oriented model to replicate the interactions between air contaminants and urban vegetation in Yerevan, illustrating how diverse pollution sources, plant life, and spatial arrangements can be integrated into scenario-driven greening evaluations [65]. Huang et al. have shown that the efficacy of urban vegetation in densely populated cities is contingent upon its spatial configuration and the adjacent constructed environment [66]. The recent study focused mostly on thermal performance instead of atmospheric chemistry, although its results are pertinent as temperature, radiation, and ventilation influence both BVOC emissions and photochemical O₃ production.

This framework allows for the integration of species-specific BVOC emission potentials, carbon sequestration, anthropogenic VOCs and NO_x, canopy structure, building geometry, meteorological factors, and pollutant dispersion within entity-based and street-scale models. These models will provide the concurrent assessment of horizontal and vertical greening scenarios regarding carbon sequestration, thermal management, O₃ and SOA generation, and public exposure. As a result, they are able to convert the initial species classifications suggested in this analysis into localized planting and spatial design approaches. Nonetheless, prior to the use of model outputs for practical planning, they require validation against empirical data about local BVOCs, anthropogenic VOCs, NO_x, O₃, climatic factors, and vegetation traits [67,68,69,70].

3.1.4. Geographic Disparities in VOC Emission by Trees and Consequences for Urban Greening

Regional data underscores discrepancies in VOC emissions by trees and their consequences for the selection of tree species (Table 3). Research conducted in Japan and Spain has shown the absorption of certain aromatic compounds and formaldehyde, respectively, whereas an investigation in Chile recorded the buildup of anthropogenic VOCs in foliage [71,72,73]. These results indicate separate mechanisms: leaf retention does not influence overall tree absorption rates, and the assimilation of one volatile organic chemical does not guarantee the efficient elimination of others. The majority of the research compiled in Table 3 evaluated BVOC emissions instead of direct air VOC absorption.

Research conducted across Asia, Europe, the Americas, Oceania, and Africa has shown emission traits and ecological reactions that are distinct to many species. Consequently, the selection of tree species must thoroughly take into account compound composition, regional climate, NO_x levels, air dispersion, and carbon sequestration. Thorough evaluations—like an Italian research project that integrates carbon sequestration, pollutant elimination, and BVOC emissions—demonstrate this methodology [74]. Nonetheless, variations in measurement techniques and environmental factors hinder a consistent cross-regional assessment. Thus, regional data enhances the extensive examination centered on China by highlighting factors pertinent to local evaluations; it neither formulates universal planting guidelines nor substantiates the suggested functional classifications outside the framework of the first study.

Table 3. Features of volatile organic compound uptake, retention, and biogenic VOC emissions by typical tree species in diverse world regions.

Macro- Region/Country	Representative Species	VOC Uptake or Retention	BVOC Emissions/ Atmospheric Effects	Implications for Species Selection	References
Asia/Japan	Quercus myrsinifolia, Q. acutissima, Osmanthus fragrans var. aurantiacus	Direct uptake of phenol, benzaldehyde, and benzyl alcohol; no detectable benzene/toluene uptake	Not evaluated	Prioritize species- and pollutant-specific VOC absorption properties	[71]
Asia/China	Fraxinus chinensis, Ligustrum lucidum, Ulmus pumila	Not assessed	No detectable isoprene or monoterpene emissions above the study threshold	Potential low-isoprenoid candidates requiring further assessment	[20]
Asia/South Korea	Quercus spp., Pinus thunbergii, Metasequoia glyptostroboides	Not assessed	High isoprene emissions in Quercus spp.; notable monoterpene emissions in the two conifers	Evaluate species-specific emissions under local light and temperature conditions	[75]
Asia/India	Dalbergia sissoo	Not assessed	Total average VOC emission reached 80.6 μg·g−1·h−1	Exercise caution with high emitters in ozone-prone environments	[76]
Asia/Iran	Platanus orientalis, Robinia pseudoacacia, Fraxinus rotundifolia	Not assessed	Late-season isoprene emissions of 45 and 17.8 μg·g−1·h−1 for the first two species; undetected in F. rotundifolia	Favor lower-isoprene alternatives when photochemical pollution is a concern	[77]
Asia–Europe/Türkiye	Broad-leaved and coniferous trees among 31 investigated species	Not assessed	Isoprene commonly dominated broad-leaved species, whereas monoterpenes dominated conifers	Develop localized, species-specific emission inventories	[78]
Europe/Spain	Quercus ilex, Pinus halepensis	Direct formaldehyde uptake above an approximately 20 ppbv compensation point	Formaldehyde emission occurred below the compensation point	Consider ambient formaldehyde concentrations and stomatal conductance	[72]
Europe/Italy	Liriodendron tulipifera, Celtis australis, Acer campestre, A. platanoides	VOC uptake not directly measured; O3 and PM10 removal assessed	VOC emissions and ozone-forming potential evaluated	Integrate carbon sequestration, O3/PM removal, and BVOC emissions	[74]
Europe/Switzerland	Populus nigra, P. tremula, P. alba	Direct VOC uptake not assessed; O3 and PM removal modeled	Among the highest total BVOC emitters in the 51-species assessment	Screen high-emitting poplars for local ozone-formation risk	[79]
Europe/Germany	Fagus sylvatica	Not assessed	Sabinene-dominated monoterpene emissions of 4–13 μg·g−1·h−1 under standard conditions	Consider light- and temperature-dependent emissions	[80]
North America/USA	Liquidambar styraciflua, Pinus radiata	Not assessed	High-BVOC sweetgum and lower-BVOC Monterey pine planting scenarios produced contrasting secondary-pollution implications	Incorporate BVOC reactivity into urban greening scenarios	[38]
South America/Chile	Liriodendron tulipifera, Platanus × acerifolia	Anthropogenic VOCs accumulated in foliage; higher concentrations in L. tulipifera	BVOC emissions not measured	Assess leaf VOC retention without equating it to whole-tree uptake rates	[73]
South America/Brazil	Ocotea odorifera, Cordia sellowiana, Byrsonima ligustrifolia	Not assessed	The first two contributed relatively little to modeled secondary pollution; pronounced seasonal variations occurred in B. ligustrifolia	Consider seasonal emissions and ozone/SOA formation potential	[11]
Oceania/Australia	Eucalyptus camaldulensis, E. globulus, E. grandis, E. viminalis	Carbonyl exchange investigated; no generalized VOC-removal ranking	Species-dependent isoprene, monoterpene, and carbonyl exchanges	Use regional measurements rather than generic Eucalyptus emission factors	[81]
Africa/South Africa	Acacia nigrescens and investigated Combretaceae species	Not assessed	A. nigrescens emitted isoprene; investigated Combretaceae species showed no significant isoprene emissions	Account for local species composition in BVOC assessments	[82]

(“Not assessed” indicates that the direct absorption of ambient VOCs was not measured.) The observations about tree species selection constitute evidence-based analyses rather than officially recorded governmental policies. Emission rates from several studies may not be directly comparable because of variations in methodology, environmental factors, and measurement units.

Table 3. Features of volatile organic compound uptake, retention, and biogenic VOC emissions by typical tree species in diverse world regions.

Macro- Region/Country	Representative Species	VOC Uptake or Retention	BVOC Emissions/ Atmospheric Effects	Implications for Species Selection	References
Asia/Japan	Quercus myrsinifolia, Q. acutissima, Osmanthus fragrans var. aurantiacus	Direct uptake of phenol, benzaldehyde, and benzyl alcohol; no detectable benzene/toluene uptake	Not evaluated	Prioritize species- and pollutant-specific VOC absorption properties	[71]
Asia/China	Fraxinus chinensis, Ligustrum lucidum, Ulmus pumila	Not assessed	No detectable isoprene or monoterpene emissions above the study threshold	Potential low-isoprenoid candidates requiring further assessment	[20]
Asia/South Korea	Quercus spp., Pinus thunbergii, Metasequoia glyptostroboides	Not assessed	High isoprene emissions in Quercus spp.; notable monoterpene emissions in the two conifers	Evaluate species-specific emissions under local light and temperature conditions	[75]
Asia/India	Dalbergia sissoo	Not assessed	Total average VOC emission reached 80.6 μg·g−1·h−1	Exercise caution with high emitters in ozone-prone environments	[76]
Asia/Iran	Platanus orientalis, Robinia pseudoacacia, Fraxinus rotundifolia	Not assessed	Late-season isoprene emissions of 45 and 17.8 μg·g−1·h−1 for the first two species; undetected in F. rotundifolia	Favor lower-isoprene alternatives when photochemical pollution is a concern	[77]
Asia–Europe/Türkiye	Broad-leaved and coniferous trees among 31 investigated species	Not assessed	Isoprene commonly dominated broad-leaved species, whereas monoterpenes dominated conifers	Develop localized, species-specific emission inventories	[78]
Europe/Spain	Quercus ilex, Pinus halepensis	Direct formaldehyde uptake above an approximately 20 ppbv compensation point	Formaldehyde emission occurred below the compensation point	Consider ambient formaldehyde concentrations and stomatal conductance	[72]
Europe/Italy	Liriodendron tulipifera, Celtis australis, Acer campestre, A. platanoides	VOC uptake not directly measured; O3 and PM10 removal assessed	VOC emissions and ozone-forming potential evaluated	Integrate carbon sequestration, O3/PM removal, and BVOC emissions	[74]
Europe/Switzerland	Populus nigra, P. tremula, P. alba	Direct VOC uptake not assessed; O3 and PM removal modeled	Among the highest total BVOC emitters in the 51-species assessment	Screen high-emitting poplars for local ozone-formation risk	[79]
Europe/Germany	Fagus sylvatica	Not assessed	Sabinene-dominated monoterpene emissions of 4–13 μg·g−1·h−1 under standard conditions	Consider light- and temperature-dependent emissions	[80]
North America/USA	Liquidambar styraciflua, Pinus radiata	Not assessed	High-BVOC sweetgum and lower-BVOC Monterey pine planting scenarios produced contrasting secondary-pollution implications	Incorporate BVOC reactivity into urban greening scenarios	[38]
South America/Chile	Liriodendron tulipifera, Platanus × acerifolia	Anthropogenic VOCs accumulated in foliage; higher concentrations in L. tulipifera	BVOC emissions not measured	Assess leaf VOC retention without equating it to whole-tree uptake rates	[73]
South America/Brazil	Ocotea odorifera, Cordia sellowiana, Byrsonima ligustrifolia	Not assessed	The first two contributed relatively little to modeled secondary pollution; pronounced seasonal variations occurred in B. ligustrifolia	Consider seasonal emissions and ozone/SOA formation potential	[11]
Oceania/Australia	Eucalyptus camaldulensis, E. globulus, E. grandis, E. viminalis	Carbonyl exchange investigated; no generalized VOC-removal ranking	Species-dependent isoprene, monoterpene, and carbonyl exchanges	Use regional measurements rather than generic Eucalyptus emission factors	[81]
Africa/South Africa	Acacia nigrescens and investigated Combretaceae species	Not assessed	A. nigrescens emitted isoprene; investigated Combretaceae species showed no significant isoprene emissions	Account for local species composition in BVOC assessments	[82]

(“Not assessed” indicates that the direct absorption of ambient VOCs was not measured.) The observations about tree species selection constitute evidence-based analyses rather than officially recorded governmental policies. Emission rates from several studies may not be directly comparable because of variations in methodology, environmental factors, and measurement units.

3.2. Differences in Carbon Sequestration and Oxygen Release Capacity Among Urban Greening Tree Species

3.2.1. Local Data on Tree Carbon Absorption and Its Consequences for Urban Flora Selection

Regional analyses offer more proof regarding urban tree carbon storage, carbon sequestration, and carbon equilibrium (Table 4). Nonetheless, their results indicate variations in tree dimensions, environmental conditions, management strategies, evaluation durations, and accounting parameters, rather than solely changes in species classification. In Beijing, the estimates of mixed species indicate the sampled populations of street trees and should not be construed as species-specific carbon sequestration rates [83]. An investigation of transplanted Ginkgo trees in Chengdu underscored the significance of planting methodologies and sustained growth, whereas a nursery examination in the Republic of Korea illustrated the role of subterranean biomass in carbon assessment [84,85]. These results advocate for the concurrent evaluation of tree health, planting environment, and species traits in assessments.

Table 4. Regional data on carbon reserves, carbon absorption, and carbon equilibrium in urban arboriculture and their consequences for local species selection.

Macroregion/Study Location	Representative Tree Species	Metrics and Main Findings	Accounting Boundaries and Limitations	Implications for Local Tree Selection	References
Asia/Beijing, China	Sophora japonica; Fraxinus chinensis; Ginkgo biloba; Populus spp.	Across 2040 street trees, mean carbon storage and annual carbon sequestration were 130.62 kg C·tree−1 and 5.85 kg C·tree−1·yr−1, respectively.	Estimates were based on biomass equations and diameter increments. A common mean diameter increment was applied across species. The mixed-species means do not represent species-specific sequestration rates.	Retain healthy mature trees and provide adequate growing space; evaluate carbon benefits in relation to tree size and community structure.	[83]
Asia/Chengdu, China	Ginkgo biloba	Mean annual carbon sequestration in the super-large transplant group declined from 7.6 kg C·tree−1·yr−1 in 2001 to approximately 1.5 kg C·tree−1·yr−1 in 2021. No declining trend was observed in the small-caliper control group.	Tree-ring reconstruction included 18 super-large transplants and five control trees. Findings are specific to the transplant treatment and study conditions.	Protect root systems and consider the effects of planting-stock size and transplant damage on long-term growth and carbon sequestration.	[84]
Asia/South Korea	Zelkova serrata; Quercus palustris; Pinus	In a study of 117 nursery-grown landscaping trees, mean annual sequestration over the preceding five years was 5.82, 5.62, 3.93, and 1.90 kg C·tree−1·yr−1 for these species, respectively.	Estimates included aboveground and belowground biomass. Sampled trees differed in size and were grown under nursery conditions; results cannot be directly extrapolated to mature street trees or interpreted as an age-standardized species ranking.	Include roots in carbon accounting and validate nursery-derived estimates under local planting conditions.	[85]
Europe/Florence, Italy	Tilia platyphyllos; Acer rubrum; Acer opalus; Cupressus sempervirens; Ulmus “Plinio”	A 50-year modeling assessment of a newly planted urban forest comprising 170 trees indicated that the cumulative carbon balance became a sink around year 13. The life cycle carbon footprint was approximately 14.7 t CO2e.	The assessment combined an assimilation model with life cycle analysis, including establishment, maintenance, and respiration. Soil carbon storage, grass assimilation, and tree mortality were excluded.	Assess long-term assimilation benefits together with establishment and maintenance emissions; the estimated compensation time should not be generalized across Europe.	[86]
Europe/Helsinki, Finland	Tilia × vulgaris; Alnus glutinosa	The annual carbon balance of the tree–soil systems changed from a source to a sink approximately 14 and 12 years after planting, respectively.	Initial respiration losses from planting substrates were substantial. An annual transition to a sink does not indicate repayment of the cumulative carbon debt; the assessment was not a complete life cycle analysis.	Incorporate substrate composition, soil respiration, and local growing conditions into species selection and carbon assessment.	[87]
North America/Los Angeles, USA	Platanus racemosa	A 50-year life cycle assessment estimated net emissions of 1.204 t CO2 tree−1 under the highest-emission scenario and net removal of 3.768 t CO2 tree−1 under the lowest-emission scenario.	Initial respiration losses from planting substrates were substantial. An annual transition to a sink does not indicate repayment of the cumulative carbon debt; the assessment was not a complete life cycle analysis.	Match species’ water requirements to local water availability and optimize irrigation, maintenance, and wood utilization.	[88]
South America/Ibagué, Colombia	Samanea saman; Ficus benjamina; Tabebuia rosea	A remeasurement study estimated mean annual aboveground biomass carbon sequestration of 111.9, 56.5, and 25.5 kg C·tree−1·yr−1 for these species, respectively.	Estimates were based on biomass carbon increments over the measurement interval. Tree-size distributions differed among species; roots and maintenance emissions were excluded.	Provide adequate space for species with large crowns while considering local suitability, species diversity, and compatibility with infrastructure.	[89]
South America/Maringá, Brazil	Cenostigma pluviosum; Tipuana tipu; Handroanthus heptaphyllus; Cordia trichotoma	A campus area established approximately 50 years earlier contained 998 trees and large palms. The estimated aboveground carbon stock was 402.76 Mg C, with a reported density of 35.25 Mg C ha−1.	This was a carbon stock assessment. Species-level aggregate contributions depended on abundance and tree size. Dividing the stock by 50 years does not provide an annual sequestration rate.	Preserve mature vegetation and use local wood density values and community structure in carbon inventories.	[90]
Oceania/Melbourne, Australia	Ulmus procera; Platanus × acerifolia; Corymbia maculata	Accounting for decayed and missing wood reduced estimated carbon stocks by 69–110 kg C tree−1. A 13% stem volume reduction was recommended for U. procera with a diameter at breast height ≥ 40 cm.	Results represent corrections to carbon stock estimates rather than measured CO2 emissions. The 13% correction is restricted to the specified species and size range.	Incorporate internal decay into carbon stock accounting rather than relying solely on visual assessments of tree vitality.	[91]
Africa/Tshwane, South Africa	Combretum erythrophyllum; Searsia lancea; Searsia pendulina	Growth models projected that 115,200 indigenous street trees could sequester approximately 54,630 t C over a 30-year period.	Results represent projected cumulative sequestration by the tree population rather than measured annual uptake per tree. Nursery production, planting, maintenance, and mortality losses were not fully accounted for.	Use locally calibrated growth models and suitable species, supported by continued monitoring of survival and maintenance requirements.	[92]

Research from Europe and North America further demonstrates how accounting limits affect projected carbon advantages. A study conducted in Florence analyzed the transition in cumulative carbon balance toward a carbon sink, whereas research in Helsinki investigated the change in yearly tree soil carbon balance; these findings denote separate processes and should not be directly compared [86,87]. In Los Angeles, several management strategies produced markedly divergent life cycle carbon results for identical species [88]. Data from Colombia, Brazil, Australia, and South Africa underscore the significance of tree size distribution, mature flora, internal wood deterioration, and regionally adjusted growth models [89,90,91,92].

Consequently, the selection of regional species ought to integrate carbon-associated characteristics with local adaptability, growth potential, survival probabilities, and maintenance needs. The data presented in Table 4 expand the geographical scope of this review, initially centered on China, although they fail to endorse a cohesive global species hierarchy. Carbon reserves, yearly biomass increases, life cycle assessments, and anticipated cumulative reserves must preserve their initial definitions and measurements. Consequently, these diverse observations offer contextual proof instead of a unified, comprehensive dataset for percentile categorization.

3.2.2. Carbon Sequestration and Oxygen Release Differences Based on Tree Life Forms and Community Structure

Under similar climatic, soil, and management conditions, deciduous trees can demonstrate greater seasonal carbon absorption rates and net oxygen output compared to some evergreen species. This phenomenon is generally linked to their fast leaf development, comparatively elevated photosynthetic efficiency throughout the growth period, and substantial seasonal biomass growth. As a result, in specific ecosystems, deciduous trees might attain more carbon absorption and oxygen production over similar durations [26,93,94,95]. Nonetheless, this trend is not to be perceived as an absolute principle, since the comparative efficacy of deciduous and evergreen trees is also affected by species characteristics, leaf phenology, water accessibility, temperature, illumination conditions, and the length of the evaluation.

According to these results, under particular environmental and temporal circumstances, deciduous tree species might exhibit superior rates of carbon sequestration and oxygen production compared to evergreen species, thereby presenting significant ecological advantages and applicability in select urban green areas. Nonetheless, this does not suggest that evergreen trees possess diminished ecological roles. Evergreen trees can sustain perennial foliage, enhance the aesthetic appeal of winter landscapes, and offer consistent ecological safeguarding. Their ecological roles could be especially significant in regions or functional areas that necessitate continuous canopy coverage, wind shielding, optical obscuration, or winter habitats. Consequently, deciduous and evergreen trees must not be regarded as universally substitutable. The choice and arrangement should be dictated by local climate, soil and water conditions, functional zoning, maintenance capabilities, and the ecological goals of the planting area.

Deciduous trees play a crucial role in carbon sequestration and net oxygen production during their growth period, whereas evergreen trees contribute to more consistent vegetation cover and ecological stability year-round. A synergistic integration of these two organisms can enhance the temporal consistency and overall ecological efficacy of urban green areas. Evergreen broad-leaved shrubs could be advantageous for carbon capture and oxygen production in urban horticultural systems, especially if their shade tolerance and perennial foliage align with local environmental factors. Integrating shrubs with trees and herbaceous flora can create a stratified plant community, although the ecological advantages hinge on species compatibility, planting density, light access, soil moisture, and ongoing care [26,28].

Trees and shrubs vary in their architecture, biomass allocation, and physiological traits; consequently, they might also exhibit differences in carbon storage, daily carbon sequestration trends, and ecological roles. Trees generally possess more expansive canopies and more aboveground biomass, potentially providing enhanced long-term carbon sequestration capabilities and rendering them vital participants in the carbon sink role of urban green areas. Conversely, shrubs possess a robust capacity to occupy space and can encourage vertical stratification, thereby aiding in the regulation of community structure and bolstering the stability of vegetation systems. The daily patterns of net carbon sequestration in shrubs can change markedly among species and taxonomic categories, with several species demonstrating unique carbon sequestration cycles. Consequently, the ability of carbon sequestration is influenced not solely by the dimensions or type of plant but also by its structural composition, physiological attributes, foliar qualities, photosynthetic efficiency, and existing environmental factors. Evaluations reliant exclusively on tree morphology or plant stature could inaccurately gauge the impact of particular tree species [24,27,93,96].

Therefore, when choosing tree species for urban environments, one must take into account the collective benefits of trees and shrubs in terms of carbon dioxide absorption, oxygen production, shading, and protection. Layered assemblages of trees, shrubs, and herbaceous flora can optimize the use of light and spatial resources, thereby augmenting the overall carbon capture capacity of urban green areas. The magnitude of this effect is contingent upon the particular location and may fluctuate based on species diversity, canopy thickness, resource accessibility, climate factors, and management strategies. Consequently, multilayered plantings ought to be tailored to local ecological contexts and functional needs, rather than implemented as a one-size-fits-all solution [97,98,99].

3.2.3. The Importance of Carbon Sequestration Capacity in Tree Species Selection and Associated Controversies

Certain academics contend that the carbon sequestration role of urban green spaces ought not to be overstated. Urban green spaces serve various functions, including carbon sequestration, temperature and humidity regulation, dust reduction, noise level attenuation, biodiversity protection, urban aesthetic enhancement, provision of recreational and health-promoting venues, and air quality improvement, among other advantages. Individuals frequently concentrate excessively on species with significant carbon sequestration capabilities, overlooking additional considerations including adaptation to diverse settings, maintenance demands, vulnerability to pests and diseases, root space necessities, and the likelihood of BVOC emissions [26,95,100,101].

Environmental elements additionally affect the actualization of carbon sequestration capacity. Certain species have considerable carbon sequestration rates in ideal growth conditions; nevertheless, their efficacy may diminish when confronted with actual metropolitan challenges such as elevated temperatures, humidity, drought, soil compaction, pollution, and several other influences. As a result, the potential for theoretical carbon sequestration may remain unactualized in reality. Consequently, depending exclusively on a species’ ability to sequester carbon is inadequate to fulfil the enduring requirements of an entire urban area [10,16,53,102,103,104].

Although the capacity for carbon sequestration is indisputable, one must also take into account ecological adaptation, environmental hazards, and additional ecosystem services [104,105].

In summary, while the carbon sequestration potential is a significant metric for assessing tree species for urban greening, it ought not to be the exclusive standard. The selection of tree species ought to consider carbon sequestration and utilization efficacy, oxygen emission potential, life form traits, vegetation structural hierarchy, environmental adaptability, and ecological hazards.

Certain species exhibit significant carbon sequestration potential and may be chosen for cultivation in carbon sink forests, ecological conservation forests, and expansive urban parks. Evergreen trees sustain their foliage throughout the year and contribute to a consistent natural habitat; shrubs enhance the diversity of various regions and fulfil additional functions in urban greening initiatives. Augmenting the quantity of deciduous and evergreen trees and shrubs, while deliberately introducing species with significant carbon sequestration abilities alongside highly adaptable varieties, can concurrently improve the carbon sink capacity and perennial ecological roles of urban green areas [106].

Consequently, forthcoming urban greening initiatives have to prioritize not just carbon sequestration but also the synergistic advancement of several additional functions. By categorizing flora according to their carbon absorption capabilities and taking into account the climatic variations of various areas, one may strategically design a city’s functional categories and community frameworks, thus establishing a robust and sustainable urban vegetation network [14,34,37,107].

3.3. Coupling Relationship Between Tree Carbon Sequestration Capacity and BVOC Emissions: A Functional Classification Framework

This review distinguishes between two provisional categories that integrate carbon performance and BVOC emissions—the “trade-off” and “synergy” categories—as well as an emission-specific descriptor known as the “provisional low-BVOC emission category.” The integrated categories require evidence for both metrics, whereas the emission-specific category describes only BVOC evidence. None of these categories establish overall environmental safety, human health benefits, or suitability for urban planting.

For the quantitative screening scenario, C represents the annual carbon increment in biomass per tree, in units of kg C·tree⁻¹·yr⁻¹, and E represents the reported total BVOC emission rate, in units of μg·gDW⁻¹·h⁻¹. Eligibility requirements include traceable source information, a verified reporting basis, a fully characterized compound coverage, and evidence suitable for a combined interpretation of the same species and relevant developmental and environmental contexts. The following dataset-specific rules are proposed: provisional synergistic category: C > 8.910 and E ≤ 7.21; provisional trade-off category: C > 8.910 and E > 7.21; provisional low BVOC emission category: E ≤ 2.52, corresponding to the lowest descriptive emission band in Table 1. These rules define an exploratory screening scenario rather than validated biological or atmospheric risk thresholds. If a species meets the synergistic criteria and simultaneously falls within the lowest emission band, its integrated category and emission descriptor are reported separately. Other combinations are not mandatorily assigned to these categories. Compound-specific measurements can identify concerns associated with specific compounds but cannot independently determine the overall BVOC category. Missing carbon or emission measurements are recorded as “not reported”; in this screening scenario, cumulative carbon stocks, short-term photosynthetic CO₂ uptake, and stand-level carbon accumulation cannot substitute for annual individual-tree biomass carbon increment. Annualized net life cycle carbon balance is interpreted separately due to its different accounting boundaries. Missing values are not considered zero, nor are they inferred from another species or measurement scale. Toxicity, sensitization, invasiveness, ecological suitability, water requirements, and exposure conditions are assessed independently of carbon and emissions descriptors. In particular, the confirmed toxicity of oleander (Nerium oleander) requires separate site-specific restrictions. The quantitative worksheet provided does not establish its total BVOC emission category; therefore, its status is recorded as “BVOC classification unresolved; toxicity restrictions apply.” The framework excludes it from children’s play areas and locations where direct contact or ingestion exposure may occur. Even if lower emission rates are subsequently verified, this restriction will not be lifted [108].

Table 5 summarizes the available evidence, tentative interpretations, unresolved classifications, and site-specific restrictions. These entries indicate priorities for further evaluation rather than confirmed planting recommendations.

Table 5. Provisional interpretation of tree carbon and BVOC evidence, classification status, and site-specific constraints.

Provisional Interpretation and Decision Status	Tree Species	Carbon Evidence and Assessment Basis	BVOC Evidence and Assessment Basis	Reported BVOC Compounds or Classes	Toxicity	Allergenicity	Water Requirements	References	Site Assessment Requirements and Restrictions
Trade-off type	Populus spp. (Poplar)	Very high	Very high	Isoprene	Low	Moderate to high	High	[3,34]	Carbon sequestration forests in remote mountainous areas, ecological public welfare forests, industrial shelterbelts, and riverbank protection forests.
	Phyllostachys edulis (Moso bamboo)	Medium	Very high	Isoprene	Low	Low	Medium to high	[3,50,109]
	Betula platyphylla (Asian white birch)	Moderately high	Very high	Isoprene	Low	High	Medium	[3,110,111,112]
	Robinia pseudoacacia (Black locust)	High	Moderately high	Isoprene	High	Moderate	Low to moderate	[3]
	Salix babylonica (Weeping willow)	Very high	Very high	Isoprene	Low	Moderate	High	[10]
Synergistic type	Pinus tabuliformis (Chinese red pine)	Moderately high	Medium	Monoterpenes	Low	Moderate	Low to moderate	[3,113]	Suburban health forests, forest parks, scenic recreation forests, hospital surrounding greening, and ecological health demonstration zones; balancing carbon sequestration and human health benefits.
	Quercus suber (Cork oak)	Moderately high	High	Isoprene and monoterpenes	Medium to high	Moderate to High	Low to moderate	[3,34]
	Platycladus orientalis (oriental arborvitae)	Medium	Moderately low	Monoterpenes and sesquiterpenes	Low	Moderate	Low to moderate	[3,34,111]
	Cunninghamia lanceolata (Chinese fir)	Very high	Medium	Monoterpenes	Low	Low to moderate	Medium to high	[114,115]
	Metasequoia glyptostroboides (Dawn redwood)	Very high	Medium	Monoterpenes	Low	Low to medium	High	[115,116]
Provisionally low BVOC-emission category	Cinnamomum camphora (Camphor tree)	High	Moderately low	Oxygenated VOCs	Medium	Medium	Moderate	[11,102,104]	Urban core areas, high-density residential areas, administrative districts, and strict air quality control zones; prioritizing the protection of local atmospheric environmental safety.
	Ginkgo biloba (Ginkgo)	Medium	Moderately low	-pinene, β-pinene, and limonene	Low to medium	Low to medium	Moderate	[112,114,115]
	Nerium oleander (Oleander)	Moderately low	Moderately low	Alkanes and esters	Extremely high	Low to medium	Low	[115]	It is worth noting that although oleander is toxic, it is considered an excellent tree species for urban greening and air purification due to its ability to withstand smoke, sulfur dioxide, chlorine, and other harmful gases, as well as its tolerance for pollution, drought, and poor soil. However, in practical situations, planting scenarios must be considered comprehensively, and its toxicity, exposure risks, ecological suitability, and maintenance requirements must be evaluated on a case-by-case basis.

Overall, classifying tree species by carbon sequestration capacity and the environmental effects of BVOCs may help shift greening planning from a single-objective model to one that optimizes multiple ecosystem service objectives [14,25,27,117]. Balanced tree species can enhance carbon sink functions and provide ecological protection; synergistic tree species are particularly adept at integrating functions such as carbon sequestration, health promotion, and landscape recreation. Such classifications help expand regional carbon sink effects, mitigate BVOC-driven air pollution, and promote greater health benefits and other services from urban green spaces and forest ecosystems [18,26,56,95,101]. This document organizes evidence related to carbon and BVOCs; the provisional classifications do not constitute proof of net atmospheric benefits, health benefits, or suitability for specific planting sites. Their practical application requires independent regional assessments of atmospheric chemistry, ecological compatibility, toxicity, sensitization potential, invasiveness, exposure, and management conditions.

4. Incorporating BVOC Emissions and Carbon Capture into Urban Greenery Strategies

4.1. Consequences for the Selection of Tree Species and Spatial Organization

This critique underscores the necessity of incorporating the evaluation of BVOC emissions alongside tree carbon sequestration in urban vegetation strategies. Investigations have progressed from defining BVOC emissions to exploring their ecological roles and effects on atmospheric chemistry and the carbon cycle. The analyzed literature reveals considerable interspecific variations in carbon efficiency, BVOC emission rates, and the makeup of released chemicals. These distinctions bolster a framework for selecting tree species that concurrently takes into account carbon sequestration, air quality, environmental adaptation, and ecosystem services.

The selection of tree species ought to correspond with local environmental circumstances and the desired purposes of the planted flora. Species exhibiting both superior carbon performance and elevated isoprene emissions may be deemed suitable for peri-urban carbon sink forests or ecological shelterbelts, contingent upon the evaluation of local pollution levels, air dispersion, and human exposure. Species demonstrating considerable carbon sequestration and monoterpene emissions could be valuable for assessing recreational or health-enhancing forests, although their prospective advantages and atmospheric impacts warrant additional verification. Lower BVOC emissions may be considered as one screening attribute. Suitability for densely populated areas, transport corridors, or ozone-sensitive locations additionally requires assessment of toxicity, allergenicity, invasiveness, ecological compatibility, exposure, and local atmospheric chemistry.

Regional planning must take into account the impact of climate on BVOC emissions and carbon sequestration. In hotter areas with elevated sun exposure, focus should be specifically on the atmospheric impacts of high-emission species. In regions with low temperatures or limited water availability, species selection should account for growing season length, drought tolerance, and the effects of environmental stress on carbon accumulation and BVOC emissions. Consequently, a simplistic north–south differentiation is inadequate for informing planting choices. Incorporating temperature zones, urban functional areas, species characteristics, and atmospheric circumstances could facilitate the equilibrium between carbon sequestration and the safeguarding of air quality along with other ecosystem services. Nonetheless, regional verification is essential prior to the generalization of these recommendations [118,119,120,121,122].

4.2. Constraints, Ambiguities, and Focal Points for Forthcoming Verification

BVOC measurements vary in terms of the reporting basis, environmental conditions, and spatial scale. Converting rates based on leaf area to a dry mass basis introduces uncertainty through the LMA and its representativeness of the sampled leaves. Differences in leaf development, canopy position, season, and water status may limit the applicability of LMA values obtained from other samples or under different conditions. Canopy-scale interpretations further introduce uncertainties related to leaf distribution, light and temperature gradients, and the relationship between leaf emissions and above-canopy exchange. Therefore, expressing values in common units does not eliminate these sources of variability [123].

Reference-condition emission potentials must also be distinguished from actual field emissions. Temperature and light normalization depend on the selected response model and its applicability to the compounds and measurement conditions. Normalized leaf-level rates cannot directly establish seasonal canopy emissions or the resulting O₃ and SOA formation.

The current compilation does not provide a complete observation-level record of conversion parameters, normalization procedures, or their uncertainties. Consequently, the worksheet alone does not allow for determining the number of fully standardized measurements or the number of observations excluded due to insufficient conversion information. The available metadata also does not support the propagation of conversion uncertainties. These limitations constrain the confidence in quantitative comparisons and species-level classifications, particularly near category boundaries. Future measurements should report raw values and units, corresponding leaf characteristics, environmental conditions, conversion parameters, and associated uncertainties.

First, the gathered data are spatially constrained. Contemporary studies are focused on particular nations, areas, and metropolitan environments, but numerous climatic zones and biogeographic regions are insufficiently recorded. As a result, recorded species-specific patterns may not be relevant to regions with varying temperatures, pollution intensities, atmospheric transport conditions, or management approaches. The quantity of observations also fluctuates considerably among species. Numerous impartial investigations endorse specific classifications, but others depend exclusively on a single local evaluation. The disparities in evidence density lead to differing levels of uncertainty among species.

Second, the carbon metrics analyzed encompass many measures, such as life cycle carbon sequestration, yearly fluctuations in tree-level carbon reserves, photosynthetic absorption in relation to leaf surface area, and carbon storage per unit area of forest stand. These markers differ in their geographical extent, time frame, and ecological significance; thus, they cannot be considered uniform, standardized assessments of carbon sequestration capacity. Relative category thresholds are established within certain datasets and ought not to be considered as general physiological or managerial standards.

Third, the suggested functional groups and cultivation scenarios have not been thoroughly substantiated by atmospheric chemistry models or empirical research. The overall rate of BVOC emissions cannot solely indicate the generation of ozone or secondary organic aerosols, as these phenomena are affected by many compound species, reactivity, NOx levels, meteorological conditions, background human emissions, and airflow. Consequently, the suggested recommendations are to be viewed as initial screening parameters instead of conclusive directives for decision-making.

Ultimately, the external validation of the three-tier framework is still restricted. The applicability should be evaluated via simultaneous measurements across various climate zones, recurrent seasonal observations, compound-specific BVOC data, standardized carbon metrics, and autonomous evaluations of ozone formation potential and secondary organic aerosol production potential.

Long-term field observations and comprehensive atmospheric chemistry simulations are essential. Independent field observations and atmospheric chemistry modeling are necessary to assess the stability and transferability of the provisional trade-off and synergy categories, as well as the provisional low-BVOC emission category, across different locations and environmental conditions [123,124,125,126,127,128,129,130,131,132].

5. Conclusions

The analyses examined in this study demonstrate that the traits of tree carbon sequestration and BVOC emissions differ markedly among species, growth phases, environmental contexts, and measuring techniques. Carbon sequestration fluctuates based on tree dimensions, environmental factors, and cultivation techniques, but BVOC emissions vary in both rate and chemical composition. The atmospheric impacts are contingent upon the reactivity of individual compounds, levels of ambient NO_x, availability of oxidants, meteorological factors, and existing anthropogenic emissions. Carbon stocks, annual biomass carbon increments, photosynthetic CO₂ uptake, and stand-level sequestration assessments denote distinct processes and accounting frameworks, necessitating individual interpretation. These results endorse the simultaneous evaluation of carbon performance and BVOC traits, while constraining direct comparisons among disparate observations.

This review proposes provisional trade-off and synergistic categories and a provisionally low BVOC-emission category, subject to the evidence requirements in Section 3.3. Missing or conflicting measurements leave assignments unresolved or context-dependent, and none of these descriptors establishes overall planting safety. The preserved descriptive BVOC emission categories, articulated in μg·gDW⁻¹·h⁻¹, are delineated as follows: moderately low, E ≤ 2.52; medium, 2.52 < E ≤ 7.21; moderately high, 7.21 < E ≤ 17.08; high, 17.08 < E ≤ 32.93; very high, E > 32.93, with E denoting the specified emission rate. The percentile thresholds for annual biomass carbon increment per tree for P20, P40, P60, and P80 were 3.950, 6.222, 8.910, and 16.482 kg·tree⁻¹·yr⁻¹, respectively. The five classifications were defined as ≤P20, >P20–P40, >P40–P60, >P60–P80, and >P80. Tree species with annual biomass carbon increments exceeding 8.910 kg C·tree⁻¹·yr⁻¹ and total BVOC emission rates ≤ 7.21 μg·gDW⁻¹·h⁻¹ may be considered provisional high-carbon, low-to-moderate-emission candidates, subject to verification using comparable carbon and BVOC measurements. The combination of an annual biomass carbon increment >16.482 kg C·tree⁻¹·yr⁻¹ and a total BVOC emission rate ≤2.52 μg·gDW⁻¹·h⁻¹ defines an exploratory scenario for investigating potential carbon–emission synergies; its suitability for species selection requires independent validation. Emission rates > 17.08 μg·gDW⁻¹·h⁻¹ may provide a basis for defining research scenarios that examine OFP, SOAP, and predicted O₃ and SOA formation under local atmospheric conditions, particularly in areas with elevated NO_x concentrations or ozone sensitivity. Similarly, emission rates > 32.93 μg·gDW⁻¹·h⁻¹ may be investigated under contrasting planting densities through modeling and field monitoring to assess their potential incremental atmospheric effects. These scenarios support hypothesis testing, while their implications for planting decisions remain dependent on regional validation.

These numerical limits delineate the aggregated evidence and do not define universal biological emission standards, confirmed atmospheric risk thresholds, or operational agricultural criteria. Functional grouping continues to be conjecture awaiting additional scrutiny rather than an established physiological process.

Prior to implementation, this paradigm necessitates several assessments of BVOC composition and carbon sequestration throughout various seasons, locations, and climatic zones for the same recognized taxonomic groups during analogous developmental phases. Reference measurements at 30 °C with a photosynthetic photon flux density of 1000 μmol m⁻² s⁻¹ must be differentiated from field observations and should include environmental metadata and uncertainty assessments. Annual biomass carbon increments must be distinguished from the net carbon balance throughout the life cycle, necessitating precise accounting limits and oversight of related emissions. Regional evaluations ought to include vegetation and atmospheric chemistry models alongside field data to measure carbon results under plausible planting conditions, in addition to gradual O₃ and SOA development.

Forecasts should be corroborated through independent locations and temporal spans, encompassing evaluations of canopy architecture, species diversity, planting density, air circulation, water accessibility, and human interaction. The suggested health advantages also necessitate proof of exposure and dose–response correlations. Such validation is essential to ascertain the circumstances under which this framework can guide practical urban design.