Particulate matter (PM) pollution is an urgent urban environmental problem. However, plants can mitigate this pollution by filtering the air. Combining the PM capture capacity with the air pollution tolerance could be better evaluate the suitability of greenbelt plants. We selected nine dominant roadside plants growing at two sites in Beijing, and compared their PM capture capacity, morphological characteristics, biochemical characteristics, and air pollution tolerance index (APTI). Sophora japonica had the highest PM capture capacity (362.98 μg cm⁻²), and its wax layers could trap large amounts of PM₂.₅; this high efficiency is important for successful phytoremediation. Sophora japonica. Sabina chinensis, Ulmus pumila, and Euonymus japonicus also showed relatively high PM capture capacity. This is due to their complex cuticular wax layers, short petioles, rough surfaces, high stomata density, and dense canopy structures which reduce the possibility of resuspension of captured PM. Amount of PM captured per unit leaf area had a significant positive effect on the degree of membrane lipid peroxidation, indicating that species with high PM capture capacity suffered higher oxidative stresses. Air pollution showed the strongest negative effect size on chlorophyll contents of E. japonicas. While, S. japonica, S. chinensis, and U. pumila could prevent chlorophyll content decline under severe oxidative stress. Sophora japonica also had the highest APTI at both sites, indicating this species had the greatest tolerance to air pollution. Our findings suggest that S. japonica would be the most suitable species for greenbelt construction in Beijing, followed by S. chinensis, E. japonicus, and U. pumila.

Trees and other vegetation have been advocated as a mitigation measure for urban air pollution mainly due to the fact that they passively filter particles from the air. However, mounting evidence suggests that vegetation may also worsen air quality by slowing the dispersion of pollutants and by producing volatile organic compounds that contribute to formation of ozone and other secondary pollutants. We monitored nanoparticle (>10 nm) counts along distance gradients away from major roads along paired transects across open and forested landscapes in Baltimore (USA), Helsinki (Finland) and Shenyang (China) − i.e. sites in three biomes with different pollution levels − using condensation particle counters. Mean particle number concentrations averaged across all sampling sites were clearly reduced (15%) by the presence of forest cover only in Helsinki. For Baltimore and Shenyang, levels showed no significant difference between the open and forested transects at any of the sampling distances. This suggests that nanoparticle deposition on trees is often counterbalanced by other factors, including differing flow fields and aerosol processes under varying meteorological conditions. Similarly, consistent differences in high frequency data patterns between the transects were detected only in Helsinki. No correlations between nanoparticle concentrations and solar radiation or local wind speed as affecting nanoparticle abundances were found, but they were to some extent associated with canopy closure. These data add to the accumulating evidence according to which trees do not necessarily improve air quality in near-road environments.

We investigated isoprene (ISO) emission and gas exchange in leaves from different positions along the vertical canopy profile of poplar saplings (Populus euramericana cv. ‘74/76’). For a growing season, plants were subjected to four N treatments, control (NC, no N addition), low N (LN, 50 kg N ha⁻¹year⁻¹), middle N (MN, 100 kg N ha⁻¹year⁻¹), high N (HN, 200 kg N ha⁻¹year⁻¹) and three O₃ treatments (CF, charcoal-filtered ambient air; NF, non-filtered ambient air; NF + O₃, NF + 40 ppb O₃). Our results showed the effects of O₃ and/or N on standardized ISO rate (ISOᵣₐₜₑ) and photosynthetic parameters differed along with the leaf position, with larger negative effects of O₃ and positive effects of N on ISOᵣₐₜₑ and photosynthetic parameters in the older leaves. Expanded young leaves were insensitive to both treatments even at very high O₃ concentration (67 ppb as 10-h average) and HN treatment. Significant O₃ × N interactions were only found in middle and lower leaves, where ISOᵣₐₜₑ declined by O₃ just when N was limited (NC and LN). With increasing light-saturated photosynthesis and chlorophyll content, ISOᵣₐₜₑ was reduced in the upper leaves but on the contrary increased in middle and lower leaves. The responses of ISOᵣₐₜₑ to AOT40 (accumulated exposure to hourly O₃ concentrations > 40 ppb) and PODY (accumulative stomatal uptake of O₃ > Y nmol O₃ m⁻² PLA s⁻¹) were not significant in upper leaves, but ISOᵣₐₜₑ significantly decreased with increasing AOT40 or PODY under limited N supply in middle leaves but at all N levels in lower leaves. Overall, ISOᵣₐₜₑ changed along the vertical canopy profile in response to combined O₃ and N exposure, a behavior that should be incorporated into multi-layer canopy models. Our results are relevant for modelling regional isoprene emissions under current and future O₃ pollution and N deposition scenarios.

The emission of mercury (Hg) from cropland soil greatly affects the global Hg cycle. Combinations of different crop cultivars and planting densities will result in different light transmittance under canopies, which directly affects the solar and heat radiation flux received by the soil surface below crops. In turn, this might lead to differences in the soil–air total gaseous mercury (TGM) exchange under different cropping patterns. However, soil–air TGM exchange fluxes in croplands under differing canopies have been poorly investigated. Here, a one-year observation of TGM exchange flux was conducted for cropland soils covering five different crop cultivars and three planting densities in North China Plain using the dynamic flux chamber method. The results showed that light transmittance under the canopies was the key control on soil–air TGM exchange fluxes. High light transmittance can enhance soil TGM emission rates and increase the magnitude of diurnal variations in soil–air TGM exchange fluxes. Furthermore, we found that there were piecewise–function relationships (Peak function–constant equation) between light transmittance under the different canopies and the numbers of days after crop sowing. The soil–air TGM exchange fluxes showed a parabolic response to changes in light transmittance under the different canopies. A second-order model was established for the response relationship between soil–air TGM exchange flux and soil Hg concentration, total solar radiation above the canopy, and numbers of days after sowing. The estimated annual average soil–air TGM exchange flux was 5.46 ± 21.69 ng m⁻² h⁻¹ at corn–wheat rotation cropland with 30 cm row spacing using this second-order model. Our results might a data reference and a promising foundation for future model development of soil–air TGM exchange in croplands under different crop cultivars and planting densities.

Atmospheric pollution could significantly alter tree growth independently and synergistically with meteorological conditions. North China offers a natural experiment for studying how plant growth responds to air pollution under different meteorological conditions, where rapid economic growth has led to severe air pollution and climate changes increase drought stress. Using a single aspen clone (Populus euramericana Neva.) as a ‘phytometer’, we conducted three experiments to monitor aspen leaf photosynthesis and stem growth during in situ exposure to atmospheric pollutants along the urban-rural gradient around Beijing. We used stepwise model selection to select the best multiple linear model, and we used binned regression to estimate the effects of air pollutants, atmospheric moisture stress and their interactions on aspen leaf photosynthesis and growth. Our results indicated that ozone (O₃) and vapor pressure deficit (VPD) inhibited leaf photosynthesis and stem growth. The interactive effect of O₃ and VPD resulted in a synergistic response: as the concentration of O₃ increased, the negative impact of VPD on leaf photosynthesis and stem growth became more severe. We also found that nitrogen (N) deposition had a positive effect on stem growth, which may have been caused by an increase in canopy N uptake, although this hypothesis needs to be confirmed by further studies. The positive impact of aerosol loading may be due to diffuse radiation fertilization effects. Given the decline in aerosols and N deposition amidst increases in O₃ concentration and drought risk, the negative effects of atmospheric pollution on tree growth may be aggravated in North China. In addition, the interaction between O₃ and VPD may lead to a further reduction in ecosystem productivity.

Urban vegetation is associated with numerous public health benefits; however, urban tree canopies may be threatened by fugitive methane exposure from leaky natural gas distribution systems. Despite anecdotal evidence of the harmful impacts of natural gas leaks on urban tree decline, the relationship between soil gas exposure and tree health has not been formally quantified in an urban setting. We conducted a case-control study to compare soil natural gas exposure in sidewalk tree pits of healthy and dead or dying trees in Chelsea, Massachusetts, during summer 2019. We measured soil concentrations of methane and oxygen at four points around the trunks of 84 case and 97 control trees. We determined that case trees had 30 times the odds of being exposed to detectable levels of soil methane relative to the control trees sampled (95% CI = 3.93, 229). Among tree pits with elevated soil gas, we also found that methane concentrations were highest on the side of the tree pit closest to the street. These results contribute evidence to support the widespread belief that soil methane exposure can negatively impact urban tree health. They also suggest that fugitive methane leakage from urban natural gas distribution systems beneath the street surface may be responsible for elevated soil gas concentrations in sidewalk tree pits and subsequent tree death.

Mosses take up nitrogen (N) mainly from precipitation through their surfaces, which makes them competent bioindicators of N deposition. We found positive relationships between the total N concentration (mossN%) of common terrestrial moss species (feather mosses Pleurozium schreberi and Hylocomium splendens, and a group of Dicranum species) and different forms of N deposition in 11–16 coniferous forests with low N deposition load in Finland. The mosses were collected either inside (Dicranum group) or both inside and outside (feather mosses) the forests. Deposition was monitored in situ as bulk deposition (BD) and stand throughfall (TF) and detected for ammonium (NH₄⁺-N), nitrate (NO₃⁻-N), dissolved organic N (DON), and total N (Nₜₒₜ, kg ha⁻¹yr⁻¹). Nₜₒₜ deposition was lower in TF than BD indicating that tree canopies absorbed N from deposition in N limited boreal stands. However, mossN% was higher inside than outside the forests. In regression equations, inorganic N in BD predicted best the mossN% in openings, while DON in TF explained most variation of mossN% in forests. An asymptotic form of mossN% vs. TF Nₜₒₜ curves in forests and free NH₄⁺-N accumulation in tissues in the southern plots suggested mosses were near the N saturation state already at the Nₜₒₜ deposition level of 3–5 kg ha⁻¹yr⁻¹. N leachate from ground litterfall apparently also contributed the N supply of mosses. Our study yielded new information on the sensitivity of boreal mosses to low N deposition and their response to different N forms in canopy TF entering moss layer. The equations predicting the Nₜₒₜ deposition with mossN% showed a good fit both in forest sites and openings, especially in case of P. schreberi. However, the open site mossN% is a preferable predictor of N deposition in monitoring studies to minimize the effect of tree canopies and N leachate from litterfall on the estimates.

Shortly after an atmospheric release, the interception of radionuclides by crop canopies represents the main uptake pathway leading to food chain contamination. The food chain models currently used in European emergency decision support systems require a large number of input parameters, which inevitably leads to high model complexity. In this study, we have established a new relationship for wet deposited radionuclides to simplify the current modelling approaches. This relationship is based on the hypothesis that the stage of plant development is the key factor governing the interception of radionuclides by crops having horizontally oriented leaves (planophile crops). The interception fraction (f) and the leaf area index normalized (fLAI) and mass normalized (fB) interception fractions were assessed for spinach (Spinacia oleracea) and radish (Raphanus sativus) at different stages of plant development and for different contamination treatments and plant densities. A database of 191 f values for Cs-137 and Th-229 was built and complemented with existing literature covering various radionuclides and crops with similar canopy structure. The overall f increased with the plant growth, while the reverse was observed for fB. The fLAI significantly decreased by doubling the contaminated rainfall deposited. Fitting a multiple linear regression to predict the f value as a function of the standing biomass (B), and the radionuclide form (anion and cation) led to a better estimation of the interception (R² = 81%) than the ECOSYS-87 model (R² = 35%). Hence, the simplified modelling approach here proposed seems to be a suitable risk assessment tool as fewer parameters will minimize the model complexity and facilitate the decision-making procedures in case of emergencies, when countermeasures need to be identified and implemented promptly.

Mangroves are highly susceptible to oil exposure. Depending on the severity, oil exposure can result in initial defoliation and eventual recovery through to mass mortality and complete loss of habitat. Some aspects of the impact of oil on mangroves and their recovery are well studied, but the focus has been on short-term responses, and the understanding of the longer-term trajectory of mangrove recovery from oiling is very limited. Here, we combine field results from sampling in the two years following a significant oiling event, with analysis of canopy cover in aerial images from before the event to 26 years afterwards. Approximately 100 ha of a monospecific stand of Avicenna marina mangroves were oiled as a result of a spill from the Era tanker in Spencer Gulf in southern Australia in September 1992. While lightly oiled trees made a full recovery, trees in heavily oiled areas experienced mass defoliation and ultimately mortality within several months of the oiling event. An analysis of aerial images indicated that there was no recovery in heavily oiled areas for 10 years following the oiling event. Between 10 and 25 years, seedling establishment and growth saw canopy cover increase to 35% of pre-oiling cover within heavily oiled areas. Predictive modelling estimates that complete recovery of mangroves to pre-oiling cover will take 55 years (median prediction in 2047). Our findings indicate that although mangroves can recover following a heavy oiling event, the rate of recovery can be slow, with full recovery in the order of half a century, much longer than has previously been anticipated.

Prior studies indicated salt marsh periwinkles (Littoraria irrorata) were strongly impacted in heavily oiled marshes for at least 5 years following the Deepwater Horizon oil spill. Here, we detail longer-term effects and recovery over nine years. Our analysis found that neither density nor population size structure recovered at heavily oiled sites where snails were smaller and variability in size structure and density was increased. Total aboveground live plant biomass and stem density remained lower over time in heavily oiled marshes, and we speculate that the resulting more open canopy stimulated benthic microalgal production contributing to high spring periwinkle densities or that the lower stem density reduced the ability of subadults and small adults to escape predation. Our data indicate that periwinkle population recovery may take one to two decades after the oil spill at moderately oiled and heavily oiled sites, respectively.