Journal of Environmental Management xxx (2018) xxx-xxx

Contents lists available at ScienceDirect

Journal of Environmental Management

journal homepage: www.elsevier.com

Research article Potential of trees leaf/ bark to control atmospheric metals in a gas and petrochemical zone

Mojgan Safaria⁠ , Bahman Ramavandia⁠ ,⁠ ∗⁠ , Ali Mohammad Sanatib⁠ ,⁠ ∗⁠ ∗, George A. Sorialc⁠ , Seyedenayat Hashemia⁠ , Saeid Tahmasebid⁠ a Environmental Health Engineering Department, Faculty of Health, Bushehr University of Medical Sciences, Bushehr, Iran b Department of Environmental Science, Persian Gulf Research Institute, Persian Gulf University, Bushehr, Iran c Environmental Engineering Program, Department of Chemical and Environmental Engineering, College of Engineering and Applied Science, University of Cincinnati, Cincinnati, OH, 45221-0012, USA d PROOF Department of Statistics, Persian Gulf University, Bushehr, Iran

ARTICLE INFO ABSTRACT

Keywords: Leaf and bark of trees are tools for assessing the effects of the heavy metals pollution and monitoring the en- Asaloyeh vironmental air quality. The aim of this study was to evaluate the presence of Ni, Pb, V, and Co metals in four Leaf tree/shrub species (Conocarpus erectus, Nerium oleander, Bougainvillea spectabilis willd, and Hibiscus rosa-sinensis) Bark in the heavily industrial zone of Asaloyeh, Iran. Two industrial zones (sites 1 and 2), two urban areas (sites 3 Heavy metal and 4), and two rural areas (sites 5 and 6) in the Asaloyeh industrial zone and an uncontaminated area as a Metal accumulation index control were selected. Sampling from leaf and bark of trees was carried out in spring 2016. The metals content in the washed and unwashed leaf and bark was investigated. The results showed that four studied metals in N. oleander, C. erectus, and B. spectabilis willd in all case sites were significantly higher than that of in the control site (p<0.05). The highest concentration of metals was found in sites 3, 4, and 6; this was due to dispersion of the pollutants from industrial environments by dominant winds. The highest comprehensive bio-concentration index (CBCI) was found in leaf (0.37) and bark (0.12) of N. oleander. The maximum metal accumulation index (MAI) in the samples was found in leaf of N. oleander (1.58) and in bark of H. rosa-sinensis (1.95). The maximum bio-concentration factor (BCF) was seen for cobalt metal in the N. oleander leaf (0.89). The nickel concentration in washed-leaf samples of C. erectus was measured to be 49.64% of unwashed one. In general, the N. oleander and C. erectus species were found to have the highest absorption rate from the atmosphere and soil than other studied species, and are very suitable tools for managing air pollution in highly industrialized areas.

small industries (Delshab et al., 2017; Naeimi et al., 2018). The emit- 1. Introduction ted heavy metals in the atmosphere then adhere to the particulate mat- ter and dust. These materials finally settle on the ground as well as Recently, unorganized industrial activities and rapid development of trees (Ugolini et al., 2013). The distribution of the heavy metals de- cities have caused the emission of various pollutants especially heavy pends on the site topography, wind velocity, particulate matter diame- metals into the atmosphere (Kafaei et al., 2017). Heavy metals like ter, and surface characteristics of the materials on which heavy metals Pb, Co, Ni, and V are serious concerns for the environment and hu- are deposited (Norouzi et al., 2016). For example, specific area, rough- man life (Alsaleh et al., 2018; Fernández-Calviño et al., 2017). Co, Ni, ness, and epicuticular waxes of leaf and bark of trees influence the par- and V are released to the environment from sources like petrochemical ticle deposition. Furthermore, the heavy metals deposited on the land and gas industries and Pb is originated from commercial activities and may be relocated by wind or runoff and thus again enter to the vegetal

Abbreviations: CBCI, Comprehensive bio-concentration index; MAI, Maximum metal accumulation index; BCF, Bio-concentration factor; AOM, Air-originated metals; PSEEZ, Pars Spe- cial Economic Energy Zone; BS, Background site; EC, Electrical conductivity; ICP-AES, Inductively Coupled Plasma Atomic Emission Spectroscopy. ∗ Corresponding author. EnvironmentalUNCORRECTEDHealth Engineering Department, Faculty of Health, Bushehr University of Medical Sciences, Mobaraki Street, 7518759577, Bushehr, Iran. ∗∗ Corresponding author. Email addresses: [email protected], [email protected] (B. Ramavandi); [email protected] (A.M. Sanati) https://doi.org/10.1016/j.jenvman.2018.05.026 Received 14 December 2017; Received in revised form 1 May 2018; Accepted 10 May 2018 Available online xxx 0301-4797/ © 2018. M. Safari et al. Journal of Environmental Management xxx (2018) xxx-xxx tissue through different sources such as soil, water, and air (Bilo et al., 2017). Therefore, it can be concluded that the human health is affected 2. Materials and methods not only by the polluted air but also by the polluted soil or water. Hence, a continuous monitoring of the environment is required. 2.1. Area of study Biomonitoring is superior to direct analyzing samples due to the use of inexpensive equipment, easy sampling procedures, and long-term The study was done in Asaloyeh area which has more than 30,000ha monitoring (Ramachandra et al., 2018; Serbula et al., 2012). Further, of industrial areas and is limited from the north by mountains, from the according to literature (Allahabadi et al., 2017; Berdonces et al., 2017), south by the Persian Gulf, from the east by the Gavbandi city, and from using of tree and shrub as biomonitors have more interesting than the west by the Tahery port. The climate of this area is warm and the lichens and mosses as they reflect the heavy metals contamination of humidity is >80% during spring, summer, and early fall seasons. Other the environment during a long time. Bark and leaf have been widely climatic characteristics of the study area include average monthly tem- used for metals biomonitoring than other parts of the tree/shrub like perature 27.3°C, average annual precipitation 250mm, and average so- buds, flowers, and needle as their structures can absorb more air pollu- lar radiation 20 MJ/m2⁠ d. The heating degree days (HDD) and cooling tants (Matin et al., 2016; Oliva and Mingorance, 2006). The bark and degree days (CDD) for Asaloyeh area are presented in Fig. S1. For these leaf could provide more surface area for deposition of the particulate cases, base degree for heating is 18°C and for cooling is 21°C. Climate 2⁠ matter. Song et al. (2015) has shown that about 148μg/cm of particu- conditions, topography, and presence of various industries have led to late matter is deposited on the leaf of studied trees. In literature (Drava the regional air pollution. The industrialized zone of Asaloyeh area is et al., 2017; Parraga-Aguado et al., 2014), the trees bark has also been also known as Pars Special Economic Energy Zone (PSEEZ). There are proposed as pollution time capsules. Thus, the leaf and bark of trees/ 20 villages, 2 cities, 18 active petrochemical companies, 14 gas refinery shrubs can be used as suitable biomonitors for heavy metals pollution companies, and about 60,000 inhabitants are living or working in this (Allahabadi et al., 2017; Baldantoni et al., 2014). However, despite of area. The map of sampling sites is depicted in Fig. 1. The sampling sites increasing research on heavy metal biomonitoring using tree/shrub in included: PROOF urban areas (Allahabadi et al., 2017; Norouzi et al., 2016; Sawidis et al., 2011), the study of heavy metals biomonitoring in the gas and petro- (I) two industrial zones (petrochemical companies site (site 1) and gas chemical industrialized zone in Asia and especially in Iran is very rare. companies (site 2) close to Nakhl Taghi city and continues up to The emitted pollutants from industrial zones not only can affect the in- near Shirinoo village); dustry itself but also the surrounding urban and rural ecosystems might (II) two urban zones (Nakhl Taghi city (site 3) and Asaloyeh city (site be affected. 4)) both are located in the south of PSEEZ; and Many trees types have been studied for biomonitoring metal pollu- (III) two rural zones (Shirinoo village (site 5), and Bozbaz village (site tion and their advantages/disadvantages for this purpose have been in- 6)) which are in the northwest and southeast of PSEEZ, respec- dicated (Birke et al., 2018; Sawidis et al., 2011). The researchers are tively (Fig. 1). looking for trees/shrubs with higher monitoring capacity to monitor pollutants. In this work the effect of heavy metals pollution in the Asa- In addition, the Gas and Petrochemical Town, where the family of loyeh industrial zone was evaluated by using native tree/shrub of Cono- petrochemical and gas company's employees live, was selected as the carpus erectus, Bougainvillea spectabilis willd, Hibiscus rosa-sinensis, and background site (BS). This site was selected due to lower natural and Nerium oleander, which have so far been rarely used as biomonitoring anthropogenic pollution sources, no dust during most of the year, less tools. The Asaloyeh industrial zone is a very busy area at the north- traffic, and reasonable distance from the city center. The Gas and Petro- ern part of the Persian Gulf which contains many gas and petrochemi- chemical Town has a population of about 8000 and an altitude of 850m. cal companies and also has highest records of unloading-loading ports The latitude and longitude of all sampling points are presented in Table among all Middle East ones. Thus, these activities as well as downstream S1. companies are held responsible for sever accumulation of pollutants like heavy metals in the region. Environmental pollution biomonitoring in 2.2. Reagents gas and petrochemical companies can provide important information for pollutant management in industrial sites with heavy metal contam- All acids and reagents used in this study such as perchloric acid inants. Heavy metals biomonitoring by trees can also provide helpful (90%) and nitric acid (65%) were of supra pure quality (Merck Co., information for the design of international networks monitoring deposi- Germany). All glassware was acid washed by soaking in dilute nitric tion. acid and rinsed with deionized water before use. The required solutions Accordingly, the aim of current work was an evaluation of air qual- for digestion process were prepared using ultrapure water (resistivity ity in the Asaloyeh industrial zone and its surroundings using bark and 18.2MΩcm). leaf of C. erectus, B. spectabilis willd, H. rosa-sinensis, and N. oleander as a bioindicator. For this purpose, the comprehensive bio-concentra- 2.3. Leaf, bark, and soil sampling method tion index (CBCI), maximum metal accumulation index (MAI), bio-con- centration factor (BCF), and air-originated metals (AOM) were calcu- In this work, the samples were included leaf and bark of tree/shrub lated based on the atmospheric contaminant load of the washed and un- and the soil of around trees trunk. The trees/shrubs with most dis- washed bark or leaf of trees/shrubs. This study tries to examine the in- persion in the studied areas such as Conocarpus erectus, Bougainvillea fluence of gas and petrochemical industries on the contaminant level of spectabilis willd, Hibiscus rosa-sinensis, and Nerium oleander were selected the rural and cities in the Asaloyeh zone. The findings are compared to (see Fig. S2). To avoid the intervention of micro-scale differences in lo- those of no contamination site in Iran. The results could be beneficial as cations, an area of 100m×100m in each site was considered for sam- preliminary reference values for heavy metals concentrations in indus- pling. trial environments for future monitoring. UNCORRECTEDIn order to have the same integration time of metal pollution, trees with the same age were used for sampling. For this purpose, each tree species with similar height, trunk diameter, and cone were selected. All selected trees had around five years old at the time of sampling. In gen- eral, the trunk diameter and height of C. erectus were 25cm and 6m, B. spectabilis willd were 4.5cm and 2.5m, H. rosa-sinensis were 5cm and

2 M. Safari et al. Journal of Environmental Management xxx (2018) xxx-xxx

PROOF

Fig. 1. Map of sampling sites and their distance to the nearest gas or petrochemical company (Image from Google Earth©⁠ software).

3m, and N. oleander were 4cm and 2.5m. For leaf sampling, leaves The leaf and bark samples were collected from all directions (N, S, E, with identical length, fully expanded, and with no spots or abnormal and W) of the trees/shrubs. After taking 25g sample of leaf and bark appearance (chlorosis or necrosis) were considered as sample. Trees of trees/shrubs with a cutter, the samples were placed in polyethylene with unusual leaf and bark (for example wrinkled or yellow leaf and bags and stored at 4°C until drying. For each site, three trees were sam- pestle bark) were excluded. Also, the trees that had honeydew were pled and the same trees were referred for the second and third sampling. not considered. Care was also taken to avoid selecting leaf with im- Sample collection was done three times in spring (from 20 March to 18 perfections like bird dropping,UNCORRECTEDinsect infestation, and pesticide treat- June) 2016 and each sampling time was done at least 25 days after rain, ment. Leaves were randomly collected from the lower two-third of the however, only one rainfall (1mm) was occurred after the first sampling. canopy of each tree/shrub. The bark sample was taken from the middle Each sampling period lasted 2 days. of the tree/shrub trunk and at about 50cm above ground. If the lichens Soil samples were also taken within a radius of 10–15cm from the were present on the bark surface, it was not considered for sampling. trunk of each tree/shrub. To do this, the top soil, including manure,

3 M. Safari et al. Journal of Environmental Management xxx (2018) xxx-xxx twigs, and dried leaves were removed and then soil samples were taken Each sample was digested three times and the average of measure- from depth of 10–15cm (in an area of 10×10cm2⁠ ). The soil samples ments is reported herein. A control sample was also used beside each were then stored in polyethylene bags and transported to the laboratory sample to determine the background pollution during digestion. To con- and kept at 4°C. firm the accuracy of the methodology and to ensure the extraction of heavy metals from the tree and soil samples, the standard solution of 2.4. Digestion of leaf, bark, and soil samples each studied heavy metal was used and the recovery rates of the metals were obtained. Only the element with determined certified value and Leaf and bark samples were divided in 2 groups; a group was washed average recovery of 80–120% was accepted for data analysis. The recov- with ultrapure deionized water to remove dust and particulate matter ery rates for all metals in this work fell within the range of 87–93%. The and the other one was left unwashed. To do the digestion process, first standard deviations of duplicate samples were not greater than 4.5%. both types of samples were dried in the oven at a temperature of 105°C All materials, acids, and solutions used for the analysis of heavy metals for 24h. Dried samples were then manually ground and passed through were of analytical grade and ultra pure. a sieve with a hole diameter of 0.5mm. The mill and sieve were thor- oughly cleaned after each grinding and sieving to avoid any cross con- 2.5. Analysis of soil sample tamination between samples. The amount of 1±0.0001g from sieved powder samples of bark and leaf was placed in the acid wash tubes For humidity measurement, the soil samples were sieved with an and 10mL of 65% nitric acid (Merck Co.) was added to it (Liang et al., ASTM sieve with diameter of 2mm and the samples were weighed. Sam- 2017). In the case of soil samples, 8mL of 65% nitric acid and 2mL of ples were then placed in the oven at 105°C for 24h and were weighed concentrated perchloric acid were added to the samples. For digestion again. The amount of moisture was calculated from the difference be- of the samples, the solution was placed at room temperature overnight tween primary and secondary weights. For the pH measurement, 10g (≈12h) after that it was placed for 4hat 100°C and then 4hat 140°C soil sample was mixed with 10mL double distilled water and after 1h the pH of solution was meaPROOFsured (Pen-Mouratov et al., 2008). For the until the solution color was clear. After cooling, the solution was di- luted by ultrapure deionized water to 50mL and then passed through electrical conductivity (EC) measurement of soil sample a ratio of 1:5 0.42μm-Whatman filter paper until 25mL of the filtrate volume was soil and double distilled water was prepared and after 1h the conductiv- provided. The stages of preparation and digestion of the samples are ity of the solution was analyzed with an EC meter (AZ 86503, Taiwan) shown in Fig. 2. (Pen-Mouratov et al., 2008). The soil pH and EC are shown in Table S1.

2.6. Heavy metals measurement

The concentration of Pb, Co, Ni, and V in soil, leaf, and bark were eventually measured using Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-AES 4100, Agilent), which induces excited atoms and ions to emit electromagnetic radiation at wavelengths char- acteristic of a particular element (167–785nm). The intensity of the emission is indicative of the concentration of the metal in the sample. Some operating conditions for the ICP-AES instrument were: plasma gas flow rate 18L/min, auxiliary gas flow rate 0L/min, nebulizer (carrier) gas pressure 2.10kg/cm2⁠ , peristaltic pump flow rate 1.4mL/min, detec- tor integration time 5s, and number of integrations per solution 3.

2.7. Data analysis

For data analysis, three factors including bio-concentration factor (BCF), comprehensive bio-concentration index (CBCI), and metal accu- mulation index (MAI) were used to evaluate the overall performance of heavy metals accumulation in the trees/shrubs in each studied area (Hu et al., 2014).

2.7.1. BCF and CBCI BCF is the ratio of total heavy metal concentration in the leaf or bark sample and soil sample at radius of 15cm from the tree/shrub trunk. This factor has been used to evaluate the ability of trees/shrubs to accu- mulate heavy metals from soil (Allahabadi et al., 2017; Shi et al., 2011). The equation of the BCF is described as follow:

(1)

UNCORRECTEDwhere [C]L,B⁠ and [C]S⁠ are the heavy metal concentration in leaf or bark and soil samples, respectively. BCF >1 shows that tree/shrub is enriched in studied metals (accumulator), BCF<1 indicates that the tree/shrub excludes the metals from the uptake, and BCF=1 indicates

Fig. 2. Flowchart for the preparation and digestion of tree and soil samples.

4 M. Safari et al. Journal of Environmental Management xxx (2018) xxx-xxx that the tree/shrub could be used as an indicator for the metal (Liang et (12.77mg/kg) and the lowest (0.14mg/kg) concentrations of nickel in al., 2017). the leaf were found in C. erectus in site 6 and BS, respectively. The high- CBCI is defined as the capability of tree for comprehensive accumu- est concentration of nickel in the bark was found in site 5 in N. olean- lation of multiple heavy metals. To estimate the CBCI, some steps are der (4.78mg/kg) and the lowest nickel concentration was also seen in required. First, the fuzzy set M, M= (M1⁠ , M2⁠ , M3⁠ …, Mi⁠ ), is estimated. M BS in the same species (0.22mg/kg). Norouzi et al. (2015) reported a denotes the level of comprehensive accumulation ability for each tree/ maximum nickel concentration of 12.70mg/kg in the unwashed leaf of shrub, and Mi⁠ denotes various heavy metal pollution factors. In the sec- Platanus orientalis L. during biological monitoring of heavy metals trans- ond step, the amount of fuzzy membership function μ(X) is calculated ported by dust in Isfahan, Iran. Permissible or normal nickel concentra- as “μ(X) = X − Xmin/Xmax − Xmin”. X is the BCF for given heavy metal, tion in trees is in the range of 0.1–5mg/kg and its toxic concentration Xmin⁠ and Xmax⁠ are the minimum and maximum amounts of the BCF for is 10–100mg/kg (Kabata-Pendias, 2010). In the present study, the con- certain heavy metal, which represent the highest and lowest compre- centration of nickel in all unwashed bark samples was less than 5mg/ hensive accumulation factors for heavy metals, respectively. It could be kg, which is below the licit range. But the nickel concentration in the considered that Xmin⁠ =0 and Xmax⁠ =1. Finally, CBCI can be calculated unwashed leaf of the C. erectus species at site 3 and 6 exceeded 10mg/ from Eq. (2). kg, so it can withstand the high concentrations of nickel. In N. oleander species at site 1, 4, 5, and 6, the concentration of nickel was higher than (2) 5mg/kg but did not reach the limit of 10mg/kg, which is the beginning of toxic concentration. Therefore, the N. oleander species has absorbed nickel more than the normal nickel concentration. where N is the total number of heavy metals and μi⁠ is μ(x) of heavy In Fig. S4, the concentration of lead in leaf and bark of 4 tree/shrub metal i. species is demonstrated. The maximum concentration of lead was seen in the N. oleander leaf located at site 4 (26.04mg/kg) and in the B. 2.7.2. MAI PROOF spectabilis willd bark in site 6 (27.75mg/kg). In contrast, the lowest con- MAI was first introduced by Liu et al. (2007) to assess the ability of centrations of lead were determined in BS in the leaf of H. rosa-sinensis different tree species to accumulate the atmospheric heavy metals. Thus, (0.08mg/kg) and in the same tree bark (0.29mg/kg). Also, in this study, MAI is used in this study to compare the ability of different tree/shrub the lead concentration in the washed leaf and bark samples was less species to accumulate the heavy metals in studied sites. MAI was esti- than that of unwashed one. In a study by Zhao et al. (2014) in China, mated using Eq. (3). the concentration of lead in the leaf of tree species from the studied sites exceeded 50mg/kg. Also, Matin et al. (2016) identified the lead concen- (3) tration of 35–52mg/kg in pine leaf samples during monitoring the cad- mium, lead, and arsenic in industrial regions in Turkey. Swaileh et al. (2006) in a study of heavy metals monitoring in several types of trees where Ij⁠ =x/Δx is the sub-index for variable j, x is the mean value of around the urban road in Oman found that lead concentration in N. ole- studied metal and Δx is the standard deviation of the measurements. ander is 2.66mg/kg and Cupressus sempervirens species have the highest Pb concentration (3.44mg/kg). Permissible or normal lead concentra- 2.7.3. Air-originated metals (AOM) tion in trees is between 0.1 and 5mg/kg and its toxic concentration is To determine the contribution ratio of the heavy metals originated 10–100mg/kg (Kabata-Pendias, 2010). In the present study, the concen- from the atmosphere, the AOM factor was used, which was estimated tration of lead in the following samples has exceeded the normal limit from the following equation: and is toxic: site 1 (N. oleander and B. spectabilis willd), site 2 (C. erec- tus and N. oleander), site 4 (N. oleander and B. spectabilis willd), site 5 (4) all species, and site 6 all species with the exception of H. rosa-sinensis. In most species, the lead concentration is higher than nickel indicating where [C]unwashed⁠ and [C]washed⁠ are the heavy metal concentrations in the more species are contaminated with lead. It has been reported that less unwashed and washed leaf or bark samples, respectively. than 2% of lead in leaves and branches of Norway spruce originates from roots and about 98% has the atmospheric origin (Hovmand et al., 2009). 2.8. Statistical analysis Other studies have shown that the transfer of lead from root to leaf is not a major trend of contamination and trees can easily absorb atmos- The statistical analyses were done by using the Statistical Package pheric lead after depositing Pb-laden particulate matter on their leaves for the Social Sciences (SPSS, v.21). The Kolmogorov-Smirnov (K-S) sta- (Ribeiro de Souza et al., 2012; Turer et al., 2001). tistic and its corresponding p-value were considered for testing the nor- The results of vanadium concentration in the leaf and bark of trees/ mality of data. The p-value greater than 0.05 indicates the probability shrubs are presented in Fig. S5. The highest concentrations of vanadium distribution of data is approximately normal. One-way ANOVA was ap- were found in site 4 in the N. oleander leaf (23.17mg/kg) and in the plied to study the significant difference between the variables in the in- bark sample of the C. erectus species in site 6 (14.43mg/kg). While the dustrial zone of Asaloyeh and the background site, different species, and minimum vanadium concentrations were seen in the B. spectabilis willd washed and unwashed leaf or bark samples. If there was a significant bark (0.20mg/kg) and in the H. rosa-sinensis leaf (0.08mg/kg) in BS. difference, multiple comparisons (Tukey test) were used to separate the Onder and Dursun (2006) measured high vanadium concentrations in different groups. Meram region (20.45ppm) and around a cement factory (53.63ppm) in 2003 and 2004, respectively, by detecting airborne heavy metals 3. Result and discussion UNCORRECTEDin Cedrus libani species in Turkey. As shown in Fig. S5, vanadium el- ement was found to be higher in most leaf samples than in bark. It 3.1. Metals concentration in leaf and bark of trees/shrubs was also observed higher V concentration in unwashed samples than in washed ones. Permissible or normal concentration range of vanadium in Fig. S3 shows the concentration of nickel in the leaf and bark of plants and trees is 0.2–1.5mg/kg and its toxic concentration is 5–10mg/ the examined tree/shrub species. As depicted in this figure, the highest kg (Kabata-Pendias, 2010). The concentration of vanadium exceeds the

5 M. Safari et al. Journal of Environmental Management xxx (2018) xxx-xxx normal limits except in the leaf sample of H. rosa-sinensis in site 4 other sites (Table S1) while the EC was vice versa. As a result, the con- (0.78mg/kg) and bark sample of B. spectabilis willd in site 1, 4, and 6. centration of metals in site 3, 5, and 6 was relatively higher than other As presented in Fig. S6, the highest concentrations of cobalt were ob- areas of study. The composition of dust is may be affecting the metals served in the leaf sample of C. erectus in site 4 (21.78mg/kg) and in N. content of tree/shrub (Al-Khlaifat and Al-Khashman, 2007). The height, oleander in site 5 (4.5mg/kg). The lowest concentration of cobalt in BS species, lifetime, and characteristics of the tree/shrub also considered was measured in the leaf of H. rosa-sinensis (0.24mg/kg) and in N. olean- to be effected on the capacity of air pollutants removal by vegetation der (0.32mg/kg). Onder and Dursun (2006) identified the highest cobalt (Hofman et al., 2013). In the study of Liang et al. (2017), the pollution of metals in denser concentrations in the Cedrus libani tree in samples around the cement traffic area was reported to be lower compared to downtown, because plant (0.44mg/kg) and samples from the Meran region (0.39mg/kg). the former is located out of town and the latter is enclosed with tall The cobalt concentration in leaf samples was notably higher than bark. buildings. Metals pollution in the atmosphere of more traffic area can The normal range of cobalt concentration in the trees is 0.02–1mg/kg easily be transmitted to downtown. This conclusion can also be gener- and the toxic range is about 15 50mg/kg (Kabata-Pendias, 2010). In – alized to this study as well, since the petrochemical industry is located this study, the amount of cobalt in most of leaf samples falls within the in the open air, while there are abundant flares in these areas and the toxic range. wind blows transmitted pollutants into urban areas. The highest average concentration of nickel, lead, and vanadium was Urbanization significantly affects the concentration of heavy metals observed in site 5 and 6 in leaf and bark samples, while the highest av- in the trees leaf. In addition to the chemical characteristics of dust par- erage concentration of cobalt was observed in leaf samples in sites 1 ticles, tree features such as leaf surface area, rugged cell wall, and crack and 4 (Table 1). Various factors may affect the content of vegetation on the surface of leaf and bark are effective in differentiating the con- metals in the sampling sites. The first possible factor is the proximity centration of metals (Liang et al., 2017). For example, the greater crack to the highway with the higher traffic of Asaloyeh-Kangan. In a study and the larger surface area of Platanus acerifolia leaf have led to more by Al-Khashman et al. (2011), the highest concentrations of lead, cop- concentration of metals (Liu et al., 2013). The N. oleander and C. erec- per, and nickel in the Phoenix dactylifera L. species were observed in sur- PROOF tus species used in this study have bigger and fluffy leaves with lower rounding of a highway. The second factor is the transfer of pollutants density than the other two species, while the species of H. rosa-sinensis from contamination sources (e.g., gas and petrochemical companies in and B. spectabilis willd have smooth and shiny leaves. For this reason, the area of Asaloyeh) by the winds. A number of studies (Al-Khlaifat the amount of metals in the former was higher than the latter. and Al-Khashman, 2007; Huang et al., 2018) measured high level of According to Table 2 and the results of one-way analysis of variance heavy metals in the leaf of trees away from the sources of contamina- (ANOVA) between the industrial zone of Asaloyeh and BS, the amount tion. Since the prevailing wind in the studied area (Asaloyeh) has the of metals in the H. rosa-sinensis species is not significant (p>0.05), but northwest direction, it is likely that the winds have transferred pollu- for the other species, the difference is significant. Also, based on statis- tants from sites 1 and 2 (sites of petrochemical and gas companies) to tical analysis, the amount of nickel, lead, and vanadium in the washed sites 3, 4, and 6. In addition, being located next to a highway (see Fig. and unwashed leaf/bark samples varied significantly (p<0.05), while 1) and the presence of downstream industries around site 5 may be the there was no noteworthy difference for cobalt metal. cause of the high pollution of metals in this station. Another factor af- According to Table 2, the results of one-way ANOVA showed that fecting the concentration of tree metals is the pH and EC parameters of there is no significant difference in the amount of nickel metal among soil where the trees grow (Alsaleh et al., 2018; Gondek et al., 2018). In- 4 tree/shrub species in leaf and bark samples (p>0.05). Based on sta- creasing the soil pH reduces the bioavailability of heavy metals to trees. tistical analysis, the difference between the lead and vanadium in leaf On the contrary, increasing the soil EC leads to an increase in the ab- of C. erectus with 3 other species is not significant (p>0.05). Yet in the sorption of metals into plants and trees. In this study, the pH of soil samples taken from site 3, 5, and 6 was found to be lower than the case of cobalt, the leaf of the C. erectus species have a significant differ

Table 1 Heavy metal contents of the leaves and barks at each site (mean±SD, mg/kg).

Place Leaf Bark

Ni Pb V Co Ni Pb V Co

Site 1 2.36±1.99 7.24±5.06 6.13±4.43 10.97±6.75 0.83±0.39 9.42±3.44 2.68±2.24 1.03±0.26 Site 2 3±0.89 9.62±4.04 8.18±3.89 10.49±1.16 1.06±0.42 7.84±2.31 3.41±1 1.26±0.16 Site 3 3.46±3.35 4.66±2.21 3.92±2.08 10.47±6.44 1.4±0.67 7.83±3.97 3.6±3.1 1.49±0.51 Site 4 2.75±2.03 10.17±9.22 8.45±7.79 11.13±7.29 2.14±0.42 11.38±3.52 3.12±2.31 2.18±0.31 Site 5 3.47±1.5 12.36±2.21 10.27±2.26 8.38±7.31 2.98±0.94 13.56±8.05 3.94±1.69 2.48±1.1 Site 6 5.15±4.12 15.34±6.21 12.75±5.33 8.73±1.51 2.21±1.09 15.06±9.6 5.97±4.65 2.28±0.88 BS 0.55±0.49 1.5±0.86 1.16±0.84 1.43±1.52 0.48±0.22 1.79±0.95 0.81±0.41 0.44±0.17

Table 2 Heavy metal contents (mean±SD, mg/kg), MAI and CBCI of the washed leaves and barks.

Part of plant Plant species Ni Pb V Co MAI CBCI

Leaf C. erectus 2.52±1.88 5.11±4.81 3.81±3.71 7.55±6.41 1.15 0.26 N. oleanderUNCORRECTED2.99±1.74 8.36±6.92 6.33±5.29 12.49±5.65 1.58 0.37 B. spectabilis willd 0.77±0.54 9.95±6.66 5.79±4.91 5.73±3.88 1.39 0.20 H. rosa-sinensis 1.03±0.44 4.11±5.26 3.14±3.99 4.15±5.3 1.17 0.18 Bark C. erectus 1.17±0.68 7.66±5.35 2.73±2 1.50±0.57 1.79 0.11 N. oleander 1.11±0.81 6.94±2.88 3.45±2.21 1.47±1.11 1.67 0.12 B. spectabilis willd 1.57±1.19 9.37±8.34 0.88±0.62 1.75±0.99 1.40 0.09 H. rosa-sinensis 1.99±0.94 5.15±6.29 1.24±0.42 1.32±0.69 1.95 0.12

6 M. Safari et al. Journal of Environmental Management xxx (2018) xxx-xxx ence (p<0.05) with H. rosa-sinensis, N. oleander, B. spectabilis willd, and H. rosa-sinensis. There is also a notable difference between the concentration of lead in the bark of B. spectabilis willd and H. rosa-sinensis. The difference in the concentration of vanadium metal between the bark of the C. erec- tus/N. oleander as well as B. spectabilis willd/H. rosa-sinensis is significant (p<0.05). The concentration of cobalt in bark of N. oleander was not significantly higher than bark of other species.

3.2. CBCI

CBCI is used to evaluate the ability of different species of trees to collect heavy metals from the soil. Table 2 lists the CBCI values for the leaf and bark of the examined tree/shrub species. The highest CBCI in leaf samples was found for N. oleander (0.37) and C. erectus (0.26) species, and the lowest CBCI was obtained for H. rosa-sinensis (0.18) and B. spectabilis willd (0.20). In bark samples, the highest CBCI was found for N. oleander (0.12) and H. rosa-sinensis (0.12), and the lowest was seen in B. spectabilis willd (0.09) and C. erectus (0.11). Therefore, the N. oleander has the ability to absorb several metals from the soil. In this study, CBCI values in leaf and bark samples were less than those re- ported by Allahabadi et al. (2017) and Zhao et al. (2014). Furthermore, PROOF CBCI in leaf samples was higher than that in bark, which well coincides with the results reported by Zhao et al. (2014). Accordingly, tree species with a higher CBCI value have the potential to be used as a phyto-reme- diation tool to remove the heavy metals from the soil (Allahabadi et al., 2017).

3.3. MAI

MAI evaluates the ability of different species of trees/shrubs to ac- cumulate atmospheric heavy metals. As shown in Table 2, in bark sam- ples, the highest MAI is for H. rosa-sinensis (1.95) and C. erectus (1.79). Fig. 3. BCF of leaf and bark of tree species at different places. In leaf samples, however, the highest MAI is for N. oleander (1.58) and B. spectabilis willd (1.39). In this study, as was in the Allahabadi et al. the bark sample of B. spectabilis willd (0.15) and N. oleander (0.12), re- (2017) study, the MAI values of bark samples were found to be higher spectively. than leaf samples. In the study of Hu et al. (2014), the MAI value in Based on the results provided above, it is implied that the leaf the leaf of the tree species was reported as 3.89 which higher than that have the higher ability to accumulate heavy metals than the bark. Con- of the leaf samples of our study. The differences in the amount of MAI sidering BCF values, the ability of tree/shrub species to accumulate largely depend on the atmospheric chemistry, atmospheric changes, and heavy metals in the leaf tissues for various metals can be shown as tree characteristics (Hu et al., 2014). Therefore, species that have more Co>V>Ni>Pb and in the bark texture as Ni>V>Pb>Co. The ab- MAI can be used as barriers between polluted and vulnerable areas sorption of metal from the soil is significantly affected by soil pH, soil (Zhai et al., 2016), for which the best tree in this study was found to be organic matter, cation exchange capacity of soil, and plant type (Zhan et the N. oleander species. al., 2014). In the observations of Allahabadi et al. (2017) the BCF value of lead in leaf samples was also seen to be the lowest. Zhao et al. (2014) 3.4. BCF and AOM have been measured the BCF in leaf of 18 tree species which the low- est BCF was form lead metal. Findings of Serbula et al. (2012) on have BCF refers to the ratio of the concentration of heavy metals in dif- showed that the absorption of heavy metals by the Robinia pseudoaca- ferent tissues of tree to the metals concentration in the soil, which can cia L. tree is not significant because the soil pH was alkaline and caused reflect the ability of trees to absorb heavy metals from soil (Zhai et al., less metals mobility. In our study, as the BCF values for all metals in the 2016). BCF results are shown in Fig. 3. For cobalt, the highest BCF was leaf and bark samples were obtained less than 1, so these trees act as ex- seen in leaf samples of N. oleander (0.89) followed by C. erectus (0.65) cluder. This low value could be attributed to the relatively alkaline pH and the lowest BCF concentration for this metal was observed in B. of the soil, which reduces the mobility of metals. spectabilis willd (0.46). For vanadium, the maximum BCF was obtained The values of AOM are plotted in Fig. 4. The AOM factor presents in the leaf of N. oleander (0.44) and the lowest in the leaf of the C. erec- the difference between the amount of metals absorbed by the tree tis- tus species (0.20). The highest BCF for nickel is also detected in the leaf sue and the dust on the leaf or bark surface (Alfani et al., 2000). Heavy of C. erectus (0.25) and N. oleander (0.24) and the lowest was found in metals can be absorbed from the soil through the roots and transferred the leaf of B. spectabilis willd species (0.07). Also, the highest BCF for to the leaves or absorbed to the leaves through the dust. Therefore, lead was found in B. spectabilis willd (0.13) and lowest was seen in the it is very difficult to distinguish the exact origin of elements in tree leaf of C. erectus species (0.06).UNCORRECTEDtissue whether the origin is from soil or air (Liang et al., 2017). The The highest BCF of cobalt metal was calculated in the bark of H. maximum AOM for nickel was obtained as 49.64% and 32.57% in leaf rosa-sinensis (0.16) and the lowest in N. oleander bark (0.14). The high- and bark samples of C. erectus, respectively. Maximum AOM for lead est BCF for vanadium metal was obtained in the N. oleander bark (0.31) was seen in leaf samples of N. oleander (39.12%) and bark samples of and the lowest in B. spectabilis willd (0.07). The highest BCF for nickel H. rosa-sinensis (44.77%). The lowest AOM of lead was, however, ob- is for the bark of H. rosa-sinensis (0.38) and the lowest for the N. olean- served in N. oleander (16.37%) and B. spectabilis willd (16.35%). The der bark (0.10). The highest and lowest BCF for lead was observed in

7 M. Safari et al. Journal of Environmental Management xxx (2018) xxx-xxx

PROOF

Fig. 4. AOM percentage of Ni, Pb, V, and Co for the different plant species. leaf samples of N. oleander (45.46%) and H. rosa-sinensis (46.68%) have metals in unwashed leaf samples (Norouzi et al., 2015; Tomašević et al., shown the highest amount of AOM for vanadium. The lowest AOM of 2011). The maximum amount of AOM was observed in the bark of H. vanadium was determined for N. oleander (34.58%) and H. rosa-sinen- rosa-sinensis. The ability of the bark of this tree to trap dust particles sis (21.02%). For cobalt, the maximum AOM was seen in leaf (16.95%) requires further investigations. The chemical and enzymatic properties and bark (14.33%) of H. rosa-sinensis. The lowest AOM of cobalt was of H. rosa-sinensis such as polyphenolic compounds and mucilages may found for the bark (10.72%) and leaf (8.68%) of C. erectus. The results have significant impact on their purifying behavior (Afify and Hassan, of AOM in many studies show that the properties of leaf surfaces play 2016; Kaleemullah et al., 2017). an important role in the amount of airborne particles taken from the air and is related to the concentration of heavy metals measured in the leaf 4. Conclusion (Simon et al., 2014). Sawidis et al. (2011) have reported that a fluffing structure on the surface of the trees increase the potential for airborne This study has been done in the Asaloyeh industrial zone (the busiest contaminants to be trapped, thereby easing the absorption of metals by gas and petrochemical companies in the Middle East) and its surround- the pores of the epithelium (stomata). In addition, the structure of the ing environment. The concentrations of Ni, Pb, V, and Co have been leaf and wax layers is very important for the accumulation of heavy met- evaluated in leaf and bark samples of N. oleander, C. erectus, H. als (Thao et al., 2014). Higher concentrations of heavy metals in the N. rosa-sinensis, and B. spectabilis willd. The evaluations were carried out indicum shrub can be attributed to the mucus layer on the leaf surface. in two industrial sites, two urban sites, and two rural sites in the Asa- Espinosa and Oliva (2006) have found that although the leaf surface of loyeh industrial zone and compared with the background site (BS). Lantana camara L. is thinner than the N. oleander species, the concentra- The highest concentration of nickel, lead, vanadium, and cobalt were tion of copper in N. oleander leaf is higher than that of for the Lantana found in the leaf of C. erectus (12.77mg/kg), N. oleander (26.04mg/ camara L. leaf. kg), N. oleander (23.17mg/kg), and C. erectus (21.78mg/kg), respec- In our study, the AOM factor for N. oleander and C. erectus species tively. The average metals concentration in all tree species (with excep- is more than the two otherUNCORRECTEDspecies. These observations could be at- tion of H. rosa-sinensis) in the Asaloyeh industrial zone was higher than tributed to the width, density, and the presence of a mucus layer on BS (p>0.05). The significant difference between the heavy metals con- the surface of leaf (which traps more atmospheric particles). It has also centrations in trees of the rural and urban areas and BS implicitly rep- been stated that a significant amount of atmospheric dust particles de- resents that industrial activities have been effective in metals contami- posit on the surface of leaves increasing the concentration of heavy nation of the surrounding area. Observing the highest CBCI for N. ole

8 M. Safari et al. Journal of Environmental Management xxx (2018) xxx-xxx ander and C. erectus suggests that this tree species have the potential to Hu, Y., Wang, D., Wei, L., Zhang, X., Song, B., 2014. Bioaccumulation of heavy metals in plant leaves from Yan an city of the Loess Plateau. China. Ecotox. Environ. Safe 110, remediate the heavy metals from the polluted soil. The highest MAI fac- ? 82–88. tor was measured for H. rosa-sinensis and C. erectus which indicates the Huang, Y., Chen, Q., Deng, M., Japenga, J., Li, T., Yang, X., He, Z., 2018. Heavy metal effectiveness of these trees for controlling the air born-metals in the pol- pollution and health risk assessment of agricultural soils in a typical peri-urban area luted areas. According to the AOM factor, N. oleander and C. erectus trees in southeast China. J. Environ. Manag. 207, 159–168. Kabata-Pendias, A., 2010. Trace Elements in Soils and Plants. CRC press. are more likely to capture vanadium and nickel from air, so planting Kafaei, R., Tahmasbi, R., Ravanipour, M., Vakilabadi, D.R., Ahmadi, M., Omrani, A., Ra- these trees in industrial areas with such atmospheric pollutants would mavandi, B., 2017. Urinary arsenic, cadmium, manganese, nickel, and vanadium lev- be beneficial. els of schoolchildren in the vicinity of the industrialised area of Asaluyeh, Iran. Envi- ron. Sci. Poll. Res. 24 (30), 23498–23507. Kaleemullah, M., Jiyauddin, K., Thiban, E., Rasha, S., Al-Dhalli, S., Budiasih, S., Gamal, Acknowledgments O., Fadli, A., Eddy, Y., 2017. Development and evaluation of Ketoprofen sustained re- lease matrix tablet using Hibiscus rosa-sinensis leaves mucilage. Saudi Pharm. J. 25, 770–779. This article was extracted from M.Sc. thesis of the first author and Liang, J., Fang, H., Zhang, T., Wang, X., Liu, Y., 2017. Heavy metal in leaves of twelve supported by Bushehr University of Medical Sciences (Grant Number: plant species from seven different areas in Shanghai, China. Urban For. Urban Green. 27, 390–398. BPUMS-H16-112). Liu, L., Fang, Y., Wang, S., Xie, Y., Yang, D., 2013. Leaf micro-morphology and features in adsorbing air suspended particulate matter and accumulating heavy metals in seven trees species. Huanjing Kexue 34, 2361–2367. Appendix A. Supplementary data Liu, Y.-J., Zhu, Y.-G., Ding, H., 2007. Lead and cadmium in leaves of deciduous trees in Beijing, China: development of a metal accumulation index (MAI). Environ. Poll. 145, Supplementary data related to this article can be found at https:// 387–390. Matin, G., Kargar, N., Buyukisik, H.B., 2016. Bio-monitoring of cadmium, lead, arsenic and doi.org/10.1016/j.jenvman.2018.05.026. mercury in industrial districts of Izmir, Turkey by using honey bees, propolis and pine tree leaves. Ecol. Eng. 90, 331–335. Naeimi, B., Foroutan, R., Ahmadi, B., Sadeghzadeh, F., Ramavandi, B., 2018. Pb (II) and Uncited reference Cd (II) removal from aqueousPROOFsolution, shipyard wastewater, and landfill leachate by modified Rhizopus oryzae biomass. Mater. Res. Exp. 5 (4), 045501. Sawidis et al., 2012. Norouzi, S., Khademi, H., Cano, A.F., Acosta, J.A., 2015. Using plane tree leaves for bio- monitoring of dust borne heavy metals: a case study from Isfahan, Central Iran. Ecol. Indic. 57, 64–73. References Norouzi, S., Khademi, H., Cano, A.F., Acosta, J.A., 2016. Biomagnetic monitoring of heavy metals contamination in deposited atmospheric dust, a case study from Isfahan, Iran. J. Environ. Manag. 173, 55–64. Afify, A.E.-M.M.R., Hassan, H.M.M., 2016. Free radical scavenging activity of three dif- Oliva, S.R., Mingorance, M., 2006. Assessment of airborne heavy metal pollution by above- ferent flowers-Hibiscus rosa-sinensis, Quisqualis indica and Senna surattensis. Asian. ground plant parts. Chemosphere 65 (2), 177–182. Pac. J. Trop. Biomed. 6, 771–777. Onder, S., Dursun, S., 2006. Air borne heavy metal pollution of Cedrus libani (A. Rich.) in Al-Khashman, O.A., Ala’a, H., Ibrahim, K.A., 2011. Date palm (Phoenix dactylifera L.) the city centre of Konya (Turkey). Atmos. Environ. 40, 1122–1133. leaves as biomonitors of atmospheric metal pollution in arid and semi-arid environ- Parraga-Aguado, I., Querejeta, J.-I., González-Alcaraz, M.-N., Jiménez-Cárceles, F.J., ments. Environ. Poll. 159, 1635–1640. Conesa, H.M., 2014. Usefulness of pioneer vegetation for the phytomanagement of Al-Khlaifat, A.L., Al-Khashman, O.A., 2007. Atmospheric heavy metal pollution in Aqaba metal (loid) s enriched tailings: grasses vs. shrubs vs. trees. J. Environ. Manag. 133, city, Jordan, using Phoenix dactylifera L. leaves. Atmos. Environ. 41, 8891–8897. 51–58. Alfani, A., Baldantoni, D., Maisto, G., Bartoli, G., De Santo, A.V., 2000. Temporal and spa- Pen-Mouratov, S., Shukurov, N., Steinberger, Y., 2008. Influence of industrial heavy metal tial variation in C, N, S and trace element contents in the leaves of Quercus ilex within pollution on soil free-living nematode population. Environ. Poll. 152 (1), 172–183. the urban area of Naples. Environ. Poll. 109, 119–129. Ramachandra, T., Sudarshan, P., Mahesh, M., Vinay, S., 2018. Spatial patterns of heavy Allahabadi, A., Ehrampoush, M.H., Miri, M., Aval, H.E., Yousefzadeh, S., Ghaffari, H.R., metal accumulation in sediments and macrophytes of Bellandur wetland, Bangalore. Ahmadi, E., Talebi, P., Fahabadi, Z.A., Babai, F., 2017. A comparative study on capa- J. Environ. Manag. 206, 1204–1210. bility of different tree species in accumulating heavy metals from soil and ambient air. Ribeiro de Souza, S.C., Adrián López de Andrade, S., Anjos de Souza, L., Schiavinato, M.A., Chemosphere 172, 459–467. 2012. Lead tolerance and phytoremediation potential of Brazilian leguminous tree Alsaleh, K.A., Meuser, H., Usman, A.R., Al-Wabel, M.I., Al-Farraj, A.S., 2018. A comparison species at the seedling stage. J. Environ. Manag. 110, 299–307. of two digestion methods for assessing heavy metals level in urban soils influenced by Sawidis, T., Breuste, J., Mitrovic, M., Pavlovic, P., Tsigaridas, K., 2011. Trees as bioindica- mining and industrial activities. J. Environ. Manag. 206, 731–739. tor of heavy metal pollution in three European cities. Environ. Poll. 159, 3560–3570. Baldantoni, D., De Nicola, F., Alfani, A., 2014. Air biomonitoring of heavy metals and poly- Sawidis, T., Krystallidis, P., Veros, D., Chettri, M., 2012. A study of air pollution with cyclic aromatic hydrocarbons near a cement plant. Atmos. Poll. Res. 5, 262–269. heavy metals in Athens city and Attica basin using evergreen trees as biological indi- Berdonces, M.A.L., Higueras, P.L., Fernández-Pascual, M., Borreguero, A.M., Carmona, M., cators. Biol. Trace Elem. Res. 148, 396–408. 2017. The role of native lichens in the biomonitoring of gaseous mercury at contami- Serbula, S.M., Miljkovic, D.D., Kovacevic, R.M., Ilic, A.A., 2012. Assessment of airborne nated sites. J. Environ. Manag. 186, 207–213. heavy metal pollution using plant parts and topsoil. Ecotox. Environ. Safe 76, Bilo, F., Borgese, L., Dalipi, R., Zacco, A., Federici, S., Masperi, M., Leonesio, P., Bontempi, 209–214. E., Depero, L.E., 2017. Elemental analysis of tree leaves by total reflection X-ray fluo- Shi, X., Zhang, X., Chen, G., Chen, Y., Wang, L., Shan, X., 2011. Seedling growth and metal rescence: new approaches for air quality monitoring. Chemosphere 178, 504–512. accumulation of selected woody species in copper and lead/zinc mine tailings. J. En- Birke, M., Rauch, U., Hofmann, F., 2018. Tree bark as a bioindicator of air pollution in the viron. Sci. 23, 266–274. city of Stassfurt, Saxony-Anhalt, Germany. J. Geochem. Explor. 187, 97–117. Simon, E., Baranyai, E., Braun, M., Cserháti, C., Fábián, I., Tóthmérész, B., 2014. Elemental Delshab, H., Farshchi, P., Keshavarzi, B., 2017. Geochemical distribution, fractionation concentrations in deposited dust on leaves along an urbanization gradient. Sci. Total and contamination assessment of heavy metals in marine sediments of the Asaluyeh Environ. 490, 514–520. port, Persian Gulf. Mar. Poll. Bull. 115, 401–411. Song, Y., Maher, B.A., Li, F., Wang, X., Sun, X., Zhang, H., 2015. Particulate matter de- Drava, G., Anselmo, M., Brignole, D., Giordani, P., Minganti, V., 2017. Branch bark of posited on leaf of five evergreen species in Beijing, China: source identification and holm oak (Quercus ilex L.) for reconstructing the temporal variations of atmospheric size distribution. Atmos. Environ. 105, 53–60. deposition of hexavalent chromium. Chemosphere 170, 141–145. Swaileh, K., Matani, M., Hussein, R.M., 2006. Heavy metals in urban roadside plants from Espinosa, A.F., Oliva, S.R., 2006. The composition and relationships between trace ele- Amman, Jordan. Bull. Environ. Contam. Toxicol. 77, 445–450. ment levels in inhalable atmospheric particles (PM 10) and in leaves of Nerium olean- Thao, N., Yu, X., Zhang, H., 2014. Deposition of Particulate Matter of Different Size Frac- der L. and Lantana camara L. Chemosphere 62, 1665–1672. tions on Leaf Surfaces and in Epicuticular Waxes of Urban Forest Species in Summer Fernández-Calviño, D., Cutillas-Barreiro, L., Paradelo-Núñez, R., Nóvoa-Muñoz, J.C., and Fall in Beijing, China, vol. 3, 12–22. Fern ndez-Sanjurjo, M.J., lvarez-Rodr guez, E., N ez-Delgado, A., Arias-Est vez, á Á í úñ é Tomašević, M., Aničić, M., Jovanović, L., Perić-Grujić, A., Ristić, M., 2011. Deciduous tree M., 2017. Heavy metals fractionation and desorption in pine bark amended mine soils. leaves in trace elements biomonitoring: a contribution to methodology. Ecol. Indic. J. Environ. Manag. 192, 79–88. 11, 1689–1695. Gondek, K., Mierzwa-Hersztek, M., Kopeć, M., 2018. Mobility of heavy metals in sandy soil Turer, D., Maynard, J.B., Sansalone, J.J., 2001. Heavy metal contamination in soils of ur- after application of composts proUNCORRECTEDduced from maize straw, sewage sludge and biochar. ban highways comparison between runoff and soil concentrations at Cincinnati, Ohio. J. Environ. Manag. 210, 87–95. Water Air Soil Pollut. 132, 293–314. Hofman, J., Stokkaer, I., Snauwaert, L., Samson, R., 2013. Spatial distribution assessment Ugolini, F., Tognetti, R., Raschi, A., Bacci, L., 2013. Quercus ilex L. as bioaccumulator of particulate matter in an urban street canyon using biomagnetic leaf monitoring of for heavy metals in urban areas: effectiveness of leaf washing with distilled water tree crown deposited particles. Environ. Poll. 183, 123–132. and considerations on the trees distance from traffic. Urban For. Urban Green. 12, Hovmand, M., Nielsen, S.P., Johnsen, I., 2009. Root uptake of lead by Norway spruce 576–584. grown on 210 Pb spiked soils. Environ. Poll. 157, 404–409.

9 M. Safari et al. Journal of Environmental Management xxx (2018) xxx-xxx

Zhai, Y., Dai, Q., Jiang, K., Zhu, Y., Xu, B., Peng, C., Wang, T., Zeng, G., 2016. Traffic-re- Zhao, X., Liu, J., Xia, X., Chu, J., Wei, Y., Shi, S., Chang, E., Yin, W., Jiang, Z., 2014. The lated heavy metals uptake by wild plants grow along two main highways in Hunan evaluation of heavy metal accumulation and application of a comprehensive bio-con- Province, China: effects of soil factors, accumulation ability, and biological indication centration index for woody species on contaminated sites in Hunan, China. Environ. potential. Environ. Sci. Poll. Res. 23, 13368–13377. Sci. Poll. Res. 21, 5076–5085. Zhan, H., Jiang, Y., Yuan, J., Hu, X., Nartey, O.D., Wang, B., 2014. Trace metal pollution in soil and wild plants from lead–zinc smelting areas in Huixian County, Northwest China. J. Geochem. Explor. 147, 182–188.

PROOF

UNCORRECTED

10