Monitoring of air pollutants using plants and co-located soil—Egypt: characteristics, pollution, and toxicity impact

The present work was conducted to evaluate the air quality in terms of inorganic pollutants and toxicity impact using two evergreen tree leaves, Eucalyptus globulus Labill (E. globulus) and Ficus microcarpa L.f., Suppl. Pl. 442 (1782) (F. microcarpa) as biomonitors. Thirty tree leaves and an equal number of co-located soil samples from different regions of Egypt (urban Greater Cairo Metropolitan (GC) and rural Menoufia Governorate (MG)) were collected. The concentrations of 34 and 40 elements were determined using instrumental neutron activation analysis (INNA) and atomic absorption spectrometry (AAS) in tree leaves and soils, respectively. Bivariate and multivariate statistical analyses were implemented. The air pollution was assessed using enrichment factor, pollution load index, potential ecological risk, and risk index. In addition, human and ecotoxicity were evaluated based on the ReCiPe method. The mean concentration values of the obtained elements in tree leave in urban Greater Cairo and rural Menoufia Governorate show that the major elements are slightly higher in F. microcarpa than in E. globulus. Likewise, the mean values of elements in soil from GC and MG show no significant difference except for major elements (Fe, Al, Mg, K, Na, and Ti) in MG. The normalized concentrations of tree leave and soil show that the accumulated elements by F. macrocarpa are slightly higher than in E. globulus in GC and MG. While in terms of the investigated area, the concentrations of elements in MG are considerably higher than in GC. Pollution load index (PLI) spatial distribution over investigated areas showed that despite high population density, heavy traffic, and urban pollution, the Cairo samples exhibit significantly lower values as compared to those from Menoufia, which is most likely due to the uncontrolled industrial and domestic waste disposal outside Cairo. Potential ecological risk (PER) was significant for As in soil and for As and Cd in tree species. Human toxicity shows higher values in urban locations. Contrariwise, in the terrestrial ecotoxicity aspect, the rural locations are much higher than in urban ones.


Introduction
Air pollution becomes one of the most serious problems caused by human activities. Air pollutants are made up of a mixture of gas, liquid, and solid compounds of the air and particulate matter (PM). They may cause adverse health effects in humans and affect plant life and cause detrimental changes in the atmosphere of the earth (Brunekreef and Holgate 2002;Kampa and Castanas 2008;Kumar 2016). Heavy metals can easily be linked to terrestrial food webs (Gall et al. 2015). They are steady in the environment and are dangerous in case of intake via different pathways (ingestion, inhalation, and dermal contact) by a human. The potential sources of pollution are dramatically increasing, for instance, processes of urbanization expansion, industrialization, and Responsible Editor: Elena Maestri economic development Hassanien 2011;Kumar 2016;Salim and Madany 1993;Yekeen et al. 2016).
The quality of soil and plants is affected mainly by the growing demand for agricultural food due to population growth and urbanization (Shah et al. 2019). Heavy metals are accumulated in trees through the foliage, and it is considered the major pathway, specifically in polluted areas (Farahat 2011). Trees are quite efficient in trapping atmospheric deposited particles. The utilization of tree leaves as biomonitors in accumulating metals has acquired great ecological and ecotoxicological importance (Farahat 2011;Yalaltdinova et al. 2018). In the recent Anthropocene, a variety of urban and rural roadside trees is now increasingly recognized as an environmentally sustainable tool for monitoring and mitigating the effects of air pollution. The deposition of PM particulates in the foliar surface of tree leaves causes structural and functional changes in plants (Kumar 2016;Panda and Rai 2015). Tree leaves as biomonitors are an accessible and inexpensive tool for the evaluation of atmospheric air deposition.
Some methods may also be utilized, such as moss and lichen. Because of their root system, they ultimately accumulate pollutants from the atmosphere, and the crustal association is minimal (Christensen et al. 2018;Steinnes 1995;Steinnes et al. 1992Steinnes et al. , 1994Steinnes et al. , 2017Vuković et al. 2016). The vascular plants, mosses, and lichens are good biomonitors. Mosses and lichens are mainly grown in wetlands and are hardly found in arid and hot countries like Egypt. That's why, the higher plants have a root system and the tree leaves act as good biomonitors (Farahat 2011;Gorelova and Frontasyeva 2017;Jiang et al. 2018;Kabata-Pendias 2011;Kumar 2016;Norouzi et al. 2015;Panda and Rai 2015;Peterson and Girling 1981;Quénéa et al. 2019;Ukpebor et al. 2010). The Egyptian population is 98 M, and almost 40 % is mainly inhabited in Delta Nile, according to CAPMS (2019). GC consists of three main governorates are Cairo, Giza, and Kalyobia, where approximately 24 M is inhabited, and population density is 51,029, 7416, and 5356 pop/km 2 , respectively (CAPMS 2019). Therefore, GC is an overpopulated city and one of the most populated cities in the world.
Because of the expensive cost of monitoring instrumentation and difficulties in associated sampling methods, the suitability of two commons, ubiquitously distributed, and ornamental tree leaves in GC and MG as biomonitors and as effective bioaccumulators of atmospheric trace metals have been evaluated. E. globulus Labill and F. microcarpa L.f ((IPNI 2021)) were selected as biomonitors, and the concentrations of 32 elements were determined. Both E. globulus and F. macrocarpa are evergreen tree leaves planted in the Mediterranean environment (Fife et al. 2008). E. globulus is a forest tree widely grown for woody biomass production in Mediterranean climates (Wilson 1993). E. globulus mature leaves are narrow, sickle-shaped with a length of 8 to 12 cm and about 1 cm wide, whereas F. macrocarpa leaves are oval with 5 to 12 cm long and 2 to 6 cm wide. Therefore, these properties could influence aerial dust retention capacity.
A variety of analytical methods can be reliably utilized for the assessment of the elemental composition in the tree leaves and soil samples. There are several analytical techniques available, each of which offers complementary advantages, such as X-ray fluorescence (XRF) and X-ray diffraction (XRD). INAA and AAS were used in the present work. INAA are an effective technique for detecting inorganic pollutants in tree leaves and soil. Likewise, the complementary analytical method AAS was employed mainly to determine Cu, Cd, and Pb in tree leaves.
To date, several researches have focused on heavy metal analysis and toxicity assessment using topsoils, biomonitors, and vascular plants, for instance, (Shifaw 2018;Tong et al. 2020) in China, (Morton-Bermea et al. 2009) in Mexico City, and(Gorelova andYalaltdinova et al. 2018) in Russia. All of these investigations reveal significant contamination, drawing scientists' attention to study the dynamics of pollution not only within their own country but also across borders. Tong et al. (2020) showed an overview of heavy metal pollution studies in the urban soils throughout 71 cities of China, based on data from online literature, during the period 2003-2019. The study focused on As, Cd, Cr, Hg, Pb, Cu, Zn, and Ni heavy metals. The results show that eight heavy metals' mean values all exceeded the soil background values in China. Similar research in China was presented by Shifaw (2018). The author reviewed and analyzed national scale heavy metals soil contamination in urban and agricultural areas and the study showed that the concentration level of heavy metals was above the natural background level in most areas of China and serious attention should be paid to solve the problem of soil toxicity and biological accumulation, which may continue to threaten the sustainability of China's development. Likewise, Morton-Bermea et al. (2009) studied heavy metal pollution (Cr, Cu, Ni, Pb, Zn, and V) in 135 urban topsoil samples from the metropolitan area of Mexico City. The study established that assessing heavy metal pollution in urban topsoil is a viable method for determining the environmental damage caused by industry and human activity.
The ability of F. microcarpa and E. globulus, bark, and soil as biomonitors in some industrial zones in Minya governorate-Egypt was investigated by El-Khatib et al. (2020). The concentrations of Pb, Cu, and Cd were determined and the highest concentrations were observed in tree leaves, which, in turn, proves the suitability of implementation of tree leaves as biomonitors for metal air quality assessment. Yalaltdinova et al. (2018) studied the atmospheric inorganic pollutants and the relevant potential toxic impacts on humans and the ecosystem in Ust-Kamenogorsk city (Kazakhstan) and identified that the highest contribution to the pollution came from Zn, and mostly because of the existence of the lead-zinc plant "Kazzinc." Similarly, Gorelova and Frontasyeva (2017) provided a basic information on how higher plants can be used for biomonitoring and bioremediation. It includes material from the authors' own research on the possibility of using woody plants for biomonitoring and phytoremediation of anthropogenic heavy metal pollution. Authors shed the light on the reliability and suitability of using higher plants for biomonitoring purposes and summarized the proper plant species and the corresponding accumulated elements.
Despite some attempts to use plant biomonitoring to address the issue of inorganic metal air pollution's impact on human and terrestrial ecosystems, there is a scarcity of information that could provide a thorough description of the investigated locations. This is because these studies are carried out on a local level. However, utilizing neutron activation analysis and an atomic absorption spectrometry, the current study provides distribution patterns of elemental composition in the air and nearby soils in GC (urban area) and MG (rural area).
The objectives of this study are to (i) determine the concentrations of trace elements in tree leaves and adjacent soils, (ii) quantify the extent of pollution using various pollution indices, (iii) use multivariate statistical analysis to extract more information about pollution sources, and (iv) assess human toxicity and terrestrial ecotoxicity.

Investigated area
The Greater Cairo contains three interlocked cities Cairo, Giza, and Kalyoubia (30° 03′ N 31° 22′ E), as shown in Fig. 1. Cairo is the capital of Egypt and the largest city in the Middle East, and the second-largest in Africa after Lagos. Its metropolitan area is the 15 th largest in the world and is located near the Nile Delta and was founded in 969 CE. Based on the recent statistics reported by CAPMS (2019), Cairo, Giza, and Kalyoubia are 100%, 60%, 42% are classified as urban zones, respectively. The continuous expansion of the urbanization around GC (satellite cities) involves significant geomorphic impacts and has unfavorable geomorphological consequences (Csima 2010). In addition to the industrial and urbanization expansion and dense population, there are 7.9 M vehicles in Egypt, Cairo has the largest number of vehicles (2.2 M), followed by Giza (921500), and Alexandria (702100). Vehicles and vehicular traffic emissions represent the main pollution sources. Moreover, GC is surrounded by two industrial zones from the northern and southern borders (Shoubra El Kheima and Helwan, respectively), which significantly increase the concentrations of air pollutants (Farahat 2011). For comparison purposes, MG (30° 38′ 14.88′′ N, 30° 54′ 53.73′′ E) was examined in terms of metal air pollution where Menoufia is a rural region and characterized by agricultural activities. The MG population is about 4.5 M, and the population density is 1807 pop/km 2 (CAPMS 2019). MG is characterized by a rural nature and an urbanized zone of 21.2%. It is located 60 km North-West of Cairo.

Sampling strategy
Thirty evergreen tree leaves samples of E. globulus and F. macrocarpa and 29 adjacent soil samples were collected from GC and MG, respectively. The samples were collected from various districts in GC and MG. A total of 20 tree leaves' samples were collected from GC, and ten samples were collected from MG. Tree leaves from the two plants, E. globulus and F. macrocarpa were selected because they are found in a widespread distribution in Egypt. E. globulus and F. microcarpa are characterized by their ability to uptake particulates more than other tree leaves at the same sampling site (El-Khatib et al. 2020;Freer-Smith et al. 2004). They are suitable for biomonitoring in dry countries (Freer-Smith et al. 2004). Leaves were randomly taken from all sides of the trees at approximately 2 m in height. The sampling process was conducted according to the recommendations and guidelines reported by Tomašević et al. (2011);Yalaltdinova et al. (2018). The full description of the sample number and the corresponding collecting points are provided in Table 1SM (Supplementary Materials).

Sample preparation
The tree leaves were triple washed with distilled water and dried at room temperature. Then, they dried in the drying oven at a temperature of 40 °C for 7 days. The dried samples were grinded into agate grinders. Two aliquots, each of 0.3 g of the obtained powder was packed in polyethylene zipped bags. Samples were pressed in the mechanical piston to obtain pellets of 1.5 cm diameter and 0.3 cm height. These pellets were packed in aluminum cups and polyethylene bags for long and short-lived isotopes irradiation. Soil samples were cleaned from plants and roots. Later on, they were dried in a current air and then dried at a temperature of 105 °C for 4 h to a constant weight, crushed with a non-metal grinder. The grinder was thoroughly cleaned and dried after every grinding to avoid any contamination among samples, as reported by Al-Khashman et al. (2011).
For AAS, an aliquot of 0.2 g of tree leaves was placed in a Teflon vessel and decomposed with 3 mL of concentrated nitric acid (HNO 3 ) and 2 mL of hydrogen peroxide (H 2 O 2 ) in a microwave digestion system (Mars; CEM, Matthews, NC, USA) for complete digestion. All of the reagents used for this study were of analytical grade: nitric acid-(69%); trace pure (Merck, Darmstadt, Germany); hydrogen peroxide-(30%), p.a. (Merck); and bidistilled water. A detailed description of the digestion mechanism was published elsewhere by Chaligava et al. (2021); Madadzada et al. (2019).

Analytical techniques
Tree leaves' samples were subjected to neutron activation analysis at the IBR-2 reactor of the FLNP JINR in Dubna, Russia. The detailed characteristics of the irradiation channels are reported elsewhere by Frontasyeva and Pavlov (2005). The samples were irradiated twice to determine the elements of short-and long-lived isotopes. For short-lived isotopes (SLI), each sample was irradiated for 3 min and measured for 15 min. Likewise, the long-lived ones were irradiated for 3 days under neutron flux of 1.8×10 11 n/cm 2 ·s. Later on, the samples were repacked and measured twice. The first measurement was performed after four days of decay for 30 min, and this is the so-called first long-lived isotope (LLI1), while the second one was achieved after 20 days of decay for 1.5 h, and this is called the second longlived isotopes (LLI2) (Chaligava et al. 2021;Madadzada et al. 2019).
AAS was used to determine the concentrations of Cu, Cd, and Pb in tree leaves samples via iCE 3300 AAS atomic absorption spectrometer with electrothermal (graphite furnace) atomization (Thermo Fisher Scientific, Waltham, MA, USA).

Enrichment factor
To avoid the significant variations in the absolute values that may mislead the interpretation of the relationships between elements, the concentrations of the entire set of elements is normalized to the corresponding values in the upper continental crust (UCC) or the soil (Badawy et al. 2017(Badawy et al. , 2020Bargagli 1989;Bargagli et al. 1995;El-Taher et al. 2019). EF is given as follows: where C x is the concentration of the elements × of concern in tree leaves, whereas C s stands for the concentration of the same element in soil. Both concentrations are normalized to Al in tree leaves and soil, respectively (Bargagli 1989). The interpretation of enrichment factor EF and the corresponding categories is provided in Table 2SM (Supplementary Materials)

Pollution load index
Before the calculation of PLI, the contamination factor (CF) for every single element was calculated as the ratio of the concentration of each element over the corresponding value in the reference plant (RP) (Markert 1992) or upper continental crust (UCC) (Rudnick and Gao 2014) for plants and soil, respectively. PLI is given as follows (Kowalska et al. 2018;Varol 2011): when PLI is greater than unity, it suggests that pollution exists and vice versa.

Potential ecological risk index PER
PER was used to categorize and prioritize which study areas should receive additional attention in terms of pollution extent (Hakanson 1980). The formula of PER i f for a single metal pollution is deduced as follows: where PER i f is a potential ecological risk index, C i f is the contamination factor, and T i f is the "toxic-response" coefficient for the given single metal/metalloid. The

Statistical data analysis
All the statistical analyses of chemical data and graphing were performed using the statistical package R (R Core Team 2016). The data were handled in MS office Excel. GIS technology was used to map the spatial distribution of the pollution patterns. The data were interpolated based on the inverse distance weighting IDW method. The so-called principal component analysis (PCA) dimensionality reduction technique was used to get more information about geochemical symmetrical elements and their origin. The data was centered log-ratio transformed (CLR) before the implementation of PCA (Badawy et al. 2018;Faith 2015). The obtained PCAs were plotted and partitioned into clusters using the K-means method.

Metal concentrations and inter-correlation
A total of 34 and 40 elements were determined in the examined two evergreens trees leaves (E. globulus and F. microcarpa) and adjacent soil samples, respectively. The full descriptive statistics of the concentrations of the determining elements are stipulated in Tables 1 and 2 for plant and soil, respectively.
The interaction of the mean values in both tree leaves in GC (urban) shows that the mean values of the obtained elements are in good agreement, except for Ca, K, Cl, Na, Al, Fe, and Mg found to be higher in F. microcarpa than in E. globulus. The major elements are shown to be accumulated in F. microcarpa higher than E. globulus except for Na.
Likewise, the mean values of elements in the two tree leaves in GM (rural) show that all the concentrations of the obtained elements in F. microcarpa are higher than in E. globulus except Na, Mn, Br, Ni, and Se are observed to be higher in E. globulus versus F. microcarpa . Overall, the concentrations of obtained elements are noticed to be higher in F. microcarpa than in E. globulus.
Similarly, the concentration mean values mg/kg of the adjacent soil samples were determined in GC (urban) and MG (rural) and the results show that the mean values are in line for the two examined governorates. The major elements namely Si, Ca, Fe, Al, Mg, K, Na, and Ti were observed to be higher in MG than in GC, except Si and Ca were higher in GC than in GM. The concentrations of the elements in F. microcarpa are higher than in E. globulus and in MG are higher than in GC.
The obtained concentrations of the elements in tree leaves and soil were normalized to the corresponding values of the elements in the reference plants (RP) reported by Markert (1992) and UCC reported by Rudnick and Gao (2014). For

Risk index
The potential toxicity response index (RI), a single-entity index combining all of the metals of interest, is calculated as follows: where RI is the sum of the PER i f for each metal of interest (Hakanson 1980). The categories are interpreted and given in Table 3SM (Supplementary Materials) (Badawy et al. 2018;Hakanson 1980;Karuppasamy et al. 2017).

Human and ecotoxicity assessment
In the life cycle impact assessment methods, we used the ReCiPe method and characterization factors (De Schryver and Goedkoop 2009;Dekker et al. 2019;Huijbregts and Van Zelm 2009;Struijs et al. 2009Struijs et al. , 2010, which comprises harmonized category indicators at the midpoint and the endpoint level. The characterization factors quantify the potential impacts that inputs and releases have on specific impact categories in common equivalence units. The below equation shows the characterization process: where mi is the magnitude of intervention i (e.g., the mass of heavy metal elements released to air), Q mi the characterization factor that connects intervention i with midpoint impact category m, and I m the indicator results for midpoint impact category m. (De Schryver and Goedkoop 2009;Goedkoop and Spriensma 2001).
In the human toxicity and terrestrial ecotoxicity category in the ReCiPe method and characterization factors, many elements are considered in the air compartment (high population and low population area). In this study, Mg, V, Cr, Mn, Co, Ni, Zn, As, Br, Ag, Sb, and Ba elements (12 elements) were matched, and characterization results were calculated. In the soil compartment (agriculture soil, forestry soil, and industrial soil), Mg, Al, Cl, As, Se, Br, Cd, Sb, I, and Ba elements (10 elements) were matched, and characterization results were calculated.
tree leaves, the obtained results show significant concentrations of Th, Ti, Ta, Al, Se, Na, Sc, As, V, and Br; Th, Ti, Na, Ta, Se, As, Al, Sc, Br, and Fe in GC samples for F. microcarpa and E. globulus, respectively, whereas significant concentrations are observed for Ti, Th, Ta, Sc, Al, Fe, Na, V, As, and Se for F. microcarpa and Th, Ta, Sc, Ti, Na, Se, Al, Fe, As, and Br for E. globulus in MG. These findings are box plotted in Fig. 2(A). Overall, the accumulated elements by F. microcarpa are slightly higher than in E. globulus in GC and MG. While in terms of the investigated area, the concentrations of elements in MG are considerably higher than in GC. Similarly, the boxplot illustrated in Fig. 2(B) shows remarkable concentrations in the studied soil in GC for Au, Cl, Ag, Br, Sb, Zn, and Yb. While, the concentrations of Cl, Ag, Br, Au, Sb, Ti, and Fe were noticed to be higher than those in the corresponding values reported by (Rudnick and Gao 2014) for UCC. To sum up, the concentrations of the elements in the studied soil are significantly higher in MG than GC.
To test the differences between the mean values of the concentrations in tree leaves and soil in different areas (rural and urban) and prove the previous findings, a Tukey test of pairwise comparisons was implemented as shown in Table 3. The Tukey test was used to identify differences in mean values of the elements in tree leaves. At a significance level of 0.05, the probabilities of rural MG vs urban GC municipalities are (p= 0.039) and the probabilities of F. microcarpa vs. E. Globulus are (p= 0.034). Both of the probabilities are lower than the significant level, and this leads to rejecting the null hypothesis and concluding that there are significant differences in mean values for these tree leaves and areas. Similarly, the probability of rural vs urban areas in the case of soil is p= 0.026, which suggests the same hypothesis.
As a summing up, the mean values of elements in tree leaves (Ta, Hf, Tb, Sm, La, Sr, Ni, Co, Cr, Ca, and Cl) are at least three times higher and (U, Br, Na, Se, As, Fe, V, Sc, and Al) are 5-10 times higher than those reported in RP reported by (Markert 1992), respectively. The mean values of Th and Ti are higher by 20 times.
Similarly, the overall results for the normalized concentrations of soil samples to the corresponding values reported by Rudnick and Gao (2014) for UCC are from 2 to 5 times higher for Sb, Cr, Hf, Ni, Ca, Co, Sm, Cl, Cd, La, Sr, Tb, and Pb, while from 5 to10 times higher for U, Br, V, Fe, and As. Finally, from 10 to 65 times higher for Se, Na, Al, Sc, Ta, Ti, Th, U, Br, V, Fe, and As, respectively. Contrariwise, Rb, Cs, Mg, I, Mn, K, Zn, Ba, Cu, and Au are less than UCC.

Principal component analysis
A total of 34 out of 40 elements were selected to perform PCA for tree leaves and soil, respectively, whereas the other elements of soil samples were excluded from the matrix because they tended to form separate groups (outliers). PCA was performed for tree leaves independent of the type of tree leaves and sampling location. For PCA, we assume that the different tree leaves accumulate the elements of the atmospheric deposition equally. This assumption was proved by plotting the first two PCAs for GC and GM, as shown in Fig  1:4 SM (Supplementary materials).
The obtained PCAs based on the aforementioned number of elements, the eigenvalues, and percentage of variance (in parentheses) for the first three dimensions were calculated to be 10.1 (33.5%), 4.2 (13.8%), 2.9 (9.7%) in tree leaves for GC and GM. The first two PCAs express 47.3 % of the data cumulative percentage of the variance. Therefore, the first two dimensions can sufficiently explain the data.
Based on the K-mean method, the first two PCAs of the variables were plotted and clustered as in Fig. 3(A). It is obvious from the figure that four clusters were created and can be described as follows: • The first cluster contains ten elements, namely Sc, Cr, Fe, Co, Cs, La, Sm, Hf, Ta, and U. It is obvious from the elements, of which this cluster was reproduced that there is a significant association of crustal elements. The presence of these elements suggests their geogenic provenance. • The second cluster includes 11 elements, namely Mg, Cl, K, Ca, Cu, Zn, Rb, Sr, Cd, Pb, and Th. These peculiarities were also proved in the boxplot as in Fig. 2, where the variations of the elements up and down the reddashed line suggest increasing or reducing of the concentration of the corresponding elements, respectively. Mg, K, Zn, and Rb do not exhibit contamination and mostly come from the crustal association or plant nutrients (geogenic origin) (Madadzada et al. 2019), whereas the other elements that contain the second cluster mostly have an anthropogenic source. Besides, the highly toxic potential elements (Cd and Pb) have a significant association with this cluster. Mainly, due to the industrialization and urbanization processes. • The third cluster has three elements: Al, Ti, and V. These elements are often found in the oil and gas production fields. Considerable concentrations of these elements were observed near to Cairo Thermal Power Plant in Shubra El Kheima and El-sayyeda Zainab. These regions are characterized by dense populations for GC. Remarkable contributions to these elements were found in Al Shuhadaa City near MG adjacent to the central railroad station. Therefore, the elevated concentrations of the elements containing this cluster may be due to the emission and maintenance processes. • The fourth cluster is formed from Na, Mn, Ni, As, Br, and I. It is apparent from the set of the clustered elements that there is a remarkable association of sea elements (Na, Br, and I), whereas the existence of Ni and As is mainly due to relevant industries to nickel such as the burning of fossil fuels, wind-blown dust, and brick kilns which was noticed in high concentrations in Shubra El Kheima power station, then in Al Shuhadaa.
It is clear from Fig. 3(B) that the first cluster has a low contribution to the first PCA, whereas a remarkable contribution was noticed for locations # 23 and 15 for Sadat City (industrial zone) and al-Amireyya (dense population), respectively. Contrariwise, the second cluster has a considerable presentation on the plane, specifically locations # 4, 7, and 27. Based on the description of the locations given in Table 1SM, the significant contributions were registered for el-Sayyeda Zainab (dense population), Shubra El Kheima (power station), and Al Shuhadaa (railway train station), respectively.
To identify the effect that may be raised from the adjacent soil, a total of 34 out of 40 elements were selected to perform PCA, as clearly shown in Fig. 4. The PCAs accounted for 14.7 (43.2 %), 4.2 (12.4 %), and 3.4 (9.9 %) of the eigenvalues and percentage of variance (in parentheses) for the first three PCAs, respectively. The first two PCAs express 55.6 % of the data cumulative percentage of the variance and can sufficiently explain the data.
In a similar manner to the plant analysis, the first two PCAs of the variables were plotted and clustered as illustrated in Fig. 4(A). Three clusters are created and can be described as follows: • The first cluster contains nine elements, namely Na, Cl, K, Ca, Zn, As, Br, Sb, and Ba. The significant contribution of Na, Cl, and Br suggests a weathering from sea elements and/or the excess use of fertilizers and pesticides that may lead to an increment of the salinity of the soil. Considerable amounts of these elements were noticed in MG to be higher than in GC. K and Ca be explained by the agricultural nature of MG, while As, Zn, and Sb most probably due to brick kilns, and vehicles and vehicular traffic emissions (Farahat 2011). • The second cluster is grouped by both geogenic and anthropogenic elements, namely Mg, Al, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Rb, Cs, La, Ce, Sm, Eu, Tb, Ta, and Th. This cluster is containing rare earth, oil production, and crustal elements. • The third grouped from Sr, Zr, Ag, Hf, W, and U. The set of elements that formed the third cluster is naturally occurring and wind-blown soil.

Discriminatory diagram
The Sc-La-Th ternary discriminatory diagram (Hallinan 2012) is illustrated in Fig. 5a and b for leaves and soil, respectively. For comparison, the upper continental crust (UCC) data by Rudnick and Gao (2014), and reference plant (RP) values by Markert (1992) are added. It is obvious that all the tree leaves points are within the vicinity of UCC, except for some samples taken from one of the highest populated regions of Greater Cairo (Shubra El Kheima) and marked by a red circle on diagram Fig. 5a. These samples were collected in the district Shubra El Kheima in GC and are located near the ring road. They are predominantly enriched in Th, which is due to the crustal association.

Assessment of pollution, human toxicity, and terrestrial ecotoxicity levels
The pollution extent was quantified using the enrichment factor, pollution load index (PLI), potential ecological risk (PER), and risk index (RI). These indices were calculated for both soil, and tree leaves samples from urban and rural regions. The corresponding values reported by Rudnick and Gao (2014) and those reported by Markert (1992) were used for soil and plants, respectively as normalizers.
The calculated results of EF are illustrated in Fig. 6. The crustal association was eliminated using the corresponding values of the adjacent soil and UCC for plants and soil, respectively. In terms of sampling locations, the results of EF show considerable values in GC and MG for Na, Cl, K, Ca, Mn, Zn, As, Br, Sr, and Ba (Fig. 6A), whereas E. globulus was observed to accumulate concentration of elements higher than in F. microcarpa (Fig. 6B). These findings are a bit different from the normalized results and mainly due to normalizing the values to the corresponding ones in UCC for soil and RF in plants, respectively. Therefore, in air metal biomonitoring, it is highly recommended to set aside the influence of crustal association using the corresponding values in the soil contamination and then calculate EF (Bargagli et al. 1995).
It is apparent that the enriched elements have a mixed origin and most probably come from fuel and oil refining (As); woodworking and papermaking or weathering from the sea or fertilizers (Na, Cl, Zn), pesticides, and herbicides (Cl, Br); weathering from cement production or brick kilns (Ca). With high possibility, the other elements are due to the fine dust weathering (Frontasyeva 2011).
The pollution load index PLI was calculated for soil and tree leaves. The results of soil samples show that the locations namely 25, 21, 20, 9, 30, 27, 29, 16, and 22 are slightly higher than unity and classified as polluted regions, whereas the tree leaves show that PLI in almost all the studied samples is higher than unity. Among all samples, location numbers 27, 7, 4, 18, 16, 2, 22, and 9 were considerably high and ranged from 3 to 9. Based on the description of the investigated locations in Table 1SM, in the case of soil, PLI is significantly higher in MG than in GC. Contrariwise, in the case of plants, except two locations 27 and 22 (train station and brick kilns), respectively. It could be explained by the densely populated, traffic, and highways in GC, which in turn, results in an elevated concentration of elements in the air rather than in MG.
Also, for a better understanding of the local situation for the soil in the investigated areas, the concentration of potentially toxic elements in soil, namely V, Cr, Mn, Co, Ni, Zn, As, Sb, and Ba (average values for Cairo and Menoufia) were compared with the corresponding alert/intervention values in different countries worldwide according to their State Regulatory Norms as shown in Table 4. The average PLI of GC and MG were calculated by normalizing the average values of potential pollutants to these alert/intervention limits for different countries. It is clear from the table that, despite the significant variability of different State Regulatory Laws, the investigated soil seems to be below the alert/ intervention thresholds. Based on these findings, the investigated soils do not need any further action. However, regular monitoring is highly recommended to be able to follow-up on the dynamics of the concentration of these elements in the soil.
Another powerful index is widely used to assess ecological risk is the potential ecological risk index PER. PER was calculated for selected elements in the studied soil and tree leaves samples. Specifically, the selected elements are Ni, Cr, Zn, and As in soil, while in tree leaves, Cr, Ni, Cu, Zn, As, and Cd were used. These elements were selected because their epidemiological data are available and given by (Hakanson 1980;Karuppasamy et al. 2017).
PER was assessed, and the results are illustrated in Fig. 7. Based on the interpretation criteria for PER and risk index (RI) given in Table 3SM and Fig. 7a for the soil, there is a low PER as the calculated PER is less than 40, whereas in tree leaves, the calculated PER for As and Cd are considerably high, and this may pose a significant hazard to the environment and hence to the humans.
In the results of urban soil locations, mostly Ba (on average, 73%), Mn (on average, 20%), V (on average, 4%), and  Markert (1992) and Rudnick and Gao (2014), respectively were added As (on average, 3%) affect the human toxicity. In the results of rural soil locations, similarly, like the results of urban soil locations, Ba (on average, 83%), V (on average, 6%), Mn (on average, 5%), and Zn (on average, 3%) showed high contributions to the human toxicity. In the results of urban soil locations, Zn (on average, 48%), V (on average, 26%), Ni (on average, 13%), and Ag (on average, 7%) affect the terrestrial ecotoxicity. In the results of rural soil locations, V (on average, 49%), Ag (on average, 25%), Co (on average, 7%), Br (on average, 7%), and Ni (on average, 5%) showed high contributions to the terrestrial ecotoxicity. For more details on the human and terrestrial ecotoxicity information maps for each of the elements studied, see Fig. 5SM in Supplementary Material).

Conclusions
The present work achieved its main objectives of the utilization of plant biomonitoring in determining the elemental composition and quantifying the air quality in terms of metal content and associated human and terrestrial toxicity impact. Plant species demonstrated their suitability and reliability for assessing metal air pollution. The normalized accumulated metals in tree leaves show that F. microcarpa is slightly higher than in E. globulus in GC and MG. The plant biomonitoring shows that the concentrations of metals in MG are considerably higher than those in GC. Despite the densely populated and traffic in GC, the adjacent soil samples' analysis shows that the concentrations of the elements in MG are significantly higher than in GC. The concentrations of Cl, Br, Sb, Ti, and Fe were noticed to be higher than those in the corresponding values for UCC. PCA partitioned the obtained elements into four groups of geogenic provenance and anthropogenic ones. The main contribution of the anthropogenic impact comes from el Darb el Ahmr, Cairo Thermal Power Plant in Shubra El Kheima, al-Amireyya (densely populated areas), and Elsayyeda Zainab in GC, while in MG comes mainly from Al Shuhadaa City near MG adjacent to the central Railroad station, Sadat City (industrial zone). Discriminatory diagram reveals that there is a considerable crustal association to the accumulated concentrations. It is highly recommended to remove out the crustal association before the calculation of the enrichment factor to avoid the overestimation and misleading of the results. These data may be used as basic guidelines by regulatory bodies in Egypt to control the inorganic metal emission in the atmosphere. In addition, planting these trees would be of high importance in inorganic metal air polluted areas.