To ensure that the test results were representative of the greening characteristics of the main climbing plants in Nanjing, we selected Himalayan ivy [Hedera nepalensis var. sinensis (Tobl.) Rehd], climbing fig (Ficus pumila Linn), and Fu Fang vine [Euonymus fortunei (Turcz) Hand Mazz] as the test subjects for preliminary research. The climbing plants were growing on bridge pillars in an overpass and on a wall in a university campus. Sampling sites were located at Nanjing Forestry University campus and Yingtian overpass in Nanjing, China. Yingtian overpass is an important section of the Nanjing Inner Ring Expressway. Traffic is heavy, and air pollution caused by vehicle exhaust is a serious problem around the sampling site. Samples were collected in May, August, and November 2021 and January 2022. To evaluate the dust retention ability of different species of climbing plants for different sizes of particulate matter, we selected 28 consecutive days without rainfall for sampling. This ensured that the dust retention of each type of climbing plant reached a saturation value for particulate matter (Liu et al., 2012).

In this study, mature individuals of three common climbing tree species were studied. To ensure that a representative sample was investigated, a range of tree forms, landscape applications, growth conditions, and leaf characteristics were included. Plants that were in good condition with luxuriant foliage, no pests, and aged between 25 and 30 years were selected. For each species of tree, three climbing trees were selected, and leaves collected randomly in four directions (east, south, west, and north) of column, and in four directions (up, down, left, and right) at 1.0–2.5 m above ground, depending on the plant structure with a pruner. Five sample trees were collected for each climbing tree species, as well as 10 pieces in each direction per plant (200 pieces in total for five plants) for H. nepalensis. E. fortune and F. pumila, from which 15 leaf samples in each direction were collected, for a total of 60 pieces (300 pieces in total for five plants) as their leaves were small. Samples for the determination of particulate retention were placed in self-sealing bags and brought back to the laboratory. The surface of each leaf was wiped clean and immediately soaked with liquid nitrogen. All samples were placed back into the self-sealing bags and stored in an ultra-low-temperature refrigerator prior to the enzyme assay.

Dust retention was determined using Przbysz’s method (Przbysz et al., 2014). Filter papers with pore sizes of 10, 2.5, and 0.2 μm were placed inside half-open weighing bottles in an oven at 60°C for 60 min (Dzierżanowski et al., 2011; Chiam et al., 2019; Kończak et al., 2021; Li X. L et al., 2021). The bottles were then covered for 60 min, removed, and placed in a drying dish until a constant weight was obtained. The dried filters were placed on an analytical balance (readability = 0.0001 g) (ME204E, Mettler Toledo, Columbus, OH, United States) and weighed to obtain both the initial weight of each filter paper (g) and the total initial weight of the weighing bottle and filter paper (g).

Particulate matter on the surface of the leaves that could be washed off by water was referred to as “particulate matter on the surface (sPM).” Leaves were placed into beakers with deionized water and stirred for 1 min with a glass rod. The beakers were placed in an ultrasonic cleaner (JP-020, Skymen, Guangdong, China) and shaken for 1 min to remove the PM adsorbed on the leaf surface. The resulting solution was passed through a standard splitting sieve (160 mesh, pore size of about 100 μm), followed by filtering in turn through filter papers with pore sizes of 10 and 2.5 μm to obtain three particle size ranges: large particles of 10–100 μm, coarse particles of 2.5–10 μm, and fine particles of 0.2–2.5 μm. The filter paper was placed into an oven to dry, removed, and left to stand in a drying dish. After reaching a constant weight, the weight of each filter paper after filtration (g) and the total weight of the weighing bottle and filter paper after filtration (g) were determined using an analytical balance. The amount of “surface PM” attached to each sample at each particle size was calculated using the weight difference method.

The mass of PM contained in wax layer (wPM) was detected the rinse and weight method (Dzierżanowski et al., 2011; Chiam et al., 2019). Clean leaf samples were rinsed with deionized water and then rinsed again for 40 s in 150 cm³ of chloroform to dissolve the epicuticular waxes and wash the micro-particles fixed in the wax to obtain the wPM fraction. The liquid obtained from the rinsing was filtered continuously through a filter according to the procedure used to filter aqueous fractions containing surface PM, with the difference that the chloroform remaining after filtration was poured into an empty pre-weighted clean flask. After the chloroform evaporation (more than 10 or even 48 h), the flasks were weighed on a laboratory microbalance. Then, the difference between the masses of the empty flask and flask containing wax was determined. Filtration was performed following the same procedure used for sPM washing. After drying, the filters were also weighed and the PM was determined based on the difference in weight.

After drying, the leaves were scanned using an electronic scanner (V370, Epson, Nagaon, Japan) and then imported into ImageJ 1.48 software to determine the leaf area, with the leaf area of each sample obtained as S (cm²).

Four leaves from each climbing plant were selected from the upper, middle, and lower well-developed mature parts of the climbing plant. The analytical instrument used for the measurements was a Li-6400 portable photosynthesis meter (Li-COR, Lincoln, NE, United States). Gas exchange parameters, including the maximum net photosynthetic rate (Pn), stomatal conductance (Gs), intercellular carbon dioxide (CO₂) concentration (Ci), and transpiration rate (Tr), were measured in May, August, and November 2021 and January 2022 for the tagged leaves. Mature leaves with healthy growth and a similar growth status were selected for measurements from 09:00 to 11:00 after eight consecutive days without precipitation.

About 0.5 g of intact and similarly colored leaf samples were randomly selected from each package of leaves, cut into small pieces, and placed in 5 ml centrifuge tubes with 2 ml of phosphate buffer. The samples were ground with an automatic sample rapid grinder (JXFSTPRP-24, Jingxin Technology, Shandong, China) and then the centrifuge tube was rinsed with extraction medium. The final volume was fixed to 5 ml with the extraction medium, and then the supernatant was extracted with a ST8R centrifuge (Thermo Fisher Scientific, Waltham, MA, United States) at 4,500 rpm for 15 min. The supernatant was then extracted with a 10 ml pipette for the determination of superoxide dismutase (SOD) and peroxidase (POD) activity. Enzyme extract (2.5 ml) was added to 2.5 ml of distilled water for the determination of malondialdehyde (MDA) and soluble protein content.

About 0.5 g of intact and similarly colored leaf samples were randomly selected from each package of leaves, cut into small pieces, and placed in 5 ml centrifuge tubes with 2 ml of phosphate buffer. The samples were ground with an automatic sample rapid grinder (JXFSTPRP-24, Jingxin Technology, Shandong, China) and then the centrifuge tube was rinsed with extraction medium. The final volume was fixed to 5 ml with the extraction medium, and then the supernatant was extracted with a ST8R centrifuge (Thermo Fisher Scientific, Waltham, MA, United States) at 4,500 rpm for 15 min. The supernatant was then extracted with a 10 ml pipette for the determination of superoxide dismutase (SOD) and peroxidase (POD) activity. Enzyme extract (2.5 ml) was added to 2.5 ml of distilled water for the determination of malondialdehyde (MDA) and soluble protein content.

The experimental data were subjected to both a one-way and two-way analysis of variance (ANOVA) using SPSS22.0 software, with multiple testing applied to detect whether statistical differences existed between the treatments. Independent sample t-tests were used to rapidly analyze the differences between the physiological and biochemical indices of plants located at the campus and the overpass. Pearson’s correlation coefficient was used to analyze the relationship between plant dust retention capacity and photosynthetic gas parameters. Plots were constructed using Origin 2016 software.

There were significant differences in the mass of the different sizes of PM among the different species in spring (Figure 1). Both campus and overpass H. nepalensis had a significantly higher adsorption capacity for total suspended particulate (TSP) and the different particle size classes of PM than the other two climbers, with overpass H. nepalensis adsorbing sPM (18.97 ± 2.03 μg cm⁻²) (Figure 1A), sPM10–100 (12.51 ± 1.56 μg cm⁻²) (Figure 1B), sPM2.5–10 (5.13 ± 0.38 μg cm⁻²) (Figure 1C), and sPM0.2–2.5 (1.33 ± 0.15 μg cm⁻²) (Figure 1D). In addition, the adsorption of PM2.5–10 was significantly higher (p < 0.05) in F. pumila at the overpass than in E. fortunei, and the adsorption of PM0.2–2.5 was significantly higher (p < 0.05) in E. fortunei than in F. pumila.

The maximum retention of the different particle sizes by H. nepalensis in summer at the overpass was sPM (15.94 ± 1.67 μg cm⁻²) (Figure 2A), sPM10–100 (11.99 ± 0.75 μg cm⁻²) (Figure 2B), sPM2.5–10 (2.60 ± 0.13 μg cm⁻²) (Figure 2C), and sPM0.2–2.5 (1.35 ± 0.11 μg cm⁻²) (Figure 2D). Campus H. nepalensis adsorbed significantly more sPM and sPM10–100 than the other two climbers, and campus H. nepalensis and E. fortunei adsorbed significantly more sPM0.2–2.5 than F. pumila (p < 0.05) There was no significant difference in the adsorption of sPM2.5–10 among the three plants. The adsorption of sPM and the different particle size classes by H. nepalensis at the overpass was significantly higher than that by E. fortunei (p < 0.05). There was no significant difference between the adsorption of sPM, sPM10–100 and sPM2.5–10 by H. nepalensis and F. pumila, but the adsorption of sPM0.2–2.5 by the three climbing plants was significantly different (p < 0.05).

In autumn, H. nepalensis at the overpass had the highest adsorption among the particles of different particle sizes, sPM (14.41 ± 2.14 μg cm⁻²) (Figure 3A), s PM10–100 (9.29 ± 1.97 μg cm⁻²) (Figure 3B), sPM2.5–10 (3.50 ± 0.45 μg cm⁻²) (Figure 3C), and sPM0.2–2.5 (1.61 ± 0.13 μg cm⁻²) (Figure 3D). A one-way ANOVA showed that there was no significant difference in sPM adsorption by the three climbers at the campus. The adsorption of sPM10–100 and sPM2.5–10 by H. nepalensis was significantly higher than that of F. pumila and E. fortunei (p < 0.05), and the adsorption of sPM0.2–2.5 by H. nepalensis and F. pumila was significantly higher than that of E. fortunei (p < 0.05). The adsorption of sPM and sPM2.5–10 by H. nepalensis and F. pumila at the overpass was significantly higher than that of E. fortunei (p < 0.05). The adsorption of sPM10–100 by all three plants at the overpass was significantly different (p < 0.05), and the adsorption of sPM0.2–2.5 by F. pumila was significantly higher than that of E. fortunei (p < 0.05).

The maximum adsorption of the different particle sizes in winter at the overpass for H. nepalensis was: sPM (21.35 ± 2.01 μg cm⁻²) (Figure 4A), sPM10–100 (14.99 ± 1.36 μg cm⁻²) (Figure 4B), sPM2.5–10 (3.49 ± 0.31 μg cm⁻²) (Figure 4C), and sPM0.2–2.5 (2.87 ± 0.15 μg cm⁻²) (Figure 4D). The adsorption patterns of the three climbers for sPM and sPM10–100 were consistent across sampling sites, with all having a significantly higher adsorption of sPM and sPM10–100 by H. nepalensis than by F. pumila and E. fortunei (p < 0.05). The adsorption of sPM2.5–10 and sPM0.2–2.5 at the campus by H. nepalensis and F. pumila was significantly higher than that of E. fortunei (p < 0.05). The adsorption of sPM2.5–10 by the three climbers at the overpass was not significantly different, and the adsorption of sPM0.2–2.5 by H. nepalensis was significantly higher than that of the other two climbers (p < 0.05).

The adsorption of PM, PM10–100, and PM2.5–10 was highest in the leaf wax layer of H. nepalensis in the spring at the overpass, with values of 13.04 ± 2.28 μg cm⁻² (Figure 5A), 7.05 ± 0.69 μg cm⁻² (Figure 5B), and 5.37 ± 0.35 μg cm⁻² (Figure 5C), respectively, and the adsorption of PM0.2–2.5 was highest in the leaf wax layer of E. fortunei (1.68 ± 0.3 μg cm⁻²) (Figure 5D). The adsorption of PM and PM2.5–10 by the waxy layer of H. nepalensis leaves in both the campus and overpass was significantly higher than that of the other two climbers (p < 0.05), and the adsorption of PM10–100 by the waxy layer of H. nepalensis and E. fortunei leaves was significantly higher than that of F. pumila (p < 0.05). The adsorption of PM0.2–2.5 by the waxy layer of H. nepalensis leaves at the campus was significantly higher than that of the other two climbers (p < 0.05). The adsorption of PM0.2–2.5 by the waxy layer of the leaves of E. fortunei at the overpass was significantly higher than that of the other two climbers (p < 0.05).

The adsorption of PM by the waxy layer of H. nepalensis leaves at both the campus and the overpass reached a maximum in summer. The adsorption of PM of different particle sizes by the H. nepalensis leaves at the overpass was: wPM (10.83 ± 0.52 μg cm⁻²) (Figure 6A), wPM10-100 (7.93 ± 0.36 μg cm⁻²) (Figure 6B), and wPM0.2–2.5 (1.03 ± 0.15 μg cm⁻²) (Figure 6D). Overall, the adsorption of PM, PM10–100 (Figure 5C), and PM0.2–2.5 by the waxy layer of campus H. nepalensis leaves was significantly higher than that of the other two climbers (p < 0.05), and the adsorption of PM, PM2.5–10, and PM0.2–2.5 by the waxy layer of H. nepalensis leaves at the overpass was significantly higher than that of the other two climbers (p < 0.05).

The adsorption of PM by the waxy layer of H. nepalensis leaves at the overpass reached a maximum in autumn for PM (10.83 ± 0.52 μg cm⁻²) (Figure 7A), PM10–100 (7.93 ± 0.36 μg cm⁻²) (Figure 7B), PM2.5–10 (1.87 ± 0.21 μg cm⁻²) (Figure 7C), and PM0.2–2.5 (1.03 ± 0.15 μg cm⁻²) (Figure 7D). The adsorption pattern of the leaf wax layer of the three climbing plants at the campus was consistent for PM, PM10–100, and PM0.2–2.5, all of which were significantly higher for H. nepalensis than for the other two plants (p < 0.05). The adsorption pattern of the three plants at the overpass was consistent for PM and PM10–100, i.e., H. nepalensis was significantly higher than the other two plants (p < 0.05). The adsorption pattern of the three climbing plants was also consistent for PM2.5–10, while there was no significant difference in the adsorption of PM0.2–2.5 among the three climbers (p < 0.05).

The adsorption of different particle sizes of PM by the leaf wax layer of all three plants in the winter at the campus was significantly higher for H. nepalensis than for the other two plants (p < 0.05). The adsorption of different particle sizes of PM by the leaf wax layer of H. nepalensis in the overpass was: wPM (10.83 ± 0.52 μg cm⁻²) (Figure 8A), wPM10–100 (7.93 ± 0.36 μg cm⁻²) (Figure 8B), and wPM2.5–10 (7.93 ± 0.36 μg cm⁻²) (Figure 8C). The adsorption patterns of PM, PM10–100, and PM2.5–10 were consistent among the three plants at the overpass, i.e., H. nepalensis was significantly higher than the other two plants (p < 0.05), and there were significant differences (p < 0.05) in the adsorption of PM0.2–2.5 for all three climbers at the overpass (Figure 8D).

The P_n, C_i, and T_r of F. pumila leaves on campus were significantly higher than those of the other two climbers in spring. The G_s of H. nepalensis leaves reached a maximum on campus, which was significantly higher than that of E. fortunei. On the overpass, the P_n of F. pumila leaves was significantly higher than that of E. fortune. The Gs and Tr of F. pumila and H. nepalensis leaves were significantly higher than those of E. fortunei. The C_i of H. nepalensis leaves was significantly higher than that of E. fortune (Table.1).

As can be seen from Table 1, the P _n , C _i , and T _R of E. fortunei leaves at the campus were significantly higher than for the other two climbers in summer. Additionally, the P _n , T _R , G, and C _i of E. fortunei leaves all reached a maximum in the summer. The P _n , G _s , and T _R of the H. nepalensis and F. pumila leaves at the interchange were significantly higher than those of E. fortunei.

The P _n , G _s , C _i , and T _R of H. nepalensis and F. pumila leaves were significantly higher than those of E. fortunei in autumn. The P _n , and G _s of H. nepalensis leaves reached a maximum at the overpass, where the P _n , of H. nepalensis leaves was significantly higher than those of the other two plants. The C _i and T _R of H. nepalensis and E. fortunei leaves were significantly higher than those of F. pumila at the overpass, and the G _s of the three plants was not significantly different.

The P _n of F. pumila leaves was significantly higher than that of E. fortunei and H. nepalensis leaves in winter at the campus. The C _i of both E. fortunei and H. nepalensis leaves was significantly higher than that of F. pumila. The G _s and transpiration rate of H. nepalensis leaves reached a maximum in winter, while the G _s of H. nepalensis and F. pumila leaves was significantly higher than that of E. fortunei, and the T _R of H. nepalensis leaves was significantly different. The T _R of H. nepalensis leaves was significantly higher than that of the other two plants.

As shown in Table 4, the interaction of tree species and sampling site had highly significant effects (p < 0.001) on the sPM, sPM10–100, sPM2.5–10, wPM, wPM10–100, POD activity, MDA content, P_n, and G_s of climbing plants. The effect of tree species on the wPM2.5–10 and wPM0.2–2.5 of all climbing plants was highly significant (p < 0.01). Sampling site had a highly significant effect on the sPM0.2–2.5 and T_r of climbing plants (p < 0.01). Tree species had a highly significant effect on the soluble protein content, C_i, and T_r of climbing plants (p < 0.001). The interaction of tree species and sampling site had highly significant effects on the POD activity (p < 0.001), soluble sugar content (p < 0.01), and sPM10–100, wPM2.5–10, wPM0.2–2.5, MDA, G_s, and T_r (p < 0.05).

AS shown in Figure 12, the SOD activity was positively correlated with MDA content, cell concentration (p < 0.01), G_s, sPM, sPM10–100, wPM, wPM10–100 (p < 0.05), and negatively correlated with the POD activity and soluble sugar content (p < 0.05). The POD activity was positively correlated with the MDA content (p < 0.05), and negatively correlated with the soluble protein content, P_n, G_s (p < 0.01), and T _R and wPM2.5–10 (p < 0.05). The MDA content was positively correlated with the soluble protein content, sPM, and sPM0.2–2.5 (p < 0.01), and sPM10–100 and sPM2.5–10 (p < 0.05), and negatively correlated with the total soluble sugar content (p < 0.01), and G_s and C_i (p < 0.05). The soluble protein content was positively correlated with the T_R (p < 0.01) and P_n (p < 0.05), and negatively correlated with the C_i (p < 0.01). The soluble sugar content was negatively correlated with the P_n (p < 0.01). The P_n was positively correlated with G_s and the T_r (p < 0.01), and negatively correlated with sPM, sPM10–100, sPM2.5–10, wPM, and wPM10–100 (p < 0.01), and sPM0.2–2.5 (p < 0.05). G_s was positively correlated with C_i and the T_r (p < 0.01), and negatively correlated with sPM, sPM10–100, and wPM10–100 (p < 0.05). The sPM was positively correlated with sPM10–100, sPM2.5–10, sPM0.2–2.5, wPM, wPM10–100, wPM2.5–10, and wPM0.2–2.5 (p < 0.01). The sPM10–100 was positively correlated with sPM2.5–10, 1 sPM0.2–2.5, wPM, wPM10–100, wPM2.5–10, and wPM0.2–2.5 (p < 0.01). The sPM2.5–10 was positively correlated with sPM0.2–2.5, wPM, wPM10–100, wPM2.5–10, and wPM0.2–2.5 (p < 0.01). The sPM0.2–2.5 was positively correlated with wPM, wPM10–100, wPM2.5–10, and wPM0.2–2.5 (p < 0.01). The wPM was positively correlated with wPM10–100, wPM2.5–10, and wPM0.2–2.5 (p < 0.01). The wPM10–100 was positively correlated with wPM2.5–10 and wPM0.2–2.5 (p < 0.01). The wPM2.5–10 value was positively correlated with wPM0.2–2.5 (p < 0.01).

The adsorption of PM by the three evergreen climbers followed the order of H. nepalensis > F. pumila > E. fortunei, indicating that H. nepalensis has a strong adsorption capacity for PM, which is related to its larger individual leaf area, sparsity of scales on the lower surface, and obvious lateral and reticulate veins on the leaf surface. E. fortunei has thin leathery leaves with fine lateral veins and inconspicuous veinlets, and displayed the fewest seasonal differences among the three climbers, indicating a consistent ability to adsorb PM. F. pumila leaves have a glabrous surface with dorsal short hairs and reticulate veins that protrude in the form of a honeycomb, which is conducive to the adsorption of PM. The roughness of the leaf surfaces of the different plants varied, and these differences had a great influence on the adsorption of PM (Wang et al., 2006; Li M et al., 2021). The ability of plants to adsorb PM is closely related to the micromorphology of the leaf surface, and dense, narrow, and deep grooves on the leaf surface could favor the adsorption of PM by plants (Li et al., 2019; Zha et al., 2019). In this study, the adsorption of PM by all three climbing plants was higher at the overpass than at the campus, which was because the overpass was located in the middle of a road segregation zone that was affected by vehicle exhaust. Atmospheric PM pollution was more serious than in other areas, which was consistent with the findings of Beckett et al., 2000 and Freer-Smith et al., 2005. The amount of PM adsorbed by leaf surfaces and wax for all three climbing plants followed the order of winter > spring > summer > autumn, which may be due to the higher level of atmospheric PM pollution in winter than in other seasons, resulting in plants adsorbing more PM. The occurrence of the maximum dust retention in winter was also reported by other researchers (Prajapati and Tripathi, 2008; Zha et al., 2019).

It was found that PM deposited on the leaf surface increased the leaf temperature and interfered with photosynthetic CO₂ exchange and transpiration rates by shading and impeding diffusion (Gimeno and Deltoro, 2000). Zia-Khan et al. (2015) found that there may be stomatal occlusion on the leaves of plants growing near coal fields, which may result in a decrease in G _s . In this study, it was found that the P _n of the three climbing plants was more strongly affected under the high PM pollution conditions at the overpass (Table 1). Compared to the overpass, at the campus, the P_n and G_s of the three climbing plants decreased and the C_i concentration increased, indicating that high levels of PM pollution inhibited photosynthesis through its effect on G _s (Guidi et al., 2011; Gajić et al., 2013). The toxic elements from particulate matter can inhibit potosynthetic gas exchange parameters of climbing plants, which consistent with Guidi et al. (2011). The Pearson correlation coefficients between the plant adsorption of different particle sizes and photosynthetic parameters showed that different particle sizes were highly significantly negatively correlated with the net P _n and G_s of plant leaves, i.e., (Table 1). A similar finding was also reported that photosynthesis nad stomatal conductance declined over time at elevated PM2.5, with large variations with PM2.5 concentrations (Yu et al., 2018). The transpiration of stomata was found to make the leaf surface wetter and enhance the deposition of PM. Among the plants investigated in this study H. nepalensis was found to have the highest transpiration rate, which is one of the reasons why it had the greatest ability to adsorb PM (Hinds, 1999). Tree species that display resistance to air pollution are reported to have the highest respiration rates (Lorencplucinska, 1980), and an analysis of the three plants in this study showed that F. pumila had the highest P_n., indicating a strong resistance to particulate pollution.

The degree of oxidative stress in plants depends on the ability of the plant antioxidant capacity system to inhibit superoxide, hydroxyl, and peroxyl radicals at the cellular level (Mroczek-Zdyrska and Wojcik, 2012; Bisoi et al., 2017). Antioxidant enzymes, such as SOD and POD, are physiological and biochemical indicators that are critical for plant stress resistance and the mitigation of oxidative stress (Radwan and Eldeen, 2012).

The three plant species in this study showed a general trend of higher SOD (Figure 10) and POD levels (Figure 11) in all seasons at the overpass than at the campus, indicating an increase in oxidative stress generated by PM pollution on plants. The increase in SOD and POD enzyme activity indicated an increase in free radicals in plant cells caused by PM pollution, with both SOD and POD scavenging these free radicals to reduce the pollution damage to plants (Verma and Dubey 2003; Xu et al., 2010; Sharma et al., 2012; Panda et al., 2018). In addition, in all three plants there was a general trend for a higher MDA content in all seasons at the overpass than at the campus (Figure 12), which was due to the higher level of cell membrane damage in the overpass plants following their exposure to higher pollution levels than the plants at the campus. The higher leaf SOD activity of the three plant species in the winter at both the overpass and campus in this study may be related to high dust levels, high reactive oxygen species production rates, high atmospheric particulate matter pollutant concentrations, and low available light levels during the winter (Singh et al., 2020).

The soluble protein content of the three climbing plants displayed different trends in each season (Table. 2), and the soluble protein content of F. pumila leaves in the summer and winter overpass environments was higher than that of the campus, which may be due to the more extreme weather conditions in summer and winter, when plants actively accumulate soluble proteins to maintain their osmoregulation and sustain higher levels of growth. Kumar and Dubey et al. found that pollutants from automobile exhaust emissions may have an inhibitory effect on protein synthesis (Kumar and Dubey, 1998). The decrease in the foliar protein content in this study may have been due to the breakdown of existing proteins or a decrease in protein synthesis from scratch (Iqba et al., 2000; Singh et al., 1998).

In the three climbing plants investigated in this study there was a general trend for lower soluble sugars in all seasons in the overpass than in the campus, indicating that the three climbing plants were more affected by stress at the overpass (Table. 3). The decrease in the soluble sugar content of damaged leaves may correspond to the photosynthetic inhibition or stimulation of respiration rate (Tzvetkova and Kolarov, 1996). Similarly, Bucker and Ballach et al. found that the level of soluble carbohydrates decreased due to the fumigation of a mixture of O₃, SO₂, and nitrogen dioxide (NO₂) in the leaves, and that the decrease in the soluble sugar content may be the result of increased metabolic depletion under stress (Bücke and Ballach, 2010). Thus, we believe that more particulate matter accumulated in the leaves leads to shade-induced decrease in photosynthetic efficiency, could be responsible for decline in sugar content in the leaves. A similar finding was also reported that heavy metals containing particulate matter can decrease the sugar levels in pistia stratiotes, spirodela polyrrhiza, Eichhornia crassipes, Lagerstomia speciose and Acacia moniliformis (Mishra and Tripathi, 2008; Gupta et al., 2011).