Leaf Functional Traits Differentiation and Its Trade-off Strategies of Urban Plant are Related to Atmospheric Particulate Pollution

Background: Functional trait-based ecological research has been instrumental in advancing our understanding of understanding of environmental changes. It is still, however, unclear how the functional traits of urban plants respond to atmospheric particulate pollution, and what trade-off strategies are shown. In order to explore the variation of plant functional traits with urban atmospheric particulate pollution gradient, we divided atmospheric particulate pollution into three levels according to road distance, and measured the variation of six key leaf functional traits and their trade-off strategies. Results: Here, we show that the functional traits of plants can be used as predictors or indicators of the response of plant to urban atmospheric particulate pollution. Within studies, there was a positive correlation between leaf thickness, leaf dry matter content, leaf tissue density, stomata density and leaf dust deposition. While chlorophyll content index and specic leaf area were negatively correlated with the leaf dust deposition. Plants improve the eciency of gas exchange by optimizing the spatial distribution of stomata of leaves. Dust deposition promotes the regular distribution of stomata. Due to the pressure of atmospheric particles, urban plant shows a trade-off relationship of economics spectrum traits at the leaf level. Taken together, these results indicate that urban atmospheric particulate pollution is the main factor causing the variation of plant functional traits. Conclusion (cid:0) Under the inuence of urban atmospheric particulate matter, plant show a "slow investment-return" type in the global leaf economics spectrum, with lower specic leaf area, lower chlorophyll content, larger leaf thickness, higher leaf dry matter content, higher leaf tissue density and higher stomatal density. This nding provides a new perspective for understanding the resource trades-off strategy of plants adapting to air pollution environment.


Background
In a world of increasing urbanization, atmospheric particulate matter pollution mitigation is currently one of the most important issues of city planning. Urban trees are of central importance for this issue because they facilitate the deposition of various gases and particles and affect microclimate and air turbulence (Seinfeld, 1989;Kulshreshtha et al., 2009;Lelieveld, J., 2002;Wang et al., 2020). The urban atmospheric particulate contains heavy metal and other harmful components, which seriously affects the health of urban residents (Lelieveld et al., 2015;Wang et al., 2020). In China, the government attaches great importance to the prevention and control of air pollution. In recent years, despite the continuous improvement of the ambient air quality in China, the treatment effect is not stable, especially in the autumn and winter seasons in Beijing-Tianjin-Hebei region and its surrounding areas. Data show that in the autumn and winter of 2018-2019, the average concentration of PM2.5 in Beijing-Tianjin-Hebei and its surrounding areas increased by 6.5% year-on-year, and the number heavy polluted days increased by 36. 8% year-on-year (Chang et al., 2019;Gao et al., 2020). Due to the special terrain and meteorological conditions, Beijing has become the gathering center of atmospheric pollutants in North China (Yang et al., 2005;Ren et al., 2008). In addition, the acceleration of urbanization and the rapid development of transportation also bring about the prevention and control of air pollution. serious challenge.
At present, it is impossible to completely rely on pollution sources control to solve environmental problems. Natural removal mechanism is an effective way to relieve the pressure of urban air pollution, and urban trees are the key factor (Nowak et  Euonymus japonicus is one of the most planted landscape trees in Beijing, which plays an important role in urban ecological and social bene ts. A study on the interception ability of 29 urban greening plants in Beijing to ne particles in the atmosphere shows that the interception ability of Euonymus japonicus to PM2.5 in shrubs is the second only to Buxusmicrophylla (Nowak et al., 2006). Based on this, this study takes Euonymus japonicus as the research object, taking the leaf functional traits as the breakthrough point. The overall goal is to study the relationship between leaf functional traits and the environment. The speci c objectives are as follows, (1) Evaluate the response and trade-off rule of plant leaf traits to different dust deposition under the in uence of atmospheric particulate matter. (2) Whether the relationship of leaf traits accords to the global leaf economics spectrum in the environment polluted by atmospheric particulate matter.

Results
Leaf dust deposition under different levels of atmospheric particulate matter pollution As shown in Fig. 1, in three different locations of urban highway, the dust collection capacity of Euonymus japonicus leaves gradually decreases from the main road. We found that the dust deposition of Euonymus japonicus leaves in the middle of the main road (T3) was (0.0052 ± 0.0022 g) > that between the relief road and the sidewalk (T2) (0.0025 ± 0.001g) > that of the outer sidewalk (T1) (0.0019 ± 0.001g). On the main road, due to the large tra c volume, not only a large amount of tra c waste gas is emitted, but also the dust on the ground was driven by the air ow of the vehicle, which was an open pollution source into the atmosphere and an important part of the total suspended particulate matter in the ambient air. Such atmospheric particulate matter mainly diffuses to both sides. The auxiliary road vehicles were mainly non-motor vehicles, and the tra c volume was small, and the suspended particulate matter in the atmosphere was relatively low. However, the main body of the sidewalk was pedestrians, and the particulate matter emitted by vehicle exhaust is greatly reduced compared with the main road. Therefore, the dust accumulation of Euonymus japonicus planted in different locations of urban streets was signi cantly different. In general, the atmospheric particulates matter gradually decreases along the center of the road to both sides.

Effect of urban atmospheric particulate matter on leaf functional traits
Leaf is an important organ for plants to obtain energy, resources and nutrition, and it is also the most  Figure 4 shows the comparison of leaf function traits (leaf thickness, chlorophyll content index, speci c leaf area, leaf dry matter content, leaf tissue density and stomatal density) of Euonymus japonicus at different dust deposition rates. In general, with the increase of dust fall, leaf thickness, leaf dry matter content, leaf tissue density and stomatal density increased, while the relative content of chlorophyll and speci c leaf area decreased. Compared with T1, the leaf thickness of T2 and T3 were signi cantly increased, and the difference reached signi cant level (P < 0.05) and extremely signi cant level (P < 0.01). The leaf dry matter content of T3 was signi cantly higher than that of T1 and T2 (P < 0.01), but the difference between T1 and T2 was not signi cant. Compared with T1, the stomatal density of T2 and T3 increased signi cantly (P < 0.01). The chlorophyll content index in T2 and T3 decreased signi cantly, and reached an extremely signi cant level between T1, T2 and T3 (P < 0.01). The speci c leaf area of T3 was signi cantly larger than that of T1 and T2 (P < 0.01), but there was no signi cant difference between T1 and T2.
Studies have shown that speci c leaf area is closely related to the growth and survival strategies of plant, which can represent the adaptability of plants to the environment and the ability to obtain resources  (Wellstein et al., 2017). In this study, the speci c leaf area of Euonymus japonicus in an environment polluted by atmospheric particulates showed a decreasing trend. This indicates that under the pollution of atmospheric particulate matter, the Euonymus japonicus enhances its nutrient deposition capacity by reducing the speci c leaf area. Most of the energy of plant photosynthesis comes from the light energy captured by photosynthetic pigments, so the chlorophyll content is closely related to the plant photosynthetic capacity (Kleinschmidt et al., 2020). In this study, the chlorophyll content decreased signi cantly due to the increase in the amount of dust trapped on the leaf surface, which may be due to the cover of the leaf surface particles, which led to a decrease in the area of light resources captured (Fig. 2). The leaves with a large amount of dust on the leaf surface have higher tissue density and dry matter content, and have stronger drought tolerance and defense capability. This shows that under the in uence of atmospheric particulate matter, Euonymus japonicus mostly uses nutrients for the construction of defense structure, and reduces the damage of atmospheric particulate matter to leaves by increasing leaf tissue density and leaf dry matter content. Stomata is an important organ for gas exchange between plants and the atmosphere, and plays an extremely important role in regulating the carbon and water cycle of the ecosystem (Woodward 1987;Masterson 1994;Allen et al., 2001). Under drought stress, stomatal density decreases to prevent water loss via transpiration (Woodward et al., 2002). In this study, we found that the dust deposition on the leaf surface increased the density of leaf stomata. This may be because atmospheric particles may block some of the pores and weaken the gas exchange function of stomata. In this case, plants can ensure normal gas exchange and balance of water circulation by increasing stomatal density.
In uence of atmospheric particulates on the spatial distribution of stomata As shown in Fig. 3, the distribution characteristics of stomatal spatial pattern of Euonymus japonicus were obviously different in different atmospheric particulate matter environments. The stomata of Euonymus japonicus were aggregated at the scales of 0 ~ 42µm(T1), 0 ~ 46µm(T2) and 0 ~ 54µm(T3), randomly distributed at the scales of 42 ~ 69µm(T1), 46 ~ 64µm(T2) and 54 ~ 60µm(T3), and randomly distributed at the scales of 69 ~ 10 µm. This indicates that atmospheric particulates have changed the spatial distribution pattern of stomata of Euonymus japonicus, showing that the spatial scale of stomata under different atmospheric particulates has changed from random distribution to uniform distribution.
With the increase of the amount of atmospheric particulate matter trapped in leaves, the spatial distribution pattern of stomata of Euonymus japonicus becomes more regular. We suspect that this may be a regulatory strategy adopted by plants to deal with atmospheric particulate matter. Euonymus japonicus can prevent the in uence of atmospheric particulate matter on water diversion of leaves by adjusting the distribution pattern of stomata.
Regression analysis of leaf functional traits and leaf dust deposition Figure 4 was a linear t between dust deposition and leaf functional traits. There was a positive correlation between leaf thickness, leaf dry matter content, leaf tissue density, stomatal density and leaf dust deposition. However, chlorophyll content index, speci c leaf area and dust deposition of leaves were negatively correlated. The R 2 values from large to small were speci c leaf area (0.2387), leaf thickness (0.1999), leaf dry matter content (0.1707), leaf tissue density (0.1391), chlorophyll content index (0.1128) and stomatal density (0.0021), and the corresponding root mean square errors were 0.0380 and 0.080, respectively. Therefore, the response of speci c leaf area to dust deposition was the most severe.

Correlation between leaf functional traits of Euonymus japonicus
By establishing the relationship between leaf functional characters of Euonymus japonicus affected by atmospheric particulate matter pollution, the relationship between leaf functional traits and their functions is further discussed, and the trade-off effect of plant traits on limited resources is clari ed. It can be seen from Fig. 5 that there is a certain quantitative relationship between different functional traits due to environmental pressure and plant trade-off strategy. Stomatal density was positively related to leaf thickness and dry matter content. The leaf tissue density was negatively correlated with speci c leaf area and positively correlated with dry matter content. The dry matter content was negatively correlated with leaf thickness, chlorophyll content index and speci c leaf area. The speci c leaf area was negatively correlated with leaf thickness, and positively correlated with chlorophyll content index. The leaf thickness was negatively correlated with chlorophyll content index.
As shown in Table 1, according to the principle that the eigenvalue was greater than 1, two principal components were extracted (the eigenvalues were 2.429 and 1.226 respectively). The contribution rates of these two principal components were 40.5% and 20.4%, respectively, and the cumulative contribution rate was 60.9%, which indicates that these two principal components were the main factors in the change of leaf functional traits. The initial factor loading matrix of the principal components (Table 1)   Stomatal is an important window for gas and water exchange between plants and the outside world (Allen et al., 2001;Field et al., 2010). This study shows that atmospheric particulate matter has a direct impact on stomatal density of plants, and the aggravation of pollution promotes the increase of stomatal density. As atmospheric particles were deposited on the blade surface, some air holes may be blocked, which may lead to an imbalance of the blade's ability to exchange gas with the outside world. Therefore, increasing the stomatal density of plants may be a regulation strategy for dealing with atmospheric particulate matter. At the same time, we also found that atmospheric particles have a signi cant impact on stomatal distribution pattern. Under the in uence of atmospheric particles, urban plants can improve their gas exchange function by adjusting stomatal structure and optimizing the spatial distribution pattern of stomata, which is bene cial to their normal physiological metabolism and growth. . In other words, in a limited resource environment, plants will optimize resource allocation among functional traits (Tomáš et al., 2012). In this study, due to the pressure of atmospheric particulate environment, Euonymus japonicus showed a trade-off strategy at the leaf level, which constituted a complex and orderly trade-off relationship of economics spectrum characteristics. Euonymus japonicus will adjust, transform or compensate its own functions according to its own resource conditions in the urban environment, so as to achieve and balance the three purposes of "survival, growth and reproduction", which is nally manifested in the functional characteristics of plant leaves. For example, due to the in uence of dust deposition, the speci c leaf area is signi cantly reduced, which indicates that plants may use a large part of materials to build protective structures, so as to reduce the damage of atmospheric particles to leaves. The decrease of speci c leaf area indicates that the greater the leaf area per unit mass of the plant, the thicker the leaf are. At this time, the more carbon the leaves were used to build the protective structure. It is manifested by the increase of leaf dry matter content and leaf tissue density at the trait level.
It was pointed out that the leaf economics spectrum was a series of interrelated and coordinated atmospheric particulate matter, Euonymus japonicus shows a "slow investment-pro t" type in the global leaf economics spectrum, which has low speci c leaf area, low chlorophyll content, large leaf thickness, high leaf dry matter content, leaf tissue density and stomatal density (See Fig. 7).

Conclusions
This study discussed the response of leaf functional traits to atmospheric particulate pollution, and provided theoretical basis for monitoring and predicting the impact of future environmental pollution changes on plants and ecosystems. The analysis leads to the following conclusions.
(1) With the increase of dust deposition, leaf thickness, leaf dry matter content, leaf tissue density and stomatal density increased, while chlorophyll relative content and speci c leaf area decreased.
(2) Under the in uence of atmospheric particulates, Euonymus japonicus can improve its gas exchange e ciency by optimizing the spatial distribution characteristics of stomata. The larger the dust deposition on leaves, the more regular the spatial distribution pattern of stomata.
(3) There is a positive correlation between leaf thickness, leaf dry matter content, leaf tissue density, stomatal density and leaf dust deposition. However, the relative content of chlorophyll, speci c leaf area and dust deposition of leaves were negatively correlated. The response of speci c blade area to dust deposition of blade was the most violent.
(4) Due to the pressure of atmospheric particulate environment, Euonymus japonicus showed a trade-off strategy at the leaf level, which constituted a complex and orderly trade-off relationship of economics spectrum characteristics. On the whole, under the in uence of urban atmospheric particulate matter, Euonymus japonicus shows a "slow investment-return" type in the global leaf economics spectrum, which is characterized by low speci c leaf area, low chlorophyll content, large leaf thickness, large stomatal density, high leaf dry matter content and high leaf tissue density.
At the leaf level, leaf functional traits may change among populations, the resource use strategy shifting to best suit the current environmental conditions. At the leaf level, leaf functional traits change in the environment of atmospheric particulate matter. The resource use strategy shifting to best suit the current urban environment. Our study found that leaf functional traits in Euonymus japonicus covaried in patterns consistent with the leaf economics spectrum. This discovery provides a profound understanding for the adaptation of leaf functional traits under the background of urban environmental change.

Materials And Methods
Sample plot setting and sampling As shown in Fig. 8, according to the characteristics of road dust, we divided the pollution degrees of atmospheric particles matter according to the distance from the main road. In September, 2019, the leaves of Euonymus japonicus were collected from the outside of sidewalk (T1), between sidewalk and relief road (T2) and the center of main road (T3). There was no rainstorm within 15 days on the sampling day, so as to avoid the in uence of rain erosion on dust deposition. We selected 60 plants for each treatment, and collected 3 mature and healthy leaves randomly from the top of each tree. And then the leaves were placed in a tray and sealed with plastic wrap. The leaf surface was not touched during the whole collection process.

Dust deposition measurement
After numbering the slow quantitative lter paper, weigh it (W 1 ) with a GH-252 electronic balance (accuracy 0.1 mg, Shanghai Youyi Instrument Co., Ltd., Shanghai, China) for later use. Filter the washed liquid with slow quantitative lter paper, and then the dry the lter paper to a constant weight by DHG-9143 electric heating blast drying cabinet (Shanghai Yitian Scienti c Instrument Co., Ltd., Shanghai, China) at a temperature of 60℃ (the difference between the two measured values does not exceed 0. 0002 g). After that, the dry lter paper (W 2 ) was weighed. At the same time, we calculate the quality difference(△W)of the blank control lter paper before and after drying (△W = W CK2 -W CK1 , W CK1 is the value of the blank control lter paper before drying, W CK2 is the value of the blank control lter paper after drying). Total dust deposition of blades W = W 2 -W 1 -△W. Medical sharp-nosed tweezers were used in the sampling and measuring process to avoid direct contact with the blade surface. The blades before and after cleaning are shown in Fig. 9.

Leaf functional traits measurement
The wight of fresh leaf (LFW) was weighed with JA1003N one-thousandth electronic balance (± 0.001g, Shanghai Jinghai Instrument Co., Ltd., Shanghai, China), and the measurement was completed within 30 minutes after removal. The leaf area (LA) was measured and automatically calculated with the V39 Temporary slide of stomata is made by imprinting method. A layer of transparent imprinting liquid is evenly coated on the back side of the leaf from the base to the tip of the leaf, avoiding the main vein, and the imprinting lm is torn off with tweezers to make a temporary Slides. Three slides are made for each leaf, and each slide is magni ed by XSP-20 optical microscope (Jiangnan Yongxin Optics Co., Ltd., Nanjing, China) and then randomly selected 5 elds of view (713.191µm×958.115µm) for image acquisition. The stomatal density is calculated using image J software.

Point pattern analysis of stomata
This analysis considers that each stomatal was a single point distributed on the blade surface, and the middle position of the stomatal opening was the position of this single point. Firstly, the selected micrographs were digitized under the same coordinate system by using the spatial distribution software ArcGIS 10.4, and the coordinate values of each pore of the selected photographs can be obtained. Then, the stomatal pattern was analyzed by using Programita Febrero 2014. Here, we used Ripley's K-Function, a spatial statistical analysis method, to carry out spatial analysis on the digitized stomatal distribution feature point. Ripley's K-Function is a kind of distribution accumulation function, which uses the secondorder matrix of all single points distances to explore the two-dimensional distribution patterns of these points on different scales. This paper adopted the paired correlation function g(r) and Ripley's K(r) function. The g(r) function is derived from the K(r) function. The K(r) function is de ned as taking any point as centering of the circle. "r" is the ratio of the expected number of points in the circle with the radius to the point density in the sample square. In the g(r) function, a circle is used to replace the circle in the traditional pattern analysis. Among them, the g(r) function can explain the neighbor density and eliminate the cumulative effect of the K(r) function, so the g(r) function is more intuitive than the cumulative calculation of the K(r) function. The expression of g(r) function is as follows (Wiegand & Moloney 2004), Under the assumption of complete spatial randomness, the g(r) value is above the envelope trace, indicating that the pores are clustered and distributed on the r scale. The g(r) value is located between the envelope traces, indicating that the pores are randomly distributed on the r scale. The g(r) value is below the envelope trace, indicating that the pores are uniformly distributed on the r scale.

Data analysis
All data were sorted in Excel 2020, and the data were analyzed and plotted with Origin 2019b. One-way analysis of variance (ANOVA) and LSD multiple comparisons were used to test the signi cance of the differences in leaf functional traits among different pollution gradients. The relationship between leaf characters was analyzed by linear regression, and the relationship between leaf characters and their comprehensive effects were comprehensively analyzed by principal component analysis. Figure 1 Dust deposition of plant leaves at different locations on urban streets. * indicates that the indicators have reached a signi cant difference at the P<0.05 level, and ** indicates that the indicators have reached a signi cant difference at the P<0.01 level. Same below.