Background

Rapid industrial development and urbanization have led to the emission of large amounts of harmful gases and particulate matter (PM) into the atmosphere [1], which has been particularly pronounced in develo** countries [2]. Atmospheric PM, especially PM2.5 (diameter ≤ 2.5 μm) and PM10 (diameter ≤ 10 μm), often contains heavy metals and can lead to urban haze, which has become the most serious air pollution issue in China in recent decades [3, 4]. Numerous studies have shown that plants can improve air quality in polluted areas by filtering PM from the air and retaining it on their unique leaf surface microstructures [22,23,24].

In this study, to elucidate the underlying relationship between the PM, foliar microstructures, and phyllosphere microbial diversity, we used ten asexual lines (to ensure a consistent genetic background) of eight distinct species, one variety, and one cultivar of Ulmus as study materials. Sanning electron microscopy (SEM) and atomic force microscopy (AFM) were applied to characterize the foliar microstructures. Using the laser particle size, we further characterized the particle size distribution of the PM retained on the foliar surface. Furthermore, we used 16 S rRNA gene and ITS high-throughput amplicon sequencing to analyze the phyllosphere microbial diversity among the lines (consist of bacteria and fungi). Based on the unique foliar microstructures of plants and their known role in retaining PM, we hypothesize that specific variations in leaf microstructures would influence the phyllosphere microbial community by providing distinct microhabitats for PM-borne microorganisms. To address this hypothesis, our study was designed to investigate: (1) how foliar surface microstructures influence PM retention capacity, and (2) how variations in PM retention shape the diversity and composition of phyllosphere microorganisms across the ten Ulmus lines.

Materials and methods

Study site and plant materials

The study site (38.1517°N, 114.4843°E) was a nursery at Hebei Academy of Forestry and Grassland Science located in ** and as shelterbelts that can effectively retain atmospheric PM, although this ability varies among species [70]. The TSP index reflects the total amount of PM retained on the leaf surface, which can be used as an indicator to evaluate the PM retention capacity of plant species [71]. In this study, the average TSP values of the ten Ulmus lines (135.76 µg·cm− 2; Fig. 1) were much higher (by 59.62–231.44%) than most common woody plants under similar climatic conditions, including Buxus megistophylla, Fraxinus pennsylvanica, and Sophora japonica [81]. Notably, we also found that, unlike the bacterial communities, certain fungal functional classes, particularly “Symbiotroph”, varied among the ten lines, with significantly higher relative abundance observed in **ye. This may be attributed to the ecological traits of the “Symbiotroph” class of fungi, potentially establishing a mutually beneficial symbiotic relationship with the plant [82]. For instance, similar to endophytes [83], phyllosphere microorganisms might provide the plant with antimicrobial substances or other beneficial compounds for defending pathogens, and in return, the plant provides the necessary resources for the fungi to thrive [84]. This may also indirectly reflect the potential higher tolerance of **ye to biotic stress than the other lines, but it needs to be further verified.

Atmospheric PM conditions significantly impact the phyllosphere microbial community of many species [19, 85, 86]. Based on the RDA analysis (Fig. 5a and b), our results suggest that fungal microorganisms could be carried by the PM2.5 attached to the dense leaf trichomes. In contrast, bacteria were not easily carried by PM and were also difficult to capture by or attach to the leaf microstructures of the Ulmus lines. Based on a correlation profile (Fig. 5c and d), we found that several microorganisms (consisting of both bacteria and fungi) were correlated (p < 0.05) with PM factors and/or leaf microstructures, suggesting that the microstructures of the leaf surfaces of the ten Ulmus lines enable them to capture PM-borne microorganisms. For example, PM2.5 carried microorganisms, such as Trichoderma and Aspergillus; the microstructures of the leaf surface, especially the dense and long trichomes, provided an ideal microhabitat for the PM2.5-microorganism complexes (Fig. 5d). The two positive feedback loops (i.e., PM2.5 -abaxial trichome lengh-Aspergillus and PM2.5-abaxial trichome density- Trichoderma) within the correlation network profile also provides evidence to support this (Fig. 6). In contrast, within the profile, we also found two negative feedback loops (PM2.5−10/PM> 100-adaxial RMS values-Cladosporium), suggesting that PM in these size ranges is difficult to be retained by the microrough leaf adaxial surface, or even if captured, it is unable to promote/even inhibit the proliferation of specific microorganisms (Fig. 6). We further assessed the significance of the aforementioned four feedback loops through multiple regression analysis and confirmed their validity (p < 0.05) (Table S9). Overall, our data prove that plant foliar microstructures can create an ideal microhabitat for PM-borne microorganisms. On the other hand, whether and how these PM-borne microorganisms subsequently affect plant growth and development remains to be further explored.

Conclusions

We demonstrate that the ten Ulmus lines investigated exhibited considerable PM retention capacities, with a mean TSP value of 135.76 µg·cm− 2. Variations in leaf surface microstructures, particularly the length and density of the trichome and the surface roughness, were the primary determinants of differential PM retention capacities among the lines. Long and dense trichomes substantially contributed to the retention of PM2.5, while larger undulations on the leaf surface enhanced the capture of PM10. Notably, these leaf microstructures provided ideal microhabitats for retaining PM-borne microorganisms, as evidenced by positive feedback loops between PM2.5, trichome characteristics, and the relative abundances of phyllosphere fungi like Trichoderma and Aspergillus. In contrast, bacterial communities were less impacted by PM. Our findings establish a three-factor network profile linking PM, leaf microstructures, and phyllosphere microbial communities, providing insights for further exploration into how different plants retain PM through foliar microstructures, thereby influencing their associated microbiomes.