Best fits were derived by incrementally increasing the number of fitted components and minimizing the fitted residual. The range for the fit was varied as a function of data quality and to test contributions from minor components. Target transforms were carried out using the SIXpack program, and XANES data analysis was carried out using the Athena program suite according to standard methods .The samples were fixed with 2.5% glutaraldehyde in phosphate buffer for more than 4 h, after which they were rinsed three times for 15 min in phosphate buffer . The samples were post fixed in 1% OsO4 in phosphate buffer for 1–2 h and rinsed three times in phosphate buffer for 15 min. The samples were dehydrated by a graded series of ethanol for ~15 min at each concentration and then dehydrated twice in alcohol for 20 min at each step. The dehydrated samples were coated with gold-palladium in a Hitachi Model E-1010 ion sputter for 4–5 min and subsequently observed via a Hitachi Model SU-8010 scanning electron microscope.To assess the overall leaf surface characteristics that affect the penetration barrier, the physiological properties of the adaxial and abaxial sides of apple leaves were determined . Surfaces of mature leaves of apple plants were observed using scanning electron microscopy , which revealed that mature apple leaves have a heterogeneous topography between their adaxial and abaxial surfaces, and the adaxial/abaxial trichome structure and morphology were similar to those described in previous studies of apple leaves. There was a smooth epicuticular wax layer covering the adaxial leaf surface of mature apple leaves; by contrast,hydroponic nft gully the abaxial leaf side was covered by nonglandular trichomes. In addition, the adaxial side was completely astomatous .

To assess the differences in Zn absorption between the adaxial and abaxial leaf surfaces, micro-XRF was used to evaluated the leaf cross-sections after foliar Zn application. The X-ray fluorescence intensity was normalized and visualized with a heat map that corresponded to differences in Zn concentration. Our results indicated a significant difference in the rate of Zn penetration between the adaxial and abaxial leaf surfaces . After continuous foliar Zn application for 2 weeks, very low signals for Zn were observed within the adaxial sides of treated leaves, while a greater distribution of Zn was observed in the abaxial sides of treated leaves.We conducted high-resolution imaging to further investigate the in vivo Zn distribution within the subcellular compartments of apple leaves after foliar Zn treatment. Foliar Zn penetration of the adaxial and abaxialleaf surfaces was visualized with merged images that showed the relative locations of Zn and K . Treatment with Zn fertilizer for 3 d led to a significant increase in Zn levels detected around the upper and lower epidermal cells, with deposition in a narrow band that corresponded to the cuticle layer and epidermal cell walls. The Zn intensities were also determined along the scan lines that passed through the epidermal cells. It is noteworthy that the detected Zn signals peaked in the cell walls and decreased sharply in the palisade parenchyma . This suggested that Zn was abundantly sequestered within the cell walls after penetrating the surface barriers. The Zn signals in the lower epidermal cell walls were ~3-fold higher than those in the upper epidermal cell walls.We further evaluated the physiological properties of leaf surfaces that influenced the adsorption and penetration of Zn after foliar application. We studied the effects of stomatal aperture and trichome density on Zn penetration at the abaxial leaf surface. Abscisic acid and darkness treatment were used prior to foliar Zn application each time to control the stomatal aperture. Methyl jasmonate treatment and mechanical removal of trichomes were used to modify lower epidermal structures.

MeJA is a phytohormone involved in plant defense systems, and it can alter the properties of plant leaf surfaces, such as trichome density. We began MeJA application at the start of new growth of young leaves until they had fully expanded. Furthermore, we examined the changes in cuticle thickness and composition, trichome density and stomatal density to assess the effects of MeJA on leaf properties. The SEM images showed that pretreatment with MeJA increased the trichome density of the abaxial leaf surface compared with that of the controls .No differences were observed in the change in stomatal density between the MeJA and control treatments . FTIR and TEM analyses were also used to investigate alterations in cuticle composition and thickness, but no significant differences were observed after MeJA treatment . Therefore, we focused on the role of trichome density in the process of foliar Zn penetration using apple leaves pretreated with MeJA and leaves with their trichomes removed.The micro-XRF results clearly showed that foliar Zn application led to a substantial increase in Zn levels in underlying leaf tissues, indicating that foliar-applied Zn penetration and absorption through the leaf surface occurred in all the treatments, and treatment-dependent effects were observed . Pretreatment with ABA and darkness to induce stomatal closure resulted in a relatively low penetration of Zn compared with that in the other treatments, suggesting foliar-applied Zn passed into the leaf interior through open stomata. Foliar Zn absorption was also related to trichome density; specifically, the penetration of Zn fertilizer increased as the trichome density decreased. Compared with MeJA treated leaves with relatively more trichomes, leaves that had their trichomes removed had more Zn across the leaf surface and retention within the treated leaves, revealing a significant reduction in foliar Zn penetration. The in vivo scanning maps for Ca, Mn, Fe, P, and S in the cross sections of leaves with different physiological surface properties after foliar Zn application are shown in Fig. S4.

The variation in elemental intensities and the correlations between different elements were further examined to reveal the possible impact of foliar Zn application on the nutrition status of treated leaves. Although it appears that the distribution patterns of other mineral elements shown in Fig. S4 were not obviously different between the treatments, the quantitation of their fluorescence yields helped to highlight the variation of individual elements after foliar Zn application. As shown in Fig. 4, the elements responded differently along with the increased Zn signals in the leaves. There is an induced effect on the tested elements after foliar application of Zn; among them,aeroponic tower garden system significantly induced enrichment of Ca and P was observed in leaves with low trichome density , in which Zn reached its highest levels. The concentrations of Mn and Fe changed depending on the Zn content, and low levels of Zn within leaves slightly promoted the accumulation of Mn and Fe, whereas high Zn concentrations in leaves had the opposite effect. The foliar application of Zn had no significant effect on the S concentration in the treated leaves . The spatial correlations between Zn and other elements were further examined. There were no significant correlations between Zn and the tested elements , in the distribution patterns. The Pearson correlation coefficients and their associated R values between Zn and P are presented in Fig. 5a–c. Our results revealed that there was an increased positive spatial correlation of Zn and P along with the penetration of foliar Zn fertilizer, suggesting a spatial overlap of Zn and P within the foliar treated leaves, and Zn was probably combined with P when the leaves were exposed to high exogenous Zn levels. We therefore performed XANES analysis on powdered frozen hydrated samples to acquire overall information about the speciation of Zn in leaves. By the use of linear combination fitting , Zn was found to be present as Zn-polygalacturonic acid, Zn-phytic acid, Zn-cysteine, and ZnSO4 after foliar Zn application . It was predicted that the majority of Zn was coordinated by Znphytic acid, with 43.6 and 52.7% for leaves after foliar Zn treatment of adaxial and abaxial leaf surfaces, respectively. This demonstrated that phytic acid may play a role in the complexation and stabilization of foliar-absorbed Zn. Znpolygalacturonic acid was the second most abundant Zn species. Polygalacturonic acid is a major component of pectin, which is most abundant in plant primary cell walls. The XANES results confirmed the sequestration of Zn within cell walls during the penetration process across the leaf surface.Currently, there is little information on the foliar penetration pathways of ionic solutes. The actual contribution of leaf surface structures to foliar-applied fertilizer absorption remains unclear. A common problem of exploring the mechanisms of foliar penetration is the technical limitations related to fluorescence and optical microscopy and observing the penetration process of nutrients through the leaf surface. In the present study, the direct visualization and high spatial resolution of XRF was suggested to be a promising and powerful strategy to investigate the distribution pattern of Zn within the plants following its application. By utilizing this technique, we were able to investigate the process of penetration of foliar-applied Zn across apple plant leaves with different physiological surface properties as well as help shed light on the possible interactions between foliar Zn concentration and the mineral nutrition status of treated leaves.

Our results showed a much higher efficiency in Zn penetration by abaxial surfaces than adaxial surfaces ofmature apple leaves . The greater nutrient absorption by the abaxial leaf surface may have resulted from the different physiological properties of the two leaf sides, as drop adherence to the leaf surface is a precondition for foliar penetration to occur. The high degree of physical variation between the abaxial and adaxial leaf surfaces of apple leaves contributed to the interactions between the liquid fertilizer and the different leaf surfaces. The adaxial surface of apple leaves is covered with hydrophobic epicuticular wax and does not have stomata, which limits the bidirectional exchange of water, solutes and gases between the plant and the environment. Based on micro-XRF imaging, Du et al. also reported that foliar penetration of Zn was much more difficult through waxy leaves of citrus, though the concentrations of Zn increased significantly in the leaf tissues directly beneath the applied sites. Very restricted nutrient solution absorption across the adaxial leaf surface also occurs in other fruit species, such as pear, lychee, and grapevine. The extent of Zn mobilization following absorption also influences the efficacy of Zn penetration. In previous research, XRF was also performed to compare the behavior of different forms of foliar-applied Zn fertilizers, and the results indicated that Zn fertilizer applied as ZnSO4 was slightly more mobile than were other Zn forms, such as Zn-EDTA and ZnO.Further, we attempt to observe the distribution patterns of Zn at the subcellular level by using nano-XRF imaging with sufficient resolution and sensitivity. For this technique, sample preparation is very important. Here, a preparation procedure involving high-pressure freezing followed by freeze substitution was used. This preparation protocol was proven in previous studies to preserve the ultra structure of samples as much as possible, which can better preserve the in vivo localization of mobile elements. This phenomenon was also confirmed in the present study, as the highly diffusible element, K , was uniformly distributed in the palisade tissue and spongy tissue, which indicates that redistribution of the element did not significantly occur during sample preparation. Recently, the cuticle was considered a lipidized epidermal cell wall region. In the present study, the subcellular distribution of Zn occurred within a narrow band that corresponded to the cuticle layer and epidermal cells, which is consistent with the interpretation that the cuticle is the extension of the cell wall. In addition, this result provided direct visual evidence that there was a strong capability for Zn fixation in the cell wall following Zn application, which limited foliar Zn penetration and led to the low effectiveness of foliar Zn penetration.This was similar to the results of previous studies showing that Zn2+ has limited mobility in wheat leaves regardless of the form in which it is applied, which suggests a poor leaf penetration and high binding capacity of Zn to leaf tissues. In contrast, other studies have shown that Zn moves across the cuticle of soybean and tomato leaves without binding to epidermal cells, which may be due to the different plant materials used in the experiments and resulted in different interactions between the Zn solution and leaf surfaces.