Will the climate of plant origins influence the chemical profiles of cuticular waxes on leaves of Leymus chinensis in a common garden experiment?

Abstract Cuticular wax covering the leaf surface plays important roles in protecting plants from biotic and abiotic stresses. Understanding the way in which plant leaf cuticles reflect their growing environment could give an insight into plant resilience to future climate change. Here, we analyzed the variations of cuticular waxes among 59 populations of Leymus chinensis in a common garden experiment, aiming to verify how environmental conditions influence the chemical profiles of cuticular waxes. In total, eight cuticular wax classes were identified, including fatty acids, aldehydes, primary alcohols, alkanes, secondary alcohols, ketones, β‐diketones, and alkylresorcinols, with β‐diketones the predominant compounds in all populations (averaged 67.36% across all populations). Great intraspecific trait variations (ITV) were observed for total wax coverage, wax compositions, and the relative abundance of homologues within each wax class. Cluster analysis based on wax characteristics could separate 59 populations into different clades. However, the populations could not be separated according to their original longitudes, latitudes, annual temperature, or annual precipitation. Redundancy analysis showed that latitude, arid index, and the precipitation from June to August were the most important parameters contributing to the variations of the amount of total wax coverage and wax composition and the relative abundance of wax classes. Pearson's correlation analysis further indicated that the relative abundance of wax classes, homologues in each wax class, and even isomers of certain compound differed in their responses to environmental factors. These results suggested that wax deposition patterns of L. chinensis populations formed during adaptations to their long‐term growing environments could inherit in their progenies and exhibit such inheritance even these progenies were exported to new environments.


| INTRODUC TI ON
Plants are sessile organisms, which have evolved the abilities to capture and utilize resources under changing environments.
During this evolution process, plastic variations of functional traits enable the plants to survive through different growth conditions and to exhibit adaptive differentiation of plant populations in response to differing climates (van Kleunen & Fischer, 2005). The cuticle, covering leaf surface, plays important roles in protecting plants from biotic and abiotic stresses (Yeats & Rose, 2013).
Understanding the way in which plant leaf cuticles reflect their environment could give an insight into plant resilience to future climate change.
Cuticular waxes are mixtures of hydrophobic compounds, which include the epicuticular wax covering on the cutin surface and the intracuticular wax embedded within the cutin matrix (Jeffree, 2006). Though the wax amount and composition vary between plant species, different organs, growing stages, and even growing conditions, most plants share similar wax biosynthesis pathways (Samuels, Kunst, & Jetter, 2008). The main wax chemicals identified in plants are long-chain fatty acids and their derivatives such as aldehydes, primary alcohols, alkanes, alkyl esters, ketones, secondary alcohols, and triterpenoids (Jetter, Kunst, & Samuels, 2006). Studies have shown that plants will alter their wax deposition and compositions under changing environments to improve their adaptations (Hatterman-Valenti, Pitty, & Owen, 2011;Mackova et al., 2013;Shepherd & Griffiths, 2006). For example, greater weighed mean alkane chain length at low and at high elevations was possibly a result of adaptation to minimize cuticular permeability due to high summer temperature at low elevation and freezing causing physiological drought at high elevations (Dodd & Poveda, 2003). An increase of alkane deposition has been observed in many drought stressed plants (Gefen, Talal, Brendzel, Dror, & Fishman, 2015). Dodd, Rafii, and Power (1998) also reported that annual rainfall was the most significant factor in regressions with the shorter-chain hydrocarbons whereas annual mean temperature was most significant for the longer-chain hydrocarbons. These reports implied that plant populations growing under different environments might exhibit intraspecific variations of wax profiles.
Plant species can display high intraspecific trait variation (ITV) under different growing conditions, reflecting heritable genetic variation and phenotypic plasticity (Henn et al., 2018). The variations between populations may be attributed to the differences in the selective pressures imposed by different ecological environments (Still, Kim, & Aoyama, 2005). Understanding the patterns of functional trait variation across environmental gradients offers an opportunity to increase inference in the mechanistic causes of plant community assembly (Fajardo & Siefert, 2018). Based on an analysis from more than 2,500 vascular plant species, Wright et al. (2005) found that 18% of variation in their study of leaf traits was explained by climate. Climate influences trait variation in part by selection for different life forms and families, and the trait values derived from climate data via redundancy analysis (RDA) showed substantial predictive power for trait values in the available global datasets (Yang et al., 2019). In a climate gradient in the east side of Qinghai-Tibetan Plateau,  found that the mean annual temperature and aridity index were significantly correlated with the averaged amounts of wax compositions and total leaf cuticular wax. Many studies have also validated that the intraspecific variation of leaf cuticular wax could be applied in distinguishing plant populations growing under different environments (Bojovic et al., 2012;Halinski, Szafranek, & Stepnowski, 2011). Therefore, understanding ITV of leaf cuticular wax across regional scales can illustrate the potential adaptive capacity of plant populations to local conditions. In addition to the large amount of information on the wax compound identifications, the relationship between wax and abiotic stresses, and gene cloning involved in wax biosynthesis, little is known in revealing genetic differentiation of plant leaf cuticular waxes from different regions and populations. In order to seek evidence for an adaptive response in wax composition and quantity across environmental and geographic gradients, Ramirez-Herrera, Percy, Loo, Yeates, and Vargas-Hernandez (2011) found that strong differentiation among regions and populations within regions was observed for wax quantity from 12 Pinus pinceana populations in a common garden test. Based on a study of Melaleuca quinquenervia across eastern Australia, Andrae et al. (2019) found that n-alkane characteristics were not a plastic response to climate variability and instead were likely fixed and could be driven by genetic differences between populations. Using quantitative genetic and quantitative trait loci analyses, Gosney et al. (2016) (Yuan, Ma, Guo, & Wang, 2017). In China, it grows across diverse soil and climate conditions such as the Songliao Plain, the Inner Mongolia Plateau, and the Loess Plateau, contributing to its wide genetic diversity (Liu, Li, Li, Yang, & Liu, 2007). In order to determine whether cuticular wax in L. chinensis exhibits similar or different patterns of local adaptation across their distributions, in this study, we selected 59 populations distributed in different environments, aiming to verify how environmental conditions could influence the chemical profiles of cuticular waxes in L. chinensis in a common garden experiment. In addition, the relationship between meteorological variables and the qualitative and quantitative wax compositions was explored in order to detect meteorological factors affecting wax profiles. We hypothesized that intraspecific variability in leaf wax production and chain length distributions of L. chinensis might be genetic differentiation of plant leaf cuticular waxes.

| Site description
We established a common garden experiment design, located in

| Sampling
In July 2016, when plants were in their heading stages, leaves (the third leaf from the top) were sampled from each plant population separately. There were no strict biological replicates for each population, thus leaves from 10 plants were mixed and regarded as one replicate, in total three replicates for each population. To avoid the difference of wax deposition in different plant development stages, we sampled plants only in heading stages. The leaves were washed gently in water to exclude dusts on leaf surface and stored in absorbent papers. The absorbent papers were changed every other day until the leaves were fully dried without going moldy.

| Cuticular wax extraction
Due to the difficulties in extracting the wax from fresh leaves in fields, dried leaves were used to extract cuticular waxes. In a pre-experiment study, the dried leaves were extracted in chloroform for 30s, 50s, 60s, and 80s. The results showed that 60s was sufficient to extract most of the cuticular waxes on plant leaves. Therefore, in this study, leaves were extracted in 50 ml chloroform containing 25 µg tetracosane as internal standard at room temperature for 1 min. The extracts were dried in a nitrogen stream at 40°C, and derivated using 50 µl of BSTFA (N,O-Bis(trimethylsilyl) Trifluoro Acetamide) and 50 µl pyridine (Aldrich) for 45 min at 70°C. The surplus solutions were evaporated under nitrogen, and the sample was redissolved in 200 µl chloroform for GC and GC/ MS analysis.
For compound identification, the GC analysis was carried out with 9790Ⅱ gas chromatograph (Fu-Li, China). The GC column was DM-5 capillary column (30 m × 0.32 mm × 0.25 µm). Nitrogen was served as the carrier gas. The GC oven was held at 80°C for 10 min and heated at 5°C/min to 260°C, where the temperature remained 10 min. The temperature was then heated at 2°C/min to 290°C and further heated at 5°C/min to 320°C, where the temperature was held for 10 min. Compounds were detected with a GCMS-QP2010 Ultra Mass Spectrometric Detector (Shimadzu Corp.) using HP-5 MS capillary column (30 m × 0.32 mm × 0.25 µm) and He as the carrier gas. Compounds were identified by comparing their mass spectra with published data and authentic standards.
Quantification was based on FID peak areas. After wax extraction, the surface areas of leaves were measured with a WinFOLIA professional leaf image analysis system (Regent Instrument, Inc.) and digitizing scanner (EPSON V750, Japan). The amount of wax was expressed in µg/cm 2 .

| Data analysis
In order to evaluate whether properties of the site of origin of an accession affected its fitness in the novel environment of Shaerqin, historical climate data for the collection sites of the accessions were obtained from National Meteorological Information Center of China (http://data.cma.cn/). Latitude and longitude for all of the accessions were obtained from https ://www.latlo ng.net/. For each site, the weather information from the grid point closet to the historical site of origin of the accession was used. An aridity index was calculated as I = P/(T + 10), where P is the annual precipitation in mm and T is the mean annual temperature in degrees centigrade (Dodd & Poveda, 2003). In order to evaluate whether chemical compositions varied between populations, one-way ANOVA was applied on total wax coverage (SPSS 18.0). Hierarchical cluster analysis was applied based on amounts of total wax coverage and wax compositions, the relative abundance of wax compositions, and the relative abundance of compound homologues within each wax class, according to furthest neighbor (SPSS 18.0). Redundancy analysis was further undertaken to visualize the relationship between climate factors and the amounts of total wax coverage and wax compositions, the relative abundance of wax compositions, and the relative abundance of compound homologues within each wax class, using R software (3.6.1) and the vegan and ggplot2 package.
The number of populations that acids had not been detected was 10, secondary alcohol was 7, ketone was 6, and alkylresorcinol was 50 ( Figure 2). Diketone was the dominant compound class in L. chinensis, accounting for 21.36% to 84.64% in total wax, followed by primary alcohol (averaged 10.63% across all population), aldehydes (5.22%), and alkanes (4.53%). The total wax ranged from 5.55 µg/ cm 2 to 40.14 µg/cm 2 , with coefficient of variance reaching 46.82%.
ANOVA analysis further indicated that the amount and the relative abundance of wax compositions varied greatly between populations in a common garden experiment, showing high ITV (Table 3).
Each compound class was consisted of series of compounds with varying chain lengths (Table 4)

| Relationship among populations
UPGMA cluster analysis, based on wax amounts, relative abundance of wax compositions and the relative abundance of homologues in each wax class, separated 59 populations into two clades F I G U R E 1 Geographic distribution of collection sites for the Leymus chinensis populations used in this study. Latitude and longitude data are from https ://www.latlo ng.net/. Each number represented one sampling site  showed that latitude, arid index, and the precipitation from June to August, which explained 8.05%-12.18% of variance (p = .005), were the most important parameters contributing to the amount of total wax coverage and wax composition and the relative abundance of wax class (Table 5).

| Relationship between wax amount and the climate factors
Pearson's correlation analysis was applied to further test the influence of climate factors on wax properties (Table 6)

| D ISCUSS I ON
A number of studies have demonstrated the responses of cuticular waxes to changing environments on many plant species. For example, drought stress would increase wax deposition (particularly alkanes) on leaf and thus reduce water loss through leaf cuticle (Gonzalez & Ayerbe, 2010), and enhanced UV-B irradiation would alter the crystal structures of epicuticular wax and thus increase the reflection of irradiation from leaf (Gordon, Percy, & Riding, 1998). Shepherd, Robertson, Griffiths, Birch, and Duncan (1995) also reported that epicuticular waxes from outdoor-grown plants were found to have higher proportions of n-alkanes, octacosanoic acid, primary alcohols, and long-chain esters but lower proportions of aldehydes, ketones, ketols, and secondary alcohols than waxes from indoor-grown plants. Such responses of plant cuticular waxes have been shown to be related to plant adaptations to changing environments (Yeats & Rose, 2013).
In this study, high ITV was observed among populations of   Besides the climate factors as we obtained, soil water conditions as well as UV-B irradiation in their original places might also influence the wax deposition pattern (Gordon et al., 1998;Schwab et al., 2015), and thus the cluster results.
To further analyze how the climate factors influenced the wax characteristics, RDA analysis and Pearson's correlation analysis were applied. RDA showed that latitude, arid index, and the precipitation from June to August, which explained 8.05% to 12.18% of variance, were the most important parameters contributing to the amount of total wax coverage and wax composition and the relative abundance of wax class. Positive or negative correlation between wax characteristics and longitudes and latitudes was also observed in Pearson's correlation analysis. These results suggested that the current characteristics of the cuticular wax were the responses to comprehensive environmental factors (Jeffree, 2006), which might partly could explain the comprehensive cluster results based on cuticular wax among different populations. This further implied that current leaf wax characteristics partly could reflect paleo-environmental conditions during which period the plant population formed and grew. For example, alkane C-23 as a robust proxy for Sphagnum mosses was used in paleoecological studies (Bush & McInerney, 2013), and aeolian-derived higher-plant lipids in the marine sedimentary record were linked with paleoclimate (Poynter, Farnimond, Robinson, & Eglinton, 1989).
In this study, the relative abundance of wax classes differed in their responses to climate factors. For example, annual temperature was negatively correlated with the relative abundance of primary alcohols and alkanes but positively correlated with β-diketone, whereas annual precipitation was positively correlated with the relatively abundance of secondary alcohol but negatively correlated with primary alcohols. This implied that historical growing environments influenced the wax biosynthesis pathways in plant populations, resulting in the changes of wax proportions, which might further contribute to plant adaptations (Shepherd, Robertson, Griffiths, & Birch, 1997;Yeats & Rose, 2013). Meanwhile, the relative abundance of wax homologues in each wax class and even the isomers of wax compound also differed in their responses to climate factors. For example, the annual precipitation was positively correlated with the relative abundance of C 26 and C 30 primary alcohol but negatively correlated with C 28 primary alcohol, and negatively correlated with C 31 β-diketone but positively correlated with hydroxylated C 31 β-diketone. This further implied that the homologue genes involved in wax biosynthesis might also differ in their sensitivity to climate factors, contributing to increased adaptability to diverse environments during plant evolution and domestication process (Zou et al., 2012).
In conclusion, the great ITV of wax characteristics among 59 populations of L. chinensis in a common garden experiment indicated that plant populations growing under certain environment might inherit their specific leaf wax deposition patterns to progenies. Such F I G U R E 4 Redundancy analysis (RDA) on data of wax and environmental factors based on amounts of total wax coverage and wax compositions (a), the relative abundance of wax compositions (b), and the relative abundance of compound homologues within each wax class (c). Tannual, annual average temperature; T 6 , T 7 , and T 8 represented average temperature in June, July and August; Pannual, annual precipitation; P 678 , rainfall from June to August TA B L E 5 The rates of total variance of wax characteristics explained by the environmental factors from redundancy analysis in Figure 4 Longitude Note: A, based on amounts of total wax coverage and wax compositions; B, the relative abundance of wax compositions; C, the relative abundance of compound homologues within each wax class. T a , annual average temperature; T 6 , T 7 and T 8 represented average temperature in June, July and August; P a , annual precipitation; P 678 , rainfall from June to August.

TA B L E 6
Spearman's correlation coefficient between climate factors and the relative abundance of wax compositions and the relative abundance of homologue in each wax class Note: T a , annual average temperature; T 6 , T 7 and T 8 represented average temperature in June, July and August; P a , annual precipitation; P 678 , rainfall from June to August. * indicates p < .05.
trait inheritance includes total wax coverage, wax compositions, and the chain length distribution patterns. RDA analysis showed that latitude, arid index, and the precipitation from June to August were the most important parameters contributing to the variations of the amount of total wax coverage and wax composition and the relative abundance of wax class. Pearson's correlation analysis further indicated that the relative abundance of wax class, homologues in each wax class, and even isomers of certain compound differed in their responses to environmental factors, suggesting that genes involved in wax biosynthesis showed heterogeneous evolution process in different environments, which contributes to the plant adaptations to growing environments.

ACK N OWLED G EM ENTS
This research was funded by National Natural Science Foundation of China (31670407, 31771694) and the National Key Basic Research Program of China (2014CB138806).

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

AUTH O R CO NTR I B UTI O N S
LY, LX, ZX, and XY collected the samples and analyzed the GC data and climate data; WZ managed the common garden experiment; HX and GY designed the experiment; GY identified the wax compounds.
All authors contributed to the drafts of the manuscript and its final approval.

DATA AVA I L A B I L I T Y S TAT E M E N T
We agree to make our data publicly available in a relevant repository (DRYAD, https ://doi.org/10.5061/dryad.v15dv 41s5) when the manuscript was accepted, including climate factors, wax amount, and wax class abundance.