Plant growth, physiological variation and homological relationship of Cyclocarya species in ex situ conservation

It is necessary to conserve Cyclocarya species, as it is a valuable, medicinal woody species. Cyclocarya paliurus responded to varying environmental factors via physiological changes and exhibited variation in growth at a resource plantation of ex situ conservation. The homological relationships revealed low genetic differentiation with only two major sub-populations.


Introduction
Forests are a major natural resource, but with the rapid growth of human population and industrialization, massive forest destruction has occurred that is much beyond regeneration, mainly because of over-exploitation, overgrazing, unsustainable practices, forest fires and environmentally unfriendly development projects (Babu and Nautiyal, 2015). Therefore, conserving forest resources, especially high-value species and those with small and vulnerable populations, is pertinent for genetic resource conservation (Holliday et al., 2017;Ratnam et al., 2014). Cyclocarya paliurus (Batal.) Iljinskaja (C. paliurus) is a native and high-value species distributed in the highlands of sub-tropical areas in China (Fang et al., 2006) that possesses a myriad of human health benefits, such as anticancer, antimicrobial, antihyperlipidemic, antioxidant and anti-inflammatory effects, which is primarily the result of the biological activities of various phytochemicals in their leaves (Xie et al., 2012(Xie et al., , 2013Wang et al., 2017;Liu et al., 2018a,b;Shang et al., 2018;Xiong et al., 2018). However, C. paliurus regenerates slowly in natural forests because of their high seed dormancy under natural conditions (Fang et al., 2006); further, the populations of C. paliurus has been subjected to severe damage due to the increasing medicinal use of leaves in recent years. To date, C. paliurus has been protected via different conservation statuses, including critically endangered, server convention and convention (http:// www.iplant.cn). Therefore, it is highly pertinent to establish an effective way to conserve C. paliurus populations.
Ex situ conservation is an effective way to preserve plant species in order to rescue or maintain the natural plant biodiversity (Corlett, 2016;Seaton et al., 2010). Seed banks and other biotechnological technologies, such as in vitro culturing, are unsuitable for conserving C. paliurus populations, because of their high rate of 'empty seed' and limitations of the in vitro regeneration system (Fang et al., 2006;Feng et al., 2020a). Thus, ex situ conservation could be a suitable method to conserve C. paliurus populations.
During the process of ex situ conservation, environmental factors affect plant growth and adaptation (Enßlin et al., 2011;Cao et al., 2018). Specifically, studies have indicated that temperature accelerates Larix chinensis or Myrsine seguinii growth (Liu et al., 2018a;Wu et al., 2019), but inhibits olive growth (Benlloch-González et al., 2016). Plants respond to changing environmental factors by change in water content (WC) or soluble sugar content (Ben Abdallah et al., 2017;Feng et al., 2020b;Wu et al., 2018), by regulating mineral element concentrations (Wu et al., 2019) or by increasing antioxidant enzyme activities (Habibi, 2017;Zhao et al., 2018). Wu et al. (2018) found that a change in leaf K concentrations can affect leaf water potential in response to warming. However, plants respond to environmental changes in a species-specific manner (Wu et al., 2019). In addition, as a characteristic of physiological response, the accumulation of secondary metabolites is also affected by environmental factors. For example, light and fertilization influence the growth and total flavonoid accumulation of C. paliurus (Deng et al., 2012;Yang et al., 2017;Liu et al., 2018a,b). Physiological responses can ultimately lead to differences in growth and adaptation among various plant species, thereby determining whether the establishment of ex situ construction is successful and valuable. Thus, understanding the physiological responses to environmental factors is critical for the successful construction of a resource plantation.
In addition, during long-term natural evolution processes, genetic differentiation of C. paliurus has occurred in natural forests via natural or human selection, including genetic drift, climate change and seed dispersal. Therefore, the relationship among Cyclocarya populations must be strictly defined in the ex situ conservation. Simple sequence repeat (SSR) has the advantage of being abundant and low-copy among the transcribed fractions of plant genomes (Uncu and Uncu, 2020) and thus has been extensively applied to analyse homological relationships in several plants, including sweet cherries Patzak et al., 2020), peaches (Dettori et al., 2015), lemons (Zhu et al., 2016) and hazelnuts (Bhattarai and Mehlenbacher, 2017). Suvi et al. (2020) investigated the homological relationship and genotypic structure of 54 rice accessions using SSR to select unique parents for breeding. Li et al. (2017) used SSR to analyse the homological relationship of C. paliurus populations from 26 provenances of 11 provinces in China. However, the homological relationship among C. paliurus populations from some province is still unknown, such as the population from Fujian Province. Hence, the definition of the homological relationship of C. paliurus populations will further our understanding and provide essential information to enable more efficient use of available genetic resources (Mohammadi and Prasanna, 2003).

Conservation Physiology • Volume 10 2022
Research article  Table S1. In total, 50 plants of each provenances were planted in each block with a spacing of 2 × 2 m. Artificial weeding was performed in May and September every year, but no supplemental irrigation was given for plant growth. Climatic factor data during the experiment period at the experimental site were collected from a local weather bureau, and all indices were summarized in Table 1. The soil at the experimental site was sandy loam soil containing 5.64 ± 0.06 (mg/kg) available nitrogen, 16.28 ± 0.06 (mg/kg) available phosphorous and 4.45 ± 0.05 (mg/kg) available potassium.
After 1 year ex situ conservation, leaf samples were collected from the middle part of the current-growth branch in April, June, August and October. The collected leaves were maintained at 4 • C and then immediately transported to the laboratory. Some leaves were dried for analysing physiological changes in WC, mineral concentration and secondary metabolite accumulation; other leaves were immediately frozen in liquid nitrogen and stored at −80 • C for analysing total soluble sugar (TTS) content and antioxidant enzyme activity. Meanwhile, leaves collected in August were analysed for homological relationship according to leaf morphological characteristics and SSR analysis.

Plant growth determination
In April, all plants were observed and the initial height (H i ) and initial basal diameter (BD i ) were measured. Then, the plant height (H a ) and basal diameter (BD a ) of all the surviving plants were measured again in December. Continuous measurements were conducted from 2017 to 2019. The growth index was calculated as follows: The height increase of plant: (NH) = H a -H i .
The basal diameter increase of plant: (NBD) = BD a -BD i .
The average increase of plant height: the sum of NH/the number of plants.
The average increase of basal diameter: the sum of NBD/the number of plants.

Determination of WC
According to the method described by Stein et al. (1975), WC in leaf was calculated as WC (%) = (FW − DW) \FW × 100, where FW is the weight of the fresh leaf, and DW is the constant weight of the dried leaf.

Extraction and determination of TTS content
TSS was extracted with distilled water from the fifth and sixth leaflets of the fresh compound leaf using the Anthrone-H 2 SO 4 method described by Li (2000), then the absorbance at 630 nm was measured with a UV-visible spectrophotometer, finally TSS content was calculated as follows: leaf TSS content (%) = (C × 25)/(W × 0.5 × 10 6 ) × 100, where C is obtained from the standard curve constructed with sugar and W is the weight of the fresh sample.

Extraction and determination of mineral nutrients
Samples were digested using the electric-heating digestion method described by Feng et al. (2020b). The content of mineral nutrients [potassium (K), sodium (Na), calcium (Ca), and magnesium (Mg)] were calculated using the following equation: (C × 0.025)/DW, where C is mineral content measured via Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES, Optima 7000DV, USA) and DW is the weight of dried sample.

Extraction and determination of antioxidant enzyme activity
Antioxidant enzyme extraction was obtained and each activity of four antioxidant enzymes [superoxide dismutase (SOD), polyphenol oxidase (PPO), peroxidase (POD), catalase (CAT)] was analysed in accordance with the method of Feng et al. (2020b).

Extraction and determination of secondary metabolite accumulation
Extraction was performed using an ultrasonic-assisted method with slight modifications (Liu et al., 2018c). Briefly, each sample (∼1.0 g) was added to 20 ml of 75% ethanol, centrifuged at 25 • C and 11 000 g for 15 min after heating at 70 • C for 60 min with an ultrasonic cleaner (KQ-800DE, China).
Total flavonoid content was determined using the method described by Liu et al., (2018c). A standard curve was constructed with rutin, and total flavonoid content was expressed as mg rutin equivalent/g dry mass.
Total triterpenoid content was assessed using the Folin-Ciocalteu colourimetric method and then expressed as mg gallic acid equivalent/g dry mass.
Polysaccharide content was determined using the method described by Liu et al., (2018a). A standard curve was constructed with glucose, and polysaccharide content was expressed as mg glucose equivalent/g dry mass.

Homological relationship definition Leaf morphological determination
Changes in leaf colour and leaf shape were observed, and total leaf area and the length of common petioles were measured for leaf samples from local provenance and five introduced provenances at resource plantation. Five compound leaves from the same plant were measured and treated as one leaf sample. And the leaf measurement of each provenance consisted of three replicates, with 10 leaf samples per replication.
Leaf cross sections were cut into ∼1 × 1 cm segments from the middle part of the fresh leaves for each provenance. The sections were fixed in 2.5% glutaraldehyde, soaked in osmic acid for 1 h and then dehydrated using a graded ethanol series (100,0, 75:25, 50:50, 25:75 and 0:100, v/v), for which samples were kept at each concentration for 15 min. The sections were critical-point dried in carbon dioxide using a critical point dryer (Leica EM CPD300, Germany), coated with goldpalladium using a vacuum coater (Leica EM ACE200, Germany) at 15 mA, viewed and photographed with a scanning electron microscope (SEM; FEI Quanta450, USA).

DNA extraction
DNA was extracted from fresh leaves of the six provenances using the improved Hexadecyl trimethyl ammonium Bromide (CTAB) method. Each sample was added to CTAB solution (100 mmol/l NaCl, 20 mmol/l Ethylene Diamine Tetraacetic Acid (EDTA) (pH 8.0), 2% CTAB (w/v) and 100 mmol/l Tris-HCl), heated at 65 • C for 30 min and centrifuged at 12000 g (TGL-16G, China) at 25 • C for 5 min. The supernatant was obtained and added to phenol-chloroform (1:1 v/v), then centrifuged at 12000 g at 25 • C for 10 min. Next, the supernatant was again obtained, added to chloroform and finally centrifuged at 12000 g at 25 • C for 10 min. This step was repeated twice. Finally, the supernatant was added to isopropanol, kept at room temperature for 15 min and centrifuged at 12000 g at 25 • C for 6 min. The deposit was cleaned with 75% ethanol and dissolved in 50 μl TE. Extracted DNA samples were stored at −20 • C.

SSR amplification
A total of 50 ng/μl DNA was used for polymerase chain reactions (PCRs) via a PCR analyser. Ten pairs of SSR primers were selected from Fan et al. (2013) and used for SSR amplification ( Table 2). The PCRs were conducted in a 20-μl reaction containing 2.0 μl 10 × Buffer, 0.1 mmol/l dNTPs, 0.3 μmol/l of each primer and 1.0 U Taq DNA polymerase. SSR amplification was performed using the following conditions: initial denaturation at 95 • C for 5 min, followed by 37 cycles at 94 • C for 30 s, an appropriate annealing temperature (from 65 • C to 55 • C using a graded decreasing temperature from the first to the tenth cycle, then cooling from 55 • C from cycles 11 to 37) for 35 s, 72 • C for 40 s and finally extension at 72 • C for 3 min. The SSR products were separated into a 6% denatured polyacrylamide gel and stained using the silver staining protocol.

Data and statistical analysis
One-way analysis of variance was used to test significant differences in plant growth from 2017 to 2019 and leaf physiological indices during the growth period by SPSS statistical software package (Version 16.0, IBM, USA) on the basis of Tukey's Highly Significant Differences at the significance level of P < 0.05. Pearson Correlation Analysis and trendsurface analysis by the performance of Origin software 9.1 (Northampton, MA01060, USA) was used to analyse plant growth and leaf physiological indices in correlation with environmental factors. Principal Component Analysis (PCA) and Homology Analysis were also performed by Origin software 9.1 to analyse the relation between different physiological indices.
An Excel original binary data matrix was constructed by calculating the presence-absence data of each amplified fragment, and these data were analysed using NTSys v.  number of polymorphic loci, percentage of polymorphic loci, Nei's gene diversity (He) and Shannon's information index (I) were also calculated using POPGen32. The generation of a clustering graph analysis among the six provenances was performed using NTSys v.2.10.

Results and analysis
Variable plant growth among the six Cyclocarya provenances Plant height and BD grew faster after 1-year ex situ conservation and significantly increased with the long-term conservation. Both of plant height and BD in 2019 reached significance as compared to that in 2017 and in 2018 (Table 3)

Water content
Leaves of FJ had the highest WC and reached significance, but those of JX showed the lowest WC in April, compared to that in other month. WC in leaf of FJ, AJ, TG, JX and WF showed similar changes from June to October, reaching minimum in August. In particular, the WC of JH decreased gradually after reaching the highest content in June, but no difference in leaf WC was observed among in June, August and October (Fig. 1).

TTS content
TSS content in leaf of FJ was more than that of JX in April. TSS content in leaf of FJ and JH increased firstly and then decreased from June to October and reached the highest value in August, but there was no difference in TSS content in leaf of JH among in June, August and October. The change of leaf TSS content of AJ was opposite to the behaviour of FJ and JH. Meanwhile, TSS content in leaf of JX and TG increased gradually, reaching significance in October, which was contrary to that of WF with the decreasing TSS content from June to October (Fig. 2).

Mineral nutrient content
K, Ca, Na and Mg were present in the leaves of all six provenances, in which the highest mineral content was Ca (≥6.0 mg/g), followed by K (≥3.0 mg/g) and Mg and Na (≤2.0 mg/g) (Fig. 3).   Ca content in leaf of FJ and JH decreased firstly and then increased, but leaf of JX showed an increasing content of Ca and a similar change was observed in that of WF, AJ, TG and JX. The change of Mg content in leaf of FJ, JX, WF and AJ decreased firstly and then increased, which was contrary to that of TG. Conversely, leaf of JH had the decreasing content of Mg. Leaf of FJ, JH, WF and AJ had similar change of Na content, while leaf of JX and TG had the increasing content of Na.

Antioxidant enzyme activity
The four antioxidant enzyme activities in leaf of the six provenances decreased in the following order: SOD > PPO > POD > CAT. However, the enzymes showed different changes in activity from April to October (Fig. 4).
In particular, SOD activity in leaf of the six provenances in April decreased in the following order: AJ > FJ > JX > WF > JH > TG. SOD activity in leaf of FJ decreased gradually from April to October, and similar changes were observed in leaf of JX and TG. Conversely, SOD activity in leaf of WF, AJ and JH decreased firstly, then increased, and finally decreased again.
PPO activity in leaf of FJ was 30.04 U/(g·min) and revealed a significant difference in June, while leaf of AJ exhibited an opposite behaviour in PPO activity with that of FJ. Both of JX and JH showed similar changes in PPO activity, peaking in August. Finally, PPO activity in leaf of TG and WF reached maximum in October and August, respectively. Different peaks of POD activity were observed in leaf of the six provenances. Specifically, both JX and FJ showed a PPO activity greater than 3.0 U/(g·min) in June, while TG and JH had the highest activity in April and October, respectively. There were similar change in PPO activity in leaf of FJ, JX, TG and JH, but leaf of AJ exhibited an opposite change in PPO activity with that of WF.  CAT activity in leaf of JX and FJ were less than 0.5 U/(g·min) from April to October. A similar change was found in leaf of TG and JH, although the activity in leaf of TG was higher than that of JH. Meanwhile, leaf of AJ and WF reached maximum CAT activity in August and June, respectively, and the peak was more than 2.0 U/(g·min).

Analysis of secondary metabolite accumulation
High contents of total flavonoid and polysaccharides (≥40 mg/g) were detected, but total tritenpenoid content was low (≤3.0 mg/g) in leaf of the six provenances. However, the accumulation of three kinds of secondary metabolite in leaf showed different changes from April to October (Fig. 5).
Total flavonoid content in leaf of FJ decreased firstly and then increased from April to October, which was similar to that of JX, WF and AJ. However, there was no significant difference in total flavonoid content observed in leaf of JH and TG from June to October.
Polysaccharide content in leaf of FJ reached 100.16 mg/g in April, and then decreased by 50.52%, 57.29% and 20.39% in June, August and October, respectively. The other five provenances had the highest content of polysaccharides in October, while both of WF and AJ had similar change in polysaccharides accumulation and similar change exit between that of JH and TG.
The highest content of total triterpenoid in August was observed in leaf of FJ, JX, JH and TG, all of which exhibited similar change, but were contrary to that of WF and AJ. Leaf of AJ showed no significant difference in triterpenoid concentration from June to October, and similar results were observed in leaf of WF and JX.

Analysis of homological relationship General morphological differences in the leaves
Leaf colour of FJ changed gradually from red to green, and similar changes were observed in leaf of AJ. However, there was no change in leaf colour of TG, WF, JX and JH (Table 4) Moreover, leaf shape was similar among FJ, AJ, WF and JH. Further, there were no differences in the shape of leaf apex and leaf base among the six provenances (Table 4).
Additionally, there were also differences in leaf number, leaf area and petiole length among the six provenances.  with other provenances. Leaf area of FJ was the lowest and had significant difference with other provenances, except for that of AJ. The common petiole length of JH was the longest, but there was no difference in that of FJ and JH (Table 4).

SEM observation on the leaf surfaces
Leaf of the six provenances had similar trichomes and stomata. However, the veins on the upper surface of the WF leaf were swollen, which was different from the others. In addition, more nectaries were observed in leaf of FJ and AJ in comparison to that of the other provenances (Fig. 6).

Allelic information based on SSR
A total of 86 alleles were detected with 10 pairs of SSR primers, of which 3-14 alleles were amplified among the primers, with an average of 8.6 ( Table 2). The highest number of alleles was 14 by S6, whereas 12 alleles were revealed by S10 ( Table 2). The mean percentage of polymorphic loci was 96.51%.
The number of alleles scored/locus ranged from 1.67 to 2 with a mean of 1.97. Meanwhile, the Ne/locus varied from 1.21 to 1.56, with a mean of 1.41, of which S2 had the highest number of effective alleles. The mean values of He and I were 0.27 and 0.43, respectively (Table 2).

Homology among the six provenances of C. paliurus
Among the six provenances, two groups were clustered with a coefficient of 0.51. In particular, group 1 contained only one provenance (WF), while group 2 included five provenances (FJ, AJ, JH, JX and TG), with FJ and AJ further clustering into one subgroup (Fig. 7).

Relationship between environment factors and plant growth
Plant growth is influenced by various environmental factors (Enßlin et al., 2011;Benlloch-González et al., 2016;Cao et al., 2018;Liu et al., 2018a;Wu et al., 2019), further demonstrated from our trend-surface analysis ( Fig. 8A and B), but which was contrary to a previous study in which Deng et al. (2015) indicated that environmental factors were not correlated with the height increase of C. paliurus. These inconsistent findings could be explained by the species-specific response of plant growth to environmental factors (Wu et al., 2019). Further, geographical differentiation developed in the Cyclocarya species, and the plants reported in Deng et al. (2015)

Physiological responses of plants to environment factors
Plants respond to changeable environment factors by a series of physiological activities (Habibi, 2017;Wu et al., 2018;Zhao et al., 2018;Wu et al., 2019;Feng et al., 2020b) as further demonstrated by our findings (Table 5, Fig. 8D). As essential substrates for plant growth, WC or TSS in leaf varied with the growth period (Figs 1 and 2) and related with environment factors (Table 5), suggesting they could take part in many biosynthetic processes for improving plant adaptation (Ben Abdallah et al., 2017;Wu et al., 2018). As essential nutrients for plant growth, mineral elements (K, Ca, Mg, and Na) play an important role in physiological functions, such as osmotic adjustments, water balance, water use efficiency improvement and stomatal control (Coskun et al., 2013;Gattward et al., 2012). A correlation analysis showed that K, Ca, Mg and Na were negatively correlated with temperature and precipitation (Table 5), suggesting that plants respond to varying environments by regulating their mineral element concentrations (Wu et al., 2018;Wu et al., 2019). In addition, plants have developed antioxidant defence systems to adapt to changing environmental factors during the growth period.   as SOD, CAT, PPO and POD, are expected to cope with the harmful effects of reactive oxygen species (ROS) and eliminate excessive H 2 O 2 and O 2 in the plant tissues (Burducea et al., 2019;Gill and Tuteja, 2010;Tang et al., 2012). Our results also showed that SOD activity in leaf of all provenances was the highest among the four examined antioxidant enzymes and was related to environmental factors during the growth period ( Fig. 4 and Table 5), inferring that SOD might be the main enzyme to interfere with the accumulation of ROS and to metabolize excess ROS produced under environmental stress (Burducea et al., 2019;Gill and Tuteja, 2010). According to our results, C. paliurus leaves were rich in flavonoids and polysaccharides, which was consistent with previous studies (Fang et al., 2011;Xie et al., 2012;Yang et al., 2017;Liu et al., 2018a, b;Shang et al., 2018). However, total flavonoid content in C. paliurus was greater than that reported in previous studies (Liu et al., 2018a, b;Zhou et al., 2019), whereas both of polysaccharide and total triterpenoid concentrations were lower than those previously reported in Deng et al., (2017) and Zhou et al. (2019). This may be because of differences in the extraction method: the extraction soluble used in our study was 75% ethanol, while that used in previous studies was water.
Variation in secondary metabolite accumulation is also influenced by environmental conditions (Liu et al., 2018a, b;Zhou et al., 2019). For example, Djerrad et al. (2015) reported that environmental conditions have an important effect on the essential oils in Pinus halepensis. Further, previous reports (Deng et al., 2015;Fang et al., 2011;Liu et al., 2018a, b) indicated that the growth environment significantly affected secondary metabolite accumulation in C. paliurus, which is in accordance with our findings (Table 5).  (Table 5), which was inconsistent with the findings of , who reported that Ca and Mg had a significant negative correlation with total flavonoid accumulation under five nitrogen fertilization levels . This discrepancy could be caused by nitrogen availability, which influences the absorption and distribution of mineral nutrients, further affecting secondary metabolite accumulation in C. paliurus .

Relationship among various responsive physiological indices
Environmental factors have an influence on physiological activities, in turn, various physiological activities are together involved in adapting changeable environment. K and Ca in leaf clustered into one group (Table 5, Fig. 8C and D) inferred that they could act on similar biological processes (Coskun et al., 2013;Gattward et al., 2012), including the promotion of osmotic protection, or the inhibition of leaf water loss via stomatal regulation (Ahmad and Maathuis, 2014;Peiter, 2011). In addition, four kinds of mineral nutrient were grouped with antioxidant enzymes (Table 5, Fig. 8C and D). For example, Mg had a positive relationship with SOD and CAT, especially the group was formed between Mg and CAT (Table 5; Fig. 8C and D). Mg directly or indirectly participates in biological processes in plants, such as the synthesis of chlorophyll (Masuda, 2008) and Ribulose-1,5-bisphosphatcarboxylase/-oxygenase activity (Portis, 2003); but Mg deficiency reduces the absorption and utilization of light energy, resulting in the production of ROS and the increase of antioxidant enzyme activity .

Homological relationship between leaf morphology and SSR analysis
Plants exhibit substantial genotypic diversity during the developmental processes. Leaves are one of the most visible and vital organs, with genotypic variations in shape, colour, margin and texture (Bruno et al., 2008), and this valuable morphological information can help identify species (Arturo et al., 2015;Kala and Viriri, 2018). In this study, leaves from the six provenances differed in shape, colour and area. However, the leaves from FJ and AJ had many similar characteristics, suggesting that their origin could be similar, which was further manifested by the results of the genetic relationship analysis.
The relatively high number of alleles generated by the SSR markers demonstrates the usefulness of the marker system for the detection of genetic diversity (William et al., 2020). Li et al. (2017) reported that 24 alleles were detected using six SSR markers. In our study, the number of alleles investigated ranged from 3 to 14, and the mean number was 8.6, which was significantly higher than the 3.83 alleles/locus reported by Li et al. (2017). This indicated that there was a good level of allelic diversity. Meanwhile, the mean gene diversity obtained in our study was different from the findings of Li et al. (2017), who reported a gene diversity of 0.09 among 26 natural provenances. This discrepancy could be due to the selection of different SSR markers. These results suggest that an SSR analysis is an efficient method for analysing genetic relationships in plants.
Homology relationships are important for understanding the evolutionary relationships among different genotypic resources and could facilitate breeding and conservation programs. In this study, according to the homology relationships, we found two groups (Fig. 7) with a coefficient of 0.51. The first of which contained only WF, and the second group included FJ, AJ, JH, JX and TG distributed in the Fujian, Zhejiang, Hubei and Jiangxi Provinces. These relationships are consistent with the results of Li et al. (2017), who reported that C. paliurus appeared to be an expanding species in subtropical China, but less genetic differentiation and a high gene flow occurred among natural populations of C. paliurus distributed in the Fujian, Zhejiang and Jiangxi Provinces, explaining why they were clustered into a larger group. Similar observations have been made in other species (Bhattarai and Mehlenbacher 2017;Dettori et al., 2015;Zhu et al., 2016;Liang et al., 2018;Patzak et al., 2020;Suvi et al., 2020). This confirmed that ex situ conservation benefits the preservation of this species' gene pool and maintains regional differences in diversity.

Conclusions
Plants of the six Cyclocarya provenances were conserved at a resource plantation in Quanzhou, Fujian Province, and studied variations in plant growth and leaf physiological response to environmental factors during the growth period and further analysed homological relationships by leaf morphological characteristics and SSR. The results showed that (i) plants of C. paliurus from the six provenances varied in growth; (ii) physiological changes during the growth period had differences in WC, TSS content, mineral content, antioxidant enzyme activity and secondary metabolite accumulation; (iii) variation in plant growth and physiological performances had significant relation with environmental factors, especially temperature and precipitation; (iv) leaf morphology among the six provenances differed in shape, colour and area. Moreover, two groups were clustered at a coefficient of 0.51 by SSR analysis, of which one contained only WF and the other included FJ, AJ, JH, JX and TG distributed in the Fujian, Zhejiang, Hubei and Jiangxi Provinces. The results of this study provide information on physiological response to environmental factors at a resource plantation of ex situ conservation and benefit to selecting suitable provenances for Cyclocarya cultivation.