Exogenous 5-azaCitidine accelerates flowering and external GA3 increases ornamental value in Iranian Anemone accessions

The Anemone genus is a tuberous geophyte which undergoes a dormancy period during unfavorable environmental conditions for growth. Five species of the Anemone genus naturally grow in several regions of Iran. The diverse uses of Anemone in gardens for landscaping, cut flowers, and potted plants indicate its high ornamental potential. Its dormancy and flowering are influenced by various factors. The present paper was conducted to explore the flowering behavior of Anemone accessions in response to different pre-treatments. For this purpose, tubers of 18 Anemone accessions (A. coronaria and A. biflora) were collected from natural regions of six provinces in Iran. These tubers were subjected to different conditions of non-chilling (20 °C, 90 days), chilling (4 °C, 90 days), GA3 (150 mgL-1; 24 h), and 5-azaCitidine (5-azaC; 40 µM; 24 h) prior to the cultivation. Most of the accessions were able to enter the flowering stage without chilling. The shortest period for the sprouting of tubers (16.89 ± 7.83 days) belonged to 5-azaC pre-treatment. In addition, this treatment accelerated the flowering time (about 30 days earlier) and diameter of the stem, bud, and flower. Morphological characteristics, such as stem height, number of leaves, bud, and petal and the longevity of flowers on the plant were significantly affected by GA3 pre-treatment. Our results indicated a positive correlation between flower length, stem height, and stem diameter with flower longevity under different pre-treatment conditions. The present study demonstrated that accessions Anm3, Anm12, and Anm18 had ornamental values higher than the population mean across four conditions.

. Analysis of variance (ANOVA) for measured morpho-phenological traits of 18 studied of Anemone accessions under different pre-treatment conditions. ns (Non-significant), *(Significant at 5% probability level) and **(Significant at 1% probability level). Pre-treatment conditions included non-chilling, chilling, 5-azaC and GA 3 . www.nature.com/scientificreports/ ing the tuber dormancy and flowering. In other words, different reactions were observed for the accessions in response to various pre-treatments. Therefore, to determine the differences between the accessions in more detail, the means and Z-scores were separately compared in each pre-treatment.

Source of variation df
Evaluation of different pre-treatments on phenological traits in Anemone accessions. The effects of four pre-treatment conditions (non-chilling, chilling, 5-azaC, and GA 3 ) on the phenological traits of 18 Anemone accessions were investigated ( Table 2). As seen, the fastest sprouting time belonged to 5-azaC and GA 3 pre-treatment. The shortest period to the sprouting of Anemone tubers (16.89 ± 7.83 days) was recorded for 5-azaC pre-treatment, which occurred approximately 18 days earlier than that in the non-chilling condition (no pre-treatment was applied to the tubers). Moreover, 24 h-immersion of the tubers in GA 3 solution accelerated sprouting to 18 days. On the other hand, the longest sprouting time was recorded for chilling and non-chilling pre-treatments lasting 34.98 ± 11.82 and 34.71 ± 15.39 days, respectively. Concerning the Z-score of the pre-treatments (Fig. 1), it was found that 5-azaC with a more negative Z-score (-0.672) was more effective on reducing the sprouting time than the flower bud formation time (Z-score = -0.29). It should be noted that the decrease or negative changes in the value of phenological traits (except for flower longevity on plant) are desirable features for horticultural researchers. Thus, the pre-treatments with further negative Z-scores were  Z-score was calculated as the difference between the mean of the trait (for each pre-treatment) and the total mean, divided by the total standard deviation (SD). The highest negative Z-scores are associated with the relative lowest value of the variable among the conditions. Since the decrease in the value of phenological traits (except for flower longevity on plant) is a desirable feature, the pre-treatments with more negative Z-scores were considered as more favorable for increasing the rate of plant growth. www.nature.com/scientificreports/ considered to be more favorable for increasing the rate of plant growth. A Z-score indicates the number of standard deviations by which the value of the mean of the treatment group is above or below the total mean value.
The results also revealed that flower bud formation in 5-azaC occurred earlier than that in other pre-treatments ( Table 2). The tubers treated with 5-azaC entered the budding stage about two weeks earlier than that in the non-chilling conditions. In contrast, flower bud color was detectable earlier in the no pre-treatment tubers (with maximum negative Z-score, -0.35); accordingly, less than 10 days following bud formation, bud color became visible. However, based on these findings, GA 3 and 5-azaC pre-treatments required more than 11 days to show bud coloring ( Table 2, Fig. 1). Additionally, the full opening of the flower in the non-chilling pre-treatment was faster (Z-score = -0.24) than that in other pre-treatments. According to the results, the longest period for flower opening after bud coloring belonged to the chilled tubers, which lasted about 8.3 days. Moreover, with 5-azaC, it took 7.6 days for the flower to fully open. On the contrary to the non-chilling pre-treatment (Z-score = -0.47), the highest longevity of flowers on the plant was recorded for GA 3 (Z-score = 0.26) and 5-azaC (Z-score = 0.21), respectively. Overall, the comparisons of the effects of pre-treatments among all the phenological traits showed that 5-azaC accelerated the flowering time of Anemone accessions the most.
Evaluation of different pre-treatments on morphological characteristics in Anemone accessions. As represented in Fig. 2, GA 3 had a more positive Z-score than that in other pre-treatments associated with most morphological traits, such as the number of flower buds, the number of petals, stem height, flower length, and leaf number. This means that GA 3 was effective on increasing the plant size. On the contrary, 5-azaC, with a negative Z-score (Fig. 2), reduced the flower bud number, leaf number, stem height, and root length. However, it had a positive Z-score for flower bud diameter, flower diameter, stem diameter, and tuber fresh weight compared to other pre-treatments. These results were obtained while the tubers with no pre-treatments before the cultivation had the most negative Z-scores for the assessed traits (Fig. 2). Therefore, the lack of chilling in the tubers resulted in a decrease in the size and dimensions of plant in comparison to other pre-treatments. In contrast, root length was longer in non-chilling conditions than that in other treatments. Overall, for morphological traits, GA 3 pre-treatment with the maximum positive Z-score was more effective on increasing the plant size as compared with other pre-treatments (Table 3; Fig. 2).

Clustering and identifying superior accessions under non-chilling condition.
In an attempt to better identify the response of each accession to different pre-treatments, the present set of accessions were separately grouped and analyzed in each pre-treatment condition. Four clusters were identified in the non-chilling condition ( Supplementary Fig. 1). The maximum number of accessions (eight accessions) were assigned to cluster I that had different sources (Table 4). Cluster I consisted of Anm3, Anm9, Anm10, Anm11, Anm12, Anm13, Anm15, and Anm18, whose mean values in terms of most traits were higher than the total mean (Table 4). In other words, in the absence of chilling, these accessions were identified to be superior. This cluster represented the accessions with a lower Z-score for sprouting time than the accessions in other clusters. Z-score comparison for different traits showed that the members of this group had a relative advantage over other accessions with larger aerial parts and early sprouting. Cluster II included accessions Anm1, Anm4, Anm7, Anm14, and Anm17; the most prominent feature of this group was the long time of sprouting. Cluster III, with two accessions of  Z-score was calculated as the difference between the mean of the trait (for each pre-treatment) and the total mean, divided by the total standard deviation (SD). The highest positive Z-scores are associated with the relative highest value of the variable among the conditions. Since increase in the value of morphological traits is a desirable feature, pre-treatments with further positive Z-scores were considered as more favorable for increasing the rate of plant growth. www.nature.com/scientificreports/ Anm6 and Anm8, had a negative deviation from the total mean for most of the traits. This means that the members of this group had low-quality traits compared to other accessions under non-chilling conditions. For note, the accessions in this group had relatively long roots. In contrast, the values in Table 4 depicts that cluster IV also contained two accessions (Anm5 and Anm16) that failed to enter the flowering phase despite early sprouting. Negative Z-score for root length and tuber fresh weight indicated that the underground organs in these accessions were limited under non-chilling conditions and their growth was slow (Table 4). Interestingly, accession Anm2 was unable to sprout under these conditions; thus, it did not fit into any of the clusters.
Clustering and identifying superior accessions under artificial chilling conditions. The dendrogram of 18 Anemone accessions under artificial chilling condition was constructed based on morpho-phenological parameters ( Supplementary Fig. 2). In general, cold exposure (90 days at 4 °C) during the storage of the tubers prior to the planting had an adverse effect on the growth and flowering of most accessions. Table 5 represents the results associated with the cluster analysis of the accessions under chilling conditions, in which the accessions are divided into four clusters. Among them, cluster II with six accessions (Anm3, Anm5, Anm12, Anm14, Anm17, and Anm18) had a positive deviation from the total mean related to all the morphological traits except for flower bud size and leaf number (Table 5). Furthermore, this cluster had a negative Z-score for sprouting time under chilling conditions. This implied the superiority of this cluster in improving the morphological traits and reducing the time for sprouting. On the contrary, the first group (cluster I with eight accessions) had values lower than the total mean for most morphological and phenological traits. The third group (cluster III with Anm1 and Anm16) was superior to the other groups regarding phenological traits and the short flowering period. The fourth group (cluster IV with one accession; Anm7) had larger underground parts and flower bud, but it had a longer sprouting time compared with other accessions (Table 5). This accession, well-separated from other groups under chilling conditions, belonged to A. biflora species. Similar to non-chilling conditions, sprouting for accession Anm2 did not occur under artificial chilling conditions while accessions Anm5 and Anm16 entered the flowering stage once they received cold treatment.
Clustering and identifying superior accessions under GA 3 immersion. Based on the cluster analysis of accessions under a 24-h immersion with GA 3 prior to the planting, the clear distinction among accessions occurred when they were grouped into three major clusters ( Supplementary Fig. 3). The 11 accessions (including Anm1, Anm3, Anm5, Anm8, Anm9, Anm10, Anm12, Anm13, Anm14, Anm15 and Anm18) characterized by early sprouting ability and positive Z-score for the aerial parts (Table 6) were assigned to the first cluster (I).
These characteristics indicate the superiority of the accessions in this group in terms of ornamental features. Moreover, the members of the same group had negative Z-scores for the tuber weight and root length, implying the weakness of the accessions in the underground parts. The prominent indices of the second group (cluster II) consisted of five accessions (Anm2, Anm4, Anm7, Anm11 and Anm17); these accessions had strong underground parts, such as long root and weighty tuber. This group did not score well for sprouting time and most of the morphological traits associated with flower section. In contrast, accessions Anm6 and Anm16 with negative Z-scores for most of the traits (except for root length) belonged to the third group (cluster III). Evidently, these accessions differed from the others, and this cluster was the weakest group under tuber immersion in GA 3 ( Table 6). An interesting point in this section was the effect of GA 3 on accession Anm2, which induced sprouting and flowering in this accession.

Clustering and identifying superior accessions under 5-azaC application. According to cluster
analysis under a 24-h immersion of tuber with 5-azaC, the biggest difference among the groups was related to www.nature.com/scientificreports/ when the accessions were divided into four main clusters ( Supplementary Fig. 4). Under these conditions, the first group (cluster I) with 13 accessions (Anm1, Anm3, Anm5, Anm7, Anm8, Anm9, Anm10, Anm12, Anm13, Anm14, Anm15, Anm17, and Anm18) was the largest one (Table 7). Based on the comparison of the Z-scores of the traits, it could be inferred that the members of this group had taller plants with larger leaves and flowers. The superior accessions were placed in cluster I and characterized by their higher ornamental value under 5-azaC application. Meanwhile, the second group (cluster II), comprising two accessions (Anm2 and Anm4), entered the flowering stage (with early sprouting) faster than that in other groups under 5-azaC application. This group had a relative superiority in terms of flower longevity on plant and tuber weight compared over the other groups. The most distinguishing feature in the third group (cluster III), with one accession (Anm11), was the large number of flower buds. Similar to the grouping of GA 3 condition, accessions Anm6 and Anm16 were clustered together (cluster IV), which is indicative of their poor response to 5-azaC. Plant size and phenology were negatively affected by 5-azaC; therefore, the general characteristics of the two accessions were late sprouting and small plant size. On the other hand, this group had longer roots once exposed to 5-azaC (Table 7). It is noteworthy that the effect of 5-azaC was similar to that of GA 3 on accession Anm2, where the 24-h immersion in 5-azaC caused sprouting and flowering in this accession.
Comparison of superior accessions under different pre-treatment conditions. The overall comparison of cluster analysis results under different pre-treatment conditions (non-chilling, chilling, GA 3 , and Table 4. Composition of the groups and the distribution of the Anemone accessions in each group based on cluster analysis under non-chilling pre-treatment. *: Z-score was calculated as the difference between the mean of the trait (for each group) and the total mean, divided by the total standard deviation (SD). The highest positive Z-scores are associated with the relative highest value of the variable among the groups. Since the decrease in the value of phenological traits (except for flower longevity on plant) is a desirable feature, the groups with further negative Z-scores were considered to be more superior. Conversely, since the increase in the value of morphological traits is a desirable feature, the groups with further positive Z-scores were considered to be superior. www.nature.com/scientificreports/ 5-azaC) showed that in all the four conditions, Anm3, Anm12, and Anm18 were relatively superior regarding most evaluated traits (Fig. 3). The results revealed that Anm3, Anm5, Anm12, Anm14, and Anm18 exhibited relatively higher values compared with the total mean under chilling, GA 3 , and 5-azaC. In addition, Anm1, Anm3, Anm5, Anm8, Anm9, Anm10, Anm12, Anm13, Anm14, Anm15, and Anm18 had higher values than the total mean in both GA 3 and 5-azaC pre-treatment. Anm7 was the single accession with a value higher than the total mean under 5-azaC application conditions (Fig. 3).
Correlation coefficient between the studied indices of accessions under different pre-treatment conditions. To determine the best phenological and morphological indices for identifying the accessions with high ornamental values, the correlation coefficients between the traits were calculated under different pre-treatment conditions (Table 8). According to the results, the longevity of flower on the plant, as an important ornamental index, had a significant positive correlation with flower length under four pre-treatment conditions, namely non-chilling (r = 0.89), chilling (r = 0.57), GA 3 (r = 0.62), and 5-azaC (r = 0.51). In addition, stem length and stem diameter showed a significant positive association with flower longevity under non-chilling, GA 3 , and 5-azaC conditions ( Table 8). Table 5. Composition of the groups and the distribution of the Anemone accessions in each group based on cluster analysis under chilling pre-treatment. *: Z-score was calculated as the difference between the mean of the trait (for each group) and the total mean, divided by the total standard deviation (SD). The highest positive Z-scores are associated with the relative highest value of the variable among the groups. Since the decrease in the value of phenological traits (except for flower longevity on plant) is a desirable feature, the groups with further negative Z-scores were considered to be more superior. Conversely, since the increase in the value of morphological traits is a desirable feature, the groups with further positive Z-scores were considered to be superior.

Discussion
Iran is a country rich in germplasm resources of the genus Anemone 1 , known in floriculture as an early flowering geophyte. This genus has a high ornamental potential and can be utilized for commercial production. Therefore, development techniques for sustainable uses, such as propagation, forcing, and breeding, could play pivotal roles in conserving the genetic resources of Anemone. There are no reports of wild Anemone variation in Iran, and the present study is the first investigation focusing on flowering behavior of Anemone accessions in response to different pre-treatments in the world. The represented information herein could be employed by researchers, ornamental plant producers, and landscape designers to improve conservation strategies. Initiating breeding programs for commercial production requires understanding the physiological behavior of Anemone in the dormancy and flowering process of tubers. The present data provided evidence for the flowering behavior among the studied wild accessions. Pre-treatment of tubers with GA 3 reduced the sprouting time about 17 days earlier than that in non-chilling and chilling conditions. However, the fastest sprouting time (16.89 ± 7.83 days) was recorded for the treated tubers under 5-azaC. Using 5-azaC treatment, flower buds were visible about 65.54 ± 19.12 days after sprouting, followed by GA 3 (75.02 ± 19.35 days), chilling (76.81 ± 21.72 days); the longest time (79.62 ± 45.56 days) until flower bud formation was detected in non-chilling treatment ( Table 2). These findings revealed that chilling is not necessary for sprouting and flowering of most Anemone accessions; thus, regrowth and flowering occurred even in chilling-free condition (albeit with delay). Notwithstanding, there were no differences in sprouting time between chilling and non-chilling conditions, and hormonal pre-treatments significantly accelerated the sprouting process. For note, in our study, 5-azaC accelerated the bud formation whereas chilling and GA 3 pre-treatment had insignificant effects (about 4 to 5 days) on bud formation. It has been confirmed that 5-azaC contributes to earlier flowering by reducing DNA methylation 32,33 . Overall, concerning the studied phenological traits of Anemone accessions, 5-azaC treatment more significantly reduced the vegetative growth period in comparison with other treatments, which is explained by the comparison of their growth period. Pre-treatment with 5-azaC lasted about 101 days from planting to flowering. Subsequently, GA 3 pre-treatment required approximately 112 days to flowering. For chilling and non-chilling, the required time from planting to flowering was about 130 days from planting. Therefore, the one-month reduction in the growth period clearly indicated the positive effect of 5-azaC (Table 2). Based on the results, the lack of chilling increased the vegetative growth period of the plants. Moreover, the root was longer under the non-chilling condition. These results are in agreement with a previous study on three cabbage cultivars 41 where 5-azaC decreased the germination and flowering time of cabbages.
Genes associated with vernalization and flowering are activated by DNA demethylation 35 . Transfer of plants from vegetative to flowering is an important process controlled by a network of genes in plants 42 . Vernalization in the plant could be effectively replaced by 5-azaC treatment and act as a flowering accelerator 43 . The 5-azaC either increases the amount of carbohydrate (leaf sugar content) through regulating the expression of carbohydrate metabolism genes and flowering genes 44 or stimulates flowering by reducing inhibitor hormones, IAA www.nature.com/scientificreports/ for instance 45 . As mentioned in the results, 40 μM 5-azaC had a negative effect on the phenotype of Anemone accessions, reducing the plant height, root length, and the number of leaves and flower buds. A similar result was reported by Li et al. 35 in regard to spinach. They reported that the presence of 5-azaC decreased the plant height and root length and accelerated flowering. Consistent with the present results, 5-azaC reduced the leaf number and length in Erodium cicutarium 46 and resulted in early flowering in Pharbitis nil 32 . Conversely, 5-azaC treatment had a positive effect on flower diameter, stem diameter, and tuber weight. Our findings did not reveal any morphological abnormalities except for the decrease in height. Reduced DNA methylation could lead to abnormal plant phenotypes 47 . Our results are in line with those of Yingduan et al. 48 who reported that phenotypic changes, such as dwarfing and small leaves, were triggered with 5-azaC treatment in wheat. Unlike our results, deformation of phenotype in Jatropha curcas 31 and Chrysanthemum 33 were reported. Based on our results, GA 3 (150 mgL -1 ) significantly increased the plant height, leaf number, petal number, and flower longevity in Anemone accessions (Tables 2, 3). It also reduced the sprouting time and vegetative growth period. External GA 3 can act as a stimulator for internal GA 3 synthesis 17 . Gibberellins play a role in breaking the dormancy, accelerating the germination and flowering time 49 and increasing the vegetative growth by affecting the synthesis and activity of hydrolyzing enzymes in storage resources 50 . In addition, gibberellins improve the photosynthetic efficiency by enhancing photosynthetic enzymes and leaf area index and increasing the nutrient uptake 51 . Ferrante et al. 52 reported that gibberellins were effective on reducing the breakdown of ribonucleic acid and protein, delaying aging, and increasing plant longevity via declining the production of ethylene. Hence, these reasons could justify the increase in the growth parameters of Anemone accessions under GA 3 pre-treatment. Table 6. Composition of the groups and the distribution of the Anemone accessions in each group based on cluster analysis under GA 3 pre-treatment. *: Z-score was calculated as the difference between the mean of the trait (for each group) and the total mean, divided by the total standard deviation (SD). The highest positive Z-scores are associated with the relative highest value of the variable among the groups. Since the decrease in the value of phenological traits (except for flower longevity on plant) is a desirable feature, the groups with further negative Z-scores were considered to be more superior. Conversely, since the increase in the value of morphological traits is a desirable feature, the groups with further positive Z-scores were considered to be superior. www.nature.com/scientificreports/ Accessions clustering under four pre-treatment conditions of tubers revealed a significant variation in the flowering behavior of Anemone accessions in response to non-chilling, chilling, GA 3 , or 5-azaC. Therefore, each area has its own management practices with regards to Anemone. The clustering of accessions from the same province or an adjacent origin into different clusters suggests the diversity of the assessed accessions. Our findings depicted that cluster analysis allowed the selection of superior accessions in all the pre-treatment conditions. The superiority of accessions Anm3, Anm9, Anm10, Anm11, Anm12, Anm13, Anm15, and Anm18 in non-chilling conditions (Table 4) indicated that these accessions could be utilized for cultivation in areas with short winters. It is noteworthy that most of the accessions superior in non-chilling conditions belonged to Abdanan region which has a subtropical climate and is located within the Zagros Mountains in the south of Ilam province, Iran. The excellent accessions in chilling conditions were Anm3, Anm5, Anm12, Anm14, Anm17, and Anm18 (Table 5), often distributed in areas with cold winters, for instance, Nahavand, Kerrend, and Ilam in Iran. It is expected that the use of these accessions in areas with long and cold winters result in optimal flowering. Ultimately, Anm3, Anm12, and Anm18 were selected as superior accessions in all the four pre-treatment conditions, which could be used as an important genetic resource in breeding programs. The distribution pattern (Kermanshah and Ilam) of these three accessions implied that they require low temperatures in winter for optimal growth and flowering. In cold periods, the above-mentioned accessions are also capable of growing without any restriction even in areas with short winters. Furthermore, these results shed light on the fact that a 90-day chilling was not sufficient for dormancy release in accession Anm2 in which sprouting was not observed in the differential response at this stage (belonging to A. biflora species). Tuber non-sprouting of this accession under non-chilling and chilling conditions indicated that further stimulus, such as longer periods of low temperature, is required for sprouting and flowering, which is confirmed by the stimulation of its growth and flowering under GA 3 and 5-azaC. This accession was collected from Razavi Khorasan located in northeastern Iran. Higher latitude, cold, dry, and relatively long winters, and the overnight cold air until mid-spring, are the prominent features of Khorasan Razavi. These climatic conditions confirm that this species (A. biflora) of Anemone requires a longer period of cold to activate the growth stimuli. Exposure to GA 3 and 5-azaC was able to meet this requirement.
Flower longevity on the plant is one of the important factors for the evaluation of ornamental plants. Understanding the relationships between the traits and identifying important and effective ornamental attributes help researchers select the superior accession. Generally, stem length and stem diameter were positively correlated with flower longevity on plants across non-chilling, GA 3 , and 5-azaC conditions. These results showed that these three indices could be considered as the criteria for selecting Anemone accessions with higher ornamental values. Thus, a tall accession with a thicker stem and a larger flower could be introduced as a superior accession.

Conclusion
In conclusion, our findings introduced a novel and practical approach to accelerating flowering in off-season Anemone production. Based on the results, 5-azaC had an acceptable potential for reducing the growth period of Anemone. However, GA 3 had a better effect on the improvement of the plant appearance. In the present study, most of the accessions did not require low temperatures for sprouting, and the flowering process was even performed without chilling; however, certain accessions for sprouting required vernalization and their flowering was found to be associated with low temperatures. The superior accessions were identified in terms of phenological and ornamental values in each pre-treatment condition. In total, Anm3, Anm12, and Anm18 were selected as superior accessions in all the four pretreatment conditions. Furthermore, this research could be a starting point for the design of novel strategies in order to develop effective conservation and management measures for a sustainable improvement in Anemone. Therefore, further studies are required to understand the behavior and mechanism of dormancy and flowering in Anemone.

Material and methods
Plant materials, experimental design, and cultivation process. All methods were performed in accordance with the relevant guidelines and legislation. From February to May 2018, intact tubers of 18 accessions related to Anemone during their flowering time were collected from six provinces of Iran (Table 9; Fig. 4). Generally, most of the Iranian accessions are geographically originated in Ilam province (12 accessions), followed by Lorestan (two accessions), Kermanshah (one accession), Khuzestan (one accession), Hamedan (one accession), and Razavi Khorasan provinces (one accession). In order to obtain representative samples and avoid collecting the clones of the selected accessions, an appropriate distance (200 m) was considered between the accessions in each site. The details of Anemone accessions are presented in Table 9. The identification of Anemone species was performed based on Flora Iranica 1 and Colorful Flora of Iran 53 . Anemone tubers of each site were dried and stored in paper bags until the onset of the experiment. Tubers of these accessions were planted and grown in the experimental greenhouse of Ilam University (Ilam, Iran) with uniform conditions. The geographical coordinates of the location are 33.65°N latitude, 46.37° E longitude, and 2068 m altitude. The planting was carried out at the beginning of October 2018 utilizing tubers. The tubers of all the accessions listed in Table 9 are currently available to other researchers in the research greenhouse of Ilam University.
The accessions were evaluated for morpho-phenological traits as a factorial experiment in a completely randomized design (CRD) with three replications. We performed the experiment in four different pre-treatment conditions (non-chilling, chilling, 5-azaC, and GA 3 ). Primarily, for artificial chilling pre-treatment, the tubers were dry stored at 4 °C (in refrigerator) for 13 weeks in dark conditions. Afterwards, 24 h before the planting, the chilling treated tubers were transferred to room temperature to start the experiment and forcing. The tubers of chilling and non-chilling pre-treatments were simultaneously soaked (24 h) in distilled water. Regarding 5-azaC (Sigma-Aldrich; 40 µM) and GA 3 (Merck; 150 mgL -1 ) pre-treatments, 24 h prior to the cultivation, the tubers were immersed in the above-mentioned concentrations. Following the immersion time, to remove any residual www.nature.com/scientificreports/ materials on the surface of the tubers, they were washed thoroughly in distilled water. Subsequently, the tubers were individually potted into a black plastic pots (14 cm × 14 cm) filled with a mixture of peat moss, perlite, and coco peat (2:1:1) as growing media and placed in a greenhouse. Forcing conditions in the growth greenhouse were 23 °C/17 °C day/night temperatures with a RH 75% until flowering time. The pots of each treatment were randomly placed in the research greenhouse. All the plants were individually monitored on a daily basis over the whole period of the experiment. Normal horticultural practices, such as fertilizer application, irrigation practices, and pesticide application were applied during the experiment. As soon as the substrate surface of the pots was dried, irrigation was carried out during the growing season according to the need of the pots. Fertilizer (NPK; 20-20-20) was applied (2 mgL -1 ) three times at one-month intervals along with irrigation water. Aphids are among the key pests of Anemone, particularly at the bud stage. To control aphids during the experiment, we performed foliar application of diazinon insecticide (1 mgL -1 ) through hand sprayers.
Phenological and morphological analysis. Phenological and morphological traits were measured once the plants reached the flowering stage. We recorded the studied phenological parameters, including days to sprouting (number of days from tuber planting to sprouting), days to bud stage (number of days from sprouting to flower bud appearance), days to bud stage in color (from bud appearance until bud showing color), days to flower anthesis (from colored bud to fully opened flower), and flower longevity on plant (from anthesis to wilting of petals). Morphological traits, such as flower bud number, bud length (mm), bud diameter (mm), leaf number, leaf width (mm), leaf length (mm), petal number, flower length (mm), flower diameter (mm), stem height (mm), stem diameter (mm), root length (mm), and tuber fresh weight (gr), were also measured. Digital caliper (Guanglu, resolution: 0.01 mm) and ruler were employed to measure these traits.
Statistical analysis. Using SAS 54 software version 9.453, analysis of variance (ANOVA) was performed on the accessions and pre-treatments as factors. Since the evaluated morpho-phenological parameters have different units of measurement scale, it is possible to compare different traits only following standardization and converting them into Z-scores. In order to avoid the effects of scale differences, the mean of each character was standardized using Z-score. Standardization data (Eq. 1) allow the comparison of different traits regardless of unit.  www.nature.com/scientificreports/ Table 7. Composition of the groups and the distribution of the Anemone accessions in each group based on cluster analysis under 5-azaC pre-treatment. *: Z-score was calculated as the difference between the mean of the trait (for each group) and the total mean, divided by the total standard deviation (SD). The highest positive Z-scores are associated with the relative highest value of the variable among the groups. Since the decrease in the value of phenological traits (except for flower longevity on plant) is a desirable feature, the groups with further negative Z-scores were considered to be more superior. Conversely, since the increase in the value of morphological traits is a desirable feature, the groups with further positive Z-scores were considered to be superior.  Table 8. Correlation coefficient between flower longevity and morphological indices of Anemone accessions under different pre-treatment conditions. *(Significant at 5% probability level) and ** (Significant at 1% probability level). www.nature.com/scientificreports/ In this equation, x, μ and σ are the raw score, mean, and standard deviation, respectively 55 . Z-scores of four treatments for different traits were used to compare the mean of the treatments and their effects on traits. A Z-score indicated the number of standard deviations by which the value of the mean of the treatment group is above or below of the total mean. The highest positive Z-scores were associated with the relative largest value of the variable among the groups.
Cluster analysis was performed based on Ward's method using squared Euclidean distance and discriminate analysis executed to identify the cutting point using SPSS 56 . In addition, the correlation coefficients between the traits were calculated using Pearson's correlation coefficients with SPSS 56 .  www.nature.com/scientificreports/