Resolving the Puzzle of Iris maackii (Iridaceae): A Morphological Insight into Its Taxonomy

Since the early 20th century, Iris maackii (Iridaceae) has been considered a synonym of I. laevigata, a synonym of I. pseudacorus, or an accepted species. The current concept of I. maackii in the literature and databases is often applied to yellow-flowered plants with prominently veined rosette leaves, which are diagnostic features of I. pseudacorus growing in Northeast Asia. Therefore, the objective was to clarify the taxonomic identity of I. maackii. This study is based on a critical examination of the literature, on the observed morphological characters in the holotype of I. maackii, and on a morphological comparison of I. maackii with living plants of I. laevigata and I. pseudacorus. Additionally, a morphometric comparison of the seed characters was carried out to clarify the morphological distinction among I. maackii, I. laevigata, and I. pseudacorus. A careful study demonstrated that the rosette leaf texture and the morphology of the flowering stem, fruit, and seeds of I. maackii are identical to or within the variation range of I. laevigata. Thus, I. maackii is morphologically non-distinct from I. laevigata and should be recognized as a taxonomic synonym of the latter. An image of the holotype of I. maackii is provided along with detailed illustrations of I. laevigata and I. pseudacorus.


Introduction
Richard Maack, an explorer and naturalist, took part in an expedition up the Ussuri River in June-August 1859 [1].The botanical material from that expedition was treated by Eduard August von Regel in St. Petersburg [2].In particular, Maack collected a fruiting specimen (Figure 1) from marshes on the Chinese left bank of the Ussuri River upstream of Shang-Ong (currently known as Hutou, northeastern Heilongjiang Province, China), opposite the mouth of the Iman River (currently known as the Bol'shaya Ussurka River, Primorsky Krai, Russia), on 15 July (27 July, according to the new calendar), 1859 [1].Originally, Regel [2] (p.148) identified this specimen as Iris pseudacorus L. However, Carl Johann Maximowicz described I. maackii Maxim.on the basis of Maack's specimen [3].
Iris laevigata and I. pseudacorus are ornamental, wetland-associated, herbaceous perennials belonging to I. ser.Laevigatae (Diels) G.H.M.Lawr., according to the conservative taxonomy of Iris [8,9,25,26].To the best of my knowledge, the blue-flowered I. laevigata (Figure 2a,b) is native to Northeast Asia, i.e., to the Russian Far East, northeastern China (Heilongjiang and Jilin provinces), the Korean Peninsula, and Japan (Hokkaido and Honshu islands).The natural distribution of I. pseudacorus, long known as the "yellow iris" and "yellow flag" (Figure 2c,d), covers Europe and extends to Western Siberia, Western Asia, and the northern fringe of Africa [27,28].low iris" and "yellow flag" (Figure 2c,d), covers Europe and extends to Western Siberia, Western Asia, and the northern fringe of Africa [27,28].Regarding I. pseudacorus, it is necessary to pay special attention to some of its biological characteristics.On the one hand, due to the ability of I. pseudacorus to remove pollutants from water [27,[29][30][31] and soil [32], it has been suggested to be used as an available and economically efficient species for phytoremediation.On the other hand, since I. pseudacorus has been extensively cultivated and naturalized, it is becoming highly invasive in North America, in the southern half of South America, southern South Africa, southeastern Australia, and New Zealand [21,33].However, I. maackii is currently illustrated with images of I. pseudacorus and, therefore, these taxa are actually considered to be identical [34][35][36][37][38][39].For this reason, the study of the morphological characters of I. maackii, as determined by its nomenclatural type, will contribute to the understanding of its taxonomy.This circumstance can undoubtedly improve the monitoring of the species' invasion and help adjust the biocontrol program for I. pseudacorus [40][41][42].Regarding I. pseudacorus, it is necessary to pay special attention to some of its biological characteristics.On the one hand, due to the ability of I. pseudacorus to remove pollutants from water [27,[29][30][31] and soil [32], it has been suggested to be used as an available and economically efficient species for phytoremediation.On the other hand, since I. pseudacorus has been extensively cultivated and naturalized, it is becoming highly invasive in North America, in the southern half of South America, southern South Africa, southeastern Australia, and New Zealand [21,33].However, I. maackii is currently illustrated with images of I. pseudacorus and, therefore, these taxa are actually considered to be identical [34][35][36][37][38][39].For this reason, the study of the morphological characters of I. maackii, as determined by its nomenclatural type, will contribute to the understanding of its taxonomy.This circumstance can undoubtedly improve the monitoring of the species' invasion and help adjust the biocontrol program for I. pseudacorus [40][41][42].
the confusion around this name.A comparison of rosette leaves, flowering stems, fruits, and seed morphology among I. maackii, I. laevigata, and I. pseudacorus, including data from the literature and field surveys, is presented.Detailed morphological illustrations of I. laevigata and I. pseudacorus based on complete material collected by the author are provided.

Plant Material and Morphological Study
Iris maackii was examined based on a single specimen (LE01010783!; Figure 1) that is a holotype of the name, consisting of two rosette leaf fragments and the upper part of the flowering stem, bearing mainly immature fruit.For plant morphology, this specimen was re-examined and 20 morphological characters, including 14 quantitative and 6 qualitative, were selected.These characters are listed in detail in Table 1.The author collected a total of 90 individuals of I. laevigata from a wild locality in the vicinity of Shtykovo Village (43°21′35″ N 132°22′1″ E, Primorsky Krai, Russia) on 27 June 2021 (Table S1).From 22 July to 3 August 2021, 63 individuals of I. pseudacorus were measured directly in the living collection of the Botanical Garden-Institute (BGI, Vladivostok, Russia).The measurements were taken during fruiting.This study aims to clarify the taxonomic identity of I. maackii in order to disentangle the confusion around this name.A comparison of rosette leaves, flowering stems, fruits, and seed morphology among I. maackii, I. laevigata, and I. pseudacorus, including data from the literature and field surveys, is presented.Detailed morphological illustrations of I. laevigata and I. pseudacorus based on complete material collected by the author are provided.

Plant Material and Morphological Study
Iris maackii was examined based on a single specimen (LE01010783!; Figure 1) that is a holotype of the name, consisting of two rosette leaf fragments and the upper part of the flowering stem, bearing mainly immature fruit.For plant morphology, this specimen was re-examined and 20 morphological characters, including 14 quantitative and 6 qualitative, were selected.These characters are listed in detail in Table 1.The author collected a total  S1).From 22 July to 3 August 2021, 63 individuals of I. pseudacorus were measured directly in the living collection of the Botanical Garden-Institute (BGI, Vladivostok, Russia).The measurements were taken during fruiting.For seed characters, 50 samples from each collection site (see Table S2) were used (all were measured with the same сaliper) Seed width (W) Seed thickness (T) Seed L/W ratio The length-to-width ratio (L/W) provides additional data on the seed shape, i.e., degree of elongation along the hilum-chalaza axis (all calculations were manual) For the seed morphology, material from eight collection sites was used (Table 2).The study was conducted on mature seeds.The seeds of I. maackii were taken from one of the locules in the only mature fruit of LE01010783.For I. laevigata and I. pseudacorus, seeds were collected from different individuals for each site.The seeds of I. laevigata were from three wild localities in Russia, one of which is located near Lake Baikal, from where the species was described [43], and from two localities in Primorsky Krai.The seeds of I. pseudacorus were from the living collection of the BGI, Vladivostok, and also from two localities in Sakhalin Island (originally identified as I. maackii), one of which is indicated in references [16,44], and from a native population in the Don River delta.
The terminology used in the descriptions follows reference [45].For the taxonomy, the Shenzhen Code (hereafter, ICN [46]) was consulted.Relevant literature, including the protologue of I. maackii [3], was also analyzed.The herbarium codes follow Index Herbariorum [47].The collection site No. 6 was mentioned in references [16,44]; all sites of I. laevigata and I. pseudacorus are from Russia.* Herbarium codes follow Index Herbariorum [47].** Seed laboratory, Botanical Gardens of Peter the Great, Komarov Botanical Institute RAS (St.Petersburg, Russia).

Morphometric Analysis
The morphometric analysis was based on four seed characters (see Table 1), and used 50 seeds from each collection site (Tables 2 and S2).All the statistical analyses were performed in the R software [48], version 4.1.2[49].The data were evaluated by one-way analysis of variance (ANOVA).After multiple statistical testing, the calculated p-values were adjusted using the procedure proposed by Benjamini and Hochberg [50].To test the ANOVA assumptions, the Shapiro-Wilk test for the normality of the distribution [49] and Levene's test for the homogeneity of the variance [51] were performed.The effect size (η2) for ANOVA and Cohen's d for the difference in the means were calculated using the respective functions of the R add-on package "lsr", version 0.6.1 [52].If the ANOVA showed a statistically significant difference among species, then subsequent pairwise comparisons were made using Dunnett's many-to-one test [53].The data for I. maackii were used as a control.The inequality of variance was taken into account by using the heteroscedastic consistent covariance estimation provided in the R add-on package "sandwich", version 2.3.0 [54,55].The differences between the mean values of each collection site (Table 2) and the control were considered statistically significant at a p-value < 0.05.
Finally, a principal component analysis (PCA) [56] was performed on the morphometric parameters of the seeds, i.e., the L, W, T, and L/W ratio (see Table 1), to visualize the distribution of species over the space of the quantitative multivariate data and to assess their delimitation.For the PCA analysis, the built-in function prcomp was used, and the results of the analysis were extracted and visualized using the respective functions of the factoextra R package [57].In the PCA scatter plot, only the first (PC1) and second (PC2) principal components were considered to represent the data.

Results
The selected characters of the species under study are shown in Figure 3 and Table 3 (also see Table S1).The holotype of I. maackii and the I. laevigata individuals were found to have no more than one lateral cluster per flowering stem, while in I. pseudacorus, there were usually more than one, and up to four, lateral clusters (Figures 1 and 3a,b).Iris maackii and I. laevigata had smoothed rosette leaves that lacked the prominent midrib, a taxonomic feature of I. laevigata, with an average width of 1.7 cm; in I. pseudacorus, these leaves were prominently ribbed and broader, on average, by 60%, to 4.4 cm wide (Figures 1 and 3c).The shoot of the upper lateral cluster was conspicuous and comparatively long in I. maackii and I. laevigata; in I. pseudacorus, it was usually inconspicuous or shorter than in I. laevigata (Figure 3a,b and Table 3).In I. maackii and I. laevigata, the upper cauline leaf was usually longer than in I. pseudacorus (Figure 3a,b).The bracts of I. laevigata were dry during fruiting; in I. pseudacorus, they were green.The number of fruit per stem was no greater than seven in I. maackii and I. laevigata, and they had smoothed surfaces and were obtuse at the apex; in I. pseudacorus, they were numerous (on average, 10), three-angled, and conspicuously beaked at the apex (Figure 3d).All three species shared the following characters: the bract and pedicel length, the number of fruit per terminal cluster and per upper lateral cluster, and the length and width of the fruit, which were oblong, cylindrical, and obtuse at the base (Figures 1 and 3, Table 3).1).Photos by the author.Images were taken using an ObjectScan 1600 scanner (Microtek International Inc., Taiwan).The seed characters of the species under study are presented in Table 4 (also see Table S2).Morphologically, the seeds from all collection sites (Table 2) were brown, glossy, and flattened with a smooth surface and a more or less fragile testa (Figure 3e).The seed shape was oblong (L/W ratio: 1.4) or D-shaped in I. maackii; oblong (L/W ratio: 1.3-1.4),D-shaped, or, rarely, almost round in I. laevigata; in I. pseudacorus, the seed shape was almost round (L/W ratio: 0.9-1.1),predominantly subacute, or, much rarer, rounded at the chalaza.The seeds of I. maackii were similar in size to those of I. laevigata (Table 4).The seed length in I. laevigata ranged from 4.5 to 7.5 mm, the width from 3.8 to 6.6 mm, and the thickness from 1.4 to 3.6 mm.In I. pseudacorus, the values of these characters were greater: the seed length ranged from 5.6 to 9.6 mm, the width from 5.2 to 10.3 mm, and the thickness from 1.7 to 4.7 mm.0.9 ± 0.1 (0.7-1.4) 8.1 ± 0.5 (6.9-9.2) 7.8 ± 0.7 (5.6-8.8)3.0 ± 0.5 (2.2-4.5)1.1 ± 0.1 (0.9-1.4) 7.4 ± 0.6 (6.4-8.9)7.5 ± 0.5 (5.2-8.4)2.8 ± 0.3 (2.2-3.5)1.0 ± 0.1 (0.9-1.5) 7.1 ± 0.6 (5.6-8.7) 7.8 ± 0.5 (6.8-8.5)2.5 ± 0.5 (1.7-3.9)0.9 ± 0.1 (0.7-1.1) The descriptions of the characters are provided in Table 1.The collection site numbers (No.) correspond to those in Table 2. Data are presented as the mean ± standard deviation (minimum-maximum).See supplementary raw data in Table S2 for more details.All measurements are in millimeters.
A statistically significant difference in all the morphometric parameters of the seeds was observed between the collection sites of I. laevigata and I. pseudacorus (see Table 1) and I. maackii (Figure 4; for ANOVA results, see Table S3).In particular, the seeds from sites 2 and 4 were approximately 10% shorter than those of I. maackii, with this difference being statistically significant (p < 0.001); the difference in the mean seed length between site 3 and I. maackii was 1.5%, being statistically non-significant (p > 0.05).The seeds from sites 5-7 were approximately 6.8-21.9%longer than the seeds of I. maackii, with this difference being statistically significant (p < 0.001).The difference in the seed length between site 8 and I. maackii was no greater than 2.7% and was statistically non-significant (p > 0.05).2) and I. maackii, selected as a control.The results of the test are presented as p-values, differences between the mean values of each experimental group (black dots), and 95% confidence intervals of these differences (whiskers) for each pairwise comparison.X-axis is differences in means.Dash-dotted line indicates zero difference.See Table S3 for more details.
No statistically significant differences in the seed width and thickness were found between collection sites 3 and 4 and I. maackii (p > 0.05), while there was a small, approximately 9.4%, but statistically significant (p < 0.001), difference between site 2 and I. maackii.There was no difference in the L/W ratio between sites 2 and 3 and I. maackii (p > 0.05).However, a small, approximately 6.8%, but statistically significant (p < 0.001), difference was found in the L/W ratio between site 4 and I. maackii.For I. pseudacorus, the mean seed width was approximately 37% greater, the seed thickness approximately 33.5% greater, and the L/W ratio approximately 41% smaller than for I. maackii and these differences were statistically significantly different (p < 0.001).
In a PCA scatter plot (Figure 5), the first two principal components explained 89.6% of the total variance and revealed two distinct groups, corresponding to I. laevigata and I. pseudacorus.The first principal component (PC1) explained 72.8% of the total variance and contributed to discriminating the species (Figure 5; also see Table S4).Based on their correlation with the PC1 axis, the L, W, and T morphometric parameters were related to I. pseudacorus on the left side.In contrast, the L/W ratio was more significant for I. maackii and I. laevigata, which completely overlapped on the right side.2) and I. maackii, selected as a control.The results of the test are presented as p-values, differences between the mean values of each experimental group (black dots), and 95% confidence intervals of these differences (whiskers) for each pairwise comparison.X-axis is differences in means.Dash-dotted line indicates zero difference.See Table S3 for more details.
No statistically significant differences in the seed width and thickness were found between collection sites 3 and 4 and I. maackii (p > 0.05), while there was a small, approximately 9.4%, but statistically significant (p < 0.001), difference between site 2 and I. maackii.There was no difference in the L/W ratio between sites 2 and 3 and I. maackii (p > 0.05).However, a small, approximately 6.8%, but statistically significant (p < 0.001), difference was found in the L/W ratio between site 4 and I. maackii.For I. pseudacorus, the mean seed width was approximately 37% greater, the seed thickness approximately 33.5% greater, and the L/W ratio approximately 41% smaller than for I. maackii and these differences were statistically significantly different (p < 0.001).
In a PCA scatter plot (Figure 5), the first two principal components explained 89.6% of the total variance and revealed two distinct groups, corresponding to I. laevigata and I. pseudacorus.The first principal component (PC1) explained 72.8% of the total variance and contributed to discriminating the species (Figure 5; also see Table S4).Based on their correlation with the PC1 axis, the L, W, and T morphometric parameters were related to I. pseudacorus on the left side.In contrast, the L/W ratio was more significant for I. maackii and I. laevigata, which completely overlapped on the right side.S4 for more details.

What Is Iris maackii from Northeast Asia According to the Literature?
The taxonomic history of I. maackii began with a single fruiting specimen (Figure 1) collected by Maack from the middle reaches of the Ussuri River, in an area of present-day China, which was originally identified as I. pseudacorus by Regel [2].Maximowicz came to the conclusion that Maack's specimen was not I. pseudacorus in the terms of Linnaeus [58] (p.38), because the leaves lacked the prominent midrib and, thus, the new species I. maackii Maxim.was described [3].In addition, the mention of I. pseudacorus from Siberia by Regel [2] was indirectly related to references [59,60], according to which this species occurred in the vicinity of Selenginsk, Republic of Buryatia, and near Lake Baikal, Russia.It was rightly noted [4,61] that I. pseudacorus did not occur in Siberia or Manchuria, and its mention by Gmelin [59] (p.31, No. 29) referred to I. laevigata, which had been described based on plants from the Baikal region and Dahuria [43].
Komarov noted that I. maackii had never been found in the type locality after Maack [62].Fedtschenko reported that the bract texture and fruit shape of I. maackii were the same as those of I. laevigata, and regarded I. maackii as a synonym of I. laevigata [4].Ivan Shishkin, a florist of the Far Eastern Branch of the USSR Academy of Sciences, made six expeditions in 1927-1929 with the aim to extend the floristic knowledge of the Iman River and surroundings [7].In particular, he came to the conclusion that only I. laevigata, including I. pseudacorus in the terms of Regel [2], and I. maackii in the terms of Komarov [62] could be found in this area.Thus, the name I. maackii was synonymized with I. laevigata [5,[8][9][10].The taxonomic history of I. maackii began with a single fruiting specimen (Figure 1) collected by Maack from the middle reaches of the Ussuri River, in an area of present-day China, which was originally identified as I. pseudacorus by Regel [2].Maximowicz came to the conclusion that Maack's specimen was not I. pseudacorus in the terms of Linnaeus [58] (p.38), because the leaves lacked the prominent midrib and, thus, the new species I. maackii Maxim.was described [3].In addition, the mention of I. pseudacorus from Siberia by Regel [2] was indirectly related to references [59,60], according to which this species occurred in the vicinity of Selenginsk, Republic of Buryatia, and near Lake Baikal, Russia.It was rightly noted [4,61] that I. pseudacorus did not occur in Siberia or Manchuria, and its mention by Gmelin [59] (p.31, No. 29) referred to I. laevigata, which had been described based on plants from the Baikal region and Dahuria [43].
Komarov noted that I. maackii had never been found in the type locality after Maack [62].Fedtschenko reported that the bract texture and fruit shape of I. maackii were the same as those of I. laevigata, and regarded I. maackii as a synonym of I. laevigata [4].Ivan Shishkin, a florist of the Far Eastern Branch of the USSR Academy of Sciences, made six expeditions in 1927-1929 with the aim to extend the floristic knowledge of the Iman River and surroundings [7].In particular, he came to the conclusion that only I. laevigata, including I. pseudacorus in the terms of Regel [2], and I. maackii in the terms of Komarov [62] could be found in this area.Thus, the name I. maackii was synonymized with I. laevigata [5,[8][9][10].
On the other hand, I. maackii was indicated for Sakhalin Island and the Kuril Islands and characterized as a plant with a prominently veined mid-rib and yellow flowers [44,[63][64][65], although these are the diagnostic features of I. pseudacorus.However, most recent authors state that I. maackii from the Russian Far East is I. pseudacorus, which was introduced there and naturalized [11,[66][67][68].Apparently, Russian settlers in this region introduced I. pseudacorus as a bactericidal rather than an ornamental plant.The essential oils from the rhizomes of I. pseudacorus were shown to have antimicrobial activity [69,70].Moreover, this plant can reduce the number of coliform bacteria by 50% and Salmonella by 70% in wastewater [27].In addition, the Far Eastern and European plants of I. pseudacorus have the same chromosome number, 2n = 34 [66] (p.127), and are not different in the noncoding regions of plastid DNA [26].
In addition, I. maackii was indicated to grow in the Liaoning, Heilongjiang, and Jilin provinces of northeastern China [18,25,[71][72][73][74][75].Zhao noted that the characteristics of I. maackii from Liaoning Province are similar to I. pseudacorus, and, therefore, he doubted that I. maackii was a true species [76].The authors of the Flora of China noted that further study was needed to determine whether or not I. maackii is separable from I. pseudacorus [25].Rodionenko suggested that I. maackii from China was the adventive I. pseudacorus [28].Recently, it has been confirmed that the plants from Liaoning Province cited as I. maackii do, in fact, belong to I. pseudacorus, while I. laevigata was not found in the province [77].

What Is Iris maackii vs. I. laevigata and I. pseudacorus According to Morphology?
An examination of the holotype of I. maackii (Figure 1) showed it to be identical to I. laevigata in the rosette leaf width and texture, flowering stem branching, length of the upper lateral shoot, bract texture during fruiting, number of fruit per flowering stem, fruit shape, and size and shape of the seeds (Tables 3 and 4, Figures 1 and 3).In addition, after the characterization of 153 individuals of I. laevigata and I. pseudacorus for 15 morphological characters, it was found that these species could be differentiated from each other, especially in the rosette leaf texture, the flowering stem branching, and the fruit shape (Figure 3 and Table 3).
Among the species studied, the seeds varied in size and shape.By using Dunnett's test, differences in the morphometric parameters of seeds were found between I. maackii and some of the collection sites of I. laevigata (Figure 4).However, these differences were no greater than 10% and were related to the origin of the seeds: the seeds of I. maackii were from a single available locule of the same fruit and, therefore, had similar sizes (Figure 3e), while the seeds of I. laevigata were from different individuals.In the PCA scatter plot, the characters of I. maackii completely overlapped with those of I. laevigata (Figure 5); the characters were taxonomically useful when the overlap was equal to or lower than a threshold of 25% [78].
Previously, it was reported that the seed characters of I. laevigata were almost identical to those of I. pseudacorus [79][80][81].However, that finding is not in concurrence with the present study.An analysis of the morphometric parameters of seeds based on Dunnett's test showed that I. laevigata and I. pseudacorus are distinct (Figure 4).In view of the results of the PCA analysis, the three taxa could unambiguously be separated into two distinct groups with clearly different features in their seed characters (Figure 5).In particular, it was confirmed that the plants listed in references [16,44] as I. maackii from the neighborhood of Shebunino Village, Sakhalin Island, belonged to I. pseudacorus (Figures 4 and 5).Thus, the morphological differences between I. laevigata and I. pseudacorus also include seed characters such as the size and shape.

Conclusions
Since the early 20th century, the taxonomic identity of Iris maackii (Iridaceae) has been unclear, and there have been various speculations as to whether it is an independent species or not.Currently, in most databases, it is regarded as a distinct species native to northeastern China and the Russian Far East [17- 24,34,[36][37][38][39].The present report provides a re-evaluation of the taxonomic identity of I. maackii based on a morphological study.In addition, an overview of the taxonomic history of I. maackii, based on numerous publications of scientists from 1861 to the present time, was conducted to establish its true identity.Since I. maackii is known on the basis of a single specimen (Figure 1), there was difficulty with the availability of material for the morphological comparison of this species with I. laevigata and I. pseudacorus, to which it is associated.However, as a result of a careful examination of the holotype of I. maackii, a total of 20 morphological characters were selected.
As argued in the present contribution, I. maackii is a taxonomic synonym of I. laevigata on the basis of a set of characters, including smoothed rosette leaves, one-branched flowering stems, an elongated shoot of the upper lateral cluster, dry bracts during fruiting, fruit that is smoothed and obtuse at the apex, and mostly oblong, D-shaped seeds.In addition, it is of equal importance that the species-specificity of the seed size and shape can be useful in the taxonomic differentiation of I. laevigata and I. pseudacorus.In order to avoid further confusion, it is here stated that the names I. maackii and I. pseudacorus must never be conflated.The present results confirm that the plants from northeastern China and the Russian Far East (viz.Sakhalin Island, Kuril Islands, and Primorsky Krai) indicated in the literature (e.g., [15,16,44]) and databases [17][18][19][20][21][22][23][24]34,[36][37][38][39] as I. maackii should be considered I. pseudacorus.An important point is that I. pseudacorus is non-native in Northeast Asia and has become highly invasive in natural and artificial waterbodies in the Neotropics, Afrotropics, Neartic, and Australasia.

Figure 4 .
Figure 4. Whisker plots showing the results of Dunnett's test of differences in the mean values of length (a), width (b), thickness (c), and L/W ratio (d) between the Iris laevigata and I. pseudacorus collection sites (see Table2) and I. maackii, selected as a control.The results of the test are presented as p-values, differences between the mean values of each experimental group (black dots), and 95% confidence intervals of these differences (whiskers) for each pairwise comparison.X-axis is differences in means.Dash-dotted line indicates zero difference.See TableS3for more details.

Figure 4 .
Figure 4. Whisker plots showing the results of Dunnett's test of differences in the mean values of length (a), width (b), thickness (c), and L/W ratio (d) between the Iris laevigata and I. pseudacorus collection sites (see Table2) and I. maackii, selected as a control.The results of the test are presented as p-values, differences between the mean values of each experimental group (black dots), and 95% confidence intervals of these differences (whiskers) for each pairwise comparison.X-axis is differences in means.Dash-dotted line indicates zero difference.See TableS3for more details.

Figure 5 .
Figure 5. Principal component analysis of the morphometric parameters of seeds from Iris maackii (red), I. laevigata (blue), and I. pseudacorus (orange).Ellipses show 95% high-density regions for normal distributions representing two groups.Arrows indicate contribution of each morphometric parameter.The codes of the morphological characters of seeds are as follows: L, length; W, width; T, thickness; and L/W, length-to-width ratio.See TableS4for more details.

Figure 5 .
Figure 5. Principal component analysis of the morphometric parameters of seeds from Iris maackii (red), I. laevigata (blue), and I. pseudacorus (orange).Ellipses show 95% high-density regions for normal distributions representing two groups.Arrows indicate contribution of each morphometric parameter.The codes of the morphological characters of seeds are as follows: L, length; W, width; T, thickness; and L/W, length-to-width ratio.See TableS4for more details.

1 .
What Is Iris maackii from Northeast Asia According to the Literature?

Table 2 .
Collection site data for seeds of the Iris species studied.

Table 3 .
Comparative morphology of the Iris species studied.

Table 3 .
Comparative morphology of the Iris species studied.descriptions of the characters are provided in Table1.Data for I. laevigata and I. pseudacorus are presented as the mean (minimum-maximum); see supplementary raw data in TableS1for more details.All measurements are in centimeters. The

Table 4 .
Morphological characters of seeds from the Iris species studied.