This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.
INTRODUCTION
There are two genera of the Balsaminaceae, and one is Impatiens. This genus is rich in germplasm resources, with more than 1 000 species in the world and more than 260 species in China (Luo et al., 2022). However, the classification of Impatiens is very challenging because it is difficult to make and preserve specimens (Yu et al., 2016). The primary classification methods of Impatiens are traditional taxonomy and molecular biology or a combination of both. In molecular biological classification, ITS sequences (Eddie et al., 2003; Yuan et al., 2004) and the chloroplast genomes (Luo et al., 2022) are commonly used for classification. Compared with molecular biology, traditional taxonomy primarily classifies plants based on morphology and palynology. The classification of plants based on pollen characteristics is considered palynology. Plants of the same genus or different genera are crossed by self-pollination or through the actions of insects (Attique et al., 2022), wind and other factors (cross-pollination) (Zavada and Hackley, 2022). This carries the genetic material of the male parent, which is a necessary condition for plants to continue their offspring. Thus, during the process of plant evolution, the pollen structure is complex to protect the genetic material (Zavialova and Nosova, 2019). To adapt to the environment, plants have formed a special pollen structure, which helps preserve the pollen. There are some differences in pollen characteristics among different plants, which is the basis of pollen classification.
Optical microscopy and scanning electron microscopy (SEM) are the primary tools to observe pollen morphology, which greatly aid the development of palynology. Researchers promoted the development of plant taxonomy by extracting pollen characteristics from families such as Campanulaceae (Khansari et al., 2012), Combretaceae (El Ghazali, 2022) and Paropsieae (Mezzonato-Pires et al., 2022). The study of plant pollen helps determine that pollen characteristics can be used as a basis for plant classification (Tuler et al., 2017; Umber et al., 2022). The importance of plant classification and pollen characteristics at the intergeneric level has been emphasised, and it has been shown that pollen data can verify the information of pollen evolution on the true stamen pedigree. In addition, it provides some explanations for the direction of plant evolution.
Among Impatiens plants, Yu et al. (2016) made an accurate analysis by creating two subgenera of Impatiens and dividing them into seven groups. Pollen micromorphology was an important index for this classification of Impatiens. Studies on the pollen morphological characteristics of Impatiens by SEM helps identify Impatiens, provide basic pollen information for breeding and improve the rate of success of breeding.
MATERIALS AND METHODS
Information collection and preservation methods of plant materials
The plant materials were primarily derived from field studies and sampling, and the collected information was recorded (Annex 1). The change in height of flower colour among the species of Impatiens increases the difficulty of classifying them (Chen, 1978). In this study, we used 1, 2, 3 and higher numbers when necessary to mark Impatiens of the same species but different colours. In the same type of Impatiens, more than two types of mature pollens were collected and stored in a test tube or collection tube filled with FAA fixatives.
Scanning electron microscope
The pollen of 35 species of Impatiens was observed by SEM. The pollen samples were prepared by first soaking the material and then gently removing the anthers with tweezers and dissection needles. The anthers were placed in a 2 mL collection tube and dehydrated with an alcohol gradient of 30%, 45%, 60%, 75%, 90% and 95%. The solution was replaced with anhydrous alcohol twice for 1 h at a time. The anthers were then removed with tweezers and dissected using a needle, so that the pollen was evenly distributed on the sample table with conductive glue. Under the vacuum conditions of an ion sputtering instrument (Cressington Scientific Instruments, Watford, UK), the gold spray was coated for 2–3 min. It was observed by a Zeiss scanning electron microscope (Zeiss, Jena, Germany).
Image analysis
The pictures were analysed using Image J (NIH, Bethesda, MD, USA), and the morphological characteristics were named based on the pollen research literature (Yu et al., 2016; Mazari et al., 2017; Hu et al., 2020; Raza et al., 2020). There were slight modifications such as pillow shape, which was named so because its shape is similar to that of ancient Chinese porcelain. When the mesh density (MD) was 12 × 1 000-fold magnification, the number of mesh was regarded as the MD.
Data statistics and analysis
The statistical data of Microsoft Excel YEAR (Redmond, WA, USA) were used, and SPSS 26.0 (IBM, Inc., Armonk, NY, USA) was used for correlation, principal component and system clustering analyses (Pérez-Gutiérrez et al., 2015; Zafar et al., 2022). The number of sepals (NS), which is one of the important bases for the classification of Impatiens, was added as a reference index to compare the importance of pollen characteristics during the analysis of indices.
RESULTS
Statistics of pollen characteristics
Based on the study and analysis of pollen micromorphology of 35 species of Impatiens, the data of 19 characteristics indexes were obtained, including pollen morphology (Table 1), pollen germination pore characteristics, pollen size and mesh characteristics (the data sheet is provided in Annex 2). The pollen characteristics were primarily divided into two parts, qualitative and quantitative traits.
Statistics of pollen micromorphological characteristics of Impatiens in the Yunnan–Guizhou Plateau.
N
Species
Epidermal ornamentation
Shape
PV
EV
1
I. siculifer
Reticulate
Rectangular circle
Oval
2
I. pinetorum
Reticulate
Pillow
Oval
3
I. rectangula
Reticulate
Rectangular circle
Oval
4
I. ruiliensis
Reticulate
Rectangular circle
Oval
5
I. holocentra
Reticulate
Rectangular circle
Oval
6
I. siculifer var.
Reticulate
Rectangular circle
Oval
7
I. racemosa
Reticulate
Rectangular circle
Oval
8
I. uliginosa1
Reticulate
Rectangular circle
Oval
9
I. uliginosa2
Reticulate
Rectangular circle
Oval
10
I. uliginosa3
Reticulate
Rectangular circle
Oval
11
I. uliginosa4
Reticulate
Rectangular circle
Oval
12
I. cyathiflora
Reticulate
Rectangular circle
Oval
13
I. dicentra
Reticulate
Rectangular circle
Oval
14
I. dicentra var.
Reticulate
Oval
Oval
15
I. noli-tangere
Reticulate
Rectangular circle
Oval
16
I. corchorifolia
Reticulate
Rectangular circle
Oval
17
I. delavayi
Reticulate
Rectangular circle
Oval
18
I. guizhouensis
Reticulate
Triangular circle
Triangular circle
19
I. auriculata
Corrugated reticulation
Triangular circle
Triangular circle
20
I. monticola var.
Reticulate
Rectangular circle
Oval
21
I. chlorosepala
Reticulate
Oval
Oval
22
I. xanthina
Reticulate
Rectangular circle
Near circle
23
I. monticola
Reticulate
Rectangular circle
Oval
24
I. yui
Reticulate spinous granule
Rectangular circle
Oval
25
I. rubrostriata
Reticulate
Rectangular circle
Oval
26
I. loulanensis
Reticulate
Oval
Oval
27
I. arguta1
Reticulate
Rectangular circle
Oval
28
I. arguta2
Reticulate
Rectangular circle
Oval
29
I. arguta3
Reticulate
Rectangular circle
Oval
30
I. fanjinica
Reticulate
Pillow
Oval
31
I. fanjinica var.
Reticulate
Pillow
Oval
32
I. reptans
Reticulate
Pillow
Oval
33
I. reptans var.
Reticulate
Pillow
Oval
34
I. pianmaensis
Reticulate
Rectangular circle
Oval
35
I. pianmaensis var.
Reticulate
Rectangular circle
Oval
EV, equatorial view.
There were six indices of qualitative characteristics (Figure 1), which included epidermis ornamentation (EO), polar view (PV), equatorial view (EV), characteristics of germination pores (CGP), mesh feature (MF) and ridge feature (RF). The characteristics under each index were different, and the proportion was also different, indicating that the characteristics of pollen qualitative indices were rich, which aided the interspecific classification of Impatiens plants (Figure 2). There were 14 quantitative characteristic indices. The NS was the reference group, and the others included the number of germination pores (NGP), germination pore length (GPL), germination pore width (GPW), germination pore length/germination pore width (L/W), MD, ridge width (RW), polar axis length (P), equatorial major axis length (E1), equatorial short axis length (E2), pollen volume (V) and P/E1, P/E2 and E1/E2. There were differences among the species, and there was an obvious difference between the maximum and minimum values (Table 2), such as volume. The maximum value was more than 14 000 μm3, the minimum value was approximately 4 000 μm3 and the average value was approximately 8 000 μm3. Based on the aforementioned analysis described, the pollen characteristics of Impatiens were rich in diversity, which can provide positive conditions to classify the members of this genus.
Analysis on the characteristics of pollen quantitative index of Impatiens.
There was some correlation among the pollen characteristics (Figure 3). Among the 19 traits of pollen, there were significant correlations among some indices, such as the SN and polar morphology, equatorial morphology, number of germination holes, MD, RW, polar axis length, equatorial minor axis length, P/E1, P/E2 and E1/E2. With reference to the NS, there was some correlation between pollen characteristics, and there was also some correlation with the NS. I. guizhouensis and I. auriculata were used as examples to show that the correlation was significant for plant classification. In the classification index, there was some relationship between the index and the index. The correlation showed that the pollen characteristics of Impatiens plants may be accompanied, and as the basis of plant identification, pollen characteristics can promote the acquisition of species information.
Principal component analysis and systematic cluster analysis
Principal component analysis
According to the principal component analysis of 20 characteristic indexes (Table 3), there were 5 main components, and the contribution rate of each component was more than 10%. The cumulative contribution rate of 5 components is 80.45% > 60%, and all characteristic indexes were included. It showed that the selected current index was of significance to the clustering results among the cluster members. The results showed that the selected characteristic indexes were of significance to the classification of Impatiens. The next step of systematic cluster analysis can be carried out. The composition matrix showed that the effect of each index on the clustering results of Impatiens was P, P/E1, E2, equatorial morphology, E1/E2, the number of germination holes and so on. The mesh condition and the width of germination holes had the least effect on the classification of Impatiens. Among the 19 traits of pollen, 11 had more taxonomic effect than the NS. Sepal was an important classification index of Impatiens, so these 11 pollen characteristics played an important role in the classification of Impatiens.
Principal component analysis table of pollen micromorphological characteristics of Impatiens.
In the clustering results of the system (Figure 4), with the red dotted line as the reference line, the system clustering had three branches, and the yellow reference line analysis had six branches. I. guizhouensis and I. auriculata were the first branch, and both types of pollens were the three-groove type. These pollens were shaped as triangular circles. They were all typical plants of the subgenus Impatiens and were reasonably classified together. However, the yellow reference line analysis indicated that they should clearly be separated because of their significant differences in the quantitative data of pollen EO, L/W and pollen size. I. yui remained a separate branch in which no reference line was used. Its EO and ridge characteristics were unique, but its quantitative index also differed significantly from that of other species.
The other species were uniformly clustered into one branch (third branch) under the red reference line, and the clustering distance was short, indicating that there was only a small difference among the species. With reference to the yellow dotted line, the third branch under the red reference line was further divided into three more careful branches, namely the fourth, fifth and sixth branches that contained 5, 4 and 23 species, respectively, and the varieties or same species of Impatiens were all divided together. On closer inspection, the Racemosae group was in the same branch (a small branch of the sixth branch), while the Fasciculatae group and the undetermined group of Impatiens gathered in another small branch. The clustering results of qualitative and quantitative traits were highly similar to those of the full characteristics (Figure 3). In summary, 35 species of Impatiens were systematically clustered with 19 pollen characteristics, and the results were satisfactory.
Impatiens grouping analysis
Based on the quantitative data analysis between groups of Impatiens (Figure 5; the data sheet is shown in Annex 3), no significant differences in the quantitative index of germination pore between groups were identified, which further verified that the quantitative index of the germination hole in the composition matrix played a low role in the classification of Impatiens. There were some differences in other quantitative indices among the groups, but there was no difference in the quantitative characteristics of pollen in some Impatiens groups, such as the Impatiens and Racemosae. The combination of Figure 4 showed that the pollen characteristics of Impatiens had some significance in the grouping of Impatiens.
The situation of the Impatiens, Uniflorae and Fasciculatae groups was relatively clear, but there was a relatively large difference between the two types of pollen in the Scorpioidae group owing to the unique pollen characteristics of I. yui. The similarity of the Clavicarpa was low. Interestingly, there was also a high degree of similarity among the six species of Impatiens that were not grouped. The grouping analysis of the similarity between Impatiens further determined the classification between them, verified the results of systematic clustering and supported the superiority of Impatiens grouping.
DISCUSSION
Impatiens germplasm resources
Impatiens is rich in germplasm resources, and there are approximately 120 species in the Yunnan–Guizhou Plateau (Luo et al., 2022). In this study of some areas of Yunnan and Guizhou, 48 species of Impatiens were collected, which is, of course, only a fraction of the local species. During the process of collection, the preliminary investigation of Impatiens shows that the there was a high interspecific coefficient of variation of Impatiens, particularly in the variation of flower colour. Three species of Lycos’s with different flower colours, and five species of Lycos’s with different flower colours were found. Their names were marked with 1, 2, 3 and 4 to distinguish them. The variation in height of Impatiens has been confirmed for a long time (Chen, 1978).
Pollen characteristics of Impatiens
We observed the pollen of Impatiens (Figure 1 and Figure 2). The results showed that the pollen of Impatiens was primarily divided into three-groove and four-groove types, and there was a significant correlation between the number of pollen germination pores and its morphology (Figure 4). The pollen characteristics included four pollen germination pores of rectangular and oval shapes and three of triangular circles. Interestingly, Impatiens with three-groove pollen had four sepals, and Impatiens with four-groove pollen had two or four sepals. This phenomenon was used as the difference between primitive and more evolved species of Impatiens plants. As shown, the pollen of the original species of Impatiens was the three-groove type with four sepals; the over-species of Impatiens pollen was the four-groove type with four sepals, and the evolutionary species of Impatiens pollen included the four-groove types with two sepals (Yu, 2012; Yu et al., 2016). The EO of pollen was the reticulate type (Janssens et al., 2012), and the MD of evolutionary species was higher than that of the original species. The addition of reticulation on the pollen surface could contribute to plant pollination and fertilisation and protect the genetic material in plant pollen. The reticulate pattern increases the structural strength and surface friction coefficient of pollen, which helps them interact with insects and increases the ability of insects to carry the pollen (Attique et al., 2022). In our classification based on pollen quantitative characteristics, we found that polar axis length, equatorial axis length and pollen size played an important role in the classification of Impatiens, which was similar to the results of most plant pollen studies (Zafar et al., 2022).
The effect of pollen characteristics on the classification of Impatiens
The PCA showed that 11 of the 19 pollen characteristics were more effective than the NS in the classification of Impatiens (Figure 2 and Table 2), indicating that the pollen characteristics of Impatiens play an important role in the classification of Impatiens. Pollen characteristics played an important role in plant classification during the research and development of palynology and have been applied to the classification of many plants (Khansari et al., 2012; Pérez-Gutiérrez et al., 2015; Ullah et al., 2022). During the process of interspecific clustering (Figure 4), the results of pollen classification of Impatiens were satisfactory, but the difference between Impatiens that was highly similar was too low to subdivide some Impatiens that only differed slightly. During the study of pollen morphology, it was also explained that the characteristics of pollen structure can support the classification between genera and higher levels of taxonomy (Umber et al., 2022). We used pollen characteristics as the basis of interspecific classification, and the results showed that, to some extent, pollen characteristics supported the interspecific classification of Impatiens. In summary, the pollen of Impatiens had some positive effect on the classification of Impatiens, whether qualitative or quantitative. However, the identification of specific species of Impatiens also required the support of other characteristics (Lu, 1991). This also indirectly showed that there was a small difference among some species of Impatiens, and the coefficient of variation of Impatiens pollen was not as high as those of other morphological indicators (Lens et al., 2005). This has important reference value for the study of ancient plants and plant evolution (Lens et al., 2012).
Analysis of the effect of Impatiens pollen on grouping
Based on the Impatiens grouping system proposed by Yu et al. (2016), the interspecific similarity (Figure 6) and grouping of Impatiens were analysed using pollen morphological characteristics and quantitative data (Figure 7). In this study, the materials that were collected involved two subgenera. There were 33 species of the Impatiens subgenus, which can be divided into six groups. In addition to these groups, six species of Impatiens did not belong to any group. They were designated undetermined groups. In the grouping, the Racemosae group had the most materials with 12 species. According to the interspecific similarity and the grouping of Impatiens, all the plants of the subgenus Clavicarpa were found to be the three-furrow type with a triangular circle. The pollens of Impatiens, Racemosae and Fasciculatae groups were all rectangular or oval. In the Scorpioidae group, there was low similarity between the two species of Impatiens, and they also differed greatly in terms of their shape and size. Among the 35 species of Impatiens, I. yui was the only one with a pollen epidermis. Interestingly, we had not determined that the basic situation of the pollen of the six species of Impatiens was highly consistent. The pollen morphology was pillow-shaped, and it was possible that there was some evolutionary relationship between them. Compared with the other groups, the pollen P/E1 of Impatiens in the Uniflorae group was smaller, and the pollen looked slender. According to the analysis of the grouping form of Impatiens, pollen has been demonstrated to play some role in the grouping system proposed by Yu et al. (2016), which supported the classification of the two subgenera. However, there were some differences in pollen morphology among the seven groups of Impatiens, and the classification system of Impatiens may require further refinement. This is particularly true for the Scorpioidae group.
Analysis of the classification results of Impatiens by pollen
The clustering results of 35 species of Impatiens showed that compared with the results of molecular markers, the clustering results of the original species were partially consistent with those of molecular markers, and the clustering distance of the same class of Impatiens was very short, or directly clustered together. This shows that the pollen morphology is similar to that of molecular markers or other morphological markers in interspecific classification (Yu et al., 2016). With the continuous improvement in the molecular marker technology, increasing numbers of researchers utilised the chloroplast genome (Janssens et al., 2006; Luo et al., 2022) or a gene fragment (Eddie et al., 2003; Swenson et al., 2007; Wang et al., 2014) to study the classification of Impatiens. The species of Impatiens is rich, and it was difficult to make and preserve specimens. The preservation of specimens greatly improved the classification of Impatiens. However, there was only a continuous breakthrough when traditional taxonomy was used, and the classification index was continuously refined. The combination of these taxonomic studies with those that utilised molecular markers and other technology overcame the classification problem of Impatiens and built a more detailed classification system of balsam.
The traditional taxonomy of Impatiens not only includes the observation of pollen by SEM but also involved the classification and analysis of Impatiens using micromorphological features, such as the leaf epidermis (Cai, 2007) and seeds (Martínez-Ortega and Rico, 2001; Janssens et al., 2009; Dadandi and Yildiz, 2015). This was a good way to combine macro- and micromorphologies (Yu et al., 2016) to find a more detailed classification system of Impatiens, but the refinement and integration of classification indicators should be more flawless.
CONCLUSIONS
The pollen of 35 species of Impatiens was studied. The analysis of pollen characteristics indicated that there was some correlation among the traits. A total of 19 pollen characteristics, except for sepals, were counted, and the qualitative characteristics showed that the pollen of Impatiens appeared rich. The comparison of quantitative characteristics showed differences among the Impatiens species. A PCA with sepals as the reference showed that the characteristic indices of Impatiens were highly important for taxonomy. In contrast, the order of taxonomic effect on Impatiens was P, P/E1, E2, EV, E1/E2, number of germinating pores, volume, reticulate ridge characteristics, PV, RW and epidermis decoration. The results of the classification of Impatiens based on pollen characteristics were reliable, and they showed that the pollen characteristics of Impatiens species were significant to the interspecific classification of Impatiens. However, the analysis of interspecific similarity showed that relying on the characteristics of pollen was not enough to support the clear classification of Impatiens. Under the condition of Impatiens grouping, it was concluded that the new Impatiens classification system was reliable, but there were some shortcomings. Thus, it was necessary to further improve the classification system. In summary, pollen micromorphology played a positive role in the classification of Impatiens, but it also had some deficiencies.