Since pollen characters can be used to help distinguish species, our aim was to determine if palynological information has taxonomic significance for Gagea species from Xinjiang, China. Gagea is widely distributed in north temperate and the subtropical zones. The genus has limited taxonomic characteristics and large morphological variation, which results in difficulty of species classification. Pollen morphology of 16 species of this genus was examined comprehensively via light microscope (LM) and scanning electron microscope (SEM). One qualitative and nine quantitative traits of the pollen grains were surveyed, followed by hierarchical cluster analysis (HCA). The pollen grains were bilaterally symmetrical heteropolar monads with a mono-sulcus and they were oblate or peroblate (Polar diameter (P) / Equatorial diameter (E) = 0.36–0.73) in shape and medium to large (P = 17.17–34.64 μm, E = 27.63–81.65 μm) in size. Three types of exine ornamentation were observed: perforate, microreticulate and reticulate cristatum. The HCA divided the 16 species into two groups. This research provides new data on pollen morphology for Gagea (the pollen morphology of eight species was reported for the first time). Pollen morphology also can be used to identify species with similar external morphology, such as G. nigra and G. filiformis. Furthermore, the study of pollen morphology not only provides new data for palynology research on Gagea, but also provides a basis for future classification of this genus.

HCA, pollen morphology, SEM, taxonomy

Pollen morphology is scarcely affected by ecological conditions; thus, it is more stable than external macrographical morphology and has high genetic stability (Zarre and Zarrei 2005; Blackmore 2007; Kadluczka et al. 2022). Pollen size and morphology, such as shape, aperture type, mesocolpium diameter, exine ornamentation and perforation size, have great significance in plant taxonomy and pollen characteristics can be used to infer genetic relationships between ecological groups (Li et al. 2020) or interspecific taxonomic ranks (Heidarian et al. 2021).

Gagea Salisb. (Salisbury 1806) is a genus in tribe Tulipeae of Liliaceae and includes ca. 300 species (Patterson and Givnish 2002; Peruzzi 2016; Peterson et al. 2019). This genus has a wide distribution in regions of South Africa, Asia and Europe (Zarrei et al. 2007; Peterson et al. 2008; Peterson et al. 2019). In Asia, the western Pamir-Alai and western Tien-Shan Mountains are two centres of diversification of Gagea, and south-western Asia is the most probable ancestral area for Gagea. (Peterson et al. 2009; Peterson et al. 2019; Kurbaniyazova et al. 2022). Species circumscription in Gagea is difficult due to overlapping primitive and advanced morphological characters, particularly if only dried specimens are available for study (Zarrei et al. 2009; Tison et al. 2013). To add to the difficulty of identifying species of Gagea, there is enormous variation in vegetative and generative characters at various stages of ontogeny under variable ecological conditions (Levichev 1990, 1999) and polyploidy, hybridisation and convergent evolution make species boundaries unclear (Zarrei et al. 2011).

The initial phylogenetic analysis of Gagea was conducted in 2008 (Peterson et al. 2008), but the phylogenetic relationship between Gagea and Lloydia have been controversial. In accordance with the weight given to morphology and/or phylogeny and various molecular markers, various sections of Gagea have been suggested (Zarrei et al. 2011). Various molecular studies, based on ITS nrDNA and a distinct plastid dataset (rbcL, ndhF, trnL-trnF, psbA-trnH, matK, trnK and rpl16 intron) have divided Gagea into seven sections (Zarrei et al. 2009), 13 sections (Levichev 2011), 14 sections (Peterson et al. 2008, 2016; Peruzzi 2012) or 15 sections (Levichev 2013; Tison et al. 2013). Molecular data reveal that Lloydia is phylogenetically nested within Gagea, forming a monophyletic group (Peterson et al. 2008; Zarrei et al. 2009). Therefore, a revision of Gagea sensu lato (including Lloydia) was proposed (Peruzzi et al. 2008; Zarrei et al. 2011) and recent data on pollen characteristics also supported the placement of Lloydia within Gagea (Hu et al. 2021).

The Flora of China includes 17 species of Gagea (Chen and Turland 2000); however, three new species from Inner Mongolia, three new species and two new records from Xinjiang recently have been described (Zhao and Zhao 2003, 2004; Zhao and Yang 2006; Peterson et al. 2011). Meanwhile, G. nigra was revised, based on morphological characters (Peterson et al. 2011). The Duocet Group (https://duocet.ibiodiversity.net/) moved eight species from Lloydia to Gagea according to APG IV and recognised that Gagea sensu lato. Xinjiang is located in northwest China which includes most of the Tien-Shan Mountains. According to statistics, there are 36 species of Gagea in China, 21 of which are naturally distributed in Xinjiang. These species are separated by morphological characters of the bulbs, leaves, flowers, fruits and seeds (Chen and Turland 2000; Peterson et al. 2011) and molecular data of cpDNA and nrDNA (Peterson et al. 2008, 2016).

At present, the pollen morphology of 46 species in Gagea has been described by various researchers from different regions in the world (Kosenko 1999; Zarre and Zarrei 2005; Wang et al. 2013; Hu et al. 2021; Sezer and Yildiz 2021). These studies demonstrated that shape and exine ornamentation have important information for taxonomic identification amongst species of Gagea. Eight species with reported pollen morphology were distributed in Xinjiang, China (Zarre and Zarrei 2005; Wang et al. 2013; Hu et al. 2021), but their taxonomic significance was not thoroughly investigated.

Through extensive field investigations in Xinjiang, we have found that G. nigra is similar to G. filiformis. The most salient resemblance characters were observed in G. nigra and G. filiformis, such as ovoid-globose bulb, brown or black tunic, leaf and flower number, umbellate or corymbose inflorescence, yellow tepals, capitate stigma, obovoid capsule and red-brown, ovoid-globose seeds. Furthermore, G. jensii shared morphological characters with G. altaica and G. alberti, including ovoid bulbs, taupe tunic, cauline leaves, corymbose or racemose inflorescence, yellow tepals, slightly 3-lobed stigma and brown flat seeds. Hence, the purpose of this research is to: (1) provide palynological information for 16 species of Gagea from Xinjiang, China, using a light microscope (LM) and scanning electron microscope (SEM); (2) distinguish species with similar morphology using palynological characters; and (3) explain the taxonomic significance of palynology in Gagea.

Within populations of each species, pollen shape and exine ornamentation were stable, but the size of pollen grains was not. One-way ANOVA showed that the means of equatorial diameter were not significantly different amongst populations of the same species (p > 0.05). Nevertheless, there was a significant difference in mean equatorial diameter amongst different species (H = 378.016, df = 15, p < 0.001). Therefore, a random population was selected as a representative material for each species for SEM and the micro-morphological characters of pollen grains were carefully observed.

In general, pollen grains of the Gagea species were similar in their morphological characters. The detailed pollen morphology data for the 16 species of Gagea are summarised in Table 2. Representative pollen grains of LM and SEM micrographs are shown in Figs 1–4.

Shape and size

Pollen grains of all species were heteropolar monads and they were mono-sulcus, bilaterally symmetrical and ellipsoidal in polar view. Pollen shape was oblate (Figs 1D, 2J, L) or peroblate (Figs 1A–C, E–H, 2I, K, M, N–P), based on the P/E ranges (Table 2). Pollen size was medium (Figs 3A, D, 4J, L, P) to large (Figs 3B, C, E, F–H, 4I, K, M–O), based on measurements of the equatorial diameter (Table 2).

Figure 1. 

Light microscope micrographs of pollen grains of eight species of Gagea from Xinjiang, China A, A1 Gagea alberti B, B1 G. altaica C, C1 G. angelae D, D1 G. bulbifera E, E1 G. divaricata F, F1 G. fedtschenkoana G, G1 G. filiformis H, H1 G. fragifera. A–H Pollen grain in polar view A1–H1 Pollen grain in equatorial view. Scale bars: 10 μm.

Figure 2. 

Light microscope micrographs of pollen grains of eight species of Gagea from Xinjiang, China. I, I1 Gagea granulosa J, J1 G. jaeschkei K, K1 G. jensii L, L1 G. neopopovii M, M1 G. nigra N, N1 G. kunawurensis O, O1 G. stepposa P, P1 G. tenera. I–P Pollen grain in polar view. I1–P1 Pollen grain in equatorial view. Scale bars: 10 μm.

Figure 3. 

Pollen grains in polar view (A–H) and exine ornamentation (A1–H1) under scanning electron microscope for eight species of Gagea from Xinjiang, China A, A1 Gagea alberti B, B1 G. altaica C, C1 G. angelae D, D1 G. bulbifera E, E1 G. divaricata F, F1 G. fedtschenkoana G, G1 G. filiformis H, H1 G. fragifera.

Figure 4. 

Pollen grains in polar view (I–P) and exine ornamentation (I1–P1) under scanning electron microscope for eight species of Gagea from Xinjiang, China I, I1 Gagea granulosa J, J1 G. jaeschkei K, K1 G. jensii L, L1 G. neopopovii M, M1 G. nigra N, N1 G. kunawurensis O, O1 G. stepposa P, P1 G. tenera.

Hierarchical cluster analysis (HCA)

The palynological groups of the species, based on their relevance, were evaluated by hierarchical cluster analysis. In this analysis, pollen morphology separated the Gagea species into two groups, based on Euclidean distance of 7.01. Group A included G. alberti, G. bulbifera, G. granulosa, G. jaeschkei, G. jensii, G. neopopovii and G. tenera. It was arranged into two subgroups (A1 and A2), based on Euclidean distance of 5.95. Whereas group B involved G. altaica, G. angelae, G. divaricata, G. fedtschenkoana, G. filiformis, G. fragifera, G. kunawurensis, G. nigra and G. stepposa (Fig. 5).

Figure 5. 

Cluster diagram (Ward’s linkage) of 16 species in Gagea from Xinjiang, China, based on one qualitative and nine quantitative pollen characters.

Our results supported previous research showing that the pollen grains in Gagea were bilaterally symmetrical heteropolar monads with a mono-sulcus (Kosenko 1999; Su et al. 2004; Zarre and Zarrei 2005; Wang et al. 2013; Hu et al. 2021; Sezer and Yildiz 2021). Grain exine ornamentation was divergent amongst the species and often served as a diagnostic character (Chung et al. 2010). Additionally, exine ornamentation was one of the most variable characters in Gagea pollen, with three types observed in our research (Table 2). Zarre and Zarrei (2005) recognised four pollen exine ornamentation types within Gagea and G. bulbifera and G. tenera were described as microreticulate and reticulate, respectively and matched type II (microreticulate) and type III (reticulate cristatum) of our research, respectively (Table 2). Pollen exine ornamentation of G. ova (synonym of G. kunawurensis) was described as foveolate (Zarre and Zarrei 2005), differing from type III (reticulate cristatum) found in our research. This result remains to be further confirmed due to lack of photographs in literature.

Pollen morphology of five Gagea species (G. alberti, G. bulbifera, G. fedtschenkoana, G. granulosa and G. nigra) collected from Xinjiang have been described (Wang et al. 2013; Hu et al. 2021). Wang et al. (2013) and Hu et al. (2021) described the exine ornamentation of G. fedtschenkoana pollen as verrucate, whereas we described it as perforate (Table 2). Moreover, the exine ornamentation of G. filiformis and G. granulosa collected from other regions was described as rugulate-perforate and reticulate, respectively (Hu et al. 2021), whereas the exine ornamentation of G. filiformis and G. granulosa described in our research was microreticulate and perforate, respectively (Table 2). Our results are consistent with those in literature when we compare our results with the published figures and the divergence in descriptions might be due to variations in terminology. Overall, the available data provide compelling evidence that pollen morphology exhibits high genetic stability.

Pollen characters have been employed as useful morphological features for the identification of species or genera and they have been applied widely to Allium (Nour et al. 2022), Praxelis (de Abreu et al. 2015), Fritillaria (Samaropoulou et al. 2022) and Lythrum (Vieira et al. 2022) to elucidate intricate taxonomic relationships. Moreover, pollen characters are effective and valuable in distinguishing species with similar morphological features (Ahmad et al. 2022). The pollen of G. nigra and G. filiformis; G. jensii, G. altaica and G. alberti displayed differences in morphological characters that could be used as identification tools, while the applicability of other morphological characters, such as shape of bulb, leaves, inflorescence, stigma and seeds, colour of tunic, tepals and seeds, number of leaves and flowers, for species identification is restricted. Pollen grains of the above five species were large and peroblate and only the pollen grains of G. alberti were medium in size (Table 2). In addition, the aperture type, porus size and pollen wall thickness of these species were similar between species. Gagea nigra pollen had reticulate cristatum exine ornamentation, whereas that of G. filiformis was microreticulate (Figs 3G1, 4M1). G. jensii and G. alberti pollen had perforate exine ornamentation, whereas that of G. altaica was microreticulate (Figs 3A1, B1, 4K1). Despite G. jensii and G. alberti pollen grains having similar exine ornamentation, they differed in size. (Table 2). Pollen size should be evaluated with circumspection, because it is known to be influenced by biological factors (Rahmawati et al. 2019).

Information on palynology should be used to provide new insight into the differences (or not) between species. For example, in the revision of G. nigra by Peterson et al. (2011) that also includes a description of three new species (including G. jensii and a comparison of the morphology of G. jensii and G. alberti), the morphology of G. jensii differs from that of G. alberti. However, we found that the pollen morphology of G. jensii and G. alberti is the same (Figs 3A1, 4K1), suggesting that they are possibly the same species. Thus, we suggest that verification of G. jensii and G. alberti as separate species must be further investigated using cytological, molecular phylogenetic or anatomical techniques. Our research provides a new way to diagnosis species with similar morphological characteristics, but they are not exhaustive.

According to recent infrageneric classification of Gagea, the species in this research belong to seven sections (Table 1) (Peterson et al. 2016). HCA was performed on the palynological data obtained from both LM and SEM. Group A included two subgroups (A1, A2) and contained three species of sect. Plecostigma (G. alberti, G. jensii and G. neopopovii), two species of sect. Bulbiferae (G. bulbifera and G. jaeschkei) and one species of sect. Minimae (G. granulosa), sect. Didymobolbos (G. tenera). The subgroup A1 enclosed four species with perforate exine ornamentation, while the subgroup A2 involved three species with medium size (Fig. 5, Table 2). Group B, which represents the large-sized and oblate-shaped pollen grains, encompassed the following sections: sect. Plecostigma (G. altaica), sect. Bulbiferae (G. stepposa), sect. Didymobulbos (G. fragifera), sect. Dschungaricae (G. kunawurensis), sect. Gagea (G. angelae and G. fedtschenkoana), sect. Minimae (G. filiformis and G. nigra) and sect. Platyspermum (G. divaricata) (Fig. 5, Table 2). Thus, the results of the HCA are equivocal as the subgroups contain species from both the same and different sections.

Gagea nigra and G. filiformis belong to the sect. Minimae, and morphology and molecular evidence suggest that G. nigra is an independent species separated from G. filiformis (Peterson et al. 2011). We have also confirmed this view through palynology results, based on two species clustered into one group (Fig. 5). In contrast, G. altaica, G. jensii and G. alberti belong to the sect. Plecostigma. However, G. altaica separated into a different group with G. jensii and G. alberti, based on palynology (Fig. 5). This may indicate that there are pollen synapomorphies related to the systematics in certain sections, but not necessarily in all sections.

Although the results are not able to offer a diagnostic key amongst groups of Gagea taxa in Xinjiang, China, they demonstrate that palynology may aid in the taxonomy of the genus by differentiating between taxa within their groups.

Pollen morphology of eight species of Gagea from Xinjiang, China was reported for the first time in our research. Gagea pollen grains were heterogenous in shape, size and exine ornamentation. Pollen characters have a certain taxonomic effect on the interspecies of Gagea, but the taxonomic relationship cannot be fully clarified only by pollen morphology. The results of the current research have provided palynological data for the classification of Gagea and also contribute to future classification of this genus.

We thank Dr Carol C. Baskin (University of Kentucky, Lexington, Kentucky, USA) for correcting the English and we are grateful to Jiancai Chi for collecting G. angelae. We also thank the Xinjiang Institute of Ecology and Geography Chinese Academy of Sciences Herbarium (XJBI) and Yang’s Herbarium for providing specimens.

This research was supported by the Natural Science Foundation of Xinjiang Uygur Autonomous Region, China (2022D01A71), the Grant of Innovation Environment Construction of the Xinjiang Uygur Autonomous Region, China (PT2224), the Innovative Team Foundation of Biology of Xinjiang Agricultural University, China (ITB202103) and the Xinjiang Agricultural University Graduate Research and Innovation Project (XJAUGRI2022042).