Bot. Bull. Acad. Sin. (2005) 46: 355-361

LI � The cytogeography of Aster ageratoides var. laticorymbus

The cytogeography of Aster ageratoides var. laticorymbus (Asteraceae), a polyploid complex endemic to China

Wei-Ping LI*

College of Life Sciences, Hunan Normal University, Changsha 410081, P.R. China

(Received July 2, 2004; Accepted June 2, 2005)

Abstract. Aster ageratoides var. laticorymbus is distrubuted from southwestern China to eastern China. Population sampling across its whole distribution area was made in its twenty-six populations. Chromosome numbers of all the populations were investigated, and six populations, representative of various ploidy levels, were analyzed karyotypically for the first time. The results show that this variety is a polyploid complex comprised of 2x, 4x and 6x. The populations are mostly hexaploid (2n = 6x = 54) and occupy an extensive area from southwestern China to eastern China while the diploid (2n = 2x = 18) and the tetraploid (2n = 4x = 36) are less frequent and limited to narrower regions of the Yunnan-Guizhou Plateau and the transitional belts from the plateau to adjacent areas. Based on the cytogeographical distribution of the complex, the analysis of its karyotypes, its morphological variations related to geography and a previous report, some hypothesis are made: (1) the polyploids of the complex might be autopolyploids, and the tetraploid might have originated independently twice; (2) in terms of some karyotypical parameters, different ploidy levels of the complex evolved at different speeds; (3) the Yunnan-Guizhou Plateau is the diversity center and the origin center of the complex, and the place from which the hexaploid dispersed eastwards to eastern China; (4) the distribution pattern that the hexaploid is dominating might have been formed by the competitive exclusion and its high capability to disperse and occupy new habits; and (5) the variety is a relative young complex.

Keywords: Aster ageratoides var. laticorymbus; Dispersal route; Distribution pattern; Karyotype; Polyploid complex; Yunnan-Guizhou Plateau.

Introduction

Aster ageratoides Turcz., a perennial herb of Asteraceae (Compositae), is widely distributed from northeastern Asia to southeastern Asia, but with its major diversity center in China (Ling and Chen, 1985; Li, 2002). This species has a complex taxonomic history and is reputed to be a difficult taxonomical subject (Huziwara, 1957; Ling and Chen, 1985; Ito and Soejima, 1995; Soejima and Peng, 1998; Soejima et al., 1999; Li, 2002). According to Ling and Chen (1985), it comprises 11 varieties in China. However, the taxonomic status of these taxa is quite controversial, and some are very difficult to distinguish from each other (Ling and Chen, 1985; Soejima et al., 1999; Li, 2002). The taxonomic confusion in this difficult species comes from the fact that very little is known of its diversity and variation in morphology and cytology (Soejima et al., 1999; Li, 2002).

Aster ageratoides var. latiocorymbus (Vant.) Hand.-Mazz. is endemic to China (Handel-Mazzetti, 1938; Ling and Chen, 1985; Li, 2002). It is characterized by poly-branched stem, oblong-lanceolate or ovate-lanceolate middle leaves, narrow phyllaries with green top, and white ray florets, the narrow phyllaries being the most important diagnostic character (Handel-Mazzetti, 1938; Ling and Chen, 1985; Soejima et al., 1999). This variety is of eco

logical importance due to its participation in the revegetation of degraded or newly open habitats (Li, 2002). Chen et al. (1992a, 1992b) reported for the first time that the variety laticorymbus has two cytotypes (4x = 36; 6x = 54) in Hunan Province and Guangxi Province. Unfortunately, mistakes in identifying specimens impaired the excellence of their work (Li, 2002). The only other report of cytology of the variety was from Soejima et al. (1999), who documented three cytotypes (2x = 18; 4x = 36; 8x = 71, 72) of this variety in the Zhaotong area, northeastern Yunnan.

While working on biosystematic studies on Aster ageratoides, the author made extensive field observations and sampling of 73 native populations in 20 provinces of China between 2000 and 2002. As a part of the work, A. ageratoides var. laticorymbus was investigated in cytogeography and morphology. This study reports the current distribution pattern of three cytotypes of var. laticorymbus, and then discusses its evolution of karyotype, its origin, dispersal route, the formation mechanism of current distribution pattern, and the age of the A. ageratoides var. laticorymbus complex.

Materials and Methods

Field observations and sampling were made in 73 populations of Aster ageratoides, of which 26 populations (Table 1; Figure 1) from 11 provinces belonged to A. ageratoides var. laticorymbus according to Li (2002). For each population, more than ten plants were transplanted to

*Corresponding author. E-mail: lwp@hunnu.edu.cn; Tel: 86-731-8871052; Fax: 86-731-8883310.

Botanical Bulletin of Academia Sinica, Vol. 46, 2005

Hunan Normal University for cytological investigations, and the specimen were prepared and deposited as vouchers in the Herbarium of Hunan Normal University (HNNU).

Actively growing root tips were cut for chromosome observations from the living plants, and pretreated with 0.1% colchicine at 8-12�C for 4 h before being fixed in Carnoy I (glacial acetic acid: 95% ethanol = 1:3) at room temperature for 12 h. They were then macerated in 1 mol/L hydrochloric acid at 60�C for 8 min, stained in 5% NH₄Fe(SO₄)₂, an intermedia, at room temperature for 3 h, stained in 0.75% hematoxylin for 3 h, washed in distilled water for 30 min, and finally depigmented and squashed in 45% acetic acid. The chromosome number of each population was determined by counting the chromosomes in at least 10 metaphases of every sample. Six populations, representative of different ploidy levels, were analyzed karyotypically, with the karyotype parameters of each population being the average values of ten plants.

The karyotype parameters used in this study included: (1) index of the karyotype asymmetry, that is, As.k (%) (As.k = the length of all long arms / the length of all chromosomes � 100); (2) ratio of the longest chromosome to the shortest; (3) average arm radio for each population.

The symbols used to describe the karyotype followed Levan et al. (1964): m = median-centromeric chromosome (arm ratio: 1-1.70), sm = submedian-centromeric chromosome (arm radio: 1.71-3.00). The karyotype symmetry was classified according to Stebbins (1971).

Figure 1. Map showing distribution of 26 sampling populations of the Aster ageratoides var. laticorymbus complex. Codes for populations correspond to those in Table 1. Triangles (�) represent diploid populations, squares (�) tetraploid populations, and solid circles (�) hexaploid populations.

LI � The cytogeography of Aster ageratoides var. laticorymbus

106�41' E to the eastern end of the Chinese mainland. Note that elevations decline from the western portion of the range to the eastern (Table 1).

For the six populations (P1-P3, P12, P16, and P20) investigated karyotypically, the micrographs of somatic metaphase chromosomes, karyotypes, and the karyotype parameters are presented in Figure 2 and Table 2. The results show that the metaphase chromosomes vary gradually in length from the longest to the shortest without showing distinct bimodality, and the frequencies of me

Results

Presented in Figure 1 and Table 1 are the chromosome numbers of twenty-six populations, which can be divided into three cytotypes: diploid (two populations: P2 and P3) with 2n = 18, tetraploid (five populations: P1 and P11-P14) with 2n = 36, hexaploid (nineteen populations: P4-P10 and P15-P26) with 2n = 54. The diploid was found in the west and the northwest of the distribution area, the tetraploid in the northwest and the middle, and the hexaploid from

Figure 2. Micrographs of somatic metaphase chromosomes and the karyotypes of Aster ageratoides var. laticorymbus (Scale bar = 4 mm). A, C. P2; B, D. P3; E, G. P1; F, H. P12; I, K. P16; J, L. P20 (Codes for populations correspond to those in Table 1).

Botanical Bulletin of Academia Sinica, Vol. 46, 2005

dian-centromeric chromosomes are high, up to 88.9-100%. The karyotypes of all these six populations are categorized as Stebbin's 1A type because the ratios of the longest and the shortest chromosome are below 2, and no arm ratios are more than 2. Two populations of the same cytotype are largely identified in chromosome parameters while different ploidy levels are significantly different in two parameters (Lt/St and A.A.R) (Table 2). Satellites were found at the terminal region of the short arms of the first pairs only in the two diploid populations (P2, P3) (Figure 2A-D).

Discussion

Karyotype Evolution of the Complex

Since the chromosome base number of Aster s.s. is consistently x = 9 (Nesom, 1994; Li and Zhang, 2004), the results reveal that Aster ageratoides var. laticorymbus possesses three cytotypes: 2x (18), 4x (36), 6x (54) (Figure 1, 2; Table 1), suggesting that in the variety there is a polyploid series.

Although two karyotypically studied populations of each ploidy level have the same or similar karyotye formula and karyotype parameters (Table 2), there exist significant morphological differences between two diploid populations (P 2 vs P3) and between two kinds of tetraploid populations (P1 vs P11-14). Especially, P2 (2x) is quite similar morphologically to P1 (4x) and P3 (2x) to P11-14 (4x). P2 and P1 have lanceolate leaves and relatively small heads with no more than 20 florets (including ray florets and disc florets) while the other populations studied here, including P3, P11-14 as well as all hexaploid populations, possess elliptic or oblong leaves and relatively large heads with usually more than 20 florets. Therefore, it is possible that these tetraploid and hexaploid are autopolyploids derived from ancestors with the same gross morphology but lower ploidy levels. In general, an autopolyploid is defined as a cytotype containing three or more sets of homologous chromosomes. However, neither in the tetraploid nor in the hexaploid, were four sets or six sets of morphologically similar chromosome complements observed, so the karyotype data do not support the hypothesis that they are of autopolyploid origin. This may be attributed to karyotype differentiation among populations of various ploidy levels, especially extensive and rapid chromosomal

restructuring and chromosomal diploidization after autopolyploidization (Soltis and Soltis, 1999; Yang, 2002) instead of an allopolyploid origin.

Soejima et al. (1999) studied var. laticorymbus in the Zhaotong area, which belongs to north the Yunnan-Guizhou Plateau and is near to P1 and P2. They found no significant morphological differences among the diploid, the tetraploid, or the octaploid of variety laticorymbus with narrower leaves and smaller heads (Soejima et al., 1999), similar to those of P1 and P2. Therefore, in the north Yunnan-Guizhou Plateau and the adjacent areas, tetraploid populations may originate from some diploid population like P2 whereas the other populations studied here, except P1 and P2, share the same morphological characters and may be on the same evolutionary lineage. If this is true, the tetraploid may have originated independently twice, and karyotype similarity between the two diploid populations and between the two tetraploid populations may have resulted from parallel evolution.

According to Table 2, diploid and polyploid demonstrate different evolutionary speeds in two karyotype parameters, Lt/St and A.A.R., which are directly proportional to karyotype asymmetry. Generally, the evolution of karyotype in the genus Aster is towards increasing asymmetry (Hong, 1990). The ratios of the longest chromosome to the shortest (Lt/St) are 1.63 and 1.57, respectively, in the two diploid populations (P2 and P3) while the values are 1.73 and 1.74 in the tetraploid, and up to 1.85 and 1.92 in the hexaploid, showing polyploidization may facilitate size differentiation among non-homologous chromosomes of a genome. In contrast, the average arm ratio is 1.38 in the two diploid populations, and it is 1.25 and 1.26 in the two tetraploid populations and 1.29 and 1.33 in the two hexaploid populations, showing higher length differentiation between the two arms of a chromosome in the diploid populations than in the polyploid populations (Table 2). However, no significant difference appears between the diploid and the tetraploid in As.k (%) or between the Ls/Lt values of Mikania micrantha H. B. K. (Asteraceae) and the Ls/Lt values of Tripleurospermum caucasicum (Willd.) Hayek (Asteraceae). Based on the data from Maffei et al. (1999) and Inceer and Beyazoglu (2004), I calculated the As.k (%) and Ls/Lt values of M. micranthathe and the Ls/Lt values of T. caucasicum (Willd.) Hayek. I found that the values of As.k (%) were similar among the seven dip

LI � The cytogeography of Aster ageratoides var. laticorymbus

dient and a migration route of the hexaploid from Yunnan-Guizhou Plateau to eastern China. This migration route is consistent with "Pattern Ii" of migration routes described by Wang (1992a). Some taxa such as Tetrastigma hemsleyanum Dieks et Gilg (Vitaceae) were considered to have originated in southwestern China and then have migrated eastwards to eastern China, belonging to the "Pattern Ii" in migration routes (Wang, 1992a). However, Wang (1992b) divided the migration route of Aster ageratoides into "Pattern IIiv", that is, "From Southwest China to Siberia or /and adjacent regions." In fact, Aster ageratoides is so complex geographically that no simple pattern can be used to describe its migration. It is more useful to pay attention to the geography of each variety of this species.

Formation Mechanism of the Distribution Pattern

Figure 1 shows that in the whole distribution region of the complex, the diploid and the tetraploid seem to be on the decline while the hexaploid is dominant in distribution. Two factors might have contributed to shaping such a distribution pattern. First, competitive exclusion may explain the phenomenon that the hexaploid and the diploid or tetraploid are never sympatric. All ploidy levels usually occur in partially shaded places such as at forest margins and roadsides (Table 1), and are apt to form aggregative distribution owing to well-developed clonal growth via rhizomes (Li, 2002), indicating that different ploidy levels share very similar ecological needs. Thus, it is possible that an intensive competition rages among different ploidy plants of the complex if they are sympatric. Although autopolyploid evolution was viewed as maladaptive, molecular data have revealed three important genetic attributes of autopolyploids compared to their diploid progenitors: enzyme multiplicity, increased heterozygosity, and increased allelic diversity, leading to the potential success of autoploids in nature (Soltis and Soltis, 1995). Thus, autopolyploidization might have made the hexaploids of the complex more competitive so that the diploid populations and tetraploid populations went extinct when they were sympatric with the hexaploid. As a result, the diploid and the tetraploid gradually decreased when the hexaploid increased in population number. Secondly, it has long been held that a polyploid has the ability to colonize a wider range of habitats. Additionally, with a well-developed pappus the hexaploid of the variety laticorymbus disperses easily to a new habitat. It holds an advantage over many other sympatric species because of its vigorous clonal growth or vegetative reproduction with horizontal rhizomes. Furthermore, well-developed clonal growth facilitates the spread of the variety with self-incompatibility (Li, 2002) when only one or a small quantity of the plants arrive at a new site. In fact, the hexaploid usually dominates the herbaceous layer where it occurs (Li, 2002). In contrast, its two closely related taxa with no horizontal rhizomes, Aster ageratoides var. micranthus Ling and A. shennongjiaensis W. P. Li et Z. G. Zhang, are narrow taxa or even dangerous species (Li, 2002; Li and Zhang, 2004).

loid populations (ranging from 60.23 to 63.76) and the four tetraploid populations (ranging from 61.83 to 63.55) of M. micranthathe. In M. micrantha the Ls/Lt values (3.1-3.9) of the four tetraploid populations were less than the variation range (3.2-5.0) of the seven diploid populations while the values were identical largely between the one diploid population (1.78) and two tetraploid populations (1.77 and 1.91, respectively) of T. caucasicum. Therefore, compared with these plants of Asteraceae, A. ageratoides var. laticorymbus is unique and notable in the karyotypical differences between the diploid and the polyploid.

In order to further test the above hypotheses, it is necessary to make karyotypic investigation and gather molecular evidence (including chloroplast DNA data) in more populations of the complex, especially in the Yunnan-Guizhou Plateau.

The Distribution Center and the Dispersal Route of the Complex

The Yunnan-Guizhou Plateau, belonging to southwestern China and covering 500,000 square km, comprises eastern Yunnan Province and most parts of Guizhou Province. It has an elevation of 1,000 to 2,000 meters and a terrain that descends from northwest to southeast. The plateau is covered with numerous mountain ridges, valleys and other rugged landforms. Yunnan-Guizhou Plateau plus Sichuan Province might have been an important center of development for angiosperms in the Northern Hemisphere during the Middle Cretaceous, and a strong evolutionary radiation might have taken place there, which resulted in the formation of many migration routes from this center to various regions in various directions (Wang, 1992a).

All ploidy levels of var. laticorymbus can be found in the Yunnan-Guizhou Plateau. The diploid occurs in the west (P3) and the north (P2; Soejima et al., 1999) of the plateau, the tetraploid and the octaploid in the north (Soejima et al., 1999), and the hexaploid in the center (P4 and P5). Moreover, the other tetraploid populations (P1 and P11-14) are located in the transitional zones from the plateau, the second step of Chinese topography, to the first step (P1) and to the third step (P11-P14). Only hexaploid extends far away from the plateau, arriving at eastern China. Therefore, the Yunnan-Guizhou Plateau can be regarded as the diversity center and the distribution center of the complex and might be its original center.

Because there is only one hexaploid in eastern China (Figure 1), it is reasonable to conclude that it originated in Yunnan-Guizhou Plateau and then from there dispersed eastwards to eastern China, forming a long and narrow distribution region (Figure 1). This conclusion is supported by morphology of the variety. The middle phyllary width of the hexaploid (P4-5) in Yunnan-Guizhou Plateau is 0.95-1.25 mm, identical with that of P1-P3, and the middle phyllaries of the hexaploid become wider and wider gradually from western populations (0.95-1.25 mm wide in P4-5) to middle longitude populations (e.g., 1.0-1.5 mm wide in P16 and P20) and to eastern populations (e.g., 1.35-1.60 wide in P24), showing a clinal variation along a longitudinal gra

Botanical Bulletin of Academia Sinica, Vol. 46, 2005

The Age of the Polyploid Complex

The present distribution pattern of the variety suggests that it represents a mature polyploid complex. In the 26 populations of var. laticorymbus sampled, only two are diploid, four are tetraploid, and the other twenty, representing 76.9% of all populations are hexaploid. The hexaploid occupies a large longitude range from 106�41' E to 120�09' E, especially exclusively from 111�36' E to 120�09' E while the diploid and the tetraploid occupy much narrower areas (Figure 1). The fact that the hexaploid dominates the distribution range of the complex seems to suggest that the complex has developed fully and arrived at a mature phase in the sense of Hong (1990).

The variety may remain a relatively young complex. First, in the Shonnongjia area two populations were samplied, of which P12 was located at 1,100 m to 1,600 m alt, and P11 occurred at 2,100-2,990 m alt., occupying mainly the mountain summit where forests are similar to those of the other sites. The population P11 is characterized by consistently subsessile, oblong leaves, and depending too much on the alpine conditions to live longer than a year in the garden of HNNU after they were transplanted three times while the plants from P12 and other populations can live longer after transplantation (Li, 2002). Such a unique population is distinguished morphologically and ecophysiologically from other populations and has not been found in other nearby mountains, suggesting this population differentiated in situ from P12 and has not begun to dispersal. Therefore, the tetrapoid is still in evolution.

Secondly, hexaploid is restricted within a relatively narrow middle latitude region from approximately 25� N to 31�N while there are no Aster ageratoides var. laticorymbus north of 31�N and south of 25� N according to my field sampling and my study of specimens in 14 herbaria (PE, WUK, SZ, CDBI, HNNU, KUN, IBSC, IBK, NAS, HGAS, CCNU, HIB, HIMC, and FUS) and Flora Reipublicae Popularis Sinicae (Ling and Chen, 1985). With well-developed pappus, the variety dispersed quickly eastward from Yunan-Gizhou Plateau along middle latitude because humidity and temperature conditions are similar at the similar latitude. It is very likely that the variety will continue expanding its latitudinal range.

Thirdly, in the Longshan, Furong and Nanyue Mountains, of Central China, the hexaploid populations (P16, P18 and P19) reach the mountain summits with altitudes of 1,514 m, 1,300 m, and 1,289 m, respectively, while the hexaploid population (P24) is restricted below 500 m alt. in Huangshan Mountain with the peak of 1,864 m, which is situated in East China. Such a difference may be interpreted as a possibility that the hexaploid arrived at Huangshan Mountain not long ago and has a potential to expand upwards in the future.

In a word, it is reasonable to conclude that Aster ageratoides var. laticorymbus is, though the hexaploid is dominated, a relatively young polyploid complex, for the tetraploid remains in evolution. The hexaploid has not

reached its greatest latitude extension and in eastern China has not reach its greatest altitude range.

Acknowledgements. I am grateful to Prof. Thomas G. Lammers and an anonymous reviewer for their critical reading of the manuscript and their valuable comments. This research was supported by the National Natural Science Foundation of China (Grant No. 30470131), the Natural Science Foundation of Hunan Province (Grant No. 02JJY4055), and the Scientific Research Fund of the Hunan Provincial Education Department (Grant No. 02C218).

Literature Cited

Chen, R.-Y., K. Irifune, W.-Q. Song, X.-L. Li, K. Taniguchi, and R. Tanaka. 1992a. Report on the cytogeographic investigation on Chinese Aster ageratoides. In T. Ryuso (ed.), Cytogenetics on Plant Correlating between Japan and China. Hiroshima University, Hiroshima, pp. 89-112.

Chen, R.-Y., K. Irifune, W.-Q. Song, X.-L. Li, K. Taniguchi, and R. Tanaka. 1992b. On the cytogeography and speciation of Chinese Aster ageratoides polyploid complex. In T. Ryuso (ed.), Cytogenetics on Plant Correlating between Japan and China. Hiroshima University, Hiroshima, pp. 113-136.

Handel-Mazzetti, H. 1938. Plantae Sinenses. Acta Horti Gothoburgensis 12: 203-226.

Hong, D.-Y. 1990. Plant Cytotaxonomy. Beijing: Science Press.

Huziwara, Y. 1957. Karyotype analysis in some genera of Compositae. III. The karyotype of the Aster ageratoides group. Amer. J. Bot. 44: 783-790.

Inceer, H. and O. Beyazoglu. 2004. Karyological studies in Tripleurospermum (Asteraceae, Anthemideae) from north-east Anatolia. Bot. J. Lin. Soc. 146: 427-438.

Ito, M. and A. Soejima. 1995. Aster. In K. Iwatsuki, T. Yamazaki, D. E. Boufford, and H. Ohba (eds.), Flora of Japan, Angiospermae, Dicotyledoneae, Sympetalae (b). Kodansha, Vol IIIb, Tokyo, pp. 59-73.

Levan, A., K. Fredga, and A.A. Sandberg. 1964. Nomenclature for centromeric position on chromosomes. Hereditas 52: 201-220.

Li, W.-P. 2002. The Biosystematic Studies on Aster ageratoides (Asteraceae) Polyploid Complex. D. Phil Thesis, Hunan Agricultural University, China.

Li, W.-P. and Z.G. Zhang. 2004. Aster shennongjiaensis (Asteraceae), a new species from central China. Bot. Bull. Acad. Sin. 45: 95-99.

Ling, Y. and Y.-L. Chen (eds.). 1985. Flora Reipublicae Popularis Sinicae, Vol. 74. Science Press, Beijing, pp. 159-167.

Maffei, E.M.D, M.A. Marin-Morales, P.M. Ruas, C.F. Ruas, and N.I. Matzenbacher. 1999. Chromosomal polymorphism in 12 populations of Mikania micrantha (Compositae). Genet. Mol. Biol. 22: 433-444.

Nesom, G.L. 1994. Review of the taxonomy of Aster sensu lato (Asteraceae: Astereae), emphasizing the New World species. Phytologia 77: 141-297.

Soejima, A. and C.-I Peng. 1998. Aster. In Editorial Commottee of the Flora of Taiwan (ed.), Flora of Taiwan, Second edition, vol. 4. Taipei, pp. 848-868.

Soejima, A., S.-G. Wu, and K. Iwatsuki. 1999. The Aster ageratoides complex (Asteraceae) in northeastern Yunan,

LI � The cytogeography of Aster ageratoides var. laticorymbus

China: cytological and morphological features. Acta Phytotaxonomica Geobotanica 50: 179-185.

Soltis, D.E. and P.S. Soltis. 1995. The dynamic nature of polyploid genomes. Proc. Natl. Acad. Sci. USA 92: 8089-8091.

Soltis, D.E. and P.S. Soltis. 1999. Polyploidy: recurrent formation and genome evolution. Trends Ecol. Evol. 14: 348-352.

Stebbins, G.L. 1971. Chromosomal Evolution in Higher Plants. London: Edward Arnold.

Wang, W.-T. 1992a. On some distribution patterns and some migration routes found in the eastern Asiatic region. Acta Phytotaxonomica Sin. 30: 1-24.

Wang, W.T. 1992b. On some distribution patterns and some migration routes found in the eastern Asiatic region (cont.). Acta Phytotaxonomica Sin. 30: 97-117.

Yang, J. 2002. The formation and evolution of polyploid genomes in plants. Acta Phytotaxonomica Sin. 39: 357-371.