The aim of this paper is primarily to publish the chromosome numbers and karyotypes of the alpine species of the Himalayan Saxifraga in order to reveal the cytological diversity of the alpine region of the Himalayas, Cytological studies of plants found above 4000 m in altitude have not yet been undertaken, though they have now become widely accepted as an important technique for revealing speciation and diversification in the higher plants. The genus Saxifraga is distributed widely in the temperate regions of the Northern Hemisphere, the Arctic and the Andes and includes more than 370 species. Saxifraga is one of the most important components of the alpine flora of the Himalayas (Ohba, 1986). Particularly the sections Ciliatae (Hirculus) and Porophyllum (Kabschia) are quite diverse in the high alpine region of the Himalayas as well as in Tibet and SW China (Yunnan & Szechuan). As regards the taxonomy of the Himalayan Saxifraga, extensive studies have been carried out by Engler & Irmscher (1916, 1919), H. Smith (1958, 1960) and Hara (1979), but our present knowledge is still far from sufficient. The great plasticity of this genus, which results in taxonomic difficulties, is derived from polyploidy, apomixis, natural hybridization, and restricted gene flow. Particularly, in the high alpine region, restricted gene flow is expected to have given rise to very narrow endemism limited to a single peak or mountain range (Ohba, 1984). These conditions are quite fascinating not only for taxonomic but also biosystematic researches. As regards cytology, the chromosome numbers of Saxifraga have been most frequently recorded from the species distributed in the temperate regions of Europe and North America and also from the Arctic regions (see Fedorov, 1969; Goldblatt, 1984). Among the Himalayan apecies, chromosome numbers of only two species have been reported (Hamel, 1953, 1960). Karyomorphological analysis has not yet been tried in this genus. During the course of identification of the material used for the present study, we were faced with three new species. These have been described in a different paper (Ohba & Wakabayashi, 1987). Morphologic and taxonomic problems of the Himalayan Saxifraga are omitted from this paper and will be published in forthcoming papers. Materials and MethodsThe materials for this cytological study were sampled from 85 populations of the alpine zone in Rolwaling Himal and Shorong Himal in Nepal. The root-tips used for cytological observation were fixed in the field. The pretreatment was made with 0.05% colchicine solution for about 3 hr at 5-15°C in a thermos bottle filled with cold water. The 0.05% colchicine solution was also made in the field, and used within 2 days. After pretreatment, the root-tips were fixed with fresh solution of 2: 1: 1 alcohol-acetic acidchloroform mixture. This mixed solution in which the root-tips had been fixed was replaced at least three times at weekly intervals with the same fresh solution in order to cleanse them. In order to macerate cell-walls and to soften the root-tips, they were soaked in solution (1: 1) of 2% pectinase and 4% cellulase for about 1 hr at 37°C. They were then stained with 2% aceto-orcein for about 24 hr and were squashed. The karyotype analysis and the measurement of chromosome size were made on the chromosomes observed at metaphase. The arm ratio (length of the long arm/length of the short arm) and the nomenclature used here for the centromeric position on chromosomes are as follows: 1.0-1.5 metacentric, 1.5-2.5 submetacentric, 2.5-7 subtelo centric, 7.0- acrocentric. The voucher specimens in this cytological study are deposited in the herbaria of University of Tokyo (TI), Department of Medicinal Plants, Nepal (KATH), and Makino Herbarium, Tokyo Metropolitan University (MAK). ResultsIn this study, the chromosome numbers of about 30 species of the Himalayan Saxifraga were counted and are presented in Table 1 together with their voucher data The chromosomes observed in a somatic cell of each species are shown in Figs. 1 & 2, Plates 3 & 4. Brief description of the cytological features of each species is as follows:
S. amabilis H. Ohba et Wakabayashi (Ohba & Wakabayashi, 1987) S. aristulata Hook. f. et Thoms. 2n=16 (Plate 3a). All of the five populations examined had the same chromosome number. Chromosomes are 6.3-3.3 µm long. The karyotype (Fig. 3B) is composed of two pairs of large metacentric chromosomes (nos. 1 & 2) and six pairs of compara tively small ones among which one is submetacentric (no. 4), four metacentric (nos. 3, 5-7) and one subtelocentric (no. 8).
S. brachypoda D. Don
S. brumonis Wall. ex Seringe S. cordigera Hook. f. et Thoms. S. diversifolia Wall. S. engleriana H. Smith S. glabricaulis H. Smith S. harai, H. Ohba et Wakabayashi (Ohba & Wakabayashi, 1987) S. hispidula D. Don S. hookeri Engl. et Irmsch. S. jacquemontiana Decne. S. lychnitis Hook f. et Thoms. S. mallae H, Ohba et Wakabayashi (Ohba & Wakabayashi, 1987) S. microphylla Royle ex Hook, f, et Thoms. S. montanella H. Smith S. moorcroftiana (Ser.) Wall. ex Sternb. S. nakaoi Kitamura S. parnassifolia D, Don S. pilifera Hook f. et Thoms. S. saginoides Hook. f. et Thoms. S. sikkimensis Engl. S. sphaeradena H. Smith S. strigosa Wall. ex Seringe S. afghanica Aitch. et Hemsl. S. andersonii Engl. S. hypostoma H. Smith S. subsessiliflora Engl. et Irmsch. S. williamsii H. Smith S. gageana W. W. Smith DiscussionOf about 30 species examined in this study, most of the species belong to sect. Ciliatae, five to sect. Porophyllum and one to sect. Micranthes. Their cytology is dis cussed by section. Sect. CiliataeThe Ciliatae (=Hirculus) is one of the most diversified sections in the Himalayas. In Nepal there are 54 species, of which about 24 were examined cytologically. Among them the somatic number 2n=16 was found in 16 species, 2n= 16 and 2n=24 in three, 2n=24 in two, 2n=32 in two, and 2n=48 and 2n=50 in one each. Thus, the basic chromosome number of this section must be x=8, and 2n=16, 24, 32 and 48 indicate, respectively, 2x, 3x, 4x and 6x. The individuals having 2n=50 may be aneuploida of 2n=48. The chromosome number 2n=2x=16 is widely prevailing in sect. Ciliatae of the Himalayan Saxifraga. Though the species in this section have a strongly diversified gross morphology, their karyotypes are generally similar. The karyotypes of nine species are shown in Fig. 3. According to Engler and Irmscher (1916 & 1919), all species are in the Hirculoideae group except S. lychnitis. S. lyclmitis, the representative species of the Lychnitideae group, is unique in having linear or narrowly elliptic petals and erect sepals covered with dense reddish or brownish multicellular hairs. S. aristu lata is apparently isolated from the remaining species in the Hircidoideae group. The karyotypes of these species have in common two pairs of large metacentric chromosomes (nos. 1 & 2) and one small subtelocentric pair (no. 8). A slight difference is found among pairs nos, 3-7, Of these five pairs: one is submetacentric and the others meta centric in S. cordigera (Fig. 3A), S, glabricaulis (B), S, aristulata (C), S. dwersifolia (D) and S. sphaeradena. (F), all are metacentric in S, parnassifolia (E); and two are submeta centric and the others metacentric in Saxifraga ap. aff. S, sphaeradena (G), S. hookeri (H, I), S. montanella (J) and S, lychnitis (K). Satellites are often found on the short arms of one pair of submetacentric chromosomes in S. cordigera (Fig. 3A, no, 4), S. hookeri (I, no, 3) and S. montanella (J, no. 3). The significance of three pairs of htstero morphic chromosome observed in a complement found in the individuals (collection no. 8520396) identified as S. hookeri (Plate 3b; Fig. 31, nos. 3, 7 & 8) is, at present, obscure. Though pictures of chromosomes clear enough for karyotype analysis have not been obtained in S. moorcroftiana and S. nakaoi, the chromosomes observed in these two species seem to indicate that their karyotypes may be similar to those of the species mentioned above. The karyotypes of the diploids found in S, mallae and S. sikkimensis (Fig. 5A1, B1), also resemble those of the species mentioned above. It seems that these 13 or so species are closely allied cytologically to one another, though there are slight karyological differences. S. brunonis and S. pilifera, of the Flagellares group, have the chromosome number 2n= 16. Their chromosomes are, as a whole, smaller, and somewhat different from those of the species described above (Fig. 2b, c) and resemble in size those of S. engleriana (Fig. 2a) of the Sediformes group. The karyotypes of these three species are shown in A, B, C of Fig. 4. The karyotype of S.pilifera (Fig. 4B) is somewhat similar to those in Fig. 3, in having two pairs of comparatively large metacentric chromosomes (nos. 1 & 2) and, among the somewhat smaller ones, two submetacentric pairs (nos. 3 & 4), three metacentric (nos. 5-7) and one acrocentric (no. 8). In S, brnnonis, there are three pairs of subtelocentric chromosomes (Fig. 4A, nos. 4, 7 & 8) among which no. 7 has satellites on the short arms, one submetacentric (no. 3) and four metacentric (nos. 1, 2, 5 & 6) in a set. The karyotype of S. brunonis (Fig. 4A) is fairly different from that of S. pilifera. S. engleriana has four pairs of submetacentric chromosomes (C, nos. 1-4), one acrocentric (no. 8) and three metacentric (nos. 5-7), thus its karyotype is quite different from those of the two species belonging to the Fidgellares group. Bused on these cytological results, these three species are considered to be rather remotely related to one another, and to be specialized cytologically in having comparatively small chromosomes. S. microphylla, which is closely related to S. engleriana, also has the chromosome number 2n=16. Though it was impossible to make a karyotype analysis, it was ob served that the chromosomes of S. microphylla were also small, being similar in size to those of the three species mentioned above, S. jacqncmontiiana is also classified in the Sediformes group. The chromosomes of this species, 2n=16, are very small (Fig. 2h). Cytologically, this species is considered to be more specialized. A chromosome number 2n=3x=24 has been found in S. mallae (Fig, 1e), S. sik kimensis (Plate 4b), S. hispidula (Fig. 2a), S. hrachypoda (Plate 3d) and S. amabilis. The first three species have also a diploid chromosome number 2n= 16 (Plates 3f & 4a; Fig. Id). Morphulogically S. mallae, belonging to the S. hirculus-complex in the Hirculoideae group, is related to S. lepida H. Smith and probably also to S, nakaoi Kitanmni, S, nigroglandulosa Engl. et Irmsch. from Yunnan, and S. chumbiensis Engl. et Irmsch, from southernmost Tibet. S, amabilis is morphologically related to S.palpe brata Hook. f. et Thoms, or S. cordigera Hook. f. & Thoms., but is isolated from the S. hirculus-complex. S. sikkimensis belongs to the Turfosae group, while S, hispidula and S. brachypoda arc of the; Genmiiparae group. The karyotypes of these species are shown in Fig, 5, but that of S, amalnlis is not presented here as the chromosomal pictures were taken at promctaphase, The chromosomes of S. mallae and S. sikkimeusis are, in general, large, whereas they are small in S, hispidula and S. brachypoda. The karyotypes of S. mallae and S, sikkimensis in, their diploid individuals closely resemble each other (Fig, 5A1, B1), and they are also similar to thosf of the species shown in Fig, 3, especially to that of S. hookeri, in having two pairs of large metacentric chromosomes (nos. 1-4) and six pairs of comparatively small ones among which two are submetacentric (nos. 5-8), three metacentric (nos. 9-14) and one subtelocentric (nos. 15 & 16). In the triploid individuals (3x=24) found in S. mallae and S. sikkimensis, the number of subtelocentric chromosomes is always two in a chromosomal comple ment (see Plate 4b; Fig. le). If these triploids (3x=24) are autotriploid, two subtelo centric chromsomes in a somatic cell of the diploids should be three in that of the triploids, but this has never been observed in the triploids. Therefore, it is presumed that the triploid individuals of these two species are not autotriploids, but allotriploids with a different genomic constitution. The karyotypes of triploid individuals in S. mallae and S. sikkimensis are shown in A2 and B2 of Fig. 5. The chromosomes nos. 1 16 in A2 and B2 are well in accord with the diploid karyotypes, A1 and B1, and the remaining chromosomes nos. 17-24 are all metacentric in A2 as well as in B2. From this fact, the genomic constitution of the triploids found in S. mallae and S. sikkimensis is presumed to be AAB1). A is the genome composed of the haploid set of chromosomes in a karyotype such as A1 or B1, being especially characterized as having one subtelo centric chromosome. B is the genome composed of eight metacentric chromosomes, though its source has not been clarified. The karyotype of the diploids in S. hispidula is shown in Fig. 5, C1. This is composed of five pairs of large chromosomes among which two are metacentric (nos. 1-4), two submetacentric (nos. 5-8) and one subtelocentric (or acrocentric) (nos. 15 & 16), and three pairs of comparatively small ones among which two are metacentric (nos. 9-12) and one submetacentric (nos. 13 & 14). This karyotype is more or less different from those of S. mallae and S. sikkimensis, and the chromosomes in a set arc considered to be rather specialized judging from their small size. The karyotype of triploid individuals in this species is shown in C2. Also in this triploid, just as in the triploids of S. mallae and S, sikkimensis, the chromosomes nos. 1-16 in C2 are well in accord with the diploid karyotype C1, and the remaining nos. 17-24 are all metacentric. Therefore, the genomic constitution of this triploid may also be presumed to be AAB. The karyotype of the triploid found in S. brachypoda is shown in Fig. 5D. This karyotype resembles that of the triploid in S. hispidula (C2), though its two subtelo centric chromosomes (nos. 15 & 16) are smaller. This triploid may also be presumed to have an AAB genomic constitution, AA comprising the chromosomes nos, 1-16 and B the metacentric ones no. 17-24. A diploid chromosome number 2n=16 was reported in this species (Hamel, 1953), It is thought that this diploid has a karyo type composed of chromosomes nos. 1-16 as illustrated in Fig. 5, D. S. brachypoda is considered to be closely related to S. hispidula, from the similarities in karyotype and chromosome size. It should be noted that there is hardly any morphological difference between the diploid (AA) and the triploid (AAB), in S. mallae, S. sikkimensis and S. hispidula. The presence of a genome B in the triploids may not play an important role in their mor phological modification, though the pollen grains in the triploids of S. mallae (8520338) and S. sikkimensis (8520367, 8520368) are mostly sterile. In the other individuals of the triploids, pollen fertility has not yet been examined. It is also noteworthy that every triploid examined in the four species, S. mallae, S. sikkimensis, S. hispidula and S. brachypoda, has in common a genome B composed of eight metacentric chromosomes. From this fact, it may be speculated that each ancestor of the species having genomes AA had once hybridized with species having the genomes BB, which also resulted in the production of AAB triploids. This triploid with the AAB genomes could often reproduce sexually, producing a male gamete with genome A and a female gamete containing the A and B genomes. Gene flow has been maintained between the triploids and the diploids, always reproducing individuals with the genomic constitution AA and AAB. This condition of ancestral diploid-triploid complex con tinues in each species. The support for this speculation may be obtained from the sev eral reports that an odd cuploid condition has often been maintained by the sexual re production (Smith-White, 1948, 1955; Fagerlind, 1940; Gustafsson, 1944; etc.). This phenomenon has been called "permanent odd polyploidy" by Grant (1981). For ex ample, Leucopogon juniperinus of the Epacridaceae (Smith-White, 1948, 1955) is a sexually reproducing triploid with 2n=3x=12 chromosomes, being recognized as an allotriploid with the genomic constitution AAB. The embryo-sacs and egg nuclei contain eight chromosomes belonging to the A and the B genomes, while the function ing pollen grains carry four chromosomes of the A genome. Fertilization restores the somatic number 2n=12 and the genomic constitution AAB. The above hypothesis about the origin of the triploids found in S. mallae, S. sik kimensis, S. hispidula and S. brachypoda is based on. the karyotypcs obtained in this study. It goes without saying that further studies of these triplokb, including S. amabilis, are needed, especially as to their meiotic behaviour, in order to clarify the real nature of these triploids. S. saginoides and S. strigosa are 2n=4x=32 in somatic chromosome number (Plate 4d; Fig. 2f). These two species are quite different and located in two different groups: that is, S. saginoides belongs to the Hirculoideae group while S. strigosa is of the Gem miparae group. Their karyotypes are shown in Fig. 6. In S. saginoides (Fig. 6A), the karyotype comprises 12 pairs of metacentric chromosomes (nos. 1-12), two subnaeta centric (nos, 13 & 14) and two subtelocentric (nos, 15 & 16). The chromosomes appear to be arranged by fours, as they are distinct in having four similar subtelocentric chro mosomes. This suggests that this tetraploid genomic constitution is AAAA, though it may have been diploidized. In S. strigosa (Fig. 6B), the karyotype is composed of eight pairs of comparatively large chromosomes among which three are metacentric (nos. 1-3) and five submetacentric (nos, 4-8), and eight pairs of smaller ones among which seven are metacentric (nos. 9-15) and one subtelocentric (no. 16). Satellites are found in metacentric chromosomes pair no. 12. This shows that this tetraploid is highly diploi dized in karyomorphology, being distinct in the presence of only one pair of subtelo centric chromosomes (no. 16) and one metacentric pair with satellites (no. 12) in a complement, and which suggests that this tetraploid may be an allotetraploid. How ever, Hamel (1960) reported that the somatic chromosome number was counted as 2n=16 in S. strigosa. Further cytological studies may be needed in this species. At any rate, the cytological evidence obtained in this study suggests that this species is remotely related to the other species examined in sect. Ciliatae. S. harai, related morphologically to S. aristulata Hook. f. & Thoms, of the Hirculoi deae group, is the hexaploid with somatic chromosome number 2n=6x=48 (Plate 4c). As shown in Fig. 7, the karyotype comprises six pairs of large metacentric chromosomes (nos. 1-6) and the remaining comparatively small ones with 12 metacentric pairs (nos. 7-18), three submetacentric (nos. 19-21) and three subtelocentric (nos. 22-24). The chromosomes in a complement appear to be arranged by sixes, as they are distinct in having six similar subtelocentric chromosomes and six submetacentric ones. The karyotype suggests that this hexaploid's genomic constitution is AAAAAA, though it may have been diploidized. The aneuploid with 2n=50 has also been found in this species. The karyotype analysis indicates that two extra chromosomes are compara tively large metacentric ones, and the karyotype of the remaining 48 chromosomes is well in accord with the foregoing. Sect. Porophyllum (= Kabschia)Five species, S. afghamca, S. andersonii, S. subsessiliflora, S, hypostoma and S. willi amsii are in this section. Each of the first three species has a somatic number 2n =26, and the chromosomes in a complement are very small, 0.7-2µm long (Fig. 2i, j, k). The basic number x= 13 and the small chromosomes are considered to be characteristic of this section. It is difficult to make a karyotype analysis because of the small size of its chromosomes. S. hypostoma and S, williamsii were classified in sect. Porophyllum (=Kabschia) by H. Smith (1958). However, the cytological evidence obtained in this study suggests that these species should belong to sect. Ciliatae rather than to sect. Porophyllum. Both of the species have the somatic number 2n= 16 and the chromosomes are, as a whole, larger than those of S. afghanica, S, andersonii and S. subsessiliflora (Fig. 2e, g). The karyotypes of these two species are very similar to each other (Fig. 4D, E), each represented by four pairs of metacentric chromosomes (nos. 1, 6-8), three submetacentric (nos. 2, 4 & 5) and one subtelocentric (no. 3). The subtelocentric chromosomes are rather large, though the chromosomes in a complement are gradually reduced in size. In sect. Ciliatae, there are, at present, no species that have a karyo type similar to this. The cytological evidence suggests that S. hypostoma and S. willi amsii are closely related to each other and rather remotely related to the species in sect. Ciliatae. The cytological results indicate that the sect. Porophyllum delimited by Engler and Irmscher is cytologically heterogeneous. Sect. Micranthes Only one species, S. gageana, is cytologically reported here. Though the S. pallida group belonging to this section was examined, the cytological results will be reported in detail in a forthcoming paper. The somatic chromosome number of S. gageana is 2n=22 (Fig. 21) and 33 (Plate 4f). The basic number is x=ll in this species, and a somatic number 2n=33 is that of the triploid (3x), The euploid series of x=ll has also been found in the S. pallida group. Therefore, the basic number x=ll is con sidered characteristic of the Himalayan species of this section. The chromosomes of S. gageana are large, 4-5 µm long, and the karyotype of the diploid ia shown in Fig. 8. This karyotype is composed of one pair of metacentric chromosomes (no. 1), six sub metacentric (nos. 2-7) and four subtelocentric (nos. 8-11). One pair of subtelocentric chromosomes (no, 11) has satellites on their short arms. The karyotype analyzed in the triploid appears to be an autotriploid, whose chromosomes are arranged by threes in a complement. In our cytological study of 30 Himalayan Saxifraga, we found that diploids were very common, but that only three of 30 species had a higher ploidy level than tetra ploidy. Thus, it might be concluded that polyploidy, generally, does not play an important role in the diversification of the genus Saxifraga in the Himalayan region. References
|
1) In this paper, the letter A or B for representing the genome is used in common among the different species. However, this does not necessarily mean that each genome A or B among the different species is identical.[return to the text] |