( author for correspondence) Key words: Allium cepa var. ascalonicum, Allium wakegi, PCR-RFLP, RAPD, shallot

Euphytica 111: 23 31, 2000. 2000 Kluwer Academic Publishers. Printed in the Netherlands. 23 Genetic diversity in Indonesian shallot (Allium cepa var. ascalonicum)and Allium wakegi revealed by RAPD markers and origin of A. wakegi identified by RFLP analyses of amplified chloroplast genes Noor Sugiharto Arifin 1, Yukio Ozaki & Hiroshi Okubo Laboratory of Horticultural Science, Faculty of Agriculture, Kyushu University 46-01, Fukuoka 812-8581, Japan; 1 present address: Department of Plant Breeding, Faculty of Agriculture, Brawijaya University, Malang, Indonesia; ( author for correspondence) Received 8 September 1998; accepted 20 April 1999 Key words: Allium cepa var. ascalonicum, Allium wakegi, PCR-RFLP, RAPD, shallot Summary RAPD and PCR-RFLP analyses were conducted to establish the phylogenetic relationships among collected accessions of shallot and Allium wakegi, and to assess the origin of A. wakegi. Twenty out of 100 primers were amplified with 112 scorable bands for cluster analysis. Two main cluster groups consisting of one group for shallot and another group for A. wakegi were clearly separated. The sub-groups of clusters reflected the phenotype differentiation in shallots and regional specificity in some A. wakegi accessions. The present results were also in agreement with previous systematics by isozymes of which the highest genetic variation of A. wakegi in Indonesia was found in West Java suggesting the possibility that this place might be one of the germplasm centers. From RFLP analysis of amplified matk gene of cpdna it was demonstrated that A. wakegi originated from shallot as a maternal plant Welsh onion as a paternal plant as well as from reciprocal crosses. Introduction Shallot (Allium cepa var. ascalonicum or A. cepa Aggregatum group), a small-bulb onion, is more popular and important than common onion (A. cepa) in Southeast Asian countries because of the difficulty to grow and produce true seeds of common onion. Allium wakegi, another small-bulb onion, is an interspecific hybrid of Welsh onion (Allium fistulosum) shallot (Tashiro, 1980; Tashiro et al., 1990) and has been found cultivated in western Japan, China, Korea and Southeastern Asia (Tashiro et al., 1982; Tashiro, 1984; Okubo & Fujieda, 1989; Inden & Asahira, 1990). Recently, Arifin & Okubo (1996) clarified by isozyme analysis that there is a high genetic variation in A. wakegi as well as in shallot collected mainly in Japan and Indonesia. They found that there is mixed cultivation with no distinction between shallot and A. wakegi in some places of Indonesia probably because of the resemblance of the two species in morphology and physiology to each other. In this study the genetic relationships in the accessions of shallot and A. wakegi and the origin of A. wakegi were investigated by RAPD and PCR-RFLP analyses, respectively. Materials and methods Plant materials Sixty-five shallot and 64 A. wakegi accessions collected mainly in Indonesia and Japan and other countries, which had been identified to be shallot or A. wakegi by isozyme analysis (Arifin & Okubo, 1996), were used. Seven Welsh onion (A. fistulosum) cultivars collected in Japan ( Kujo-hoso, Kujofuto, Ishikura, Shimonida and Matsumoto ), Hongkong ( Hongkong ) and Indonesia ( Baraka-4 ), and three common onion (A. cepa) cultivars purchased

24 Figure 1. Collection sites of shallot and A. wakegi in Indonesia. Numbers in the map correspond to those in Table 3. in Fukuoka, Japan were used as control. Collection sites in Indonesia are numbered on the map as shown in Figure 1. Total DNA isolation Total DNA isolation was carried out by the modified CTAB method by Murray & Thompson (1980). Approximate 60 mg young leaf materials were ground in liquid nitrogen in a mortar, and 1.5 ml of extraction buffer 1 (50 mm Tris-HCl ph 8.0, 5 mm EDTA, 0.35 M sorbitol, 13.3 mm PEG and 0.1% (v/v) mercaptoethanol) was added to each of the powdered samples. The samples were centrifuged for 5 min at 12,000 rpm in a microcentrifuge. After the supernatant was discarded, the residue was resuspended in 300 µl extraction buffer 2 (50 mm Tris-HCl ph 8.0, 5 mm EDTA, 0.35 M sorbitol, 1% sodium sarcosyl and 0.1% (v/v) mercaptoethanol) and 200 µl of extraction buffer 3 (1.8 M NaCl, 0.25% cetyl trimethylammonium bromide (CTAB)), and incubated at 65 C for 30 min. Approximate 500 µl of chloroform / isoamylalchohol (24:1, v/v) was added to the suspension, and centrifuged for 5 min at 12,000 rpm. The upper supernatant was mixed with 1% CTAB by shaking gently for an hour, and DNA was reprecipitated with 400 µl of 1 M CsCl and 800 µl of 100% ethanol for more than 20 min in 20 C. After reprecipitation, the extracted DNA was purified by the Geneclean III Kit (BIO 101, USA). Quantity of the extracted DNA was determined by analyzing the sample DNA on agarose gel alongside λ/hind III standard, and dissolved in the appropriate amount of TE to prepare for the following PCR experiments. RAPD assay The template DNAs from the accessions were subjected to screening with 100 different 12-base primers from CMN-A00 to CMN-A99 (BEX, Tokyo). Amplifications were performed in 25 µl reaction solution containing 100 µm each of datp, dctp, dgtp and dttp, 0.15 µm primer, 20 ng template DNA and 0.5 unit Tth DNA polymerase in 1 PCR buffer. The reaction solution was overlaid with one drop of mineral oil. Amplification was performed in an ASTEC Program Temp Control System PC 700 programmed as one cycle of preliminary denaturation by 94 Cfor 30 s, 41 cycles of 94 C for 30 s, 45 C for 1 min 30 s and 72 C for 2 min, and then closed by 72 C10min for final extension. Five µl amplified products were subjected to electrophoresis on a 1.5% agarose gels in the presence of ethidium bromide, and the gels were photographed under UV light. PCR-RFLP assay Three regions in cpdna (rbcl, 16S and matk) were amplified by PCR. Primers were designed as in Table 1. PCR amplification of these cpdna regions was performed in a total volume of 25 ml solution containing 10 ng template DNA, 0.5 mm of each primer, 100 µm of each dntp and 0.5 unit of Taq DNA polymerase. PCR amplification was carried out by one cycle for 4 min at 94 C, followed by 41 cycles of 30 sec at 94 C, 1 min at 50 C and 3 min at 72 C, and one cycle for 10 min at 72 C with an Astec Program Temp Control System PC 700. The amplified DNAs were purified by Geneclean III Kit (BIO 101, USA) and dissolved in 50 ml TE. The PCR products were digested by 2.5 units of nine different restriction en-

25 Table 1. Primer information for PCR amplification of three specific genes in chloroplast DNA Gene Primer (forward and reverse) Size (bps) Source rbcl z 5 -gtcggattcaaagctggtgt-3 1,309 Pinus thunbergii 5 -TCACAAgCAgCAgCTAgTTC-3 16S y 5 -ACgggTgAgTAACgCgTAAg-3 1,375 Nicotiana tabacum 5 -CCAgTACggCTACCTTgTTAC-3 matk x 5 -ggggttgctaactcaacgg-3 900 Nicotiana tabacum 5 -AACTAgTCggATggAgTAg-3 z Shiraishi & Watanabe, 1995. y Shinozaki et al., 1986. x Johnson & Soltis, 1994. zymes, Afa I, Alu I, Bgl I, Hha I, Hinf I, Mva I, Msp I, Sau3A I and Taq I. Each restriction fragment of each region was separated by agarose gel electrophoresis as described above. Data analysis The RAPD profiles were scored manually from photographs of the gels, by assigning a value of 1 for band presence and 0 for band absence. Only the stable (reproducible) bands after several times repetition were scored. The scores of band presence or absence were used to calculate a pairwise genetic distance matrix using the formula of Nei & Li (1979). A dendrogram based on UPGMA cluster analysis of the genetic distance matrix (personally programmed) was constructed. Results and discussion RAPD Twenty of the 100 primers tested were polymorphic, whereas 12 primers did not show any amplification and 68 primers showed either monomorphic or very complex banding patterns. Of the polymorphic primers, four of them revealed an interspecific differentiation only, whereas the remaining revealed both inter- and intraspecific differentiation (Table 2). The number of polymorphic bands per primer varied from 1 to 14 in the size ranging from 240 to 2200 bps. In total, polymorphism was detected from 112 markers with a mean of 5.6 different bands per primer. Amplification with different primers, however, produced some accession-specific patterns. With the primer CMN-A45, there was no polymorphism of banding patterns in all the shallot and Welsh onion accessions, although they were not identical. Only three accessions of A. wakegi Wataes from South Sumatra (Site No. 13), Panundaan from West Java (Site No. 19) and one accession from Davao, Philippines had the bands that both shallot and Welsh onion had (Figure 2a). Other accessions of A. wakegi had the same banding pattern as Welsh onion. Polymorphism having three bands at 750, 1100 and 1360 bps detected with the primer CMN-A31 was found only in shallot accessions collected in West Java, West Sumatra and Thailand. Eighty nine percent of A. wakegi accessions, including the collection in Japan, had the same banding pattern as shallot (Figure 2b). The results may indicate that many Japanese, A. wakegi accessions are derived from the shallot in southeast Asia. With the primer CMN-A21, only one accession of A. wakegi Panundaan from West Java (Site No. 19), showed the same banding pattern as shallot, whereas the remaining 58 out of 59 accessions showed the same banding pattern as Welsh onion (Figure 2c). Bands at 630 and 820 bps with the primer CMN-A08 were specific in shallot of West Java, and not found in any other places (Figure 2d). Only one accession of A. wakegi, Panundaan, had both bands of the same sizes. The results may indicate that A. wakegi Panundaan originated in West Java and may suggest that A. wakegi accessions have originated in different places. Forty-nine cluster trees were divided into two main cluster groups i.e., shallots (including common onion) (cluster types A R) and A. wakegi (cluster types a ab) and Welsh onion as an out group (cluster types fa fc) (Figure 3). The distance was 0.5 between shallot and A. wakegi, and 0.75 between A. wakegi and Welsh onion. The accession names and their origin associating with cluster types in Figure 3 are presented in Table 3. Number of cluster types consisting of only

26 Table 2. Primers causing polymorphism and number of scored bands Primer Sequence Polymorphism No. of scored bands CMN-A04 gccccgttagca Interspecific 5 CMN-A08 TTCggACgAATA Inter- & intraspecific 5 CMN-A13 CTCAgCgATACg Inter- & intraspecific 4 CMN-A15 ATCgCggAATAT Interspecific 7 CMN-A21 gtgaccgatcca Inter- & intraspecific 5 CMN-A30 CCTTTCCgACgT Inter- & intraspecific 11 CMN-A31 ggtggtggtatc Inter- & intraspecific 10 CMN-A32 CTTgTCATgTgT Inter- & intraspecific 5 CMN-A45 TggCCTCTTggA Inter- & intraspecific 4 CMN-A50 ATTggTgCAgAA Inter- & intraspecific 5 CMN-A53 gacgcccattat Inter- & intraspecific 14 CMN-A54 AAggCgTgTTTA Inter- & intraspecific 7 CMN-A58 gtcatgcctgga Inter- & intraspecific 6 CMN-A59 CAgTgggAgTTT Inter- & intraspecific 2 CMN-A60 CAggTgggACCA Inter- & intraspecific 2 CMN-A69 AAgCCTATACCA Interspecific 2 CMN-A71 ggtgccggagca Interspecific 1 CMN-A75 ggcggttatgaa Inter- & intraspecific 4 CMN-A98 gacggttctaca Inter- & intraspecific 7 CMN-A99 gcggtcagcaca Inter- & intraspecific 6 Total 112 Figure 2. RAPD patterns with primers CMN-A45 (a), CMN-A31 (b), CMN-A21 (c) and CMN-A08(d). 2a. Lanes 1 8; A. wakegi Lane 9; λ/hindiii, Lanes 10 11; shallot, Lane 12; A. wakegi Lane 13; Welsh onion, Lane 14; shallot, Lane 15; A. wakegi. 2b. Lane 1; Welsh onion, Lanes 2 3; A. wakegi, Lane 4; onion (A. cepa), Lane 5; λ/hindiii, Lane 6; A. wakegi, Lanes 7 9; shallot. 2c. Lane 1; shallot, Lane 2; onion, Lane 3; shallot, Lane 4; A. wakegi, Lane 5; Welsh onion, Lane 6; λ/hindiii, Lane 7; A. wakegi Shunsen, Lane 8; onion, Lane 9; A. wakegi Panundaan, Lane 10; shallot, Lane 11; onion. 2d. Lane 1; shallot, Lanes 2 5; A. wakegi, Lane6;A. wakegi Shunsen, Lane 7; shallot, Lane 8; λ/hindiii, Lane 9; A. wakegi Panundaan, Lanes 10 15; A. wakegi. (Lane numbers do not correspond to collection site numbers in Figure 1 and Table 3.)

Figure 3. Dendrogram constructed from the results of RAPD analysis. 27

28 Table 3. Classification of shallot and A. wakegi accessions with regard to the cluster type constructed from RAPD Cluster Accessions collected in type z Indonesia (Collection site No.) y Japan other countries A Samosir (2) B Amurang (39), Tlekung-1 (47) C Manado (40), Lambadoko (41) D Gunung Rajo-1 & -8 (5), Kawunggirang (21), Girimulyo-1 (26), Tonsewer- Thailand 1 (38), Probolinggo (51), Kedak (53), Pare (54), Batur (59), Selat (60), Gel-gel (61) E Pojongmanggu (18), Majalengka (20), Cidahu-1 & -2 (29), Saruran-1 (43), Barakan (44) F Medan Biasa (9), Cikareo-1 (16), Cilastari (17), Losari (30), Purwokerto (34), Colo (36), Harso (45), Arsopuro (46) G Cirebon (9) H Tongging-2 (4), Kayutanduk (6), Air Bertumbuk (8), Muara Enim (10), Ciledog-1 & -3 (28), Tlekung-3 (47), Karang Tengah (50), Kepung (55), Mampen (56), Pringgabaya (57), Sembalun (58) I Cingkir (9), Medan Merah (9) J Kemurang (31), Prupuk (32), Tasik (35) K China-1 L Siniang (25) M Batu-1 & -2 (48), Wonorejo (49) Ipoh N Dieng (33) O Batu-3 & -4 (48) P Phillipine-2 & -8, Taherpur, Gitka Q A. cepa (common onion) R Bhatia, Phillipine-9 & -11 a Sibolangit (3) Bise Taiwan Koba, Viscaino b Yonabaru, limori, Genkai c Shimonoseki, Hakata d Daisen, Mihara Chegidong e Cikareo-2 (16) Kihara f Amami, Toubaru, Shigatsubori, Murasakidane g Tongging-1 (4), Girimulyo-3 (26), Ciledog-2 & -6 (28), Saruran-7 (43) h Taketomi, Yakena i Alam Indah (15), Saruran-6 (43) Sera j Miyazaki, Kunisaki k Shunsen, Shunegi l Zairai, Shirodane m Ginoza-1 & -T n Teruma, Uehara, Haebaru o Motobu, Ikemi p Naha-2 q Isahaya China-2 & -3, Myanmar r Ciledog-4 (28), Bawang Merah x s Yatsushiro t Tonsewer-2 (38) u Phillipine-1 v Karatsu, Naha w Kihara Wase, Ishigaki, Yonakuni x Shodon, Inazawa y Shaden z Bandung x, Wonokrio (14) aa Wataes (13) ab Panundaan (19) fa Kujo Hoso, Kujo Futo, Ishikura Hongkong fb Shimonida, Matsumoto fc Baraka-4 (44) z Cluster types A R and a fc correspond to those in Figure 3. Cluster types A P and R; shallot, Q; common onion (A. cepa), a ab; A. wakegi and fa fc; Welsh onion (A. fistulosum). y Numbers in parentheses following accession names correspond to those in Figure 1 (collection site number in Indonesia). x Old introductions to our laboratory from Indonesia. Places of collection are not known.

29 Table 4. Number of fragments in 16S and matk genes restricted by several enzymes Restriction enzyme No. of fragments 16S matk AfaI 3 3 AluI 4 4 z BglI 1 3 HhaI 3 1 HinfI 4 4 MvaI 5 1 MspI 4 2 Sau3AI 4 6 z TaqI 5 5 z Polymorphisms occurred. one accession (unique accession) was five (A, G, L, K and N) for shallot and seven (p, s, t, u, y, aa and ab) for A. wakegi. Four of the five unique shallot accessions were collected in Indonesia. Three of the seven unique accessions of A. wakegi were from Indonesia, two from Japan and two from other countries. The maximum number of individuals grouped within a tree in shallot was 12, while that in A. wakegi was only five. Cluster types A to J include many common Indonesian shallot with many daughter bulbs (data on bulb number and size are not shown), whereas cluster types L to O consist of Java highland shallot which produces a rather big size bulb with or without a daughter bulb (Collection site Nos. 25, 33, 48 and 49). The grouping of shallot by RAPD analysis may thus reflect the morphological traits of shallot. A. wakegi was divided into four subgroups: cluster types a to j, k to q, r to y and z to ab. Nine out of 16 Indonesian and 18 out of 37 Japanese A. wakegi accessions are in the first subgroup. The remaining Indonesian accessions are scattered in the first, third and fourth subgroups, and the remaining accessions from Japan and other countries are distributed in the first, second and third subgroups. In A. wakegi, no relationship, however, is found between the subgrouped accessions and the sites of collection. PCR-RFLP Of the three genes tested, amplification of rbcl by PCR showed unstable bands, whereas 16S and matk genes were well amplified. Screening all the accessions with 10 restriction enzymes resulted in 33 and 30 fragments for 16S and matk genes, respectively Figure 4. RFLP patterns of specific PCR-amplified matk gene of cpdna digested with Sau3A I. Lane 1; shallot, Lane 2; Welsh onion, Lane 3; A. wakegi, Lane 4; A. wakegi Panundaan, Lane 5; λ/hindiii, Lane 6; onion. (Lane numbers do not correspond to collection site numbers in Figure 1 and Table 3.) (Table 4). Neither inter- nor intraspecific variation, however, was found in the digested fragments of 16S. In the fragments of matk, interspecific polymorphism between common onion and Welsh onion was detected with Sau3A I and Alu I, but no intraspecific polymorphism was found in these two species. Fragments of 580 and 1030 bps for Sau3A I were found only in shallot and common onion, while those of 380 and 860 bps were only detected in Welsh onion (Figure 4). Sixty-one of 64 A. wakegi accessions showed the same banding pattern as Welsh onion, but the remaining three accessions, Panundaan (West Java, Indonesia, Collection site No. 19), Tonsewer- 2 (North Sulawesi, Indonesia, Collection site No. 38) and Shunsen (Korea), had the patterns of shallot. Other polymorphismswerearesultof Alu I restriction which shared the bands of 342 and 345 bps for shallot and Welsh onion, respectively (data not shown). The A. wakegi accessions agreed with the shallot pattern and the Welsh onion pattern restricted by Alu Iwas completely the same as that restricted by Sau3A I. It has been proven that A. wakegi originated from natural interspecific hybridization between Welsh onion and shallot (Tashiro, 1984; Hizume, 1994). Harris & Ingram (1991) summarized the modes of inheritance of plastids of various plant species, among which both Welsh onion and common onion have the maternal inheritance. Tashiro et al. (1995)

30 also demonstrated that Welsh onion and shallot have maternal inheritance of plastid DNA in their reciprocal hybrids. They also confirmed, by the analysis of RFLP in the cpdna, that A. wakegi is an allodiploidplant which is homologous with the genome of Welsh onion as a maternal plant and with that of shallot as a paternal plant. They analyzed only nine accessions of A. wakegi, although they collected widely from Japan, Korea, China, Taiwan and Myanmar. The results of the present study confirmed the findings of Tashiro et al. (1995). It was also confirmed that some A. wakegi has the derived from the hybridization of shallot Welsh onion. This is the first report that proved that two types of A. wakegi exist and are cultivated. The reason why the Welsh onion shallot is more common than shallot Welsh onion is not known. Emsweller & Jones (1935) reported that Welsh onion and A. cepa cv. California Early Red were planted in adjacent plots and open pollinated seed was collected. Seventeen hybrids out of 836 plants grown from seeds harvested from Welsh onion were obtained. In contrast, 5,000 plants grown from seeds harvested from California Early Red yielded no hybrids. They concluded that these species cross more readily in one direction than in the other but they did not discuss the reason. Two possible reasons are that Welsh onion tends to be more outcrossed than shallot and that the flowering period is longer in Welsh onion than in shallot, but this has not yet been clarified. Relations of DNA data to collection (cultivation) sites in Indonesia In a very small area in West Java, from collection site Nos. 16 to 26, there were fields of shallot, Welsh onion and A. wakegi both from the cross of Welsh onion shallot and the reciprocal combination. Cikareo-l and Cikareo-2 were collected in the same field (Site No. 16), but the former was shallot and the latter was A. wakegi (Welsh onion shallot). At site No. 19, very close to site No. 16, A. wakegi originating from shallot Welsh onion ( Panundaan ) was found. East of site No. 19, there was a field where Welsh onion was in cultivation together with shallot ( Siniang ) (Site No. 25). Near site No. 25, there was a mixed cultivation of shallot ( Girimulyo-1 ) and A. wakegi (Welsh onion shallot) ( Girimulyo-3 ) in the same field (Site No. 26). The genetic distance among A. wakegi collected in this area (cluster types e, i, r and ab in Figure 3) was very large, and all other accessions of A. wakegi, including those collected in Japan, were genetically more similar than the accessions from West Java. West Java is considered as an area where the genetic variation of A. wakegi was very wide. It is supported by the previous report concluded by isozyme analysis (Arifin & Okubo, 1996). Many of the shallot accessions collected in these areas, by contrast, constructed the cluster types D ( Kawunggirang from Site No. 21 and Girimulyo-1 from Site No. 26), E ( Pojongmanggu from Site No. 18 and Majalengka from Site No. 20) and F ( Cikareo-1 from Site No. 16 and Cilastari from Site No. 17), among which the genetic distance is very close. This may indicate that the large genetic variation in A. wakegi accessions found in these areas is caused by Welsh onion populations having different genetic backgrounds. Based on the results West Java may be considered as one of the areas where A. wakegi was born. However, further studies are necessary. Besides in West Java, mixed cultivation of shallot with A. wakegi was found also in North Sumatra (Site No. 4) and North Sulawesi (Site No. 38). Mixed cultivation of shallot with Welsh onion occurred in South Sulawesi (Site No. 44), Welsh onion with A. wakegi in South Sumatra (Site No. 13) and of the three species together in the same field in South Sulawesi (Site No. 43). Considering the fact that there are fields of mixed cultivation of Welsh onion and shallot, it is also possible to imagine that new A. wakegi is still being created there, since the farmers do not discriminate A. wakegi from shallot. The species are cultivated and sold together. References Arifin, N.S. & H. Okubo, 1996. Geographical distribution of allozyme patterns in shallot (Allium cepa var. ascalonicum Backer) and wakegi onion (A. wakegi Araki). Euphytica 91: 305 313. Emsweller, S.L. & H.A. Jones, 1935. An interspecific hybrid in Allium. Hilgardia 9: 265 273. Harris, S.A. & R. Ingram, 1991. Chloroplast DNA and biosystematics: The effects of intraspecific diversity and plastid transmission. Taxon 40: 393 412. Hizume, M., 1994. Allodiploid nature of Allium wakegi Araki revealed by genomic in situ hybridization and localization of 5S and 18S rdnas. Jap J Genet 69: 407 415. Inden, H. & T. Asahira, 1990. Japanese Bunching Onion (Allium fistulosum L.) In: J.L. Brewster & H.D. Rabinowitch (Eds), Onions and Allied Crops Vol. III. Biochemistry, Food Science, and Minor Crops, pp. 159 178. CRC Press, Boca Raton, Florida. Johnson, L.A. & D.E. Soltis, 1994. matk DNA sequences and phylogenetic reconstruction in Saxifragaceae sensu stricto. Systematic Bot 19: 143 156.

31 Murray, M.G. & W.F. Thompson, 1980. Rapid isolation of high molecular weight plant DNA. Nucleic Acid Res 8: 4321 4325. Nei, M. & W.H. Li, 1979. Mathematical model for studying genetic variation in terms of restriction endonucleases. Proc Natl Acad Sci USA 76: 5269 5273. Okubo, H. & K. Fujieda, 1989. Intraspecific differentiation and isozyme patterns in Allium wakegi Araki. J Japan Soc Hort Sci 58: 401 406. Shinozaki, K., M. Ohme, M. Tanaka, T. Wakasugi, N. Hayashida, T. Matsubayashi, N. Zaita, J. Chunwongse, J. Obokata, K. Yamaguchi-Shinozaki, C. Ohta, K. Torazawa, B.Y. Meng, M. Sugita, H. Deno, T. Kamogashira, K. Yamada, J. Kusuda, F. Takaiwa, A. Kato, N. Tohdoh, H. Shimada & M. Sugiura, 1986. The complete nucleotide sequence of tobacco chloroplast genome: its gene organization and expression. EMBO J 5: 2043 2049. Shiraishi, S. & A. Watanabe, 1995. Identification of chloroplast genome between Pinus densiflora Sieb. et Zucc. and P. thunbergii Parl. based on the polymorphism in rbcl gene. J Jpn For Soc 77: 429 436. Tashiro, Y., 1980. Cytogenetic studies on the origin of Allium wakegi Araki. I. Meiosis in A. wakegi and the interspecific hybrid between A. fistulosum L. and A. ascalonicum L. Agr Bull Saga Univ 49: 47 57. Tashiro, Y., 1984. Genome analysis of Allium wakegi Araki. J Japan Soc Hort Sci 52: 399 407. Tashiro, Y., S. Miyazaki & K. Kanazawa, 1982. On the shallot cultivated in the countries of southeastern Asia. Bull Fac Agric Saga Univ 53: 65 73. Tashiro, Y., K. Amakawa, T. Mizutani, M.A. Quadir & S. Miyazaki, 1990. Genetic analysis of isozymes in Allium. 1. Aspartate transaminase isozymes in A. fistulosum, A. cepa, A. ascalonicum and A. wakegi. J Japan Soc Hort Sci 59 (suppl. 1): 232 233. Tashiro, Y., T. Oyama, Y. Iwamoto, R. Noda & S. Miyazaki, 1995. Identification of maternal and paternal plants of Allium wakegi Araki by RFLP analysis of chloroplast DNA. J Japan Soc Hort Sci 63: 819 824.