792 Phylogenetic relationships among Secale species revealed by amplified fragment length polymorphisms T. Chikmawati, B. Skovmand, and J.P. Gustafson Abstract: Amplified fragment length polymorphism (AFLP) data were utilized to analyze the phylogenetic relationships among 29 accessions representing 14 of the most commonly recognized ranked species or subspecies in the genus Secale. We observed 789 AFLP markers of 1130 fragments utilizing 18 P-/M- and E-/M- primer combinations. All polymorphic fragments were used to construct phenetic and phylogenetic trees. The resulting phenogram and cladogram had similar tree topologies. Cluster analysis showed that Secale sylvestre was the most distantly related to all other ryes. Annual forms were grouped together, and the perennial forms appeared more closely related to each other. This suggested that life cycle could have played an important role in determining the relationships among Secale species. Secale sylvestre was considered to be the most ancient species, whereas Secale cereale was the most recently evolved species. Amplified fragment length polymorphism analysis clearly separated all Secale species into only 3 major species groups, within the genus Secale: S. sylvestre, Secale montanum (syn. Secale strictum) for perennial forms, and S. cereale for annual forms. This study demonstrated that the AFLP approach is a useful tool for discriminating species differences, and also gave a much better resolution in discerning genetic relationships among Secale species as compared with previous studies using other approaches. Key words: AFLP, Secale, phylogenetic relationship. Résumé : Le polymorphisme de longueur des fragments amplifiés (AFLP) a été employé pour analyser les relations phylogénétiques entre 29 accessions représentant 14 des espèces ou sous-espèces les plus communément reconnues au sein du genre Secale. Les auteurs ont obtenu 789 marqueurs AFLP parmi 1130 fragments en employant 18 combinaisons d’amorces P-/M- and E-/M-. Tous les fragments polymorphes ont été employés pour produire des arbres phénétiques et phylogénétiques. Les phénogrammes et dendrogrammes résultants avaient une topologie semblable. Une analyse de groupement a montré que le Secale sylvestre était l’espèce la plus éloignée de tous les autres seigles. Les formes annuelles ont été groupées ensemble et les formes pérennes étaient plus apparentées les unes aux autres. Ceci suggère que le cycle vital pourrait avoir joué un rôle majeur dans l’établissement des relations au sein des espèces du genre Secale. Le S. sylvestre est considéré comme étant l’espèce la plus ancienne, tandis que le Secale cereale serait l’espèce la plus récente. L’analyse AFLP a clairement séparé toutes les espèces entre seulement 3 groupes majeurs : S. sylvestre, Secale montanum (syn. Secale strictum) chez les formes pérennes, et S. cereale chez les formes annuelles. Cette étude montre que les AFLP constituent un outil utile pour discerner les relations génétiques au sein des espèces du genre Secale par comparaison avec les études antérieures faisant appel à d’autres approches. Mots clés : AFLP, Secale, relation phylogénétique. [Traduit par la Rédaction] Chikmawati et al. 801 Introduction The genus Secale L. is a typical representative of Mediterranean flora. It has a wide distribution from central Europe and the western Mediterranean through the Balkans, Anatolia, Israel and the Caucasus to Central Asia, with an isolated population in South Africa (Sencer and Hawkes 1980). This genus includes perennial or annual, self-incompatible or self-compatible, and cultivated, weedy, or wild species (Vence et al. 1987). Cultivated rye (Secale cereale L.) is an Received 26 January 2005. Accepted 3 May 2005. Published on the NRC Research Press Web site at http://genome.nrc.ca on 18 October 2005. Corresponding Editor: P. Donini. T. Chikmawati. Department of Agronomy, University of Missouri-Columbia, Columbia, MO 65211, USA; and Department of Biology, Bogor Agricultural University, Bogor 16144, Indonesia. B. Skovmand. The Nordic Gene Bank, P.O. Box 41, SE230 53 Alnarp, Sweden. J.P. Gustafson.1 Department of Agronomy, University of Missouri-Columbia, Columbia, MO 65211, USA; and USDA-ARS, Plant Genetics Research Unit, University of Missouri-Columbia, Columbia, MO 65201, USA. 1 Corresponding author (e-mail: pgus@missouri.edu). Genome 48: 792–801 (2005) doi: 10.1139/G05-043 © 2005 NRC Canada Chikmawati et al. important source of bread, especially in parts of Northern and Eastern Europe with poor soils and severe winters (Leonard and Martin 1963). In addition, wild and cultivated ryes also have a great potential as a source for value-added trait genes such as those for high protein content, disease resistance, and other morphological and biochemical traits, for wheat (Triticum spp. L.) and triticale (×Triticosecale Wittmack) improvement. Despite its economical importance, taxonomy and phylogenetic relationships within the Secale genus have long been the subjects of controversy. Taxonomists have tried to discriminate each species within Secale using different approaches; however, identification of many taxa has been difficult because of a lack of diagnostic characters. As a result, many varying classification efforts have been reported. The first reports based on morphological characteristics, life cycle, and geographical distribution (Vavilov 1917, 1926) accepted 4 species in the genus Secale, for example, Secale africanum Stapf., S. cereale L., Secale fragile Marsch, and Secale montanum Guss. Roshevitz (1947) distinguished as many as 14 species based on their crossability. However, Khush (1962) did not find any cytogenetic support to classify perennial ryes (S. montanum, S. africanum, and Secale kuprijanovii Grossh.) as different species and proposed that they should be taken as subspecies of S. montanum, whereas the weedy ryes (Secale ancestrale, Secale afghanicum Vav., Secale dighoricum Vav., and Secale segetale Zhuk.) were to be considered subspecies of S. cereale. Lastly, Frederiksen and Petersen (1998) made a taxonomic revision of Secale based on examination of material in several herbaria and comments on the application of many of the species names that were used. They recognized only 3 species: Secale sylvestre, Secale strictum (syn. S. montanum), and S. cereale. Phylogenetic relationships among Secale species have also been studied using many different approaches, including morphological analyses (Frederiksen and Petersen 1997), isozymes (Vence et al. 1987), thin-layer chromatography patterns (Dedio et al. 1969), ribosomal DNA spacer lengths (Reddy et al. 1990), restriction fragment length polymorphism (RFLP) of plastid genome (Murai et al. 1989), and chloroplast DNA variation (Petersen and Doebley 1993), as well as the internal transcribed spacer sequences of the 18S– 5.8S rDNA (ITS-1) region of cultivated and wild species (De Bustos and Jouve 2002). In contrast with the classification systems, all phylogenetic studies demonstrated similar results. Secale sylvestre showed distinct characteristics and was the most distant species, whereas other taxa have been and continue to be more difficult to distinguish from each other. Amplified fragment length polymorphism (AFLP) analysis is a technique through which selected fragments from the digestion of total plant DNA are amplified by PCR (Vos et al. 1995). This technique produces DNA fingerprints that provide a large number of genetic markers, which can be used as a satisfactory alternative to morphological and biochemical trait analyses. Amplified fragment length polymorphism markers have been used for phylogenetic studies to uncover closely related taxa that had been impossible to resolve with morphological or other molecular systematic characteristics (Russell et al. 1997; Janssen et al. 1997; Mueller and Wolfenbarger 1999; Zhang et al. 2001; Hodkinson et al. 793 2000; Beardsley et al. 2003; Laurence et al. 2003). This method could be suitable for the analysis of relatedness, especially because AFLP markers are virtually free of artifacts, which is an acute problem of anonymous markers for relatedness estimation, and because comigration of nonallelic fragments occurs at extremely low levels (Waugh et al. 1997). Previous studies showed that AFLP technology also provided better resolution in discerning phylogenetic relationships as compared with isozymes, nuclear RFLPs and chloroplast DNAs (Sharma et al. 1996). However, the use of AFLPs for studying phylogenetic relationships in the genus Secale has not been previously reported. The present research was designed to demonstrate the use of AFLPs in discriminating Secale species and to understand the phylogenetic relationships among them. The objectives of the present study were to (i) apply AFLP technology to perform detailed analyses of polymorphism in the genus Secale, (ii) discriminate Secale species, and (iii) describe the phylogenetic relationships among Secale species based on AFLPs. Materials and methods Plant materials and DNA isolation Twenty-nine accessions of weedy/wild rye and cultivated rye were used (Table 1), representing 14 of the most commonly recognized taxa ranked as species or subspecies in the genus Secale. Seeds were kindly provided by the germplasm collections of the Plant Breeding and Acclimatization Institute (IHAR), Poland; Germplasm Resources Information Network (GRIN) of the United States Department of Agriculture (USDA); and International Maize and Wheat Center (CIMMYT), Mexico. Different numbers of accessions per species were used, because that is what was available from the germplasm collections. One accession from each of Agropyron cristatum (L.) Gaertn and Triticum monococcum L. were used as outgroups (Monte et al. 1995) and were obtained from the USDA-Sears collection, University of Missouri. All of the accessions are carefully maintained in seed form by the various collections and can be obtained from the collection’s curator. DNA was extracted from young freeze-dried tissue collected from 5–10 plants of each accession using a cetyltrimethylammonium bromide method (Saghai-Maroof et al. 1994). Five to ten plants were used based on preliminary results, which showed that 5–10 samples per accession represented the diversity within each species. AFLP analyses Amplified fragment length polymorphism analyses were carried out according to the technique developed by Vos et al. (1995). From each accession, a 500-ng sample of total genomic DNA was digested with the combination of EcoRI and MseI or PstI and MseI restriction enzymes and ligated with the respective adaptors for each enzyme. Preamplification reactions were performed using EcoRI and MseI or PstI and MseI primer combinations with 1 nucleotide extension being added to the 3′ end of each primer. The pre-amplification products were used as templates for selective amplification, which involved 3 additional nucleotides being placed at the 3′ end of EcoRI or PstI and MseI prim© 2005 NRC Canada 794 Genome Vol. 48, 2005 Table 1. List of species name, accessions number (existing germplasm bank location), country of origin, type, life cycle of Secale species studied. No. Accession No.* Species† 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 31209 (IHAR) 31082 (IHAR) 31319 (IHAR) 31344 (IHAR) CIse 107 PI 283971 (USDA) 31322 (IHAR) 31323 (IHAR) CIse 1 PI 534929 (USDA) PI 534962 (USDA) PI 534965 (USDA) PI 535008 (USDA) 30186 (IHAR) 31211 (IHAR) 31083 (IHAR) 31328 (IHAR) 31351 (IHAR) CIse 105 PI 283982 (USDA) 31359 (IHAR) 31366 (IHAR) 31367 (IHAR) PI 573649 (USDA) PI 383756 (USDA) PI 383757 (USDA) PI 205222 (USDA) PI 401404 (USDA) PI 440654 (USDA) S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. anatolicum Boiss. afghanicum (Vav.) Roshev. africanum Stapf ancestrale (Zhuk.) Zhuk. ancestrale (Zhuk.) Zhuk. ancestrale (Zhuk.) Zhuk. chaldicum Fed. ciliatoglume (Boiss.) Grossh. cereale L. cereale L. cereale L. cereale L. cereale L. cereale L. dighoricum (Vav.) Roshev. dighoricum (Vav.) Roshev. kuprijanovii Grossh. segetale (Zhuk.) Roshev. segetale (Zhuk.) Roshev. segetale (Zhuk.) Roshev. sylvestre Host turkestanicum Bensin vavilovii Grossh. vavilovii Grossh. montanum Guss. montanum Guss. montanum Guss. montanum Guss. montanum Guss. Country of origin Type‡ Life cycle Poland Afghanistan South Africa Turkey Japan Algeria Russia Federation Poland Sweden Italy US (Mississipi) US (Florida) Canada (Alberta) Poland Poland Poland Poland Russian Federation Italy Former USSR Poland Turkey Poland Afghanistan Turkey Turkey Turkey Iran Hungary Wi We Wi We We We Wi Wi C C C C C C We We Wi We We We Wi C Wi Wi Wi Wi Wi Wi Wi P A P A A A P P A A A A A A A A P A A A A A A A P P P P P *Accession numbers followed by existing germplasm bank location. †Botanical names of accessions that originated from Poland were verified using Secale monographs (Hammer et al. 1987; and Kobylyanskyi 1989). Botanical names of accessions that originated from GRIN were based on old names. ‡Wi, wild; We, weedy; A, annual; C, cultivated; and P, perennial. ers. The EcoRI or PstI primers were labeled with 33P prior to amplification. The amplification products were resolved on a 5% polyacrylamide sequencing gel in 1× Tris–borate–EDTA buffer for 2.5–3 h. The gels were dried for 2 h and exposed to X-ray film for 48 h. Data analysis Polymorphism was scored manually. DNA fragments located on the same migration position were read as 1 locus and scored 1 for present or 0 for absent. The data matrices for each primer combination were converted into Nei–Li’s or Dice similarity matrix and then compared among each other using Mantel’s test (Mantel 1967). The data matrices showing good relationships were then combined and used for cluster analysis. To detect potentially weakly supported lineages, 2 different algorithms were used in the data analysis. The 1st approach utilized phenetic analysis involving various similarity coefficients and clustering methods to obtain the shortest tree. Similarity coefficients and clustering methods were tested using similarity of qualitative data (SIMQUAL), sequential agglomerative hierarchical nested (SAHN) clus- tering routine, and TREE from program NTSYSpc version 2.1 (Rohlf 2000). The unweighted pair group method with arithmetic means (UPGMA), weighted pair group method with arithmetic means (WPGMA), complete-link, and singlelink clustering methods were applied in all possible combinations with both the Dice and Jaccard similarity coefficients. Cophenetic correlation coefficients (r) were calculated and compared for each of the combinations using COPH and MXCOMP from NTSYSpc 2.1 procedures. Principal coordinate analysis (PCO) was performed using the procedures DCENTER, EIGEN, and MXPLOT from NTSYSpc 2.1k. The 2nd approach involved a phylogenetic analysis using a maximum parsimony method that was performed with the PAUP* (phylogenetic analysis using parsimony) program (Swofford 1998). Bootstrap tests were performed using 5000 replications for phenetic analysis and 1000 replications for phylogenetic analysis to assess confidence in producing tree topologies (Felsenstein 1985). The results from both phenetic and phylogenetic analyses were compared to determine the best phylogenetic relationship among the species. Fisher’s exact test was used to test the correlation between life cycle © 2005 NRC Canada Chikmawati et al. 795 Table 2. Primer combination, total bands, polymorphic bands, level of polymorphisms, and average of r value with respect to the genus Secale. Primer combination Total bands Polymorphic bands Polymorphism (%) Average of r value* P-ACA/M-AAC P-ACA/M-ACT P-ACA/M-ATC P-ACA/M-CCA P-ACA/M-CCT P-AGT/M-AAC P-AGT/M-ACG P-AGT/M-AGA P-AGT/M-ATC P-AGT/M-ATG P-AGT/M-CCT Average 86 42 77 63 38 53 59 67 67 92 57 64 73 32 68 59 31 36 48 46 36 60 43 48 85 76 88 94 82 68 81 69 54 65 75 76 0.80 0.65 0.74 0.77 0.69 0.79 0.71 0.79 0.70 0.77 0.79 — E-ACA/M-ACG E-AGC/M-ACG E-CCT/M-CCA E-CCT/M-CAC E-CCT/M-CAG E-CCT/M-CCT E-CCT/M-CTG Average 54 64 75 72 51 63 50 61 33 39 48 40 33 34 30 38 61 61 64 56 65 56 60 60 0.72 0.73 0.80 0.75 0.67 0.77 0.60 — 1130 789 — — Total *r value from Mantel’s test. character and AFLP data (Fisher 1970). Genetic structure of the Secale genus was analyzed using Wright’s F test (Hartl and Clark 1997). Results Eighteen primer combinations were utilized in analyzing phylogenetic relationships among the various Secale species (Table 2). Amplified fragment length polymorphism analysis revealed a very large distinct number of fragments per primer pair in the members of the genus Secale (Fig. 1). The number of fragments present varied with each primer and ranged from 38 to 86. Each primer combination demonstrated approximately 54%–94% polymorphism. In total, 789 polymorphic markers of 1130 fragments were observed. Primer combinations P-/M- and E-/M- generally yielded similar numbers of fragments, 64 and 61 bands, respectively; however, P-/M- primer combinations produced a higher level of polymorphism (76%) than the E-/M- primer combinations (60%). The most efficient primer combination, which showed the largest number of polymorphic fragments, was the PACA/M-AAC (Table 2). The Mantel test was performed to establish the correlation among primer combinations before constructing phenetic and phylogenetic trees. The correlation values ranged from 0.60 to 0.80 (Table 2) and were significant in all primer combinations. The 789 polymorphic markers resulting from the 18 primer combinations were then used to construct phenetic and phylogenetic trees. Cophenetic correlation coefficients measured the goodness of fit of the cluster analysis to the similarity matrix, and the results showed that cophenetic correlation coefficients were >0.90 (Table 3) indicating that the goodness of fit for all combinations of the similarity matrix and cluster analysis were very good. Therefore, all those combinations were appropriate for use in analyzing the phenetic relationships. Among the cluster methods, UPGMA yielded the highest cophenetic correlation in all cases. The combinations of UPGMA with the Dice and Jaccard coefficients also yielded identical tree topologies. Therefore, they were considered the most suitable combinations for data analysis. The phenogram and the cladogram, from combinations of UPGMA and Dice similarity, had similar tree topologies (Figs. 2 and 3). In both cluster analyses, S. sylvestre was first separated (100% support), followed by Secale ciliatoglume (70% and 62% support in dendrogram and cladogram, respectively). The rest of the accessions were clustered into 2 major groups. Group III, S. montanum, Secale anatolicum, S. kuprijanovii, Secale chaldicum, and S. africanum, consisted of all perennial taxa (42% and 72% support in dendrogram and cladogram, respectively). Group II, S. ancestrale, S. afghanicum, S. cereale, S. dighoricum, Secale turkestanicum, S. segetale, and Secale vavilovii, consisted of annual taxa (91% and 37% support in dendrogram and cladogram, respectively). The results of PCO analysis showed that the first 2 axes accounted for 16.38% and 7.24% of the data variance (Fig. 4). Cumulatively, the PCO results represented 24% of the data, which were sufficient to resolve all the analyzed accessions into 3 distinct groups. Group 1 contained S. sylvestre and the 2 outgroups, group II contained the annual taxa, and group III contained the perennial taxa. The © 2005 NRC Canada 796 Genome Vol. 48, 2005 Fig. 1. An example of the AFLP polymorphism among Secale species, detected with the PstI-ACA and MseI-AAC primer combination. Table 3. Cophenetic correlation coefficients for amplified fragment length polymorphism data of all Secale species used. UPGMA, unweighted pair group method with arithmetic means; WPGMA, weighted pair group method with arithmetic means. Clustering/Similarity Jaccard Dice UPGMA WPGMA Complete-link Single-link 0.98 0.96 0.95 0.97 0.98 0.96 0.96 0.97 PCO results were congruent with the results of cluster analysis. Genetic differentiation among annual and perennial taxa was analyzed using the Wright’s F test (Table 4). Since S. sylvestre showed really distinct characters, this taxon was excluded from the analysis. The genetic differentiation level of annual taxa was smaller than that of perennial taxa; however, the genetic diversity within the annual taxa was much higher than that of perennial taxa. Discussion Phylogenetic relationships among Secale species The AFLP technique produced a wide range of variability among Secale taxa that was sufficient to clearly resolve all analyzed accessions into 3 major groups: group 1 contained of S. sylvestre, group II contained all of the perennial taxa, and group III contained all of the annual taxa. This indicated that the perennial versus annual life cycle probably played an important role in determining the relationships among the Secale species. Further analysis using Fisher’s exact test supported this by showing that 24% of the AFLPs detected were associated with the character life cycle. Our results supported the validity of 3 major series within Secale as recognized by Roshevitz (1947). Secale montanum and all perennial forms constituted 1 major group, the series Kuprijanovia Roshev.; S. cereale and all weedy annual relatives constituted the series Cerealia Roshev.; and S. sylvestre stood alone as an annual and constituted the series Silvestria Roshev. Among the annual taxa, S. cereale had a closer relationship to S. ancestrale, S. afghanicum, S. dighoricum, and S. segetale than to S. vavilovii. Even though both S. turkestanicum and S. cereale are cultivated plants, they had a distant relationship with each other. The breeding system differences between the 2 taxa (S. turkestanicum is selfpollinated and S. cereale is cross-pollinated) may offer a possible explanation. Among perennial taxa, S. ciliatoglume showed the most distant relationship from the others. Secale ciliatoglume is an isolated weedy population with pubescent culms that appear to be endemic to orchards and vineyards near Mardin, Turkey. It is possible that the very limited distribution of this taxon allowed it to maintain a distinct identity from the others. Within group II, S. africanum had the furthest relationship from S. montanum, whereas S. anatolicum and S. kuprijanovii had the closest relationship to S. montanum. Somewhat surprisingly, S. sylvestre was closer to the perennial taxa than to the annual taxa. Since S. sylvestre had the closest relationship to the outgroups, it can be considered as the most ancient among Secale species, and S. cereale can be considered as the youngest of the Secale species. Based on cytological, ecological, and morphological © 2005 NRC Canada Chikmawati et al. 797 Fig. 2. A dendrogram constructed from the AFLP data, using Nei–Li’s distance and UPGMA clustering. Numbers on the branches are bootstrap and range from 16% to 100%. studies, Stutz (1972) demonstrated that cultivated rye (S. cereale L.) originated from the weedy progeny derived from introgression of S. montanum (syn. S. strictum) into S. vavilovii, S. africanum, Secale dalmaticum, S. ciliatoglume, and S. kuprijanovii as they had close relationships with each other and appeared to be only slightly modified isolated populations of S. montanum. Populations of S. anatolicum were thought to be weedy forms of S. montanum, genetically and chromosomally distinct from the weedy annual forms. In general, the species relationships within genus Secale as based on AFLP data were consistent with Stutz (1972). However, Stutz (1972) also suggested that S. montanum was the common ancestor of all the Secale species, which conflicts with the present AFLP data. The present data clearly demonstrated that S. sylvestre was the most ancient species that split off first from the common ancestor, whereas S. montanum split off after the separation of S. sylvestre. Molecular taxonomy of Secale based on AFLP Secale sylvestre is a low growing plant with a fragile © 2005 NRC Canada 798 Genome Vol. 48, 2005 Fig. 3. A cladogram resulting from the maximum parsimony method that was conducted using heuristic search methods with TBR branch swapping, collapse of zero-options, and weighting of all characters equally. Numbers on branches are bootstrap values and range from 5% to 100%. Consistency index (CI) = 0.18, retention index (RI) = 0.35, rescale consistency index (RCI) = 0.06. rachis, widely distributed from central Hungary eastward throughout the sandy steppes of southern Russia. This taxon can be easily distinguished from other taxa by its long awned glumes (Stutz 1972). Khush and Stebbins (1961) showed that S. sylvestre was cytogenetically very distant from S. cereale, and was geographically, ecologically, and reproductively isolated from S. montanum (Sencer and Hawkes 1980). In addition, S. sylvestre also has other unique characteristics, such as distinct chloroplast DNA (Petersen and Doebley 1993), a spacer length variant of the ribosomal DNA (Reddy et al. 1990), and an internal transcribed spacer of the 18S–5.8S–26S rDNA (ITS-1) region, as compared with the other Secale species. Given the strong distinction of S. sylvestre from other taxa, it is easy to consider S. sylvestre as a distinct species. Amplified fragment length polymorphism analysis showed that S. sylvestre also demonstrated a © 2005 NRC Canada Chikmawati et al. 799 Fig. 4. Principal coordinate analysis of Secale species based on AFLPs. I = S. sylvestre and 2 outgroups; II = annual taxa; and III = perennial taxa. Table 4. Total genetic diversity, genetic diversity within population, genetic differentiation and genetic identity of Secale accessions. Ht, total genetic diversity; Hs, genetic diversity within population; Gst, genetic differentiation; GI, genetic identity. Groups Ht Hs Gst GI Annual taxa Perennial taxa 0.25±0.03 0.30±0.03 0.09±0.0062 0.02±0.0009 0.63 0.90 0.68–0.89 0.24–0.48 very distinct profile in all primer combinations, and it was well separated from others in all analyses. Thus, AFLP analyses confirm S. sylvestre as a distinct species. The presence of a high degree of similarity among wild, weedy annual forms and cultivated rye has been demonstrated. Khush (1963) showed that there was no evidence of structural differences between the genome of cultivated rye and several weedy ryes (S. cereale, S. vavilovii, S. ancestrale, S. afghanicum, S. dighoricum, and S. segetale), which had previously been recognized as varieties, subspecies, or even species. They all readily crossed and produced vigorous F1s, which had similar chromosome arrangements, breeding habit, periodicity, and crossability, and they also demonstrated geographical continuity. Therefore, Khush (1963) proposed all annual forms are subspecies of S. cereale. Further study based on morphometrical analyses concluded that it was impossible to recognize each annual taxon based on their morphology (Frederiksen and Petersen 1997). They proposed 2 intraspecific taxa within a single species (S. cereale), which are S. cereale subsp. cereale for cultivated rye and S. cereale subsp. ancestrale for weedy and wild annual rye taxa. The most recent study based on ITS-1 region also found no differences between the weedy forms and cultivated rye (Bustos and Jouve 2002). Thus, all previous studies demonstrated that morphologically and genetically all annual taxa were similar to each other and that it was impossible to discriminate them clearly. The AFLP results showed that 6 accessions of S. cereale, which originated from different locations, composed a monophyletic group. The S. dighoricum accessions also clustered together, which was not too surprising since they both originated from the same location. Thus, it is highly possible that those accessions are duplicated accessions, whereas other annual taxa that were represented by more than 1 accession did not cluster together. Principal coordinate analyses intermingled them with each other. Nei–Li’s distances yielded very small genetic-distance values among them indicating they had a high degree of AFLP similarity. This was also shown from the intermediate level of genetic differentiation (Gst = 0.63). Except for cultivated rye, it was still difficult to discriminate between wild and weedy rye using AFLPs, therefore, AFLP analysis supported Frederiksen and Petersen (1997). Among perennial species, only S. ciliatoglume did not cluster together with the others. Secale ciliatoglume stood alone between annual and perennial taxa in cluster analyses, but the separation was intermediately supported (62% in phylogenetic and 70% in phenetic analysis). The PCO analysis placed this accession in the same quadrant with the other perennial taxa. The information about S. ciliatoglume from previous studies was very limited, and only dealt with morphological data. This taxon has been shown to be morphologically similar to S. montanum, and only deviated by having a dense cover of hairs over the internodes, leaf sheaths, and blades (Frederiksen and Petersen 1998). Frederiksen and Petersen (1997) suggested that S. ciliatoglume should be given an intraspecific rank. Stutz (1972) demonstrated that several perennial forms (S. anatolicum, S. africanum, S. dalmaticum, and S. mon© 2005 NRC Canada 800 tanum) readily crossed to each other, and that crossing among them yielded normal chromosome configurations indicating no reproductive barrier. Khush (1962) proposed all perennial taxa as subspecies of S. montanum. Furthermore, Sencer and Hawkes (1980) showed that all the wild perennial forms had a fairly common morphological resemblance. Our analysis, based on AFLPs, clustered the 5 accessions of S. montanum together with low bootstrap support (47% in cladogram). The separation of the other perennial taxa was with intermediate bootstrap support (47%–74% in cladogram). However, the genetic differentiation level among perennial taxa was very high (Gst = 0.90). This result suggests that AFLP marker polymorphism levels within the perennial taxa were sufficient to discriminate and place them in an intraspecific rank, instead of in an interspecific rank. In summary, this is the first phylogenetic study of the genus Secale based on AFLP analyses. Analyzing AFLPs among 29 accessions representing 14 of the most recognized ranked species or subspecies in the genus Secale demonstrated that AFLP marker technology is a better tool for analyzing phylogenetic relationships among Secale species, because it produced high levels of polymorphism that were sufficient to resolve all accessions into 3 distinct groups. Group I contained S. sylvestre, group II contained all perennial taxa, and group III contained all annual taxa. (Secale ciliatoglume stood alone between annual and perennial taxa in cluster analyses.) The phylogenetic relationships among Secale taxa based on AFLPs strongly supported Stutz (1972). The AFLP data were clearly able to distinguish 3 Secale species: S. sylvestre, S. cereale, and S. montanum. In addition, AFLPs were also able to recognize 2 intraspecific taxa of S. cereale, namely S. cereale subsp. cereale for cultivated rye and S. cereale subsp. ancestrale for wild and weedy rye annual taxa, and 6 intraspecific taxa of S. montanum, namely S. montanum subsp. montanum; S. montanum subsp. africanum; S. montanum subsp. anatolicum; S. montanum subsp chaldicum; S. montanum subsp. ciliatoglume; and S. montanum subsp. kuprijanovii. These results were consistent with the last revision of the genus Secale by Frederiksen and Petersen (1998) who recognized only 3 species within the genus Secale: S. sylvestre, S. strictum (syn. S. montanum), and S. cereale. Acknowledgements The authors thank Miftahudin for help and advice on data analysis and Kathleen Ross for helpful discussions on plant preparation. We are also grateful to Dr. J. Bockelman from Germplasm Resources Information Network (GRIN) of the United States Department of Agriculture, Dr. M. Niedzielski and Dr. W. Podima from the Plant Breeding and Acclimatization Institute (IHAR), Poland, for supplying us with seeds. Tatik Chikmawati is a Ph.D. student funded by the Quality for Undergraduate Education (QUE) project from the Department of Biology, Bogor Agricultural University, Indonesia. References Beardsley, P.M., Yen, A., and Olmstead, R.G. 2003. AFLP phylogeny of Mimulus section Erythranthe and the evolution of hummingbird pollination. Evol. 57(6): 1397–1410. Genome Vol. 48, 2005 De Bustos, A., and Jouve, N. 2002. Phylogenetic relationships of the genus Secale based on the characterization of rDNA ITS sequences. Plant Syst. Evol. 235: 147–154. Dedio, W., Kaltsikes, P.J., and Larter, E.N. 1969. Numerical chemotaxonomy in the genus Secale. Can. J. Bot. 47: 1175– 1180. Felsenstein, J. 1985. Confidence limit on phylogenies: an approach using the bootstrap. Evolution, 39: 783–79. Fisher, R.A. 1970. Statistical methods for research workers. 14th ed. Oliver and Boud, Edinburgh, U.K. Frederiksen, S., and Petersen, G. 1997. Morphometrical analyses of Secale (Triticeae, Poaceae). Nord. J. Bot. 17: 185–197. Frederiksen, S., and Petersen, G. 1998. A taxonomic revision of Secale. Nord. J. Bot. 18: 399–420. Hammer, K., Skolimowska, E., and Knueffer, H. 1987. Vorarbeiten zur monographischen Darstellung von Wildpflanzensortimenten: Secale L. Kulturplanze, 35: 135–177. [In German.] Hartl, D.L., and Clark, A.G. 1997. Principles of population genetics. Sinauer Associates, Inc., Suntherland, Mass., U.S.A. Janssen, P., Maquelin, K., Coopman, R., Tjernberg, I., Bouvet, P., Kersters, K., and Dijkshoorn, L. 1997. Discrimination of Acinetobacter genomic species by AFLP fingerprinting. Int. J. Syst. Bacteriol. 47: 1179–1187. Khush, G. 1962. Cytogenetic and evolution studies in Secale II: interrelationships of the wild species. Evolution, 16: 484–496. Khush, G. 1963. Cytogenetic and evolutionary studies in Secale III: cytogenetics of weedy ryes and origin of cultivated rye. Econ. Bot. 17: 60–71. Khush, G., and Stebbins, G.L. 1961. Cytogenetic and evolution in Secale I: some new data on the ancestry of S. cereale. Am. J. Bot. 48: 723–730. Kobylyanskyi, V.D. 1989. Flora of cultivated plants of USSR. Vol. II. Part 1. Agronomizdat, Leningrad, Russia. Laurence, D., Ludovic, G., Benedicte, R., and Pierre, T. 2003. Using AFLP to resolve phylogenetic relationships in a morphologically diversified plant species complex when nuclear and chloroplast sequences fail to reveal variability. Mol. Phylogenet. Evol. 27: 185–196. Leonard, W.H., and Martin, J.H. 1963. Cereal crops. The Macmillan Company, New York. Mantel, N. 1967. The detection of disease clustering and a generalized regression approach. Cancer Res. 27: 209–220. Monte, J.V., McIntyre, C.L., and Gustafson, J.P. 1995. Analysis of phylogenetic relationships in the Triticeae tribe using RFLPs. Theor. Appl. Genet. 86: 649–655. Mueller, U.G., and Wolfenbarger, L.L. 1999. AFLP genotyping and fingerprinting. Trends Ecol. Evol. 14: 389–394. Murai, K., Naiyu, X., and Tsunewaki, K. 1989. Studies on the origin of crop species by restriction endonuclease analysis of organellar DNA. III. Chloroplast DNA variation and interspecific relationships in the genus Secale. Japan. J. Genet. 64: 36–47. Petersen, G., and Doebley, F. 1993. Chloroplast DNA variation in the genus Secale (Poaceae). Plant Syst. Evol. 187: 115–125. Reddy, P., Appel, R., and Baum, B.R. 1990. Ribosomal DNA spacer-length variation in Secale spp. (Poaceae). Plant Syst. Evol. 171: 205–220. Rohlf, F.J. 2000. NTSYSpc. Numerical taxonomy and multivariate analysis system. Version 2.1. Exeter Software, Setauket, New York. Roshevitz, R.I. 1947. Monographie of the genus Secale L. Acta Inst. Botanici Acad. Scientiarum U.R.S.S. I(6): 105–163. Russell, J.R., Fuller, J.D., Macaulay, M., Hatz, B.G., Jahoor, A., Powell, W., and Waugh, R. 1997. Direct comparison of levels of © 2005 NRC Canada Chikmawati et al. genetic variation among barley accessions detected by RFLPs, AFLPs, SSRs, and RAPDs. Theor. Appl. Genet. 95: 714–722. Saghai-Maroof, M.A., Biyashev, R.M., Yang, G.P., Zhang, Q., and Allard, R.W. 1994. Extraordinarily polymorphic microsatellite DNA in barley: species diversity, chromosomal locations, and population dynamics. Proc. Natl. Acad. Sci. U.S.A. 91: 5466–5470. Sencer, H.A., and Hawkes, J.G. 1980. On the origin of cultivated rye. Biol. J. Linn. Soc. 13: 299–313. Sharma, S.K., Knox, M.R., and Ellis, T.H.N. 1996. AFLP analysis of the diversity and phylogeny of Lens and its comparison with RAPD analysis. Theor. Appl. Genet. 93: 751–758. Stutz, H.C. 1972. On the origin of cultivated rye. Am. J. Bot. 59: 59–70. Swofford, D.L. 1998. Phylogenetic Analysis Using Parsimony (PAUP*). Version 4.0b10 [computer program]. Sinauer Associates, Sunderland, Mass., U.S.A. Vavilov, N.I. 1917. On the origin of cultivated rye. Bull. Appl. Bot. Genet. Plant Breed. 10: 561–590. 801 Vavilov, N.I. 1926. Studies on the origin of cultivated plants. Bull. Appl. Bot. Genet. Plant Breed. 16: 1–248. Vence, F.J., Vaquero, F., and Perez de la Vega, M. 1987. Phylogenetic relationships in Secale (Poaceae): an isozymatic study. Plant Syst. Evol. 157: 33–47. Vos, P., Hogers, R., Blecker, M., Reijans, M., van de Lee, T., Hornes, M., Frijters, A., Peleman, J., Kuiper, M., and Zabeau, M. 1995. AFLP: a new technique for DNA fingerprinting. Nucleic Acids Res. 23: 4407–4414. Waugh, R., Bonar, N., Baird, E., Thomas, B., and Graner, A. 1997. Homology of AFLP products in three mapping populations of barley. Mol. Gen. Genet. 255: 311–321. Zhang, L.-B., Comes, H.P., and Kadereit, J.W. 2001. Phylogeny and quaternary history of the European montane/alpine endemic Soldanella (Primulaceae) based on ITS and AFLP variation. Am. J. Bot. 88: 2331–2345. © 2005 NRC Canada