Crop Science Illumina
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 30 July 2007
Published in Crop Sci 47:1744-1753 (2007)
© 2007 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tamiru, M.
Right arrow Articles by Maass, B. L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Tamiru, M.
Right arrow Articles by Maass, B. L.
Agricola
Right arrow Articles by Tamiru, M.
Right arrow Articles by Maass, B. L.
Related Collections
Right arrow Other Tubers
Right arrow Plant Genetic Resources

PLANT GENETIC RESOURCES

Genetic Diversity in Yam Germplasm from Ethiopia and Their Relatedness to the Main Cultivated Dioscorea Species Assessed by AFLP Markers

Muluneh Tamirua,c, Heiko C. Beckerb and Brigitte L. Maassa,*

a Dep. of Crop Sciences: Agronomy in the Tropics, Georg-August-Univ. Göttingen, Grisebachstr. 6, D-37077 Göttingen, Germany
b Dep. of Crop Sciences: Plant Breeding, Georg-August-Univ. Göttingen, Von-Siebold-Str. 8, D-37075 Göttingen, Germany
c Hawassa Univ., P.O. Box 05, Awassa, Ethiopia

* Corresponding author (bmaass{at}gwdg.de).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Yam (Dioscorea spp.) is widely cultivated in southern Ethiopia, where farmers maintain many named landraces. Nevertheless, little is known about their diversity and species identity. In this study, amplified fragment length polymorphism (AFLP) markers were used to investigate the genetic diversity in 48 yam accessions from Ethiopia, and to assess their relatedness to yam species commonly cultivated in West Africa such as D. alata L., D. bulbifera L., D. cayenensis Lam., and D. rotundata Poir. Ten AFLP primer combinations generated 900 fragments, of which 97% were polymorphic. Cluster and principal coordinate analyses revealed that the Ethiopian accessions are distinct from the Dioscorea species widely cultivated in West Africa. A separate analysis of the Ethiopian accessions gave six clusters that represented the various maturity groups and the nonflowering accession in the collection. Analysis of molecular variance (AMOVA) showed that 81% of the variation detected was found within collecting areas, while the variation among collecting areas contributed only 19%. The groups detected by AFLP markers were highly consistent with the local yam classification system and also reflected the main structure of morphological diversity.

Abbreviations: AFLP, amplified fragment length polymorphism • AMOVA, analysis of molecular variance • PCR, polymerase chain reaction • RAPD, random amplified polymorphic DNA • UPGMA, unweighted pair-group method using arithmetic means algorithm


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
YAMS (Dioscorea spp.) are important crop plants widely distributed throughout the humid and subhumid tropics (Coursey, 1967). The major food species originated in three isolated regions: Southeast Asia, West Africa, and Tropical America, which are also considered main centers of yam domestication and diversity (Asiedu et al., 1997). Yam has great economic and social significance in sub-Saharan Africa, where more than 95% of the world yam is produced (Degras, 1993). Ethiopia is an isolated center of yam cultivation in East Africa (Norman et al., 1995), and the crop plays a vital role in local livelihood particularly in the densely populated areas of southern, southwestern, and western parts of the country (Hildebrand et al., 2002; Tamiru, 2006).

Yam shows considerable diversity both at inter- and intraspecific levels (Okoli, 1991). The diversity under cultivation is further enhanced by the ongoing domestication of wild yam in various countries (Mignouna and Dansi, 2003; Scarcelli et al., 2006). Nevertheless, the extent of genetic diversity in many Dioscorea species and their relationships is yet to be investigated in detail. Attempts to characterize yam using morphological (Hamon and Touré, 1990b; Dansi et al., 1999) and isozyme (Hamon and Touré, 1990a; Dansi et al., 2000b) markers did not give conclusive results due to their high degree of variability. Chromosome counts are also variable in yams, ranging from 2n = 20 to 2n = 140 in the common food species (Hahn, 1995).

Molecular markers such as restriction fragment length polymorphism (RFLP), random amplified polymorphic DNA (RAPD), and amplified fragment length polymorphism (AFLP) have been applied in yams for taxonomic, phylogenetic, and diversity studies (Terauchi et al., 1992; Asemota et al., 1996; Ramser et al., 1996, 1997; Mignouna et al., 1998; Dansi et al., 2000a). This enabled detection of differences among cultivars that were considered to be similar based on morphological and isozyme markers, and demonstrated their usefulness as discriminative tools in yam (Dansi et al., 2000a). Efforts are now under way for the broader application of molecular markers in genetic improvement of the crop (Mignouna et al., 2003b). The AFLP technique (Vos et al., 1995) combines the use of restriction enzymes and polymerase chain reaction (PCR). This method has been successfully used in yams for diversity studies (Mignouna et al., 1998; Malapa et al., 2005) and construction of genetic linkage maps (Mignouna et al., 2002). Compared to other markers such as RAPD, the AFLP technique generated more polymorphic markers in yams, providing a robust DNA fingerprinting technique for yam genomic analysis, including the detection of duplicates in germplasm collections (Mignouna et al., 2003a).

The diversity of yams in Ethiopia is poorly understood. A previous study revealed the presence of a substantial number of named landraces with varying degree of abundance and distribution in Wolayita and Gamo-Gofa zones, southern Ethiopia (Tamiru et al., 2007). Most of these named landraces represented phenotypically distinct materials, and the overall structure of morphological diversity was largely consistent with the local yam classification system (Tamiru et al., unpublished data, 2007). The actual diversity at DNA level is, however, yet to be investigated. Moreover, the use of standard descriptors did not allow accurate classification of the majority of the landraces from Ethiopia into any of the known cultivated Dioscorea species (S. Demissew, personal communication, July 2004; Tamiru, 2006).

The main objectives of this study were to use AFLP markers for a detailed analysis of genetic diversity within yam accessions collected from southern Ethiopia, and to determine the species of these accessions by including elite genotypes representing the main cultivated Dioscorea species as reference materials. Previously, the use of molecular markers in yam diversity studies mostly concerned materials from West Africa. Therefore, as the first of its kind on yams from Ethiopia, this study strives to generate information that is crucial in guiding improvement and conservation programs in the country.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Plant Materials
Sixty-two yam accessions were considered in this study (Table 1). Of these, 53 accessions were selected among the materials assembled from the major yam growing areas in southern Ethiopia and previously used for morphological characterization (Tamiru et al., unpublished data, 2007). The materials represent the range of variability that exists within yam in the collecting area for morphological traits and maturity characteristics, as well as representing numerous different landraces based on traditional classification. Five accessions belong to the species of aerial yam (D. bulbifera L.), while the identity of the remaining 48 accessions could not be established using conventional taxonomic methods for Dioscorea species. An additional nine elite genotypes, representing three major cultivated yam species, were obtained from the International Institute of Tropical Agriculture (IITA) in Nigeria and included as reference materials.


View this table:
[in this window]
[in a new window]

  		Table 1. Yam accessions studied.

 
The Ethiopian accessions were transported to Germany as seed tubers and were established in a greenhouse of the Department of Crop Sciences at Georg-August-University, Göttingen. The genotypes from IITA were received as tissue culture plantlets and were transferred to fresh medium for furnishing young leaves for DNA extraction.

DNA Isolation
Total genomic DNA was extracted from 100 mg fresh leaf samples using Nucleon PhytoPure plant and fungal DNA extraction kit for small samples (Amersham Biosciences, Freiburg, Germany) following the extraction and purification protocols of the manufacturer with the following minor modifications. Mercaptoethanol and RNase were added to Reagent 1 at concentrations of 10 mM and 20 µg mL–1, respectively. Following precipitation of the DNA with cold isopropanol, samples were stored overnight at 4°C. After DNA was resuspended in TE buffer (10 mM Tris-HCl pH 8.0, 1mM EDTA), samples were incubated at 65°C in a shaking water bath for 1 h to ensure a good resuspension. DNA was quantified using a Versa Fluro flurometer (Bio-Rad Laboratories, Hercules, CA) with flurochrome Hoechst 33258 stain. DNA quality was checked on 1% agarose gels prepared with TAE buffer (40 mM Tris-acetate, 1 mM EDTA, pH 8.0), and run for 2 h at 100 V. Samples were finally diluted to a standard concentration of 50 ng µL–1 with TE buffer.

AFLP Analysis
Restriction–Ligation
AFLP analysis was performed mostly following the procedures described by Vos et al. (1995). Approximately 250 ng DNA sample was digested with 4 units of both the restriction endonucleases EcoRI and MseI for 1.5 h at 37°C in a final reaction volume of 30 µL containing 1x restriction–ligation buffer (10 mM Tris-acetate acid, 10 mM magnesium acetate, 50 mM potassium acetate, 5 mM DTT, pH 7.5).

Ligation of adapters followed immediately using 5 pmol EcoRI and 50 pmol MseI adapters. The adaptor ligation mixture contained 30 µL of the restriction digestion aliquot to which 1 U T4 DNA Ligase (Promega GmbH, Mannheim, Germany), 1x restriction–ligation buffer, PCR grade water, and 0.25 mM ATP were added giving a final reaction volume of 40 µL. The ligation was performed in a Biometra T-Gradient thermo cycler (Biometra, Göttingen, Germany) with a program of 37°C for 3 h and 10 min, 33.5°C for 3 min, 30°C for 3 min, 26°C for 4 min, and 22°C for 15 min. An aliquot of the digested–ligated template DNA was diluted 1:5 with TE buffer, and 8 µL of the dilution was used as a template in the preamplification reaction.

Preamplification
Preamplification reactions were performed with primers having single selective nucleotides at their 3' end (E01 and M02) (Table 2). The 20 µL reaction mix was made of 8 µL of the digested and ligated template DNA and 12 µL preamplification mix containing 0.3 mM dNTPs, 1.5 U Taq-DNA-polymerase (Solis BioDyne, Tartu, Estonia), 1x PCR buffer (10 mM Tris-HCl, 1.5 mM MgCl2, 50 mM KCl, pH 8.3), 10 pmol EcoRI (A-3), and 8.7 pmol MseI (C-3). From 25 mM stock solution, 2.5 mM MgCl2 was added to bring the final concentration of MgCl2 to 4 mM. The temperature–time profile of the cycles was an initial 94°C for 30 s for denaturing, and then 20 cycles of 94°C for 30 s denaturing, 56°C for 30 s annealing, and 72°C for 60 s extension. The preamplification product was finally diluted 1:10 with TE buffer and used in the amplification reaction.


View this table:
[in this window]
[in a new window]

  		Table 2. Adapters and primers used in the study of Dioscorea spp. and their sequences.

 
Amplification
Ten primer combinations, including those previously used in yam genome analysis (Mignouna et al., 1998, 2002, 2003a), were employed in this study. Amplification reactions were performed using 2 pmol and 7 pmol EcoRI and MseI primers, respectively, having three selective nucleotides at their 3' ends (Table 2). EcoRI primers were labeled with an infrared fluorescent dye (IRD-800, MWD-Biotech AG, Ebersberg, Germany). The 20-µL reaction mixture contained 14 µL of amplification mix (0.24 mM dNTPs, 0.6 U Taq-DNA-polymerase, 1x PCR-buffer, and 4 mM MgCl2) and 6 µL of the diluted preamplification product as a template. The PCR was performed with initial denaturation at 94°C for 30 s, one cycle with a temperature–time profile of 94°C for 30 s (denaturing), 65°C for 30 s (annealing), and 72°C for 60 s (extension) followed by 11 cycles of touch down protocol with a similar temperature–time profile as in the previous cycle but the annealing temperature was lowered by about 0.7°C in each cycle. This was followed by a further 24 cycles with a temperature–time profile of 94°C for 30 s, 56°C for 30 s, and 72°C for 60 s (extended by 1 s per cycle).

Gel Electrophoresis
Gel electrophoresis and detection of the AFLP amplification products were performed on an automated DNA sequencer (LI-COR 4200 IR2, LI-COR GmbH, Bad Homburg, Germany). The AFLP fragments were mixed with a loading dye (98% [v/v] formamide, 10 mM EDTA, 0.025% bromophenol blue, and 0.025% xylene cyanol) at a 2:1 ratio. The mixtures were denatured for 4 min at 95°C and then quickly cooled on ice before loading. The fragments were resolved on 6% denaturing polyacrylamide gels containing polyacrylamide (acrylamide and bisacrylamide), 1.386 M urea (NF-Urea Rotiphore, Karl Roth GmbH, Karlsruhe, Germany), 10x TBE buffer (1.34 mM Tris-HCl, 450 mM boric acid, 25 mM EDTA, pH = 9.2), and 12% Long Ranger (50% gel solution; Biozym Scientific, Oldendorf, Germany). Polymerization was started by the addition of 0.01% TEMED. The gels were prerun (45°C, 1000 V, and 37 mA) for 15 min before loading the samples. About 1.4 µL of each sample was loaded and fragment mobility measured by a real-time laser for 6.5 h using the same temperature, current, and voltage profiles as in the prerun. All 62 samples were run on the same gel and, thus, one gel was used per primer combination. To facilitate data scoring, a 50- to 700-bp DNA sizing standard (LI-COR GmbH) was used. Gel images were stored electronically for further analysis.

Data Scoring and Analysis
Polymorphic bands were manually scored as 1 (present) and 0 (absent) with the help of the Adobe Photoshop software (Adobe Photoshop 7.0, Adobe Systems Inc., San Jose, CA). A band was considered polymorphic if it was present in at least one accession and absent in others. Mostly, clearly scorable bands were considered. In a very few cases, bands that could not be clearly considered as present because of low intensity were given a score of 9 and treated as a missing data point later in the analysis.

The data matrix for all the 62 accessions and the 10 primer combinations, excluding monomorphic bands, was used to calculate pairwise genetic similarity based on Jaccard coefficient (GSJ). The resulting similarity matrix was subjected to clustering by applying the unweighted pair group method using arithmetic means algorithm (UPGMA), and principal coordinate analysis with the help of the computer program NTSYSpc Version 2.1 (Rohlf, 2000). The matrix was subjected to bootstrapping with the software WinBoot (Yap and Nelson, 1996) using 1000 randomly drawn samples to assess the solidity of the genetic relationships among the groups constituted by cluster analysis. To test the goodness of fit between the UPGMA dendrogram and the original similarity matrix, the cophenetic correlation coefficient was calculated by means of the MXCOMP function of the NTSYSpc. The 48 unidentified accessions from Ethiopia were further subjected to similar analyses separately based on 165 polymorphic bands out of the total 376 bands present within these accessions. Moreover, analysis of molecular variance (AMOVA) was performed with the help of the GenAlEx6 software (Peakall and Smouse, 2006) using collecting area (i.e., zones) as a grouping criterion.


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
AFLP polymorphism
AFLP fingerprinting of the 62 accessions with 10 EcoRI/MseI primer combinations detected a substantial level of polymorphism (Table 3). Overall, 900 fragments were amplified, of which 877 (97.4%) were polymorphic. Total number of fragments scored per primer combination varied from 73 to 119, while the proportion of polymorphic fragments ranged from 94.5 to 100.0%. The size of fragments scored ranged from about 50 to 600 nucleotides. The polymorphism detected within the unidentified Ethiopian accessions and the other Dioscorea groups studied is summarized in Table 4. The proportion of polymorphic bands detected per primer combination within the accessions from Ethiopia varied from 34.2 to 61.8%. On average, the primers used revealed a relatively higher proportion of polymorphic bands within the unidentified accessions from Ethiopia (44%) than within the accessions of D. rotundata Poir. (40%) or D. cayenensis Lam. (38%). A very low level of polymorphism was detected within the accessions of D. bulbifera (6%). Nevertheless, this comparison is not exhaustive due to unequal group sizes.


View this table:
[in this window]
[in a new window]

  		Table 3. Total number of fragments scored and level of polymorphic bands detected in Dioscorea spp. with 10 amplified fragment length polymorphism (AFLP) primer combinations.

 

View this table:
[in this window]
[in a new window]

  		Table 4. The level of polymorphism revealed with 10 amplified fragment length polymorphism (AFLP) primer combinations within the unidentified yam accessions from Ethiopia and three known Dioscorea species.

 
Genetic Diversity within and between Groups
The comparison of pairwise genetic similarity values revealed broad genetic diversity within the materials studied (data not shown). About 52% of these comparisons represented genetic similarities greater than 0.8, 26% showed similarities less than 0.4, while the remaining 22% had values between 0.4 and 0.8. Overall, pairwise genetic similarity values ranged from 0.088 to 1.000 with a mean of 0.645. The accessions AKA 001 (D. bulbifera) and TDr-932 (D. rotundata) were the most dissimilar, whereas WOL 014 and WOL 016, and SID 011, SID 017 and AKA 004 appeared genetically identical.

The mean genetic similarities among and within the various Dioscorea species studied are presented in Table 5. The mean intraspecific genetic similarity varied between 0.734 and 0.973 for D. cayenensis and D. bulbifera accessions, respectively. The most similar species were D. rotundata and D. cayenensis with a mean similarity coefficient of 0.552. The Ethiopian accessions were genetically closer to D. rotundata (0.419) and D. cayenensis (0.362), while D. bulbifera was relatively closer to D. alata L. (0.275) than to any of the other species considered.


View this table:
[in this window]
[in a new window]

  		Table 5. Matrix of mean genetic similarity estimates (Jaccard coefficient) among and within the unidentified accessions from Ethiopia and the other Dioscorea species studied (numbers in parenthesis refer to within group genetic similarity ranges).

 
The genetic similarity within the Ethiopian accessions ranged from 0.695 to 1.000. Analysis of molecular variance for these accessions, using collecting area as a grouping criterion, revealed that within group variance accounted for 81% of the total variance, while the variation among collecting areas contributed only 19% (Table 6).


View this table:
[in this window]
[in a new window]

  		Table 6. Analysis of molecular variance (AMOVA) of 165 polymorphic amplified fragment length polymorphism (AFLP) markers from 48 unidentified yam accessions from Ethiopia.

 
Phenetic Analysis
UPGMA clustering based on Jaccard similarity coefficient separated the 62 accessions into five distinct clusters at about 60% of genetic similarity (Fig. 1 ). Each cluster was supported by a bootstrap value greater than 99% based on 1000 permutations. The existence of distinct clusters was also confirmed by the high cophenetic correlation coefficient (r = 0.99). The unidentified accessions from Ethiopia formed a distinct group, first cluster. The second cluster represented all the genotypes of D. rotundata, whereas those of D. cayenensis were clearly separated into the third cluster. TDa-9801174 (D. alata) was distinct from all the other accessions. Accessions of aerial yam (D. bulbifera) were grouped separately into the fifth cluster.


Figure 1
View larger version (33K):
[in this window]
[in a new window]

  		Figure 1. Dendrogram of 62 Dioscorea accessions evaluated based on Jaccard similarity coefficient and unweighted pair-group method using arithmetic means algorithm (UPGMA) clustering. Numbers in percentage indicate bootstrap values based on 1000 replicate analyses; cophenetic correlation coefficient, r = 0.99.

 
A separate clustering of the 48 unidentified accessions from Ethiopia based on 165 polymorphic AFLP fragments gave six distinct groups at about 70% genetic similarity, and three of these groups represented individual accessions (Fig. 2 ). All the groups were supported by high bootstrap values (>99%). Cluster 1 included 12 of the 14 accessions representing the late-maturing (single-harvested) landraces. Accession WOL 012, known by the local name ayina, was clearly distinct from the other late-maturing accessions. The same was true for the accession GGF 004. Cluster 4 comprised all accessions of the early-maturing (double-harvested) landraces and five of the accessions obtained from Areka Agricultural Research Center (AKA). Data on maturity time was missing for the latter group. Cluster 5 represented WOL 004b (one of the accessions with distinct panicle inflorescence) and AKA 013 (one of the three accessions with female inflorescence). The other accessions with female inflorescence were GGF 004 (cluster 3) and AKA 004 (cluster 4). The only nonflowering accession in the collection, AKA 014, was clearly separated from all the other accessions. Of the 165 polymorphic AFLP bands used for this clustering, 25 were specific to AKA 014.


Figure 2
View larger version (30K):
[in this window]
[in a new window]

  		Figure 2. Dendrogram of 48 unidentified Dioscorea accessions from Ethiopia based on Jaccard similarity coefficient and UPGMA clustering. Numbers in percentage indicate bootstrap values based on 1000 replicate analyses; cophenetic correlation coefficient, r = 0.99.

 
The relationship among the 62 accessions studied was further illustrated by results of the principal coordinate analysis. The first three principal coordinates accounted for 40.8, 18.9, and 12.6% of the total variance. Plotting of the first and second coordinates clearly separated accessions of D. alata and D. bulbifera, as well as those from Ethiopia (Fig. 3 ), and was highly consistent with results of the cluster analysis (Fig. 1). Although the species D. rotundata and D. cayenensis were distinct from the other groups, the separation between the two was relatively weak. Following principal coordinate analysis of the 48 Ethiopian accessions, a plot of the first and second principal coordinates (Fig. 4 ) detected similar groups revealed by cluster analysis (Fig. 2). However, clusters 5 and 6 were not clearly separated.


Figure 3
View larger version (14K):
[in this window]
[in a new window]

  		Figure 3. Plot of the first and second principal coordinates for 62 Dioscorea accessions based on 877 polymorphic bands derived from 10 amplified fragment length polymorphism (AFLP) primer pairs.

 

Figure 4
View larger version (16K):
[in this window]
[in a new window]

  		Figure 4. Plot of the first and second principal coordinates for 48 unidentified Dioscorea accessions from Ethiopia based on 165 polymorphic bands derived from 10 AFLP primer pairs. Numbers in parentheses refer to the clusters from UPGMA analysis; {circ} = late-maturing; {blacktriangleup} = early-maturing; {triangleup} = accessions from Areka Agricultural Research Center for which data on maturity time is not available; and • = nonflowing.

 

   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Interspecific Genetic Variability
Species identity of yams in Ethiopia has always been a subject of controversy and speculation. For example, Gebre-Mariam and Schmidt (1998) referred to yam tubers collected near Dilla, Gedeo zone of southern Ethiopia, as belonging to D. abyssinica Hochst. ex Kunth. However, the accessions we collected from the same area were morphologically distinct from D. abyssinica (S. Demissew, personal communication, July 2004; Tamiru et al., unpublished data, 2007). Hildebrand et al. (2002) used the designation "D. cayenensis complex" for all "the poorly understood set of species: D. cayenensis, D. rotundata, D. abyssinica, D. praehensilis and D. sagittifolia" from Sheko, Southwest Ethiopia. In the present study, cluster and principal coordinate analyses showed that the accessions from Ethiopia are distinct from the Dioscorea species commonly cultivated in West Africa. On the other hand, the accessions are genetically closer to D. cayenensis and D. rotundata than to any of the other species studied. Although this seems to suggest the possible involvement of similar wild species in the process of yam domestication both in Ethiopia and West Africa, the distinctiveness of the Ethiopian accessions may represent a divergent evolutionary pathway isolated from the widely known center of yam diversity in West Africa.

The differentiation of the various species and the description of their genetic relationships as determined by this study are generally consistent with established taxonomical relationships of the common Dioscorea species. The species D. rotundata, D. cayenensis, and D. alata form part of section Enantiophylum, while D. bulbifera belongs to section Opsophyton under the genus Dioscorea (Onwueme and Charles, 1994). D. bulbifera is also a distinct member of the genus in that it produces aerial tubers, and it is the only species that originated both in Asia and Africa (Ramser et al., 1996). Nevertheless, our observation that D. alata is more closely related to D. bulbifera than to other species of the same section contradicts the established taxonomy as well as earlier molecular studies involving both species (Malapa et al., 2005). On the one hand, this finding need to be further confirmed by including more accessions representing both species from different geographic origins. On the other hand, the fact that some cultivars of D. alata produce aerial tubers may support the observed closeness of the species to D. bulbifera.

Diversity within the Accessions from Ethiopia
The diversity detected within the accessions from Ethiopia mainly reflected their relatedness based on maturity time. The two major groups, cluster 1 and 4 (Fig. 2 and 4), roughly correspond to the main yam categories recognized by local farmers: hatuma boye ("male" yam) and macha boye ("female" yam). The so-called female yams include all the early-maturing landraces that are harvested twice or double-harvested (Tamiru et al., 2007). The so-called male yams mature late and are harvested once. This classification is also partly in agreement with findings of an earlier study involving morphological characterization of the same accessions (Tamiru et al., unpublished data, 2007). Although these clearly distinct yam types are well known in the main African species (Onwueme and Charles, 1994), the variation between them has not been well investigated at DNA level.

The separate clustering of WOL 012, representing the landrace ayina or ayino, further substantiates the relationship between maturity time and the actual genetic diversity. Ayina is harvested only once in most localities across the study area and is also morphologically close to the late-maturing landraces. Nevertheless, it matures earlier than all the landraces in this group (Tamiru et al., unpublished data, 2007). As a result, it is managed as an early-maturing landrace with two harvests per cropping season in some localities of Damot Woyde district in Wolayita zone, southern Ethiopia. In effect, this landrace is an intermediate type with respect to maturity time, which corresponds to its value for the first principal coordinate for molecular difference (Fig. 4). GGF 004 (bune or buna) has a very limited distribution, but is widely grown in Kucha district of Gamo-Gofa zone (Tamiru et al., 2007). It is perceived to be highly drought tolerant, and the tubers store well in the soil for a long time. It is usually harvested late in the season (November to December) after the harvesting season is over even for the late-maturing landraces and is distinguished by its highly branched tubers (with shape of "dog's feet") and distinct female inflorescence (Tamiru et al., unpublished data, 2007).

Accessions WOL 004b and AKA 013 are distinct with respect to their type of inflorescence. WOL 004b has panicle inflorescences, while about 94% of the unidentified accessions from Ethiopia possess raceme inflorescence. On the other hand, AKA 013 is one of the three accessions that have pistilate inflorescence (Tamiru et al., unpublished data, 2007). The commonly cultivated Dioscorea species are mostly dioecious with separate male and female inflorescences on different plants (Bai and Ekanayake, 1998; Degras, 1993). Nonetheless, this cannot fully explain the separate clustering of WOL 004b and AKA 013 from the rest of the accessions. WOL 013b and WOL 023b, despite having similar inflorescence morphology as WOL 004b, were grouped in cluster 1. Likewise, AKA 004 and GGF 004, the other accessions with female inflorescence, were not grouped with AKA 013. Similar reports are available in clonally propagated crops such as plantain (Musa spp. subgroup AAB) (Ude et al., 2003), where inflorescence morphology, although a key trait in conventional taxonomy, did not always show association with observed groupings of genotypes based on molecular markers.

Previously, we reported that although most named landraces are morphologically distinct, there were cases where either no morphological differentiation was observed between differently named landraces or some landraces known by similar local names were morphologically distinct (Tamiru et al., unpublished data, 2007). This is more common within the early-maturing group, where farmers use tuber flesh color as a key trait in distinguishing individual landraces. These observations were supported by the structure of genetic diversity revealed in the present study. GGF 003, WOL 002, WOL 007, and WOL 005 are all known by the same vernacular name hatiye, and are characterized by white tuber flesh color. The same is true for SID 001 and SID 005 that are known by the local name ado in the Sidama zone. But tuber color was not associated with the grouping of these accessions within cluster 4. It is known that significant morphological variations may be the result of differences in only few genes (Bradley et al., 1997). Accessions GGF 001, GGF 002, WOL 001, WOL 004a, and WOL 004b represented the late-maturing landrace wadala. Although they were collected from different localities, they were grouped in the same cluster. WOL 004b, which was sampled from the same field with WOL 004a, was well separated from this group, supporting the fact that landraces are often genetically heterogeneous (Harlan, 1975). This also represents a case where farmers' classification underestimates the actual genetic diversity.

Although the accessions were collected across different geographic areas, no geographic pattern of variation was detected. The early-maturing landraces are particularly well distributed throughout the collecting area. However, their grouping within cluster 4 does not relate to their area of origin. This observation is supported by the partitioning of genetic variation by AMOVA (Table 6), which showed very low variability among the collecting areas. This finding is also in line with the oral history that yam and its culture was originally introduced from Gamo-Gofa to Wolayita, and then to Sidama and Gedeo zones (Tamiru, 2006), which might have followed extensive exchange of germplasm among the various areas.

Implications for Conservation and Improvement
The distinctiveness the Ethiopian accessions from commonly cultivated yam species is an important finding of this study. Future studies must include wild species both from Ethiopia and other African countries for phylogenetic studies to resolve the species identity of the materials. Determination of ploidy level may also provide crucial information regarding their genome organization. If the distinctiveness of these materials is further confirmed by such studies, it may lead to the conclusion that Ethiopia represents a distinct center of yam diversity.

The pattern of diversity across the study area revealed that Gamo-Gofa and Woalyita zones host considerable levels of yam diversity. Accordingly, future collection activities must consider more sampling in these areas besides extending to other localities in the southern and southwestern parts of the country that were not covered by this study.


   ACKNOWLEDGMENTS
 
The authors gratefully acknowledge the yam genotypes provided by local farmers in Southern Ethiopia and the International Institute of Tropical Agriculture (IITA). We also thank the DU (Debub University)-NORAD (Norwegian Agency for Development Cooperation) project for the generous support during the fieldwork in Ethiopia, Berisso Kebede for assistance with AFLP analysis, and PD. Dr. Wolfgang Ecke for his help in setting up the experiment and critical comments on the manuscript. This study was sponsored by the German Academic Exchange Service (DAAD).


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Received for publication November 15, 2006.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 




This article has been cited by other articles:


Home page
 jashsHome page
J. C. Kuhl and V. L. DeBoer
Genetic Diversity of Rhubarb Cultivars
J. Amer. Soc. Hort. Sci., July 1, 2008; 133(4): 587 - 592.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tamiru, M.
Right arrow Articles by Maass, B. L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Tamiru, M.
Right arrow Articles by Maass, B. L.
Agricola
Right arrow Articles by Tamiru, M.
Right arrow Articles by Maass, B. L.
Related Collections
Right arrow Other Tubers
Right arrow Plant Genetic Resources

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Vadose Zone Journal
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome