Vol. 14(7), pp. 270-279, July 2020 DOI: 10.5897/AJPS2020.2020 Article Number: 332CF7B64301 ISSN 1996-0824 Copyright © 2020 Author(s) retain the copyright of this article http://www.academicjournals.org/AJPS African Journal of Plant Science Full Length Research Paper Morphological and molecular characterization of cultivated yam (Dioscorea species) in selected counties in Kenya Valentine Atieno*, Grace W. Gatheri, Joseph W. Kamau and Morris Muthini Department of Plant Sciences, Kenyatta University, P. O. Box 43844-00100, Nairobi, Kenya. Received 15 May, 2020; Accepted 19 June, 2020 This study was conducted to characterize Dioscorea spp. in Kenya using morphological and molecular characteristics. Data on 22 morphological traits were subjected to cluster analysis and multivariate analysis using principal component (PCA). The dendrogram of cluster analysis revealed three main groups: Species distribution based on PC-1 and PC-2 showed the distantly related species in each st nd rd quarter; D. alata L. (1 quarter), D. bulbifera L. (2 quarter), D. cayenensis Lam. (3 quarter) and D. th minutiflora Engl. (4 quarter). In molecular characterization, one sub-cluster grouped D. minutiflora Engl. and D. burkilliana J. Miege as one genetic group. However not all D. minutiflora Engl. species were in one specific cluster showing that there may be variation within the species. D. alata L. and D. bulbifera were seen to be potentially related because they shared a common origin. D. bulbifera L. and D. cayenensis Lam. genotypes clustered together, indicating that the species might be closely related. Generally, the rbcL marker demonstrated the phylogeny of Kenyan Dioscorea spp L. Comparison of morphological and molecular data analysis gave almost similar results. From the study, the phylogenetic relationships of Kenyan Dioscorea spp. were established and morphological and molecular characterization was efficient in establishing species relatedness among Dioscorea spp. Key words: Dioscorea spp., rbcL, principal component analysis, molecular characterization, morphological characterization, yams. INTRODUCTION Yams (Dioscorea spp.) are important monocotyledonous tuberous plants belonging to the order Dioscoreales, family Dioscoreaceae and the genus Dioscorea (Tamiru et al., 2008; APG III, 2009). The genus contains about 644 species distributed throughout the tropics in West Africa, South East Asia and Tropical America (Asiedu and Sartie, 2010; Couto et al., 2018). More than 8 species are important staples D. rotundata Poir. (White yam), D. alata L. (Water yam), D. cayanensis Lam. (Yellow yam), D. bulbifera L. (Aerial yam), D. dumetorum (Kunth) Pax. (Trifoliate yam), D. esculenta (Lour) Burk. (Chinese yam) D. nummularia Lam., D. pentaphylla L., D. hispida Dennst. and D. trifida L. (Ihediohanm et al., 2012). They are annual or perennial herbaceous vines, *Corresponding author. E-mail: valatieno16@gmail.com. Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License Atieno et al. with edible underground and aerial tubers (either stem or root depending on species) and are the world’s second most significant tuber crop. Yams are essential sources of food consumed as vegetables boiled, baked or fried. Yams bring food security to about 300 million people in Africa, Asia, parts of South America, Caribbean and the South Pacific Islands (Nanbol and Namo, 2019). Some species contain medicinal components useful in the pharmaceutical industries. For example, D. nipponica, D. alata L. and D. zingiberensis contain diosgenin helpful in relieving arthritis and muscle pain and lowers cholesterol levels (Chandrasekara and Kumar, 2016; Jesus et al., 2016). Purple yam contains anthocyanin that slows down lipid peroxidation and prevents the onset of cardiovascular disease (Blesso, 2019; Reis et al., 2016). In Kenya, the diversity of yams has been evolving over the years as numerous generations in many parts of the country select and domesticate different species and types independently according to their local cultivation practices and needs. In a recent report by the Kenya National Strategy on Genetic Resources (2016-2020), yams were listed among the underutilized and neglected crops in the country. The cultivated yams in Kenya include D. rotundata Poir., D. minutiflora Engl., D. bulbifera L., D. dumetorum (Kunth) Pax., D. alata L. and D. cayenensis Lam. They are mainly cultivated by elderly farmers basically for food in counties of Eastern, Central, Western and Coastal regions of the country (Muthamia et al., 2013). Molecular studies done on Kenyan yams have been minimal. Previous studies have investigated the genetic diversity using polymorphic Simple Sequence Repeats (SSR) markers that distinguished the landraces Muthamia et al. (2013) and ploidy levels; this revealed variable ploidy levels among the local yam landraces (Muthamia et al., 2014) and not the phylogeny of the species. Both studies recommended further work on the phylogeny of Kenyan yams. Other studies have solely utilized morphological characters to infer relationships within and between the Dioscorea species in Kenya (Mwirigi et al., 2009). This study aims to establish the relationships of Kenyan Dioscorea species using morphological and molecular characterisation, taking into account the recommendations of previous research. MATERIALS AND METHODS 271 8´35.1168´; E 37° 50´52.20852´). Specimens were also collected in Embu county (S 00° 27´53.36388´; E 37° 29´56.65272´), TaitaTaveta county (S 03° 24´2.46888´), Busia county (N 00° 29´47.86548´; E 34° 12´7.0272´) and Bungoma county (N 00° 34´10.300´; E 34° 33´31.1536´). Purposive sampling was used to select representative study sites with respect to the potential of yam production. This was done with the help of agricultural officers in each county who identified farmers farming yams. Collection of Dioscorea specimens Leaves and voucher specimens of Dioscorea species were collected from the various geographical regions of Kenya in the year 2018 (September to November). Collected specimens were identified and voucher specimens deposited in Kenyatta University Herbarium. Silica-gel dried leaves were collected for each sample for molecular characterization. Morphological characterization Twenty-four yam specimens were used for this study. Morphological data were observed directly on living plants under field conditions from farms where yams were grown. Twenty-two characteristics obtained from the International Plant Genetic Resources Institute’s (IPGRI) descriptors of yam (Dioscorea species) were considered (IPGRI, 1997) (Table 1). DNA extraction DNA was extracted from 0.2 g silica-gel dried leaves obtained from 17 randomly selected representative specimens and collected into Eppendorf tubes. Normal saline was added, and centrifuged. 400 µl lysis buffer was added and incubated for 1 h at 55-60°C with occasional mixing. The specimens were crushed and incubated again at 37°C for 3-4 h to deactivate lysozyme in the lysis buffer. They were cooled for 30 min, and afterwards centrifuged at 13,000 revolutions per minute for 5 min; an equal amount of chloroform was added gently and mixed thoroughly. The specimens were centrifuged again at 13,000 revolutions per minute for 8 min, using a large-bore pipette. The supernatant was transferred to another labelled Eppendorf tube, 600 µl isopropanol was added and mixed gently until the DNA was precipitated. The specimens were kept at 4°C for 20 min to precipitate the DNA further, centrifuged at 12,000 revolutions per minute for 5 minutes and the supernatant was discarded. The DNA pellets were washed by adding 70% ethanol and centrifuged again at 13,000 revolutions per minute for 2 min. The supernatant was discarded and the pellets were air-dried at room temperature. DNA yield was checked by running 3 µl of freshly extracted DNA specimens on 1% agarose gel stained with 3 µl loading dye and 1µl SYBR® green stain; it was visualized under an ultraviolet transilluminator at the Kenyatta University Tissue Culture laboratory. The quality and concentration of all DNA specimens were determined using Agarose gel electrophoresis. Study area PCR and sequencing The study was conducted in Meru, Embu, Taita Taveta, Busia and Bungoma counties. These counties were selected based on information gathered from the Kenya Agricultural and Livestock Research Organization on where Dioscorea species are mainly grown. Dioscorea specimens were collected from six farms from three sub-counties in Meru county; Imenti North (N 00° 4´32.43684´; E 37° 38´54.29688´), Imenti Central (S 0° 1´34.56264´; E 37° 38´37.65588´) and Tigania Central (N 00° PCR was achieved using rbcl marker (H1f F: CCACAAACAGAGACTAAAGC and Fofana R: GTAAAATCAAGTCCACCGCG (Fofana et al., 1997) and synthesized from Inqaba Biotec East Africa (IBEA), SouthAfrica. This primer marker was selected as a result of ease of PCR amplification and discriminatory power among yam species (Girma et al., 2015a). rbcl codes for ribulose 1, 5 bisphosphate carboxylase/oxygenase. This was carried out in a 25 µl reaction 272 Afr. J. Plant Sci. Table 1. Character and character states scored for morphological studies. Character Twining direction Stem colour Absence/presence of spines Absence/presence of wings Wing position Spine shape Leaf colour Leaf margin colour Vein colour Position of leaves Leaf type Leaf margin Leaf shape Leaf apex shape Petiole colour Flowering Flower colour Inflorescence type Aerial tuber shape Skin colour Surface texture Flesh colour Character state 1-Clockwise (climbing to the left) 2-Anticlockwise (climbing to the right) 1-Green; 2-purplish green; 3-brownish-green; 4-dark brown; 5-purple and 6-other. Absent/ Present Absent/ Present At the base/ Above base 1-Straight; 2-Curved upwards; 3-Curved downwards 1-Yellowish; 2-Pale green; 3-Dark green; 4-Purplish green; 5-Purple; 6-Other 1-Green; 2-Purple; 3-Other 1-Yellowish; 2-Green; 3-Pale purple; 4-Purple; 5-Other 1-Alternate; 2-Opposite; 3-Alternate at base/opposite above; 4-Other Simple/ Compound Entire/ Serrate 1-Ovate; 2-Cordate; 3-Cordate long; 4-Cordate broad; 5-Sagittate long; 6-Sagittate broad; 7Hastate; 8-Other 1-Obtuse; 2-Acute; 3-Emarginate; 4-Other 1-All green with purple base; 2-All green with purple leaf junction; 3-All green with purple at both ends; 4-All purplish-green with purple base; 5-All purplish-green with purple leaf junction; 6-All purplish-green with purple at both ends; 7-Green; 8-Purple; 9-Brownish green; 10-Brown; 11-Dark brown; 12-Other 1-No flowering; 2-Flowering in some years; 3-Every year 1-Purplish; 2-White; 3-Yellowish; 4-Other 1-Spike; 2-Raceme;3- Panicle 1-Round; 2-Oval; 3-Irregular (not uniform); 4-Elongate 1-Greyish; 2-Light brown; 3-Dark brown; 4-Other 1-Smooth; 2-Wrinkled; 3-Rough 1-White; 2-Yellowish white or off-white; 3-Yellow; 4-Orange; 5-Light purple; 6-Purple; 7-Purple with white; 8-White with purple; 9-Outer purple/inner yellowish; 10-Other volume containing 2.5 µl of 10x standard Taq, reaction buffer; 0.5 µl of 10 mM dNTPs; 0.5 µl of 10 µM primer H1F; 0.5 µl of 10 µM primer Fofana; 1 µl of template DNA, 0.125 µl of Taq, DNA polymerase, 19 µl nuclease-free water and 0.5 µl of Triton X. The PCR reaction was carried out in Techgene thermocycler FTGENE5D model (Techne- UK). The PCR reaction conditions for amplification consisted of initial denaturation at 94⁰C for 2 min followed by 35 cycles (denaturation at 94°C for 30 s, primer annealing at 46°C for 30 s, extension at 72⁰C for 90 s) and a final extension at 72°C for 7 min. The PCR products were stored at 4°C until used. PCR products were stained with SYBR green and separated by gel electrophoresis in 1% (w/v) agarose gel in 0.5X TBE buffer at 80 V for 30 min. After gel electrophoresis, the PCR products were visualized using an Ultra-violet trans-illuminator lamp. One hundred base pair (100bp) ladder was used for estimation of the molecular sizes of the bands. Gels were photographed using a Samsung digital camera. PCR products were then sent to South Africa for bidirectional sequencing at Inqaba Biotec East Africa (IBEA). Data analysis Data analysis based on morphological data Data on morphological characteristics from 24 specimens were coded into numerical values and used for cluster analysis. The dendrogram was drawn based on a hierarchical cluster analysis using single linkage (nearest-neighbour) procedure using DARwin computer software version 6. The dendrogram obtained was used in comparison with rbcL phylogenetic tree. Standardized data for qualitative characters were subjected to multivariate analysis and principal component analysis to identify the most discriminating morphological character using MVSP 3.2 and Conoco 5 software, respectively. Phylogenetic analysis The obtained sequences were exported to Finch TV Version 1.4.0 for base-calling. A consensus sequence was then created using DNA Baser Assembler v5.15.0; then a contig was created in comparison with the reference sequence using Gene studio Professional Edition. BLAST analysis was done to find identities that match the species. rbcL sequences were subjected to multiple alignments using the muscle alignment method in MEGA X to identify gaps and similar and mismatch regions among the two molecular characters. Maximum Likelihood (ML) and neighbourjoining algorithms were applied in phylogeny reconstruction. UPGMA was the statistical method used. The aligned sequences after subjection to the above parameters resulted in the construction of rbcL maximum likelihood phylogenetic trees. Atieno et al. 273 Table 2. Eigenvalues. Parameter Eigenvalues Percentage Cum. Percentage Axis 1 16.392 46.451 46.451 Axis 2 6.558 18.583 65.033 Axis 3 5.222 14.798 79.831 Axis 4 2.534 7.181 87.012 Table 3. PCA variable loadings. Traits A-Twining direction B-Stem colour C-Spines D-Spine shape E-Spines on stem base F-Wings G-Wing position H-Leaf colour I-Leaf margin colour J-Vein colour K-Leaf position L-Leaf type M-Leaf margin N-Leaf shape O-Leaf apex shape P-Distance between lobes Q-Petiole colour R-Flowering S-Tuber shape T-Skin colour U-Surface texture V-Flesh colour PC 1 0.073 0.111 0.113 0.259 0.579 -0.094 -0.094 0.005 0.115 -0.063 0.103 0.000 0.000 -0.225 0.000 0.559 -0.350 -0.063 -0.029 0.100 0.115 -0.087 RESULTS Principal component analysis The PCA results established that the first four principal components together described 87.01% of the overall variance present in the data set (Table 2). Scores on the first principal component (PC-1) which explained 46.45% of the total dissimilarity were vastly correlated to stem colour, presence of spines, spine shape, spines on stem base, leaf margin colour, leaf position, the distance between lobes and surface texture (Table 3). The second principal component (PC-2) described 18.58% of the overall dissimilarity and was vastly correlated to spines on stem base, the distance between lobes and petiole colour (Table 3). The third component (PC-3) which described 14.78% of the dissimilarity was primarily related to the distance between lobes and flesh colour. The fourth principal component (PC-4) described PC 2 -0.044 -0.007 0.074 0.058 0.273 0.024 0.024 -0.003 0.032 -0.124 -0.070 0.000 0.000 0.034 0.000 0.270 0.902 0.039 0.022 -0.053 0.024 -0.013 PC 3 0.007 -0.102 -0.071 -0.160 -0.558 -0.011 -0.011 -0.049 -0.154 0.056 0.010 0.000 0.000 0.081 0.000 0.767 -0.032 -0.039 -0.012 -0.048 -0.068 0.114 PC 4 0.047 0.223 -0.029 -0.066 -0.351 -0.044 -0.044 -0.096 0.613 -0.038 0.223 0.000 0.000 -0.398 0.000 -0.013 0.127 -0.112 0.104 0.065 0.068 -0.416 7.18% of the total distinction and was determined by leaf margin colour, stem colour, leaf position, petiole colour and shape of the tubers. The distribution of species based on the first and second principal components shows dissimilarity among the species and how extensively dispersed they are along both axes (Figure 1). The two components explain a cumulative variability of 65.03%. Based on the distribution of specimens in the first quarter, D. alata L. is the most distantly related to that group; whereas in the second quarter D. bulbifera L. is the least similar in the group. The most distant in the third quarter is D. cayenensis Lam. The last quarter is made up of a D.minutiflora Engl. that is least similar to the group (Figure 1). Correlation between the variables related to the first and second principal components are presented in Figure 1. From the correlation circle in Figure 1, petiole colour has a significant effect on the variables as a result of the arrow being long. There is a positive correlation between Afr. J. Plant Sci. 1.0 274 DAlatT DMinut DAlatT DMinut VeinColo DAlatT DAlatT DMinut TwinDirc DMinut LeafPost DMinut DMinut SkinColo DMinut DMinut StemColo LeafMargLeafColo DMinut DMinut SurfText LeaMarCl LeafType LeaApxSh TubrShap SpinShap SpinOnSt DisBetLb DAlatT FlesColo LeafShap Wings WingPost Flowerin DBulbf DBulbf DBulbf Spines DBulbf DBulbf DMinut DMinut -1.0 DMinut DMinut DCayen PetiColo -1.0 1.0 Figure 1. Correlation circle of the first two principal components (PC1 and PC2). the shape of leaves and the presence of wings. However, there is a negative correlation between the shape of leaves and the presence of wings on one hand and twining direction and tuber skin colour. Petiole colour and spines are not correlated as well as petiole colour and vein colour. Dendrogram based on morphological characters The 24 yam specimens included in the morphological study were grouped into three clusters (Figure 2). Cluster 1 grouped D. minutiflora Engl. species collected from different areas; Teso North, Embu and Meru. This cluster had two sub-clusters; 1a and b respectively. Sub-cluster 1a is a group of D. minutiflora Engl. characterised with many spines on stem base, spines curved upwards, leaf veins yellow, yellow leaf margins, leaves alternate at base/ opposite above, green petioles and rough tuber surface texture. Sub-cluster 1b is a group of D. minutiflora Engl. with many spines on the stem base, spines curved downwards, vein green, leaf margin green, leaf position opposite, petioles all green with a purple base and tuber surface texture rough. This suggested a close relationship between the D. minutiflora Engl species collected from the different areas (Teso North, Embu and Meru) based on similar morphological traits. Cluster 2 contained three sub-cluster groups Cluster 2(I), Cluster 2(II) and Cluster 2(III). Sub-cluster 2(I) is a group of D. alata L. species from Taita taveta and Busia Atieno et al. 275 a 1 b I 2 II III 3 Figure 2. Dendrogram showing the relationship in yams (Dioscorea s) species based on morphological characteristics. (Teso North) that twine to the right in an anticlockwise direction, stem purplish-green, leaf margin purple, leaf shape cordate long, petioles purplish-green with a purple base, tuber flesh purple and white. D. alata L. species collected from Teso North had sagittate long leaves, and tuber flesh purple in colour whereas D. alata L. tuber flesh colour from Taita Taveta was white. Sub-cluster 2(II) is a group of D. bulbifera L. species from Bungoma, Busia (Teso North and South) and Embu counties with stem twining to the left in a clockwise direction, spines absent, wings present on the stem, flowering in some years and presence of aerial tubers. However, only D. bulbifera L. from Bungoma and Busia showed flowering and clustered together whereas that from Embu did not. Sub-cluster 2(III) had D. cayenensis Lam. from Busia (Teso North) characterised by yellowish veins on the leaves, cordate broad leaf shape and a cylindrical tuber with tuber flesh colour yellow. Cluster 3 is a group of D. minutiflora Engl. from Meru County. Few spines on stem base, green stems, pale green and dark green leaves, brown leaf margins, green leaf veins, cordate leaves and white tuber flesh colour characterised this cluster. These characters were key in distinguishing this cluster from Cluster 1. Molecular characterisation of Kenyan yam (Dioscorea species) Six species’ identities were used to construct the dendrogram among the 17 selected genotypes. The dendrogram based on rbcL markers distinguished the seventeen yam genotypes into two main cluster groups (Figure 3) Cluster 1 consisted of two main subclusters (a) and (b). Subcluster 1(a) and cluster 2 comprised D. minutiflora genotypes. This is similar to the cluster 1 and 3 of the morphological analysis which consists of D. minutiflora species clustered together (Figure 2). However, D. minutiflora genotypes were in different clusters; 1(a) and cluster 2, showing that there may be variation in the genotypes (Figure 3). Subcluster b(I) consisted of D. alata genotypes similar to the cluster 2(i) of the morphological analysis. Subcluster b(II) consisted of D. bulbifera and D. cayenensis genotypes, indicating that the two species might be closely related as shown in morphological studies in cluster 2(II and III) (Figure 3). The results showed a high correlation between the morphological and molecular data in the study of Kenyan yams (Figures 2 and 3). The evolutionary history was inferred using the Neighbor-Joining method (Saitou and 276 Afr. J. Plant Sci. a I 1 b II 2 Figure 3. Evolutionary relationships of taxa. Nei, 1987) (Table 4). The bootstrap consensus tree inferred from 1000 replicates is taken to represent the evolutionary history of the taxa analyzed (Felsenstein, 1985). Branches corresponding to partitions reproduced in less than 50% bootstrap replicates are collapsed. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches (Felsenstein, 1985). The evolutionary distances were computed using the Maximum Composite Likelihood method (Tamura et al., 2004) and are in the units of the number of base substitutions per site. The analysis involved 23 nucleotide sequences. Codon positions included were st nd rd 1 +2 +3 +Noncoding. All positions with less than 95% site coverage were eliminated. That is, fewer than 5% alignment gaps, missing data, and ambiguous bases were allowed at any position. There were a total of 593 positions in the final dataset. Evolutionary analyses were conducted in MEGA X (Kumar et al., 2018) (Table 5). DISCUSSION Morphological traits that had a paramount role in discriminating between the yam species in this study were stem colour, leaf margin colour, leaf position, the distance between lobes, petiole colour, tuber shape, tuber surface texture and tuber flesh colour. These results are in congruence with results obtained by Jyothy et al. (2017). They revealed that morphological variability score on the first principal component (PC-1) was highly correlated with characters related to tuber shape and tuber flesh colour. Similarly, Mwirigi et al. (2009) reported that PC-2, PC-3 and PC-4 were mainly correlated with characters related to leaf position and tuber flesh colour similar to the results of PC-4 and PC-3 from this study. Results obtained from Sheikh and Kumar (2017) revealed that variability scores on the first principal component (PC-1) were highly correlated with characters related to stem colour. This was also similar with the results obtained in this study on the first principal component (PC-1) being highly correlated with stem colour. From the dendrogram, morphological characterisation of Kenyan yams from 5 geographical regions indicated that most species from the Eastern area (Meru and Embu) are closely related despite their geographic location being widespread and some showing a few morphological variations. This is as a result D. minutiflora Engl. from the Atieno et al. 277 Table 4. Gene bank species identities. Lab designation Species identification V005, V007 V001,V012,V013, V014, V019 V009,V010,V011, V023 V018, V025, V026 V002, V003 V022 Dioscorea burkilliana Dioscorea togoensis Dioscorea alata Dioscorea bulbifera Dioscorea cirrhosa Dioscorea cayennensis The accession number of nearest neighbour MG805605.1 NC_039856.1 NC_039707.1 MG805604.1 HQ637842.1 NC_039836.1 Percentage identity (%) 98.83 98.83 99.63 99.82 98.83 99.46 Table 5. Laboratory species identities. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Lab designation V001 V002 V003 V005 V007 V009 V010 V011 V012 V013 V014 V018 V019 V022 V023 V025 V026 Species identities Dioscorea minutiflora Dioscorea minutiflora Dioscorea minutiflora Dioscorea minutiflora Dioscorea minutiflora Dioscorea alata Dioscorea alata Dioscorea alata Dioscorea minutiflora Dioscorea minutiflora Dioscorea minutiflora Dioscorea bulbifera Dioscorea minutiflora Dioscorea cayennensis Dioscorea alata Dioscorea bulbifera Dioscorea bulbifera two regions clustering together. This indicates a likelihood of numerous exchange of planting materials among and between farmers from different zones. It is also likely that constant vegetative propagation and selection have contributed to the wide phenotypic variability of D. minutiflora Engl. (Mwirigi et al., 2009). However, there are four accessions of D. minutiflora Engl. in Meru and Embu distinguished by the size of the tuber and spiny stem base. It can be seen that D.alata L. (Taita Taveta and Busia) and D. bulbifera L. (Embu, Bungoma and Busia) are very closely related and distant to D. cayenensis Lam (Busia). The dendrogram from molecular data was prepared by using the neighbour-joining method. In the cluster analysis D. minutiflora Engl. and D. burkilliana J. Miege from West Africa were grouped, indicating that they might be considered as one genetic group, as stated by Chaïr et al. (2005). In another study, Magwé-Tindo et al. (2018) identified Guinea Yam wild relatives using the whole plastome phylogenetic analyses which clearly showed Area of collection Meru Meru Meru Meru Meru Taita-Taveta Taita-Taveta Taita-Taveta Embu Embu Embu Embu Embu Teso North Teso North Teso North Bungoma that D. minutiflora Engl. and D. burkilliana J. Miege formed two strongly supported groups and clustered together. This is in agreement with results obtained by Ramser et al. (1997) who found them in the same habitat. Miège (1968), in his study, established D. burkilliana J. Miege and D. minutiflora Engl. as two morphologically similar species that differ only by the characteristics of their below-ground parts. These results are in agreement with the results of this study as a result of D. burkilliana J. Miege and D. minutiflora Engl. clustering together. D. alata L. and D.bulbifera L. are seen to be potentially related from Figure 2 because they share a common origin. This, however, contradicts established taxonomy as well as earlier molecular studies involving both species stating that D. alata L. and D.bulbifera L. are not closely related (Malapa et al., 2005). On the other hand, the fact that some cultivars of D. alata L. produce aerial tubers may support the observed closeness of the species to D. bulbifera L. (Tamiru et al., 2007). The input of both morphological and molecular data is critical in 278 Afr. J. Plant Sci. producing well-resolved species delimitation. In this study, results showed a correlation between morphological and molecular data analysis, indicating that molecular data supported morphological species delimitation. Caddick et al. (2008) in his study stated that higher sampling of taxa and morphological and molecular characters for Dioscoreales had produced resolved topologies that corroborate the circumscription that was proposed by APG (1998). His study also concluded that increased bootstrap support in analysis indicated high congruence between independent morphology and molecular data sets and demonstrated that both morphological and molecular data are essential in resolving the relationships within Dioscoreales. Sartie et al. (2012) in their study on genotypic and phenotypic diversity of cultivated tropical yams using phenotypic and SSR markers established an improved understanding about the genetic and phenotypic relatedness among D. rotundata Poir., D. cayenensis Lam., D. alata L. and D. dumetorum (Kunth) Pax. This is similar to what was done in this study using phenotypic and molecular markers to establish phylogeny of Dioscorea in Kenya. Girma et al. (2015b) in their study of morphological and SSR analysis of D. alata L. indicated that combining SSR markers and phenotypic data were useful for identification of D. alata L. accessions likewise to combining morphological data and molecular markers in characterizing Kenyan Dioscorea species. Conclusion Dioscorea species grown in Kenya exhibited morphological variations. Phylogenetic relationships of Kenyan Dioscorea species were established with D. alata L. and D. bulbifera L. seen to be closely related and D. minutiflora Engl. and D. burkilliana J. Miege from West Africa grouping together as one genetic group. Molecular and morphological characterization was efficient in establishing species relatedness among Dioscorea species. Future studies should consider collections from other localities in addition to Meru, Embu, Taita-Taveta, Busia and Bungoma counties and more than one molecular marker should be used. CONFLICT OF INTERESTS The authors have not declared any conflict of interests. REFERENCES APG (1998). An ordinal classification for the families of flowering plants. Annals of the Missouri Botanical Garden 85: 531-553. APG III (Angiosperm Phylogeny Group) (2009). An update of the angiosperm phylogeny group classification for the orders and families of flowering plants: APG III. Botanical Journal of the Linnean Society. 161:105-121. Asiedu R, Sartie A (2010). Crops that feed the World 1. Yams. Food Securit 2(4): 305-315. Blesso C (2019). Dietary Anthocyanins and Human Health. Nutrients 11(9):2107. Caddick LR, Rudall PJ, Wilkin P, Chase MW (2008). Yams and their allies: systematics of Dioscoreales. In: Wilson KL, Morrison D(eds) Systematics and evolution of monocots. C.S.I.R.O Clayton South VIC Australia, pp. 475-487. Chaïr H, Perrier X, Agbangla C, Marchand J, Dainou O, Noyer J (2005). Use of cpSSRs for the characterisation of yam phylogeny in Benin. Genome 48(4):674-684. Chandrasekara A, Kumar JT (2016). Roots and Tuber Crops as Functional Foods: A Review on Phytochemical Constituents and Their Potential Health Benefits. International Journal of Food Science 2016:1-15. Couto R, Martins A, Bolson M, Lopes R, Smidt E, Braga J (2018). Timecalibrated tree of Dioscorea (Dioscoreaceae) indicates four origins of yams in the Neotropics since the Eocene. Botanical Journal Of The Linnean Society 1(1). Fofana B, Harvengt L, Bandoin J, Jardin P de (1997). New primers for the polymerase chain amplication of cpDNA intergenic spacers in phaseolus phylogeny. Belgian Journal of Botany 129:118-122. Girma G, Spillane C, Gedil M (2015a). DNA barcoding of the main cultivated yams and selected wild species in the genus Dioscorea. Journal of Systematics And Evolution 54(3):228-237. Girma G, Gedil M, Spillane C (2015b). Morphological, SSR and ploidy analysis of water yam (Dioscorea alata L.) accessions for utilisation of aerial tubers as planting materials. Genetic Resources And Crop Evolution 64(2):291-305. Ihediohanm N, Onuegbu N, Peter-Ikec A, Ojimba N (2012). A Comparative Study and Determination of Glycemic Indices of Three Yam Cultivars (Dioscorea rotundata, Dioscorea alata and Dioscorea dumetorum). Pakistan Journal of Nutrition 11(6):547- 552. IPGRI/IITA (1997). Descriptors for Yam (Dioscorea spp.). International Institute of Tropical Agriculture, Ibadan, Nigeria/International Plant Genetic Resources Institute Rome Italy. Jesus M, Martins A, Gallardo E, Silvestre S (2016). Diosgenin: Recent Highlights on Pharmacology and Analytical Methodology. Journal Of Analytical Methods In Chemistry 2016 1-16. Jyothy A, Sheela MN, Radhika NK, Anwar I (2017). Morphological Characterisation of Greater Yam (Dioscorea alata L.) Landraces in Kerala. Journal of Root Crops 43(1):3-10. Magwé-Tindo J, Wieringa J, Sonké B, Zapfack L, Vigouroux Y, Couvreur T, Scarcelli N (2018). Guinea yam (Dioscorea spp., Dioscoreaceae) wild relatives identified using whole plastome phylogenetic analyses. Taxon 67(5):905-915. Malapa R, Arnau G, Noyer JL, Lebot V (2005). Genetic diversity of the greater yam (Dioscorea alata) and relatedness to D. nummularia Lam. and D. transversa Br. as revealed with ALFP markers. Genetic Resouces and Crop Evolution 52:919-929. Miège J (1968). Dioscoreaceae. In: Hepper FN (ed) Flora of West Tropical Africa. Hutchinson J, Dalziel JM Vol 3 Millbank London UK. pp. 144-154. Muthamia ZK, Morag FE, Nyonde AB, Mamati EG, Wanjala BW (2013). Estimation of genetic diversity of the Kenyan yam (Dioscorea spp.) using microsatellite markers. African Journal of Biotechnology 12(40):5845- 5851. Muthamia Z, Nyende A, Mamati E, Ferguson M, Wasilwa J (2014). Determination of ploidy among Yam (Dioscorea spp.) landraces in Kenya by flow cytometry. African Journal of Biotechnology 13(3):394402. Mwirigi M, Kahangi E, Mamamti E, Nyende A (2009). Morphological variability within the Kenyan yam (Dioscorea spp.). Journal of Applied Biosciences 16:894- 901. Nanbol K, Namo O (2019). The Contribution of Root and Tuber Crops to Food Security: A Review. Journal Of Agricultural Science And Technology B:9(4). Ramser J, Weising K, Lopez-Peralta C, Terhalle W, Terauchi R, Kahl G (1997). Molecular marker-based taxonomy and phylogeny of Guinea yam (Dioscorea rotundata – Dioscorea cayenensis). Genome 40:903915. Reis J, Monteiro V, de Souza Gomes R, do Carmo M, da Costa G, Atieno et al. Ribera P, Monteiro M (2016). Action mechanism and cardiovascular effect of anthocyanins: A systematic review of animal and human studies. Journal of Translational Medicine 14(1). Sartie A, Asiedu R, Franco J (2012). Genetic and phenotypic diversity in a germplasm working collection of cultivated tropical yams (Dioscorea spp.). Genetic Resources and Crop Evolution 59(8):17531765. Sheikh N, Kumar Y (2017). Morphological Characterisation of Meghalayan Dioscorea spp. North-East India. Journal of Agriculture Science and Technology 19:487-497. 279 Tamiru M, Becker H, Maass B (2007). Genetic Diversity in Yam Germplasm from Ethiopia and Their Relatedness to the Main Cultivated Species Assessed by AFLP Markers. Crop Science 47(4):1744. Tamiru M, Heiko CB, Brigitte LM (2008). Diversity, distribution and management of Yam landraces (Dioscorea spp.) in Southern Ethiopia. Genetic Resources and Crop Evolution 55:115-131.