ISSN: 2276-7770                  ICV: 6.15


Submission Date: 24/08/2015

Accepted: 09/09/2015

Published: 21/09/2015


Research Article (DOI http://doi.org/10.15580/GJAS.2015.5.082415121)


Dichotomous key determining varietal groups of yam species Dioscorea alata L.


KOUAKOU Amani Michel1*, DIBI Konan Evrard Brice1,

N’ZUE Boni1, ESSIS Brice Sidoine1, ZOHOURI Goli Pierre1


1Centre National de Recherche Agronomique (CNRA), Côte d’Ivoire 01 BP 1740 Abidjan, Tél. (225) 22 48 96 24, Fax. (225) 22 48 96 11, www.cnra.ci.


*Corresponding Author’s E-mail: amkouakou@ yahoo. fr




Yam is an important food crop in Côte d'Ivoire. Nevertheless, its production is facing many constraints including the inability to distinguish cultivated varieties from one another. Thus, different varieties and species are designated by the same names from one region to the other and vice versa, the same varieties are called differently. The absence of major descriptors, which are easy to use is the main cause of this confusion. The objective of this study is to propose agro-morphological descriptors easy to identify, for a better determination of yam varieties of D. alata. A total of 91 accessions of yam of D. alata species were used. These accessions were observed using morphological parameter. Eight simple agro-morphological descriptors which are easy to describe were identified as major dichotomous keys for the determination of the D. alata species. These characters have been selected based on the discrimination they permit to operate. Using them, four varietal groups were formed: Pyramid, Douoblé, Brazo and Bêtê-Bêtê.


Keywords: Yam, Dioscorea alata, morphological characterization, dichotomous keys, varietal groups.





Yams of the Dioscorea genus are among the most widespread tropical food crops (Coursey, 1967). The main producers of this crops are in West Africa. Nigeria is the first producer in the world with 37.11 million of tonnes of yams. It is followed in Africa by Côte d'Ivoire with a production of 5.8 million of tonnes in 2013 (FAOSTAT, 2014). There are over 600 species of yams in the world, some are wild, while others are grown. This genetic diversity is an asset for the sustainability of yam cultivation. Indeed, it offers the possibility to make selections and create new varieties responding to the ever evolving concerns of the populations. Therefore, for thousands of years, producers exploit this variability to develop their crops (Geoffrey, 2000). The desired traits of yams are: high yield, good resistance to diseases and pests and good organoleptic and nutritional qualities. However, new needs arise with the industrialization such us production of starch, flour, pharmaceutical products and admixtures.

The importance of each species depends on the country. In Ivory Coast, the species Dioscorea alata is the most cultivated and consumed, in contrast to other African countries where Dioscorea rotundata is yam species primarily found. D. alata represent more than 55 to 60% of the national production (Doumbia et al., 2006). However, most research on the genetics of yams in Africa and in the world have focused on the species D. rotundata (Hamon et al, 1986;. Zoundjihékpon 1993; Dansi et al., 1997; Mignouna, 2002). The works on the genetics of D. alata have essentially focused on the molecular aspects (Egesi et al., 2007, Obidiegwu et al., 2009).

Despite their socio-economic importance and the high number of yams species cultivated, several obstacles limit the integration of yam in a modern agricultural system. Among these constraints, there is the inability to distinguish cultivated varieties from one another. Thus, it is found different varieties and species designated by the same names from one region to the other and vice versa, the same varieties are called differently. The absence of major descriptors, which are easy to use by all is the main cause of this confusion.

The overall objective of this article is to propose agro-morphological descriptors easy to identify, for a better determination of yam varieties of D. alata. Specifically, this will involve the determination of genetic groups of D. alata yams, by the use of major dichotomous keys.





Plant material


The plant material is composed of 91 accessions of yam of D. alata species from the CNRA yams collection in Ivory Coast.




The work was conducted at the CNRA research station of food crops located in Bouake, in the central of Côte d'Ivoire, between 7 ° 40 'N and 5 ° 2 O. The region is dominated by the Guinean savannah and a lateritic soils moderately desaturated. The average annual rainfall is between 1100 and 1200 mm.

The approach used in this study was to observe characters describing the stem and leaves for the aerial parts, roots and tubers for underground parts and reproductive organs that are bulbils and flowers. The stem, leaves and tubers were described by their shape and colour, while the absence or presence of flowers, bulbils and roots were recorded. The number and weight of products tubers were evaluated. The severity of anthracnose and viruses as well as attacks of mealybugs were recorded.

The culture was made without any fertilization. Yam plants were staked individually in order to avoid mixing between stems and leaves during observations.


Experimental design


The experimental design used is one block without repetition. The accessions were planted on mounds in strips of 10 m. Each yam accession has been represented by 10 plants spaced one meter on lines and one meter between the lines.


Observations and measurements


The observations were made on ten (10) plants per accession and averages were calculated for each character. Only accessions having a number of living plants greater than or equal to 5 were selected. Data collection focused on 27 agro-morphological characters as described in Table 1. The characters that describe the stem, leaves and leaf diseases were observed during the vegetative phase of the plant and those describing the tubers were recorded at harvest. Data were collected using the descriptors defined by IPGRI-IITA (1997).

The characters were separated according to whether they are related to:


-     the stem and the leaf,

-     the health status of the plant,

-     reproduction (Table 2)

-     Morphology of tuber and yield components (Table 3).


The description of young plants was made two months after planting, and those of the adult plants was done between five and six months after planting. The ratings of severity of anthracnose and viruses were made five months after planting by the method of Nicole et al., (1990). The severity is the percentage of the leaf surface of the plant affected by the disease. The terms are:


1 = 0 to 2% of affected leaf area,

2 = 3-25% of affected leaf area,

3 = 25 to 50% of affected leaf area,

4 = more than 50% of affected leaf area,

5 = plant has died from the attack.


Regarding the attacks of mealybugs on tubers at harvest, terms are:


0: no mealybugs on the tuber,

1: presence of a small number of mealybugs on the tuber

2: presence of a large number of mealybugs on the tuber.


At harvest, tubers were weighed and separated into small and large tubers tubers. Large tubers are those whose weight is greater than or equal to 1 kg. The others are considered as small tubers. Each batch was then counted and weighed. Before frequencies calculation, the results recorded were converted into numbers of tubers classes according to the rule of Sturge (Scherrer, 1984). Similarly, the weight of tubers weights obtained after weighing, were transformed into tuber weight classes.

The class determination formula is:


Number of Classes = 1 + (3.3 log n); n being the number of samples (91 in this case).

The class intervals were obtained by applying the rule of Yule (Scherrer, 1984):

Intervals of classes = (Xmax - Xmin) / Number of classes, Xmax and Xmin respectively being the largest and the smallest value in the statistical series.

Descriptive analyses were performed with the EXCEL software to characterise the accessions at the agro-morphological level.












Aerial vegetative characters


Variability of characters related to the stem


The absence of thorns on the stem, the presence of wings on the stem and the quadrangular section with the angle in wings are the phenotypes mostly observed (Table 4). They are present in 98.90% of accessions representing 90 of the 91 analysed genotypes. In contrast, the presence of thorns on the stem and the absence of wings on the stem as well as the circular section of the stem are rare characters for D. alata and are present only on one accession in the study population; this represents 1.10%.

Regarding the colour of the young stem, green, purple and brown are the three observed characteristics. Fifty seven (57) of 91 accessions have purple colour at the base of the stem of young plants, against one brown accession and 33 accessions whose base is green. But at the top of the stem, accessions with green colour are most numerous with 78 accessions of 91, while those who had the top of the stem purple are 11. Two (2) the accessions have the top of the stem brown.







Variability of characters related to the leaf


The young leaves have purple or green colour. Ninety percent (90%) of young leaves are green and 10% are purple. For the adult leaves, 72 of the 91 accessions have green leaves, while leaves of the 19 another accessions are yellowish.

The adult leaves have a strung shape for 57.14% of the accessions and a sagittate shape for 42.86% of these accessions; one accession has hastate shape leaves.


Variability associated with the colour of the petiole


Three colour of petiole were observed in the population examined. They are completely green petiole or green petiole with both ends purple and green petiole with their junction on the stem purple. Accessions with entirely green petioles represented 76% of the observed accessions, against 19% of accessions which petiole were completely green with purple colour at the two ends, and 5% with green petiole having purple base.


Variability between characters related to reproduction


Only 10 of the 91 accessions observed have flourished (10.99%). Among the flowering clones, 7 male accessions were observed against 03 female accessions. Only 12% of accessions gave bulbils (Table 5).







Variability of traits related to yield components


The average number of small tubers produced per plant was 0.94 for a maximum of 3.4 small tubers per plant. Some accessions did not produce small tubers.

The average number of large tubers per plant was 0.54; the maximum for that character being 1.44. Some accessions did not produce large tubers.

As for the total number of tubers per plant, the minimum was 0.55 and the maximum 4.2; the average being 1.5 for the population studied.

The average weight of tubers per plant was 1 kg; the minimum was 0.16 kg and the maximum, 4 kg (Table 6).

For each yield components, seven classes were defined. The intervals between two classes were 0.2 for the number of large tubers per plant and 0.5 for the number of small tubers per plant, the total number of tubers per plant and the total weight of tubers per plant. For practical use of the results, these classes were grouped into large and small tubers, according to their individual weight greater than 1000 g or not.

All accessions produced at least one tuber per plant. Accessions with only produces large tubers represent 13.19%, while those who have produced only small tubers constitute 8.79% of the population (Figure 1). In the category of large caliber tubers, the following proportions were obtained (Figure 2):


- 8.79% no big tuber

- 86.81% one large tuber per plant;

- 4.40% two large tubers per plant.


Among the accessions that produces small tubers, the breakdown is as follows (Figure 3):

 - 13.19% no small tuber

 - 60.44% one tuber

 - 20.88% two tubers

 - 3.3% 3 tubers.

 - 2.2% over than 3 tubers.


As for the total number of tubers produced per plant, there was 14.29% of accessions with a single tuber ; 62.64% with two (2) tubers ; 18.68 with three (3) tubers and 3 4.4% with more than 3 tubers (Figure 4).

The proportions for the weight of production is 74.73% for accessions whose tuber weight is less than one kilogram per plant ; 24.18% for those with tuber weight per plant between 1 and 2 kg, and 1.1% for  accessions with a weight of tuber per plant between 2 and 3 kg (Figure 5).













Characters related to the external morphology of the tuber


The presence of roots on the tuber, the level of branching and tuber shape describe the external appearance of the tuber (Table 7). There is a predominance of accessions that have no roots on the tuber. They represent 70.33% of accessions, against 24.18% who have very little root on the tuber and 5.49% of accessions which tuber is covered with roots. Seventy eight (78) accessions, or 85.71% have unbranched tubers and 23 accessions or 25.87% have branched tubers. Forty (40) accessions, or 43.95% showed round-shaped tubers and 32 genotypes, or 35.16% had cylindrical tubers. These two forms are the most frequent. The collection also includes two accessions (2.2%), whose tubers are flattened; 4 (4.4%) with conical tubers and 13 (14.29%) with irregularly shaped tubers.





Characters related to the internal aspect of the tuber


The oxidation of the amylaceous parenchyma, its texture and colour, as well as the uniformity of the colour and the thickness of phelloderm describe the internal aspect of the tuber in cross section. Less than 10% of accessions showed oxidation after cutting (Table 8). The texture of the amylaceous parenchyma varied from fine texture to the moderately grainy and the very grainy texture. Most of the accessions (72.5%) presented tubers with a fine texture amylaceous parenchyma and 16.5% had a moderately grainy texture. The other 11% of the accessions showed a very grainy texture.

Three (3) colours were observed for the fresh flesh of the tuber. White flesh predominated, with almost 75% of the genotypes. Pure white flesh of the tuber represented 32.97% of the accessions whereas 40.66% of them were yellowish-white. Orange flesh was noted for 16.48% of the accessions and the purple colour was shown by 7.69% of them (Figure 6).

These uniform colours of the flesh of the tuber were observed in 85.71% the cases. The other accessions have white and purple or white and orange flesh.

The thickness of the phelloderm was less than one centimetre for 81% of the accessions.







Traits related to disease susceptibility


Few accessions of the collection were susceptible to viruses and anthracnose (Table 9).

For yam virus, 48.35% of the accessions showed any symptom; 32.97% of the accessions had 3 to 25% of their leaf area affected, while 13.19% had 25 to 50% of their leaf area with symptoms. Only 5.49% of the accessions had more than 50% of their leaf area attacked. Considering anthracnose, 64.84% of the accessions didn’t show any symptom. Only one accession died after the attack of Colletotricum gleoesporioides while for 13.19% of the accessions, the attack has concerned between 3 and 25% of the leaf area. The proportion of accessions with 25 to 50% of leaf area attacked by anthracnose was 13.19%.

Mealybugs, were present on the tubers of 58.24% of accessions at the harvest time (Table 9).





Dichotomous key for the determination of the varietal groups


In the light of the characters observed, we propose a key of determination for D. alata varietal groups. It consists of eight simple agro-morphological descriptors which are easy to determine. These descriptors are: the presence of thorns on the stem, the presence of wings on the stem, the shape of the cross section of the stem, the presence of the purple colour on the stem, the presence of the purple colour in leaves of the young plant, the presence of the purple colour in the tuber, the colour of the flesh of the tuber and the colour of the leaf.

These traits have been selected on the base of their facility to be used. Thus four varietal groups were formed (Figure 7). In fact, the presence of thorns on the stem, the presence of wings on the stem, the shape of the cross section of the stem permitted to separate distinctively the clone IB26 from the other D. alata. The presence of thorns and the absence of wings on the stem and also the circular shape of the cross section of the stem characterise the IB26 accession. This genotype forms therefore alone a varietal group to be called Pyramid group.

               After this first group of characters, the presence of the purple colour in the tuber and the aerial vegetative organs are used to separate the other accessions. Accessions having purple stems, purple leaves of


purple young plants and the purple flesh tuber can be opposed to the others. These are the accessions of the group called Douoblé.

               Among D. alata, apart from the accessions that belongs to the two groups already described some others have green petiole with purple extremities; orange flesh, green stems and grenen mature leaves. These genotypes are grouped in the Brazo group.

               The accessions of the fourth group have an entirely green mature stems, green petiole, green leaves and white flesh: they form the Bêtê-Bêtê group. This Bêtê-Bêtê group can be separated into three subgroups. One is formed of accessions having dark-green leaves and piriform tubers. The second subgroup includes accessions which leaves are pale-green and have arrowhead form their tubers have oval form. The third subgroup has accessions with large and heart-shaped leaves, Their tubers are elongated. The first accessions compose the N'Za subgroup, the second subgroup is called Florido and the last is Bêtê-Bêtê







The accessions of D. alata of the yam collection of CNRA in Côte d'Ivoire can be separated into four varietal groups. Bêtê-Bêtê and Brazo groups contain the largest numbers of accessions. The first group is characterised by the green colour of the aerial part of the plant and a white flesh of the tuber. This group was identified by Dumont and Jeanteur in 1988 as the one including the most cultivated yams varieties in Côte d’Ivoire. In New Caledonia, Lebot and al. (1998), identified a group of D. alata cultivars including 85 out of the 131 examined. This group of yams has very wide green leaves, and their plants are vigorous and have late maturity. These descriptions are identical to those observed in Bêtê-Bêtê, especially regarding the colour of the leaves. This group of New Caledonia can be assimilate to Bêtê-Bêtê of Côte d’Ivoire. Similarly, this group could also correspond to Pacala group of Caribbean (Degras, 1986). The genotypes composing this group have good organoleptic qualities (Degras, 1986). They also have a good postharvest storability (Girardin, 1996). However, most of the Bêtê-Bêtê are marked by the presence of Internal Brown Spot (IBS), which is a viral disease that attack the tubers and which origin is not well known (Girardin, 1996). When this spots are abundant, the tuber cannot be eaten (Nindjin et al., 2007). The Bêtê-Bêtê group is separated into three subgroups: Florido, N'Za and Suidié. Florido are characterized by fully green vegetative organs and white flesh of the tuber. These tubers have a fine texture. The round or oval shape of their tubers and their arrowhead-shaped leaves distinguish them from the other accessions of the Bêtê-Bêtê group. This subgroup is very uniform within the Bêtê-Bêtê group like the N'Za marked by their piriform tubers and dark green leaves. Thus, Contrary to the conclusions of Rodriguez (1983), who considered that Florido belongs to N'za group, it appears that Florido and N'Za are inside the large group Bêtê-Bêtê, but they constitute two different subgroups.

The Brazo group is characterized by the genotypes with green petiole and purple tips. Their flesh has orange colour. Most of them have recently been introduced in Côte d’Ivoire from Puerto Rico (Rodriguez, 1983; Dumont and Jeanteur 1988). Some accessions like EA12 and EM10, produce large tubers that bear roots over the whole surface. They therefore, have two characters, one of which (production of large tubers) is desires by the farmers and the other (presence of roots on the tuber) greatly depreciates the market value of the variety. However, the presence of roots on the tuber is influenced by the physical quality of the soil and can therefore be reduced. The genotypes of Brazo group are less sensitive to anthracnose and viruses and have very high yields often being above 30 tonnes per hectare. Most of them produce flowers and can be used in breeding schemes for the research of high yielding varieties resistant or tolerant to viruses and anthracnose. The coarse texture of the flesh of Brazo tubers could be an indication of their poor cooking quality when they are freshly harvested.

The Douoblé group represents only seven percent of the total number of accessions of the collection. This group is very homogeneous with the purple colour of the leaves, the stem of the young plant and the flesh of the tuber. The uniformity of this group can be explained by a monoclonal origin. Indeed, Hamon and al. (1986) observed that the Gnan group of the species D. rotundata of Côte d’Ivoire is monomorphic and attributed this characteristic to its monoclonal nature. These authors also noted the poverty of the vernacular nomenclature of this group. Similarly, yams of Douoblé group are designated in all regions of Côte d'Ivoire by names which translation means the violet coloration of their organs.

The Pyramid group is numerically the least important of the collection. It is represented by a single accession of the collection on which this study focused.

The varietal groups described in this work are different from those of Rhodes and Martin (1972) who, using 100 morphological characters have regrouped 30 accessions of D. alata in three groups after analyses using a principal component analysis and a hierarchical clustering. At the end of their studies, they classified Brazo Fuerte cultivars and Pyramid in the same group while Florido and Farmlisbon were regrouped in another one. In this study, Farmlisbon and Brazo Fuerte landraces are in the group Brazo, while Pyramid alone constitutes another different group. Indeed, the presence of thorns and the circular section of the Pyramid stem oppose it to the other accessions of D. alata. Considering the diversity of the organs that contain the purple colour within the D. alata species, it is not proper to consider that all the accessions that have a purple organ constitute a single varietal group. Only accessions which have the purple colour in the same organ can be classified in the same group.





The authors address deep gratitude to Mr ZAMBLE Tchambi and Mr COULIBALY Dogoba, technicians at CNRA and to CNRA.





Ayensu ES and Coursey DG (1972). Guinea yams. The botany, ethnobotany, use and possible future of yams in west Africa. Economy and Botany, 26 (4): 301-318.

Coursey DG (1967). Yams: An account of the Nature, Origins, Cultivation and Utilisation of the Useful Members of the Dioscoreaceae. Tropical Agriculture Series. Ed. Longmans,  230 p.

Degras L (1986). L’igname : Plante à tubercule tropicale. Techniques Agricoles et Productions Tropicales. Edition Maisonneuve et Larose et A.C.C.T.,  409 p. 

Doumbia S, Touré M and Mahyao A (2007). Commercialisation de l’igname en Côte d’Ivoire : état actuel et perspectives d’évolution. Cahiers Agriculture, 15 : 273 – 277.

Dumont R and Janteur P (1988). Bilan de cinq années de production, en grande culture, sur la variété florido (D. alata) dans la région centre de la Côte d’Ivoire. VIIth  symposium of the international Society for Tropical Root Crops, Gosier (Guadeloupe), 1- 6 july 1985, Ed. INRA, Paris ;179-193.

Egesi CN, Odu BO, Ogunyemi S, Asiedu R and Hughes J (2007).  Evaluation of Water Yam (Dioscorea alata L.) Germplasm for Reaction to Yam Anthracnose and Virus Diseases and their Effect on Yield Phytopathology 155, 536–543, Journal compilation Blackwell Verlag, Berlin.

FAOSTAT (2011). Faostat.fao.org/site/567 consulté le 02 juin 2011.

Foua-Bi K (1982. Etude de Aspidiella hartii Ckll (Homoptera, Diaspididae). Déprédateur des ignames en Côte d’Ivoire. Thèse Doctorat d’Etat, Université d’Abidjan, 209 p.

Girardin O (1996). Technologie après-récolte de l’igname : Etude de l’amélioration du stockage traditionnel en Côte d’Ivoire. Thèse de Doctorat d’Etat. Ecole Polytechnique Fédérale de Zurich, 137 p. Thèse EPFZ n° 11710.

Hamon P, Hamon S and Touré B (1986). Les ignames cultivées du complexe Dioscorea cayenensis-rotundata de Côte d’Ivoire. Inventaire des cultivars traditionnels, Ed IBPGR, FAO, 63 p.

IPGRI / IITA (1997). Descripteurs de l’igname (Dioscorea spp.). Institut international d’Agriculture Tropicale, Ibadan, Nigeria / Institut International des Ressources Phytogénétiques, Rome, Italie, 65 p.

Lebot V, Trilles B, Noyer JL and Modesto J (1998). Genetic relationships between Dioscorea alata L. cultivars. Genetic Resources and Crop Evolution, 45: 499-509.

Martin FW, (1976). Tropical yams and their potential. Part 3. Dioscorea alata, Agriculture handbook, 495: 40 p.

Nicole M, Nandris D, Digbeu S and Zohouri P (1990). Pathologie fongique de l'igname en Côte d’Ivoire: enquête 1989. Rapport technique, 105 p.

Nindjin C, Otokoré D, Hauser S, Tschannen A, Farah Z and Girardin O (2007). Determination of relevant sensory properties of pounded yams (Dioscorea spp.) using a locally based descriptive analysis methodology. Food Quality and Preference, 18: 450 – 459.

Obidiegwu JE, Asiedu R, Ene-Obong EE, Muoneke CO and Kolesnikova-Allen M (2009). Genetic characterization of some water yam (Dioscorea alata L.)  accessions in West Africa with simple sequence repeats. Journal of Food, Agriculture & Environment 7(3&4): 634-638.

Orkwor G (1998). The importance of yam, in Food yams: Advances in research, ed Orkwor G.C., Asiedu R. et Ekanayake I. J., p 13-17.

Rhodes AM and Martin FW (1972). Multivariate studies of variations in yam (Dioscorea alata L.). Journal of American Society of Horticulture, 97 (5) : 685-688.

Rodriguez H (1983). Intérêt d’une variété d’igname portoricaine en Côte d’Ivoire: Le Florido. Agronomie Tropicale, 38 (2): 153-157.

Scherrer (1984). Biostatistique. Eds Chicoutimic  Gaetan Morin, 850 p.



Cite this Article: Kouakou AM, Dibi KEB, N’zue B, Essis BS, Zohouri GP (2015). Dichotomous key determining varietal groups of yam species Dioscorea alata L. Greener Journal of Agricultural Sciences, 5(5): 190-203, http://doi.org/10.15580/GJAS.2015.5.082415121.