Parataxonomy and diversity of local varieties of taro (Colocasia esculenta (L.) Schott) and macabo (Xanthosoma sagittifolium (L.) Schott) grown in Benin (West Africa)

Abstract

Taro and macabo are plants with high food, medicinal and economic value, but are under-utilized in Benin. The aim of this study was to identify local varieties of taro and macabo grown in Benin, with their vernacular names, farmers' recognition criteria and distribution. Data were collected between May 2021 and May 2023, through interviews using semi-structured questionnaires. The sample surveyed consisted of 349 taro/macabo producers, distributed in 40 villages belonging to 9 departments. The variables collected concerned producers' socio-demographic characteristics, local appellations, farmers' recognition criteria and the distribution of local varieties of the two species, and were processed by calculating relative citation frequencies and establishing a generalized linear model in R software. In addition, QGIS software was used to represent the distribution of the varieties surveyed. A total of 42 local appellations have been registered to designate all local varieties. Ten local varieties, including six of C. esculenta (V1 to V6) and four of X. sagittifolium (V7 to V10), were identified on the basis of seven (07) farmers' criteria, the main ones being tuber color (41.88 %) and leaf blade color (23.30 %). Local varieties V1, V3, V4 and V5 were the most widespread for C. esculenta, while varieties V7 and V9 were the most widely distributed from the south to the north of the country for X. sagittifolium. The socio-demographic parameters such as age category, occupation and ethnic groups tested had a significant influence (p < 0.05) on the number of local varieties held by growers. Agro-morphological characterization will provide data for the selection of elite varieties.

1. Introduction

In Africa, roots and tubers make a significant contribution to the livelihoods of the poorest local communities [1]. These roots and tubers belong to the varied species such as: Manihot esculenta (cassava), Ipomoea batatas (sweet potato), Dioscorea cayenensis (Cayenne yam, yellow yam), Dioscorea rotundata (Guinea yam, white yam), Colocasia esculenta (taro) and Xanthosoma sagittifolium (macabo) [[2], [3], [4]]. Among these root and tuber species with high food potential, taro and macabo are the two species not only with low agricultural production but also having benefited from little scientific research [5,6].

In Benin, after maize, roots and tubers, including taro and macabo, are the second main source of calories in the diet [7]. However, the work carried out on roots and tubers, apart from cassava, yams and sweet potatoes, is not very diversified and/or focused on a restricted geographical area. This work has focused on varietal diversity, production constraints, cultivation practices and some ethnobotanical aspects of taro in a few municipalities in southern Benin [[8], [9], [10], [11]]. For these authors, two cultivars of taro are produced in Benin and their large-scale production is subject to numerous constraints despite its socio-economic importance. These constraints are mainly agronomic (low productivity, poor soils, long plant development cycle, lack of quality seeds, shrinking arable land, diseases, pest attacks, weeding difficulties), climatic (flooding) and economic (poor financial support). They therefore justify the scant attention paid to taro cultivation in Benin [12].

The analysis of this bibliographic synthesis testifies to the fact that studies carried out on the roots and tubers of Benin have obscured macabo (X. sagittifolium) which, however, has also been reported as a source of hydrocarbons like yams in Benin [3]. Therefore, the conclusions reported by the above authors (in terms of varietal diversity, production constraints, cultural practices and ethnobotanical aspects) could go beyond C. esculenta and also concern X. sagittifolium because the two species have the same vernacular names (taro in french and english) according to Akoègninou et al. [3]. In addition, cultivars of X. sagittifolium are sometimes presented as those of C. esculenta [8,9]. This presupposes the existence of a need for clarification in the work on the classification of these local varieties of taros held by producers.

In such a context, it is necessary to define approaches for managing constraints. Such approaches require detailed knowledge of the plant genetic resources of taro and macabo. Better still, [8,9,[13], [14], [15]]; and [11] have reported the existence of two cultivars (C. esculenta var. esculenta and C. esculenta var. antiquorum) of C. esculenta, with differences in tuber size and structure (main and secondary), leaf and bud color and distribution. Increased knowledge of the varieties of the two species will help to solve the problems of food insecurity and contribute to improving the cash income of growers through the use of high-productivity varieties. To achieve this, it is necessary to take stock of the current varietal diversity of taro and macabo grown in Benin. The aim of this study was to identify the local varieties of taro and macabo grown in Benin, with their vernacular names, farmers' recognition criteria and distribution.

2. Material and methods

2.1. Study area

The study was carried out in Benin, bordered to the north by Niger and Burkina Faso, to the south by the Atlantic Ocean, to the east by Nigeria and to the west by Togo (Fig. 1). Benin is subdivided into three phytogeographical zones: the Guinean zone in the south, the Sudano-Guinean zone in the center and the Sudanian zone in the north. Three rainfall regimes have been identified in Benin. In the north, the climate is dry tropical with 2 seasons, characterized by a long dry season from November to May and a long rainy season from June to October. Average annual rainfall is 1000 mm. In the center, the climate is transitional humid tropical, with either 4 or 2 climatic seasons. In the south, the climate is sub-equatorial, with 4 climatic seasons: a long dry season from mid-November to mid-March, a long rainy season from mid-March to mid-July, a short dry season from mid-July to mid-September and a short rainy season from mid-September to mid-November. Average annual rainfall in this part of the country is 1200 mm [3].

Fig. 1.

Map showing the geographical distribution of the villages surveyed.

Two main soil types are encountered. The first type is made up of drained soils, including ferralitic soils, ferruginous soils, concretioned or not, with or without cuirasses, rough mineral soils and eutrophic brown soils. The second type comprises hydromorphic soils with temporary or permanent hydromorphy, including vertisols and halomorphic soils [3]. The population is estimated at 12,909,041 [16], divided into nine (9) sociolinguistic groups and 60 national languages [17].

2.2. Data collection

Data were collected from May 2021 to May 2023. The first stage consisted of investigations at the level of the structures in charge of agriculture in Benin, such as the 07 “Agences Territoriales de Développement Agricole (ATDA)" and the 12 “Directions Départementales de l'Agriculture, de l'Elevage et de la Pêche (DDAEP)". This made it possible to identify the taro and macabo producing departments in Benin. The survey revealed that taro and macabo are produced mainly in nine (09) departments: Atlantique, Zou, Collines, Mono, Couffo, Ouémé, Plateau, Atacora and Donga. The municipalities and villages surveyed were chosen on the basis of their long history of production of these species, ethnic diversity and production habitat (shallow, flooded land).

It was planned to select 5 villages in each department out of the nine (09) identified, and then choose 10 informants per village, i.e. a total of 50 informants per department and 450 informants for all these departments. However, due to an insufficient number of informants in some villages, the number of villages was reduced to 4 in some departments, and in the end, 349 informants took part in the survey (Table 1).

Table 1.

Distribution of sample size by socio-demographic characteristics of informants.

Characteristics	Modalities	Number of informants	Relative frequency
Gender	Woman	67	19.20
	Man	282	80.80
Age categories	Young (age ≤35)	128	36.68
	Adult (35 < age ≤60)	202	57.88
	Old (age >60)	19	5.44
Education level	Illiterate	175	50.14
	Primary	98	28.08
	Secondary	72	20.63
	Higher	4	1.15
Occupation	Farmer	303	86.82
	Craftsman	31	8.88
	Trader	3	0.86
	Civil servant	11	3.15
	Traditional practitioner	1	0.29
Ethnic groups	Adja	70	20.06
	Aïzo	33	9.46
	Fon	50	14.33
	Goun	32	9.17
	Holli	52	14.90
	Kotafon	8	2.29
	Lokpa	10	2.87
	Nago	19	5.44
	Ottamari	25	7.16
	Wémè	50	14.33

Data were collected using face-to-face individual interviews based on a semi-structured questionnaire. The goal of the study in terms of gathering knowledge on C. esculenta and X. sagittifolium for a PhD research (this study was part of the PhD research of the first author) was first explained to village authorities and next to each of the informant as to have their consent for participation before starting the interviews. Only individuals that consented to participate in the study were considered. Interviews were conducted with assistance of a local translator when necessary. The variables collected were: the socio-demographic characteristics of the informants, the local varieties grown and their vernacular names, their meanings and the farmers' varietal recognition traits, as well as their distribution in Benin. Each interview lasted between 1 h and 1 h 30 min. Accessions (tuber shoots and leaves) were collected for photography and the installation of an experimental field.

2.3. Data analysis

The relative frequency of citation (RFC) were calculated for the various parameters considered using the formula adapted from Friedman et al. [18]:

$$RFC=\frac{x}{n}\times 100$$ (1)

where x is the number of informants who mentioned a parameter and n is the total number of informants.

As the number of varieties grown is a count data, a generalized linear model with a Poisson distribution was used to test its variability according to the socio-demographic parameters (gender, age category, education level, occupation and ethnic groups) of informants [19].

The software used for this purpose was R version 3.5.0 [20]. In addition, QGIS software version 3.28 was used to produce distribution maps of local varieties of the two species.

3. Results

3.1. Vernacular nomenclature of local varieties of taro and macabo grown in Benin

A total of 42 local names (13 for taro and 29 for macabo) were identified for the local varieties of taro and macabo grown in Benin (Table 2). Most of these names (71.88 %) were made up of two words, the first meaning “taro” and the second highlighting the color of the tuber's flesh. This was followed by single-word designations (22 %) simply translating “taro” in X. sagittifolium. In C. esculenta, the single-word name means both “taro” and its production habitat. The Aïzo, Fon and Goun, for example, refer to C. esculenta as “toglin”, i.e. taro from water, rivers or swamps. Finally, there were names composed of three words (6.11 %), the first two of which reflect the cases illustrated above, and the third of which refers to the size of the tuber. The Fon, for example, refer to X. sagittifolium as “glin wéwé kpêvi", translating as “small white taro” (Table 2).

Table 2.

Vernacular nomenclature of local taro and macabo varieties with their meanings.

Ethnic groups	Vernacular names of the two species and their meanings
	Colocasia esculenta	Xanthosoma sagittifolium
Adja	–	Makana, glin, glin wéwé5, glin vovo6, makana djin ou makana édjonto6, makana épépé5
Aïzo	Toglin1	Glin wéwé5, glin vovo ou glin vêvê ou glin vêê6
Fon	Toglin1, toglin wéwé2, toglin vovo3	Glin wéwé5, glin wéwé kpêvi8, glin vovo6
Goun	Toglin1	Glin wéwé5, glin vêvê6, glin vêvê kpêvi9
Holli	Ikoko iré4, ikokoodo1	Ikoko foufou ou ikoko founfoun5, ikoko founfoun koukourou8, ikoko kpikpa6, ikoko ognibo10, ikoko kpikpa gougou11, ikoko kpikpa koukourou9
Kotafon	–	Bagali ou glin ou glin woué12, glin vê6, glin houé5
Lokpa	–	Mankéni kouhouloum5, mankéni kousém6
Nago	Ikokoodo founfoun2, ikoko iré4	Ikoko founfoun5, ikoko kpikpa6, ikoko kpikpa gougou11, koukou
Ottamari	Eyéibé, yacorgo, yèkotonko ou dikotonko, yakoukou-pia5, yakoukou-ouan6	–
Wémè	Toglin ouéto7, toglin wéwé2, toglin1	Glin vêvê6, glin vêê gaa11, glin vêvê kpêvi9, glin wéwé5, ayoglin ou glin wéwé awouéto ou kouvitoglin13

The superscript numbers of the vernacular names reflect their meaning (1: river taro, 2: white water taro, 3: red water taro, 4: true taro, 5: white taro, 6: red taro, 7: false river taro or second river taro, 8: small white taro, 9: small red taro, 10: white man taro, 11: long red taro, 12: yam, 13: second white taro or false white taro or taro of the resurrected).

These different names were composed of synonyms within each species. For instance, Ottamari, refer to C. esculenta as Eyéibé, Yacorgo, Yèkotonko or Dikotonko, Yakoukou-pia. The Wémè also refer to X. sagittifolium by several names (Ayoglin or Glin wéwé awouéto or Kouvitoglin).

3.2. Diversity of local varieties of taro and macabo

A total of ten (10) local varieties were identified. Of these, six belong to the C. esculenta species (V1, V2, V3, V4, V5 and V6) and the other four to X. sagittifolium (V7, V8, V9, V10). Table 3 summarizes the botanical characteristics and Fig. 2, Fig. 3 illustrate each of these local varieties.

Table 3.

Vernacular names and botanical characteristics of local varieties of taro and macabo.

Local varieties and their vernacular names	Botanical description of local varieties
V1 (toglin (3,2), ikokoodo (8,5), toglin wéwé (3))	Leaf: peltate; leaf blade dark green with a medium-sized violet dot on the blade at the petiolar junction; petiole with violet apex on the outside and whitish-green on the inside; pale-green sheath with brown border; white ring. Tuber: white flesh mottled with violet. Cycle length: 6–12 months
V2 (toglin, toglin vovo (3))	Leaf: peltate; leaf blade dark green with a violet dot at the petiolar junction; petiole with violet apex on the outside and green on the inside; pale-green sheath; light pink ring. Tuber: beige flesh with purple flecks Cycle length: 6–12 months
V3 (yakoukou-ouan (9), yacorgo, yèkotonko, dikotonko (9), ikoko iré (5))	Leaf: peltate; leaf blade dark green with small violet dot at petiolar junction; petiole purple; sheath purple; white ring Tuber: yellowish-white flesh Cycle length: 6–12 months
V4 (éyéibé (9), yacorgo, yèkotonko, dikotonko(9), ikoko iré (5))	Leaf: peltate; green leaf blade; green, violet-striped petiole; green, purple-striped sheath darker than the petiole; white ring Tuber: white or beige flesh Cycle length: 6–12 months
V5 (yakoukou-pia (9), yèkotonko ou dikotonko, ou yacorgo (9), ikoko iré (5))	Leaf: peltate; leaf blade yellowish-green; petiole yellowish-green; sheath yellowish-green; ring white Tuber: white flesh Cycle length: 6–12 months
V6 (toglin ouéto (4))	Leaf: peltate; leaf blade dark green with a large violet dot at the petiolar junction; petiole with violet apex on the outside and yellowish green on the inside; slightly brownish green sheath; whitish ring Tuber: white flesh spotted with brown. Cycle length: 6–12 months
V7 (makana, makana djin, makana édjonto (1), glin vovo (3), glin vêvê, glin vêê (4,2,10), ikoko kpikpa (8,5), Mankéni kousèm (7), koukou (8))	Leaf: sagittate; leaf blade dark green; petiole green ± whitish; violet sheath with purple inner border; pink ring Tuber: pink flesh Cycle length: 6–12 months
V8 (glin vêvê ou glin vêê (4,10))	Leaf: sagittate; dark green leaf blade; green petiole; light-violet sheath with purple inner border; light pink ring. Tuber: light pink flesh Cycle length: 6–12 months
V9 (makana, makana épépé (1), glin wéwé (3,2,4), glin woué (6), Mankéni kouhouloum (7), ikoko foufoun (8 et5))	Leaf: sagittate; light-green leaf blade; pale-green petiole; pale-green sheath; white ring Tuber: white flesh Cycle length: 6–12 months
V10 (ayoglin, glin wéwé, awouéto, kouvitoglin (4))	Leaf: sagittate; dark-green leaf blade; green ± white petiole; violet sheath with purple inner edge; light-pink ring Tuber: beige flesh Cycle length: 6–7 months

V1–V6: Local varieties of C. esculenta, V7–V10: Local varieties of X. sagittifolium. The superscript numbers of vernacular names indicate the ethnic groups (1: adja, 2: aïzo, 3: fon, 4: goun, 5: holli, 6: kotafon, 7: lokpa, 8: nago, 9: ottamari, 10: wémè).

Fig. 2.

Illustrations of local varieties of C. esculenta.

(V1: variety 1, V2: variety 2, V3: variety 3, V4: variety 4, V5: variety 5, V6: variety 6; a. plant, b. leaf blade, c. petiole and sheath, d. tuber slice).

Fig. 3.

Illustrations of local varieties of X. sagittifolium.

(V7: variety 7, V8: variety 8, V9: variety 9, V10: variety 10, a. plant, b. leaf blade, c. petiole and sheath, d. tuber slice).

The number of local varieties grown ranged from 1 to 4. Single-variety growers were the most numerous, accounting for 79.94 % of all those surveyed. They were followed by those who grow two varieties each (18.91 %). Those who individually grow the largest number of local varieties (3 and 4) were in the minority (0.86 % and 0.29 % respectively).

Socio-demographic characteristics showed that men cited a total of ten local varieties, while women cited only four. In terms of age categories, adults held all (ten) varieties, compared with six among the young and four among the old. Concerning education level, illiterates and those enrolled at primary level each cited 9 varieties; those at secondary level cited 8 varieties, and those at higher level cited only two.

Regarding of occupation, farmers cited the greatest number of varieties (ten). Next came traders (6 varieties), craftsmen (5 varieties), traditional practitioners (4 varieties), and civil servants (3 varieties). In terms of ethnic groups, the Holli and Wémé recorded the highest varietal richness (6 local varieties each). They were followed by the Fon (4 local varieties), the Aïzo, Goun, Nago and Ottamari (3 local varieties each). The lowest varietal richness was observed among the Adja, Kotafon and Lokpa (2 local varieties each).

The socio-demographic parameters such as age category, occupation and ethnic groups tested had a significant influence (p < 0.05) on the number of local varieties held by growers (Table 4).

Table 4.

Linear models showing the effects of socio-demographic parameters on the number of local varieties.

	Estimate	Std. Error	z/t value	Pr (>|z|)
(Intercept)	0.03	0.15	0.20	0.001
Gender	0.07	0.15	0.47	0.637
Age category	−0.01	0.12	−0.80	0.042
Occupation	−0.01	0.23	−0.04	0.001
Ethnic groups	0.48	0.19	2.46	0.014
Education level	−0.13	0.59	−0.21	0.834

Note. Std.: standard error.

3.3. Farmers' criteria for recognizing local varieties of taro and macabo

Farmers cited seven and six criteria respectively for recognizing local varieties of taro and macabo (Fig. 4). The most frequently cited criteria for C. esculenta were: color of tuber flesh (25.07 %), aspect of leaf blade cleft (18.88 %), leaf blade color (17.70 %) and color of petiolar junction (13.86 %). For X. sagittifolium, the most frequently cited criteria were: color of tuber flesh (42.69 %), color of petiole (23.43 %) and color of leaf blade (23.28 %). Color diversity at the petiole junction is exclusively characteristic of local varieties of C. esculenta.

Fig. 4.

Variation in the relative frequency of citing local variety recognition criteria.

3.4. Modality variation within the recognition traits of local varieties of taro and macabo

Informants cited several variables or modalities per recognition trait for local varieties of C. esculenta and X. sagittifolium (Table 5). In the case of C. esculenta, the presence of a cleft, but not reaching the petiolar junction, was strongly cited (70.20 %). Leaf blades were dark green (24.2 %), light green (18.74 %) and pale green (15.11 %). The presence of three different colorations at the petiolar junction was reported, with violet (43.42 %) and green (30.31 %) as the predominant colors. The petiole and sheath were generally purple (25.16 %) and pale green (10.26 %). White (30.26 %) and light pink colors have been reported on the ring or collar of plants. Varieties of this species produce tubers with white flesh (38.16 %) or beige flesh spotted with purple (13.16 %).

Table 5.

Distribution of modalities within the recognition traits of local varieties of taro and macabo.

Recognition traits	Modalities	Relative frequency of citation (%)
		C. esculenta	X. sagittifolium
Aspect of leaf blade cleft	Cleft not reaching petiole	70.2	3.7
	Cleft reaching petiole	1.8	61.84
Color of leaf blade	Light green	18.74
	Dark green	24.2	30.04
	Pale green	15.11	23.92
	Yellowish green	1.31	–
Color of petiolar junction	Violet	43.42	–
	Green	30.31	10.11
	Yellow	11.84	–
Color of petiole	Light green	–	5.19
	Whitish green	–	24.78
	Purple	25.16	–
	Pale green	13.16	–
	Yellowish green	1.31	–
Color of sheath	Pale green	10.26	–
	Light green	–	2.44
	Mottled purple green	2.63	–
	Violet	–	9.91
	Purple	3.95	–
	Yellowish green	1.31	–
Color of ring	White	30.26	9.15
	Pure pink	1.6	25.56
	Light pink	13.95	2.59
Color of tuber flesh	White	38.16	32.85
	Rose	–	48.7
	Beige	–	0.86
	Light white	1.22	0.86
	Yellowish-white	2.63	–
	Purple-spotted beige	13.16	–
	Beige spotted with intense purple	3.94	–

In X. sagittifolium, the presence of a cleft reaching the petiolar junction was reported (61.84 %). The leaf blades were dark-green (30.04 %) and pale-green (23.92 %), with whitish-green (24.78 %) and violet (9.91 %) petioles and sheaths respectively. The pink color (25.56 %) of the ring of most varieties of this species was mentioned. Tubers with pink (48.70 %) and white (32.85 %) flesh were reported in most cases for this species (Table 5).

3.5. Distribution of local varieties taro and macabo in the departments surveyed

The greatest diversity of local varieties (6 local varieties) was observed in the Oueme department (Fig. 5). These were varieties V1, V6, V7, V8, V9 and V10. In addition, 6 local varieties were recorded in the Plateau department (V1, V3, V4; V5; V7 and V9). Next in line was the Zou department, with 4 local varieties (V1, V2, V7 and V9). This department was followed by Atlantique: 3 local varieties (V1, V7 and V9); Atacora: 3 local varieties (V3, V4 and V5); Mono and Donga: 2 local varieties each (V7 and V9) and finally the departments of Couffo and Collines: 1 local variety (V7) each.

Fig. 5.

Map showing the distribution of local varieties of taro and macabo in the departments surveyed.

The local variety V7 was present in 08 of the 09 departments surveyed and represents the most widely grown local variety. It was followed by V9, present in 06 departments. Next come variety V1, present in 04 departments, and varieties V3, V4 and V5, present in 02 departments. Varieties V2, V6, V8 and V10 were found in only one department each (Fig. 5).

4. Discussion

This study provided a directory of vernacular names for local varieties of taro and macabo grown in Benin. This diversity of vernacular names, recorded for each of the two species across the country, testifies to their importance to local communities [21]. The name “toglin” attributed to C. esculenta by some ethnic groups, meaning water taro, highlights the species' preferred habitat. Similar observations have already been reported on the species by Quenum et al. [11]. In the case of X. sagittifolium, the common name used by most ethnic groups is “glin”, which simply refers to taro. This suggests that X. sagittifolium, unlike C. esculenta, has no edaphic affinity for flooded land. This is in line with the conclusions of Spichiger and Jeanmonod [22], who argue that the use of macroscopic ecological, utilitarian or botanical characters in nomenclature and species identification dates back to antiquity, not to the age of scientists, and continues to be applied by local communities to this day. In addition to these specific clarifications, growers use local names made up of two or even three words to designate local varieties of both species. These findings once again confirm the importance of parataxonomy as a taxonomic basis for distinguishing plant species [23].

The present study also collected ten local varieties, including six of C. esculenta and four of X. sagittifolium, based respectively on seven and six farmers' criteria for recognizing these two taxa. This varietal richness recorded for C. esculenta is higher than that of 3 reported for the same species in Côte d'Ivoire by Koffi and Koffi [24]. It is also significantly higher than the two cultivars mentioned by Houngbo et al. [8]; Akplogan et al. [9] and Quenum et al. [11] as those grown in Benin. Moreover, the two cultivars of C. esculenta revealed by the work of Houngbo et al. [8] and Akplogan et al. [9] are rather for those of X. sagittifolium. Furthermore, the six local varieties of C. esculenta identified in this study are grouped into the two botanical varieties of C. esculenta (C. esculenta var. esculenta and C. esculenta var. antiquorum) also reported by Quenum et al. [11]. The discrepancies in the number of local varieties between this study and those of Akplogan et al. [9]; Koffi and Koffi [24] and Quenum et al. [11] could be justified by the use of a relatively higher number of identification criteria (6–7) in the present study (whole Benin) compared with the five common criteria used for both species in the work of Koffi and Koffi [24] and the four criteria used by Akplogan et al. [9]. Added to this would be the limited extent (Southern Benin) for the work of Akplogan et al. [9] and Quenum et al. [11].

Compared with the work of Koffi and Koffi [24] in Côte d'Ivoire, the findings showed 06 common varieties, including 03 in C. esculenta (V1, V3 and V5 corresponding to M4, M5 and M6) and 03 in X. sagittifolium (V7, V8 and V9 corresponding to M1, M2 and M3). On the other hand, the M7 variety identified by Koffi and Koffi [24] has not been recorded in Benin, while V10 is a new variety in relation to the work of Koffi and Koffi [24]. The divergence in botanical characteristics observed between V10 from Benin and M7 from Côte d'Ivoire could be linked to environmental factors, as Onomo et al. [25] have already pointed out, for whom biotic and abiotic factors can govern phenotypic variations in plants in their respective environments. This suggests the need for genetic characterization to elucidate the dissimilarities [26].

The recognition traits of local varieties, reported by growers in both countries (Benin and Côte d'Ivoire), retain the colorations of the leaf in its various parts and those of the tuber flesh. This corroborates the findings of Ouédraogo et al. [27] in Burkina Faso, regarding the importance of the discriminating character of these plant parts. The criterion “aspect of the leaf blade cleft”, which discriminates between the two species with a cleft not reaching the petiolar junction, characteristic of a peltate leaf in C. esculenta, and one reaching the petiolar junction, characteristic of a sagittate leaf in X. sagittifolium, is in line with the work of Jennings [28], regarding the use of leaf shape as a distinctive criterion between the genera Colocasia and Xanthosoma.

In C. esculenta, V1, V3, V4 and V5 varieties appear to be the most widespread across the country. As for X. sagittifolium, the V7 and V9 varieties are rather the most distributed from the South to the North of the country. These findings are consistent with those of Koffi and Koffi [24], who mentioned the wide distribution of the V7 variety in Côte d'Ivoire. According to Lebot [1], the distribution and diversification of “taro” varieties can be explained by the influence of some factors, in particular the geographical isolation of wild “taro” populations over long periods, natural selection and human selection.

This study showed the importance of age category, occupation and ethnic group in the variation of local varieties held by growers. Similar observations were reported in the work of Tumuhimbise et al. [29] on root and tuber cultivation in Uganda. Indeed, it has been shown in ethnobotanical studies that adults and the elderly hold more knowledge unlike the young [30]. In addition, occupation proved significant in the present study. This would be justified by the fact that among the five types of occupation studied, farmers are strongly represented (303 out of 349). They remain the custodians of endogenous knowledge linked to taros and macabos.

By taking stock of local varieties of C. esculenta and X. sagittifolium, this study highlighted the botanical characters that distinguish each of them. Thus, the results clarified the botanical confusions that remained in the work of Houngbo et al. [8] and Akplogan et al. [9]. However, socio-demographic variables such as price, yield, distance, social network, access to credit, not tested in this study, could influence knowledge about C. esculenta and X. sagittifolium [29].

5. Conclusion

Ethnobotanical surveys identified 42 vernacular names for the 10 local varieties, including 06 for C. esculenta and 04 for X. sagittifolium. The two species C. esculenta and X. sagittifolium are called “toglin” and “glin” respectively. Seven farmer criteria are used to determine local varieties of C. esculenta and six criteria are used for local varieties of X. sagittifolium. Criteria such as leaf blade color, tuber flesh color, leaf cleft appearance, petiole color, sheath color and ring color are common to both species. The local varieties recorded show different distribution spectra in Benin. Socio-demographic parameters such as age category, occupation and ethnic groups tested had a significant influence (p < 0.05) on the number of local varieties held by growers. These results are a first step towards future research into agro-morphological and molecular genetic characterization, with a view to identifying elite varieties. Such identifications are essential for reducing poverty and combating food insecurity, particularly in sub-Saharan Africa.

Data availability

The datasets generated during the current study are available from the authors upon request.

Additional information

No additional information is available for this paper.

Funding statement

This work has not received specific funding, however Gbèdomèdji Hurgues Aristide Houénon was supported by the International Science Foundation through a research grant (No. I-1-D-6644-1), which was instrumental in the statistical analysis and writing of the original version.

CRediT authorship contribution statement

Affiavi Aurore Sylvie Tamadaho: Writing – original draft, Visualization, Validation, Resources, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Gbèdomèdji Hurgues Aristide Houénon: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation. Rose Fernande Fagbédji: Writing – original draft, Investigation. Dègninou Yélognissè Innocent Ahamidé: Writing – original draft, Investigation. Hounnankpon Yédomonhan: Validation, Supervision, Methodology.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.