Cultivation and possible domestication of feral and possibly wild yams (Dioscorea spp.) in Southwest Ethiopia: ethnobotanical and morphological evidence

ABSTRACT

The far Southwest Ethiopians transplant wild plant species to their gardens. One of such plant is the Dioscorea that we studied to assess the knowledge of wild yam and process of domestication. The study links two types of evidence to obtain insight about the process of yam domestication. We analyze two data sets derived from (1) ethnobotanical survey using 231 semi-structured interviews; and (2) morphological study in 47 yam accessions. Our study revealed that domestication is still active in some villages. Knowledge of yam domestication was shared by 44% of the farmers’ even by those that have never practiced its domestication. Farmers who can describe the trend of domestication and the morphotypes of domesticate represented 21 and 28%, respectively. Farmers who have recent transplants in their garden varied from 4% in Bench to 10% in Sheko. The domestication process described by the two ethnic groups is similar. The duration of domestication can take up to six years, but with most of the individuals, it only takes three to five years. By linking the two types of evidence, two evolutionary processes are distinguished: (1) populations of recent domesticate expressing a domestication syndrome possibly belongs to the wild D. abyssinica or D. praehensilis, and (2) plants of incipient domesticate that might be derived from volunteers or diverse types of hybrids. Each of these processes can lead to integration of wild genotypes into the cultivated gene pool, and hence, enhance genetic diversity of cultivated yams. The domestication practices of traditional farmers should thus be taken into account if yam conservation and improvement plans need to be established.

1. Introduction

The loss of diversity of cultivated crops has been the subject of considerable concern in the past three decades.¹ Genetic erosion in the form of local crop varieties has been described as a loss of plants with potential agricultural value, and has implications for food security. Food security is not just about ensuring that food is available; it also includes ensuring that there is sufficient diet diversification. The knowledge and management of local farmers is a key for the maintenance of local agro-biodiversity. The traditional farmer has maintained a large number of crop diversities through domestication.^2–6

Domestication is a continuous evolutionary process guided by humans, mainly through artificial selection.⁷ Domestication is the outcome of a selection process that leads to an increased adaptation of plants to cultivation and utilization by humans.⁸ It involves a variety of selection episodes, but of all cultivated crop species globally, such episodes have been cataloged for relatively few species. Numerous ongoing processes of domestication have yet to be studied. They could be a valuable source of information to understand the diversity of forms through which humans drive plant evolution and to reconstruct and understand the earlier phases of domestication.⁹

Domestication usually involves drastic genetic changes in the material, but in the case of yam, it involves the adaptation of wild plants into cultivated forms without genetic changes.¹⁰ To avoid confusion, Scarcelli et al.^11,12 proposed the term spontaneous for plants growing without farmers help, and is a general term referring to hybrids, volunteers, and even wild plants. In accord, many authors regarded such plants as wild plants.^5,10 The occurrence of distinct yam species in Africa predates human history, and domestication of these species appears to have been done by local human populations. Studies carried out in different parts of Africa showed that several wild yams have been brought under cultivation by African farmers through domestication .^11–17 D. cayenensis and D. rotundata are one of the earliest domesticated yams in Africa, and D. abyssinica and D. praehensilis have been proposed as probable wild progenitors.^5,10 Studies by ¹⁴, and Scarcelli et al.^11,12 obtained molecular evidence in support of this opinion.

The adaptation process in yam thus involves intense vegetative multiplication and selection procedures of those plants to obtain modifications mainly in form, size, and taste of its tuber. It can take a variable period of time, but with most of the individuals, it takes only three to five years.⁵ During this period, farmers submit the tuber of spontaneous plants to stress in order to induce some modifications. If farmers are satisfied with the modifications, the tubers are mixed with tubers of a similar variety, or they may rename it to a variety of domestic yams it resembles most closely.^11,12 This trend differs from one ethnic group to another within a given area. It is thus useful to document the process among the distinct ethnolinguistic groups within a given geographic locations.

This paper presents the practices of wild yam domestication in Ethiopia. The definition of domestication in the case of yam was adopted from earlier studies.^5,10,11 Accordingly, ‘domestication’ is defined as the adaptation of spontaneous plants to cultivation contexts without genetic changes. According to this definition, ‘spontaneous’ is a general term referring to feral (volunteers or hybrids) and possibly wild plants growing without farmers help. Here, we use the term ‘wild’ in reference to plants growing spontaneously, while we use the name ‘wild transplant’ and ‘domestic’ in reference to plants that are under domestication and cultivation, respectively.

Ethiopia represents one of the main centers in the world where plant diversity is noticeably high and some of them concerned became domesticated.^19–22 The better-known root and tuber crops such as Oromo potato (Plectranthus edulis), anchote (Cocciniaa abyssinica), enset (Ensete ventricosum), and bagana (Amorphophallus abyssinicus) were domesticated in Ethiopia.²³ Furthermore, several yam species might have their origin in Ethiopia and are one of those crops with wild relatives in the country.^24,25 These and other reports contributed a lot toward exploring yam species and their distributions in Ethiopia. Among others,²⁶ reported that the genus Dioscorea has eleven species in Ethiopia, some of which exist both in wild and cultivated forms. This provides an excellent setting for studying the effect of selection in a gradient of management intensity.

Compared with other areas in Ethiopia, those of the far Southwest Ethiopians have a strong tradition of yam farming and excellent knowledge of domestication. But, this great wealth of native knowledge has not been well documented. In an attempt¹⁵,has documented the motives and opportunities of domestication by the Sheko farmers. Nothing has been documented regarding the domestication trends by the Bench, upland Omotic speaking peoples to the east of Sheko. Recently, this knowledge is beginning to deteriorate as farming reorients toward cash crops and in fact convinced evidences confirmed that yam farming is in danger of being replacement by coffee farming. Authors own observation also suggest that yam is one of the most understudied root crops, with high conservation and improvement concern in Southwest Ethiopia. Thus, the practice of wild yam manipulation by the local farmers emerges as an essential research direction to focus on.

The results of this study will be compared with our ongoing molecular study to link this knowledge with the scientific parameters and will be published later in more details. Here, we describe ethnobotanical and morphological evidence of domestication. The objective of the present study was (1) to investigate the knowledge of wild yam and wild yam manipulation; (2) to assess the process of yam domestication; and (3) to link ethnobotanical and morphological evidences in order to obtain insights into its evolution under domestication.

2. Materials and methods

2.1. Ethnobotanical evidence of domestication

The data derived from ethnobotanical assessment allow for the recognition of three main groups of yam population. These are (a) yams that are growing wild; (b) yams that have been recently transplanted from wild location to garden; and (c) yams that are of under cultivation.

Some farmers in Southwest Ethiopia manipulate wild growing yams, bringing them from uncultivated areas to transplant into their gardens. Wild-growing yams are found in un-cut forest, secondary forest, fallow lands, and ravines or stream banks with disturbed vegetation. They have no known history of human husbandry, although it is possible that they are feral or have genetic material derived from nearby populations of cultivated yams. Many gardens contain a few yams that seem on the state of transition, yams transplanted from wild contexts where many of them had no known history of cultivation. They are usually grown in gardens beneath tree. Unlike the domestic yams that are replanted in an annual cycle, they left in the same place for many years. Hence, they qualify as yams under domestication, and not just as yams under cultivation. We, therefore, designate them as ‘wild transplants’ to differentiate it from wild and domestic yams.

Domestic yams, in contrast, have a long history of cultivation and propagation in gardens by humans. They are usually grown in small plots of open field along rows of stakes and are usually replanted in an annual cycle. Some domestic yams such as Baidai-kachi, Chabsha, Torbay and Tolubab are known to be recently transplanted to small plots of open field. Some of these names can also be used interchangeably for group of yams that have been recently transplanted from wild contexts. Farmers relate certain wild, wild transplant, and domestic varieties to one another. They are also capable of reporting both specific cases and general patterns of adoptive transplantation. If the wild transplant proves to resemble a domestic variety, farmers may give it the domestic variety name a few years after transplantation.

This study was conducted in Bench-Maji and Sheka zones of Southwestern Ethiopia. Ethno-botanical evidence of domestication was assessed based on two groups of data derived from (1) 231 semi-structured interviews aimed at documenting the knowledge of wild yam and wild yam manipulation; and (2) 32 semi-structured interviews aimed at assessing the detail process of yam domestication. The knowledge of wild yam and domestication were assessed based on questions like (i) knowledge of wild yam (ability to identify the different types of wild yams, and knowledge on the existence of wild yam in their environment, abundance and distribution of wild yams, time of wild yam availability and use wild yams as food), (ii) knowledge of domestication (to know that wild yams can become cultivated yams by domestication, to know a simple knowledge of domestication or ability to describe the technique of domestication), (iii) ability to describe the morphotypes or product of domestication (ability to describe the distinct morphotypes between wild and domesticable yams or ability to describe cultivars currently known as the result of domestication) and (iv) actual practice of it.

Semi-structured interviews were conducted with households who are currently practicing domestication. The details of the domestication process were assessed by collecting data on the type and source of wild material used for domestication, how these materials were gathered, the areas where these wild yams were collected and the procedure of the domestication itself. Questions asked to farmers were included; years of cultivation needed to obtain morphological types similar to those of cultivated types, the morphological variations that were observed during the process and the techniques used to transform the wild yam into cultivated types.

2.2. Morphological evidence of domestication

Domestication of yam was assessed based on the data derived from morphological study in 47 accessions corresponding to 17 differently named landraces of the D. cayenensis complex. Morphological evidence was assessed in each of the accessions in two replicates and performed according to the standard procedures previously employed to characterize yams. Formal descriptors for yam had already been developed and have been improved at several instances.^27,28 In this study, we have used the descriptors suggested by 18. The local name and geographic location of the accessions considered in our study are presented below (Table 1). Characters which are frequently used by farmers for identification were considered in this study. Accordingly, we have selected 24 characters; six are related to leaf traits, seven are related to stem traits and eight are related to the tuber traits. For each accession, flowering ability, sex, and phenological related traits were also recorded (Table 2). Data collection procedures were performed in three different stages including establishment, growth, and harvest stage. Data on establishment stage were started shortly after emergence, which was then regularly monitored throughout the growing period. Data on young stems and leaves were assessed after about 20–30 d from emergence. Data on mature stems and leaves were assessed before senescence, while data on tubers were assessed by harvesting tubers at full senescence. The measurements made on qualitative traits were transformed into different classes, attributing codes to each class. Each character was scored as variable classes ranging from 1 to N, where N is the number of scales (Table 2).

Table 1.

Region and local names of yam accessions considered in this study

No.	Accession code	Local name	Area of collection	Altitude (m asl)	Latitude (N)	Longitude (E)
1	BMJ 001	Karka-Kachi (WT)	Sheko	1473	7°2ʹ21.16”	35°31ʹ10.35”
2	BMJ 002	Karka-Kachi (WT)		1696	7°2ʹ34.05”	35°31ʹ10.65”
3	BMJ 003	Karka-Kachi (W)		1300	7°2.435’	35°33.452’
4	BMJ 004	Karka-Kachi (WT)		1677	7°2.435’	35°31.180’
5	BMJ 005	Karka-Kachi (WT)		1680	7°2.358’	35°31.147’
6	BMJ 006	Karka-Kachi (W)		1266	7°1.179’	35°33.498’
7	BMJ 007	Yasind (WT)	Semen-Bench	2083	7°4ʹ40.25”	35°42ʹ34.79”
8	BMJ 008	Beri	Sheko	1645	7°2ʹ18.30”	35°31ʹ10.29”
9	BMJ 009	Torbay		1645	7°2ʹ18.30”	35°31ʹ10.29”
10	BMJ 012	Tsano		1680	7°2.358’	35°31.147’
11	BMJ 013	Dizzu Kachi		1296	7°2ʹ34.05”	35°31ʹ10.65”
12	BMJ 014a	Tsano		1673	7°2ʹ21.16”	35°31ʹ10.35”
13	BMJ 014b	Tsano		1673	7°2ʹ21.16”	35°31ʹ10.35”
14	BMJ 015	Tsano		1696	7°2ʹ34.05”	35°31ʹ10.65”
15	BMJ 018	Beri		1645	7°2ʹ18.30”	35°31ʹ10.29”
16	BMJ 019	Dal-boy	Debub-Bench	1399	6°49ʹ51.71”	35°29ʹ10.23”
17	BMJ 020	Shapinsin		1380	6°49ʹ47.39”	35°29ʹ17.20”
18	BMJ 021	Tsidboy		1380	6°49ʹ47.39”	35°29ʹ17.00”
19	BMJ 022a	Tsidboy	Semen-Bench	1855	7°3ʹ58.84”	35°40ʹ2.59”
20	BMJ 022b	Tsidboy		1855	7°3ʹ58.84”	35°40ʹ2.59”
21	BMJ 023	Tsidboy		1810	7°4.104’	35°39.825’
22	BMJ 024	Chabsha		1857	7°4ʹ00.36”	35°40ʹ4.26”
23	BMJ 025a	Bandaboy	Debub-Bench	1383	6°49ʹ46.30”	35°29ʹ07.96”
24	BMJ 025b	Bandaboy		1383	6°49ʹ46.30”	35°29ʹ07.96”
25	BMJ 026	Bandaboy		1400	6°49ʹ56.16”	35°29ʹ15.19”
26	BMJ 027	Tolubab		1400	6°49ʹ56.16”	35°29ʹ15.19”
27	BMJ 028a	Shamut	Debub-Bench	1395	6°49ʹ46.32”	35°29ʹ07.96”
28	BMJ 028b	Shamut		1395	6°49ʹ46.32”	35°29ʹ07.96”
29	BMJ 029	Tsidboy		1407	6°49ʹ57.76”	35°29ʹ13.01”
30	BMJ 030	Tsidboy		1706	7°3ʹ58.93”	35°38ʹ38.15”
31	BMJ 031	Tsidboy		1390	6°49ʹ51.07”	35°29ʹ09.26”
32	BMJ 032	Bandaboy		1390	6°49ʹ51.07”	35°29ʹ09.26”
33	BMJ 033a	Shamut		1402	6°49ʹ54.95”	35°29ʹ12.34”
34	BMJ 033b	Shamut		1402	6°49ʹ54.95”	35°29ʹ12.34”
35	BMJ 034a	Karka-Kachi (W)	Sheko	1241	7°01.173’	35°33.498’
36	BMJ 034b	Karka-Kachi (W)		1241	7°01.173’	35°33.498’
37	BMJ 034 c	Karka-Kachi (W)		1241	7°01.173’	35°33.498’
38	SHK 001a	Tsanseb	Yeki	1293	7°12ʹ42.02”	35°22ʹ24.06”
39	SHK 001b	Tsanseb		1293	7°12ʹ42.02”	35°22ʹ24.06”
40	SHK 002	Tsanseb		1312	7°12ʹ44.45”	35°22ʹ25.06”
41	SHK 003	Gatinseb		1305	7°12ʹ46.05”	35°22ʹ27.51”
42	SHK 004	Gatinseb		1293	7°12ʹ42.02”	35°22ʹ24.06”
43	SHK 007	Konkay		1152	7°8ʹ39.20”	35°24ʹ12.89”
44	SHK 008	Konkay		1202	7°8ʹ40.20”	35°24ʹ12.89”
45	SHK 009a	Tsanseb		1162	7°8ʹ44.88”	35°24ʹ17.00”
46	SHK 009b	Tsanseb		1162	7°8ʹ44.88”	35°24ʹ17.00”
47	SHK 010	Kaibab-kachi		1278	7°12ʹ45.23”	35°22ʹ26.56”

Table 2.

Morphological descriptors

Tuber traits	N	Stem traits	N	Leaf traits	N
Tuber shape	4	Color of the young stem	4	Color of the young leaf	4
Spiny roots on the tuber surface	7	Spines on the young stem	2	Color of the adult leaf	2
Density of spiny roots	9	colored spot at spine base	5	Leaf Shape	5
Tuber color, proximal end	3	Color of adult stem	4	Leaf petiole color	4
Tuber color, middle section	4	Spines at adult stem base	6	Leaf length (cm)
Tuber color, distal end	3	Spines at stem above base	8	Leaf width (cm)
Tuber length (cm)		Spine shape	3	Others
Tuber width (cm)				Flowering ability	3
				Sex of plants	2
				Maturity time (days)

N – Number of scales

2.3. Data analysis

Data on qualitative traits were subjected to multivariate analysis using cluster analysis in Minitab software. Cluster analysis was used to generate a hierarchical dendrogram through an un-weighted pair group method average using complete linkage, Euclidean similarity index.

3. Result and discussion

3.1. Ethnobotanical evidence of domestication

At least three species, i.e., D. alata, D. bulbifera and D. cayenensis complex grow in the study areas. The later is a provisional name for a swarm of species whose botanic relations are currently being examined: D. cayenensis, D. rotundata, D. abyssinica, D. praehensilis, and D. sagittifolia.²⁹ Member species are indigenous to Ethiopia and occur all over Sub-Saharan Africa.³¹ In the study areas, they grow wild, are cultivated beneath a large tree in a home garden, and are farmed in small fields along row of stakes. The D. cayenensis complex is the only yam species known to be subject to domestication. Farmers recognize 49 landraces belonging to this complex, ³⁰ Of these, 27 represented actually grown landraces while the rest reported verbally (Worojie TB, Asfaw BT, Mengesha WA; Diversity, distribution and farmers management of Yam Landraces (Dioscorea spp.) in Southwest Ethiopia, submitted).

3.1.1. The knowledge of wild yam

Sheko and Bench farmers have excellent knowledge of transforming wild-growing yams into cultivated contexts. Knowledge about wild yam is important for its utilization and also to conserve the diversity. It has been realized in this survey that 167 (72%) farmers in southwest Ethiopia have a simple knowledge of wild yam existence in their environment. Significant proportions of the individuals also know the time of wild yam availability (48%) and abundance and distribution (56%) of wild yams in their environment. According to the farmers, wild yams are widely distributed in the open wooded grasslands and in the disturbed forest areas, and the best time for gathering wild yams varies between Septembers to February. The use of wild yam as food is common, and about 30% of the farmers consumed it during their short hunting journeys in the forest (Table 3). The data obtained in this survey is a proof of local knowledge in using and selecting wild yams for domestication. In Benin, researches have reported similar information that supports our view.^5,11,12

Table 3.

Knowledge of wild yams and wild yam domestication in Southwest Ethiopia

Of a total of 231 interviewees, respondents in which those:	Answers in:
	Frequency	%age
Know the existence of wild yam in their environment	167	72
Know abundance and distribution of wild yams	130	56
Know time of wild yam availability	112	48
Use wild yams as food	69	30
Have a simple knowledge of domestication	101	44
Can describe the technique of domestication	49	21
Can describe the distinct morphotypes of wild and domesticated yams	65	28
Practice techniques of domestication	32	14

The knowledge of wild yam domestication is still alive even among farmers who have never recently practicing domestication. About 44% of the farmers know that wild yams can become cultivated yams by domestication. Moreover, about 21% of the individuals can describe the techniques of domestication in spite of stopping this practice a long time ago (Table 3). They have described the techniques as follows. First, they collect tubers of wild yams from forest areas and bring it under cultivation through intensive vegetative multiplication and selection procedures in home gardens. Farmers explained that these procedures induced changes in plant characteristics, mainly in its tuber. Among other studies providing information in this respect are those by¹⁸, who reported the simple knowledge of domestication is shared on average by 82.5% of the farmers in Benin and 62.5% of the farmers in Nigeria. The same source also reported that the proportions of farmers capable of describing the techniques range from 22% to 47% in Benin and from only 10% to 77% in Nigeria.

This survey also indicated that 28% of the respondents are capable of describing the distinct morphological differences between domesticated and non-domesticated yams (Table 3). According to farmers what distinguishes domesticated yams from their progenitors, in general, includes decrease in prickliness of stem and root, loss of dormancy, changes in size and shape of the tuber, loss of bitter properties, and changes in the propagation system. This data would have an immense importance in studying how conscious or unconscious selection process by human has affected yam adaptation under domestication. It becomes apparent that conscious selection may have been responsible for most of the changes. Yet, changes in dormancy and propagation system of domesticates may occur, despite the domesticators consciousness. In view of this Abbo et al. ³¹ reported that domestication is unlikely to have been resulted by the way of unconscious selection rather it is due to a well-focused and highly conscious selection episode.

Based on the domestication syndrome described above, we have identified two distinct morphotypes for domesticated yams across locations. According to some Bench farmers, non-domesticated yam is known by its thicker stem bearing several prickles that are long and generally curved. Its tuber has patches of purple to gray pigmentation near the stem junction. The morphotype of domesticable yam in Bench area is recognized by its intermediate stem diameter bearing prickles with variable length. The tuber has dark grayish flesh and its top is protected by a considerable number of prickly roots. Similar morphological distinctions have also been made by most Sheko farmers. However, some Sheko farmers in the lowland area described additional morphological differences between domesticable and wild yams.

These farmers recognized the non-domesticable yam by its medium-sized stem bearing short prickles. Farmers said it did not exhibit very prickly roots in the tuber crown. The morphotypes of domesticated yams in the Sheko area was recognized by its small stem diameter bearing few and short prickles. Its tuber did not exhibit prickly roots on tuber crown. The tuber flesh color varies from purple to white, is less fibrous, and the top of the tuber crown is protected by very few prickly roots. Cognizant of the observed morphotypes, we suspect that the wild ancestor of domesticated yam belongs either to D. abyssinica or D. praehensilis. In agreement with our results, 6, reported that farmers in northern Benin identify four different morphotypes of domesticable yams and one morphotypes of non-domesticable yams.

The current study also indicates that only few farmers are currently practicing the actual techniques of domestication. This represents only 14% of the 231 farmers interviewed in Sheko and Bench areas (Table 3). The domestication trend however varies with respect to the two ethnic groups. It seems to be more widespread in Sheko compared with Bench ethnic groups. For example, among the 32 experienced domesticators, only 9 (4%) domesticators were from the Bench area. This is less than half of the experienced yam domesticators identified in Sheko area. The few attempts made in Southwest Ethiopia confirmed that this trend is declining.¹⁵ A similar result was obtained by ⁵, in Benin, where only 5% of Nago and Fon ethnic groups practice yam domestication. Similarly, 10, estimated the proportion of domestication is about 3.7% by Bariba group in northern Benin.

3.1.2. The process of domestication

All farmers who are currently practicing domestication were considered as they know the detail process of domestication and in fact most of them are still experienced domesticators. All those interviewed said that tuber is the sole type of original material used during domestication. Most farmers obtained it from forest areas but, nearly 33% of the individuals meet their demand by collecting tubers growing spontaneously in the forest (Table 4). During gathering wild materials, farmers left pieces of tubers in situ for ensuring the survival of wild growing populations.

Table 4.

The process of yam domestication in bench and sheko area, Southwestern Ethiopia

Questions			Responses	Frequency	Percentage
Which type of planting material you use for domestication?			Tuber	32	100
			Seed	0	0
Where do you get the original planting material?			In the Forest	21	67
			Near the village	11	33
How long it take to transform it into cultivated context?			2 y	0	0
			3 y	6	19
			4 y	11	34
			5 y	13	41
			6 y	2	6
Do you know history of transplantation?			Yes	15	47
			No	17	53
If yes	When did you bring it to the home garden?		3 y ago	1	7
			4 y ago	2	13
			5 y ago	4	27
			>10 y ago	5	33
			>15 y ago	3	20
Do you use any obstacle to limit tuber growth?			Yes	0	0
			No	32	100
Why do you domesticate Yam?		For dietary reasons		17	53
		To enrich the existing diversity		9	28
		For curiosity/to sustain native knowledge		6	19

The duration of domestication, in general, can take up to six years but with most of the individuals, it only takes three to five years. During the domestication of wild plants, farmers try to obtain some modifications in the form, size, and taste of tubers. Farmers were often splaying their fingers toward the ground to mimic the small, branching tubers of wild growing yams, while they credit wild transplant yams as having broader tubers. Farmers also report that wild transplant yam retains its wild traits for the first few years after transplantation. It begins to take on the traits of domestic yams after four or five years of cultivation (Table 4). Domestic yams are characterized as broader and tastier. Other studies have described similar distinctions between wild and domestic yams, and parallel changes in the morphology of tubers taken from the wild over the course of five years of cultivation^{5,11,12,15,32}.

Understanding the evolutionary history of wild transplantation may be valuable for inferring how yam diversity was shaped under domestication. In this regard, 47% of the farmers had a known history of adoptive transplantation. Of these, nearly 2(13%) and 4(27%) of them said they had transplanted it about four and five years ago, respectively. While 53% of the respondents were not precise in their response, some said it happened 10 y earlier, while others said it was 15 y earlier (Table 4). Some of the yams are known to be recently transplanted to an open field and grow along rows of stakes; suggesting that at least some of the known domestic yams in southwest Ethiopia are of local origin. Knowledge on crop evolution under domestication and the diversity patterns of their progenitors are highly relevant for crop improvement through efficient utilization of wild germplasms.¹¹

When interviewees were asked to describe the techniques of domestication; all said that no particular technique is used to limit the length of the tubers during domestication. Farmers simply undertook an intensive vegetative multiplication of the wild transplants (Table 4). The wild transplants were planted in the same place for several years. Farmers left it to grow for one or two years before the first harvest. The tubers were repeatedly harvested, but pieces of tubers were left in the soil for subsequent re-growth. This process induces changes in the morphology of tubers. No genetic changes are expected during the process since only vegetative propagation is used. The mechanism underlying the phenotypic modifications is unknown, but they could be resulted from phenotypic plasticity, epigenetic modifications, or somatic mutations.¹⁷ A similar result was obtained by⁵, with yams domesticated by Nago and Fon groups in Benin. However, in Benin,¹⁸, and ³², reported that the Bariba, Yom, Kpakpaza, and Yagbo farmers place some obstacles in the soil mounds in order to restrict the tuber length.

When respondents were asked the reason for domestication, they said that they do it for simple curiosity, for dietary reasons, and in order to enrich the existing crop’s diversity. Among these, those who said that they did it for dietary reasons and for enriching the existing diversity represent 53 and 28%, respectively. But, 19% of the farmers were not precise in their response; some said they do it to maintain what they heard from their parents, whereas others said they do it for simple curiosity (Table 4). Among other studies providing information in this respect are those by Scarcelli et al.^11,12 with the yam cultivated by traditional farmers in Benin. The traditional farmers there enrich their cultivated stock of variation with wild materials. In Mexico⁹, reported that local management of columnar cactus has promoted morphological divergence between wild and managed populations.

3.1.3. Contribution and current trends of domestication

The observed manipulation of wild growing yams by the far Southwest Ethiopians suggests that this practice allows gene flow between the wild and cultivated populations, thereby enhances the genetic diversity of domestic yams. Genes from the wild populations can enter into cultivated populations through wild transplantations. This occurs when the female wild transplants are fertilized with the pollen from the nearby cultivated male parents. Many gardens here contain a few wild transplants, most of which are morphologically very close to some domestic yams. This can be taken as a clear evidence for the integration of wild genotypes into cultivated gene pool. The morphological patterns observed here is similar to those reported by^11,15, for yams cultivated by local farmers in Benin and Ethiopia, respectively. A similar trend of local selection has also been documented for species columnar cacti in Mexico.^9,33

Yam domestication is a constantly repeating process in Southwest Ethiopia. Traditional farmers there maintain high levels of genetic diversity through domestication. However, the majority of the cultivated landraces are undergoing sever genetic erosion and some have almost disappeared; suggesting that the knowledge of domestication is beginning to deteriorate. Several factors that varied across locations are responsible for this. The decline in domestication appears to be high in some areas of Sheko where non-native yams are common. The declining trend of domestication became more conspicuous in most areas of the Bench where pressure on the limited available land is high, and only few forest reserves remained. Earlier studies by Scarcelli et al.^11,12,18, also described similar reasons for the declining trend of domestication in Benin. The latter authors further added that in the Bariba zone of northern Benin, wild yam domestication is considered as a shameful act. In some lowland areas, we observed that it is in danger of being replaced by cassava varieties. Most farmers considered cassava as a non-staked yam and usually known as Enchet-Kachi. In accord to our result,³² reported locally threatened yam varieties in Benin by the introduction of other varieties with different characteristics that are used for the same purpose.

3.2. Morphological evidence of domestication

3.2.1. Relationship between wild and cultivated yams

Evidence derived from the morphological study of 47 accessions corresponding to populations of the D. cayenensis complex has suggested that the process of domestication is still active in some villages; confirming earlier reports^15,34,35. These studies have indicated that D. abyssinica and D. praehensilis are the likely wild progenitors that express clear morphological similarities to domestic yams. These species have been assigned to have different putative roles in the domestication process.

Many of the accessions, although collected under different contexts were morphologically similar and therefore classified together, hence allowing the establishment of close morphological relationships among wild and domestic yams (Table 5). Morphologically diverse individuals of the wild D. praehensilis were encountered in the collection, yet certain traits remain constant for most of the individuals: large sized and green leaves, tubers with very high prickly roots on the crown, very prickly stem, and female flowers. These traits can be also recognized in yams that have been recently transplanted from wild habitats to gardens (Karka-Kachi and Yasind) or within the cultivated forms (Kaibab-Kachi, Tolubab, Torbay, Dal-boy and Shapinsin) (Table 5). Some of the domestic yams used by Sheko and Bench farmers may thus be the possible domestication products of D. praehensilis.

Table 5.

Major morphological characteristics of yam landraces used by Sheko and Bench farmers

Folk names	Major morphological characteristics
Karka-kachi***	Young stem: green with many spines at the base and colored spot. Mature stem: green with very high curved spines at and above the base. Leaf: medium sized, dark green and cordate broad. Petioles: green with brown at both ends. Female flowers. Tubers: cylindrical long, exhibit very high spiny roots on the tuber crown
Karka-kachi**	Young stem: green with many spines at the base without colored spot. Mature stem: green with high to very high curved spines at and above the base. Leaf: medium sized, dark green and cordate broad. Petioles: green with brown at both ends. Female flowers. Tubers: cylindrical long, exhibit high to very high spiny roots on the tuber crown
Yasind**	Young stem: Brownish green, few spines without colored spot. Mature stem: Brownish green, many spines at and above the base. Leaves: large sized, green, sagittate long and broad. Petioles: green with brown at both ends. Female flowers. Tuber cylindrical long, very high spiny roots on the tuber.
Torbay*	Green stems and leaves, few spines at young stem base without colored spot, high to very spines at and above adult stem base, straight spine shape, sagittate long leaves, green petioles, male flowers, cylindrical broad tubers, high to very high spiny roots on tuber surface
Kaibab-kachi*	Green stems and leaves, few spines at young stem base without colored spot, high to very high spines at adult stem base, high spines at adult stem above the base, straight spines, sagittate leaves, green petioles, female flower, cylindrical broad tubers, high to very high spiny roots on tuber.
Chabsha*	Brownish green young stems, few spines at stem base without colored spot, brownish green young leaves, green mature stems, high to very high spines at stem base, high spines at stem above the base, straight spines, dark green and, sagittate broad mature leaves, green petioles with brown at both ends, male flower, irregular tuber, high spiny roots on tuber surface.
Tolubab*	Brownish green young stems, many spines at stem base without colored spot; brownish green young leaves; green adult stems and leaves; high to very high spines at stem base; high spines at stem above the base; straight spines; large sized, cordate broad leaves; green petioles, male flower, cylindrical broad tuber, high to very high spiny roots on tuber surface.
Beri	Green stems and leaves, few spines on young stem without colored spot, high to very high spines at and above adult stem base, straight spine shapes, sagittate long leaves, green petioles, female flowers, cylindrical tubers, high spiny roots on tuber surface
Shapinsin	Green stems, few spines at young stem base without colored spot, light green young leaves, high to very high spines at adult stem base, intermediate spines at adult stem above the base, curved spines, large sized, green and sagittate broad adult leaves, green petioles with brown at both ends, female flower, cylindrical tuber, high to very high spiny roots on tuber surface.
Banda-boy	Green mature stems, few spines at young stem base with colored spot, high to very high spines at adult stem base, high spines at adult stem above the base, straight spines, green and sagittate broad adult leaves, green petioles, female flower, cylindrical broad tuber, intermediate to very high spiny roots on tuber surface.
Dal-boy	Brownish green young leaves and stems, few spines at young stem base without colored spot, purplish green mature stems, high to very high spines at adult stem base, high spines at adult stem above the base, straight spines, large sized dark green and sagittate broad leaves, green petioles with purple at both ends, female flower, cylindrical broad tuber, high to very high spiny roots on tuber.
Konkay	Brownish green young stems, few spines at the base without colored spot, purplish green young leaves and mature stems, few spines at adult stem base, very few to few spines at adult stem above the base, straight spines, dark green and sagittate long leaves, purple petioles with purple at both ends, male flower, oval long tuber, few spiny roots on tuber surface.
Kachi-Tsaa’nseb, Tsano	Purplish green young stems, few spines at stem base without colored spot, brownish green young leaves, purplish green mature stems, dark green leaves, few to intermediate spines at adult stem base, few spines at adult stem above the base, straight spines, sagittate long leaves, purple petioles with purple at both ends, male flower, cylindrical tubers, very few to few spiny roots on tuber.
Tsid-boy	Purplish green young and mature stems and young leaves, many spines at stem base without colored spot, dark green leaves, intermediate to high spines at stem base, intermediate spines at stem above the base, straight spines, sagittate long leaves, green petioles with purple at both ends, male flower, cylindrical tuber, intermediate to very high spiny roots on tuber surface.
Shamut	Purplish green young stems, few spines at stem base with colored spot, brownish green young leaves, purplish green mature stem, dark green leaves, intermediate to high spines at stem base, few to intermediate spines at stem above the base, straight spines, sagittate long leaves, green petioles with purple at both ends, male flower, cylindrical broad tuber, medium to high spiny roots on tuber.
Kachi ga’nseb	Purplish green young stems, many spines at stem base without colored spot, brownish green color of young leaves and mature stems, dark green leaves, few to intermediate spines at stem base, few spines at stem above the base, straight spines, sagittate long leaves, purple petioles with purple at both ends, male flower, cylindrical broad tuber, very few to few spiny roots on tuber surface.
Dizzu-kachi	Green young stems, few spines at stem base without colored spot, brownish green young leaves and mature stem, small sized dark green adult leaves, few spines at stem base, very few spines at stem above the base, sagittate long leaves, green petioles with purple at both ends, male flower, round tuber, very few to few spiny roots on tuber surface.

NB: *** = Wild yams; ** = yams that have been recently transplanted from forest area to home garden and grow beneath a large tree; * = individuals of cultivated yams that are known to be recently transplanted to open field and grow along row of stakes; landraces with non-stars represented populations of longtime cultivars.

The wild D. abyssinica is also the most diverse species in some savanna areas. Individuals of D. abyssinica with various foliar shapes, ranging from the cordate to sagittate type, have been encountered. The prickliness of the stem and root varies widely from slightly armed to very prickly. The morphology of the tuber is also highly diverse. However, certain traits remain constant for most of the observed individuals. They exhibit small-sized green leaf and slightly armed stems with or without a little spot at the base. They did not exhibit very prickly roots in the tuber crown. These characters can also be recognized in some domestic yams. Four varieties, namely, Dizzu-kachi, Kachi-Tsaa’nseb, Konkay and Kachi ga’nseb were morphologically very similar to D. abyssinica (Table 5), and might have been domesticated from this species. Others might be derived from hybridization within cultivated species and between cultivated and wild species. Five cultivated yam types, namely, Banda boy, Shamut, Beri, Tsid-boy, and Chabsha combine the characters of both D. praehensilis and D. abyssinica and might be interspecific hybrids (Table 5). Leaf characters are much more likely to be of the ‘abyssinica’ type while the stem and tuber characters are typically of the ‘praehensilis’ type. The observed relationship suggests that at least some of the domestic yams cultivated in Southwest Ethiopia are of local origin and are the possible products of incipient or recent domestication. A similar relationship between cultivated and wild species has been reported for yam in Ethiopia^15,13 and West Africa. ^{5,11,12,14,36,37}

3.2.2. Relationship between the three yam populations

The cluster analysis of qualitative traits separated the 47 yam accessions into six major clusters (Figure 1, Table 6). Many of them, although recorded under different names were morphologically related and thus classified together. The wild and wild transplant yam accessions are clustered together in the first cluster. Some domestic yams that are known to be recent transplants (accessions 9, 22, 26, and 47) and that the farmers considered as longtime varieties (accessions 8, 15, 16, 17, 23, and 24) have been clustered together with the wild and wild transplant yams (Figure 1, Table 6). This indicates that some cultivated yams may grow in the fallow lands and farmers may collect them when gathering spontaneous plants. Two hypotheses could explain the presence of cultivated populations in the fallow lands. First, these plants could be volunteers derived from some cultivated plants which grow spontaneously in the fallow lands. Some tuber fragments can be left behind in the ground during harvest. Then, they could grow the following year and be harvested by farmers, since they frequently collect spontaneous plants from fallow lands near their cultivated plants.⁵ Second, these plants could be the result of intervarietal hybridization among the numerous varieties that are simultaneously cultivated in the same field.¹⁴ Similar distinctions have been made with yams managed by farmers in Ethiopia,¹⁵ and in Benin.³⁶ The situation we describe here in yam seems to be quite similar to situations reported for potato in the Andes, ³⁸ and species of columnar cacti in Mexico.⁹ The local farmers there collect plants that grow spontaneously in the fallow lands.

Figure 1.

UPGMA dendrogram of yam accessions based on complete linkage, squared Euclidean similarity

Table 6.

Leaf and tuber dimensions and physiological characteristics of the different landraces

Clusters	NA	NL	Accession codes	Local names	Sex	MT	TL	TW	LL	LW	FA
							cm	cm	cm	cm
I	16	9	35, 36, 37	Karka-kachi**	F	L	51.3	18.5	8.82	5.7	H
			1, 2, 4, 5	Karka-kachi*	F	L	49.5	17.3	8.6	6.7	H
			7	Yasind*	F	L	53	15.5	11	7.4	H
			17	Shapinsin	F	L	44.5	15	9.2	7.1	H
			8, 15	Beri	F	L	42.5	9	8.3	6.4	M
			47	Kaibab-kachi	F	L	41	25	8.4	5.8	H
			9	Torbay	M	L	39	21	10.1	6.1	H
			23, 24	Bandaboy	F	L	37	8	8.3	5.6	M
			26	Tolubab	M	L	30	20.5	12.7	10.1	M
II	4	3	3, 6	Karka-kachi**	F	L	46.3	18	8.82	5.7	H
			16	Dal-boy	F	L	40.5	18	10.5	6.1	M
			22	Chabsha	M	L	37	8	7.7	7	M
III	7	3	10, 12,13, 40	Kachi-Tsaa’nseb	M	E	32	12.9	9.4	5.2	P
			11	Dizzu-kachi	M	E	25.5	17	8.8	5.1	P
			43, 44	Konkay	M	E	35.5	12.5	7.5	5.2	P
IV	2	1	25, 32	Bandaboy	F	L	24	12	7.8	5.2	P
V	9	4	14,38, 39, 45, 46	Kachi-Tsaa’nseb	M	E	31.2	16.5	9	5	P
			41,42	Kachi-ga’nseb	M	E	39	20.5	10	5.7	P
			33	Shamut	M	E	41	27	9.2	5.1	P
			21	Tsid-boy	M	E	27.5	13	9	5.5	P
VI	9	2	18,19,20,29,30,31	Tsid-boy	M	E	28.9	17.6	9.5	5.7	P
			27,28,34	Shamut	M	E	41	21.2	9.4	5.3	P

Abbreviations: NA – Number of Accessions; NL – Number of Landraces; F – Female; M – Male; MT – Maturity Time; E – early maturing; L – late maturing; TL – Tuber Length; TW – Tuber Width; LL – Leaf Length; LW – Leaf Width; LA – Leaf Area; FA – Flowering Ability; H – High; M – Medium; P – Poor;

Accessions in the later four clusters represented cultivated yams that have a longtime history of cultivation and propagation by humans (Figure 1, Table 6). These yams have shown a sizable morphological variation when compared with the wild yams. Morphological divergence among the two populations is clear in characters such as tuber size, prickliness, canopy size, and plant cycle. Variation in tuber characters was recognized by most people, and was the principal target of selection favoring abundance of the preferred phenotypes. Plants producing thicker and tastier tubers with fewer prickles are more abundant in domestic populations than in wild and wild transplant populations. Our analysis also reveals that plants from domestic populations are earlier maturing than those from wild and wild transplant parents (Table 6). A decrease in flowering ability is more prevalent in domestic populations than in wild populations. Domestic yams are propagated vegetatively, while the wild populations are mainly propagated by seeds. Two hypotheses could explain these divergences. First, it may be resulted from the differential seed diffusion systems of farmers for attractive phenotypes. Second, domestication involves the feral or possibly wild types produced by sexual reproduction. Therefore, through the domestication practice, sexual reproduction contributes to the evolutionary dynamics of yam. Other studies have reported similar morphological divergence between wild and domestic yams.^5,11,17

4. Conclusion

The domestication of ³⁰yam is an active process in Sheko and Bench provinces. The knowledge of feral and wild yams is still alive even among farmers who have never domesticated yams. These farmers are also capable of describing the distinct morphological differences between wild and domestic yams. Several wild yams have been brought under cultivation by Sheko and Bench farmers, but this trend is tending to decline. Yet, the observed manipulation of wild yam by the far Southwest Ethiopians has shown an interesting implication about the history of its cultivation and utilization by humans in prehistoric times.

Ethnobotanical and morphological evidence allows for the recognition of two divergent evolutionary processes. First, a population that is positioned medially along the domestication gradient and express elements of domestication syndrome belonging to D. abyssinica or D. praehensilis. Second, populations that might be derived from cross-fertilization between the two wild species or between wild and domestic types. These populations could also be volunteers that escaped into wild in the past involuntarily. The two divergent processes could reflect the selection effort of farmers in the gradient of management intensity. From this study, it emerges that the plants which were used by the local farmers as starting material in the domestication process were composed of diverse types of hybrids, escapes of known landraces, and possibly wild plants. No genetic changes are expected during the domestication process since only vegetative propagation is used. As a result, the term “domestication” here could be replaced by “adoptive transplantation”.

Given the fact that the diversity pattern of Dioscorea species agrees with the Vavilovian view, we can speculate Ethiopia as an independent center of yam domestication. But, different yam species might have been domesticated independently in Africa or domestication may have occurred simultaneously in each areas. No sound generalizations can thus be drawn based on this inquiry and should be supported with molecular and archeological evidences to reveal the evolutionary dynamics of yam. If it is supported by such studies, Ethiopia may represent one of the main centers in Africa where yam diversity is noticeably high and some of the yam species could become domesticated.

Funding Statement

This study was funded by Ministry of Science and Higher Education, Ethiopia. The funding body has no role in the design of the study, analysis and interpretation of data and in writing the manuscript;Ministry of Education, Ethiopia [around 400 USD];

Competing interests

The authors declare that they have no competing interests.

Availability of data and materials

The ethnobotanical data gathered during the current study are included in the manuscript. Other supplementary materials such as photographs and field notes of varieties are available by request to the authors.

Author contributions

The field survey, interpretation, and analysis of the data and writing the original draft were made by the corresponding author; while the co-authors contribute in conceptualization, review, and editing the paper. All authors have read and approved the final version of the manuscript.