Socioeconomic characteristics of respondent households

A sample of respondents on socioeconomic characteristics is described in Table 2. Among the respondents, 82.1% of families were male-headed households, while only 17.9% were female-headed households. About 50.4% of the heads of households were between the ages of 45 and 65, while 25.4% of the respondents were over 65. Approximately 41% of respondents were illiterate, whereas 28.8% had informal education and could read and write. However, 51% of the respondents overall in the studied administrative zones who participated in the interview were female. They are knowledgeable enset cultivators who have a great deal of knowledge about planting, managing in the field, harvesting, and using enset products. They rely on enset products for most of their food needs, medical requirements, needs for fodder, and environments. They also gain benefits from the rich ecosystem of goods and services created by the enset agrosystem.

Table 2 Socioeconomic characteristics of respondent householdsExtent of richness and diversity of enset landraces

In this study, we identified and recorded 282 locally named enset landraces in the Hadiya, Kembata-Tembaro, Gurage, and Silte zones of southern Ethiopia. Enset growers can easily distinguish one enset landrace from the other by observing the external (leaf structure, size, orientation, midrib color, and other) and internal features (leaf and midrib anatomy and fiber structure) of the enset plants, and they give distinct vernacular names for each landrace. Each local farmer in the studied area was observed cultivating a diverse of enset landraces in his or her homegarden, which shows a considerable variation in the number of enset landraces on individual homegardens. It ranges from two to thirty-two in this study (Table 2). According to farmers’ knowledge of local names: 86 enset landraces from Hadiya, 73 from Kembata-Tembaro, 66 from Gurage, and 57 from Silte were recorded. The highest and lowest number of landraces per homegarden was documented in Hadiya and Silte zones, respectively (Table 3). In comparison with earlier reports, a relatively larger number of landraces have been identified in this study. The literature shows that Tsegaye [17] recorded 146 different enset landraces including 59 from Hadiya, 55 from Wolaita, and 52 from Sidama while Negash [16] reported the same total number including 65 from Kefa-Sheka, 30 from Sidama, 45 from Hadiya and 6 from Wolaita. Likewise, Birmeta [18] described 111 enset landraces from nine enset-growing localities of Ethiopia that contrasted with the findings of the present study as in some other previous studies. For instance, Yemataw et al. [13] and Zeberga et al. [19] described the same numbers of (312) different enset landraces from eight ethnic groups, out of these 69 from Silte, 66 from Kembata-Tembaro, 63 from Gurage, and 51 from Hadiya. Furthermore, Yemataw et al. [24], who described 218 different enset landraces from seven zones, came up with 59 landraces from Hadiya, 43 from Kembata, 41 from Dawuro, 39 from Wolaita, 34 from Gamo Goffa, 31 from Gurage, and 30 from Sidama. Some of these values are slightly comparable to the findings of the present study but such records are impossible to make a direct comparison of the number of enset landrace diversity with the results of the current study due to variations in the method and size of the sampling area. However, in most cases, the richness of enset landraces recorded in the current study is far higher than the reports of the previous studies which is likely to be related to the rigor and intensity (including the sampling frame) as well as the knowledge of the men and women informants that participated in the present study. The number of enset landraces in the present study could be attributed to the technique of sampling, the area the study covered, and the nature of the agroecological condition of the study area, which embraces midland and highland that is suitable for enset cultivation. Moreover, the study zones like Hadiya are bordered by all the other study zones, so the exchange and earning of suckers are common traditions among farmers. In the same manner, Tsegaye [17] and Yemataw et al. [24] stated that the exchange of enset landraces from the neighboring ethnic groups perhaps contributed to the richness of enset landrace diversity in Ethiopia.

Table 3 Enset landrace diversity in the four administrative zones, richness, Simpson (1-D), Shannon (H') diversity indices, and evenness (E)

The Shannon diversity index (H′) ranged from 3.73 (Silte) to 3.96 (Hadiya), this signifies the existence of a high richness of enset landraces in the study zones. Even though zones varied in richness, they revealed a very narrow range of variances in Simpson’s 1-D and evenness indices. The Simpson’s 1-D ranged from 0.963 (Silte) to 0.978 (Hadiya) and evenness indices ranged from 0.89 to 0.92. All these results specify the presence of high enset landraces diversity in these four zones (Table 3). This finding is in line with earlier reports [13, 19]. According to [39], the value of a diversity index increases when both richness and evenness increase and is maximized when all species are nearly equally abundant. In biodiversity studies, Shannon diversity indices (H′) typical values range between 1.5 and 3.5 and the index is rarely greater than 4 [40]. The higher the value of H′, the more diverse the communities, and the Shannon index increases as both the richness and evenness of the communities increase.

Similarities and differences of enset landraces diversity among zones

The similarity among pairs of zones (taking two zones at a time) concerning farmers-named landraces was evaluated using Sorenson’s similarity index (Table 4). Generally, the similarity index ranged from 0.24 to 0.73, and the number of commonly shared landraces varied from 16 to 47. Hadiya and Kembata-Tembaro were the most similar zones, followed by Gurage and Silte about enset landraces (Table 4). Hadiya also shared 38 and 35 enset landraces with Gurage and Silte zones, respectively. This high sharing of enset landraces among zones may be due to sociocultural and linguistic similarities, and geographical locations. For instance, Hadiya is bordered by all study zones, so the informal exchange of enset suckers from the adjacent zones possibly contributed to the highest similarity of enset landrace diversity among zones in the present study. This agrees with the work of [19] and [24], who reported the existence of a high amount of sharing similar enset landraces among Hadiya and Kembata, Gurage and Silte, and Wolaita and Dawuro zones of Ethiopia. On the other hand, pairs of zones with relatively least similarity were Kembata-Tembaro and Silte, and Gurage and Kembata-Tembaro 0.24 and 0.25 for each pair, respectively. This may be due to the geographical distance between the two zones and also variations in sociocultural factors.

Table 4 Enset landraces shared (bold) and Sorensen similarity indices between pairs of zonesDistribution and abundance of enset landraces

Distribution of the enset landraces throughout the study sites varied across zones. Out of 282 enset landraces recorded, 15 (5.3%) were widely distributed in all four zones. These were Agade, Astara, Bededete/Badade, Gimbo/Gimbuwa, Heniwa/Hiniba/Enba, Kasete, Manduluqa/Mande, Mariye, Merza, Mesmesia, Moche, Separa/Sebera, Torora/Xorore, Weshemeja and Zobira (Table 4). Similarly, 33 (11.7%) farmers’ named enset landraces were commonly cultivated and found in three (Hadiya, Gurage, and Silte) out of four zones (Table 4). Likewise, 72 (25.5%) of the enset landraces had a narrow distribution and were specific to a single zone (Table 4). But the remaining 210 (74.5%) were recorded in more than one administrative zone. The finding of this study was in line with the previous study of [17, 19], and [24] from the same or different zones in Ethiopia.

The abundance of enset landraces also differed among the study zones in addition to their distribution. Few enset landraces such as Gimbo, Hiniba, and Separa were relatively high in abundance at all four study zones. Agade, Bedededa, and Zobira were also other most frequent enset landraces in three out of the four zones (Table 5). Some landraces were well encountered in two zones but virtually absent from the other study zones. For example, Sisqella and Gishira were the most abundant landraces of the enset homegardens visited in Hadiya and Kembata-Tembaro zones but were almost absent or rare in other zones. Moreover, some landraces such as Abatemerza, Degomerza, Dirbo, and Unjame in Kembata-Tembaro, Amerate and Lemat in Gurage, Shewrad in Silte, Disho, and Bequcho in Hadiya zones were dominant but outside these zones, they were found with a low abundance. A similar observation was reported by [13] and [19], they indicated that landrace Agade in Silte, Amerate in Gurage, Shododenia in Dawuro, and Addo and Genticha in Sidama encountered a high local abundance at each studied zones. This may be due to the environmental adaptability of the landraces or/and different attributes of farmers. Negash [16] and Tsegaye [17] also reported that enset landrace diversity and distribution were influenced by factors such as household resources, cultural background, population pressure, and agroecology. Enset landraces, namely Manduluqa, Mariye, Mesmesia, Moche, and Torora, described in this study were found in a limited number of homegardens but widely spread in each zone. In the same manner, [13] and [19] indicated that household features, the distance between locations, and ethnic preference contribute to the landrace diversity and abundance.

Table 5 List of farmers-named landraces and their richness in the four administrative zonesDiverse local names of the enset landraces among zones

The local names of enset (Ensete ventricosum) and its different growth stages vary from one ethnic group to another. Enset is called wessa in Hadiya and Kembata-Tembaro, wesse in Silte, and aset in Gurage. Moreover, each growth (transplanting) stage has a distinct name by which it is identified. The Hadiya and Kembata-Tembaro farmers share almost the same local names for all sucker stages. These are known as dubbo, simma, ero/kiniba, and balwesa, but in Silte 1 -and 2-year-old suckers are called bosho and daporo, respectively, and the other two stages are nearly similar to the Hadiya and Kembata-Tembaro zones (Fig. 2a–d). In Gurage, 1-year-old sucker is fonfo, but the second and third stages are named the same as other studied zones.

Fig. 2

Examples of the main enset propagation stages and its end products: a mother corm; b multiple suckers from mother corm; c young enset ready to give mother corm and transplanted to permanent field; d matured enset; e mass of processed product; f and g example of dishes that prepared from the primary products

According to the interviewed farmers, the same enset landraces are sometimes known by different names in different administrative zones (Table 6). In this study, 11 farmer-named landraces identified with the help of knowledgeable farmers in each zone indicated that the same enset landraces were known by different names in the other studied zones (Table 6). The role of knowledgeable men and women enset farmers was so critical in this research since they are experts in the landrace identification and description of ethnobotanical methodology. The landrace names given by enset farmers mostly reveal distinct morphological appearances or other culinary characteristics such as taste or use values (data not shown). Each ethnic group has its series of local names for enset landraces. For example, the landrace Shate in Hadiya, and Shirteye in silte and Gurage are the same landrace with different local names often representing the bitter-tasting characteristics of all its parts. Enset landrace Xiggo in Hadiya, called Qeqile-nech in Kembata-Tembaro, is well known to the enset farmers as its bleeding (red liquid) when parts are cut. The origin of certainly cultivated enset is evident from the name. One such example in this study is Kembat which may be originated from Kembata; however, its name in Kembata-Tembaro and Hadiya is called Disho (Table 6).

Table 6 Different local (vernacular) names for the same enset plants within or among zones

In addition, according to farmers, some landraces were named based on the color of pseudostem and leaf (Bushawese in Silte meaning red enset), but this landrace in Hadiya is given the name Meqelwesa, meaning placental enset, which is related to use characters. Similarly, landraces Soqido is salt (taste of boiled corm or amicho) in Hadiya while in Silte and Gurage it is Kemele meaning Ape (maybe the color of the pseudostem or petiole) (Table 6). In general, this is observed due to the use of various local names in the different communities of the study area, having their specific characters and method of perceiving by the local farmers. Based on key informants' responses and focus group discussion, some cultivated enset landraces were named with minor or slight dialect differences forms in the local names among study zones. Those include landraces: Gimbo/Gimbuwa, Hiniba/Heniwa/Enba, Jegirada/Z’girad, Hyro/Ahiro, Qibnare/Qinare/Qiniwara, and Guary/Gariya. This reveals that sometimes the same landraces are often known by different names in different or the same regions. The method of the naming of landraces as indicated by farmers in our study is also similar to what has been reported in other enset-growing zones. For instance, [12, 15], and [22] reported that the naming criteria of some enset landraces in the Wolaita, Sidama, and Ari respectively, are mostly based on morphological and agronomic traits, place of origin, various uses, and culinary attributes. In the study areas, farmers use their local language in everyday speech and communication in each zone. There are numerous enset landrace names and synonyms in these different languages and dialects were recorded throughout the study zones (Table 5). For instance, in this study 15 identically named enset landraces were identified from all four studied zones. In the same manner, three zones (Hadiya, Silte, and Gurage) commonly share 33 of the same named enset landraces in the present study. A similar observation was notified by [12, 13], and [15], they also described the existence of identically named enset landraces in more than one ethnolinguistic community. This may occur due to getting the enset planting materials and a long-lasting practice of farmers in sharing with their respective landrace names from adjacent administrative zones. Similarly, [12] stated the presence of 'borrowed' landrace names between ethnolinguistic groups. Similar trends were also observed in different traditional crops such as sorghum [11], banana [26], sweet potato [27], cassava [41], and common bean [42, 43]. Our study has also shown that enset growers sometimes delivered various names for the same landrace within the zones. For instance, the landrace named Ayase is known as Hella in Hadiya Duna woreda, Qombotira is called Asheqit in Silte, and also Gegered is known as Heniwa in Endegegn woreda of Gurage zone. Bareke et al. [42] and Abera et al. [43] also reported similar results from Ethiopia for common beans. Likewise, the different names for the same enset landrace also exist among zones (e.g., Disho in Hadiya and Kembata-Tembaro is known as Kembat in Silte and Gurage, Tem-wese in Silte is also called Xebere in Kembata-Tembaro or Qebere in Hadiya) (Table 6). Moreover, the names of some enset landraces have the same meaning but it was locally known with different folk names throughout study zones. For instance, Xiggo in the Hadiya, Dem-worad in the Silte and Demyetertqey in the Gurage refer to bleeding because of exuding red fluid when any part of the enset is cut. This is similar to the findings of [43] who found that common bean producers provided different names in terms of seed color in two areas but the names have the same meaning.

Pattern of use and management practices undertaken by farmers

Traditionally, farmers in the study area were familiar with the utilization and management of enset from earlier generations to meet their food, drug, and other requirements. In the study area, all enset landraces were primarily cultivated for food and feed use, except landrace Meqelwesa or Qeqile-qey which was rarely used as food. This landrace is one of the most traditionally preferred medicinal enset landraces recommended for human and cattle ailments (Table 7). Based on the information we acquired during the individual interview and focus group discussion, enset farmers preferred landraces with early maturity and vigorous growth, easily harvestable, early fermenting, high qocho and bulla yielding, and good cooking qualities. In addition, in all four zones, generally, multi-use enset landraces were highly chosen and more cultivated than specific-use landraces. However, in some situations, there was regional or ethnic preference across the study zones.

According to a result of the key informants ranking from the five commonly shared and the other two, Gimbo became the first, Separa the second, and Agade the third most preferred enset landraces for their qocho and bulla quality; Astara and Agade scored the highest points for both their amicho (cooked corm) tasty and medicinal value, and Sisqella, Bededede, and Gimbo stood first to third, respectively for their fiber quality (Table 8). For instance, extracting bulla from other harvested masses of enset (Fig. 2e) in Gumer woreda of Gurage zone by women is not common practice, unlike other woredas and zones. But they purchase it from other adjacent woreda markets for different purposes. In the same pattern, the use and production of fiber, which is another enset product obtained from the decorticating of petiole and pseudostem are decreasing in most of the studied zones. Because it employs a traditional production method that requires more time and labor. In addition, nowadays most of the traditional fiber-made products are replaced by other plastic materials. However, some enset farmers in Hadiya and Kembata-Tembaro preferred more droughts tolerant and high fiber quantity and quality (Table 9) in addition to qocho and bulla yield, while those in Gurage and Silte favored easy harvesting and processing, early fermenting, and less fibrous landraces (Table 10). The present study also indicated that there were slight differences in terms of perceiving enset end-users across the study zones.

Table 7 Enset landraces selected for medicinal purposesTable 8 Preference in direct matrix ranking of five commonly shared and other two localized enset landraces cultivated in southern EthiopiaTable 9 Enset landraces selected for strong and long fiberTable 10 Enset landraces selected by farmers for amicho

Moreover, interviewed farmers in Kembata-Tembaro grouped enset landraces into two major sex categories: female enset and male enset. The division of male and female is not linked to biological reproduction but it is based on perceived features of the landraces. The female groups are known for ease of decorticating, early fermentation, corm palatability, more susceptibility to different diseases, and low strength of fiber whereas the male groups contrast to these characteristics. In contrast, farmers in Hadiya, Gurage, and Silte did not tend to classify enset plants into sex designation. Tsegaye [17] also reported the relationship to the difference in food culture, sociocultural preferences for different enset products, and farming systems of the regions. Similarly, [44] described the influence of cultural background on plant species diversity and the uses of plant species for different purposes. Enset landrace diversity within the same and different cultural groups nicely demonstrates that cultural needs and requirements are key factors in the diversification of crop varieties. In particular, the unique landraces recorded in the different ethnic communities indicate the origin and maintenance of those landraces by specific ethnic groups because they need them for their food, medicine, and other uses.

According to the farmers' report, we identified a total of 32 landraces which were applied in different proportions by each ethnic group: 10 in Hadiya, 9 in Silte, 7 in Kembata-Tembaro, and 6 enset landraces in Gurage as traditionally medicinal use to treat various health problems in human and cattle (Table 7). Out of the total listed, 12 medicinally used enset landraces shared the identical name in at least two zones, so the total number decreased to 20. Landrace like Astara mentioned by the farmers is an example of enset that has multiple uses of traditional medicinal purposes in the all study area. Furthermore, landraces such as Qinare/Qiniwara, Gishira, Guary, Xessa, Hayiwona, and Agade were also the most frequently used medicinal enset present in homegardens of two or more ethnic communities (Table 7). On the other hand, some medicinal landraces (Cherquwa and Wolegella) were identified as having narrow distribution in the study zones. However, in some cases the same kinds of enset are known with alternative local names used as medicines for different problems among the study communities (Table 7).

For instance, landrace Xiggo in Hadiya is mainly traditionally used to treat kidney and liver problems, whereas the same variety with different names (Qeqile-nech in Kembata and Dem-worad in Silte) quoted by many farmers to remove the delayed placenta and for aborification (used to cause/facilitate abortion) purposes (Table 7).

This may be due to each ethnic community having its ways, practice, and beliefs to utilize enset plants. All of the traditionally medicinal enset landraces were also selected for sweet amicho (cooked corm) production except landraces Gishira, Dare, and Bedededa. In the same manner, the most chosen part of enset for medicinal use was corm but the landrace Meqelwesa (in Hadiya) or Qeqile-qey (in Kembata) were all part used as traditional medicine. In some cases, farmers also used cooked qocho or porridge prepared from bulla to treat different health problems in the study zones (Fig. 2f and g). In terms of connection to the ailments shared by the farmers and the medicinal enset landraces used in their treatment, we observed that bone fracture, swelling of the pus and to expel the delayed placenta from humans and cattle were the most shared health problems in the study area and among the communities. We observed that some ethnic groups (e.g., Silte and Gurage, Hadiya and Kembata-Tembaro) share more medicinal enset landraces and show greater similarity in patterns of using enset crop (Table 7). In the same manner, [44] stated that intercultural sharing may be explained by the pharmacological effectiveness of shared medicinal plants among ethnic groups.

Most enset-growing farmers in the study area are familiar with maintaining and use of their different preferred landraces to stabilize many situations over a long period without external support and inputs of planting materials. Farmers in the study zones frequently produce their planting materials or suckers from homegardens but few farmers obtain them freely from neighbors, family, and friends as a gift or by purchasing from other farmers. This was in line with the reports of [7, 13].

During our discussions with farmers and field observation, we observed that in two local markets: Alicho in Silte and Gumer in Gurage zones, enset suckers were purchased from January to April. These two sites are situated at a higher altitude than other studied woredas (Table 1). Moreover, some elder farmers mentioned that enset cultivation practice and its distribution into their woreda and villages relatively late than others. They said that “We haven’t been familiar with enset production and managing before 65 years ago.” To some extent, this verifies that enset farming systems in the studied area are not equally and uniformly experienced within and among communities. Enset cultivation and use culture has been gradually and slowly moving to the peripheral areas from region to region, from zone to zone, and from district to district due to farmer “experimentation” and horizontal transfer of indigenous knowledge. In a similar vein, [2, 5], and [18] indicated that the distribution of cultivated enset in Ethiopia appears to be expanding, especially after periods of devastating famines of the 1980s, when people in other regions learn about the benefits of this crop and attempt to incorporate it into their farming system. They have also shown that enset moves some inches into the Oromia region and this is observed in southwest Shewa and southeast Arsi and may be in the western part of Bale. Negash [16] also noted that during the drought period, many farmers migrated from their villages to as far away in search of food, and there they learned about enset production. When they came back to their homesteads, they introduced enset. Furthermore, [45] and [46] show that smallholder farmers expand the production area of the perennial crop enset as a climate coping strategy in a drought-prone indigenous agrisystem.

Farmers in the studied area maintain great enset landraces diversity within traditional cultivation and production systems insight toward meeting domestic subsistence requirements. Yemataw et al. [13] also described that farmers observe and select the landraces based on their planting intentions for the coming year than the proportion to the quantity they have. The on-farm maintenance of biodiversity requires understanding by the farmer of how specific varieties should be grown, stored, and maintained to maximally realize the characteristics these farmers value [47].