Author contributions

A. Zemede proposed, designed, and conducted the trial, analyzed data, drafted and finalized the manuscript. F. Mekbib, K. Assefa and Z. Bishaw made valuable contributions to and commented on the final manuscript.

Conflict of interest statement

The authors declare that they have no competing interest.

Introduction

Ethiopia is the largest wheat producer in Sub-Saharan Africa with about 1.6 million ha of durum and bread wheat (Central Statistical Agency, 2017). Wheat is one of the major cereal crops in the Ethiopian highlands, which range between 6 and 16 N latitude, 35 and 42 E longitudes, and from 1500 m to 2800 m above sea level (masl). Ethiopia has already been identified as an important center of diversity for different crops. Durum wheat is a tetraploid wheat variety traditionally grown on heavy black clay soils (Vertisols) of the Ethiopian highlands between 1800-2700 masl, where it is mainly produced by small scale farmers. Durum wheat is suitable for manufacturing pasta products (e.g. macaroni, spaghetti); however, in Ethiopia it is also used for making “injera”, a white leavened Ethiopian bread made from grain flour, and other local foods. Durum wheat grown in Ethiopia is constituted by a few improved varieties and a large number of traditional farmers varieties, commonly referred to as landraces (Bechere, Kebede, & Belay, 2000; Bishaw, Struik, & Gastel, 2014; Eticha, Belay, & Bekele, 2006). Durum wheat is extensively cultivated and grown by a large number of farmers and a significant number of traditional farmers’ varieties have been recorded in the central highlands of Ethiopia in the past 20 to 30 years (Negatu, Mwangi, & Tesema, 1992). Statistics on durum wheat production in Ethiopia from the past decades are difficult to obtain, they are usually lumped with bread wheat in reporting. The total area under wheat production in 1983was reported to be 625,590 ha of which 60-70% was estimated to be used forthe production of durum (Tesema, 1991). The following decades saw a drastic reduction in area coverage for durum wheat which dropped to around 50% of total wheat production area (Gashawbeza et al., 2003). Recent reports indicated that the area coverage for durum has further declined to not more than 15-20% from a total 1.6 million ha of lands (Mengistu, Kidane, Fadda, & Pè, 2016).

Introduction of modern varieties, improved production practices and problems associated with biotic stresses like crop pests and diseases as well as abiotic stresses such as droughts in traditional farming systems were among the major problems faced by the farmers leading to gradual replacement of traditional farmers’ varieties. Genetic erosion is defined as “the loss of genetic diversity, in a particular location and over a particular period of time, including the loss of individual genes and the loss of particular combinations of genes such as those manifested in landraces or varieties” (FAO/IPGRI, 2002). It could be considered as a consequence of the loss of diversity in the farmer’s field and will have a major effect on the future crop productions and productivity for indigenous crop species of Ethiopia. The problem has been increasing and is expected to be more aggravated for durum wheat than other cereal crops.

Several research reports from survey studies and farmers’ discussion indicated loss of diversity on tetraploid wheat has increased faster than expected in different environments of Ethiopia. The reduction in number of farmers, traditional farmer varieties or landraces and area coverage confirmed the danger and extent of genetic erosion on tetraploid wheat, as reported in multiple studies. Teklu and Hammer (2006) reported genetic erosion of 88%, 100% and 78% for the durum wheat species Triticum durum, T. turgidum and T. dicoccon, respectively. In Harar Zuria, the same authors detected genetic erosion of 88.9 % for T. durum and 100% for both T. turgidum and T. dicoccon. In the western Shoa zone of Ethiopia, genetic erosion reached 75% and 62% for tetraploid wheat species in Ambo and Dandi districts, respectively (Geleta & Gausgruber, 2013). In a similar study farmers identified 26 tetraploid wheat land races (21 from Akaki and 17 from Ejere), which were once widely grown in the area and of which only six were currently available; the average loss of diversity was thus estimated to be 77% (Tsegaye & Berg, 2007). Besides primary sources collected from a survey using semi-structured questionnaires and farmers’ group discussions, this study also used direct field assessment to identify recently cultivated landraces and compare them to previously available landraces in the similar districts, calculating levels of genetic integrity and erosion based on the procedures of Hammer, Knüpffer, Xhuveli, and Perrino (1996). The reduction in numbers of farmers growing landraces and area of cultivation were identified as useful indicators for the existence of genetic erosion (Brown & Hodgkin, 2015).

The objectives of this present study were to calculate and estimate on farm genetic erosion and identify associated causes in three districts of central Ethiopia using different approaches. Results obtained from direct field assessments and observations as well as information collected from farmers’ interviews and discussions were used to investigate the current status of durum landraces and to determine trends based on the number of farmers using landraces and the area of cultivation of the landraces.

Materials and Methods

Study areas

The study was conducted in three districts of the central highlands of Ethiopia based on direct observation, a semi-structured survey questionnaire and group discussions. Direct observation was conducted across three different routes which covered Gimbichu, Ada, and Lume districts during the 2017 cropping season (Figure 1). A survey questionnaire and group discussions were conducted in Ada, Lume and Gimbichu districts of the central highlands of Ethiopia in 2018. The three districts were purposely selected to represent the durum wheat agro-ecology environment of the East Shoa Zone: 1) Ada represents a midland wheat growing agro-ecology area,where modern crop varieties are assumed to be dominantly cultivated; 2) Lume represents a midland area where both landraces and modern wheat cultivars are believed to exist due to interventions by the Ethiopian Biodiversity Institute to recover the status of local in situ conservation; and 3) Gimbichu district represents wheat growing agro-ecology at high elevation and is an area where durum landraces are still extensively cultivated.

Results

Direct observation

Genetic erosion occurred on tetraploid wheat after 25 years based on direct farm assessment (Table 1). The variation was significant on loss of diversity on tetraploid wheat at 0.01%. The average genetic erosion across districts was 75.8%, which is lower compared to previous reports (Geleta et al., 2013). The loss of diversity on tetraploid wheat was found to be the highest in Ada (87.5%) and lowest in Gimbichu (60%) based on direct field assessment (Table 1).

Table 1: Genetic erosion on tetraploid wheat since 1992 in three districts of the central highlands of Ethiopia

Districts	landraces cultivated in 1992	landraces cultivated in 2017	Genetic Integrity (%)	Estimated genetic erosion (%)
Gimbichu	15	6	40	60
Lume	15	2	20	80
Ada	16	2	12.5	87.5

The proportion of landraces compared to modern wheat varieties was lower in lower altitude than high altitude growing environments (Figure 2). In 2017 the proportion of landraces in altitudes above 2400m were 70% while below 2400m this dropped to about 30% compared to modern wheat.

Number of farmers and area of cultivation

Modern bread and durum wheat as well as tetraploid landraces area coverage were estimated based on random field visits and identification of wheat species found in the field and calculated as percentage of each wheat species area to the total wheat area in a given district.

In 2017 only 8.3% of farmers in Lume, about 23.5% in Ada and 40.7% in Gimbichu were growing local varieties. When the three districts were considered, only 24.2% of farmers were growing local durum wheat varieties. On average 24.5% of wheat growing area was occupied by landraces across the three districts compared to modern wheat varieties. The highest area coverage was in Gimbichu (52.3%) while the lowest was 2.5% in Lume (Figure 3).

Farmers’ survey

In order to obtain a more pronounced picture of the estimated loss of genetic diversity in central Ethiopia, semi-structured interviews were conducted with 149 farmers from three communities in the three districts. Based on interviews with farmers, significant genetic erosion was observed in tetraploid wheat after 25 years (Table 2). The loss of diversity in tetraploid wheat reached up to 100% in Ada and was lowest in Gimbichu (67.7%), with an average of 87%, based on the survey.

Table 2: Genetic erosion of tetraploid wheat since 1992 in three districts of the central highlands of Ethiopia based on semi-structured interviews with 149 farmers.

Districts	Durum wheat varieties cultivated in 1992	Durum wheat varieties cultivated in 2017	Genetic Integrity (%)	Estimated genetic erosion (%)
Gimbichu	15	5	33.3	67.7
Lume	15	1	6.7	93.3
Ada	16	-	-	100.0

In the 2018 growing season only 7.3% of farmers in the Lume district but 68.4% in Gimbichu were maintaining landraces of wheat. None of the farmers interviewed in Ada district were growing wheat landrace varieties. A similar trend was observed in cultivation area of landraces compared to modern varieties. The sample area coverage was 13.5% in Gimbichu and only 1.5% in Lume. No farms were cultivating landraces in Ada district (Figure 4).

Discussion

The results indicated that on average 75.8% and 87% genetic erosion was observed on tetraploid wheat after 25 years based on direct farm assessment and survey study, respectively. The levels of loss observed from both approaches appear to be similar to that reported in Western Ethiopia (Geleta et al., 2013). The loss observed was higher in Ada and Lume, situated at lower altitude, than in the high altitude zone above 2400 masl of Gimbichu district, where significant areas of land are still used to grow landraces. This difference can be attributed to the fact that both Ada and Lume districts have similar agro-ecology, access to big market centers and are also near to the Debre-Zeit Agricultural Research Center where a lot of improved wheat cultivars are developed, demonstrated and distributed. Modern wheat and tef crops are more expanded (Assefa et al., 2015) and could also be major causes for loss of diversity in the Ada and Lume districts. The presence of higher diversity in Gimbichu district might be associated to the merits of landrace adaptation and yield stability (Berg, 2009). Similarly, (Tsegaye et al., 2007) reported 77% genetic erosion and relative diversity for tetraploid wheat in Ejere, a high elevation district of central Ethiopia.

Our study demonstrated a correlation between the extent of genetic erosion and altitude. This indicates the relative importance of focusing further landrace collection and conservation efforts on altitudes above 2400 masl to capture more diversity. The study further confirmed the existence of better occupation in terms of area coverage and maintenance of landraces by large numbers of farmers in Gimbichu as compared to Ada and Lume. This could be due to the expansion of modern wheat varieties at the expense of the landraces in the latter districts as illustrated in Figure 3 and Figure 4 and confirmed by farmers’ group discussions. (Negatu et al., 1992) reported that 64% of sample farmers grew local durum wheat varieties in the same districts of central Ethiopia. The low percentage of farmers currently growing local varieties in the three districts are likely the result of commendable expansion of well performing modern varieties.

The results found in the current survey and previous research reports were comparable to those obtained in direct farm assessment. The different approaches followed may account for the large variations observed on the extent of genetic erosion although the trends were similar between the altitudes. The findings of the study have a significant impact for in situ conservation of landraces for future use in breeding to combat the challenges of climate change. Farmers’ opinions gathered in group discussions indicated that drought contributed to high loss of landraces in Ada and Lume and the result contradicts the existing theory that landraces are well adapted and coevolved in the environment they originated.

The results contribute to the setting of targeted strategies in high altitude zones for in situ conservation and exploitation of landraces in future breeding programs. Future studies should take into account direct observations as alternative approaches to assess and estimate genetic erosion of tetraploid wheat and other indigenous crop species.