Introduction

Teff (Eragrostis tef Zucc., 2n = 4x = 40) is a C4 cereal crop that has been cultivated in the Horn of Africa since millennia (Harlan 1969). In Ethiopia, teff is a staple crop for about 70 million people (Assefa et al. 2011), as well as a source of feed (Yami 2013) and a cash crop generating incomes for about $0.5 B per year in the local smallholder farming system (Bayissa 2018). Teff flour is rich in proteins and minerals (Bultosa 2007), making it prized as a gluten-free superfood in western countries (Tietel et al. 2020; Gebru et al. 2020). As a consequence, teff is increasingly under the lens of local and international research to support its cultivation and commercialization (Chanyalew et al. 2019).

In Ethiopia, teff stands first in the total area cultivated and second to maize for total grain production and number of households producing it, but last in term of yield per area unit (Central Statistical Agency 2018). Its current average productivity is well below its genetic potential: improved cropping technologies in non-lodging conditions can result in yields exceeding 4.5 t ha−1 (Tefera and Ketema 2001), more than double the national average (Central Statistical Agency 2018). However, teff yield potential can be fully achieved only when applying appropriate sowing and plant curation (Ben-Zeev et al. 2020; Mihretie et al. 2021; Bogale et al. 2013), management of soil fertility (Wato 2019), and appropriate agronomic practices (Gezahegn et al. 2019; Berhe et al. 2013) including weed management (Rezene and Zerhun 2000) and pest management (Gemechu Degete 2021; Gyan et al. 2020).

Notwithstanding a large body of knowledge has been developed around the best practices for teff management, much progress can still be done in regards of its genetic improvement. Teff can be well considered a neglected and underutilized species (NUS) (Bachewe et al. 2019; Tadele 2018), a crop for which much potential exists still undisclosed by modern breeding approaches. It did not benefit of the leap forward of the green revolution that revolutionised the yields of other cereals, including the closely related rice and wheat. This depends on many factors, including the fact that only recently teff has been brought under the focus of the international scientific community. Nowadays, a complete teff genome produced using third generation genomics is available (VanBuren et al. 2020), and complements a draft genome published a few years ago (Cannarozzi et al. 2014).

As is the case of other NUSs, teff germplasm shows high variation for several useful traits that are seldom present in improved lines derived by formal breeding efforts (Jifar et al. 2018; Woldeyohannes et al. 2020). The diversity in teff germplasm, including wild relatives, landraces, and farmer varieties, is a reservoir of allelic diversity that once properly characterized may boost teff breeding (Girma et al. 2014; Cannarozzi et al. 2018) (Fig. 1). Teff landraces diversity is connected with the climatic variation existing across its growing area and to the socio-cultural process linked to its cultivation and trade (Woldeyohannes et al. 2020). Being mainly cultivated in a smallholder farming system with negligible use of agronomic inputs, teff germplasm has evolved at the interface of human and natural selection, accumulating variation useful to make it adapted to a range of abiotic and biotic stresses (Woldeyohannes et al. 2020), and including resistance to various pests (Chanyalew et al. 2019).

Fig.1
figure 1

An experimental field with 3850 teff landraces sourced from the EBI collection, at grain filling stage (West Gojam, Amhara, Ethiopia, 11° 16′ 32″ North, 37° 29′ 30″ East) (Woldeyohannes et al. 2020)

Full size image

This review emphasizes the breeding significance of teff genetic resources as a possible approach to address known challenges of teff farming. The first part of the review discusses the value of teff cultivation, its main constraints, and the status of conservation of teff genetic resources. The second part of the review provides a detailed description of the potential of teff genetic resources to improve key target traits. The concluding part of the review details molecular studies conducted in teff to support advanced breeding methods. We conclude discussing how teff diversity may be accessed with modern research approaches to maximize its agronomic performance, local adaptation, and farmers’ appreciation.

Status of teff cultivation and agrobiodiversity conservation

Appeal of teff cultivation and consumption

Teff is currently cultivated in Ethiopia by about 6.7 million rural households over 3 M ha, more than double the area that was allocated to its production in the 1990s (Central Statistical Agency 2018). The appeal of teff cultivation in the Ethiopian highlands has several reasons. Due to its capacity for local adaptation, teff is considered a low-risk crop by local farmers (Yihun et al. 2011). Early maturing cultivars are commonly used in areas with a short growing period, often as a replacement crop at times of failures of higher yielding long-season crops (e.g. maize). Early maturing teff cultivars have also a practical functionality for doubling with other cropping systems in high rainfall areas allowing for cultivation of pulses and oil crops (Ketema 1997). However, farmers may choose teff also for economic and nutritional considerations. In formal and informal markets, teff grain and straw fetch higher prices as compared to those of other cereal crops. Its flour is used for food preparations including injera (traditional circular, thin, fermented pancake), kitta (unleavened bread), porridge, muk (a kind of soup) and talla (local beer) (Ebba 1969). Teff flour is prized as it is highly nutritious and rich in minerals, fat and proteins, and micronutrients (Bultosa 2007; Gebru et al. 2020). Its straw, when used for feed, is also desirable due to low lignin content and high quality for crude protein content, in vitro dry matter digestibility, and energy value (Yami 2013).

The manifold advantages of teff cultivation and consumption make it a valuable resources to contrast malnutrition (Abewa et al. 2019) as well as a crop with high potential for global health food consumers (Lee 2018). The demand for gluten free foods is growing and expanding as more people are diagnosed with celiac disease and other types of gluten sensitivity (Bascuñán et al. 2020), making teff-derived bakery products ever more desirable in the western world and even a candidate for malting and brewing (Cela et al. 2020; Gebremariam et al. 2014).

Challenges of teff farming

Teff mean yield across Ethiopia, at 1.76 t ha−1, is much lower than that of other cereals cultivated in the same area (e.g. maize 3.9 t ha−1, wheat 2.7 t ha−1, sorghum 2.7 t ha−1 and barley 2.1 t ha−1) (Central Statistical Agency 2018). Teff low productivity is due to several production challenges that exist unchanged since thousands of years, some of which exacerbated by climate change (Table 1).

Table 1 Extent of estimated yield losses by cause in teff growing areas in Ethiopia
Full size table

Lodging is arguably the major bottleneck for teff farming (van Delden et al. 2010; Assefa et al. 2015; Wrigley et al. 2006). It is ultimately caused by the insufficient resistance of sclerenchyma in the culms, leading to failure to support grain bearing panicles when they become too heavy. Lodging is exacerbated by supplements of fertilizer, damage to the root system, heavy rates of seedling, and lack of nutrients (Rajkumara 2008), and affects yield quality and quantity by interfering with water and nutrient transport as well as with light interception (Ketema 1997). Soil features, including water availability, are the second most prominent limitation to teff cultivation. Low moisture deficit can indeed negatively impact teff growth, particularly at flowering and grain filling stages (Araya et al. 2011). In the western part of Ethiopia and in the highlands, soil acidity severely affects teff productivity, lowering teff response to fertilizer application (Abate et al. 2017). Biotic factors may also challenge teff cultivation, although they are less studied. Teff diseases include rust (Uromyces eragrostidis Tracy), head smudge (Helminthosporium miyakei Nisikado), damping-off (Drechslera spp.) and leaf spot (Helmithosporium spp.) (Badebo 2013; Gemechu 2018; Gemechu Degete 2021). Insect pests can also impact teff production, especially teff grasshoppers (Aiolopus longicornis, Aiolopus thalassinus), teff shoot flies (Elachiptera simplicipes, Melanochaeta vulgaris, Oscinella nartschukiana), teff red worm (Mentaxya ignicollis), wollo bush cricket (Decticoides brevipennis Ragge), and termites (Macrotermus subhyalinus and Odontotermus spp.) (Damte 2013).

The climate crisis may exacerbate current teff cultivation constrains, exposing the agroecosystems in the Horn of Africa to abiotic stresses potentially altering their productivity and function. In Ethiopia, farmers already favour drought tolerant crops and varieties to adapt to climate change, a pattern expected to consolidate (Marie et al. 2020). Climate change may also have an indirect effect through changes in the number, distribution patterns and virulence of pests and diseases (Black et al. 2011). Recent studies shown that teff cultivation suitability may diminish by 2070, urging long-term planning of breeding decisions (Woldeyohannes et al. 2020, 2021). In this scenario, the lack of improved varieties for specific environments (Assefa et al. 2015) calls for a more coherent characterization and utilization of teff genetic resources.

Ex situ and in situ conservation of teff germplasm

Ethiopia is the domestication center for E. teff (Harlan 1928) and a systematic collection, evaluation, and utilization of teff germplasm began in Ethiopia in the late 1950s. For the past five decades, the Ethiopian Biodiversity Institute (EBI) has collected and conserved a significant number of teff accessions, much exceeding those available ex situ in other gene banks (Table 2).

Table 2 Summary of 7235 teff accessions currently conserved ex situ in germplasm banks worldwide, with an estimate of passport data completeness
Full size table

Although the large number of teff accessions currently conserved at EBI, the exploitability of this resource may be further improved. Inside this gene bank, that operates at the highest international quality standards (Thomas et al. 2019), several accessions lack part of the passport information (Girma et al. 2014). A recent study reported that out of the 3850 teff accessions representing the active collection from the EBI, amounting to about 60% of the full collection, complete passport information was available for 1754 accessions (Woldeyohannes et al. 2020). Teff ex situ accessions can be in some cases duplicates, as the genetic redundancy of accessions between and within institutions is not fully known. Furthermore, most accessions lack information related to traditional name and farmers knowledge on the specific landrace, a feature that may be considered useful in designing further sampling campaigns to assess ethnographic significance of the accessions (Roncoli 2006). Wild species are seldom included in sampling campaigns and are not featured in the EBI collection: out of the 350 species in the genus Eragrostis, 14 are endemic to Ethiopia (Costanza et al. 1979) and may have high genetic diversity relevant for teff conservation and improvement (Girma et al. 2018).

Expanding teff collections is still a critical endeavor. In situ, the quick spreading of improved varieties of teff (e.g. Quncho) may permanently replace teff landraces in several agroecologies (Assefa et al. 2011). Moreover, in the absence of teff breeding materials with enhanced tolerance to biotic stresses such as soil acidity, the switch towards adoption of acidophilic crops may further accelerate the loss of teff genetic resources in regions where these stresses are prominent (Abate et al. 2017). All these developments call for urgent and constant updating and a further expansion of teff germplasm collections. Today, most of the EBI teff collection derives from sampling associated with the main roads of Ethiopia. A geographic information system (GIS) analysis looking at the intersection of EBI teff georeferenced accessions and the road network in Ethiopia shows that 61.7% of the accessions with GPS coordinates were collected within 500 m and 87.7% within 2000 m from the nearest road, leaving behind potentially relevant teff adaptation zones in remote areas (Fig. 2). This is a quite common feature of ex situ collections, and for good reasons of cost-effectiveness of sampling campaigns (Kasso and Balakrishnan 2013). However, future sampling campaign may focus on more remote areas, targeting the extremes of teff distribution (e.g. exceptionally low or high-altitude ranges) to harness additional adaptive teff variation of breeding relevance.

Fig. 2
figure 2

A map showing sampling points of a selection of 1754 teff accessions in the EBI collection overlaid to the Ethiopian road network. The EBI ex situ teff collection is the largest in the world, and a remarkable display of Ethiopian teff agrobiodiversity. Future national and international sampling efforts may focus on secluded areas far from the road network to capture additional teff diversity. Data from (Woldeyohannes et al. 2020)

Full size image

Teff breeding and diversity in ex situ collections

Teff research and breeding started in 1956 at the Jimma Agricultural and Technical High School, now Jimma Junior College of Agriculture. In 1960, it was transferred to the Central Agricultural Experiment Station, now the Debre Zeit Agricultural Research Centre. Since then, 49 improved varieties have been released (Ministry of Agriculture and Livestock Resources 2019). Of these, 25 were derived from farmer cultivars through mass selection, while the rest were obtained via conventional hybridization programs (Table 3). Throughout the teff breeding program, grain yield increased at an average of 0.8% to 0.9% per year (Teklu and Tefera 2005; Dargo et al. 2016). Varieties developed with hybridization yield 9% greater than those obtained through direct selection from germplasm, indicating that grain yield can be enhanced by active breeding (Assefa et al. 2013).

Table 3 List of modern teff varieties developed and released in Ethiopia by national and regional Agricultural Research Center until 2019 (Ministry of Agriculture and Livestock Resources 2019)
Full size table

In any crop, the success of varietal improvement is function of the combination and interaction of several components of agronomic performance. Literature reports show that most of teff traits are interrelated with one another and often changes in one trait are likely to influence others, so that the net gain obtained by selecting for a phenotype may be counter balanced or even negated by a simultaneous change in the others (Table 4).

Table 4 Trait associations reported in Eragrostis literature
Full size table

Due to the enormous diversity in teff genetic resources available to breeders, and due to the relatively early stage of its improvement, several reports have focused on the diversity of teff collections. We have discussed how much of teff cultivation in the Horn of Africa depends on landraces and traditional varieties that farmers select and propagate since centuries. As a result, landraces acquired traits for local adaptation that could be very relevant for breeding. The diversity included in these landraces is very large for agronomic traits, adaptation traits, and farmer preference (Woldeyohannes et al. 2020, 2021). Below, we discuss extant variation reported in landrace collections and breeding materials for key improvement traits.

Lodging

Lodging is arguably the most important bottleneck for teff improvement. This issue becomes more prominent with increased yield, panicle size, and biomass (Tefera et al. 2003; Muluken et al. 2020) and is exacerbated by fertilization in high input areas (Assefa et al. 2015; van Delden et al. 2010; Chanyalew et al. 2019). Lodging in teff is mainly due to stem failure, as shown by the fact that root lodging is seldom present (Muluken et al. 2020) and that varieties with compact panicles and reduced height have increased lodging resistance (Blösch et al. 2020). Reduced plant stature is therefore a main breeding target in teff. Teff genetic resources bear large variation for stem biomechanical traits that can contribute to lodging resistance (Muluken et al. 2020). Culm internode diameter may vary substantially (from 1.2 to 5 mm) (Ebba 1975), and thicker stems may also contribute to support higher panicle weight. Lodging resistant genotypes were produced though mutagenesis, successfully reducing plant height (Cannarozzi et al. 2018) and achieving high yield potential (Jifar et al. 2017).

Panicle traits

Teff yield is positively associated with panicle size, floret abundance, and shoot biomass (Chanyalew et al. 2009; Jifar et al. 2015; Ferede 2013). Longer panicles are preferred by farmers in agroecologies that allow longer vegetative growth, as they may result in higher yields (Tefera et al. 1990). Indeed, breeders may predict yield potential from panicle features (Adnew et al. 2005). Still, the rate of change of panicle traits in breeding is not sufficient per se in enhancing overall grain yield (Teklu and Tefera 2005). Large variation exist in teff collections for panicle related traits, including spikelet length (3 to 15 mm), spikelet width (1 to 3 mm), lemmas length (2 to 3 mm), lemmas width (1.3 to 2.03 mm) (Ebba, 1975) and number of grains per panicle (1520 to 6652) (Tefera et al. 1990). The EBI teff collection features accessions exhibiting very different panicle types, from very compact to extremely loose (Woldeyohannes et al. 2020) (Fig. 3). Compact types have a high spikelet number per panicle and are frequently cultivated under more favourable conditions (Woldeyohannes et al. 2020), however loose types can also be high yielding (Tefera et al. 1990).

Fig. 3
figure 3

Example specimens showing panicle type variability existing in the EBI teff collection. Photos taken in open field using white paper in the background to enhance contrast (Woldeyohannes et al. 2020)

Full size image

Plant and root architecture

At present, plant stature of teff is positively correlated with achieved grain yield (Tadele et al. 2013; Chanyalew et al. 2009; Tefera et al. 2003; Muluken et al. 2020). This is likely contributed by positive associations of plant height with panicle length (Jifar et al. 2015), thousand seed weight and harvest index (Tadele et al. 2013), days to maturity, culm length and diameter (Tefera et al. 2003). However, an increased plant height has not been a target for teff improvement, also due to its implications to lodging (Teklu and Tefera 2005). Biomass and particularly straw yield are priority traits for teff smallholder farming in Ethiopia, where straws are used for feed and house thatching (Jifar et al. 2018). When compared to barley straw, teff straw has a lower lignin content and higher quality for crude protein content, in vitro dry matter digestibility, and energy value (Yami 2013). In modern teff breeding, biomass increased at an average of 0.9% per year (Dargo et al. 2016) and showed that it may significantly contribute to enhance grain yield (Teklu and Tefera 2005). Plant vigour and production in teff can also be put in relation with root depth. Root traits in teff germplasm show wide genetic variation, including for root depth (59.3 to 116.5 cm), root number (18.3 to 72.8), and root shoot ratio (0.07 to 0.30) (Ayele et al. 2001). Increased root length may be associated with leaf turgor pressure under drought stress (Degu et al. 2008), as well as with salinity and acidity tolerances (Abate et al. 2013; Asfaw et al. 2011). Osmotic traits related to roots are also highly variable: in particular, osmotic adjustment (0.44 to 1.02 MPa), relative water content (97.8 to 99.5%) and osmotic potential (− 0.88 to − 1.15 MPa) (Ayele et al. 2001).

Seed traits

It is believed that the name teff derives from the Amharic word

(Teffa) for lost, possibly referring to the remarkably tiny size of the seeds and the ease to lose them (Fig. 4). However, seed colour over seed size has been a target for teff breeding in the past decades, as colour is a primary trait for selection of grains in both formal and informal markets (Belay et al. 2008). Teff seed color varies from dark brown to white (Woldeyohannes et al. 2020), but white seeds fetch higher market prices and indeed most of teff varieties developed by breeding are white in colour (Table 4). Still, brown seeded teff genotypes are reportedly associated with aluminium toxicity tolerance (Abate et al. 2013), and may have higher nutritional content, supporting the need for their valorisation. Seed weight improvement did not result in a significant increase since the 1970s (Dargo et al. 2016). Though small overall, seed size is highly varied in teff collections, e.g. for grain length (0.9 to 1.7 mm) and thousand grain weight (0.19 to 0.42 g) (Assefa et al. 2001; Ebba 1975), suggesting untapped potential for improvement.

Fig. 4
figure 4

Contrasting seed colour of representative teff genotypes in the EBI collection

Full size image

Phenology

Earliness is among the main teff adaptive mechanisms to prevent yield losses due to terminal drought. Possibly as a result of adaptation, teff accessions sampled in areas with lower rainfall have a shorter life cycle (Woldeyhannes et al. 2020). However, longer span of growth and later maturation are associated with increased yield and yield related traits (Tadele et al. 2013; Chanyalew et al. 2009; Tefera et al. 2003; Jifar et al. 2015; Assefa et al. 2002). In the EBI collection, a large variation exists for phenology traits: when evaluated in the same location, teff accessions mature with a span of 40 days from earliest to latest genotypes (Woldeyohannes et al. 2020). It is thus important that improvement for high grain yield should focus on maturity groups targeting different agroecologies.

Leaf traits

Leaf traits are related to photosynthetic efficiency as well as to water balance in the plant. Leaf size in teff collections show large variation, including in flag leaf area (2 to 26 cm2), leaf blade length (5 to 55 cm), and total leafiness of the plant (Ketema 1993; Ebba 1975). Across the genus Eragrostis, drought tolerance has been associated with increased leaf tensile properties (Balsamo et al. 2006). Tensile strength is higher in wild relatives than in E. tef (Balsamo et al. 2005, 2006), yet teff shows high variation for excised leaf water loss, drought deficit and leaf water potential, leaf relative water content and stomata conductance (Teferra et al. 2000). Differential responses to drought stress were observed among cultivars in association with leaf canopy temperature at anthesis, with higher temperatures associated to lower yields (Takele 2001).

Resistance to pests

Teff is regarded as relatively resistant to biotic stresses. Head smudge (Helminthosporium miyakei Nisikado) is arguably the most economically important disease in teff farming, and teff genotypes in collections showed some degree of resistance to it (Gemechu Degete 2021). No complete resistance is yet available for teff rust (Uromyces eragrostidis Tracy), another disease with broad diffusion in Ethiopia (Gemechu 2018; Badebo 2013). Teff rust typically occurs after heading stage, yet causes relatively little grain yield losses as compared to other constraints (Dawit and Andnew 2005). Resistance to aphids (Rhopalosiphum padi) is also available (Zafar et al. 2020), yet further evaluations are required to determine the status of genetic resistance alleles in teff germplasm collections.

Nutritional features

Much of the national and international success of teff is due to its unique flavour and nutritional properties. Starch makes up about three-quarters of teff flour. Its amylose content (20 to 26%) is comparable to that of most cereals (Bultosa 2007), but the total dietary fibre content of whole grain teff (9.8%) is higher than that of major cereals and even higher than that of quinoa (7.1%) (Gebru et al. 2020). Teff germplasm displays high variability in nutritional properties yet the highest iron and calcium contents are recorded in brown seeded varieties (Table 5) (Gebru et al. 2020; Yami 2013; Baye 2014). Fat (2% to 3%) and protein content (8% to 11%) in teff grain is similar, in some instances better, than that of other more common cereals (Moharram and Abu-foul 1992; Baye 2014), with a balanced amino acid composition and relatively high concentration of lysine (Ketema 1997). In teff flour, riboflavin ranges from 0.13 to 0.14 mg/100 g, niacin from 1.7 to 1.8 mg/100 g, and thiamine from 0.3 to 0.6 mg/100 g, higher than in most common cereals (Ketema 1997). Polyphenols and phytates are also present in high concentration (Baye 2014). This is also related to the fact that, due to its small size, the grain cannot be divided into germ, bran and endosperm during processing and it is consumed as a whole (Ketema 1997). The total phenolic content (mg GAE/gr) in teff ranges from 0.89–1.2 to 1.04–1.27 in white and brown grains, respectively (Tietel et al. 2020). Feed traits are associated with food quality traits and yield traits, confirming the possibility of improving feed quality traits without significantly affecting grain yield (Jifar et al. 2018).

Table 5 Selected micro and macro nutrient concentrations (mg/kg) of white seeded teff compared with brown seeded teff types
Full size table

Perspectives in teff breeding

The role of traditional knowledge

Smallholder farming is the most dominant form of agriculture in the Horn of Africa, contributing significantly to food security in the region. Teff is no exception. Smallholder teff growers accumulate and propagate indigenous knowledge that could be useful for a robust crop breeding program targeting local adaptation. In smallholder farming settings, traditional knowledge often drives selection and maintenance of germplasm for local adaptation (Fadda et al. 2020). In marginal cropping environments, farmer priority traits could be different from traits targeted by formal breeding objectives. Farmers’ choice of varieties is indeed related to their desire to meet local economic, social and agroecological conditions (Ngonkeu et al. 2017; Mancini et al. 2017; Christinck et al. 2017; Fadda et al. 2020).

An integrated participatory characterization of crop genetic resource may be useful to design and address farmer preference traits in modern varieties to increase the adoption of genotypes and thereby increase productivity (Rahman et al. 2015; Ngonkeu et al. 2017), enhancing in situ conservation of indigenous knowledge. This approach has been reported in Ethiopian durum wheat accessions that showed the smallholder farmers’ evaluation processes are quantifiable and repeatable (Mancini et al. 2017), and their knowledge can be harnessed in breeding programs (Kidane et al. 2017, 2019). In teff, preliminary results about farmer priority traits have been reported (Woldeyohannes et al. 2021), yet further research is required to fully translate this knowledge to breeding decisions.

Genetic variation and molecular breeding

Genetic variation is the raw material to fuel teff improvement. Although teff germplasm is highly diverse, some traits in the currently surveyed collections may still lack desirable variation, e.g. lodging tolerance (Assefa et al. 2013). Mutagenic agents may then be used to generate novel genetic variation from which desired individuals may be selected, and this approach was successfully used in teff (Cannarozzi et al. 2018). Among mutagenized teff lines, promising candidate lines were identified for seed size, herbicide tolerance, drought, soil acidity and salinity tolerance via subsequent phenotypic screening (Table 6).

Table 6 Valuable traits developed in mutagenized teff populations (Cannarozzi et al. 2018)
Full size table

Analyzing the molecular diversity encompassed in teff genetic resources is a prerequisite for their efficient exploitation in breeding and for the development of conservation strategies of its genetic diversity. In the past two decades, molecular markers technologies were used to characterize the diversity of teff and related species in manifold studies (Table 7). Yet, advanced molecular breeding in teff has seen a very limited use as compared to other key cereals grown in Ethiopia (Girma et al. 2014, 2018; Teshome et al. 2020; Assefa et al. 2015). Most of teff priority traits highlighted above have not yet been exploited using modern molecular techniques. There is in general very limited information on high throughput discovery of single nucleotide polymorphisms (SNPs) and other molecular markers relative to Eragrostis species diversity with genomic based tools (Girma et al. 2018).

Table 7 A selection of genomic and molecular studies performed on teff and related species in the last two decades
Full size table

However, the accumulation of genomic information for teff (Cannarozzi et al. 2014; VanBuren et al. 2020; Woldeyohannes et al. 2021; Gebre et al. 2016) make teff ripe to be brought into the era of molecular breeding. Quantitative trait loci (QTL) have been described on teff for several traits including yield components and morphology (Chanyalew et al. 2005; Yu et al. 2007), also in relation to drought (Degu and Fujimura 2010). Recently, a large collection of teff landraces have been used to conduct a genome wide association study unveiling genomic loci potentially responsible for agronomic performance, climatic adaptation, and farmers’ appreciation (Woldeyohannes et al. 2021). Once genes underlying QTL are discovered, genome editing may be used to enhance the agronomic performance of teff varieties. Efficient methods for transformation and regeneration of transgenic lines as those developed for cereal crops such as sorghum, maize, wheat, and rice (Numan et al. 2021) are needed before the potential of genome editing can be fulfilled in teff. However, early reports of transformation of teff to induce dwarfism suggest that efficient transformation protocols may be achieved also in this species (Gebre et al. 2013). In rice, the editing of a handful of genes involved in yield determination generated mutants with increased grain number, tiller number, dense erect panicles and large grain size and plant architecture (Rakshit et al. 2020). Similarly, panicle architecture may be edited to generate tiller spreading phenotype enhancing crop yield in rice (Zafar et al. 2020), suggesting similar applications for teff breeding.

Conclusion

Teff breeding is approaching a golden age contributed by the emergence of knowledge and tools deriving from its vast and untapped diversity. In Ethiopia, teff breeding may be conducted with two generations per year. Even so, more than 10 years are required today to develop and release an improved teff variety using hybridization (Chanyalew et al. 2019). Genomic innovations supporting molecular breeding may be put at use with alternative breeding methods including speed breeding (Chiurugwi et al. 2019; Watson et al. 2018) and 3D-breeding (van Etten et al. 2019; de Sousa et al. 2021) to speed up the development of teff varieties with enhanced local adaptation and farmers’ uptake. The enhancement of teff productivity, nutritional quality, and farmer appreciation may leverage the great diversity existing in teff collections and use modern molecular tools to open a new era of teff breeding. The appropriate combination of this wealth of information is needed to revolutionize teff cropping and propel it towards the international market.