«Eragrostis tef(Zucc.) Trotter Seyfu Ketema Biodiversity Institute Addis Abeba, Ethiopia 2 Tef. Eragrostis tef (Zucc.) Trotter The International Plant ...»
Although on a limited scale, researchers on tef have always used tef germplasm in their improvement programmes. The limited use was imposed by the lack of comprehensive characterization data for various essential traits. Prior to 1993 characterization data were not available for more than 200 accessions (Seyfu 1991).
As indicated above, a more systematic and comprehensive use of tef germplasm in tef breeding and other research work has commenced (Seyfu 1993).
27 Promoting the conservation and use of underutilized and neglected crops. 12.
7 Breeding Breeding activities Applied breeding work to improve tef included direct selection from the landraces, interspecific and intraspecific hybridization and mutation breeding, while at the basic research level investigations were made in the area of biotechnology The applied research attempts in the areas of mutation and interspecific hybridization programmes have not yet contributed to the development of improved cultivars.
On the other hand, the direct selection from the landraces and the intraspecific hybridization programme which was employed to effect gene recombination were successful in developing several improved cultivars of tef with desired traits. In the intraspecific hybridization programmes the pedigree and modified pedigree selection methods were used to handle the segregating population. The improved cultivars developed include: cultivars that have high grain yield with wide or specific adaptation, cultivars with acceptable high grain quality, and early maturing, highyielding varieties. All the improved cultivars were accepted by farmers and currently are in production. Direct selection from the landraces, mutation breeding and intraspecific hybridization were tried for developing lodging-resistant varieties.
However, so far no success has been achieved. Lodging is still one of the production constraints and therefore the breeding programme has the development of lodgingresistant varieties as one of its objectives. Other production constraints are: lowyielding cultivars, low moisture stress resistance, waterlogging, frost, weeds, poor soil fertility, diseases and insects. Generally the crop improvement programme in tef attempts to solve these production constraints through a multidisciplinary research approach. Specifically, the breeding programme should overcome the problems of low grain yield, and also develop cultivars that are resistant to low moisture, waterlogging and disease as there is a wealth of genetic diversity within tef germplasm.
Hybridization technique The small size of the floret, its autogamous nature and its unique pollination habit
- pollination occurs only during the early hours of the morning - require the employment of an appropriate artificial hybridization technique. Under Ethiopian climatic conditions, tef flowers open and pollinate only between 6:45 and 7:45 a.m.
(Tareke 1976). Techniques for delaying flower opening were developed to control pollination and to effect hybridization whenever required during the late hours of the day. These included putting potted plants overnight under cold temperature (5°C) or under complete darkness at temperatures of 14-22°C (Seyfu 1983). The process of initial flower opening to anther dehiscence takes about 30-40 minutes and once pollination is effected the pollen grains take only 3-4 minutes to germinate on the stigmas (Seyfu 1983). The suggestion of Tareke (1981) to use pollen from stigmas as a source for hand-pollination for an hour and a half is not recommendable. This is because pollen grains taken off the stigma are useless after 3-4 minutes since they 28 Tef. Eragrostis tef (Zucc.) Trotter would have already germinated and sent their pollen tubes down the style and thus will not be viable and functional for pollinating another stigma. Detaching the florets, which contain the anthers and their pollen intact, from a spikelet when they just begin to open for natural self-pollination, and keeping them in moist vials could prolong the viability of the pollen grains for use in pollination (Seyfu 1983). Anther dehiscence can also be delayed by removing undehisced anthers with jeweller’s forceps from florets as the flowers open under the influence of light conditions and increased temperature and transferring them to damp filter paper in a petri dish.
The humidity here delays dehiscence for a considerable time (Ponti 1978).
During artificial crossing, the state of the stigma assists in identifying the receptive and ready florets for pollination. The receptive florets have fluffy, feathery and turgid stigmas, while the nonreceptive ones have sticky and nonturgid stigmas (Seyfu 1983).
The following procedure for the artificial hybridization of tef is suggested by
Grow, one or two plants of each parent in separate pots of about 13-cm diameter.
Put the female plant and the pollen donor plant into separate light-tight dark boxes at about 2:00 p.m., 8-18 days after anthesis begins on the central or any other tiller. Keep the boxes away from direct sunlight at a temperature well below 28°C; lower temperatures improve the degree of control over flowering. Next day, crossing may be done any time before early afternoon.
Take the pollen donor plant out first and wait for 3-5 minutes. As soon as it starts to open its flowers, detach the spikelets with open florets using forceps and attach them to the moistened inner wall of a vial. Label the vial with a code number of the plant (these can be used as the pollen source later).
Take out the female plant, lay it horizontally under a binocular microscope (x15), and begin to emasculate as soon as the flowers start to open and before the anthers dehisce. Only the basal florets should be emasculated, the other florets on the spikelet being removed. This enables identification of the treated flowers.
Keeping the emasculated flowers under observation, remove a spikelet from the vial. Detach an individual anther with forceps while observing beneath the binocular microscope or while the anther is still attached to the flower; gently squeeze out the pollen directly onto the stigma of the emasculated floret.
Label each plant for crossing records.
A skilled operator can avoid hand-emasculation by applying the donor pollen before the anthers in the flowers of the female parents dehisce, although this procedure involves some risk.
Hybrid identification One of the following means can be used to identify successful hybrids from accidental selfs after artificial hybridization in the F1 generation: dominance is displayed by loose panicle, coloured lemma and coloured seeds over compact panicle, yellowish white lemma and white seed, respectively. Therefore seed colour, lemma colour and 29 Promoting the conservation and use of underutilized and neglected crops. 12.
panicle form can serve as the useful genetic markers on mature F 1 tef plants (Tareke 1975). In addition, the following techniques can help distinguish the hybrids from the accidental self-pollinated seeds in a very short time, mainly because the techniques do not require growing mature plants. Plants with florets having red lemmas have red coleoptiles when their seeds are germinated under direct sunlight.
These show dominance over plants that have florets with yellowish white lemmas and green coleoptile during germination. Hence coleoptile colour can serve as a genetic marker at earlier growth stages (Seyfu 1983). Similarly, the dark embryo that develops only in mature caryopsis and shows dominance over the pale embryo can serve as a marker to identify the F1 hybrid seeds (Ponti 1978).
Inheritance of characters and selection
The inheritance patterns of tef for the following three traits (Tareke 1981) were:
lemma colour - purple, grey, red, and yellowish brown seed colour - dark brown, medium brown, yellowish white and greyish white panicle form - loose and compact.
Four pairs of genes were found to be involved in the inheritance pattern of the lemma colour. Gene action and interactions of dominance, codominance, complementarity and epistatic nature have been observed on tef plants. Thus the following genotypes were given to the four cultivars having different colours: fesho (purple) = CCPPP2P2GG; bursa (grey) = ccPPP2P2GG; key murri (red) = CCppp2p2gg;
and Trotteriana (yellowish-white) = ccPPP2P2gg. Duplicate pairs of genes were identified in the inheritance of seed colour and the gene action between these genes and their alleles was simple dominance with additive effects. Panicle form was conditioned by duplicate pairs of genes for degree of looseness and another pair of genes for branching pattern.
The fact that tef showed a disomic inheritance pattern for lemma colour, seed colour and panicle form indicates that it is an allotetraploid; thus the quantitative genetic theory developed for diploids, including the procedures used to estimate the genetic variance and heritability of diploids, is applicable to tef.
A preliminary genetic experiment using the F2 and F3 generations of a single cross indicated that gene actions for most agronomic traits of tef were not simple additive/dominance only (Hailu et al. 1992). For example, in the case of yield and yield components of tef, the effect of nonadditive genes has been verified (Hailu 1993).
A study on two sets of crosses indicated that additive and dominance by dominance gene effects control the inheritance of grain yield. The magnitudes of the additive terms were lower than the dominance by dominance interaction term, which signals that selection based on grain yield per se might be difficult. Triple test cross analysis also revealed the presence of epistasis for grain yield in this crop.
Nonallelic gene interactions were also reported for yield per panicle, panicle weight, tiller number, harvest index, plant weight, plant height, panicle length, culm thickness, kernel weight, days to heading and days to maturity using different mating designs (Hailu 1993).
Tef. Eragrostis tef (Zucc.) Trotter 30 Epistasis has been found to be part of the genetic architecture of grain yield and other agronomic traits of tef and has been detected by different procedures unequivocally. Previously, gene interaction in tef was reported for some traits (Tareke et al. 1989a, 1989b, 1989c) and therefore it is not surprising to detect the same for more complex quantitative traits. Therefore, future quantitative genetic analysis in the tef plant must consider procedures that may allow the detection of epistasis to avoid biased estimates of additive and dominance components.
In small cereals, several researchers have reported that traits like kernel weight, number of kernels per head and tiller number have a strong association with grain yield. These characters are widely accepted as yield components. In the tef plant, Melak Hail et al. (1965) computed a positive correlation between seed yield and panicle length at Purdue in the USA. Under an Ethiopian environment, Hailu (1988) observed kernel weight per main panicle and productive tillers positively associated with grain yield. In a study using early generations of a single cross, Hailu et al.
(1992) obtained panicle weight per plant (a trait closer to true grain yield) positively associated with panicle weight per primary tiller and productivity index.
Subsequent studies confirmed that yield per panicle and panicle weight showed strong correlation with grain yield per plant (Hailu 1993). These traits have high heritability compared with grain yield per se and therefore they are useful criteria for grain yield selection in tef.
Biotechnology All of the cytological literature on tef and other Eragrostis species reports the difficulties in studying tef chromosomes because of their small size (0.2-l µm) and their refractory nature to staining techniques, and recommends improvement in the technique before any useful result can be expected. The meristems of the main roots are small and divisions are relatively few; the small roots of seedlings grown in petri dishes were also studied but were not suitable. Meiosis is difficult to observe owing to the few pollen mother cells in each anther. This suggests that detailed studies of the genome composition of tef and other Eragrostis species to obtain meaningful results are not possible at this stage. However, the more efficient techniques now available (image analyis, flow cytometry and banding techniques) coupled with molecular techniques should make it possible to analyze individual chromosomes and the genome architecture of tef more accurately Establishment of plant cell and tissue culture techniques for tef is a prerequisite to successful plant transformation and regeneration. Embryogenic cell lines in particular can provide a suitable source of embryogenic protoplasts for culture and regeneration, and the development of a single cell system which is a sine qua non for direct gene transfer techniques such as electroporation and micro-injection.
Somaclonal variation has potential as a new option to generate additional genetic variability in co-adapted, agronomically useful cultivars without the need to resort to hybridization and so far has provided breeders with new sources of variability to incorporate into conventional breeding programmes. Agronomically useful traits 31 Promoting the conservation and use of underutilized and neglected crops. 12.
such as increased tolerance to physiological stress and pests and diseases have been recovered from such materials. In vitro mutation techniques can be useful to identify lodging-resistant tef lines.
Tesfaye (1991) reported preliminary results in cell and tissue culture and genetic transformation and also outlined biotechnological approaches for tef research.
According to Tesfaye (unpublished), embryogenic callus from leaf explant and seeds was induced with 1.0-2.5 mg/L of 2,4-D in the culture media. Similarly, Firew (1991) regenerated tef plantlets through somatic embryogenesis originating from cultured leaf explants. Appropriate culture media for zygotic embryos were studied for use for embryo-rescue in tef (Cheverton 1985). Callus was obtained from the fusion product of tef and sorghum (Endeshaw 1992). In vitro manipulation of the spikelets of tef showed that tef spikelets develop successfully in Wheat Spikelet Medium (WSM) (Hailu 1993). Furthermore, spikelets of the F1 (varieties: Fesho x Kaye Murri) showed good early development.