ITEMS FROM ETHIOPIA
INSTITUTE OF BIODIVERSITY CONSERVATION AND RESEARCH
P.O. Box 30726, Addis Ababa, Ethiopia.
Introduction. The loss of crop genetic resources in countries like Ethiopia can be linked to the spread of modern agriculture in two major ways. First, the introduction of high-yielding, genetically narrow, uniform cultivars (often less dependable than the landraces they replaced when grown under the traditional agricultural management), resulted in abandonment of the broad genetic base in farmers' varieties. Second, the planting of vast areas to uniform cultivars with a narrow genetic base made agricultural productivity extremely vulnerable to yield-limiting factors such as disease and pests. According to Loutte et al. (1997), farmers also were given several socioeconomic incentives to replace varieties that evolved within their agro-ecosystem with improved/introduced varieties in many regions of the world.
Efforts to conserve crop diversity and cope with the rapid loss of genetic resources to date have focused on maintaining genetic diversity in static ex situ gene banks. However, a variety of problems with reliance on ex situ conservation strategies have been acknowledged, such as inadequate sampling procedures during field collection and the lack of representation in gene banks of the entire range of diversity of a given crop and its close genetic relatives (Altieri et al. 1987). The complex interaction of genetically diverse, indigenous varieties (farmers' verities) with their associated pests, predators, and pathogens has been arrested. The traditional farmer fails to retain knowledge associated with farmers' varieties.
Recognizing this situation was the basis for a conservation and enhancement strategy for in situ (on-farm) landraces/farmers' varieties in 1989 in Ethiopia through the project 'A Dynamic Farmer Based Approach to the Conservation of Ethiopia's Plant Genetic Resources' funded by Global Environment Facility (GEF). Through a novel strategy of establishing community gene banks, the Institute of Biodiversity Conservation and Research (IBCR) in Ethiopia has played a leadership role internationally in establishing an on-farm conservation mechanism to complement ex situ conservation. The project was executed in six zones in the country including the Bale and eastern Shewa zones. Durum wheat is one of the crops under the conservation program in both of these sites.
The uniqueness of the Ethiopian durum wheat germ plasm was observed by Vavilov (1957) and later confirmed by others (Pecetti et al. 1992). Important features for crop improvement have been reported in the Ethiopian wheats of farmers' varieties, including resistance to leaf rust, powdery mildew, and glume blotch; long coleoptile short culm, early ripening, high protein content, and adaptation to low soil fertility and waterlogged conditions; and resistance to drought and Hessian fly (Proceddu et al. 1975; Amri et al. 1990; Pecetti et al. 1992; Bechere and Tesemma 1997).
Determining the magnitude of genetic variability and the pattern of distribution in different in situ crop conservation sites and comparing it with ex situ collection is essential for the successful conservation and sustainable utilization of the genetic materials. Such a study also will contribute tremendously to ex situ gene bank management, the identification of sites for in situ conservation, and serve as a benchmark for future assessment of the extent of genetic erosion. The present study is, therefore, aimed at assessing the level of genetic diversity in durum wheat by comparing two in situ crop conservation sites and populations from ex situ collections.
Material and Methods. Ten and 19 durum wheat farmers' varieties (locally named varieties) currently under the in situ conservation program in the Bale and eastern Shewa sites, respectively, and 48 durum wheat accessions (representative samples of the original materials that were collected in the Bale zone where the current in situ site is located) were used.
A randomized, complete-block design with two replicates was used. Each of the farmers' varieties were grown in a replicate in two 4-m rows with a 0.2-m row spacing. Seeds were sown at a rate of 125 kg/ha. Altogether, a total of 77 populations (2,816 individuals) were used for this study.
Five qualitative traits, beak awn length (short (1), intermediate (2), or long (3)), glume color (white to yellow (1), red to brown (2), or purple to black (3)), glume hairiness (absent (1) or present (2)), seed color (white to yellow (1), red to brown (2), or purple to black (3)), and spike density (lax (1), intermediate (2), or dense (3)) were used to compare the two in situ and one ex situ sites. These characters were chosen because their expression is little or unaffected by environment, which makes them reliable morphological markers for characterization of wheat germ plasm. Farmers who use folk taxonomy to identify their varieties also use most of the traits selected in this study. These traits attract the eye of the wheat germ plasm curator and signal the presence and/or absence of phenotypic variation in the field according to Belay (1997).
Percentage frequencies of each phenotypic class of each character was used to estimate the Shannon-Weaver diversity index, H, which is defined as:
where n is the number of phenotypic classes per character and pi is the proportion of the total number of plants in the ith class. H was standardized by converting to the relative index, H' = H / Hmax. A correspondence analysis (on multivariate basis) was used to plot the populations in a low-dimensional graphic presentation. Correspondence analysis is a weighted principal component analysis of a contingency table. The computations were conducted using SAS version 8.002 (1999).
Result and Discussion. Estimates and analysis of diversity. Table 1 gives the estimation of H' for each of the five characters in the two in situ conservation sites and the ex situ gene bank. The highest mean-diversity index (') pooled over characters was obtained for the eastern Shewa and Bale in situ conservation sites ('= 0.81 and 0.64, respectively), whereas populations from the ex situ had the lowest level of diversity ('=0.49). The overall mean diversity index for both the in situ sites and the population from the gene bank was 0.73 ± 0.09. The correspondence analysis (on multivariate basis) also was very accurate in discriminating between the populations. Plotting the first two dimensions obtained from the correspondence analysis indicated that the populations from the gene bank and the in situ had distinct morphological attributes (Fig. 1). Populations from the two in situ sites seemed to be more diverse than their ex situ counterpart. Most of the populations sampled from the gene bank were negative in the first dimension and showed less variability. We also observed that the materials from the gene bank have poor germination potential and vigor in the field.
| Seed source | Beak length | Glume color | Glume hairiness | Seed color | Seed density | ± SE |
|---|---|---|---|---|---|---|
| Bale (in situ) | 0.76 | 0.82 | 0.40 | 0.95 | 0.28 | 0.64 ± 0.13 |
| E. Shewa (in situ) | 0.74 | 0.95 | 0.92 | 0.84 | 0.62 | 0.81 ± 0.06 |
| Gene bank (ex situ) | 0.48 | 0.84 | 0.00 | 0.93 | 0.18 | 0.49 ± 0.18 |
| Total | 0.66 | 0.91 | 0.55 | 0.99 | 0.52 | 0.73 ± 0.09 |
The relative contribution of the five qualitative characters for H' varied considerably (Table 1), with high variation for seed color (H' = 0.99), glume color (H' = 0.91) and beak length (H' = 0.66). Seed color was the most diverse character in both in situ sites and the gene bank populations. A high level of diversity for seed color also has been reported in a world collection of Ethiopian tetraploid wheat (Jain et al. 1975) and in collections from the central highlands (Tesemma and Belay 1991) and northern and northcentral regions of Ethiopia (Bechere et al. 1995). A possible reason for the occurrence of such a high level of diversity of seed color could be the utilization of different seed-color types in traditional consumption for different purposes in Ethiopia.
ANOVA revealed significant differences between the conservation strategies (in situ and ex situ conservation, see Table 2). Traits such as spike density, glume hairiness, and seed color played major roles in distinguishing populations.
| Characters |
|
|
|---|---|---|
| Beak length | 0.14 | 0.09 |
| Glume color | 0.09 | 0.09 |
| Glume hairiness | 2.07** | 0.05 |
| Seed color | 0.31* | 0.09 |
| Spike density | 0.66** | 0.03 |
These results are further supported by partitioning of the phenotypic diversity within and between the conservation strategies (Table 3). Eighty-seven percent of the total variation was found between the conservation strategies, whereas 13 % was found within the conservation strategies. Spike density, glume hairiness, and seed color contribute relatively more (30 %, 20 %, and 8 %, respectively) to the differentiation between the in situ and ex situ conservation sites. These traits are more useful in discriminating the two in situ sites from the ex situ in this particular study.
| Characters | H'di' | H'c' | H'c' / H'di' | (H'di'- H'c') / H'di' |
|---|---|---|---|---|
| Beak length | 0.66 | 0.66 | 1 | 0 |
| Glume color | 0.91 | 0.87 | 0.96 | 0.04 |
| Glume hairiness | 0.55 | 0.44 | 0.80 | 0.20 |
| Seed color | 0.99 | 0.91 | 0.92 | 0.08 |
| Spike density | 0.52 | 0.36 | 0.69 | 0.31 |
| Mean | 0.73 | 0.65 | 0.87 | 0.13 |
H'di' = diversity index for each character calculated from the entire data set; H'c' = average diversity index of each character for the two in situ and the ex situ sites; H'c' / H'di' = Proportion of diversity within the two conservation strategies; (H'di'- H'c') / H'di' =Proportion of diversity between the two conservation strategies in relation to the total variation.
According to Belay et al. (1991), spike density is highly associated with environmental conditions especially with altitude. They noted that lax spike is a dominant trait in areas at high altitude and high rainfall. In these conditions, lax spikes may confer resistance to diseases that attack the spike and susceptibility may be associated with the degree of compactness (Belay et al. 1991; Parry et al. 1995). In the arid condition in Syria, dense spikes were reported to be dominant (Ellings and Nachit 1991). The association of glume hairiness with resistance to Karnal bunt (Warham 1988) and powdery mildew (Negassa 1986) also have been reported. Among the seed-color groups, the purple-black type was reported to have an earlier maturity and higher tillering capacity than the other groups, making them better suited to waterlogged soil conditions in areas of high altitude (Belay et al. 1995).
These results clearly show that traits, especially the spike density, glume hairiness, and seed color, are highly associated with environmental factors such as temperature and rainfall. This, in turn, indicates the traits have an adaptive significance.
Considerable evidence indicated that damage to chromosomes, some of it resulting in heritable changes, takes place as seeds loose their viability. Studies in barley and wheat showed that as storage age increases, chromosome aberrations (per cell) increases (Gundhardt et al. 1953). Changes in the properties of DNA associated with loss of viability in rye seeds, namely the loss of DNA-template activity (Holden and Williams 1984) and decreases in the molecular size of extractable DNA (Cheah and Osborne 1978), also have been observed.
In conclusion, conservation at the on-farm level allows for continuing farmer selection, interaction with the environment, and gene exchange with wild species so that evolution of landraces may continue. Under this system, many cultivated crops species have coexisted with local environmental factors, and the flow of genes among genotypes has taken place with minimum interference. The development of new variations and increasing the diversity within the crop genotypes has been aided. Therefore, ex situ conservation needs to be complemented in a way that will maximize the retention and continued evolution of the adaptive qualities of landraces, which will avoid the loss of variation that occurs in sampling and maintenance. The present study indicates the necessity of strengthening and expanding the in situ conservation programs in a broad range of agroecological conditions to obtain maximum diversity and utility as source materials in crop-improvement programs.
References.