Tải bản đầy đủ - 0trang
V. Evaluation of Wild and Primitive Forms of Wheat and Barley
EVALUATION OF GENETIC RESOURCES IN CEREALS
germplasm available for evaluation, and the desirable traits revealed, and
whether repeated backcrossing is needed to eliminate undesirable traits
introduced from the wild or primitive parent. The evaluation of wild
relatives of crop plants, including wheat and barley, has also been reviewed by Harlan (1984).
Many regions in the world have succeeded in improving crop production
through the introduction of high-yielding varieties. However, these varieties have not met with great success in West Asia and North Africa because
of their intolerance to an environment where moisture is limited and inputs
are very low (Srivastava and Damania, 1989). To improve and stabilize
cereal production in the drylands, characters such as earliness plus tolerance to drought, temperature extremes, low plant nutrients, and diseases
need to be incorporated into varieties. The genes for these characters can
probably be found in wild relatives and primitive forms that are well
adapted to such environments through independent survival over a very
long period of time.
Wheat and barley plant breeders in developed countries consider landraces and primitive forms as unadapted germplasm and Hallauer and
Miranda (1981) termed as exotics all germplasm that does not have immediate use without selection for adaptation to a given area. The exploitation
of wild and primitive forms in wheat breeding has been insufficient for four
reasons. First, collections of wild progenitors in the past have been fragmentary as well as scanty and material available in collections is not
representative (Croston and Williams, 1981). Second, work on wild forms
has primarily concentrated on evolutionary (Feldman and Sears, 1981) and
taxonomic studies (Bowden, 1959; Chennaveeraiah, 1960; Morris and
Sears, 1967; Kimber and Feldman 1987). Third, wild relatives are not well
adapted to Europe and North America and hence are used mainly as
single-gene donors. Finally, variability between and within populations of
wild species has not been looked at in adequate detail and utilization has
hardly commenced (Srivastava et al., 1988).
The principal reason for utilizing wild species such as Aegilops, Agropyron, and Triticum dicoccoides Korn. in wheat breeding has been the
transfer of genes for disease resistance, salinity tolerance, and highprotein content, respectively, when these desirable traits are not found in
the available genetic stocks. However, confirmation of the sources of
resistance and cataloging of genes for resistance should precede any attempt at transfer from the wild to the cultivated forms. This would involve
collecting and evaluating in-depth relevant germplasm collections for genetic variability. In-depth evaluation would consist of replicated tests at
different locations, resulting in genetic analyses and observations on traits
such as disease resistance.
A. B. DAMANIA
It is also useful to gain knowledge of genetic relationships between the
wild donor and cultivated recipient species because without ready crossability, desirable traits cannot be easily transferred. Thus, wild progenitors
of the cultivated species should be considered an important source of
variability for broadening the genetic bases of cultivated crops (Harlan,
1976; Hawkes, 1977; Lange and Balkema-Boomstra, 1988). Brown (1978)
states that the genetic resources of wild relatives of crop plants should be
systematically evaluated, for these sources of genes will supplement, and
even rival, the primitive landraces in their effectiveness in crop improvement programs. There is an opinion among certain workers, especially
those involved in breeding for favorable environments, that the variability
in landraces of wheat and barley has been fully exploited and for further
progress in introducing useful genes one should turn to wild relatives and
other alien germplasm of the secondary and tertiary gene pools.
Strampelli, working in Italy, was among the first plant breeders in
Europe to utilize wild and primitive genetic resources, especially Triticum
villosum Schur. and T . spelta L., to improve wheat in 1906 (Maliani and
Bianchi, 1979). Elliot (1957) transferred stem rust resistance to common
wheat from Agropyron elogatum and Riley et al. (1968) transferred yellow
rust resistance from Aegilops comosa Sibth. to cultivated wheat by genetically induced homologous recombination.
One of the most detailed evaluation studies of morphological, physiological, genetic, and cytological characteristics of Aegilops and Triticum
species was carried out at the University of Kyoto (Kihara et al., 1965). A
geographical survey of species of wheat and its wild relative Aegilops was
conducted following the University of Kyoto expedition to Afghanistan,
Iran, and Pakistan in 1955. Emphasis was placed on collecting Aegilops
squarrosa L., the probable donor of the D genome to cultivated wheat;
other species were also sampled whenever possible. There was evidence
that bread wheat arose as a hybrid between cultivated emmer ( T . dicoccum
Schub.) and A . squurrosu. This study was not only illustrated with photographs, but also contained valuable information on utilization of the
germplasm. Cox et a/. (1989) evaluated 212 accessions of A. squarrosa
from the University of Kyoto collection using polyacrylarnide gel electrophoresis and found gliadin diversity to be higher than in the D genome of
cultivated bread wheat.
Genetic variability in Aegilops species and primitive forms of wheat has
been studied by several researchers (Hillel et al., 1973; Sharma et a/.,
1981; Dhaliwal et al., 1986; Waines et al., 1987), whereas in the past more
emphasis was placed on Triticum dicoccoides, the wild progenitor of
wheat (Lawrence et al., 1958; Avivi, 1979) as reviewed by Poyarkova
(1988). Also, Gerechter-Amitai and Stubbs (1970) found the source of
resistance to yellow rust in T. dicoccoides from Palestine.
EVALUATION OF GENETIC RESOURCES IN CEREALS
In another evaluation of primitive and wild types of wheat comprising 75
accessions of T . boeoticum Boiss., T . araraticum Jakubz., and T . dicoccoides; 42 of T . urartu Tuman.; and 34 of A . squarrosa seeking resistance to leaf rust, Hessian fly, and greenbug, Gill et al. (1989) found that all
species except T . dicoccoides had resistance to both leaf rust and Hessian
fly. Only A . squarrosa contained some resistance to greenbug, T . boeoticum had an intermediate response, but all other species were susceptible.
Multiple resistance lines to leaf rust and Hessian fly were identified, but
only A . squarrosa had two lines resistant to all three pests.
Hillel et al. (1973) evaluated A . longissima Schweinf. and A . speltoides
Tausch, two loosely related species that differ in their mating systems, to
assess the direct effect of the mating system on the amount of genetic
variability. They found that for most of the 36 quantitative characters
examined, the differences between populations, the total variances of the
populations, and the mean within-species variances were greater in the
selfer (longissima) than the outbreeder (speltoides). These differences
were attributed to the low probability of a successful gene flow in
A . longissima, with each isolated population adapted to a specific microenvironment.
Ninety-three accessions of cultivated emmer wheat ( T . dicoccum), five
each of two wild tetraploid wheats ( T . dicoccoides)and T . araraticum, and
the cultivated varieties ‘Modoc’ ( T . durum) and ‘Anza’ ( T . aestiuum) were
evaluated for plant height, seed weight, flour protein content, and flour
lysine content, as well as several morphological and grain quality characters by Sharma et al. (1981). Variability among lines for each trait in
different species was significant except plant height in dicoccoides and
araraticum. Primitive and wild wheats were higher in protein and lysine
content, but lower in spike weight and 1000-kernel weight than the two
modern cultivars. It seems that selection for larger kernels has resulted in a
drop in the protein content of seeds from the wild species to the modern
wheats. Lawrence et al. (1958) also found higher protein and lysine contents in wild wheats than in cultivated bread and durum wheats, but only
one accession of T . dicoccoides and T . monococcum L. were analyzed.
Avivi (1978) evaluated 47 samples of T . dicoccoides for protein content in
kernels and found it to be highly variable, ranging from 19% to 28%.
Waines (1983) did a comparative study of primitive diploid wheats such
as T . monococcum with modern polyploid varieties and advocated the
former’s direct usage as commercial varieties. Among the several positive
characteristics of diploid wheats mentioned were less extraneous genetic
material, only seven linkage groups and hence easier to manipulate than
polyploid wheats, greater ecogeographic distribution and hence wider
adaptability, and resistances to pests and diseases. However, Waines
(1983) did not fully evaluate grain quality aspects such as pasta and dough
A. B. DAMANIA
products, which are vital to the success of any wheat variety, although one
may tend to agree with him that exploratory research should be carried out
to see if diploid wheats have a future as commercial varieties.
Hordeum spontaneum C. Koch. is now recognized as the only progenitor of cultivated barley (Harlan, 1979). Wild barleys, H. spontaneum and
H . murinum, and cultivated varieties from Afghanistan were collected by
the Kyoto University Scientific Expedition. The morphological, physiological, and genetic characterization are described by Takahashi et al.
(1965). Resistance to barley powdery mildew was also found in H. spontaneum, but the reaction was classified as being different from that of the
cultivated forms. The overall diversity in this collection was reported to be
low and the wild forms were less variable than the cultivated. Jana et al.
(1987) studied genetic diversity in morphological characters in H. spontaneum and cultivated barleys from the Near East and also found that the
latter were more variable than the former. However, it must be considered
that this wild form in the near East has a long history of survival, has
undergone millenia of natural selection pressures and therefore is better
adapted to harsh environments (if not as variable as the cultivated landraces), and hence is a very valuable genetic resource for abiotic stress
tolerance genes (Grando et al., 1985).
An evaluation of H. spontaneum accessions was also carried out by
Ceccarelli and Grando (1987) to assess the amount of useful genetic variability within this species. The results indicated that the progenitor is a
useful source of genes for a number of economically important characters
for breeding barley in the dry areas. Bakhteyev (1979) evaluated a collection of 77 samples of the same wild species originating from Iran, Iraq, and
Turkey at an experimental farm in the Soviet Union. Considerable variability was observed in this study and the species merits breeders’ interest
especially for adaptation to dry areas.
Landraces of primitive wheats have been evaluated for several economically important traits. Blum et al. (1987) evaluated 13 accessions of
T. compactum Host., a form of wheat cultivated in Syria and Palestine in
the prepottery neolithic period (Renfrew, 1969), for grain quality. It was
found, while comparing these with some modern cultivars of bread and
durum wheat, that the latter were superior in all characters tested except
protein content and quality.
In a comparative study of wild and primitive forms of Triticum, Asins
and Carbonell (1986) provided information for future use on intraspecific
variability differences among species. Robertson et al. (1979) evaluated
the genetic variability in primitive wheats for seedling root numbers, as it is
well known that this character is probably responsible for drought resistance in wheats. Among the 16 species studied, the highest root numbers
EVALUATION OF GENETIC RESOURCES IN CEREALS
were found in T. durum and the lowest in T . araraticum; seed weight was
positively correlated with root number. A test of progenies showed that
this character was stable from one generation to the next over two environments. O’Toole and Bland (1987) have reviewed genotypic variation in
root systems of cultivated wheat and Aegilops.
Electrophoretic techniques also have been applied to the study of variability of storage proteins in wild and primitive forms of wheat (Cole er al.,
1981; Damania er al., 1988) and barley (Jana et al., 1987). Using this
technique, Nevo et al. (1979) showed that there was greater variability in
wild and weedy barleys ( H . sponraneum) collected from Palestine than in a
composite cross of cultivated barley that included more than 6000 varieties
in its parentage. This result was surprising since cultivated barley is conspicuously more variable than its wild and weedy relatives from the Near
East (Harlan, 1984). However, Jana and Pietrzak (1988) reported almost
identical levels of variability between the two in material collected from
Greece, Jordan, Syria, and Turkey.
Nevo et al. (1982, 1988) studied genetic diversity within and between
Turkish populations of wild emmer, T . dicoccoides, utilizing electrophoretic and statistical analysis, and reported that climatic selection
played an important role in genetic differentiation of populations of this
species and that the wild gene pool represents a significant genetic resource for utilization in wheat improvement. In another study of the same
material, Nevo and Payne (1987) reported variability in seed storage
protein and the use of certain high-molecular-weight glutenins for improvement of bread-making qualities of wheat.
Variability for kernel proteins in 841 accessions of T . dicoccoides has
been reported also by Mansur-Vergara et al. (1986) using gel electrophoresis. The protein content measured ranged from 15% to 25% with
some accessions having high protein content as well as large kernel size. A
high percentage of protein content was also reported by Srivastava and
Damania (1989) in dicoccoides accessions from Syria and by Avivi (1978)
in those from Palestine.
In a study of 167 accessions of T . dicoccum from 23 different countries
of origin using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), Vallega and Waines (1987) identified a total of 20
alleles out of which 9 were different from those reported by earlier work.
The newly discovered alleles enhance the genetic variability available to
improve the industrial quality of wheats, and some of them may facilitate
basic research on the relationship of industrial quality with highmolecular-weight glutenin subunit number.
Success in future crop improvement depends largely on the ability to
exploit existing genetic resources in wild relative and primitive forms,
A. B . DAMANIA
especially for less favorable environments. A better utilization and exploitation of these resources requires greater understanding of critical issues
related to evolutionary pathways, geographical distribution, genetic diversity, and multicharacter associations. Variability studies, such as those
described earlier, have provided this vital information. However, much
more work needs to be done before evaluation can be instrumental in
greater use of genes from wild relatives and primitive forms.
VI. DOCUMENTATION OF GENETIC RESOURCES
An efficient system for dissemination of evaluation data on genetic
resources material held in genebanks is essential if it is to be of use to
breeders. Databases of genetic resources information on world collections
have been established and contain a formidable amount of evaluation
information that needs analysis and documentation (Ford-Lloyd, 1978).
For example, 12,129 accessions of barley from a world collection were
evaluated by ICARDA at one location (Somaroo et al. 1986, 1988); there
was significant variability among the landraces for such characters as days
to heading, plant height, 1000-grain weight, protein/lysine ratio, and resistance to diseases.
In recent years much effort has been devoted to making these databases
as comprehensive and mutually compatible as possible. However, less
effort is being channeled toward considering how the results of these
studies might be put to use. This is because most users are interested
primarily in a very restricted aspect of the data (Williams et af., 1980).
Plant breeders may be interested in an immediate problem such as resistance to a particular pathogen or a specific agronomic trait. Alternatively,
their interest may reflect locally prevailing environmental conditions. For
example, breeders at ICARDA are not interested in plants adapted for
favorable conditions, as such environments are not common in West Asia
and North Africa, where the Center operates (Damania and Srivastava,
1989). However, an international genetic resources center might be expected to take a wider view and Williams et al. (1980) contend that genetic
resources evaluation data presently stored in computer memory banks
contain more information of practical value than is immediately apparent
without proper analysis. A detailed study of network analysis of genetic
resources data has been attempted (Williams et af.,1980; Robinson et af.,
1980; Burt et al., 1980).
Internationally agreed descriptor states are not used when evaluation
data are exchanged between genetic resources workers. Considerable time
EVALUATION OF GENETIC RESOURCES IN CEREALS
and effort would be saved if a lengthy description of the method and
intervals used for recording each character is avoided. The IBPGR has
been convening small working groups of scientists for the purpose of
arriving at an internationally agreed upon list of descriptors for describing
information for wheat (IBPGR, 1985) and barley (IBPGR, 1982). However,
experience has shown that it is too time-consuming to record observations
on all descriptors mentioned in the descriptor lists, as the method of
recording and the selection of characters are very much dictated by the
region and the needs of local breeders at each institution concerned. It is
also necessary to set out data in a standard format using a generally agreed
upon series of descriptors and descriptor states for the crop. In this way
the data can be entered into computers, retrieved, and exchanged among
institutions with the least possible confusion and optimum efficiency
For the purpose of utilization, systematic analysis and description of
samples is useful in both distinguishing between populations and identifying duplicates, as well as in providing information on the extent of
variation within a given germplasm collection. It is axiomatic that the more
documentation on a collection, the greater the chance of its rational utilization. Information from the site where a particular sample was collected
may be extremely important. For instance, at ICARDA, germplasm that is
described as having a short maturity period receives immediate attention
of the breeders as this trait is very useful to escape drought and high
temperatures during grain filling in the dry areas. Therefore, information
recorded by germplasm collectors at site would be very valuable later
when the samples are evaluated.
Peeters (1988) studied statistically a large barley germplasm collection
at Cambridge and reported that despite extensive collecting activity in
recent years and subsequent exchange between countries, combinations
of characters have remained substantially different in germplasm by country gene pool. Material from the United States now contains more variability in toto than material from any other country. Subsequently, Peeters
and Martinelli (1989) used hierarchical cluster analysis to classify entries
from this collection according to their degree of similarity and concluded
that this statistical analysis procedure could be used as a tool to classify
entries to their respective gene pools even when no passport data are
Often those responsible for entering data recorded at a collecting site
into a computer data base believe that lengthy descriptive notes made at
the collection site, for example, notes on disease observations or peculiarities of the habitat, are not relevant and hence should be omitted. Nothing
could be more erroneous. Although it is recommended that passport data