Tải bản đầy đủ - 0trang
III. Genetic Mechanisms of Tolerance to Acid Soils
ACID SOIL TOLERANCE IN WHEAT
in a given soil or crop season. Given their physiological independence in wheat,
the genetic mechanisms of A1 and Mn tolerance will be addressed separately in
A. BASISOF ALUMINUMTOLERANCE
The genetic control of Al tolerance in wheat, until recently, was poorly understood and clouded with conflicting evidence of simple vs. complex inheritance.
One early report in wheat indicated that the moderately Al-tolerant cultivar ‘Druchamp’ and the sensitive cultivar ‘Brevor’ differed by a single gene controlling
seedling root growth in a 0.06 mM Al solution (Kerridge and Kronstad, 1968).
Additional genes were postulated for genotypes with tolerance to higher levels of
Al toxicity such as Atlas 66. It is inconceivable that the wide genetic range in A1
tolerance, with varying intermediate degrees (Foy et al., 1965; Lafever et al.,
1977), can be explained on the basis of single gene theory. Yet, single cross populations may segregate for a single gene pair at a prescribed level of stress, as in
the Waalt (moderately A1 tolerant) X Warigal (A1 sensitive) cross examined by
Larkin (1 987) and later verified by Wheeler et al. (1992). The bioassay for measuring tolerance, however, may not have the necessary precision to detect minor
genes when a single gene pair with large effects segregates. Berzonsky (1992)
recognized the potential ambiguity in classifying plants as sensitive and tolerant
based on empirically derived criteria.
The multitude of possible physiological mechanisms of A1 tolerance described
in Section I1 should inspire geneticists to consider multiple genetic mechanisms.
Different genetic systems may operate in seedlings vs. adult plants or at different
levels of Al stress. Different systems may be detected by different bioassays (visual vs. quantitative). Not until genome and chromosome location studies were
performed did it become certain that several genes can influence phenotypic expression of Al tolerance and that these genes may impart effects of varying magnitude with possible interactions.
1. Gene Number and Location
Slootmaker (1974) first attempted to roughly locate genes for acid soil tolerance in wheat by comparing the response of diploid, tetraploid, and hexaploid
species relative to their genome constitution. The various hexaploid genotypes
(AABBDD) had the highest degree of tolerance. The A genome species exceeded
the B genome species but not the tetraploids (AABB) at a lower acidity level. The
importance of the D genome for acid soil tolerance was demonstrated by increased
sensitivity of a tetraploid derivative lacking the D genome from ‘Canthatch,’ a
B. F. CARVER ANID J. D. OWNBY
hexaploid cultivar, and restoration of tolerance in the reconstituted hexaploid by
addition of the D genome from several sources. Even greater tolerance is provided
by the R genome from rye (Secale cereale L.), either by itself or in combination
with durum or hexaploid wheat genomes as hexaploid or octoploid triticales.
Summarizing, the genomes can be ranked in decreasing order of tolerance: R >
D > A > B. It could not be confirmed if the measured responses in acid soil were
induced by Al toxicity alone.
A similar experiment was conducted by Berzonsky and Kimber (1986) using
several diploid, tetraploid, and hexaploid Triticum species, except plants were exposed to 0.44 mM Al in nutrient solutions. Surprisingly, no root regrowth (a measure of tolerance) was observed among diploid A or D genome species classified
as tolerant by Slootmaker ( 1 974). This does not necessarily imply the lack of A1
tolerance genes on A and D genomes of hexaploid wheat; rather, the accessions
sampled from the progenitor species merely lack tolerance. A unique source of
tolerance was found in the tetraploid 7: ventricosum Ces. (Dun genomes) and in
other species sharing the Un genome. Further evidence is needed to verify their
hypothesis that a genetic mechanism different from 7: aestivum exists in T ventricosum. The prospect is encouraging for transferring tolerance from 7: ventricosum
to 7: aestivum via a bridge cross to a tetraploid species. Tetraploid hybrids have
been obtained by crossing 7: turgidum with 7: ventricosum (Maan, 1987).
The genomic location of Al tolerance genes in the tolerant hexaploid cultivar
Atlas 66 indicates that chromosomes other than those of the D genome confer
tolerance (Berzonsky, 1992).Camargo (198 1) had previously concluded that two
dominant genes controlled tolerance in Atlas 66 in relatively low Al concentrations (c0.22 mM Al). A tetraploid derivative of Canthatch was crossed with
Atlas 66, and recombinant inbred lines comprised of D genome chromosomes
exclusively from Atlas 66 were isolated. All lines should show tolerance to
0.44 m M Al (to which Atlas 66 is tolerant) if genes for tolerance are located
strictly on the D genome chromosomes inherited from Atlas 66. However, some
F, lines segregated for sensitive plants and backcrossing to the susceptible hexaploid parent produced an increased proportion of susceptible plants. Therefore,
segregation likely occurred on chromosomes of the A and/or B genomes.
Aluminum tolerance genes were more precisely located on individual chromosomes using aneuploid lines lacking a specific chromosome or chromosome arm.
This method lacks some appeal for locating Al tolerance genes because cultivars
for which aneuploid stocks are available do not possess desirable levels of tolerance for genetic improvement. Surprisingly, several genes exist in Chinese
Spring for a cultivar generally regarded considerably less tolerant than Atlas 66
(Table 111). The B genome is the least important source of tolerance genes, at least
for Chinese Spring, while the most frequent source of tolerance genes is the D
genome. Tolerance genes have consistently been identified on chromosomes 2DL
ACID SOIL ‘TOLERANCE IN WHEAT
Chromosomal Location of Genes ControllingA1 Tolerance
in the Moderately Tolerant Cultivar, Chinese Spring
2L, 3L, 4L, 7
Takagi et al. (1983)
Aniol and Gustafson (1984)
Aniol ( 1990)
4L,” 6L, 7s
“Originally reported as 4BL.
“Original citation indicated larger effect on tolerance (location of “major”
‘ Suppresses Al tolerance of Chinese Spring.
and 4DL. All loci identified confer tolerance except one, 6BS, which in one report
suppressed tolerance in Chinese Spring.
2. Gene Expression and Heritability
Gene expression at one locus may not be independent of the gene(s) present at
other loci, particularly when combined from different species or genera. For example, gene expression of A1 tolerance in wheat is altered in the presence of genes
from rye and vice versa. Genes located on chromosomes 3R, 4R, and 6R confer
Al tolerance, but the level of tolerance is greatly reduced when added to a wheat
background (Aniol and Gustafson, 1984). Apparently, certain wheat genes suppress the expression of Al tolerance genes from rye, yet others allow expression
of rye Al tolerance (Table IV). Most chromosome arms of wheat which control
Al tolerance in wheat also activate or suppress Al tolerance of rye. On the other
hand, some wheat chromosomes may not confer Al tolerance in wheat, at least in
Chinese Spring, but influence rye gene expression.
Gene interactions are not restricted to wheat-rye combinations. Epistatic interactions are implied when segregation in susceptible X tolerant parent crosses
does not fit a simple additive-dominance model. Evidence of epistasis, however,
does not surface with discrete classification of segregating progeny. Discrete classification of A1 responses may be an oversimplistic description of a complex segregation pattern. The traditional bioassay of root staining with hematoxylin is
commonly used for the discrete classification of A1 tolerance (Polle et al., 1978),
but genotypic differences in stainability may be more aptly described using a
quantitative scale. Further, the inheritance of A1 tolerance is usually determined
from F2 populations instead of by more sophisticated mating designs needed to
B. F. CARVER AND J. D. OWNBY
Chromosomal Location of Wheat Genes Controlling
A1 Tolerance in Chinese Spring Wheat and/or Blanco Rye
in wheat 0
aAs reported by Anion and Gustafson (1984) and Aniol
(1990) for Chinese Spring.
bAs reported by Gustafson and Ross (1990) for Blanco rye.
PChromosomearm not reported.
detect gene interactions. Such was the case for the cross, ‘Cardinal’ (A1 tolerant) x ‘GK Zombor’ (A1 susceptible), in which epistatic effects at two loci were
hypothesized based on root length measurements in nutrient solutions (Bona
et al., 1994). Inheritance of root length in acidic soil was also not monogenic.
Although Al tolerance in wheat is generally regarded as a dominant trait, the
importance of additive gene action has also been emphasized (Aniol, 1984b;
Campbell and Lafever, 1981; Ruiz-Torres and Carver, 1992; Bona et al., 1994).
The relative magnitude of dominance effects is further complicated by changes in
magnitude and direction of dominance for different allelic combinations. Two
crosses examined by Bona et al. (1994) represented a wide range in degree of
dominance for root length in acidic soil. Dominance was complete in one cross
(Cardinal X Zombor) but absent in another (‘Becker’ X ‘GK Kincso’). Using a
ACID SOIL TOLERANCE IN w
Relative Root Length" of Four Tolerant and Three Susceptible Wheat
Parents and Their F , Progeny Averaged across Two Al Concentrations",'
Atlas 66 (83%)
Susceptible" parent (and RRL) (%)
83 ( + ) ''
77 ( + I
79 ( + )
I 3 (+)
68 ( + )
74 ( + )
"RRL, c/r of root length at 0 mM Al.
"0.36 and 0.72 mM Al.
I Data taken from Ruiz-Torres and Carver ( 1992).
"Classification based on hematoxylin staining of seedling roots (Carver el d ,
''Indicates positive (+) or negative ( - ) heterosis of F, relative to midparent; no
sign indicates F , was not distinguished from midparent.
factorial series of tolerant X susceptible crosses, Ruiz-Torres and Carver (1 992)
examined the consistency in gene expression for Al tolerance in a series of susceptible backgrounds. The genotype of the susceptible parent influenced the degree of tolerance expressed in the F,. For example, the phenotype of the F, resembled either the tolerant parent or the susceptible parent when Atlas 66 was
crossed with either 'Chisholm' or 'Century,' two susceptible parents (Table V). In
only three crosses, the hybrid phenotype was intermediate to the two parents. In
most crosses where heterosis occurred, the hybrid resembled the tolerant parent
in relative root length.
This variation in heterotic pattern is not indicative of a single recessive gene governing sensitivity (Kerridge and Kronstad, 1968; Lafever and Campbell, I978),
but may reflect allelic variation among susceptible parents (or variation in closely
linked loci) since Chisholm has slightly more field tolerance in acid soils than
Century (Carver et ul., 1993). Allelic variation at a single locus controlling A1
tolerance has been observed in Hordeum vulgare L. (Minella and Sorrells, 1992)
and Zea mays L. (Rhue et af., 1978). Variation in heterotic pattern may also serve
as evidence of additive X additive interactions, with or without dominance.
Greater consistency was found for the F, phenotype relative to the midparent in a
factorial series of soft wheat crosses (Campbell and Lafever, 1981). but their results are not directly comparable to those in Table V because lower Al concentrations were used in reporting root length per se (unadjusted for root growth in the
absence of Al stress).
B. F. CARVER AND J. D. OWNBY
Figure 2. Frequency of tolerant seedling plants detected by hematoxylin staining in two F2wheat
populations at five Al concentrations ranging from 0.09 to 0.90 mM in solution culture. The hypothesized segregation ratio of tolerant: susceptible plants (or tolerant:intermediate:susceptible plants) is
given for each Al concentration. Gene expression is influenced by the concentration of A1 in the
nutrient solution and the unique combination of parents. Data taken from Bona er al. (1994).
The genetic expression of A1 tolerance is also influenced by the severity of A1
stress. Camargo (1981) noticed different inheritance patterns at different A1 concentrations,and Campbell and Lafever ( 1981) recognized the importance of characterizing A1 tolerance at more than one A1 concentration. More specifically,
Aniol(1984b) observed a significant decrease (more than one-half) in the proportion of tolerant F3 plants from susceptible X tolerant crosses, as A1 concentration
increased from 0.30 to 0.59 mM Al. Similarly, Bona ef al. (1994) observed a
major break in the proportion of tolerant plants between 0.36 and 0.72 mM Al,
using hematoxylin staining to classify segregating F2 plants as tolerant, intermediate, and susceptible (Fig. 2). A single genetic model did not consistentlyexplain
the inheritance of A1 tolerance across A1 concentrations in one population. In
another population, inheritance was monogenic, but as A1 stress increased, the
direction of dominance changed from positive to negative, with no dominance at
an intermediate A1 concentration. Minella and Sorrells (1992) also observed a
change in the direction of dominance with increasing A1 stress in barley; this pattern has been reported for other types of stress resistance in wheat (Sutka and
Veisz, 1988). Either expressivity of tolerance genes decreases at higher A1 stress
or different levels of A1 stress induce expression of different gene systems. Obviously some lethal stress level exists at which expressivity ultimately dissipates;
ACID SOIL TOLERANCE IN WHEAT
the expression of different gene systems at different sublethal concentrations,
however, would be of practical value in selection.
Direct evidence of this phenomenon is provided by localization of tolerance
genes to specific chromosomes at various Al concentrations (Aniol, 1990). The
gene(s) located on chromosome 5AS of Chinese Spring was expressed in the
range of 0.037 to 0.074 mM Al, whereas genes located on 2DL and 4DL were
expressed only at 0.074 mM. Aniol ( 1 990) speculated that genes located on 5AS,
2DL, and 4DL control Al uptake, but 5AS may also regulate detoxification of A1
inside the root tissue. This study also provided evidence that different genes may
impart different effects (major vs. minor) because aneuploid lines which lacked
5AS were classified susceptible even though tolerance genes on 2DL and 4DL
were present. Genetic mechanisms of Al tolerance in wheat must be defined with
reference to the level of A1 stress applied and the parental combinations used to
develop experimental populations.
The extent to which the Al tolerance phenotype is determined by genotype is
well documented in wheat (Table VI). Estimates of heritability vary widely depending on the method of estimation and pedigree of the population. This further
illustrates the genetic complexity of A1 tolerance in wheat and the futility in adopting a simple genetic model to uniformly describe it. Nonadditive gene action,
particularly dominance, plays a major role in gene expression, but its value may
be limited to F, hybrid breeding systems unless additive epistatic interactions can
be fixed in a homozygous genotype. The potential role of epistasis warrants further investigation.
B. BASISOF MANGANESE
The genetic literature is much less developed for Mn tolerance of wheat than
for A1 tolerance. There are several reasons for this information gap. The perception that Mn toxicity is less severe than Al toxicity and may be limited to poorly
aerated soils has resulted in less concern to characterize genetic variation in Mn
tolerance. Plants may develop symptoms of Mn toxicity (chlorotic, erect, or brittle
leaves) with no obvious reduction in vegetative growth (Foy et al., 1988). The
toxic effect of Al, on the other hand, can be observed much earlier in plant development due to immediate root damage and subsequent retardation of vegetative growth. Fisher and Scott (1993a) questioned whether Mn tolerance is a
worthy wheat-breeding objective for southern New South Wales, given the neutral
effect on grain yield and the difficulty in screening for Mn tolerance. Experiments designed to assess Mn toxicity often require a treatment period of several
weeks before assessing plant damage (usually shoot and/or root yields). The lack
of a rapid bioassay, which uses a readily identifiable characteristic(s) of Mn tolerance, has probably contributed to the information gap as well. Large germ plasm
Estimation of Heritability for Al Tolerancein Nutrient Solutions (Refs. 1-4) or Acidic Soils (Ref. 5)
Relative root length
Relative root length
Relative root length
Method of estimation
Correlation of F, and F2
Correlation of midparent and F,
Variance components: General
Correlation of midparent and F
Variance components in F2, BC
Variance components in F,, BC
Variance components in F,
"Lower concentration applied for longer duration ( 1 -8 days).
' ( I ) Lafever and Campbell (1978). (2) Campbell and Lafever (1981 ), (3)Ruiz-Torres and Carver (1992). (4) information reported in Proc.Australian Plant Breeding
Conf., loth, Gold Coast, Australia, Vol. 2, pp. 78-79, and (5) Bona et al. (1944).
' Considered quasi-estimate over several crosses.
ACID SOIL TOLERANCE IN WHEAT
collections have not been screened to identify Mn-tolerant parents for selection
programs or for constructing experimental populations. These populations, in
turn, are needed to characterize inheritance of Mn tolerance.
The largest collection screened thus far (Macfie er al., 1989) was composed of
30 genotypes primarily of Canadian origin previously assembled for A1 tolerance
screening (Briggs et al., 1989). Genotypes showed a wide and continuous tolerance range to a 14-day treatment of 0.50 mM Mn (above basal level), based on
relative root and shoot weights, The lack of distinct genotypic differences indicated that several genes control Mn tolerance. Whether these differences reflect
the cumulative action of several genes, or possibly a few genes with large environmental effects, is still uncertain. The Mn treatment reduced root weight more than
shoot weight relative to the corresponding control weights (39% vs. 68%, averaged across 30 genotypes), but differences in relative root weight were highly
consistent with differences in shoot weight ( r = 0.88). Consistency in root and
shoot weights after excess Mn treatment has been reported for a smaller set of
genotypes of Brazilian and Australian origin (Burke et al., 1990), with greater
biomass reduction occurring also in roots. These results suggest that a common
gene system controls, at least in part, Mn tolerance in both root and shoot tissue.
Foy et al. (1988) also recognized the continuous and wide variability for Mn
tolerance in wheat, with the qualification that a few genes may explain the majority of the variation. This was demonstrated by backcrossing Mn tolerance from
Carazinho to Egret; yet, transfer of tolerance was incomplete. Cultivars shown to
have exceptional levels of Mn tolerance are the Brazilian cultivar Carazinho, the
Australian cultivar Warigal, and the Canadian cultivar ‘Norquay’ (Macfie er al.,
1989; Burke et al., 1990). All are classified as spring types. Manganese tolerance
has not been widely surveyed in winter wheat.
Depending on soil classification and pH, a wheat crop may experience both A1
and Mn toxicity over the course of a season. This possibility raises an important
question relevant to wheat improvement: To what extent is tolerance to either
factor genetically related? Notwithstanding the possibility of physical linkage of
genes controlling tolerance, do genes which control A1 tolerance also influence
Mn tolerance by pleiotropy? It is plausible that plant defense mechanisms may
operate against both elements when present in toxic concentrations (e.g., exclusion, compartmentation, or detoxification). One line of indirect evidence suggests that genetic control is dichotomous. Accumulation of unusually high leaf
tissue concentrations of Mn in some genotypes, without significant loss in dry
matter yield, is a unique mechanism of Mn toleraiice which appears independently
B. F. CARVER AND J. D. OWNBY
Cultivar comparisons also offer indirect evidence that different gene systems
influence A1 and Mn tolerance (Neenan, 1960; Foy et al., 1973; Burke et al.
1990), but the converse is not necessarily proven, i.e., that no gene(s) exists which
coregulates tolerance. Several examples can be cited to support differential control. Atlas 66 and BHI 146 are tolerant to A1 but sensitive to Mn toxicity, each to
various degrees; likewise, Monon and Warigal are A1 sensitive but Mn tolerant.
However, examples of concurrent tolerance can also be cited. Carazinho and Norquay are tolerant to both A1 and Mn. Obviously, all combinations exist, which
may lead to different conclusions depending on which genotypes are sampled.
Macfie et al. ( I 989) quantified the phenotypic relationship across 29 cultivars
using relative root weight as an indicator of each tolerance. As might be expected,
the correlation was positive but intermediate ( r = 0.57). Any genetic interpretation of a correlation estimated in this manner is not advised because it may reflect
simultaneous selection pressures (inadvertent or intentional) during cultivar development, particularly if field testing occurred where A1 and Mn toxicities coexisted. A direct approach to ascertain a genetic relationship would be to apply
selection pressure for only one element (A1 or Mn) and examine the correlated
response in tolerance to the other element. Fisher and Scott (1993a) generated
pairs of closely related lines differing for Mn tolerance but uniform for A1 tolerance, implying that tolerance to A1 and Mn was inherited independently from the
original parent, Carazinho.
Limited data suggest that selection should focus on both A1 and Mn tolerance
if tolerance to acid soils is to be fully realized. Aluminum-tolerant genotypes
might be identified first to reduce the number of genotypes to a manageable level
before using more tedious procedures to screen for Mn tolerance.
IV. BREEDING FOR ACID SOIL TOLERANCE
Genetic improvement of acid soil tolerance in wheat must be economically and
biologically reasonable, as mandated for any plant-breeding objective (Simmonds,
1979, p. 29). While economic justification is not widely documented, the monetary benefit of acid soil tolerance measured in southern New South Wales (Fisher
and Scott, 1993a) may be extended to other wheatland soils undergoing gradual
acidification, like the southern Great Plains of the United States and the summer
rainfall region of South Africa. Economic rationale for breeding acid soil tolerance can be predicated on other terms than monetary, as discussed in this section.
The extent to which a useful level of heritable variation exists for components
of acid soil tolerance and the availability of practical techniques to identify tolerance or susceptibility largely determine whether breeding for acid soil tolerance in wheat is biologically reasonable. Still, others may argue that short-term
ACID SOIL TOLERANCE IN WHEAT
genetic approaches to long-term environmental problems are unjustified. While
the “Adapt-a-Plant” philosophy embraced by C. D. Foy in Kaplan (1989) is difficult to refute, sustainability of wheat production is eventually threatened as
acidification increases with time, to the point that nutritional disorders in addition
to Al and Mn toxicity could conceivably render genetic tolerance ineffective.
A. WHENIs GENETIC
Genetic improvement is most often justified where corrective lime applications
are ineffective or impractical in acidic subsoil layers (Foy et al., 1965; Aniol and
Kaczkowski, 1979). Lime distribution below the surface layer is not impossible
but is generally cost prohibitive. Liming the surface layer may offer a partial yield
benefit, but genetic potential is still not reached if the root system does not penetrate the acidic subsoil. A wheat cultivar with improved tolerance may tap critical
water and nutrient supplies below the surface layer until subsoil pH declines to
levels which exceed the tolerance range of the genotype. The level at which pH
stabilizes in the subsurface layers primarily depends on mineralogy of a particular
soil, but pH (H,O basis) generally can be as low as 3 (Baas Becking et al., 1960).
This lower pH limit corresponds to the hydrolysis of ferrous oxide minerals which
are the ultimate product of mineral weathering.
Genetic improvement of acid soil tolerance may also be justified even when soil
acidity is ameliorated by surface applications of lime, as for some soils in southern
and western Australia and the southern Great Plains of the United States (Westerman, 1981 ; Dolling er al., 199 1). Fisher and Scott (1 993a) estimate that the current benefit of Al tolerance to grain yield in the southern NSW wheat belt is l .4%
and could increase to 3.2% if their soils acidify at the current rate over the next
10 years. No benefit was expected above pHca 4.4 ( 1 : 2 soil:O.Ol M CaCl,), and
no apparent yield benefit was found with Mn tolerance in their soils. The economic benefit to grain yield derived from breeding for disease resistance still exceeded that of breeding for Al tolerance.
The relative importance of acid soil tolerance is magnified if the economic
benefit of increased forage production is also considered. Unfortunately, the effects on forage production have not been widely researched. A soil pH below 5.0
(H,O basis) in Oklahoma reduces forage production more severely than it reduces
grain yield (E. G. Krenzer, Jr., 1994, personal communication). Preliminary data
collected for hard winter wheat cultivars indicate that acid soil tolerance can dramatically improve potential grazing capacity (Fig. 3), particularly during early
vegetative growth (prior to winter dormancy) when Al toxicity stress on shoot
growth is most noticeable. As soil pH drops from 5.0 to c4.5, early season forage
growth may be reduced by as much as 85%, hardly enough to support a wheat
pasture program. Because even the tolerant genotype suffers some forage yield