Effects of Zinc Deficiency on Rice Growth and Genetic
Factors Contributing to Tolerance
Matthias Wissuwa*, Abdelbagi M. Ismail, and Seiji Yanagihara
Japan International Research Center for Agricultural Sciences, Crop Production and Environment Division,
Tsukuba, Ibaraki 305–8686, Japan (M.W.); Crop and Environmental Sciences Division, International Rice
Research Institute, Metro Manila, Philippines (A.M.I.); and International Research Division, Agriculture,
Forestry and Fisheries Research Council Secretariat, Ministry of Agriculture, Forestry and Fisheries,
Chiyoda-ku, Tokyo 100–8950, Japan (S.Y.)
Zinc (Zn) deficiency is the most widespread micronutrient disorder in rice (Oryza sativa), but efforts to develop cultivars with
improved tolerance have been hampered by insufficient understanding of genetic factors contributing to tolerance. The
objective of this paper was to examine alternative evaluation methods and to identify the most informative traits that would
provide realistic information for rice breeders and to map quantitative trait loci (QTLs) associated with tolerance. Screening
experiments in low-Zn nutrient solution and in a Zn-deficient field did not produce similar tolerance rankings in a set of
segregating lines, which suggested that rhizosphere effects were of greater importance for lowland rice than internal Zn
efficiency. The most severe symptom in the field was high plant mortality. The occurrence of leaf bronzing, usually regarded as
indicative of susceptibility, did not necessarily concur with high plant mortality, which implied that both were under
independent genetic control. The QTL mapping experiment conducted in the field with a population derived from a cross of
IR74 (intolerant) with Jalmagna (tolerant) largely confirmed this. Four QTLs associated with plant mortality were detected, and
only one of those colocalized with one of the four QTLs detected for leaf bronzing. The two most influential QTLs for plant
mortality were detected on chromosomes 2 and 12. They explained 16.6% and 24.2% of the variation, and alleles of the tolerant
donor parent Jalmagna reduced mortality by 16.6% and 14.8%, respectively. QTLs for plant mortality acted in a purely additive
manner, whereas digenic epistatic interactions were important for leaf bronzing.
Zinc (Zn) deficiency was first diagnosed in rice
(Oryza sativa) on calcareous soils of northern India
(Nene, 1966; Yoshida and Tanaka, 1969). It was subsequently found to be a widespread phenomenon in
lowland rice areas of Asia, and, next to nitrogen (N)
and phosphorus (P) deficiency, Zn deficiency is now
considered the most widespread nutrient disorder in
lowland rice (Neue and Lantin, 1994; Quijano-Guerta
et al., 2002). Zn deficiency can be corrected by adding
Zn compounds to the soil or plant, but the high cost
associated with applying Zn fertilizers in sufficient
quantities to overcome Zn deficiency places considerable burden on resource-poor farmers and it has
therefore been suggested that breeding efforts should
be intensified to improve the tolerance to Zn deficiency in rice cultivars (Quijano-Guerta et al., 2002;
Singh et al., 2003).
Zn deficiency causes multiple symptoms that usually appear 2 to 3 weeks after transplanting (WAT) rice
seedlings; leaves develop brown blotches and streaks
that may fuse to cover older leaves entirely, plants
* Corresponding author; e-mail [email protected]; fax 81–29–
838–6354.
The author responsible for distribution of materials integral to the
findings presented in this article in accordance with the policy
described in the Instructions for Authors (www.plantphysiol.org) is:
Matthias Wissuwa ([email protected]).
www.plantphysiol.org/cgi/doi/10.1104/pp.106.085225
remain stunted and in severe cases may die, while
those that recover will show substantial delay in
maturity and reduction in yield (Yoshida and Tanaka,
1969; Van Breemen and Castro, 1980; Neue and Lantin,
1994). One major obstacle to the improvement of
tolerance to Zn deficiency in rice is that it is fully
understood neither what causes these symptoms nor
what makes plants tolerant. Zn deficiency has been
associated with a wide range of soil conditions: high
pH (.7.0), low available Zn content, prolonged submergence and low redox potential, high organic matter and bicarbonate content, high magnesium (Mg) to
calcium (Ca) ratio, and high available P (for review, see
Neue and Lantin, 1994). High soil pH appears to be the
main factor associated with the widespread Zn deficiency in the calcareous soils of the Indo-Gangetic
plains of India and Pakistan (Qadar, 2002), whereas
perennial wetness is the major cause for Zn deficiency
on peat soils and in coastal saline soils (Neue and
Lantin, 1994; Quijano-Guerta et al., 2002). On our experimental site, located close to the International Rice
Research Institute (IRRI) in the Philippines, Van Breemen and Castro (1980) associated Zn deficiency with
low redox potential due to perennial flooding and a
high organic matter content. Both factors were thought
to act more or less independently. Scharpenseel et al.
(1983) reexamined this evidence and found high correlations between Zn deficiency and soil bicarbonate and
Mg contents but not with organic matter.
Plant Physiology, October 2006, Vol. 142, pp. 731–741, www.plantphysiol.org Ó 2006 American Society of Plant Biologists
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2006 American Society of Plant Biologists. All rights reserved.
731
Wissuwa et al.
Table I. Average (SD) for plant growth and Zn nutritional status of a set of 10 genotypes as affected by Zn deficiency
Genotypes were grown in a Zn-deficient field plot without Zn fertilizer (minus Zn) and in a neighboring control plot that received an annual Zn
application of 20 kg Zn ha21 as ZnSO4 (plus Zn). Data are from the first field experiment in the dry season 2003.
Height
TDM
cm
g plant21
Plus Zn
69.4 (12.1)
9.61 (1.3)
Minus Zn
33.4 (9.3)
1.93 (1.4)
Effect of Zn
251.9
280.1
deficiency (%)
Shoot Zn Concentration
Root Zn Concentration
mg g21
mg g21
43.7 (6.5)
10.7 (1.9)
275.6
122.1 (40.7)
27.5 (2.9)
277.5
Zn Content
Plant Mortality
Leaf Bronzinga
mg plant21
%
Score
522.3 (119.6)
6.4 (2.8)
32.1 (22.8)
69.2 (19.2)
293.9
162.8
0 (0)
5.3 (2.2)
a
Phenotypic evaluations on a scale from 0 (no bronzing) to 9 (severe bronzing).
Numerous studies investigated potential mechanisms for tolerance to Zn deficiency in rice. Tolerant
cultivars may have lower Zn requirements or translocate relatively more Zn from roots to shoots (Cayton
et al., 1985). However, nutrient balances within plant
shoots, such as iron (Fe) to Zn, Mg to Zn, P to Zn,
manganese (Mn) to Zn, and copper (Cu) to Zn ratios,
appear to be more important than tissue-Zn concentrations (Cayton et al., 1985; Qadar, 2002). It was concluded that high translocation of Zn to shoots and
reduced translocation of Fe, Mg, P, Mn, and Cu would
be an important tolerance mechanism. How these
nutrient imbalances affect plant growth remains unresolved, but one likely explanation is a disturbance of
enzyme functions (Neue and Lantin, 1994; Cakmak
et al., 1997). Studies on genotypic differences in the
ability to increase Zn availability in the rhizosphere for
subsequent uptake have focused on the active release
of Zn-mobilizing substances from rice roots. Zhang
et al. (1998) detected phytosiderophores in root exudates of rice, and Hoffland et al. (2006) found higher
rates of organic acid excretion in genotypes tolerant to
Zn deficiency. Yet another interpretation of cause-andeffect mechanisms under Zn deficiency focuses on the
negative effect of secondary stress factors, such as high
bicarbonate concentrations in the soil solution, on root
growth with subsequent negative effects on Zn acquisition. (Yang et al., 1994; Hajiboland et al., 2003).
Hacisalihoglu and Kochian (2003) reviewed the evidence from studies conducted in a range of crops and
concluded that efficient Zn utilization in the shoot was
of higher importance than rhizosphere processes. This
conclusion was partly based on the observation that
tolerance rankings obtained in field trials and in nutrient solution screens were generally in good agreement. These partly contradictory findings highlight
the need to reinvestigate the reasons for genotypic
differences in tolerance to Zn deficiency in rice.
Several factors could be responsible for the poor
understanding of tolerance to Zn deficiency in rice. As
stress factors associated with Zn deficiency vary between soil types (high pH on calcareous soils; low
redox potential on perennially flooded soils), it is
likely that genotypes require a different set of tolerance mechanisms to overcome the site-specific stress
factors. Furthermore, most studies examined differences between only two contrasting genotypes and
focused on a single tolerance mechanism rather than
on interactions between multiple factors. Mapping
quantitative trait loci (QTLs) for tolerance to Zn deficiency is one approach to overcome these limitations.
By identifying QTLs associated with symptoms of Zn
deficiency, it will ultimately be possible to dissect the
Figure 1. Effect of Zn deficiency on TDM (A) and plant height (B) in a
set of 10 genotypes. Data are from the first field experiment in the dry
season 2003. Horizontal bars represent SEs of genotypic means.
732
Plant Physiol. Vol. 142, 2006
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2006 American Society of Plant Biologists. All rights reserved.
Effects of Zinc Deficiency on Rice Growth
RESULTS
Field Experiment with a Set of 10 Genotypes
Figure 2. Genotypic differences in plant mortality (A) and leaf bronzing
(B) as a result of Zn deficiency. Data are from the first field experiment in
the dry season 2003. Horizontal bars represent SEs of genotypic means.
overall response to Zn deficiency into distinct genetic
factors, each associated with a physiological mechanism conferring tolerance.
In initial screening trials in a Zn-deficient field plot,
parents of an existing mapping population were found
to differ in tolerance to Zn deficiency (Hoffland et al.,
2006). The traditional variety, Jalmagna, was considered tolerant, whereas the modern variety, IR74, was
highly susceptible. In this study, the response to Zn
deficiency of both parents was further characterized
together with a small subset of recombinant inbred
lines (RILs) from the mapping population to: (1)
confirm the suitability of this population for mapping
QTLs associated with tolerance to Zn deficiency; (2)
test the hypothesis that efficient Zn utilization in the
shoot is the most important factor and that screens in
low-Zn nutrient solutions would therefore yield similar results to field test as observed with some upland
crops; and (3) design and conduct a QTL mapping
experiment based on conclusions drawn from experiments conducted under objectives 1 and 2.
The 10 genotypes used in this study were the
parents of a QTL mapping population (Jalmagna and
IR74), six RILs (474, 507, 512, 597, 614, and 639) from
that mapping population, and the two unrelated genotypes Dular (intolerant control) and M79 (tolerant
control). Plants in the unfertilized treatment plots
showed typical symptoms of Zn deficiency, such as
stunting, reduction in total dry matter (TDM), high
plant mortality, and severe leaf bronzing (Table I). That
these symptoms were indeed due to Zn deficiency was
evident from a reduction in Zn content per plant of
more than 90% relative to the fertilized control. Low
plant mortality rates and the absence of leaf bronzing
in the fertilized control further showed that Zn deficiency was the main factor responsible for the observed phenotype. While Table I gives averages over
all 10 genotypes, Figures 1 and 2 and Table II show
phenotypic values per genotype of that same experiment. Genotypes studied differed in growth habit with
Dular and Jalmagna displaying the tall plant type
typical of traditional varieties (Fig. 1). Both also had
the highest TDM in the control plot, but they differed
considerably in their response to Zn deficiency. Dular
was one of the most intolerant genotypes in this field
trial with regard to reductions in plant height and
TDM of 70% and 95%, respectively. Other highly intolerant genotypes include IR74, 512, and 597. Jalmagna
together with M79 remained the most vigorous genotype under Zn deficiency in absolute terms, but 507
was more tolerant if minimal reductions in height and
TDM relative to the plus Zn control are used to
evaluate tolerance.
Genotypic variation for seed Zn content, Zn
translocation from root to shoot, and for shoot Zn
Table II. Zn content in seeds used for experiments and Zn
concentrations and content of 10 genotypes grown in a
Zn-deficient field
Data are from the first field experiment in the dry season 2003.
Genotype
M 79
Jalmagna
507
474
614
639
512
597
IR74
Dular
HSDa
Seed Zn
Shoot Zn
Total Plant Relative Zn
Zn
Content Concentration Zn Content
Content Translocationb
mg seed21
mg g21
mg plant21
%
%
0.7
0.72
0.58
0.55
0.55
0.62
0.6
0.79
0.47
0.77
0.12
11.4
8.5
11.7
10.4
10.8
11.5
10.8
10.8
11.5
9.3
2.7
65.2
50.2
66.5
33.6
37.2
29.5
8.6
7.4
12.0
10.5
24.9
14.6
11.2
18.9
8.1
7.4
5.9
2.2
1.6
4.0
2.3
4.0
60.8
43.6
44.7
48.5
56.9
48.9
52.0
52.3
49.1
52.2
13.0
a
Tukey’s HSD (P , 0.05) for the comparison of genotypic
b
Zn translocation was calculated as the proportion of total
means.
plant Zn found in the shoot.
Plant Physiol. Vol. 142, 2006
733
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2006 American Society of Plant Biologists. All rights reserved.
Wissuwa et al.
Table III. Correlation coefficients for the association between traits potentially associated
with tolerance to Zn deficiency
Data are from the first field experiment in the dry season 2003. * and **, Significant at P , 0.05 and 0.01,
respectively.
TDM
(per Row)
TDM
(per Plant)
Leaf
Bronzing
Zn Content
(per Plant)
Shoot Zn
Concentration
Root Zn
Concentration
Mortality
20.72**
TDM (per row)
TDM (per plant)
Leaf bronzing
Zn content (per plant)
Shoot Zn concentration
20.48**
0.83**
0.29
20.59**
20.56**
20.38*
0.48**
0.93**
20.23
0.07
20.37*
20.38*
0.43*
20.02
0.01
20.18
20.29
0.44*
0.07
0.57**
Traits
concentrations was low and not consistently associated with tolerance to Zn deficiency (Table II). Zn
deficiency reduced shoot Zn concentrations on average to around 10.7 mg g21, which is well below the range
of 15 to 20 mg g21 considered critical for lowland rice
(Dobermann and Fairhurst, 2000). The two landraces,
Jalmagna and Dular, had the lowest shoot Zn concentrations and therefore appear to be more efficient in
producing dry matter with limited Zn compared to
modern varieties. However, both differed dramatically
in the extent of leaf bronzing and total Zn content;
higher internal efficiency was therefore not an important trait for tolerance to Zn deficiency under the experimental conditions. Far more important was the
ability to maintain higher Zn content under Zn deficiency, both in absolute terms and relative to Zn
content in the plus Zn control.
The most severe effect of Zn deficiency was a high
plant mortality rate that could exceed 90% in the
intolerant genotypes (Fig. 2). Affected seedlings usually died within 5 to 7 WAT. The tolerance ranking of
genotypes based on mortality generally agreed well
with the ranking for stunting and TDM, with 507 being
the most tolerant and 597, 512, and IR74 being the most
sensitive. However, the classification of genotypes as
tolerant based on mortality, stunting, and TDM deviated from classifications for the leaf-bronzing score.
Several highly intolerant genotypes (Dular, IR74, and
597) did exhibit a high degree of leaf bronzing, but this
was also observed for the otherwise tolerant line M79
(Fig. 2). Other mismatches included lines 614 (severe
leaf bronzing despite relatively high TDM) and 512
(little leaf bronzing but high mortality and low TDM).
Correlation coefficients between measured traits
were determined to further examine the relation between plant growth parameters and plant Zn status
(Table III). Plant mortality was not significantly correlated with Zn concentrations in shoots or roots, but
TDM per plant and total plant Zn content were very
tightly associated, confirming that genotypic differences in Zn acquisition were more important than
genotypic differences in internal Zn efficiency. Correlations between shoot Zn concentration and leaf
bronzing were positive, which suggested that intolerant plants (more leaf bronzing) had higher Zn concentration in shoot tissue. This is surprising and may
indicate that leaf symptoms were caused by additional
stress factors that become important at low shoot Zn
concentrations but not when the internal Zn status of
plants is adequate, as in the plus Zn control.
Data presented in Table III were based on plant
samples taken 12 WAT, when most intolerant plants
had died. However, it should be more informative to
relate the Zn status of plants sampled at the beginning
of the critical stage for survival to subsequent appearance and mortality, and for that reason plants were
sampled twice in a follow-up experiment. The first
sampling was made 5 WAT when symptoms of Zn
deficiency were visible but mortality was low (,10%).
A second sample was taken 10 WAT when average
mortality had reached 62%. Correlation coefficients
between tolerance parameters measured at 10 WAT
Table IV. Correlation coefficients for the association between traits measured 5 WAT and parameters associated with tolerance to Zn
deficiency at 10 WAT
Data are from a second field experiment in the wet season 2003. * and **, Significant at P , 0.05 and 0.01, respectively. ns, Not significant. Traits
measured at 5 WAT are in bold (column 1), and traits measured 10 WAT are in normal type.
Traits (5 WAT)
Survival
TDM (per Row)
TDM (per Plant)
Leaf Bronzing
Zn Content (per Plant)
Shoot Zn Concentration
RDM (per plant)
ShDM (per plant)
Root Zn concentration
Shoot Zn concentration
Zn content (per plant)
Zn translocationa
Root:shoot ratio
0.50**
0.51**
0.09 ns
20.21 ns
0.48**
20.09 ns
20.12 ns
0.56**
0.59**
0.18 ns
20.46**
0.55**
20.19 ns
20.11 ns
0.50**
0.53**
0.26 ns
20.40**
0.52**
20.21 ns
20.11 ns
20.16 ns
0.18 ns
20.27 ns
0.49**
20.11 ns
0.35*
0.06 ns
0.38*
0.09 ns
0.06 ns
20.17 ns
0.20 ns
20.38*
0.51**
20.32*
20.48**
20.15 ns
0.37*
20.35*
0.01 ns
0.31*
a
Zn translocation was calculated as the proportion of total Zn found in the shoot.
734
Plant Physiol. Vol. 142, 2006
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2006 American Society of Plant Biologists. All rights reserved.
Effects of Zinc Deficiency on Rice Growth
Figure 3. Dry weight accumulation of 10 rice genotypes grown in
chelator-buffered nutrient solution containing very low (0.1 mM Zn) or
adequate (2.0 mM Zn) levels of total Zn. Excess chelators (HEDTA) reduced
the concentration of free Zn to 7.75 E210 and 3.15 E211, respectively.
(leaf symptoms, survival, TDM) and several traits
measured at 5 WAT are shown in Table IV. Results
largely confirmed earlier observations. High shoot Zn
concentrations at 5 WAT did not convey an advantage
during the critical phase that followed. On the contrary, plant survival, TDM, and low leaf symptoms
were associated with lower shoot Zn concentrations.
Higher total plant Zn content (Zn acquisition), on the
other hand, was positively associated with survival
and TDM. High Zn content at 10 WAT was associated
with factors favoring root growth (high root to shoot
ratio and lower Zn translocation to shoots). Leaf
symptoms were again positively correlated with high
Zn concentrations in shoots.
Solution Culture Experiment
Experiments in chelator-buffered nutrient solution
yielded very different results compared with field
experiments. Two of the genotypes that were highly
intolerant under field conditions, Dular and 597, produced more than twice the TDM in the low-Zn solution compared to the most tolerant genotype 507 (Fig.
3). Dular and 597, together with Jalmagna, were the
only genotypes with a relative TDM above 50%. All
three also had the highest seed Zn reserves (Table II),
which could imply that genotypic differences in seed
Zn content contributed to differences in growth in Zndeficient nutrient solutions. A second experiment with
a subset of genotypes and slightly higher Zn concentrations in the low-Zn treatment but otherwise identical conditions produced very similar results (Table
V). The reduction in Zn concentrations of more than
60% and of total Zn content of more than 70% in the
low-Zn treatment relative to the control indicated that
growth reductions were primarily due to Zn deficiency. Genotypic differences for these relative parameters were not significant. Dular and 597, the most
intolerant genotypes in the field, again had the highest
TDM, and this concurred with high root length and
high total Zn content. Both genotypes were also free of
leaf symptoms. This was generally the case for most
plants, which indicated that severe bronzing observed
in field-grown plants was caused by additional stress
factors that probably only have an effect when combined with low Zn availability. Correlation analyses
(Table VI) showed tight associations between seed Zn
content and all traits related to general plant vigor
(TDM, height, root length), whereas traits of more
specific relevance to tolerance to Zn deficiency (shoot
Zn concentrations, leaf bronzing) were unrelated.
QTL Mapping
The average plant mortality for the 110 lines of the
QTL mapping population was 69.9% with a range of
13.5% to 96% (Table VII) in the wet season of 2004
(WS-04). The average in the dry season of 2005 (DS-05)
was 68.2%, with a range of 24.7% to 100%. In both seasons, the frequency distribution of lines was skewed
toward the intolerant parent IR74. Differences between
seasons were more pronounced for the leaf-bronzing
score, with more severe bronzing in the wet season
Table V. Plant growth and Zn nutritional status of seven genotypes grown in chelator-buffered nutrient solutions at a low level of Zn supply
(0.3 mM Zn) and their relative performance compared to a 2.0 mM Zn control treatment
ns, Not significant.
Genotype
Jalmagna
507
474
614
597
IR74
Dular
HSDa
a
TDM
Relative Dry
Matter
Root
Length
Shoot Zn
Concentration
Relative Zn
Concentration
Total Zn
Content
Relative Zn
Content
Leaf
Bronzing
g plant21
%
cm
mg g21
%
mg plant21
1.24
0.60
0.57
0.60
1.31
0.69
1.29
0.21
57.5
54.5
58.2
65.2
59.1
62.7
59.4
ns
40.9
24.1
27.6
30.4
41.6
33.9
42.8
8.4
12.7
14.7
14.7
14.9
16.4
13.5
12.6
ns
28.0
36.3
35.9
32.7
31.9
32.8
25.0
ns
%
Score
25.2
13.1
13.7
14.3
34.4
15.9
26.4
6.7
23.4
23.1
24.0
28.8
26.1
26.5
22.0
ns
0
0
0
0.3
0
0.2
0
ns
Tukey’s HSD (P , 0.05) for the comparison of genotypic means.
Plant Physiol. Vol. 142, 2006
735
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2006 American Society of Plant Biologists. All rights reserved.
Wissuwa et al.
Table VI. Correlation coefficients for the association between traits potentially associated with
tolerance to Zn deficiency in chelator-buffered nutrient solution containing 0.3 mM Zn
* and **, Significant at P , 0.05 and 0.01, respectively.
Traits
Height
TDM
0.95**
Height
Root length
Leaf bronzing
Shoot Zn concentration
Zn content (per plant)
Root Length
Leaf Bronzing
0.81**
0.78**
20.31
20.27
20.13
(average of 4.8) compared to the dry season (average
of 3.5). Phenotypic scores varied from 0.8 to 9 and 1 to
7.6 in the wet and the dry season, respectively. Mean
values for shoot dry matter (ShDM) were 4.2 g plant21
and 35.7 g row21 in WS-04 and 22.1 g plant21 and 115.2
g row21 in DS-05. Higher values for the dry season
were due to a later harvesting time. Plants that survive
Zn deficiency typically enter a recovery phase (Neue
and Lantin, 1994), and plants sampled at 12 WAT (WS04) were in the early recovery phase, while recovery
had further progressed at 17 WAT (DS-05). Transgressive segregation was observed toward both extremes for all traits.
In WS-04, three QTLs were detected for plant mortality on chromosomes 1, 2, and 12 (Table VIII). Individually, they explained between 10.9% and 16.5% of
the variation for plant mortality. For the QTL Zmt1
(Zinc-deficiency-induced mortality) on chromosome 1, the
tolerance allele came from the intolerant parent IR74,
whereas the tolerant parent Jalmagna contributed tolerance alleles for QTLs Zmt2 and Zmt12. The same three
QTLs were again detected in DS-05, but, in addition, one
minor QTL on chromosome 7 was mapped (Table IX).
The portion of overall variation explained by individual
QTLs increased in the dry season, particularly for those
QTLs for which tolerance alleles came from Jalmagna.
The most influential QTL, Zmt12 linked to marker
RG543-1 on chromosome 12, alone explained 24.2% of
the variation. The mode of action of all detected QTLs
was purely additive. Epistatic effects between detected
QTLs (PLABQTL) or between any marker pair (command scantwo in R/qtl) were not detected.
Three minor QTLs for leaf bronzing were detected
on chromosomes 1, 4, and 7 in the wet season (Table
VIII). For the QTL Zbz4 (Zinc-deficiency-induced leaf
bronzing) on chromosome 4, the tolerance allele came
from Jalmagna, whereas intolerant parent IR74 contributed tolerance alleles for QTLs Zbz1a and Zbz7. In
the dry season, two additional QTLs were detected on
chromosomes 1 and 12 together with the common
QTLs Zbz1a and Zbz4 (Table IX). As in the case of plant
mortality, QTLs detected in the dry season explained a
higher proportion of the total variation for leaf bronzing. Several epistatic effects were detected in addition
to main-effect QTLs.
Only two of the five QTLs associated with leaf
bronzing were also associated with plant mortality.
Shoot Zn
Concentration
20.10
0.11
20.07
0.10
Seed Zn
Content
Zn Content
0.92**
0.95**
0.69**
20.26
0.24
0.88**
0.91**
0.64**
20.23
0.05
0.86**
QTLs Zbz1b and Zmt1 were both mapped to marker
interval RG220 to RG109, and the IR74 alleles reduced
plant mortality by 15.4% and leaf bronzing by 1.2
points (Table IX). The second common QTL (Zbz7/
Zmt7) was mapped to interval P2M7-7 to P2M6-5, but
the effect of this QTL differed between seasons and
traits. Zbz7 was detected in the wet season with IR74
contributing the tolerance allele, whereas Zmt7 was
only detected in the dry season with IR74 contributing
the negative allele. The most influential QTL for plant
mortality (Zmt12; DS-05) was not detected as a main
effect for leaf bronzing but was involved in several
epistatic interactions. The Jalmagna allele at Zmt12
(associated with reduced mortality) also had beneficial
effects on leaf bronzing when lines of the mapping
population carried either the IR74 allele at position
125 cM on chromosome 3 (QTL Zdm3; see below) or
the Jalmagna allele at position 107 cM on chromosome
11 (Table VIII). In both cases, the beneficial allelic
combination reduced the leaf-bronzing score by about
2 points.
In mapping QTLs for ShDM, we made the distinction between ShDM per plant or per entire row. Of the
main-effect QTLs detected for ShDM, only locus Zdm7
(Zinc-deficiency-related-dry matter) for ShDM per plant
in DS-05 was not related to either plant mortality
or leaf bronzing (Table IX). Zdm3 was detected for
the first time as a main-effect QTL for ShDM per row
(WS-04; Table VIII), but this locus had been involved
in several epistatic interactions for leaf bronzing. IR74
alleles increased ShDM at both QTLs. The remaining
Table VII. Phenotypic means and distribution among 110 RILs
and parents (IR74 and Jalmagna) of the mapping population
for traits used in QTL mapping
Traits
Mean
WS-04
Mortality (%)
69.9
Leaf bronzing
4.8
ShDM (g plant21)
4.2
ShDM (g row21)
35.7
DS-05
Mortality (%)
68.2
Leaf bronzing
3.4
ShDM (g plant21)
22.1
ShDM (g row21) 115.2
SD
Min
Max
IR74
Jalmagna
19.5 13.5
96.0 75.2
1.9
0.8
9.0
5.8
2.3
0.7
10.4
2.8
31.2
2.4 168.4 22.7
38.4
1.8
9.6
132.2
16.8 24.7 100
78.2
1.4
1.0
7.6
5.5
5.2
0.0
32.4 17.4
65.5
0.0 282.9 55.0
45.4
2.3
32.2
209.1
736
Plant Physiol. Vol. 142, 2006
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2006 American Society of Plant Biologists. All rights reserved.
Effects of Zinc Deficiency on Rice Growth
Table VIII. QTL detected under Zn deficiency in WS-04 for leaf bronzing, plant mortality, and dry matter per plant or per row
Marker Interval
(Epistatic Interaction)
Chromosome
1
4
7
1 3 3
12 3 3
12 3 11
1
2
12
2
2 3 3
2 3 4
12 3 3
12 3 6
3
12
12 3 3
12 3 7
Leaf bronzing
RZ154-3–P1M10-14
RG788–P2M5-13
P2M7-7–P2M6-5
(Zbz1b 3 Zdm3)
(Zmt12 3 Zdm3)
(Zmt12 3 C11/107 cM)
Full model with epistasisd
Plant mortality
RG220–RG109
P2M9-8–P2M6-8
CDO344-1–RG543-1
Full modeld
ShDM per plant
P2M9-8–P2M6-8
(Zmt2 3 Zdm3)
(Zmt2 3 Zbz4)
(Zmt12 3 Zdm3)
(Zmt12 3 C6/185cM)
Full model with epistasisd
ShDM per row
RZ675–P1M9-10
CDO344-1–RG543-1
(Zmt12 3 Zdm3)
(Zmt12 3 Zmt7)
Full model with epistasisd
QTL
Position(cM)
Support Interval
LODa
R2b
Allelic Effectc
Tolerance Allele
Score
Zbz1a
Zbz4
Zbz7
76
92
0
72–86
84–96
0–22
3.2
2.6
2.7
12.8
12.9
10.7
10.8
6.1
6.0
8.2
42.4
11.52
21.68
11.20
IR74
Jalmagna
IR74
IR74 3 IR74
Jal 3 IR74
Jal 3 Jal
%
Zmt1
Zmt2
Zmt12
136
44
150
132–154
24–60
142–156
4.2
2.7
2.9
8.0
16.5
10.9
11.6
29.1
Zmt2
44
28–76
2.5
10.6
10.4
6.7
13.8
12.2
39.4
115.4
216.4
212.4
IR74
Jalmagna
Jalmagna
g
11.4
1.52
Jalmagna
IR74 3 Jale
Jal 3 IR74
Jal 3 IR74
IR74 3 Jale
g
Zdm3
Zmt12
125
150
100–134
144–162
4.4
2.7
9.7
18.1
11.6
12.6
5.6
35.8
227.2
18.8
IR74
Jalmagna
Jal 3 IR74
Jal 3 Jal
a
b
Log10 of the likelihood odds ratio. The LOD score is calculated from the F value in the multiple regression.
Coefficient of determination or
c
the percentage of the phenotypic variance, which is explained by the detected QTL.
Estimated phenotypic effect of substituting both IR74
d
alleles with Jalmagna alleles at QTL.
Results of a multiple regression analysis that includes all main effects and digenic epistatic
e
interactions.
Epistatic interaction between this allele combination results in worse phenotype than expected if effects were purely additive.
two main-effect ShDM-QTLs (Zmt2 and Zmt12) had
already been detected for plant mortality, and as for
those two traits, Jalmagna alleles increased tolerance
to Zn deficiency. Digenic epistatic interactions played
a relatively greater role for ShDM than for mortality
and bronzing. Zmt12 was involved in six out of nine
interactions and therefore appears to be highly influential. One epistatic interaction in particular occurred
more than once: Zmt12 3 Zdm3 was associated with
leaf bronzing (both seasons) and with ShDM per plant
and ShDM per row (WS-04), and the Jal 3 IR74 allelic
combination consistently contributed to tolerance.
DISCUSSION
Experimental Conditions
The detailed characterization of Jalmagna, IR74, and
several derived RILs in a Zn-deficient field confirmed
the tolerance ranking obtained previously (Hoffland
et al., 2006) and suggested that RILs of the mapping
population varied sufficiently to warrant a full QTL
mapping effort. Tolerance rankings obtained in lowZn nutrient solutions did not concur with tolerance
rankings in the field. One reason was the lack of
association between field tolerance and parameters
related to internal Zn efficiency (shoot Zn concentration or Zn translocation), which typically determine
genotypic differences in nutrient solution experiments. One other reason why lines with low field
tolerance did well in culture solution may have been
related to differences in Zn reserves contained in the
seed. The correlation between biomass accumulation
and seed Zn was high in nutrient solution (r 5 0.88)
but low in the field (r 5 0.18). High seed Zn may have
increased seedling growth rates under Zn deficiency,
and with a bigger root system (Table V), genotypes
like Dular and 597 could have had an advantage in
Zn acquisition in competition with other genotypes
when grown in nutrient solution. However, it is
also possible that seedling vigor was a general characteristic of tall traditional varieties such as Dular,
Jalmagna, and RIL597 (597 resembles Jalmagna more
than the second parent, IR74). At this point, it is not
possible to confirm either hypothesis, but our results
indicate that high seed Zn reserves alone are of no
lasting advantage under the conditions encountered at
our field site. It remains to be seen if high seed Zn
would be beneficial if combined with additional tolerance mechanisms.
Regardless of these considerations, it can be concluded that tolerance to low Zn availability alone, as
Plant Physiol. Vol. 142, 2006
737
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2006 American Society of Plant Biologists. All rights reserved.
Wissuwa et al.
Table IX. QTL detected under Zn deficiency in the DS-05 for leaf bronzing, plant mortality, and dry matter per plant or per row
Chromosome
1
1
4
12
12 3 3
12 3 9
1
2
7
12
7
4 3 7
2
12
2 3 12
7 3 12
Marker Interval (Epistatic
Interaction)
Leaf bronzing
P1M10-14–P2M10-11
RG220–RG109
RG788–P2M5-13
P3M5-9–P1M7-4
(Zmt12 3 Zdm3)
(Zmt12 3 C9/115cM)
Full model with epistasisd
Plant mortality
RG109–Sd1
P2M9-8–P2M6-8
P2M7-7–P2M6-5
CDO344-1–RG543-1
Full modeld
ShDM per plant
P3M7-11–P3M1-11
Zbz4 3 Zdm7
Full model with epistasisd
ShDM per row
P2M9-8–P2M6-8
CDO344-1–RG543-1
Zmt2–Zmt12
Zdm7 3 Zmt12
Full model with epistasisd
QTL
Position (cM)
Support Interval
Zbz1a
Zbz1b
Zbz4
Zbz12
80
124
92
70
76–100
122–130
86–98
66–96
LODa
4.4
4.1
4.7
3.6
R2b
18.3
17.8
16.5
18.7
14.6
12.4
5.7
55.6
Allelic Effectc
Tolerance Allele
11.64
11.20
21.84
10.92
IR74
IR74
Jalmagna
IR74
Jal 3 IR74
Jal 3 IR74
Zmt1
Zmt2
Zmt7
Zmt12
138
42
20
150
122–142
18–58
0–42
134–154
2.9
4.2
2.7
6.4
11.7
11.8
16.6
11.3
24.2
39.7
19.8
216.6
29.2
214.8
IR74
Jalmagna
Jalmagna
Jalmagna
Zdm7
90
82–96
3.4
14.1
17.9
30.0
24.5
IR74
Jal 3 Jale
10.6
10.3
4.8
6.0
27.7
70.9
46.2
Jalmagna
Jalmagna
Jal 3 Jal
IR74 3 Jal
8.1
Zmt2
Zmt12
34
149
8–62
134–154
2.6
2.5
7.4
a
b
Log10 of the likelihood odds ratio. The LOD score is calculated from the F-value in the multiple regression.
Coefficient of determination or
c
the percentage of the phenotypic variance, which is explained by the detected QTL.
Estimated phenotypic effect of substituting both IR74
d
Results of a multiple regression analysis that includes all main effects and digenic epistatic
alleles with Jalmagna alleles at QTL.
e
interactions.
Epistatic interaction between this allele combination results in worse phenotype than expected if effects were purely additive.
simulated in buffered nutrient solutions with low free
Zn activity, was not suitable for screening genotypes
for tolerance to the more complex stress encountered
at our field site. This conclusion is in contrast to the
one reached by Hacisalihoglu and Kochian (2003) that
mechanisms related to internal efficiency would be
more important than Zn acquisition and that nutrient
solution experiments generally reproduce results seen
in the field. Their conclusion was based on a review of
studies conducted with upland crops (wheat, barley,
and beans). It is therefore likely that the more complex
interactions of additional stress factors (low redox
potential, high organic matter and bicarbonate content,
high Mg to Ca ratio) in the perennially flooded soil of
our study site required additional tolerance mechanisms, possibly associated with the rhizosphere, that
cannot be reproduced in solution experiments. The
excretion of organic acids (Kirk and Bajita, 1995) or
phytosiderophores (Zhang et al., 1998) capable of increasing the bioavailability of Zn in the rhizosphere are
potential mechanisms of importance in lowland rice.
Hoffland et al. (2006) detected very low citric acid
exudation rates in our most intolerant genotypes
(Dular, IR74), which suggested that organic acid
excretion could explain genotypic differences in Zn
acquisition observed among the genotypes used in this
study.
Genetic Factors Associated with Tolerance
to Zn Deficiency
Studies on the effect of low Zn availability in rice
have usually regarded the occurrence and extent of leaf
bronzing as indications of Zn deficiency (Cayton et al.,
1985; Qadar, 2002). Genotypes were considered tolerant to Zn deficiency if they showed few leaf symptoms,
and this was generally accompanied by lower mortality and higher dry matter. If we had limited our initial
study to the two parents of the QTL mapping population, IR74 and Jalmagna, we would have come to a
similar conclusion. However, using an additional set
of segregating RILs of the QTL mapping population
showed that leaf bronzing did not necessarily concur
with other more important indicators of tolerance to
Zn deficiency. Based on the nonsignificant correlation
between plant mortality and leaf bronzing, we hypothesize that both symptoms are under independent
genetic control and conclude that leaf bronzing alone
is not a reliable criterion to evaluate genotypic differences in tolerance to Zn deficiency in rice.
To test this hypothesis, we initially focused on plant
mortality and leaf bronzing in the QTL analysis, and
results largely confirmed our hypothesis. Only one
minor QTL (Zmt7/Zbz7) was mapped as a main effect
for both mortality and bronzing, whereas more
738
Plant Physiol. Vol. 142, 2006
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2006 American Society of Plant Biologists. All rights reserved.
Effects of Zinc Deficiency on Rice Growth
influential QTLs were specific for either trait. However, the plant mortality QTL Zmt12 was involved in
several digenic epistatic interactions for leaf bronzing.
In each case, it was the Jalmagna allele that contributed
to a reduction in bronzing when in combination with
the suitable allele at the epistatic locus. It appears that
the positive effect of Zmt12 on plant mortality can also
lead to reduced leaf bronzing if other genetic factors
are present that allow plants to take advantage of this
positive effect.
Lines of the mapping population showed more
severe symptoms of Zn deficiency in the wet season
compared to the dry season (Table VII). This is unusual
because it is generally expected that the high irradiance typically encountered in the dry season increases
overall stress and particularly enhances leaf bronzing
due to the activation of radical oxygen species (Marschner
and Cakmak, 1989). That we did not observe this in the
study could have been due to very heavy rainfall
during a typhoon in the WS-04 that led to flooding of
field plots during the seedling stage. Flooding may
have caused additional damage to the plants that were
already stressed by Zn deficiency. The presence of
such an additional stress factor may explain why the
portion of the overall variation explained by QTLs (R2)
associated with tolerance to Zn deficiency was higher
in the dry season. This could largely be attributed to an
increase in R2s of those QTLs for which the tolerant
parent Jalmagna contributed the positive allele. Results of the dry season could therefore be considered
more representative of Zn deficiency.
Fewer main-effect QTLs were detected for ShDM
than for other traits, and all but one (Zdm7) had
already been detected as main or interaction effects
for plant mortality or leaf bronzing. Digenic epistatic
interaction effects were often more important for
ShDM than main effects, and in almost all cases these
interactions were detected between main-effect QTLs
previously identified for plant mortality or leaf bronzing. That these loci were more prominent for individual Zn deficiency symptoms (mortality, leaf bronzing)
is to be expected because dry matter should be the
result of interactions between multiple stress factors
and multiple tolerance mechanisms. As the complexity of a trait increases, such as in the case of ShDM
under stress compared to leaf bronzing or mortality,
one can expect a reduction in the R2s of those QTLs.
However, that epistatic interactions often involved the
same loci detected as main-effect QTLs (e.g. Zmt12 and
Zmt2) can be seen as strong confirmation of their
importance under Zn deficiency.
Based on the first physiological test with lines from
the mapping population, one candidate mechanism
potentially linked to QTLs is the ability to maintain Zn
acquisition, possibly as a result of organic acid excretion. An alternative but less direct way to improve Zn
acquisition from Zn-deficient soils characterized by
high bicarbonate concentrations, as encountered in
our field site, is to overcome the bicarbonate-induced
inhibition of root growth (Yang et al., 1994). Simple
parameters of internal Zn efficiency such as tissue Zn
concentrations were not associated with tolerance to
Zn deficiency in the subset of genotypes studied in
more detail. However, more precise physiological
studies could yet reveal a role of more efficient biochemical utilization of cellular Zn in tolerance to Zn
deficiency in rice (Hacisalihoglu and Kochian, 2003).
Candidate mechanisms that could be associated with
leaf-bronzing QTLs, in particular, are the activity of
Zn-requiring enzymes (Cakmak et al., 1997) or the
degree to which nutrient imbalances interfere with
efficient biochemical utilization of Zn under Zn deficiency (Qadar, 2002).
To our knowledge, this is the first published report
on QTLs associated with tolerance to Zn deficiency in
any crop; however, it is not possible to relate any of the
tolerance mechanisms discussed above with the loci
detected so far. Nevertheless, having identified distinct genetic factors for two of the typical symptoms
associated with Zn deficiency represents the first important step toward gaining a more complete understanding of required tolerance mechanisms. Once
appropriate genetic material (substitution lines, nearisogenic lines) has been developed for the QTLs
identified in this study, it will be possible to examine
cause-effect relationships for the complex factor interactions that lead to the Zn-deficient phenotype without confounding effects of a multitude of unrelated
loci. The first steps in that direction have been undertaken by backcrossing one of the most tolerant RILs
from this mapping population with the intolerant
parent IR74 for the purpose of developing a secondary
mapping population.
Table X. Total concentration of added nutrients and resulting free
activities in chelator-buffered nutrient solution (pH 5.5), based
on calculations made using the software Geochem-PC (Parker
et al., 1995)
Element
Reagent
Concentrationa
Free Activitya
mol L21
mol L21
Ca
Mg
K
NH4
NH4
PO4
Fe
Mn
Cu
B
Mo
Zn (2.0)b
Zn (0.3)b
Zn (0.1)b
NTA
HEDTA
MES
CaCl22H2O
MgSO47H2O
KCl
(NH4)2SO4
NH4Cl
KH2 PO4
FeCl37H2O
MnSO4H2O
CuSO45H2O
H3BO3
Na2MoO42H2O
ZnSO47H2O
ZnSO47H2O
ZnSO47H2O
2.50E-03
1.00E-03
2.51E-03
0.50E-03
6.50E-03
2.00E-05
2.00E-04
2.00E-06
5.00E-07
1.00E-05
5.00E-07
2.00E-06
3.00E-07
1.00E-07
9.42E-05
1.82E-04
3.00E-03
1.28E-03
5.23E-04
2.16E-03
2.06E-06
4.47E-14
4.95E-15
5.54E-07
3.41E-13
2.01E-09
2.69E-07
7.76E-10
4.56E-11
3.15E-11
a
The concentration and free activity refers to elements, not reb
Treatment levels refer to total concentrations added: 2.0,
agents.
0.3, or 0.1 mM Zn.
Plant Physiol. Vol. 142, 2006
739
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2006 American Society of Plant Biologists. All rights reserved.
Wissuwa et al.
Field experiments conducted in this study were
under conditions where Zn deficiency is caused by
continuous flooding and high bicarbonates. However,
Zn deficiency could also be induced by other soil
factors not related to prolonged flooding, such as high
soil pH in saline/sodic soils. Additional mechanisms/
QTLs may, therefore, be important under such conditions.
block for an additional 5 min. Drop-wise addition of H2O2 was repeated until
the solution became clear; we then continued the digestion for one final hour.
The digest was filtered, brought to a 25-mL volume, and the Zn concentration
in the filtrate analyzed by atomic absorption spectrometer (AAS 3100; Perkin
Elmer).
Within Zn treatments, the experiment was arranged in a randomized
complete block with five replications. Data from the plus-Zn control were
used to calculate relative values. The main data analysis then focused on the
Zn-deficient plot. For ANOVA, the PROC MIXED of SAS was used, and for
separation of genotypic means we used Tukey’s HSD test with a significance
threshold of P , 0.05.
MATERIALS AND METHODS
Solution Culture Experiment
Plant Material
The same 10 lines were evaluated in a nutrient solution experiment in a
greenhouse during the dry season of 2003. A chelator-buffered solution was
used to maintain a constant and low concentration of Zn in solution and to
neutralize possible contamination with traces of Zn. The activities of added
nutrients in buffered solutions were calculated with the software Geochem-PC
(Parker et al., 1995). Chelators NTA and HEDTA were added in excess of
ligands. The pH was buffered at 5.5 through MES. Concentrations and free
activities of nutrients are shown in Table X.
Seeds were germinated on moist filter paper and transferred to netted
styrofoam sheets floating on a solution containing 0.5 mM CaCl2. Ten days
after seeding (DAS), seedlings were transferred to the experimental setup:
styrofoam sheets covering large 32-L basins filled with buffered nutrient
solution. Plants were fastened to holes in the styrofoam with the help of foam
strips. Each basin had 24 planting holes that contained three plants each. The
treatment commenced after 10 d of pretreatment in zero-Zn, half-strength
nutrient solution. Two levels of Zn (2.0 mM, normal; 0.1 mM, deficient) were
applied. Nutrient solutions were changed weekly and the pH was adjusted to
5.5 daily. Plants were harvested after 5 weeks of Zn treatment (55 DAS) at a
stage when field-grown plants would have shown very clear symptoms of Zn
deficiency such as leaf bronzing. This experiment was repeated in the wet
season of 2003 with some modifications: a smaller set of genotypes was used
(omitted M79, 512, 639), Zn concentrations were increased to 0.3 mM in the
deficient treatment, and plants were harvested after 7 weeks of Zn treatment
(70 DAS). Tissue samples were processed as described for samples from field
experiments.
A mapping population had been developed at IRRI for the purpose of
identifying QTLs for flooding tolerance (Sripongpangkul et al., 2000). The
parental lines were two indica cultivars, IR74 and Jalmagna. IR74 is a highyielding, modern semidwarf cultivar, whereas Jalmagna is a traditional, tall,
low-yielding cultivar adapted to deepwater conditions in India. Parental
lines were also found to contrast in their tolerance to Zn deficiency, which
suggested that this mapping population might be suitable for detecting QTLs
associated with tolerance to Zn deficiency. The original mapping population
consisted of 165 F8 RILs derived by single-seed descent.
For the purpose of further characterizing the response to Zn deficiency in
that population, we chose 10 genotypes based on an observational trial
conducted in the dry season of 2002. Six RILs (lines 474, 507, 512, 597, 614, and
639) representing the whole range of response to Zn deficiency were selected
in addition to both parents and the two unrelated genotypes Dular (traditional
variety from India, intolerant of Zn deficiency) and M79 (modern breeding
line of Chinese origin, tolerant of Zn deficiency).
Field Experiment
All field experiments were conducted on a highly Zn-deficient soil (Typic
Hydraquent) located in Tiaong, Quezon province, Philippines. The field is
characterized by low Zn availability (0.1 mg kg21; DTPA extraction method),
high pH (7.8; 1:1 w/v water), high organic matter (2.8%), and a low redox
potential. The field was kept permanently flooded because temporary drying
was found to reduce the severity of Zn deficiency. During field preparations,
the standard recommended dose of NPK was applied as a compound fertilizer
(14-14-14) at a rate of 136 kg ha21. A neighboring plus-Zn control plot received
20 kg Zn ha21 annually as ZnSO4 together with the NPK fertilizer. At the
booting stage, a second dose of N fertilizer was applied as urea (68 kg ha21) to
both field plots. A detailed description of soil conditions at the field site can be
found in Van Breemen and Castro (1980) and Scharpenseel et al. (1983).
Experiments were conducted in the dry season (February–May) and wet
season (August–December) of 2003.
Seeds of the 10 genotypes were germinated in seeding trays with grids
filled with approximately 6 g of soil from the IRRI farm. Seedlings were raised
for 20 d and soil was washed off roots prior to transplanting to field plots. An
experimental unit consisted of a single row of 20 plants per genotype with
spacing of 20 cm within and between rows. The experiment was conducted
with five replications. Five and 12 WAT, the number of surviving plants per
row was determined, plant height was measured, and surviving plants were
scored for leaf bronzing. Leaf symptoms were scored based on the following
classification: 0, no symptoms; 1, very few small brown lesions on oldest
leaves; 3, brown lesions on oldest leaves enlarged, very few brown lesions on
middle leaves; 5, more than 50% of old leaves covered by brown lesions,
brown lesions on middle leaves enlarged; 7, old leaves entirely covered by
brown lesions, more than 50% of middle leaves covered by brown lesions; and
9, all but the youngest leaf completely brown.
Remaining plants were collected 12 WAT. One modification in the wet
season experiment in 2003 was that each experimental unit consisted of a
double row with 15 plants per row. One row was sampled 5 WAT before plant
number was reduced in intolerant genotypes; the second row was sampled
10 WAT. Roots and shoots were separated, oven dried at 70°C for 4 d, and their
dry weights recorded. Plant samples were then ground to a fine powder, and
0.5 g tissue was digested in 5 mL of concentrated H2SO4 at 300°C for 2 h. Tubes
were then removed from the digestion block to cool for 15 min, after which
2 mL of H2O2 (30%) was carefully added. Tubes were returned to the digestion
QTL Mapping
The phenotypic evaluation of 110 RILs of the IR74 3 Jalmagna population
was conducted in the same field plots and under the same fertilizer regime as
described above. An experimental unit consisted of a single row of 20
individual 20-d-old seedlings, transplanted with a spacing of 20 cm within
and between rows. Four replications were planted to both the Zn-deficient
and the fertilized control plots. Experiments were conducted in WS-04
(September–December) and in DS-05 (February–June). Plant mortality and
leaf bronzing were scored 12 WAT in both seasons. In 2004, plants were
harvested after scoring (12 WAT). In 2005, we intended to obtain data on grain
yield; however, in most of the lines, flowering was either delayed substantially
or did not occur at all within the growing season. For that reason, we were able
to harvest mostly shoot material (17 WAT). Shoot dry weight per row was
determined after oven-drying shoot samples of all harvested plants for 4 d at
70 C. Shoot dry weight per plant was derived by dividing total row weight by
number of plants harvested.
QTL mapping was based on the linkage map of 142 AFLP and two gene
markers constructed by Sripongpangkul et al. (2000). The 144 markers covered
a total length of 2,339.9 cM, with an average interval of 17.9 cM between
markers. For the majority of markers, the distribution was skewed toward
IR74 (64% of alleles were from IR74). QTL analysis was performed with the
software package PLABQTL (Utz and Melchinger, 1996), which uses a multiple regression approach as suggested by Haley and Knott (1992). In a first
step, simple interval mapping was performed and cofactors selected. For
cofactor selection, F-to-enter and F-to-drop thresholds were set at 6.0 to avoid
selecting multiple markers linked to one QTL as cofactors. Using these
cofactors to reduce the residual variation, QTLs were detected with composite
interval mapping. A log of the odds (LOD) score .2.50 was considered as
indicative of the presence of a QTL. For the detection of digenic epistatic
interactions, we used the command ‘‘scantwo’’ of ‘‘R/Qtl’’ written by Broman
and Wu (2005) for the free statistical software R. The command scantwo
740
Plant Physiol. Vol. 142, 2006
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2006 American Society of Plant Biologists. All rights reserved.
Effects of Zinc Deficiency on Rice Growth
performs a two-dimensional genome scan with a two-QTL model and is able
to detect interactions between loci that may not be detected as main factors.
Interactions detected with scantwo were added to the model statement in
PLABQTL using the command seq/s. Interactions were retained if they were
declared significant in the multiple regression analysis of PLABQTL.
Received June 16, 2006; accepted July 21, 2006; published August 18, 2006.
LITERATURE CITED
Broman KW, Wu H (2005) R/qtl: a QTL mapping environment; version
1.01-9. Department of Biostatistics, Johns Hopkins University. http://
www.biostat.jhsph.edu/;kbroman/qtl/
Cakmak I, Öztürk L, Eker S, Torun B, Kalfa HI, Yilmaz A (1997)
Concentration of zinc and activity of copper/zinc-superoxide dismutase in leaves of rye and wheat cultivars differing in sensitivity to zinc
deficiency. J Plant Physiol 151: 91–95
Cayton MTC, Reyes ED, Neue HU (1985) Effect of zinc fertilization on the
mineral nutrition of rices differing in tolerance to zinc deficiency. Plant
Soil 87: 319–327
Dobermann A, Fairhurst TH (2000) Rice: Nutrient Disorders and Nutrient
Management. Potash and Phosphate Institute, Potash and Phosphate
Institute of Canada and International Rice Research Institute, Singapore
Hacisalihoglu G, Kochian LV (2003) How do some plants tolerate low
levels of soil zinc? Mechanisms of zinc efficiency in crop plants. New
Phytol 159: 341–350
Hajiboland R, Yang XE, Römheld V (2003) Effects of bicarbonate and high
pH on growth of Zn-efficient and Zn-inefficient genotypes of rice, wheat
and rye. Plant Soil 250: 349–357
Haley CS, Knott SA (1992) A simple regression method for mapping
quantitative trait loci in line crosses using flanking markers. Heredity
69: 315–324
Hoffland E, Wei C, Wissuwa M (2006) Organic anion exudation by lowland
rice (Oryza sativa L.) at zinc and phosphorus deficiency. Plant Soil 283:
155–162
Kirk GJD, Bajita JB (1995) Root-induced iron oxidation, pH changes and
zinc solubilization in the rhizosphere of lowland rice. New Phytol 131:
129–137
Marschner H, Cakmak I (1989) High light intensity enhances chlorosis and
necrosis in leaves of zinc, potassium, and magnesium deficient bean
(Phaseolus vulgaris) plants. J Plant Physiol 134: 308–315
Nene YL (1966) Symptoms, cause and control of Khaira disease of paddy.
Bull Indian Phytopathol Soc 3: 97–191
Neue HU, Lantin RS (1994) Micronutrient toxicities and deficiencies in
rice. In AR Yeo, TJ Flowers, eds, Soil Mineral Stresses: Approaches to
Crop Improvement. Springer-Verlag, Berlin, pp 175–200
Parker DR, Norvell WA, Chaney RL (1995) GEOCHEM-PC—a chemical
speciation program for IBM and compatible personal computers. In RH
Loeppert, AP Schwab, S Goldberg, eds, Chemical Equilibrium and
Reaction Models. Soil Science Society of America, Madison, WI, pp
253–269
Qadar A (2002) Selecting rice genotypes tolerant to zinc deficiency and
sodicity stresses. I. Differences in zinc, iron, manganese, copper,
phosphorus concentrations, and phosphorus/zinc ratio in their leaves.
J Plant Nutr 25: 457–473
Quijano-Guerta C, Kirk GJD, Portugal AM, Bartolome VI, McLaren GC
(2002) Tolerance of rice germplasm to zinc deficiency. Field Crops Res
76: 123–130
Scharpenseel HW, Eichwald E, Haupenthal C, Neue HU (1983) Zinc
deficiency in a soil toposequence, grown to rice, at Tiaong, Quezon
Province, Philippines. CATENA 10: 115–132
Singh B, Natesan SKA, Singh BK, Usha K (2003) Improving zinc efficiency
of cereals under zinc deficiency. Curr Sci 88: 36–44
Sripongpangkul K, Posa GBT, Senadhira DW, Brar D, Huang N, Khush
GS, Li ZK (2000) Genes/QTLs affecting flood tolerance in rice. Theor
Appl Genet 101: 1074–1081
Utz HF, Melchinger AE (1996) PLABQTL: a program for composite interval
mapping of QTL. J Quant Trait Loci 2: Paper 1
Van Breemen N, Castro RU (1980) Zinc deficiency in wetland rice along a
toposequence of hydromorphic soils in the Philippines. II. Cropping
experiment. Plant Soil 57: 215–221
Yang X, Römheld V, Marschner H (1994) Effect of bicarbonate on root
growth and accumulation of organic acids in Zn-inefficient and
Zn-efficient rice cultivars (Oryza sativa L.). Plant Soil 164: 1–7
Yoshida S, Tanaka A (1969) Zinc deficiency of the rice plant in calcareous
soils. Soil Sci Plant Nutr 15: 75–80
Zhang X, Zhang F, Mao D (1998) Effect of iron plaque outside roots on
nutrient uptake by rice (Oryza sativa L.). Zinc uptake by Fe-deficient rice.
Plant Soil 202: 33–39
Plant Physiol. Vol. 142, 2006
741
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2006 American Society of Plant Biologists. All rights reserved.