Brief Communications
Genetic Relationships
Between Loci for Palmitate
Contents in Soybean
Mutants
S. M. Rahman, T. Kinoshita,
T. Anai, and Y. Takagi
Elevated palmitate content in soybean
[Glycine max (L.)] oil may be important for
the production of some food and industrial
products. The palmitate content of commercial cultivars is approximately 11.0%.
Soybean mutants with elevated and reduced palmitate have been developed. Elevated palmitate in J10 (ø15.6%) and
C1727 (ø17.3%), and reduced palmitate
in C1726 (ø8.6%) were controlled by
sop2, fap2, and fap1, respectively. The
genetic system for elevated palmitate in
KK7 (ø13.5%) was unknown. Reciprocal
crosses were made as Bay 3 J10, Bay 3
C1727, Bay 3 KK7, C1727 3 J10, C1726
3 J10, J10 3 KK7, C1727 3 KK7, and
C1726 3 KK7 to determine if the sop2 allele in J10 was different from fap2 and
fap1, and if the allele in KK7 was different
from those of sop2, fap2, and fap1. No
maternal effect was observed in any of the
crosses. The phenotypic segregation ratio
and the genotypic evaluation of F2 plants
from C1726 3 J10, J10 3 KK7, C1727 3
KK7, and C1726 3 KK7 indicated that the
sop2 allele in J10 was at a different locus
from fap1, and that the allele in KK7 was
at a different locus from sop2, fap2, and
fap1. The lack of transgressive segregation in F2 seeds and their F2 plants with
putative lower and higher palmitate content of C1727 3 J10 indicated that the
sop2 allele was at the same locus as fap2.
The sop2 allele in J10 was given the permanent designation of fap2 (J10) and the
allele in KK7 was temporarily designated
fapx. Lines with the fap2fap2fapxfapx genotype may provide important sources of
elevated palmitate in soybean seed oil.
The fatty acid composition of vegetable
oils influences their use in industrial food
applications. Commercial oil seed crops
cannot satisfy all the demands of the food
industry. The food industry may benefit
from vegetable oils with unique fatty acid
profiles (Rattray 1991).
Alteration of the fatty acid composition
of seed oil can be done in different ways
(Robbelen et al. 1989). Although natural
variability for these traits has been observed in most of the oilseed crops
( Knowles 1989; Stefansson et al. 1961),
natural genetic variability for palmitate
content in commercial soybean (Glycine
max L.) is limited (Rattary 1991). Therefore additional variability for this trait
must be developed by mutagenesis, the
recombination of different mutant genes,
or genetic engineering (Ohlrogge et al.
1991).
Palmitate is one of the two major saturated fatty acids of soybean oil that is important for the production of plastic fat
products, such as shortening and margarine ( Vles and Gottenbos 1989). The palmitate content of commercial cultivars is
approximately 11.0%. Lines with elevated
or reduced palmitate have been developed by different breeding techniques
( Bubeck et al. 1989; Erickson et al. 1988;
Fehr et al. 1991a,b; Wilcox et al. 1994).
Erickson et al. (1988) found that the elevated palmitate content (ø17.3%) of
C1727 was controlled by fap2, while the
reduced content (ø8.6%) of C1726 was
controlled by fap1. Fehr et al. (1991a) determined that the A21 line, with approximately 20.0% palmitate, had an allele
(fap2-b) at the same locus as the fap2 or a
tightly linked locus. Schnebly et al. (1994)
crossed A24 with approximately 17.7% palmitate to the reduced palmitate lines
C1726 (fap1; Erickson et al. 1988) and A22
(fap3; Fehr et al. 1991b), and found that
A24 had an allele for elevated palmitate
content at a locus independent of fap1 and
fap3. The allele was given the permanent
designation fap4.
The mutant line J10, with approximately
15.6% palmitate, was developed at Saga
University by X-ray irradiation of the
seeds of Bay ( Takagi et al. 1995). Rahman
et al. (1996) temporarily designated the allele for palmitate in J10 as sop2 because
it’s relationship to other alleles was unknown. Another mutant line, KK7, with approximately 13.5% palmitate was developed from a different M4 population of
Bay. The genetic system of elevated palmitate in KK7 has not been determined.
The objective of this study was to determine if the sop2 allele in J10 was different
from fap2 and fap1, and if the allele in KK7
was different from those of sop2, fap2, and
fap1.
Materials and Methods
The soybean lines used as parents for this
study were the mutants J10, C1727, C1726,
and KK7 and the cultivar Bay. The following crosses and their reciprocals were
made in the greenhouse at 228C–308C with
a 12 h day length at Saga University in
1994: Bay 3 J10, Bay 3 C1727, Bay 3 KK7,
C1727 3 J10, C1726 3 J10, J10 3 KK7,
C1727 3 KK7, and C1726 3 KK7.
Plants of the parents used for crossing
were identified, and selfed seeds were harvested from a node adjacent to one from
which the F1 seed was obtained. Individual
F1 and parent seeds were analyzed for fatTable 1. Number of F2 seeds analyzed for fatty
acid composition from each cross
Cross
No. of F2 seeds
Bay 3 J10
Bay 3 C1727
Bay 3 KK7
C1727 3 J10
C1726 3 J10
J10 3 KK7
C1727 3 KK7
C1726 3 KK7
80
100
80
120
64
176
192
64
423
Table 2. Mean palmitate content of F 1 seeds from
mutant 3 Bay and mutant 3 mutant crosses, and
of seeds from the parents
Parent or cross
Palmitate
content
(%)
Parent or cross
Palmitate
content
(%)
Bay
Bay 3 J10
J10 3 Bay
J10
LSDa
Midparent
F1 meanb
LSDc
Replications
Bay
Bay 3 C1727
C1727 3 Bay
C1727
LSD
Midparent
F1 mean
LSD
Replications
Bay
Bay 3 KK7
KK7 3 Bay
KK7
LSD
Midparent
F1 mean
LSD
Replications
C1727
C1727 3 J10
J10 3 C1727
J10
LSD
Midparent
F1 mean
LSD
Replications
10.6
12.9
12.8
15.4
0.38
13.0
12.9
0.24
12
10.6
13.3
13.1
16.3
0.35
13.5
13.2
0.20
12
10.4
11.1
11.1
13.5
0.32
12.0
11.1
0.20
12
16.2
16.1
16.2
15.4
0.25
15.8
16.2
0.17
16
8.3
10.8
10.7
15.5
0.22
11.9
10.8
0.17
16
15.2
12.5
12.6
13.5
0.33
14.4
12.6
0.22
16
16.2
14.0
13.9
13.5
0.27
14.9
14.0
0.19
16
8.2
10.6
10.7
13.4
0.17
10.8
10.7
0.12
16
C1726
C1726 3 J10
J10 3 C1726
J10
LSD
Midparent
F1 mean
LSD
Replications
J10
J10 3 KK7
KK7 3 J10
KK7
LSD
Midparent
F1 mean
LSD
Replications
C1727
C1727 3 KK7
KK7 3 C1727
KK7
LSD
Midparent
F1 mean
LSD
Replications
C1726
C1726 3 KK7
KK7 3 C1726
KK7
LSD
Midparent
F1 mean
LSD
Replications
Figure 1. Distribution of palmitate content in seeds of J10 and KK7 and in F2 seeds of J10 3 KK7.
Least significant difference (P 5 .05) for comparison of
parent and F1 values.
b
Average of reciprocal crosses used for comparison
with the midparent value.
c
Least significant difference (P 5 .05) for comparison of
the midparent value with the F1 mean.
a
ty acid composition in a randomized complete-block design. Each cross was an independent test, and each replicate of a
test consisted of one seed from each of
the parents and one seed from each of the
reciprocal hybrids. Each seed was cut into
two parts with a razor blade. The part
with the embryonic axis, used for planting, was two-thirds of the seed. The part
lacking the embryonic axis was used for
fatty acid analysis. The palmitate content
of the reciprocal F1 and parent seeds were
compared using a standard analysis of
variance for a randomized complete-block
design to determine the presence of maternal effects and dominance relationships.
The part of each F1 and parent seed containing the embryonic axis was planted in
the field at Saga University in July 1994.
The seeds were planted 20 cm apart within the rows with 60 cm between rows.
424 The Journal of Heredity 1999:90(3)
except for C1727 3 J10, were used to determine the genotype of the F2 plants. In
C1727 3 J10, the range of palmitate in F2
seeds was found to be slightly higher than
the parental range. Therefore 10 F2 seeds
from each of the putative lower and higher
sides and 10 seeds of each parent were
selected. Seeds of these crosses were
planted in the field at Saga University in
July 1995, 20 cm apart within rows and 60
cm between rows. Each plant was harvested individually. For the genotypic
evaluation of the F2 plants, a random sample of 15 individual F3 seeds from each F2
plant and 10 seeds from each of three
plants of each parent for C1726 3 J10, J10
3 KK7, C1727 3 KK7, and C1726 3 KK7
were analyzed for fatty acid composition.
In C1727 3 J10, a 20-seed sample for each
F2 plant and parent plant was also analyzed. There were three F2 plants from
C1727 3 J10 with a palmitate content higher than the range of C1727 and J10. There-
Each F1 and parent plant was harvested
individually. To reduce the effects of within-plant variability, seeds of parents and F2
seeds were collected from pods at the fifth
through eighth nodes of the main stem.
The number of F2 seeds analyzed for fatty
acid composition from each cross is
shown in Table 1. Thirty seeds from each
parent were also analyzed.
For the determination of phenotypic ratios, the entire F2 seed was used for Bay
3 J10, Bay 3 C1727, and Bay 3 KK7. The
F2 seeds for the other crosses were cut
into two parts, as described in the analysis of F1 seeds. The part of each F2 seed
with the embryonic axis was used for
planting, and the part lacking the embryonic axis was used for fatty acid analysis.
The identity of all F2 seeds and their progeny was maintained during fatty acid analysis, planting, and harvest.
All F2 seeds with the embryonic axis and
seeds from the parents for these crosses,
Table 3. Range of palmitate content of parent and F 2 seeds
Parent or
cross
Range of
palmitate
content
(%)
Bay (P1)
J10 (P2)
F2 seeds
Bay (P1)
C1727 (P2)
F2 seeds
Bay (P1)
KK7 (P2)
F2 seeds
C1726 (P1)
J10 (P2)
F2 seeds
C1726 (P1)
KK7 (P2)
F2 seeds
9.5–11.1
14.9–16.8
9.3–16.3
9.5–11.1
14.8–16.9
9.5–16.2
9.5–11.1
12.5–14.2
9.7–14.5
6.5–7.9
14.9–16.8
6.6–16.5
6.5–7.9
12.5–14.2
6.3–14.3
Frequency distribution
x2a
5P1
.P1 to ,P2 5P2
3:1
18
42
20
0.30
.0.75
25
56
19
2.16
.0.25
54
26
b
2
P
1:2:1
x2c
P
1:14:1
P
.0.10
7
53
4
27.84
,0.001
2.41
.0.25
6
53
5
27.59
,0.001
1.41
.0.25
x for goodness-of-fit to a 3:1 ratio.
x2 for goodness-of-fit to a 1:2:1 ratio.
c 2
x for goodness-of-fit to a 1:14:1 ratio.
a
2.40
x2b
KK7 were selected. Twelve F3 seeds from
each F2 plant and from each of two plants
of Bay, J10, and C1727 were planted in the
field in July 1996. The F3 plants of each line
and plants of the parents were harvested
individually. A 20-seed sample from 10 F3
plants for each line and from 10 plants for
each parent was analyzed for fatty acid
composition.
Fatty acid composition was determined
by gas chromatography, as described by
Takagi et al. (1989). Chi-square analyses
were calculated to test the best fit of the
data to the expected genetic ratio. A single
gene model was used to evaluate the segregation ratio for palmitate in F2 seeds of
Bay 3 J10, Bay 3 C1727, and Bay 3 KK7.
In other crosses, except C1727 3 J10, a
two-gene model was used for the evaluation of F2 seeds and plants. No genetic
model was assumed for C1727 3 J10.
Figure 2. Distribution of palmitate content in seeds of C1727 and KK7 and in F2 seeds of C1727 3 KK7.
Results and Discussion
Figure 3. Distribution of palmitate content in seeds of C1727 and J10 and in F2 seeds of C1727 3 J10.
fore 15 individual F3 seeds from each of
these F2 plants and 15 seeds each from
C1727 and J10 were analyzed. None of the
F2 plants had all F3 progeny with a palmitate content higher than that of C1727 and
J10. Duncan’s multiple range test was used
to compare the means of palmitate in putative lower and higher F2 plants with the
parental plants.
The palmitate contents of the parents
grown under the same field conditions as
the F1 plants were used to classify F2
seeds. The range of palmitate content of
each parent was used to determine the parental classes for each cross, for example,
an F2 seed was considered similar to one
parent (P1) when its palmitate content
was within the range exhibited by that P1
seed. For example, in Bay 3 J10, the F2
seeds were classified as 5P1, .P1 to ,P2,
and 5P2. The same classification was used
to evaluate the segregation of F3 progeny
from individual F2 plants.
For further confirmation of the geno-
types of segregates with a palmitate content similar to that of Bay and higher than
that of C1727 or J10, five F2 plants from
both classes of J10 3 KK7 and C1727 3
Maternal effects for palmitate content
were not observed in the analysis of F1
seeds from reciprocal crosses ( Table 2).
Thus the genotype of the embryo determined its palmitate content and not the
genotype of the maternal plant. The lack
of maternal effects made it possible to
evaluate genotypic differences among segregating seeds from heterozygous plants.
No dominance effects for palmitate content were observed in Bay 3 J10 and
C1726 3 KK7. In these crosses, the F1
mean was significantly different from the
parents, but not significantly different
from the midparent value ( Table 2). There
was partial dominance for palmitate content in Bay 3 C1727, Bay 3 KK7, C1726 3
J10, and C1727 3 KK7. The four crosses
Table 4. Classification of 64 F 2 plants from the cross C1726 (fap1fap1Fap2Fap2) 3 J10
(Fap1Fap1fap2fap2) based on the phenotypic pattern of 15 F 3 seeds from each F2 plant
Genotypic frequency of
F2 plants
Expected F3 phenotypic patternb
Proposed F2 genotype
Expected
5P1c
fap1fap1Fap2Fap2
fap1fap1Fap2fap2 or
Fap1fap2Fap2Fap2
fap1fap1fap2fap2 or
Fap1Fap1Fap2Fap2
Fap1fap1Fap2fap2
Fap1fap1fap2fap2 or
Fap1Fap1Fap2fap2
Fap1Fap1fap2fap2
4
4
3
16
22
3
3
8
16
10
13
3
3
3
16
4
12
3
Observeda
.P1 to ,P2
3
5P2
3
3
3
Observed genotypic frequency based on the F3 progeny evaluation satisfactorily fit the expected frequency based
on a chi-square test (x2 5 4.38, P . .25).
b
The expected F3 phenotypic patterns are based on a model for alleles with additive gene action at two independent
loci controlling palmitate content, which predicts a 1:4:2:4:4:1 F2 genotypic ratio.
c
P1 5 C1726 (6.8–8.0% palmitate) and P2 5 J10 (15.4–17.3% palmitate).
a
Brief Communications 425
Table 5. Classification of 64 F 2 plants from the cross C1726 (fap1fap1FapxFapx) 3 KK7
(Fap1Fap1fapxfapx) based on the phenotypic pattern of 15 F3 seeds from each F 2 plant
Proposed F2 genotype
fap1fap1FapxFapx
fap1fap1Fapxfapx or
Fap1fap1FapxFapx
fap1fap1fapxfapx or
Fap1Fap1FapxFapx
Fap1fap1Fapxfapx
Fap1fap1fapxfapx or
Fap1Fap1Fapxfapx
Fap1Fap1fapxfapx
Genotypic frequency of
F2 plants
Expected F3 phenotypic patternb
Expected
5P1c
Observed
a
.P1 to ,P2
4
5
3
16
12
3
3
8
16
7
21
3
3
3
16
4
17
2
3
5P2
3
3
3
Observed genotypic frequency based on the F3 progeny evaluation satisfactorily fit the expected frequency based
on a chi-square test (x2 5 4.00, P . .50).
b
The expected F3 phenotypic patterns are based on a model for alleles with additive gene action at two independent
loci controlling palmitate content, which predicts a 1:4:2:4:4:1 F2 genotypic ratio.
c
P1 5 C1726 (6.8–8.0% palmitate) and P2 5 KK7 (13.7–15.0% palmitate).
a
had F1 seed with mean palmitate contents
that differed significantly from the parents
and were significantly lower than the midparent value. For C1727 3 J10, the mean
palmitate content of F1 seeds was similar
to C1727, indicating complete dominance.
An exceptional result was observed in J10
3 KK7. The mean palmitate in this cross
was significantly lower than the parents
and it tended toward the value of Bay ( Table 2). Thus it appears that the alleles in
these lines were at different loci.
The lack of reciprocal differences in F2
seeds for palmitate content indicated no
cytoplasmic effects in any of the crosses,
and consequently the data for reciprocal
F2 seeds from each cross were combined.
The range of palmitate content for the
parents ( Table 3) grown in the same environment as the F1 plants was used to determine the parental classes of F2 seeds
for each cross. The palmitate contents in
F2 seeds of Bay 3 J10 and Bay 3 C1727
were distributed into three divisions. The
observed data of 18:42:20 for Bay 3 J10
and 25:56:19 for Bay 3 C1727 satisfactorily
fit to a phenotypic ratio of 1:2:1 ( Table 3).
The segregation of the F2 seeds from Bay
3 KK7 showed two divisions for palmitate
content and the data of 54:26 satisfactorily
fit a 3:1 ratio. The results of these crosses
indicated that palmitate content in each of
J10, C1727, and KK7 was controlled by a
recessive allele at a single locus. The segregation of the F2 seeds from C1726 3 J10
and C1726 3 KK7 showed three divisions
for palmitate content. The data from each
of two crosses was initially evaluated with
a 1:2:1 ratio that would be expected with
two alleles at one locus, and none of the
crosses satisfactorily fit the ratio. When
the segregation patterns were evaluated
with a 1:14:1 ratio that would be expected
with alleles at two independent loci, both
Table 6. Classification of 176 F 2 plants from the cross J10 (fap2fap2FapxFapx) 3 KK7
(Fap2Fap2fapxfapx) based on the phenotypic pattern of 15 F3 seeds from each F 2 plant
Genotypic frequency of
F2 plants
Expected F3 phenotypic patternb
Proposed F2 genotype
Expected
.P1c
fap2fap2fapxfapx
fap2fap2Fapxfapx or
Fap2fap2fapxfapx
fap2fap2FapxFapx or
Fap2Fap2fapxfapx
Fap2fap2Fapxfapx
Fap2fap2FapxFapx or
Fap2Fap2Fapxfapx
Fap2Fap2FapxFapx
11
8
3
44
38
3
3
22
44
25
47
3
3
3
44
11
48
10
Observed
a
5 P1 to P2
3
,P2
3
3
3
Observed genotypic frequency based on the F3 progeny evaluation satisfactorily fit the expected frequency based
on a chi-square test (x2 5 2.70, P . .50).
b
The expected F3 phenotypic patterns are based on a model for alleles with additive gene action at two independent
loci controlling palmitate content, which predicts a 1:4:2:4:4:1 F2 genotypic ratio.
c
P1 5 J10 (15.4–17.3% palmitate) and P2 5 KK7 (13.7–15.0% palmitate).
a
426 The Journal of Heredity 1999:90(3)
of the crosses had a satisfactory fit ( Table
3). The F2 seeds from J10 3 KK7 and C1727
3 KK7 were also segregated into three divisions ( Figures 1 and 2), but the divisions
were .P1, 5P1 to P2, and ,P2. The observed data of 29:110:37 for J10 3 KK7
( Figure 1) and 37:114:41 for C1727 3 KK7
( Figure 2) were evaluated with all possible
ratios for a two-gene model, and none of
the expected ratios satisfactorily fit the
observed ratio. However, the transgressive segregation in these crosses indicated
that alleles in J10 and KK7 were at a different loci and also that alleles in C1727
and KK7 were at a different loci; J10 and
C1727 could be the same locus.
A progeny test was conducted for palmitate content in C1726 3 J10, C1726 3
KK7, J10 3 KK7, and C1727 3 KK7 by analyzing F3 seeds from each F2 plant. The
palmitate content of the parents grown in
the same environment as the F2 plants was
used to evaluate the segregation among F3
seeds from each F2 plant for each of these
crosses. The range of palmitate content
for each parent ( Tables 4–7) was used to
determine the parental classes. In C1726 3
J10 and C1726 3 KK7, the F3 seeds were
classified as 5P1, .P1 to ,P2, and 5P2 ( Tables 4 and 5). In J10 3 KK7 and C1727 3
KK7, the F3 seeds were classified as .P1,
5P1 to P2, and ,P2 ( Tables 6 and 7). The
segregation of F3 seeds from these crosses
supported the theoretical model of two alleles for palmitate content that occur at
independent loci, with the alleles at each
locus exhibiting additive gene action. According to this model, six patterns would
be expected for F3 seeds from F2 plants.
These six patterns would be expected to
occur in a genotypic ratio of 1:4:2:4:4:1.
The observed genotypic frequencies of 64
F2 plants from each of C1726 3 J10 and
C1726 3 KK7 were 4:22:10:13:12:3 and 5:12:
7:21:17:2, respectively, which satisfactorily
fits the expected ratio of 4:16:8:16:16:4 ( Tables 4 and 5). The results of these crosses
indicated that the alleles controlling palmitate content in J10 and KK7 were at a
different locus from the fap1 allele in
C1726 ( Erickson et al. 1988). The observed genotypic frequency of 176 F2
plants from J10 3 KK7 was 8:38:25:47:48:
10, which satisfactorily fit the expected ratio of 11:44:22:44:44:11 ( Table 6). The genotypic frequency of 192 F2 plants from
C1727 3 KK7 was 9:53:27:47:42:14, which
satisfactorily fit the expected ratio of 12:
48:24:48:48:12 ( Table 7). The results of
these crosses indicated that the allele for
palmitate content in KK7 was at a different
locus from the sop2 allele in J10 (Rahman
Table 7. Classification of 192 F 2 plants from the cross C1727 (fap2fap2FapxFapx) 3 KK7
(Fap2Fap2fapxfapx) based on the phenotypic pattern of 15 F3 seeds from each F 2 plant
Genotypic frequency of
F2 plants
Expected F3 phenotypic patternb
Proposed F2 genotype
Expected
.P1c
fap2fap2fapxfapx
fap2fap2Fapxfapx or
Fap2fap2fapxfapx
fap2fap2FapxFapx or
Fap2Fap2fapxfapx
Fap2fap2Fapxfapx
Fap2fap2FapxFapx or
Fap2Fap2Fapxfapx
Fap2Fap2FapxFapx
12
9
3
48
53
3
3
24
48
27
47
3
3
3
48
12
42
14
Observed
a
5 P1 to P2
3
,P2
3
3
3
Observed genotypic frequency based on the F3 progeny evaluation satisfactorily fit the expected frequency based
on a chi-square test (x2 5 2.75, P . .50).
b
The expected F3 phenotypic patterns are based on a model for alleles with additive gene action at two independent
loci controlling palmitate content, which predicts a 1:4:2:4:4:1 F2 genotypic ratio.
c
P1 5 C1727 (15.9–17.5% palmitate) and P2 5 KK7 (13.7–15.0% palmitate).
a
et al. 1996) and the fap2 allele in C1727
( Erickson et al. 1988).
The F2 seeds and their F2 plants with putative lower and higher palmitate from
C1727 3 J10 were evaluated to determine
whether the alleles in C1727 and J10 were
at the same locus or different loci. The F2
seeds of this cross did not segregate for
palmitate content ( Figure 3). Therefore no
genetic ratio for allelic effects could be applied. Moreover, none of the F2 seeds had
palmitate content transgressively lower or
higher than the parents. The mean palmitate content of F2 seeds was similar to that
of C1727 and J10 (data not shown), but the
range was slightly large, compared to the
parental ranges ( Figure 3). The mean palmitate of putative lower or higher F2
Figure 4. Distribution of palmitate content in plants of J10 and Bay and in F3 plants with Fap2Fap2FapxFapx and
fap2fap2fapxfapx genotypes from J10 3 KK7.
plants did not differ significantly from the
parental plants, based on the Duncan’s
multiple range test. Therefore the extension observed in the range of F2 seeds was
due to environmental effects, and not due
to additional major gene effects, which
would confirm that C1727 and J10 had the
same allele at a single locus.
The results of this study indicated that
J10 had an allele at a different independent locus than C1726, but at the same locus as C1727, and KK7 had an allele at a
different locus than C1726, C1727, and J10.
Therefore the temporary designation of
sop2 for the allele in J10 has permanently
been designated by fap2 (J10), and the allele in KK7 has temporarily been designated by fapx because it’s relationship to fap3
allele in A22 ( Fehr et al. 1991b) and fap4
allele in A24 (Schnebly et al. 1994) was unknown. Further studies will be needed to
determine if the allele in KK7 is at fap3 or
fap4.
On the basis of the F3 progeny test, 10
F2 plants with the genotype Fap2Fap2FapxFapx and 8 F2 plants with the genotype fap2fap2fapxfapx were found in J10 3
KK7 ( Table 6), and 14 F2 plants with the
genotype Fap2Fap2FapxFapx and 9 F2 plants
with the genotype fap2fap2fapxfapx were
found in C1727 3 KK7 ( Table 7). In J10 3
KK7, the F3 plants with the genotype
Fap2Fap2FapxFapx had a mean palmitate
content of 11.3% and a range of 10.5–
12.6%, compared with a mean of 11.5% and
a range of 10.7–12.2% for the Bay. The F3
plants with the genotype fap2fap2fapxfapx
had a mean of 22.5% and a range of 21.1–
24.4%, compared with a mean of 17.5% and
a range of 16.2–18.5% for the J10 ( Figure
4). In C1727 3 KK7, the F3 plants with the
genotype Fap2Fap2FapxFapx had a mean
palmitate content of 11.9% and a range of
10.5–12.8%, compared with a mean of
11.5% and a range of 10.7–12.2% for the
Bay. The F3 plants with the genotype fap2fap2fapxfapx had a mean of 22.9% and a
range of 21.0–24.8%, compared with a
mean of 17.3% and a range of 16.3–18.4%
for the C1727 ( Figure 5). The lines with
.22.0% of palmitate content obtained in
this study could potentially be used for
specific edible purposes.
From the Department of Genetics and Breeding, Rajshahi University, Rajshahi, Bangladesh (Rahman), Saga
Prefectural Agricultural Research Center, Nanri Kawasoe, Saga, Japan ( Kinoshita), and Laboratory of Plant
Breeding, Faculty of Agriculture, Saga University, Saga
840–0027, Japan (Anai and Takagi). We express sincere
thanks to Dr. J. R. Wilcox, Purdue University, for providing the seeds of C1726 and C1727. Address correspondence to Dr. Y. Takagi at the address above.
Figure 5. Distribution of palmitate content in plants of C1727 and Bay and in F3 plants with Fap2Fap2FapxFapx
and fap2fap2fapxfapx genotypes from C1727 3 KK7.
q 1999 The American Genetic Association
Brief Communications 427
References
Bubeck DM, Fehr WR, and Hammond EG, 1989. Inheritance of palmitic and stearic acid mutants of soybean.
Crop Sci 29:652–656.
Erickson EA, Wilcox JR, and Cavins JF, 1988. Inheritance of altered palmitic acid percentage in two soybean mutants. J Hered 79:465–468.
Fehr WR, Welke GA, Hammond EG, Duvick DN, and
Cianzio SR, 1991a. Inheritance of elevated palmitic acid
content in soybean seed oil. Crop Sci 31:1522–1524.
Fehr WR, Welke GA, Hammond EG, Duvick DN, and
Cianzio SR, 1991b. Inheritance of reduced palmitic acid
content in seed oil of soybean. Crop Sci 31:88–89.
Knowles PE, 1989. Safflower. In: Oil crops of the world,
their breeding and utilization (Robbelen G, et al., eds).
New York: McGraw-Hill; 363–374.
Ohlrogge JB, Browse J, and Somerville C, 1991. The genetics of plant lipids. Biochem Biophys Acta 1082:1–26.
Rahman SM, Takagi Y, and Kinoshita T, 1996. Genetic
analysis of palmitic acid contents using two soybean
mutants, J3 and J10. Breed Sci 46:343–347.
Rattray J, 1991. Plant biotechnology and the oils and
fats industry. In: Biotechnology of plant fats and oils
(Rattray J, ed). Champaign, Illinois: American Oil and
Chemical Society; 1–35.
Robbelen G, Downey RK, and Ashri A, 1989. Oil crops
of the world, their breeding and utilization (Robbelen
G, et al., eds). New York: McGraw-Hill.
Schnebly SR, Fehr WR, Welke GA, Hammond EG, and
Duvick DN, 1994. Inheritance of reduced and elevated
palmitate in mutant lines of soybean. Crop Sci 34:829–
833.
Stefansson BR, Hougen FW, and Downey RK, 1961. Note
on the isolation of rape plants with seed oil free from
erucic acid. Can J Plant Sci 41:218–219.
Takagi Y, Hossain ABMM, Yanagita T, and Kusaba S,
1989. High linolenic acid mutant in soybean induced by
X-ray irradiation. Jpn J Breed 39:403–409.
Takagi Y, Rahman SM, Joo H, and Kawakita T, 1995. Reduced and elevated palmitic acid mutants in soybean
developed by X-ray irradiation. Biosci Biotechnol Biochem 59:1778–1779.
plants that were aneuploid (2n 5 17). Cbanding analysis of root-tip cells showed
that the extra chromosomes were highly
heterochromatic B chromosomes. The
Giemsa banding pattern of the A-chromosome complement in the aneuploid
cells was not altered from the banding patterns of normal cells.
The occurrence of B chromosomes has
been reported in many plants and animals
(Jones and Rees 1982). They are different
from normal or A chromosomes and are
also termed supernumerary or accessory
chromosomes. B chromosomes typically
have the following characteristics: they
are usually smaller than A chromosomes
and are generally heterochromatic; they
normally do not influence the viability and
phenotype of the organism; they vary between different cells, tissues, individuals,
and populations; they do not pair with A
chromosomes; and they affect mitotic behavior by lagging and elimination, polymitosis, or preferential distribution (Rieger et al. 1991).
During a routine cytogenetic investigation of diploid M. sativa ssp. falcata ( L.)
Arcangeli accessions from the U.S. National Plant Germplasm System, we discovered two accessions in which three plants
each were aneuploid with 2n 5 17. Masoud et al. (1991) showed that it was pos-
sible to detect heterochromatin of alfalfa
using the C-banding technique. The purpose of this study was to determine the
nature of the extra chromosome using Cbanding.
Materials and Methods
Seeds of 13 accessions of diploid M. sativa
ssp. falcata ( Table 1) were obtained from
the U.S. National Plant Germplasm System,
Pullman, Washington. One accession, UAG
1806, was obtained from the Karl Lesins
collection ( E. Small, Agriculture and AgriFood Canada, Ottawa, Ontario, Canada).
Two accessions, PI 115365 and PI 486207,
contained aneuploid plants (2n 5 17).
Both accessions originated in Russia and
were originally obtained from the Vavilov
Institute, St. Petersburg, Russia.
Root tips were obtained 3 days after germination was initiated, pretreated in an
ice bath for 20 h, and fixed in Farmer’s fixative (3:1, v/v, 95% ethanol : glacial acetic
acid) for at least 30 min. The C-banding
procedure of Bauchan and Hossain (1997)
for alfalfa was followed. Thirty-five seedlings per accession were studied and a
minimum of 50 cells were observed from
each seedling. Photomicrographs were
taken using a Zeiss Axiophot Microscope
using Kodak Technical Pan 2415 film. The
photomicrograph was obtained through
Vles RO and Gottenbos JJ, 1989. Nutritional characteristics and food uses of vegetable oils. In: Oil crops of
the world, their breeding and utilization (Robbelen G,
et al., eds). New York: McGraw-Hill; 63–86.
Wilcox JR, Burton JW, Rebetzke GT, and Wilson RF,
1994. Transgressive segregation for palmitic acid in
seed oil of soybean. Crop Sci 34:1248–1250.
Received May 7, 1998
Accepted December 7, 1998
Corresponding Editor: William Tracy
Identification of B
Chromosomes Using Giemsa
Banding in Medicago
M. A. Hossain and G. R. Bauchan
This is the first verified report of the existence of B chromosomes in the genus
Medicago. During a routine cytogenetic investigation of diploid M. sativa ssp. falcata
(L.) Arcangeli accessions obtained from
the U.S. National Plant Germplasm System, we discovered two accessions, PI
115365 and PI 486207, each with three
428 The Journal of Heredity 1999:90(3)
Figure 1. The normally C-banded A chromosomes from PI 486207 plus the highly heterochromatic B chromosome
(arrow).
Table 1. List of accessions observed and the
country where the accession was collected
Accession number
Country of origin
PI 115365
PI 262532
PI 263154
PI 307398
PI 405064
PI 440527
PI 464727
PI 464728
PI 467970
PI 486207
PI 494662
PI 577557
UAG 1806b
Russia
Israel
Russia
Sweden
USA
Russia
Turkey
Turkey
USA
Russia
Romania
Bulgaria
Canada
a
The plant introductions (PI) were obtained from the
U.S. Plant Introduction Station in Pullman, Washington.
b
UAG 1806 provided by E. Small from the Karl Lesins
Collection, Agriculture and Agri-Food Canada, Ottawa,
Ontario, Canada.
a
the use of a computerized image analysis
system as described by Bauchan and
Campbell (1994).
Results and Discussion
Three individual seedlings from two different accessions contained a B chromosome. The interphase nucleus of a normal
C-banded cell contains small darkly
stained heterochromatin blocks located
on either side of the centromere. In the
aneuploid plant a single, condensed heterochromatic structure in addition to normal centromeric blocks of heterochromatin was detected ( Figure 1). The Giemsa
banding pattern of the A chromosome
complement in the aneuploid cells was
not altered from the banding patterns in
normal cells.
There have been two apparently erroneous reports of B chromosomes in the
genus Medicago. The first was in M. granadensis Willd. in plate 1 in Heyn’s (1963)
monograph. The second was in M. ciliaris
All., M. intertexta Mill., M. littoralis Rohde,
and M. murex Willd. (Agarwal and Gupta
1983). From the photomicrographs given
in these two reports, it appears that the
structures reported as B chromosomes
were the detached satellites of satellited
chromosomes. Many Medicago species
have chromosomes that have large nucleolus organizer regions ( NOR) which can
cause the satellited portion of the chromosome to be located away from the main
body of the chromosome. The satellites of
the species M. rugosa Desr. and M. scutellata ( L.) Mill. were incorrectly identified as
whole chromosomes until Bauchan and Elgin (1984) determined that they were satellites and thus corrected the chromosome number for these species to 2n 5 30.
In conclusion, this is the first report of
the existence of a B chromosome in the
genus Medicago. We are uncertain about
the origin of the B chromosomes as the Achromosome complement does not appear to have a reduction in DNA content
as observed by the normal C-banding pattern of the A chromosomes. We isolated
two plants, one from each accession, from
a total of 500 seedlings possessing a B
chromosome. However, both of the plants
died before flowering, thus we have not
been able to study meiosis in these plants.
We are unsure if the plants died due to
genetic load, environmental stresses, or
both.
From the U.S. Department of Agriculture, Agricultural
Research Service, Plant Sciences Institute, Soybean and
Alfalfa Research Laboratory, Beltsville, MD 20705-2350
( Bauchan) and the University of Maryland, Natural Resources and Landscape Architecture Department, College Park, Maryland ( Hossain). Mention of a trademark
or proprietary product does not constitute a guarantee
or warranty by the U.S. Department of Agriculture and
does not imply its approval to the exclusion of other
products that may also be suitable.
q 1999 The American Genetic Association
References
Agarwal K and Gupta PK, 1983. Cytogenetic studies in
the genus Medicago Linn. Cytologia 48:781–793.
Bauchan GR and Campbell TA, 1994. Use of an image
analysis system to karyotype diploid alfalfa (Medicago
sativa L.). J Hered 85:18–22.
Bauchan GR and Elgin JH, 1984. A new chromosome
number for the genus Medicago. Crop Sci 24:193–195.
Bauchan GR and Hossain MA, 1997. Karyotypic analysis of C-banded chromosomes of diploid alfalfa: Medicago sativa ssp. caerulea and ssp. falcata and their hybrid. J Hered 88:533–536.
Heyn CC, 1963. The annual species of Medicago. Jerusalem, Israel: Hebrew University of Jerusalem.
Jones RN and Rees H, 1982. B chromosomes. London:
Academic Press.
Masoud SA, Gill BS, and Johnson LB, 1991. C-banding
and alfalfa chromosomes: standard karyotype and analysis of a somaclonal variant. J Hered 82:335–338.
Rieger R, Michaelis A, and Green MM, 1991. Glossary
of genetics: classical and molecular, 5th ed. New York:
Springer-Verlag; 50.
Received March 3, 1998
Accepted December 7, 1998
Corresponding Editor: Prem Jauhar
Allozyme Polymorphisms
Discriminate Among Coast
Redwood (Sequoia
sempervirens) Siblings
D. L. Rogers
The ability to identify genotypes of coast
redwood [Sequoia sempervirens (D. Don)
Endl.] using allozyme polymorphisms was
tested using pedigreed germplasm. In
this, the first study of genotypic identification in this species, 10 allozyme loci
were found to be more than sufficient to
discriminate among closely related (selfedsib) individuals within four pedigreed families. Commonly, selfed-sibs within each
family differed at four or five loci. Thus the
occurrence of type II errors (i.e., erroneously assigning nonidentical sibs to the
same clone) was found to be zero. The
probability of type I errors is discussed
and assumed to be small. Some traits of
coast redwood, such as its hexaploid condition and (supposedly) largely outcrossing breeding system, further suggest that
allozyme polymorphisms may adequately
discriminate among clones in natural populations. With further improvement in technique, such that dosages of various alleles
could be discriminated, isozyme analysis
could prove even more powerful for clonal
identification in this species. Clonal identification is a prerequisite for unbiased
studies of genetic structure and mating
systems in natural populations of asexually reproducing plant species.
In population biology, the basic unit of
concern is the individual (Schmid 1986).
However, in the case of clonal plants, the
individual may be represented by numerous growth units such as tillers or branches—the common sampling entity. Because
of this potential confusion between sampling units and genetically distinct individuals, the ability to discriminate among individuals is fundamental to studies of spatial patterning of genetic variation, gene
flow, demography, mating systems, and
natural selection in asexually reproducing
plant species. Failure to distinguish clonal
growth units from individuals can seriously confound descriptions of genetic structure and estimates of population parameters such as inbreeding (e.g., Lokker et al.
1994).
In field studies, clonal identification may
be challenging. Two related issues are
What is the most appropriate method?
and What is the resolving power of this
Brief Communications 429
method? Before the advent of biochemical
or molecular markers, putative clones
were identified by root excavation or by
otherwise following physical connections.
Such methods are labor intensive, expensive, appropriate only for certain (i.e., stoloniferous or rhizomatous) species, and
subject to pitfalls. Tangled webs of roots
or rhizomes may make physical identification difficult (e.g., Carex lasiocarpa;
McClintock and Waterway 1993), and
small sections of severed rhizomes may
produce new shoots (e.g., Calamagrostis
canadensis; MacDonald and Lieffers 1991)
or connections among ramets may disintegrate over time (e.g., Vallisneria americana; Lokker et al. 1994), thereby creating
disjunct clones. Furthermore, root grafting
of plants can occur, causing mistaken
identification as members of a clone. Finally, the destructive process of root excavation itself may jeopardize clonal identity.
Molecular markers are often employed
in the identification of domestic germplasm, such as clones, varieties, or cultivars. For example, Demeke et al. (1993)
made use of random amplified polymorphic DNAs (RAPDs) to distinguish commercial potato cultivars and clonal variants of cultivars. Similarly a DNA typing
system using sequence-tagged microsatellite site markers has been developed to
differentiate cultivars of grapevine (Vitis
vinifera and related Vitis species; Thomas
et al. 1994). However, in agricultural or
horticultural applications, commonly the
number of genotypes is finite and the
range of variation known. Hence, in those
applications, genotypic identification can
be more confidently conferred. The use of
pedigreed germplasm as a measure of type
II error (i.e., falsely identifying different genotypes as members of one clone) in natural plant populations has been limited in
the past as this material is generally not
available for most clonal plant populations studied. Pedigreed germplasm offers
two advantages in such an investigation.
First, as known distinct genotypes, there
is an absolute measure of type II error.
Second, the (known) degree of relationship among the genotypes can be indirectly used as a measure of the resolving
power of the genetic markers.
Coast redwood (Sequoia sempervirens D.
Don) is a culturally, economically, and ecologically significant forest tree species
with a narrow natural range along the
western coast of Oregon and California. It
reproduces asexually to some extent in nature, although the factors affecting recruit-
430 The Journal of Heredity 1999:90(3)
ment of sexually and asexually derived
members are not well understood. Furthermore, although circular configurations
of redwoods commonly are assumed to be
clonal, this assumption requires genetic
corroboration. The ability to identify
clones in nature would contribute to our
understanding of the reproductive biology
of the species and provide a basis for
studies of genetic structure. This species
possesses an unusual characteristic for a
gymnosperm—that of being a hexaploid—
making it a likely candidate for successful
clonal identification via allozyme polymorphisms, as well as compelling scientific
and socioeconomic reasons to investigate
this possibility.
The objective of this study was to determine the suitability of allozyme polymorphisms for distinguishing clones of
coast redwood, taking advantage of the
availability of pedigreed genetic materials
to test for the occurrence of type II errors.
Materials and Methods
Four selfed families of coast redwood were
produced by controlled pollination in the
late 1970s, and the resulting selfed-sib
progeny were installed in a hedge orchard
in 1981 ( Libby et al. 1981). The four parental trees originated in Humboldt County, California, and included the charismatic individual ARC154, the world’s tallest
tree. The other three parent trees were
designated R37, S1, and R17. Further details concerning the breeding activities
and handling of seeds and seedlings may
be obtained in Libby et al. (1981). Stecklings (i.e., field-plantable rooted cuttings)
were obtained from the selfed-sib seedlings and incorporated into a field trial
and hedge orchard located at the University of California’s Russell Reservation, located 15 km east of Berkeley, California.
For the present study I collected foliar
samples from the Russell field trial, 12
years old at the time of sampling, supplementing the samples when necessary with
hedge orchard material. Nine (selfed-sib)
trees were available for two families, R37
and S1, and eight were available for each
of the other two.
Foliar samples were prepared and electrophoretically studied according to procedures described in Rogers (1994). Previous studies (Rogers 1994) showed no
difference in individual clones’ isozyme
banding patterns among their respective
ramets of different ages (since cloning) or
growing on different sites. Thus there was
assumed to be no impact in this study
from sampling in both the field trial and
the hedge orchard. Seven enzyme systems
were studied—all having been shown previously to produce high-quality banding
patterns consistently in foliar samples
(Rogers 1994): isocitrate dehydrogenase
( IDH), malate dehydrogenase (MDH), 6phosphogluconate dehydrogenase (6-PGD),
shikimate dehydrogenase (SKD), phosphoglucose isomerase (PGI), triose phosphate isomerase ( TPI), and UDP-glucose
pyrophosphorylase ( UGP). Twelve isozyme loci were scored, 10 of which were
polymorphic in this set of families. A complete dataset was highly desirable for this
study. Most gels produced clear and distinct bands for most assays. However,
when interpretation was difficult, assays
were repeated until clear and consistent
results were obtained.
Inheritance has been studied using fullsib pedigreed seeds for six of the seven
enzyme systems in earlier analyses, and
results were consistent with hexasomic inheritance (Rogers 1997). The seventh enzyme system, TPI, was not part of that
study as it stained too faintly in seed tissues to be interpretable. Although a robust inheritance study for this enzyme is
still required, confidence was given to its
genetic interpretation with a modest fullsib foliar study (Rogers 1994).
A 10-locus genotype was assigned to
each individual. Allozyme genotypes were
scored conservatively, noting the presence of alleles but not their dosage. Within
each of the four selfed-sib families the genotype of each individual was compared
with that of every other, resulting in 28
pairwise comparisons for the eight-member families [n(n 2 1)/2], where n is the
number of family members, and 36 for
those with 9 members. For each comparison, the number of differing loci was noted. This array of difference scores for each
family was then converted to a relative frequency distribution so as to allow comparisons among the four families.
The type II error (i.e., erroneously assigning nonidentical sibs of coast redwood
to the same multilocus genotype) was calculated as the ratio of repeated genotypes
to total sample size. This value is sometimes used as an estimate for the otherwise incalculable type II error in samples
from natural populations (e.g., Hunter
1993). However, in the present study, since
the genotypic identity of every sample
was known a priori, this error can be calculated directly.
Results and Discussion
Table 1. Allozyme signatures of the 34 clones of coast redwood assayed for this study a
Allozyme locus
Parent
Clone
1
2
3
4
5
6
7
8
9
10
S1
22
23
26
31
32
33
35
45
13
6
7
8
12
3
9
10
11
2
3
12
13
15
16
19
21
11
2
9
14
15
16
20
24
21
11
12
11
11
12
12
11
12
11
12
12
12
12
13
13
13
13
12
12
11
11
12
12
13
12
11
11
12
11
11
13
11
11
12
12
123
123
11
13
13
123
123
13
13
12
123
123
123
12
123
12
12
12
12
12
12
12
123
13
11
12
12
12
12
12
12
11
12
11
11
11
11
11
12
11
11
12
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
12
11
12
11
12
12
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
12
12
11
12
11
11
12
13
14
11
11
13
14
13
11
11
11
11
11
11
12
12
12
12
13
11
11
13
13
11
11
13
13
12
12
12
11
12
12
12
11
12
12
12
12
11
11
11
12
12
23
23
23
22
23
23
23
23
11
11
23
23
23
23
23
23
11
23
11
12
23
23
11
12
11
11
11
12
11
12
12
12
12
11
11
12
12
12
12
12
12
11
11
11
12
12
12
11
12
12
12
11
12
12
11
12
11
12
12
11
11
11
11
13
11
11
11
11
11
11
11
11
11
11
13
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
12
11
12
12
11
12
12
11
12
12
12
14
11
12
12
12
12
12
11
12
11
12
11
12
12
11
12
11
12
12
11
11
11
11
13
11
11
11
13
11
13
11
13
13
13
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
13
11
11
11
11
13
11
11
R17
R37
ARC154
a
b
b
Only the (10) polymorphic loci are presented. PGI-1 and TPI-1 are monomorphic for this set of clones.
Key: 1 5 isocitrate dehydrogenase; 2 5 shikimate dehydrogenase-1; 3 5 shikimate dehydrogenase-2; 4 5 UDPglucose pyrophosphorylase; 5 5 phosphoglucose isomerase-2; 6 5 malate dehydrogenase-1; 7 5 malate dehydrogenase-2; 8 5 malate dehydrogenase-3; 9 5 6-phosphogluconate dehydrogenase; 10 5 triose phosphate isomerase-2.
Figure 1. The relative frequency number of nonmatching loci between pairs of selfed-sib progeny in four selfed
families of coast redwood. Comparisons are based on 10-locus genotypes.
In this clonal identification study, 10 polymorphic isozyme loci were found to be
sufficient to uniquely identify every individual within four selfed families of coast
redwood ( Table 1). The occurrence of
type II errors was thus 0%. The number of
nonidentical loci among individuals in
three of the four families approximates a
normal distribution (family S1 does not),
with a mode of four or five nonidentical
loci between individuals ( Figure 1). Even
at a single locus there was often considerable variation within the selfed-sib families.
Certain features of coast redwood enhance the probability of distinguishing
clones on the basis of allozyme differences. In the first allozyme study of natural populations a high degree of polymorphism has been revealed within populations (Rogers 1994). Second, its hexaploid
nature confers a great range in possible
allozyme genotypes. While the addition of
each new allele increases the number of
possible diploid genotypes in a fundamentally additive fashion, it increases the
number more quickly in hexaploids. For
example, with four alleles at a locus, a diploid organism could display a maximum of
10 single-locus genotypes and a hexaploid,
74 (i.e., if allele dosage could also be determined). In situations such as this study,
where genotypes could only be distinguished that differed in the inclusion/exclusion of alleles (not dosage), the comparable number of genotypes is 15. See
Rogers (1994) for further discussion of allele dosage visualization in coast redwoods. Furthermore, with the addition of
loci to the genotypic characterization, the
hexaploid possibilities expand even more
rapidly relative to the diploid. Thus hexaploidy strongly enhances the resolving
power of differentiating individuals in
coast redwood.
Another feature of coast redwood that
may enhance the ability to distinguish
clones in natural populations is its apparent breeding system. Coast redwood
seems to be similar to other conifers in
being predominantly outcrossing, there
being some evidence of inbreeding depression under natural conditions ( Libby
et al. 1981). By lowering relatedness
among sexually derived recruits (i.e., relative to selfing), outcrossing would tend
to lessen the likelihood of type II errors in
clonal identification efforts.
In general, the higher the level of withinpopulation genetic variation that a species
Brief Communications 431
possesses, the more likely it is that clones
can be distinguished with genetic markers. Hamrick and Godt (1990) have found
that plant traits that are correlated with
high degrees of genetic variation within
populations include taxonomic status
(gymnosperms), breeding system (outcrossing, wind pollinated), life form (woody,
long-lived perennial), and geographic
range (widespread). Coast redwood possesses all of these traits with the exception of the last, being restricted in geographic range. However, its high ploidy
level may add another degree of potential
variability. (Ploidy level was not investigated as a correlate with genetic variation
in the Hamrick and Godt study.) In an earlier study comparing life-history characteristics and allozyme variation in plant
species, Hamrick et al. (1979) noted that
species with high numbers of chromosomes tended to have high levels of genetic variation.
In this clonal identification study the
ability to uniquely identify all 34 individuals
within the four selfed-sib families using
only 10 polymorphic loci bodes well for the
ability to make conclusive clonal identifications in natural populations of coast redwood. Others have attempted, without pedigreed data, to quantify the resolving power
of isozyme analysis to differentiate unique
genotypes. Hunter (1993) estimated the resolving power of seven-locus (isozyme) genotypes to differentiate coral clones (Porites compressa). Her method first involved
calculating the mean number of unique genotypes resolved by each locus individually. Next, the mean number of unique genotypes resolved by all combinations of
two loci was determined, and so on, for all
combinations of loci up to seven. Then a
graph was produced plotting the cumulative (mean) number of genotypes resolved
versus the number of loci used (i.e., one to
seven). An asymptote was approached between six and seven loci, this being the inferred measure of resolving power. However, as stated by Hunter (1993), ‘‘variations in local gene frequencies dictate that
more loci may be required to identify and
distinguish between all clones in all populations.’’ This is particularly true when two
or more of the loci involved in the assay
are tightly linked. Furthermore, in the Hunter (1993) study, as the asymptote coincided with the maximum number of loci assayed, the number of loci required for optimum resolution remained questionable.
Another approach to the estimation of
resolving power comes from an electrophoretic study of a rare rhizomatous plant
432 The Journal of Heredity 1999:90(3)
in the northeastern United States. In a
study of the clonal structure of Queen of
the Prairie (Filipendula rubra), Aspinwall
and Christian (1992) calculated the average probability of ramets with the same
multilocus allozyme genotype belonging
to the same clone (i.e., the probability of
not making a type II error). Their statistic
simply reflects the proportion of samples
with any given single-locus phenotype,
summed over all samples and all polymorphic loci. This approach is a simpler version of the methods commonly employed
in paternity analysis—to determine the
statistical likelihood that a given male is
the paternal parent (e.g., Meagher 1986,
Meagher and Thompson 1986). However,
estimation of clonal identity, like paternal
identity, is a function of the number of (unlinked) loci examined, population allelic
frequencies, and the underlying genetic
structure, for example, Hardy–Weinberg
proportions (Meagher 1986). If this information is missing or population polymorphism is low or genetic structure is complicated, it is difficult to calculate the resolving power of clonal identification.
Type I error, the probability of assigning
genetically identical samples to different
multilocus genotypes, is generally the likelihood of environmental, physiological, or
laboratory-related factors influencing the
interpretation of banding patterns. The
probability of these events in coast redwood allozyme studies has been previously explored and found to be negligible
(Rogers 1994). Type I errors could also
embrace somatic mutations (i.e., two samples from the same individual differ because of a somatic mutation in one part of
the plant). The rate of, and hence likelihood of sampling, somatic mutations in
this species, particularly one distinguishable at the electrophoretic level, are unknown. Previous studies have failed to distinguish among sports (somatic variants)
in domestic species of banana (Musa acuminata) and apple (Malus domestica) using
isozyme analysis [Jarret and Litz (1986)
and Weeden and Lamb (1985), respectively]. One study of coast redwood found no
within-tree variations in leaf monoterpene
composition ( Hall 1985), suggesting there
were no underlying somatic mutations expressed at this (monoterpene) level. However, dramatic somatic mutations at the
morphological level have been described
in this species ( Libby 1984). An investigation of the rate of somatic mutations in
coast redwood is beyond the scope of this
study. As such, it is accepted as a study
condition that they may contribute to
type I error.
There have been few studies of clonal
structure in natural populations of asexual
woody plants (e.g., Comtois et al. 1986;
Montalvo et al. 1997) and fewer yet of
asexual, woody, gymnosperm species.
Ellstrand and Roose (1987), in a comprehensive survey of 27 genetic studies of
clonal plant species, found a highly significant correlation between the number of
characters scored (whether isozyme loci
and/or morphological traits) and the number of genotypes detected. This suggests,
regardless of species attributes, that sensitivity of clonal differentiation (or resolving power) is generally enhanced with the
addition of characters.
The only other study of an asexually reproducing woody plant of which I am
aware that made use of pedigreed material
to determine the probability of detecting
clones in the field was with Osage orange
(Maclura pomifera) (Schnabel et al. 1991).
Of interest, this species is also a polyploid—an autotetraploid. Nevertheless,
within the five full-sib families assayed for
four loci, every individual did not have a
unique allozyme identity. In fact, the most
common four-locus genotype accounted
for up 10% of the progeny within each family. In open-pollinated families, the number
of duplicated multilocus genotypes was
reduced dramatically, leading the authors
to conclude that clumped individuals in
natural populations are unlikely to be
falsely assigned to the same clone with
this set of molecular markers.
Resolution between clones in coast redwood on an electrophoretic basis could be
further improved by any one of three conditions. The first, and most readily invoked, is to assay additional polymorphic
loci. The second and third require further
technical development. The ability to
identify allele dosage effects would greatly
improve resolving power. Finally, resolution would also be increased with the ability to distinguish among genotypes that
had the same number of each allele at
each locus but differed in the location of
alleles on the three homologous pairs of
chromosomes (Alpert et al. 1993).
From the Genetic Resources Conservation Program,
University of California, 1 Shields Ave., Davis, CA 95616.
An early version of this manuscript received the attention and helpful criticism of W. J. Libby, C. I. Millar, J.
R. McBride, and P. T. Spieth. I thank the University of
California at Berkeley for permission to sample reserves of coast redwood in their Russell Reservation
and W. J. Libby for his activities that made available
pedigreed material for this study. Address correspondence to D. L. Rogers at the address above or e-mail:
[email protected].
q 1999 The American Genetic Association
References
phoretically detectable genetic variation in plants.
Annu Rev Ecol Syst 10:173–200.
Alpert P, Lumaret R, and Di Giusto F, 1993. Population
structure inferred from allozyme analysis in the clonal
herb Fragaria chiloensis (Rosaceae). Am J Bot 80:1002–
1006.
Hunter CL, 1993. Genotypic variation and clonal structure in coral populations with different disturbance histories. Evolution 47:1213–1228.
Aspinwall N and Christian T, 1992. Clonal structure, genotypic diversity and seed production in populations
of Filipendula rubra (Rosaceae) from the northcentral
United States. Am J Bot 79:294–299.
Jarret RL and Litz RE, 1986. Isozymes as genetic markers in bananas and plantains. Euphytica 35:539–549.
Comtois P, Simon JP, and Payette S, 1986. Clonal constitution and sex ratio in northern populations of balsam poplar Populus balsamifera. Holarctic Ecol 9:251–
260.
Libby WJ, McCutchan B, and Millar CI, 1981. Inbreeding
depression in selfs of redwood. Silv Genet 29:15–24.
Demeke T, Kawchuk LM, and Lynch DR, 1993. Identification of potato cultivars and clonal variants by random amplified polymorphic DNA analysis. Am Potato J
70:561–569.
Libby WJ, 1984. United States patent number: plant 5,
345; November 20, 1984.
Lokker C, Susko D, Lovett-Doust L, and Lovett-Doust J,
1994. Population genetic structure of Vallisneria americana, a dioecious clonal macrophyte. Am J Bot 81:
1004–1012.
Ellstrand NC and Roose ML, 1987. Patterns of genotypic
diversity in clonal plant species. Am J Bot 74:123–131.
MacDonald SE and Lieffers VJ, 1991. Population variation, outcrossing, and colonization of disturbed areas
by Calamagrostis canadensis: evidence from allozyme
analysis. Am J Bot 78:1123–1129.
Hall GD, 1985. Leaf monoterpenes of coast redwood
(Sequoia sempervirens [D. Don] Endl.) (PhD dissertation). Santa Cruz, California: University of California.
McClintock KA and Waterway MJ, 1993. Patterns of allozyme variation and clonal diversity in Carex lasiocarpa
and C. pellita (Cyperaceae). Am J Bot 80:1251–1263.
Hamrick JL and Godt MJW, 1990. Allozyme diversity in
plant species. In: Plant population genetics, breeding,
and genetic resources ( Brown AHD, Clegg MT, Kahler
AL, and Weir, BS, eds). Sunderland, Massachusetts: Sinauer; 43–63.
Meagher TR, 1986. Analysis of paternity within a natural population of Chamaelirium luteum. 1. Identification
of most-likely male parents. Am Nat 128:199–215.
Hamrick JL, Linhart YB, and Mitton JB, 1979. Relationships between life history characteristics and electro-
Meagher TR and Thompson E, 1986. The relationship
between single parent and parent pair genetic likelihoods in genealogy reconstruction. Theor Popul Biol
29:87–106.
Montalvo AM, Conard SG, Conkle MT, and Hodgskiss
PD, 1997. Population structure, genetic diversity, and
clone formation in Quercus chrysolepis ( Fagaceae). Am
J Bot 84:1553–1564.
Rogers DL, 1994. Spatial patterns of allozyme variation
and clonal structure in Coast redwood (Sequoia sempervirens) (PhD dissertation). Berkeley, California: University of California.
Rogers DL, 1997. Inheritance of allozymes from seed
tissues of the polyploid gymnosperm, coast redwood
(Sequoia sempervirens ( D. Don) Endl.). Heredity 78:
166–175.
Schmid B, 1986. Spatial dynamics and integration within clones of grassland perennials with different growth
form. Proc R Soc Lond B 228:173–186.
Schnabel A, Laushman RH, and Hamrick JL, 1991. Comparative genetic structure of two co-occurring tree species, Maclura pomifera (Moraceae) and Gleditsia triacanthos ( Leguminosae). Heredity 67:357–364.
Thomas MR, Cain P, and Scott NS, 1994. DNA typing of
grapevines: a universal methodology and database for
describing cultivars and evaluating genetic relatedness. Plant Mol Biol 25:939–949.
Weeden NF and Lamb RC, 1985. Identification of apple
cultivars by isozyme phenotypes. J Am Soc Hort Sci
110:509–515.
Received January 7, 1998
Accepted December 7, 1998
Corresponding Editor: James Hamrick
Brief Communications 433