Genetic Analysis of Variation for
Auxin-Induced Adventitious Root
Formation Among Eighteen Ecotypes of
Arabidopsis thaliana L. Heynh.
J. J. King and D. P. Stimart
Eighteen ecotypes and two inbred lines of Arabidopsis thaliana L. Heynh. were
analyzed for variation in the number of adventitious roots formed (hereafter referred to as rooting) on seedling hypocotyls in response to auxin treatment. Mean
root counts varied from 1.7 to 23.1. Stable high (HA) and low (LA) rooting lines
selected from ecotype Columbia, a low rooting ecotype (Mt-0), and unselected Columbia populations were evaluated for vegetative and reproductive growth parameters to determine correlated phenotypic effects of selection for rooting response.
High rooting in HA correlated with compact, highly branched shoot growth. Genetic
analysis of HA, Mt-0, and their F1, F2, and reciprocal backcross generations indicated that high and low rooting responses in this population may be controlled by
several genes acting independently in additive-dominance fashion. Genetic variance partitioned into principally additive effects, with dominance favoring low rooting.
From the Department of Horticulture, University of
Wisconsin, 1575 Linden Dr., Madison, WI 53706. Joseph
King is currently at Seminis Vegetable Seeds, 37437
State Highway 16, Woodland, CA 95695. Address correspondence to Dr. Stimart at the address above. Research was supported by USDA funds, project number
3125.
q 1998 The American Genetic Association 89:481–487
Development of adventitious roots in higher plants involves perception of morphogenetic signals and de novo formation of
root meristems by cells of shoots or
leaves ( Esau 1977). This process is the essential first step in forming root systems
of lower vascular plants, monocots, and
clonally reproduced forestry and horticultural crops ( Davies et al. 1994; Esau 1977;
Kovar and Kuchenbuch 1994; Ritchie
1994). In an agricultural sense, adventitious root formation on vegetative propagules allows clonal multiplication and rapid fixation of superior genotypes, expediting their incorporation into production or
breeding programs. This strategy is often
exploited for long-lived woody species.
However, the inability to initiate adventitious roots on otherwise superior individuals in many species remains an obstacle
to ubiquitous application of this procedure.
Competence to form adventitious roots
is quantified as percent rooted cuttings or
root count per rooted cutting, and economically important genotypes of Malus
pumila Mill. (Alvarez et al. 1989), M. domestica L. ( Harbage 1991), Eucalyptus
(Grattapaglia et al. 1995), and other woody
species are classified as easy or difficult
to root. Although variation in ability to
form roots is known in herbaceous species ( Hardwick 1979), difficult-to-root genotypes are better characterized among
woody species where clonal propagation
is used more commonly as an alternative
to breeding in these long-lived perennials.
Data from a limited number of genetic
studies suggest that competence to form
adventitious roots is a quantitative trait
(Riemenschneider 1993). Four quantitative trait loci (QTL) were associated with
percent rooted cuttings in Eucalyptus
(Grattapaglia et al. 1995). Three of these
were detected in an easy-to-root species—
E. urophylla S. T. Blake—and the fourth
was detected in a difficult-to-root species—Eucalyptus grandis W. Hill ex. Maiden. A narrow-sense heritability of 0.15 for
percent rooted cuttings was calculated for
Pinus taeda L. ( Foster 1990), and broadsense heritabilities for root count per
rooted cutting were estimated to be 0.44–
0.56 in Populus deltoides Bartr. (Wilcox and
Farmer 1968) and 0.92 ( Foster et al. 1984)
and 0.30 (Pounders and Foster 1992) in
Tsuga heterophylla (Raf.) Sarg. In the herbaceous species Phalaris arundinacea L.,
narrow-sense heritabilities were 0.36 for
percent rooted nodes, and 0.78 for root
count per node (Casler and Hovin 1980).
Analysis of tissue culture responses in Trifolium pratense L. found mostly additive
genetic variance for root count and percent rooting. Dominance effects were significant but accounted for only a small
portion of genetic variance ( Keyes et al.
1980).
481
Despite potential benefits of improved
rooting ability in woody species, their long
life cycles limit the acquisition of basic information on inheritance and the types of
gene action involved in adventitious root
formation. The shorter life cycles of herbaceous species permit more rapid development of family structures favoring genetic analyses. While dramatic structural
and developmental differences exist between herbaceous and woody species,
some basic information on root development is likely to be transferable between
plant types.
Recent characterization of root development ( Dolan and Roberts 1995), and
identification of genes altering ( Benfey et
al. 1993; Cheng et al. 1995; King, et al.
1995), or associated with (Smith and Federoff 1995) this process in Arabidopsis thaliana L. Heynh. make this an attractive organism for analysis of adventitious root
formation. The objectives of this study
were to determine the extent of variation
in auxin-induced adventitious root formation among ecotypes of A. thaliana, analyze the genetic basis of this variation, and
characterize correlated phenotypic effects
of this variation.
Methods
Plant Growth Conditions
For rooting assays, plants were grown on
basal medium ( Haughn and Somerville
1986) supplemented with 1% (w/v) sucrose, 0.6 mM thiamine hydrochloride,
0.28 mM myoinositol, and 0.7% (w/v) Difco
Bacto agar, hereafter referred to as medium. We poured 25 ml of medium into 125
ml round glass jars, referred to as rooting
jars, added a polycarbonate ring, ;3 cm
tall and of a diameter ;5 mm less than the
inside diameter of the jars, covered the
jars with 60 mm 3 10 mm glass petri dish
tops, and autoclaved. Immediately after
autoclaving, rings were centered in each
jar, leaving a channel of 2–3 mm between
rings and inner walls of jars. We disinfested seeds by soaking them for 20 min in
1.6% (v/v) NaOCl plus 0.1% (v/v) Tween20 followed by three rinses in sterile
deionized H2O, then sowed 50 seeds into
the channels between the polycarbonate
rings and the inner walls of the jars as a
suspension in 1 ml of cooled liquid medium containing 0.7% (w/v) agar. This distributed seeds in an approximate monolayer along the inner walls of the jars allowing us to make root counts and observe the seedlings easily with a dissecting
microscope. Newly sown seeds were strat-
482 The Journal of Heredity 1998:89(6)
ified at 48C in darkness for 24 h, then germinated in darkness at 238C for 5 days,
causing hypocotyls to etiolate to 7–10 mm
in length. We then opened the jars in a
laminar air flow hood and added 4 ml of
cooled liquid medium containing 35 mM
indole-3-butyric acid ( IBA) and 0.7% (w/v)
agar which flooded seedling hypocotyls to
a depth of 5 mm to induce adventitious
root formation. Jars were resealed and
placed at 238C under a daily light regime
from 0800 to 2400 h at 50 mE/m2/s provided by cool-white fluorescent bulbs. After
14 days we counted emerged adventitious
roots on seedling hypocotyls. This procedure is hereafter referred to as the rooting
protocol.
In the greenhouse, we grew plants in 1:
1:1 soil : peat : perlite (v/v), hereafter referred to as potting mix. The greenhouse
temperature was set to 208C. A daily light
regime, as above, at a minimum of 550 mE/
m2/s was provided by natural light supplemented by 1,000 W high-pressure sodium
lamps.
Initial Selections
Adventitious roots on 100 randomly selected seedlings of A. thaliana ecotype Columbia were counted, and the lowest rooting individual ( LA), with 1 root, and the
highest rooting individual ( HA), with 13
roots, were transplanted to potting mix
and allowed to self-pollinate. We advanced
these lines by single seed descent to S6. In
each generation, 40 seedlings per line
were evaluated and the lowest and highest
rooting individuals in LA and HA lines, respectively, were selected and self-pollinated to produce the next generation.
Evaluation of Root Counts Among
Ecotypes
To increase seed availability, we sowed approximately 50 seeds from each of 18 ecotypes of A. thaliana onto the surface of
potting mix in 10 cm square plastic pots,
arranged the pots in a greenhouse in a
randomized complete block design with
four blocks and one pot per ecotype per
block. We allowed the plants to self-pollinate, then harvested and bulked seeds
within each ecotype. Seeds derived from
these self-pollinations were used to initiate rooting assays (as described above)
on 36 seedlings of each ecotype, S6 families of HA and LA, and a bulk of self-pollinated seeds from 20 randomly selected
Columbia seedlings that had passed twice
through the rooting protocol (C235). We
included C235 to detect changes in root
counts due to factors other than selection
for high or low rooting parents. The experiment was a randomized complete
block design with 3 blocks and 12 samples
per genotype per block. We selected individuals producing no adventitious roots
from ecotype Mt-0, which produced the
lowest mean root count, and the lowest
and highest rooting individuals from LA
and HA, respectively, transplanted these
seedlings to potting mix, and allowed
plants to self-pollinate in a greenhouse.
Evaluation of Multiple Trait Variation
Among Selected Lines
To examine possible trait correlations
with root count, phenotypic variation for
10 traits was quantified among the following five genotypes: S7 families of LA and
HA; unselected ecotype Columbia, an S1
family of Mt-0, and self-pollinated progenies of 20 randomly selected Columbia
plants that had passed three times
through the rooting protocol (C335). Thirty-two seeds of each line were sown, one
seed per pot, into potting mix in 10 cm
square plastic pots. Pots were arranged in
a greenhouse in a randomized complete
block design with four blocks, each containing eight pots per genotype. Greenhouse conditions were as described
above. We collected data for days corresponding to the following life-history
traits—germination, visible flower bud,
bolting, anthesis, and apical arrest [defined as ‘‘cessation of generative activity
at the inflorescence meristems’’ ( Hensel et
al. 1993)]—and compared plant form and
size based on rosette leaf counts and fresh
and dry weights at anthesis, and height
and lateral inflorescence branch counts at
apical arrest. Data were subjected to analysis of variance.
In this experiment, plants of HA were
short with many elongating lateral branches (suggesting weak apical dominance),
whereas plants of Mt-0 were tall with few
elongating lateral branches (suggesting
strong apical dominance) ( Figure 1). To
ascertain the relationship between formation of lateral branches and adventitious
roots, we counted adventitious roots in
the rooting environment, then transplanted HA, Mt-0, and 71 F2 (synthesis described below) seedlings to 10 cm square
plastic pots and grew them in the greenhouse as described above. When plants
reached apical arrest, we counted the lateral inflorescence branches and performed a correlation analysis for root and
branch counts.
among putative F1 individuals. Flowers not
crossed on the F1 plant were allowed to
self-pollinate to generate the F2.
We counted adventitious roots on seedlings of each generation in two experiments, performed at different times, using
randomized complete block designs consisting of four blocks with 10 to 20 samples per genotype per block depending on
generation. More samples were counted
for segregating generations ( F2 and backcrosses) to equalize variances with nonsegregating generations.
We transformed raw data as (root count
1 1)½, analyzed experiments separately,
then tested for homogeneity of error variances to determine if data could be
pooled across experiments. Generation 3
block/experiment variance was tested
against sampling variance and, if significant, was used as the error term for testing generation and generation 3 experiment effects. We partitioned generation
sum of squares into additive effects as (P1
2 P2)/2 and dominance effects as F1 2 (P1
1 P2)/2, and discerned differences among
generation means with Duncan’s multiple
range test using Kramer’s modification for
unequally replicated treatments ( Kramer
1956). Using mean root counts for P1, P2,
and F1, we calculated the degree of dominance according to Mather and Jinks
(1977) as
Dominance
5
5 (F1 mean 2 Midparent value)
4
had been maintained as a laboratory
strain and propagated by self-pollinations
for several generations; the variance for
root count in the Mt-0 parental line was
the lowest among the six generations tested, implying absence of segregation at loci
affecting root count; and the Mt-0 line was
homozygous at 20 microsatellite loci on
four of the five Arabidopsis chromosomes
( Innan et al. 1997). After determining that
root counts were not different in F1 families derived from reciprocal crosses (data
not shown), we chose HA 3 Mt-0 as the F1
parent because F1 plants were identical
morphologically to Mt-0, but were distinctly different from HA. This allowed us to
detect HA self-pollinated contaminants
1
(High parent mean
2
]6
Figure 1. (A) Phenotypes of representative plants of A. thaliana: (a) ecotype Columbia; ( b) low adventitious
rooting line, LA; (c) high adventitious rooting line, HA; (d) low adventitious rooting line, Mt-0. ( B) Lateral branch
production in HA and Mt-0.
Genetic Analysis of Root Count
We developed six populations for analysis
of variation in rooting: parent 1 (P1) was
HA at S7; parent 2 (P2) was Mt-0 at S1, F1
( HA 3 Mt-0), F2; backcross to parent 1
( BP1) was HA 3 [HA 3 Mt-0]; and backcross to parent 2 ( BP2) was Mt-0 3 [HA 3
Mt-0]. Ecotype Mt-0 was chosen because it
yielded the lowest mean root count. The
fact that the Mt-0 parent in the F1 was an
S1 from the ecotype presents the possibility of heterozygosity at loci affecting root
count and would therefore violate assumptions basic to a generation means
analysis. We assumed homozygosity at
these loci in Mt-0 based on three factors:
the Mt-0 line used in these experiments
[
2 Low parent mean)
We performed a generation means analysis with an additive-dominance model for
least squares estimation of generation
root count means (Mather and Jinks
1977). Generation means were modeled in
terms of mean effects (m), pooled additive
effects (d), and pooled dominance effects
(h). A joint scaling test (Cavalli 1952) was
used to calculate a chi-square value to determine adequacy of the additive-dominance genetic model. The chi-square value
was tested with three degrees of freedom
with significance indicating inadequacy of
the model.
Results
Mean root counts for LA, HA, C235, and
the 18 ecotypes ranged from 1.7 to 23.1
King and Stimart • Adventitious Root Formation in Arabidopsis 483
Table 1. Adventitious root counts at 14 days
after treatment with 35 mM IBA for A. thaliana
lines selected for high (HA) and low (LA) rooting,
an unselected line from ecotype Columbia (C235),
and unselected ecotypes
Table 2. Days to visible flower bud, bolting, and anthesis; leaf counts and fresh and dry weights at
anthesis; days to apical arrest; lateral branch counts; and height at apical arrest in A. thaliana ecotype
Columbia and its derived line C335, and in lines selected for high (HA) and low (LA and Mt-0) rooting
Genotype
Days to
visible
Buda
Columbia
C335c
HA-S7
LA-S7
Mt-0
21.9
21.9
24.0
23.1
23.4
Adventitious root count
a
Genotype
n
Minimum
Maximum
Mean 6 SE
HA-S6
LA-S6
Tsu-0
En-2
Landsberg
Bor-0
Edi-0
Pi-0
Sv-0
Mh-0
No-0
Ni-0
Lip-0
C235a
Columbia
Ge-0
Ct-1
Oy-0
An-1
Wil-2
Mt-0
32
36
24
36
36
36
36
36
36
36
36
31
36
36
36
36
36
36
36
36
35
8
2
1
3
2
0
1
0
1
1
1
2
1
1
0
0
0
0
0
0
0
36
19
17
14
11
11
12
13
10
9
11
9
15
8
8
8
14
8
7
8
7
23.1
10.9
8.1
7.2
6.4
6.0
5.6
4.9
4.9
4.8
4.7
4.6
4.2
4.2
3.8
3.8
3.6
3.3
2.2
2.0
1.7
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
1.5
0.5
0.7
0.4
0.4
0.6
0.6
0.4
0.4
0.4
0.3
0.3
0.5
0.3
0.3
0.3
0.5
0.4
0.3
0.3
0.3
Bulk of self-pollinated seed of 20 randomly selected
plants of ecotype Columbia passed twice through the
rooting protocol.
( Table 1). The HA line was highest at 23.1,
LA was second highest at 10.9, and the
other 19 genotypes ranged from 1.7 to 8.1.
Columbia and C235 ranked consecutively
with means not different statistically. Ecotype Mt-0 had the lowest root count and
was evaluated further as a potential low
rooting line.
Evaluation of Multiple Trait Variation
Among Selected Lines
Seeds of all genotypes germinated in 4
days. Flower buds were visible on all
plants at 22–24 days after sowing. Bolting
occurred at 24–27 days and anthesis at
28–32 days ( Table 2). Mt-0 produced the
largest and LA the smallest plants based
on leaf count and fresh and dry weights at
anthesis ( Table 2). The most distinct differences were in fresh and dry weights of
Mt-0, being 1.8 and 1.6 times greater, respectively, than the next highest genotype, HA. Genotype rankings for days to
apical arrest were consistent with days to
anthesis: Columbia and C335 were earliest,
Mt-0 was intermediate, and HA and LA
were latest ( Table 2). Apical arrest occurred in all genotypes at 48–53 days. Lateral branch count was highest at 15.1 in
HA, being 1.7 times greater than next
ranked Mt-0 ( Table 2, Figure 1). Branch
counts for genotypes other than HA were
7.6 to 8.7. Mean heights at apical arrest
ranged from 31.6 cm for HA to 46.8 cm for
484 The Journal of Heredity 1998:89(6)
cab
c
a
b
ab
Days to
bolting
Leaf
count at
anthesis
Fresh
weight at
anthesis
(g)
Dry
weight at
anthesis
(g)
Days to
apical
arrest
Branch
count at
apical
arrest
Height at
apical
arrest (cm)
24.7
25.1
26.7
26.4
25.9
9.9
8.9
9.8
8.7
11.0
1.1
1.1
1.2
0.8
2.2
0.1
0.1
0.2
0.1
0.3
48.4
48.8
52.6
53.0
50.6
8.0
7.8
15.1
7.6
8.7
41.6
43.7
31.6
35.9
46.8
c
c
a
b
b
b
c
bc
c
a
bc
bc
b
c
a
bc
bc
b
c
a
c
c
a
a
b
bc
bc
a
c
b
c
b
e
d
a
Values are means of 20 ( Days to Bud and Bolting) or 16 (all other variables) observations.
Duncan’s multiple range test rankings. Means within a column with same letter are not different at P 5 .05.
c
Bulk of self-pollinated seed of 20 randomly selected plants of ecotype Columbia passed three times through the
rooting protocol.
a
b
Mt-0 ( Table 2). Heights of all genotypes
were different statistically, with intervals
between means of 2–6 cm. The smallest
difference was between Columbia at 41.6
cm and C335 at 43.7 cm.
In the F2 population, counts of lateral inflorescence branches and adventitious
roots correlated significantly (r 5 0.31).
Lateral branch counts ranged from 9 to 18
for HA, 4 to 9 for Mt-0, and 2 to 28 for the
F2 (data not shown). The F2 distribution
was unimodal and skewed positively, with
a mean of 7.7. Branch count distributions
of HA and Mt-0 overlapped at 9. Dividing
the F2 population at this value yielded two
segregating classes containing individuals
with branch counts of 2 to 8 and 10 to 28
(n 5 48 to 51 and 22 to 19, respectively,
depending on placement of the three F2 individuals with branch counts of 9 into the
high or low class). These distributions fit
a 3:1 segregation ratio with chi-square values of 1.54, 0.93, 0.47, and 0.17, suggesting
recessive, single-gene inheritance of the
highly branched phenotype.
Genetic Analysis of Root Count
The root count distribution for P1 ( HA at
S7) ranged from 0 to 28, and the distribution for P2 (Mt-0 at S1) ranged from 0 to 15.
Root count distributions for F2 and BP2
generations resembled the distribution for
P2 ( Figure 2). The F-test of error variances
of the two separate experiments was not
significant (P . .05) indicating homogeneity of variances, so data were pooled
over experiments and analyzed ( Table 3).
In the pooled analysis, generation 3
block/experiment was significant and was
used as the error term for testing generation and generation 3 experiment effects
which were significant (P , .01) and not
significant (P . .05), respectively. Partitioning generation sum of squares showed
significant additive and dominance effects
( Table 3).
Ranking of means by Duncan’s multiple
range test did not differ between analyses
of transformed and untransformed data.
Therefore distributions ( Figure 2) and
means ( Table 4) of untransformed data
are presented. The mean adventitious root
count for HA was 14.5 and for Mt-0 was 4.9
( Table 4). The F1 mean at 7.9 was 1.8 less
than the midparent value of 9.7, yielding
an estimate of 20.38 for degree of dominance. The BP1 mean at 10.5 was 0.8 greater than the midparent value. Means for P2,
F2, and BP2 at 4.9, 4.7, and 4.9, respectively,
were not significantly different, and
ranked lowest among generation means.
The chi-square test of estimated generation means was nonsignificant ( Table 5),
indicating adequacy of an additive-dominance model for explaining rooting responses in this population. Estimated generation means were accurate to within
10% of observed means for all generations
except the F1 and F2, where estimated
means deviated from observed by 24%
and 56%, respectively.
Discussion
This study was initiated to evaluate genetic variation for auxin-induced adventitious
root formation in A. thaliana. Among 21 genotypes, the number of adventitious roots
formed in response to exogenous auxin
ranged from 1.7 to 23.1. The generation
means analysis for HA, Mt-0, and their F1,
F2, BP1, and BP2 generations suggested that
rooting response in this population is under relatively simple genetic control. An
additive-dominance model predicted generation root count means accurately, indicating little or no epistasis. Phenotypic
analysis of Columbia, C335, HA, LA, and
Mt-0 distinguished HA and Mt-0 by the
high leaf count and high fresh and dry
weights at anthesis for Mt-0, and the high
Table 3. Analysis of variance of adventitious
root count data a taken 14 days after application
of 35 mM IBA to 6-day-old seedlings of A. thaliana
representing generations derived from the cross
of high (HA) 3 low (Mt-0) rooting lines
Source
df
Experiment ( E)
1
Block/experiment ( B/E)
6
Generation (G)
5
Additive effects
1
Dominance effects
1
Residual
3
Generation 3 experiment
5
30
G 3 B/E
Error
676
a
b
mon in plants, and environmental modifications enhance rooting in recalcitrant genotypes (Andersen 1986; Loach 1988). The
structure and abundance of roots change
in response to flooding (McNamara and
Mitchell 1990), nitrogen form, solution pH
( Finn et al. 1990), temperature, light, and
mechanical impedance ( Lake 1987). The
variation in HA could reflect environmental sensitivity of root development that is
detectable principally in genotypes with
high root counts.
Dominance effects in the analysis of
variance were significant but accounted
0.05
3.21
214.14
99.97
3.36
110.82
0.05
0.54
42.83
99.97
3.36
36.94
3.69
21.62
250.61
nsb
ns
**
**
*
**
0.74 ns
0.72 **
0.37
Data transformed as (root count 1 1)½.
ns, *, ** 5 not significant and significant at P 5 .05 or
P 5 .01, respectively.
for only 1.6% of the sum of squares for
generations, whereas additive effects accounted for 46.7% ( Table 3). Similarly, in
Trifolium pratense L., additive effects were
the principal component of genetic variance for adventitious root count while
dominance effects were significant but
less important ( Keyes et al. 1980). In Cucumis sativus L. (Ghaderi and Lower 1979)
and in Phaseolus vulgaris L. ( Fawole et al.
1982), root system dry weight was determined mainly by dominance effects. The
root systems in these species include embryonic and lateral roots. The extent of genetic correlation among embryonic, lateral, and adventitious root formation is unknown, although in a few cases development of these root types has been
uncoupled genetically ( Zobel 1975).
Dominance gene action appeared to favor low rooting. This was suggested by
similarities among root count means for
P2, F2, and BP2 ( Table 4); the F1 distribution
being shifted toward P2 ( Figure 2); and degree of dominance being 20.38. This calculation of degree of dominance likely represents a minimum estimate of dominance
gene action because it cannot account for
the situation in which opposite domi-
Figure 2. Adventitious root counts at 14 days after treatment with 35 mM IBA of high ( HA) and low (Mt-0) rooting
lines of A. thaliana and their F1, F2, BP1 ( backcross to HA), and BP2 ( backcross to Mt-0) generations.
lateral branch count at apical arrest for
HA ( Table 2, Figure 1).
The generation sum of squares for root
count data consisted of 48% additive and
dominance effects and 52% residual. The
residual is due principally to the extreme
variation in HA and the backcross to HA
( BP1) ( Figure 2). Unless selection during
self-pollination of HA to S7 favored heterozygosity at loci controlling root formation,
this line was likely homozygous, leaving
mainly environmental effects contributing
to phenotypic variance. Phenotypic plasticity in response to environment is com-
Sum of Mean
squares square
Table 4. Adventitious root counts 14 days after
applying 35 mM IBA to 6-day-old seedlings of A.
thaliana representing generations from the cross
of high (HA) 3 low (Mt-0) rooting lines
a
Generation
N
Mean 6 SE
P1 ( HA-S7)
P2 (Mt-0-S1)
F1
F2
BP1
BP2
88
91
124
157
127
137
14.5
4.9
7.9
4.7
10.5
4.9
6
6
6
6
6
6
0.6
0.3
0.3
0.2
0.4
0.3
aa
d
c
d
b
d
Duncan’s multiple range test rankings. Means with
same letter are not different at P 5 .05.
King and Stimart • Adventitious Root Formation in Arabidopsis 485
Table 5. Variances of generation means, components of the model for least squares estimation of
expected generation means, and x2 value of the joint scaling test for adequacy of the model to predict
adventitious root counts at 14 days after treatment with 35 mM IBA in the cross of A. thaliana high (P1)
3 low (P2) rooting lines
Means
Generation
n
Vxa
Weight
(1/Vx)
P1 ( HA-S7)
P2 (Mt-0-S1)
F1
F2
BP1
BP2
x2 5 1.82 nse
88
91
124
157
127
137
29
7.1
12.8
7.9
19.1
12.3
0.035
0.141
0.078
0.126
0.053
0.081
Model
mb
dc
hd
Difference
Observed Expected (O 2 E)
1
1
1
1
1
1
1
21
0
0
0.5
20.5
0
0
1
0.5
0.5
0.5
14.5
4.9
7.9
4.7
10.5
4.9
11.3
4.1
6.3
7.7
10.1
5.2
3.2
0.8
1.6
23.0
0.4
20.3
Variance of generation means.
m 5 9.0 6 0.46.
c
d 5 4.95 6 0.45.
d
h 5 22.72 6 0.85.
e
Not significant at P 5 .05.
b
486 The Journal of Heredity 1998:89(6)
Berleth T and Jürgens G, 1993. The role of the monopteros gene in organising the basal body region of the
Arabidopsis embryo. Development 118:575–587.
Biddington NL and Dearman AS, 1982. The involvement
of the root apex and cytokinins in the control of lateral
and root emergence in lettuce seedlings. Plant Growth
Reg 1:183–193.
Casler MD and Hovin AW, 1980. Genetics of vegetative
stand establishment characters in reed canarygrass
clones. Crop Science 20:511–515.
Cavalli LL, 1953. An analysis of linkage in quantitative
inheritance. In: Quantitative inheritance (Reeve ECR
and Waddington CH, eds). London: HMSO; 135–144.
a
nance effects of multiple loci cancel
(Mather and Jinks 1977).
High rooting in HA appears to be controlled by several independent genes and
environmental influences. Multigene control is suggested by absence of high rooting segregants in the F2, however, the shift
of BP1 toward high rooting implicates relatively few loci that could be introgressed
through backcrossing. These conclusions
for Arabidopsis are consistent with results
of Grattapaglia et al. (1995) who found
four QTL could account for 63% of the
phenotypic variance for percent rooted
cuttings in a cross of high 3 low rooting
Eucalyptus species. While these two experiments quantified adventitious root formation differently (root counts in Arabidopsis; percent rooted cuttings in Eucalyptus), both of these measures could be valid
assessments of competence to form adventitious roots, and could be detecting
effects of analagous sets of genes.
The high lateral branch count in HA correlated significantly with high adventitious
root count, and appeared to be controlled
by a single recessive gene. The altered
meristem program (amp1) mutant of A.
thaliana produces more lateral shoots
than wild type and contains elevated levels of endogenous cytokinin (Chaudhury
et al. 1993). Cytokinins applied at low concentrations enhance responses to auxin
( Dominov et al. 1992) including formation
of lateral and adventitious roots ( Biddington and Dearman 1982; Van Staden and
Harty 1988). Cytokinins are also known to
delay chlorophyll loss and leaf senescence
(Galston and Davies 1970), and leaf retention has correlated positively with adventitious root counts in Phaseolus vulgaris L.
and Glycine max L. Merrill ( Hardwick
1979). Although not quantified, plants of
Benfey PN, Linstead PJ, Roberts K, Schiefelbein JW,
Hauser M-T, and Aeschbacher RA, 1993. Root development in Arabidopsis: four mutants with dramatically altered root morphogenesis. Development 119:57–70.
Chaudhury AM, Letham S, Craig S, and Dennis ES, 1993.
Amp1—a mutant with high cytokinin levels and altered
embryonic pattern, faster vegetative growth, constitutive photomorphogenesis and precocious flowering.
Plant J 4:907–916.
Cheng J-C, Seeley KA, and Sung ZR, 1995. RML1 and
RML2, Arabidopsis genes required for cell proliferation
at the root tip. Plant Physiol 107:365–376.
HA were observed in the greenhouse to
remain green longer than plants of other
genotypes. If HA contains elevated cytokinin causing enhanced responsiveness to
auxin, exogenous application of auxin in
the rooting protocol may overcome root
inhibitory effects of cytokinin and increase root initiation. Further analyses of
HA could elucidate factors contributing to
this phenotype.
The results reported here represent an
initial characterization of variation for
auxin-induced adventitious root formation
in A. thaliana. Future analyses could benefit from expanding the genetic base being
analyzed by intermating within high and
low rooting ecotypes, extracting unselected inbred lines from intermated populations, then characterizing and attempting
to control the distribution of responses
within inbreds before crossing between response groups. Analysis of multiple crosses would permit broader inferences about
rooting in this species. Extraction of recombinant inbred lines from F2 populations would provide greater degrees of
freedom for more complex models of generation means analysis, allow estimation
of heritability by parent offspring regression, and allow more accurate estimation
of environmental variance from inbred
lines.
Grattapaglia D, Bertolucci FL, and Sederoff RR, 1995.
Genetic mapping of QTLs controlling vegetative propagation in Eucalyptus grandis and E. urophylla using a
pseudo-testcross strategy and RAPD markers. Theor
Appl Genet 90:933–947.
References
Harbage JF, 1991. Anatomy and physiology of adventitious root formation in Malus domestica microcuttings
(PhD dissertation). Madison, Wisconsin: University of
Wisconsin.
Alvarez R, Nissen SJ, and Sutter EG, 1989. Relationship
between indole-3-acetic acid levels in apple (Malus
pumila Mill.) rootstocks cultured in vitro and adventitious root formation in the presence of indole-3-butyric
acid. Plant Physiol 89:439–443.
Anderson AS, 1986. Environmental influences on adventitious rooting in cuttings of non-woody species. In New
root formation in plants and cuttings (Jackson MB, ed).
Dordrecht, The Netherlands: Martinus Nijhoff; 223–254.
Davies FT, Davis TD, and Kester DE, 1994. Commercial
importance of adventitious rooting to horticulture. In:
Biology of adventitious root formation ( Davis TD and
Haissig BE, eds). New York: Plenum; 53–60.
Dolan L and Roberts K, 1995. Plant development: pulled
up by the roots. Curr Opin Genet Dev 5:432–438.
Dominov JA, Stenzler L, Lee S, Schwarz JJ, Leisner S,
and Howell SH, 1992. Cytokinins and auxins control the
expression of a gene in Nicotiana plumbaginifolia cells
by feedback regulation. Plant Cell 4:451–461.
Esau K, 1977. Anatomy of seed plants, 2nd ed. New
York: John Wiley & Sons; 215–256.
Fawole I, Gabelman WH, and Gerloff GC, 1982. Genetic
control of root development in beans (Phaseolus vulgaris L.) grown under phosphorus stress. J Am Soc
Hort Sci 107:98–100.
Finn CE, Rosen CJ, and Luby JJ, 1990. Nitrogen form
and solution pH effects on root anatomy of cranberry.
HortScience 25:1419–1421.
Foster GS, Campbell RK, and Adams WT, 1984. Heritability, gain and C effects in rooting of western hemlock
cuttings. Can J For Res 14:628–638.
Foster GS, 1990. Genetic control of rooting ability of
stem cuttings from loblolly pine. Can J For Res 21:1361–
1368.
Galston AW and Davies PJ, 1970. Control mechanisms
in plant development. Englewood Cliffs, New Jersey:
Prentice-Hall.
Ghaderi A and Lower RL, 1979. Gene effects of some
vegetative characters of cucumber. J Am Soc Hort Sci
104:141–144.
Hardwick RC, 1979. Leaf abscission in varieties of Phaseolus vulgaris ( L.) and Glycine max ( L.) Merrill—a correlation with propensity to produce adventitious roots.
J Exp Bot 30:795–804.
Haughn GW and Somerville C, 1986. Sulfonylurea-resistant mutants of Arabidopsis thaliana. Mol Gen Genet
204:430–434.
Hensel LL, Grbic V, Baumgarten DA, and Bleecker AB,
1993. Developmental and age-related processes that in-
fluence the longevity and senescence of photosynthetic tissues in Arabidopsis. Plant Cell 5:552–564.
(Gregory PJ, Lake JV, and Rose DA, eds). Cambridge:
Cambridge University Press; ix–xiii.
root formation ( Davis TD and Haissig BE, eds). New
York: Plenum; 37–52.
Innan H, Terauchi R, and Miyashita NT, 1997. Microsatellite polymorphism in natural populations of the wild
plant Arabidopsis thaliana. Genetics 146:1441–1452.
Loach K, 1988. Controlling environmental conditions to
improve adventitious rooting. In: Adventitious root formation in cuttings ( Davis TD, Haissig BE, and Sankhla
N, eds). Portland, Oregon: Dioscorides; 248–273.
Skoog F and Miller CO, 1957. Chemical regulation and
organ formation in plant tissues cultured in vitro. Symp
Soc Exp Biol 11:118–131.
Keyes GJ, Collins GB, and Taylor NB, 1980. Genetic variation in tissue cultures of red clover. Theor Appl Genet
58:265–271.
Mather K and Jinks JL, 1977. Introduction to biometrical genetics. Ithaca, New York: Cornell University
Press.
Smith DL and Federoff NV, 1995. LRP1, a gene expressed
in lateral and adventitious root primordia of Arabidopsis. Plant Cell 7:735–745.
McNamara ST and Mitchell CA, 1990. Adaptive stem
and adventitious root responses of two tomato genotypes to flooding. HortScience 25:100–103.
Van Staden J and Harty AR, 1988. Cytokinins and adventitious root formation. In: Adventitious root formation in cuttings ( Davis TD, Haissig BE, and Sankhla N,
eds). Portland, Oregon: Dioscorides; 185–201.
Kovar JL and Kuchenbuch RO, 1994. Commercial importance of adventitious rooting to agronomy. In: Biology of adventitious root formation ( Davis TD and
Haissig BE, eds). New York: Plenum; 25–36.
Pounders CT and Foster GS, 1992. Multiple propagation
effects on genetic estimates of rooting for western hemlock. J Am Soc Hort Sci 117:651–655.
Wilcox JR and Farmer RE, 1968. Heritability and C effects in early root growth of eastern cottonwood cuttings. Heredity 23:239–245.
Kramer CY, 1956. Extension of multiple range tests to
group means with unequal numbers of replication. Biometrics 12:307–310.
Riemenschneider DE, 1993. Genomic manipulation of
plant materials for adventitious rooting research. In: Biology of adventitious root formation ( Davis TD and
Haissig BE, eds). New York: Plenum; 61–76.
Lake JV, 1987. Interactions in the physical environment
of plant roots. In: Root development and function
Ritchie GA, 1994. Commercial importance of adventitious rooting to forestry. In: Biology of adventitious
King JJ, Stimart DP, Fisher RH, and Bleecker AB, 1995.
A mutation altering auxin homeostasis and plant morphology in Arabidopsis. Plant Cell 7:2023–2037.
Zobel RW, 1975. The genetics of root development. In:
The development and function of roots ( Torrey JG and
Clarkson DT, eds). New York: Academic Press; 261–275.
Received March 1, 1996
Accepted February 18, 1998
Corresponding Editor: Jonathan F. Wendel
King and Stimart • Adventitious Root Formation in Arabidopsis 487