Blackwell Science, LtdOxford, UKEREEcological Research0912-38142003 Ecological Society of JapanSeptember 2003185599609Original ArticleClonal dominance in aquatic bladderwortS.

Araki and Y. Kadono

Ecological Research (2003) 18, 599609

Restricted seed contribution and clonal dominance in a

free-floating aquatic plant Utricularia australis R. Br. in
southwestern Japan
Satoru Araki* and Yasuro Kadono

Department of Biology, Faculty of Science, Kobe University, Nada, Kobe, 657-8501, Japan
Utricularia australis R. Br. is an aquatic angiosperm species common in natural and irrigation ponds
in temperate regions. This species reproduces both sexually and vegetatively, but in southwestern
Japan the occurrence of male-sterile populations, in which plants produce no pollen and propagate
only vegetatively, has been recorded. We studied the reproductive contribution of seeds in normal
pollen-producing populations using isozyme analyses, a pollination experiment under culture and
field observations. Although seedlings obtained from controlled mating indicated segregation of
isozyme, polymorphism of the isozyme genotype was detected mainly among populations, but rarely
within each pond population. This suggested clonal dominance and rarity of seed or seedling survival
in natural populations. In the pollination experiment, the mean seed set ratio in cross-pollination
between plants of the same isozyme genotype (7.6%) did not differ significantly from self-pollination
(7.6%), but was lower than cross-pollination between plants of different genotypes (45.7%). The low
ratio in crossing between the same genotype plants was ascribed to the clonality of the parents. In
general, these results corresponded with the low ratios in seed setting observed in natural populations (7.913.7%). All the male-sterile populations we surveyed showed the same genotype, thus
male sterility in the study area was considered to have the same origin.
Key words: isozyme; male sterility; sexual reproduction; Utricularia australis; vegetative
reproduction.

INTRODUCTION
Vegetative reproduction by the root system or
other organs occurs in various taxa of angiosperms
(Klimes et al. 1997). Most of these species also
reproduce in a sexual way through seed production. The relative contribution of vegetative and
sexual modes to reproduction has been a subject of
ecological studies because it is essential to know
this information to adequately explain the reproductive system of the species (Ellstrand & Roose
1987; Eriksson 1997). The amount of seedling
recruitment varies among species and/or habitats

*Author to whom correspondence should be

addressed. Present address: Research Center for Coastal
Lagoon Environments, Shimane University, Matsue
690-8504, Japan. Email: araki@soc.shimane-u.ac.jp
Received 30 October 2002.
Accepted 26 March 2003.

(Eriksson 1989), thus the relative importance and

biological role of the vegetative and sexual systems
are considered to vary among species and their
environments. In particular, a strong reliance on
vegetative propagation is found in aquatic habitats. Many studies on aquatic macrophytes have
suggested single, or very few, genet compositions
in a population or through neighboring populations, although most of these species produce seed
or at least have the ability to produce seed through
sexual processes (Les 1988; Barrett et al. 1993;
Grace 1993). This remarkable clonality is, in part,
because these seedlings have a higher risk of dying
in inadequate light and dissolved oxygen
conditions in bottom water or sediment
habitats and the chance of seedling recruitment is
restricted (Barrett 1980; Smolders et al. 1995a,
1995b; Barrat-Segretain 1996).
In some groups, the occurrence of sterile strains,
such as species hybrids (Waterway 1994; Hollingsworth et al. 1996a) or triploids (Nakamura &

600

S. Araki and Y. Kadono

Kadono 1993; Nakamura et al. 1998), that propagate only vegetatively have been reported. If
seeds cannot contribute to reproduction because of
high mortalities of seeds or seedlings, mutations
that prevent seed production may be maintained
in these species because seed loss has no disadvantage. Furthermore, it may be advantageous if there
is a trade-off between seed production and vegetative reproduction. Irrespective of whether such a
trade-off exists (Muir 1995; Verburg & During
1998), it is predicted that sterile mutants can be
easily maintained in populations that reproduce
mainly vegetatively (Eckert et al. 1999; Eckert
2002).
The aquatic bladderwort, Utricularia australis R.
Br. (Lentibulariaceae), includes both fertile and
sterile plants. This species inhabits natural and
irrigation ponds in a submerged, free-floating
form. In summer, well-grown plants develop
scapes above the water surface and bear approximately four to six flowers per scape. These flowers
are self- or cross-pollinated by small insects, such
as aphids or small dipterous species (Yamamoto &
Kadono 1990; S. Araki and Y. Kadono, pers. obs.,
1996). In autumn, the plants form a vegetative
hibernating organ, turion, on the tips of the main
shoot and each branch. Poorly grown plants do not
develop scapes during the flowering season and
reproduce only vegetatively by turions. Turions
and seeds remain dormant until the following
spring, whereas other parts of the plant body disappear in winter. Yamamoto and Kadono (1990)
reported the occurrence of male-sterile populations
of this species, in which plants produce no pollen
grains, in the southern part of Hyogo Prefecture,
southwestern Japan. Plants in those populations
receive no pollen and produce no seed, thus it is
not a case of gynodioecy. If seeds are essential for
reproduction of this species, it is difficult to
explain how male-sterile plants could spread. But
if seeds give no or rare reproductive contribution
in this species, abnormality in pollen production is
not disadvantageous.
In the present study, we examined whether or
not seeds contribute to reproduction in normal,
pollen-producing populations of U. australis
through field observations of fruit and seed production, pollination experiments under culture
and an investigation of the population genetic
structure using isozyme analysis.

In the present paper we describe the study species as U. australis following Kadono (1993),
although there has been some confusion in the taxonomy of this species. Formerly two taxa,
U. japonica (Makino 1914) and U. tenuicaulis (Miki
1935), were reported as novel species in Japan.
They differ from each other in some morphological
features of the turion (Miki 1935; Yamamoto &
Kadono 1988). The former occurs mainly in
northern Japan, especially in Hokkaido, and rarely
in other regions of Japan, whereas the latter is distributed widely in Japan and rarely in Hokkaido
(Kadono 1994). These two taxa are now treated in
three ways according to different taxonomic opinions. Komiya and Shibata (1980) regard both of
them as synonyms of U. australis and describe
U. australis form. australis and U. australis form.
tenuicaulis, respectively. However, Tamura (1981)
describes the former as U. vulgaris var. japonica and
the latter as U. tenuicaulis. Kadono (1993)
describes the former as U. vulgaris var. japonica and
the latter as U. australis. To date, no consensus has
been achieved in nomenclature of these taxa. We
consider, from the features mentioned above, and
recently reported isozyme differences, that these
two taxa are genetically differentiated from each
other (Araki 2000). Thus, our present study does
not include plants that have been referred to as
U. japonica or U. vulgaris var. japonica, although
they are sometimes included in U. australis.

METHODS
Field observations
We define a population as all the plants growing in
a pond and refer to each population using the name
of the place in which the pond is located. If more
than two ponds are located in the same locality,
each of the populations in the area is numbered
(e.g. Yutani-1 and Yutani-2).
We conducted field observations to confirm
flowering and pollen production for 32 populations, namely Yutani-1 to Yutani-29 in Yokawa
Town, Heisou in Kakogawa City, and Biwakoh-1
and Biwakoh-3 in Kasai City, Hyogo Prefecture.
We visited these irrigation ponds weekly or
biweekly in the flowering seasons of 1995 and
1996, and recorded the occurrence or absence of
flowering. Pollen production was checked for

Clonal dominance in aquatic bladderwort

selected plants by observing anthers under microscope for each population.
Fruit and seed set ratios were investigated in five
populations, Yutani-3, Yutani-25, Heisou, Biwakoh-1 and Biwakoh-3. We visited those populations on one to four occasions in the flowering
season of 1996, and counted fruits and peduncles
that had already dropped their corollas. The fruit
set ratio was calculated for each population as the
number of fruits divided by the total number of
peduncles without corolla. Next, we sampled all or
some of the fruit-setting plants and dissected
fruits. The seed set ratio was calculated for each
mature fruit as the number of seeds divided by the
total number of ovules (approx. 100 per flower). In
cases where plants could not be collected, only the
fruit set ratio was determined in the field.

Pollination experiment
For the purposes of isozyme analysis, plants collected from 47 populations in Hyogo Prefecture
were cultivated in containers placed in the campus
of Kobe University. Among them, plants from
nine populations, Yutani-3, -8, -11, -12, -23, -25,
Heisou, Biwakoh-1 and -3, flowered during cultivation and were used in the pollination experiment. We paid attention to the possibility of crosspollination between the same clonal individuals,
which is equal to self-pollination. To avoid mixing
true cross-pollination and pseudo-cross-pollination, we employed two types of cross-pollination,
namely, mating with the plant which indicated the
same genotype from an isozyme study and with a
plant that had a different genotype. Plants were
treated in one of the following four ways when
they were in bloom. The treatments were:
1 Bagging. Each scape was bagged during the
flowering period.
2 Selfing. The flower was self-pollinated artificially and the whole scape was bagged.
3 Crossing with the same genotype. The flower
was cross-pollinated with pollen grains obtained
from an individual of the same genotype and the
whole scape was bagged.
4 Crossing with different genotype. The flower
was cross-pollinated with pollen grains
obtained from an individual with a different
genotype and the whole scape was bagged.

601

Plants collected from Yutani included male-sterile

plants. Those plants were used only in the fourth
treatment as maternal plants. Seed set ratio in each
fruit was transformed to its arcsine square root for
statistical analysis.

Isozyme analysis
We collected plants for isozyme analysis from 47
irrigation ponds located in the southern part of
Hyogo Prefecture. These included the 32 populations that were used for the field observations,
except for Yutani-16 and Yutani-20. Each of the
floating plant bodies was picked up intact from the
water. Sample sizes between populations ranged
from 1 to 146 (mean 13) because the amount of
growing plant differed among ponds. Samples
could include ramets of the same genet because
each individual can produce plural turions and
consequently plural ramets. When genotypic
variation within a population was detected in the
isozyme study, many plants were sampled additionally from the pond and the population genetic
structure was intensively examined. Seedlings
obtained from the pollination experiment were
also analyzed.
We used horizontal starch gel electrophoresis
following the method of Soltis et al. (1983). TrisHCl bufferPVP solution was employed as a
grinding buffer with a modification of mercaptoethanol concentration to 0.3%. Histidine-citrate
buffer, corresponding to system 9 in Soltis et al.
(1983), and its one-fourth dilution was used as an
electrode buffer and gel buffer, respectively. The
pH of the electrode buffer was adjusted to 6.5 by
an alteration in the amount of citric acid. A 1
5 cm length of vegetative shoot tip was ground in
0.3 ml grinding buffer and centrifuged at
50 000 g for 3 min. The supernatant was soaked
into wicks and inserted into the slit of the gel,
which was then electrophoresed for 15 min. After
the wicks have been drawn, the electrophoresis was
carried out again under constant current (35 mA)
for 6 h. The staining method of the gel followed
the method outlined by Soltis et al. (1983).
In preliminary tests, we analyzed the following
14 enzymes: aconitase (ACN), adenylate kinase
(ADK), alcohol dehydrogenase (ADH), formate
dehydrogenase (FDH), glutamate dehydrogenase
(GDH), hexokinase (HK), isocitrate dehydrogenase

602

S. Araki and Y. Kadono

(IDH), malate dehydrogenase (MDH), malic

enzyme (ME), mannose-6-phosphate isomerase
(MPI), 6-phosphogluconate dehydrogenase (6PG),
phosphoglucoisomerase (PGI), phosphoglucomutase (PGM), and shikimate dehydrogenase (S kDH).
Among these, clear and stable banding patterns
were detected in PGI and PGM of adult plants and
in PGI of seedlings. Phosphoglucomutase activity
in seedlings was low and could not be detected as
a banding pattern. Thus, we used electrophoretic
patterns of PGM and PGI in the genetic analysis of
adult plants and PGI for seedlings.

RESULTS
Pollen production
Among the 29 populations at Yutani, no flowers
were observed in 17 populations during the two
flowering seasons under natural conditions. However, plants collected from two populations,
Yutani-12 and -23, in which no flowers occurred
in natural conditions, bloomed in culture for the
pollination experiment (Table 1). All the flowering
plants sampled in Yutani-3 and some from Yutani25 were confirmed to produce pollen grains. Plants
in Heisou, Biwakoh-1 and -3 bloomed and produced pollen grains. In contrast, all the plants
sampled from Yutani-7, -8, -11, -17, -19, -21,
-24, -27, -28 and -29, and the majority of plants in
Yutani-25, did not produce pollen grains, thus
showing male sterility. Also plants of Yutani-12
and -23, in which flowers were observed only in
cultivation, showed male sterility.

Fruit and seed production in natural

populations
In natural populations, fruit setting was observed
only at Yutani-3, -25, Heisou and Biwakoh-1.
Fruit and seed set ratios are shown in Table 2.
Fruit set ratio in these populations ranged from 0
to 50%. In Heisou, fruit set ratio varied significantly with time (G-test: G = 25.0, d.f. = 2,
P < 0.005). A significant difference among populations on the same date was also observed. The
fruit set ratio on 24 September differed significantly between Biwakoh-1 and -3 ( G = 22.8,
d.f. = 1, P < 0.005), although both were pollenproducing populations. Seed set ratio was approx-

imately 10% (7.913.7%) in the three populations

investigated (Yutani-25, Heisou and Biwakoh-1)
and did not differ significantly (ANOVA: F3,19 =
0.262, P > 0.05).

Pollination experiment
The results of the pollination experiment are
shown in Table 3. There was no fruit set in the
bagging treatment. In contrast, more than 80% of
fruit set was observed in the three types of hand
pollination. There was no significant difference in
fruit set ratios among the three types of selfing
treatments, crossing with the same type and crossing with different types (G-test: G = 2.21,
d.f. = 2, P > 0.05). However, mean seed set ratios
differed significantly among the three types of
hand pollination (ANOVA: F2,100 = 77.7, P < 0.01).
Multiple comparisons using Tukey tests indicated
that the mean seed set ratios in the cases of selfing
and crossing with the same type, were approximately 10% and did not differ significantly from
each other (q100,3 = 0.157, P > 0.05). In contrast,
the mean seed set ratio in the case of crossing with
a different type, which was more than 40%, differed significantly from selfing (q100,3 = 16.4,
P < 0.01) and from crossing with the same type
(q100,3 = 16.3, P < 0.01).

Isozyme analysis
Six (AF) and three (13) types of electrophoretic
patterns were detected in PGM and PGI, respectively, from plants in natural populations (Fig. 1).
Although the ploidy level of Utricularia species is
not clear (Taylor 1989), two staining bands of
PGM in the anodal side can be interpreted as an
allozymic pattern because PGM is considered to be
a monomer enzyme (Weeden & Wendel 1989).
Thus, the pattern of types A and F indicate heterozygotes and the other four patterns indicate
homozygotes about the locus (PGM-1). In addition, the two bands in the cathodic side, which are
detected in types D, E, and F, may indicate allozymes at another locus (PGM-2). However, plants
indicating a homozygotic pattern, which has only
the nearest band to the cathode, were not found in
our study, thus, the two bands in the cathodic side
shown in types D, E, and F may be a consequence
of gene duplication. The electrophoretic patterns

Clonal dominance in aquatic bladderwort

603

Table 1 The occurrence or absence of flowering (Fl), pollen production (PP) and genetic types (GT) of two enzymes,
phosphoglucomutase and phosphoglucoisomerase, in the aquatic bladderwort, Utricularia australis R. Br., in each
pond studied
Localities of the ponds
Yokawa Town (3452 N, 13506 E)

Kakogawa City (3448 N, 13451 E)

Kasai City (3457 N, 13452 E)

Akashi City (3440 N, 13457 E)

Takarazuka City (3455 N, 13520 E)

Sanda City (3458 N, 13515 E)

Yutani-1
Yutani-2
Yutani-3
Yutani-4
Yutani-5
Yutani-6
Yutani-7
Yutani-8
Yutani-9
Yutani-10
Yutani-11
Yutani-12
Yutani-13
Yutani-14
Yutani-15
Yutani-16
Yutani-17
Yutani-18
Yutani-19
Yutani-20
Yutani-21
Yutani-22
Yutani-23
Yutani-24
Yutani-25
Yutani-26
Yutani-27
Yutani-28
Yutani-29
Heisou
Biwakoh-1
Biwakoh-2
Biwakoh-3
Abiki
Ohkubo
Oharano
Sakaino
Kamisasori-1
Kamisasori-2
Ono-1
Ono-2
Yamada
Aimoto-1
Aimoto-2
Ochibara-1
Ochibara-2

Fl

PP

GT

+
+

+
+

+
+
+

+
+
+
+
+
nd
+
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd

nd
nd
+
nd
nd
nd

nd
nd

nd
nd
nd
nd

nd

nd

nd

/+
nd

+
+
nd
+
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd

A1 (5)
A1 (3)
B2 (5)
B2 (5)
A1 (7)
A1 (5)
A1 (14)
A1 (8)
A1 (5)
A1/B2 (5)
A1 (10)
A1 (10)
A1 (10)
B3 (5)
E3 (12)
nd
A1 (10)
A1 (9)
A1 (10)
nd
A1 (1)
A1 (5)
A1 (5)
A1 (10)
A1/A3 (88)
A1 (7)
A1 (10)
A1 (19)
A1 (5)
C2/F3 (146)
A2 (10)
A2 (5)
A2 (5)
A1 (10)
D1 (10)
A1 (10)
A1 (25)
A1 (10)
A1 (17)
A1 (5)
A1 (24)
A1 (10)
A1 (15)
A1 (9)
A1 (4)
A1 (15)

(+), Flowers occurred or pollen production was confirmed; (), no flowering or pollen grains were produced; (nd), no data. The
notations of GT are the same as in Fig. 1 and the sample sizes in the enzyme studies are in parentheses.

604

S. Araki and Y. Kadono

Table 1 Continued.
Localities of the ponds
Yashiro Town (3456 N, 13502 E)

Simokume
Ureshino
Manose

Fl

PP

nd
nd
nd

nd
nd
nd

GT
A3 (5)
A3 (5)
A3 (5)

(+), Flowers occurred or pollen production was confirmed; (), no flowering or pollen grains were produced; (nd), no data. The
notations of GT are the same as in Fig. 1 and the sample sizes in the enzyme studies are in parentheses.

Table 2 Percentage of fruit and seed set under natural conditions

Seed set (%)
Population
Yutani-3

Yutani-25
Heisou

Biwakoh-1
Biwakoh-3

Date

Fruit set (%)

Range

Mean

5 September
20 September
25 September
2 October
22 August
6 August
20 August
18 September
24 September
24 September

3.3 (30)
50.0 (4)
0.0 (4)
0.0 (8)

48.1 (27)
0.0 (11)
4.8 (63)
18.0 (61)
0.0 (109)

2.013.0 (4)
0.029.2 (12)

0.030.0 (3)
5.410.6 (4)

9.3
7.9

13.7
8.2

All dates are in 1996.

Numbers in parentheses are sample sizes. (), data unavailable.

Table 3 Percentages of fruit and seed set in the pollination experiment under cultivation
Seed set (%)
Treatment
Bagging
Selfing
Crossing-same
Crossing-different

Fruit set (%)

Range

Mean

0.0 (56)
93.1 (58)
84.3 (51)
85.7 (21)

0.017.2 (45)
0.029.4 (41)
10.891.0 (17)

7.6
7.6
45.7

Crossing with plants of the same genotype.

Crossing with plants of a different genotype.
Numbers in parentheses are sample sizes. (), data unavailable.

of PGI could not be interpreted for genotype, but

variation in the banding pattern was considered to
reflect genotypic variation. The multilocus genotype of each plant was referred to as type A1,
type C2 etc. based on the combination of electrophoretic patterns of the two enzymes.
The genotypes detected in each population are
shown in Table 1. Intrapopulational variations of
multilocus genotype were only detected in three

populations (i.e. Heisou, Yutani-10 and Yutani25). In Heisou, 122 plants were type F3 and 24
were type C2 among the 146 plants tested. In
Yutani-25, 71 plants were type A1 and 17 were A3
among the 88 plants tested. In Yutani-10, the
coexistence of plants of type A1 and B2 was confirmed in a preliminary survey, but all plants in the
pond died out in the summer of 1995 for unknown
reasons, thus further data were unavailable. To the

Clonal dominance in aquatic bladderwort

contrary all the plants collected from male-sterile
populations were shown to be type A1. Plants of
other types that flowered (A2, A3, B2, C2 and F3)
produced pollen grains (Table 1).
The geographic distribution of each genotype in
the southern part of Hyogo Prefecture is shown in
Fig. 2. Different patterns of distribution were
found for each genotype. For example, populations

Fig. 1. Variations in the electrophoretic pattern of

two enzymes, phosphoglucomutase (PGM) and phosphoglucoisomerase (PGI) of Utricularia australis collected from 47 populations in the southern part of
Hyogo Prefecture, Japan. Each staining pattern is
referred to as type AF for PGM and type 13 for PGI.
The Rf values of each band in type F of PGM are 1.18,
1.12, 1.00 and 0.94, and in type 1 of PGI are 1.00,
0.88, 0.75, 0.63 and 0.50, respectively, in order from
the anode to the cathode.
Fig. 2. Geographical distribution of Utricularia australis
categorized by the genotype in
two enzyme systems, phosphoglucomutase and phosphoglucoisomerase. (
), A1; (), A2;
(), A3; (), B2; (), B3; (),
C2; (), D1; (), E3; (), F3.
The symbol characters of genotypes follow Fig. 1. Neighboring populations with the
same genotype are represented
by a single mark. (Kas.), Kasai;
(Kak.), Kakogawa; (Ak.),
Akashi; (Ya.), Yashiro; (Yo.),
Yokawa; (Sa.), Sanda; (Ta.),
Takarazuka.

605

of type A2 were distributed only in Kasai City, and

three of the four populations of type A3 were in
Yashiro Town. In eastern areas (Sanda and Takarazuka) all the studied populations were type A1. In
Yutani, where small ponds were densely located,
five types were found to occur. Among these five
types, type A1 was the most common, and the
other types (A3, B2, B3, and E3) were restricted to
several ponds (see Table 1).
Three types of polymorphism were detected
from 10 seedlings originated in self- or cross-pollination among plants of type 2 in PGI (Fig. 3).
One of these types, which is referred to as type 4,
was a type that was not found in natural populations. Figure 2 does not show a typical segregation
pattern of dimeric enzyme, although PGI is considered to be a dimer (Weeden & Wendel 1989).
There may be invisible or overlapping bands.

Fig. 3. Zymograms
of
phosphoglucoisomerase
detected in seedlings obtained from mating among type
2 parents. The symbolic number of each pattern is the
same as in Fig. 1. A pattern that was not found in the
natural populations is expressed as type 4 here.

606

S. Araki and Y. Kadono

Zymograms of seedlings that have parents of types

other than type 2 could not be determined because
of the limited availability of seedlings.

DISCUSSION
As all the male-sterile plants were type A1 in the
enzyme study, we inferred that male sterility in
the study area has the same origin and that all the
male-sterile plants belong to a single clonal
group. In addition, isozyme analysis suggested
clonal dominance not only in male-sterile populations, but also in seed-producing populations.
Seedlings of types 2, 3 and 4 of PGI occurred from
mating between type 2 parents. However, plants
of type 3 or 4 were not found in six of the seven
ponds in which plants of type 2 occurred
(Table 1). This suggests that recruitment from
seeds is rare and only clonal propagation from
turions has been prevalent in those populations. In
Heisou, plants of both type F3 and type C2
occurred. However it was not possible that the
type F3 plants had originated from mating among
type C2 plants because type F of PGM has two
alleles that type C does not share. In the same way,
type C2 cannot be the descendant of type F3
because type 2 of PGI cannot derive from recombination of the genotype of type 3. Furthermore, if
mating had occurred between these two types,
plants other than types C2 and F3 would have
originated, but such plants were not found in the
population. These results suggest that genetic
polymorphism in the Heisou population is not a
consequence of sexual reproduction, but of independent colonization of two different clonal lineages into the pond.
Clonal dominance in this species can be inferred
from genetic variation of PGM. Plants of types A
and F were heterozygotes in PGM-1. However,
coexistence of plants of type A and other homozygotic types was not found in our study. The coexistence of plants of type F and homozygotic type C
was found in Heisou, but we have already discussed above that type C2 cannot be a descendant
of type F3 in this population. Thus, the trace of
segregation of the two alleles of PGM-1 was not
found within each of the populations, although
polymorphism among populations may be a consequence of sexual recombination.

Why is recruitment from seeds rare? Clonal

dominance occurs in many aquatic macrophyte
populations (Verkleij et al. 1983; Van Wijk 1988,
1989; Les 1991; Barrett et al. 1993; Grace 1993;
Hollingsworth et al. 1996b). The reasons for this
are not necessarily clear, but some explanations
have been proposed, including inadequate light
intensity or dissolved oxygen concentration for
seedlings. In the case of U. australis, most of the
turions in cultivation remained floating throughout winter, whereas the seeds sunk when they were
suspended in water (S. Araki and Y. Kadano, pers.
obs., 1996). If seeds are buried in sediment at the
bottom of the pond, seed and/or seedling survival
may be reduced because of low light and oxygen.
Even if seeds are not buried, seedlings will be less
vigorous because the resource storage capabilities
of a seed are far less than a turion. The dry weight
of a turion varies according to its size. The smallest
turion, which is approximately 1 mm long, weighs
approximately 0.07 mg, whereas the largest one,
approximately 10 mm, weighs approximately
40 mg. However, all seeds are approximately
1 mm in size and weigh approximately 0.04 mg
(S. Araki & Y. Kadono, unpubl. data, 1996), which
is 1/1000 in comparison with a turion.
A mosaic pattern in the geographic distribution
of the multilocus genotype was found in the southern part of Hyogo Prefecture. It is considered that
this pattern has been derived by clonal dominance
in reproduction and the colonization processes of
the species. Waterfowl movement may play a role
in colonization of aquatic plants by trapping seeds
in feathers (Cook 1990; Vivian-Smith & Stiles
1994). In the case of U. australis, the occurrence of
the same genotype in nearby populations and the
spread of the male-sterile type A1 suggest that dispersal of vegetative propagules has occurred. It is
highly likely that each genet succeeds in establishment as a founder spreads into neighboring habitats by vegetative means. Thus, metapopulation
becomes a mosaic of genets as illustrated by genotype distribution.
In addition to the rarity of seed or seedling survival, the chance of sexual reproduction is
restricted by low seed production. The fruit set
ratio in natural populations is low when compared
with hand pollination and sometimes decreases to
zero. More intensive research on the relationship
between the amount of fruit production and

Clonal dominance in aquatic bladderwort

pollinator activity or weather conditions is needed
for clarification of the reasons for fruit set ratio
fluctuation in natural populations.
Seed set ratio appears to be diminished for some
genetic reasons. The pollination experiment
showed that seed set ratios were low in the cases of
selfing and crossing between plants of the same
multilocus genotype. In contrast, crossing
between plants of different genotypes resulted in
higher seed set ratios. If we assume that plants
with the same multilocus genotype belong to the
same clonal lineage, crossing with the same type
corresponds to intraclonal pollination, which is
equal to self-pollination. Crossing with a different
type corresponds to interclonal pollination. Thus,
the results of the pollination experiment indicate a
low seed set ratio in intraclonal mating and a high
seed set ratio in interclonal mating. Low seed set
ratios observed in natural populations are also considered to be consequences of the clonality of the
population. There are two explanations for low
seed set ratio in intraclonal pollination:
1 Inbreeding depression. Inbreeding depressions
may have prevented normal seed development
(Seavey & Bawa 1986; Husband & Schemske
1996; Manicacci & Barrett 1996; Mahy &
Jacquemart 1998). Even if the parent plants
grow normally, homozygosis that prevents seed
development may occur through mating.
2 Self-incompatibility. Self-incompatibilities in
many plant species are genetically controlled by
a single multi-allelic S locus (Haring et al.
1990; Lee et al. 1994; Murfett et al. 1994;
Richman et al. 1995). If U. australis is selfincompatible, mating within the same clonal
individuals results in low seed set ratios because
the parents have the same S allele.
Because habitats such as ponds and lakes are isolated from each other, various types of genetic features of the population are easily affected by
founder effects or genetic bottlenecks (Hedrick &
Parker 1998; Johnson & Black 1998; Schug et al.
1998). Genetic bottlenecks affect the number of
genets in a population and/or the number of S alleles within a population (Imrie et al. 1972; Reinartz
& Les 1994). In the case of U. australis, it is probable that the number of genets in one population is
one or a few at the time of establishment of a new
population. If such a founder effect is repeated dur-

607

ing the process of colonization, the number of genets in a population decreases further and finally
becomes one. In this situation, all the individuals
in a population belong to the same genet and have
the same S allele, thus cross-pollination within the
population is equal to self-pollination, which may
result in inbreeding or incompatible effect.

ACKNOWLEDGEMENTS
We are grateful to Mr T. Suzuki, Museum of
Nature and Human Activities, for his suggestions
for improving the techniques of isozyme detection
and to the owners of the irrigation ponds investigated.

REFERENCES
A RAKI S. (2000) Isozyme differentiation between
two infraspecies taxa of Utricularia australis R. Br.
(Lentibulariaceae) in Japan. Acta Phytotaxonomica et
Geobotanica 51: 3136.
B ARRAT -S EGRETAIN M. (1996) Germination and
colonisation dynamics of Nuphar lutea (L.) Sm. in a
former river channel. Aquatic Botany 55: 3138.
B ARRETT S. C. H. (1980) Sexual reproduction in
Eichhornia crassipes (water hyacinth). II. Seed production in natural populations. Journal of Applied
Ecology 17: 113124.
B ARRETT S. C. H., E CKERT C. G. & H USBUND B.
C. (1993) Evolutionary processes in aquatic plant
populations. Aquatic Botany 44: 105145.
C OOK C. D. K. (1990) Seed dispersal of Nymphoides
peltata. Aquatic Botany 37: 325340.
E CKERT C. G. (2002) Loss of sex in clonal plants.
Evolutionary Ecology 15: 501520.
E CKERT C. G., D ORKEN M. E. & M ITCHELL S. A.
(1999) Loss of sex in clonal populations of a flowering plant, Decodon ferticillatus (Lythraceae). Evolution 53: 10791092.
E LLSTRAND N. C. & R OOSE M. L. (1987) Patterns
of genotypic diversity in clonal plant species.
American Journal of Botany 74: 123131.
E RIKSSON O. (1989) Seedling dynamics and life
histories in clonal plants. Oikos 55: 231238.
E RIKSSON O. (1997) Clonal life histories and the
evolution of seed recruitment. In: The Ecology and
Evolution of Clonal Plants (eds H. De Kroon & J.
Groenendael) pp. 211226. Backhuys Publishers,
Leiden.

608

S. Araki and Y. Kadono

G RACE J. B. (1993) The adaptive significance of

clonal reproduction in angiosperms: an aquatic
perspective. Aquatic Botany 44: 159180.
H ARING V., G RAY J. E., M C C LURE B. A.,
A NDERSON M. A. & C LARKE A. E. (1990) Selfincompatibility: a self-recognition system in
plants. Science 250: 937941.
H EDRICK P. W. & P ARKER K. M. (1998) MHC
variation in the endangered Gila topminnow. Evolution 52: 194199.
H OLLINGSWORTH P. M., P RESTON C. D. &
G ORNALL R. J. (1996a) Isozyme evidence for the
parentage and multiple origins of Potamogeton
suecicus (P. pectinatus P. filliformis, Potamogetonaceae). Plant Systematics and Evolution 202: 219
232.
H OLLINGSWORTH P. M., P RESTON C. D. &
G ORNALL R. J. (1996b) Genetic variability in
two hydrophilous species of Potamogeton, P. pectinatus and P. filiformis (Potamogetonaceae). Plant Systematics and Evolution 202: 233254.
H USBAND B. C. & S CHEMSKE D. W. (1996) Evolution of the magnitude and timing of inbreeding
depression in plants. Evolution 50: 5470.
I MRIE B. C., K IRKMAN C. J. & R OSS D. R. (1972)
Computer simulation of a sporophytic self-incompatibility breeding system. Australian Journal of
Biological Sciences 25: 343349.
J OHNSON M. & B LACK R. (1998) Increased genetic
divergence and reduced genetic variation in populations of the snail Bembicium vittatum in isolated
tidal ponds. Heredity 80: 163172.
K ADONO Y. (1993) Lentibulariaceae. In: Flora of
Japan, Volume IIIa (eds K. Iwatsuki, T. Yamazaki,
D. E. Boufford & H. Ohba) pp. 400404. Kodansha, Tokyo.
K ADONO Y. (1994) Aquatic Plants of Japan. Bunichi Sogo Shuppan, Tokyo (in Japanese).
K LIMES L., K LIMESOVA J., H ENDRIKS R. & V AN
G ROENENDAEL J. (1997) Clonal plant architecture: a comparative analysis of form and function.
In: The Ecology and Evolution of Clonal Plants (eds
H. De Kroon & J. Van Groenendael) pp. 129.
Backhuys Publishers, Leiden.
K OMIYA S. & S HIBATA C. (1980) Distribution of
the Lentibulariaceae in Japan. Bulletin of Nippon
Dental University, General Education 9: 163212.
L EE H., H UANG S. & K AO T. (1994) S proteins
control rejection of incompatible pollen in Petunia
inflata. Nature 367: 560563.
L ES D. H. (1988) Breeding systems, population
structure, and evolution in hydrophilous
angiosperms. Annals of Missouri Botanical Garden
75: 819835.

L ES D. H. (1991) Genetic diversity in the monoecious hydrophile Ceratophyllum (Ceratophyllaceae).

American Journal of Botany 78: 10701082.
M AHY G. & J ACQUEMART A. (1998) Mating system of Calluna vulgaris: self-sterility and outcrossing estimations. Canadian Journal of Botany 76:
3742.
M AKINO T. (1914) Observations on the flora of
Japan. The Botanical Magazine, Tokyo 28: 130.
M ANICACCI D. & B ARRETT S. C. H. (1996) Fertility differences among floral morphs following
selfing in tristylous Eichhornia paniculata (Pontederiaceae): inbreeding depression or partial incompatibility? American Journal of Botany 83: 594
603.
M IKI S. (1935) New water plants in Asia Orientalis
III. The Botanical Magazine, Tokyo 49: 847852.
M UIR A. M. (1995) The cost of reproduction to the
clonal herb Asarum canadense (wild ginger). Canadian Journal of Botany 73: 16831686.
M URFETT J., A THERTON T. L., M OU B., G ASSER
C. S. & M C C LURE B. A. (1994) S-RNase
expressed in transgenic Nicotiana causes S-allelespecific pollen rejection. Nature 367: 563566.
N AKAMURA T. & K ADONO Y. (1993) Chromosome number and geographical distribution of
monoecious and dioecious Hydrilla verticillata
(L. f.) Royle. (Hydrocharitaceae) in Japan. Acta
Phytotaxonomica et Geobotanica 44: 123140.
N AKAMURA T., S UZUKI T. & K ADONO Y. (1998)
A comparative study of isoenzyme patterns of
Hydrilla verticillata (L. f.) Royle in Japan. Journal of
Plant Research 111: 581585.
R EINARTZ J. A. & L ES D. H. (1994) Bottleneckinduced dissolution of self-incompatibility and
breeding system consequences in Aster furcatus
(Asteraceae). American Journal of Botany 81: 446
455.
R ICHMANN A. D., K AO T., S CHAEFFER S. W. &
U YENOYAMA M. K. (1995) S-allele sequence
diversity in natural populations of Solanum carolinense (Horsenettle). Heredity 75: 405415.
S CHUG M. D., D OWNHOWER J. F., B ROWN L. P.,
S EARS D. B. & F UERST P. A. (1998) Isolation
and genetic diversity of Gambusia hubbsi (mosquitofish) populations in blueholes on Andros Island,
Bahamas. Heredity 80: 336346.
S EAVEY S. R. & B AWA K. S. (1986) Late-acting
self-incompatibility in angiosperms. Botanical
Review 52: 195219.
S MOLDERS A. J. P., D EN H ARTOG C. & R OELOFS
J. G. M. (1995a) Observations on fruiting and
seed-set of Stratiotes aloides L. in the Netherlands.
Aquatic Botany 51: 259268.

Clonal dominance in aquatic bladderwort

S MOLDERS A. J. P., D EN H ARTOG C. & R OELOFS
J. G. M. (1995b) Germination and seedling development in Stratioites aloides L. Aquatic Botany 51:
269279.
S OLTIS D. E., H AUFLER C. H., D ARROW D. C. &
G ASTONY G. J. (1983) Starch gel electrophoresis
of ferns: a compilation of grinding buffers, gel and
electrode buffers, and staining schedules. American
Fern Journal 73: 927.
T AMURA M. (1981) Lentibulariaceae. In: Wild Flowers of Japan III (eds Y. Satake, J. Ohwi, S. Kitamura, S. Watari & T. Tominari) pp. 137139.
Heibonsha, Tokyo (in Japanese).
T AYLOR P. (1989) The Genus Utricularia: A Taxonomic Monograph. Her Majestys Stationery Office,
London.
V AN W IJK R. J. (1988) Ecological studies on Potamogeton pectinatus L. I. General characteristics,
biomass production and life cycles under field conditions. Aquatic Botany 31: 211258.
V AN W IJK R. J. (1989) Ecological studies on Potamogeton pectinatus L. III. Reproductive strategies
and germination ecology. Aquatic Botany 33: 271
299.
V ERBURG R. W. & D URING H. (1998) Vegetative
propagation and sexual reproduction in the woodland understorey pseudo-annual Circaea lutetiana
L. Plant Ecology 134: 211224.

609

V ERKLEIJ J. A. C., P IETERSE A. H., H ORNEMAN

G. J. T. & T ORENBEEK M. (1983) Comparative
study of the morphology and isoenzyme patterns
of Hydrilla verticillata (L.f.) Royle. Aquatic Botany
17: 4359.
V IVIAN -S MITH G. & S TILES E. W. (1994) Dispersal of salt marsh seeds on the feet and feathers
of waterfowl. Wetlands 14: 316319.
W ATERWAY M. J. (1994) Evidence for the hybrid
origin of Carex kunieskernii with comments on
hybridization in the genus Carex (Cyperaceae).
Canadian Journal of Botany 72: 860871.
W EEDEN N. F. & W ENDEL J. F. (1989) Genetics of
plant isozymes. In: Isozymes In Plant Biology (eds D.
E. Soltis & P. S. Soltis), pp. 4672. Dioscorides
Press, Portland.
Y AMAMOTO I. & K ADONO Y. (1988) Distribution
of Utricularia vulgaris L. var japonica (Makino)
Tamura and Utricularia tenuicaulis Miki in the
southern part of Hyogo Prefecture, southwestern
Japan. The Journal of Phytogeography and Taxonomy
36: 7275 (in Japanese with English summary).
Y AMAMOTO I. & K ADONO Y. (1990) A study on
the reproductive biology of aquatic Utricularia
species in southwestern Japan. Acta Phytotaxonomica et Geobotanica 41: 189200 (in Japanese with
English abstract).