全 文 ：广 西 植 物 Guihaia 30(3)：306— 315 2OlO年 5月
海滨锦葵花内五柱头裂片弯曲的独自调节
阮成江，周丽君，曾方玉
(大连民族学院 生物技术与资源利用国家民委教育部重点实验室，辽宁 大连 116600)
摘 要：海滨锦葵花柱有 5柱头裂片，若传粉失败，柱头裂片向下弯曲使柱头与自身花粉接触，发生延迟自交。
传粉不足条件下发生的延迟自交可能提供繁殖保障。该研究定量分析了花粉搁置和授精胚珠数对花内未授粉
柱头裂片运动的影响，并测定 了花粉一胚珠比及柱头可授性和花粉生活力。结果表明．花内未授粉柱头裂片的
弯曲不受其他柱头接受花粉量及授精胚珠数的影响，仅响应于其自身是否接受到花粉，花份一胚珠比值显示海
滨锦葵交配系统属兼性异交；未授粉柱头经弯曲与自身花粉授触时的强柱头可授粉和高花粉生活力为自交授粉
的发生提供了可能。包括柱头裂片运动在内的多个花性状有机地展示了一种新花内混合交配系统，且花内柱头
裂片弯曲的独自调节为从花水平验证被广泛接受的自交进化解释——繁殖保障假说提供了可能。
关键词：海滨锦葵；独自调节；柱头裂片弯曲；花粉一胚珠比；混合交配系统；繁殖保障
CLC number：Q949 Document code：A Article ID：1000—3142(201O)03—0306—10
Independent regulation of curvature of five stigma
lobes within Kosteletzkya virginica flowers
RUAN Cheng-Jiang，
(Key Laboratory of Biotechnology& Bio-Resources
ZHOU Li—Jun，ZENG Fang-Yu
Utilization，Dalian Nationalities University，Dalian 116600，China)
Abstract：Kosteletzkya virginica has a pistil with five stigma lobes．If un—pollinated，individual stigma lobes
downward curve to bring them contact with a flower’S own anthers，possibly resulting in delayed selfing，
which might provide reproductive assurance when pollinators are scarce．It was observed，but not experimen—
taly tested，that each of the five lobes within flowers acts independently．Here，we quantitatively tested the
effects of pollen loads and seed set on the curvature of un—pollinated lobes within flowers，pollen—ovule ratio
and stigma receptivity and pollen viability．Results showed that the curvature of un—polinated lobes within
flowers did not appear to be a response tO the amount of polen grains deposited on the adjacent stigmas and
seed set，depending on if each received polen．The value of polen—ovule ratio indicated that mating system be—
longed tO facuhative xenogamy．When un—pollinated stigmas were in contacting with its own anthers via style
movements，high stigma receptivity and pollen viability provided a possible for resulting in delayed selfing．
Multiple floral traits including independent curvature of stigma lobes harmoniously exhibit a mixed mating sys—
tem within flowers．Our data provided the first empirical demonstration that the curvature of the five stigma
lobes within flowers regulates independently，which provides incentive for testing at the flower level the repro—
ductive assurance widely accepted explanation for the evolution of selfing．
Key words：Kosteletzkya virginica；independent regulation；stigma lobe curvature；pollen—ovule ratio；mixed
mating system ；reproductive assurance
Received date：2008～1 2-06 Accepted date：2009—05—10
Foundation item：Supported by the National Natural Science Foundation of China(30500071)
Biography：RUAN Cheng一]iang(1972一)，male，Ph．D，Professor，mainly engaged in plant evolutionary adaptation and genetics and breeding
(E-mail)ruan@ dlnu．edu．cn．
3期 阮成江等：海滨锦葵花内五柱头裂片弯曲的独自调节 307
1 Intr0duction
Flowers exhibit a wide array of anatomical
specializations and behaviors which are strongly
linked to plant pollination and mating system．
These consist of pistil(style)movements such as
stigma closure(Fetscher & Kohn，1999；Yang et
“Z．，2004)，late curling of stigma lobe(Buttrose et
。Z．，1977；Klips＆ Snow，1997；Ruan et a1．，2004)，
flexistyly(Li et a1．，2001)and style curvature(Cul—
ley，2002)． They also include stamen(pollen)
movement．such as stamen curvature(Rathcke &
Rea1， 1993)， stamen elongation (Eckert &
Schaefer，1998)，pollen sliding(Wang et a1．，2004)，
pollen catapult(Edwards et a1．，2005)and anther
rotation(Liu et a Z．，2006)．Other floral activities
include corolla abscission(Dole，1990)and wilting
(Sun et a1．，2005)．The putative adaptive signifi—
cance of floral behavior includes reduction in intra—
floral male—female interference(Edwards et a1．，
2005)，promotion of outcrossing(Verma et a1．，
2004；Petew& Johnson，2006)，achievement of de—
layed autonomous selling(Culley，2002；Etcheverry
et al。，2003)and response to harsh environments
(Bynum & Smith，2001；Huang et a1．，2002)．
The facuhative nature of autonomous pollina—
tion through curvature of un-pollinated styles to—
wards its anthers in M alvaceae is intriguing．It has
been reported in Hibiscus laevis(Klips& Snow，
1997)，H．trionum (Buttrose et(If．，1977；Seed et
a1．2006)and Kosteletzkya virginica(Ruan et a1．，
2005a)． Various hypotheses addressing the adap—
tive significance of this curvature have been pro—
posed，such as promotion of outcrossing(Stephens，
1 948)，achievement of delayed selfins(Blanchard，
l976；Klips& Snow，1997；Ruan et a1．，2004)and
first preference to outcrossing but facilitating de—
layed selfins if outcrossing fails(Buttrose et a1．，
1 977)． The lobed curvature facilitates un—pollina-
ted stigmas to receive cross—pollen by positioning
the stigmas in the flight path of pollinators subse—
quent to a predominantly male phase of anthesis
when the stigmas are held together，projecting for—
ward．After this，it can lead to delayed self—pollina—
tion if they are in contacting with the anthers when
the opportunity of outcrossing has passed．
In H．trionum var．trionum ，curvature is re—
versible，with partially curved styles regaining an e—
rect position after receiving pollen；and the effect of
pollination is broadly inhibitory such that applica—
tion of pollen to a single stigma lobe is sufficient to
prevent curvature of adj acent lobes(Buttrose et
“Z．，1977)．In contrast，the five styles within H．
trionum var．vesicarius flowers act independently
(Seed et a1．，2006)，and if un—reversible curvature
ceases depend on the amount of pollination that
each receives． Pollinated styles of H ．1aevis still
curve(moved angle less than 135。)，although to a
lesser extent than un-pollinated styles(moved an—
gles more than 135。)(Klips& Snow，1997)．In
contrast，the curvature of un—pollinated stigma
lobes in K．virginica permanently ceases once pol—
lination takes place，and it is un—reversible(Ruan et
nZ．，2004)．It was observed，but not experimental—
ly tested，that the curvature of each of the five stig—
ma lobes within K．virginica flowers act independ—
ently；namely pollination of one to four stigmas
does not prevent the curvature of the adjacent un—
pollinated lobes．
Kosteletzkya virginica(Malvaceae)，a peren—
nia1 herbaceous plant，is native to salt marsh in A—
merican(Gallagher，1985)． In l993，it was intro—
duced into China as a potential species for impro—
ving soil and developing ecologically sound saline
agriculture．The fruit capsule has five ventricles，
each containing a single ovule．Flowers of K．vir—
ginica have a monadelphous column(1 8—5 7 stamens
are fused into fl column)，surrounding pistil with
five stigma lobes(Ruan et a1．，2005b)．Each plant
produces one to twenty～five flowers everyday，and
each flower only lasts one day． Polination of K．
virginica includes insect pollinators(e．g．Bornbus
specious，Colias hyale and Manduea floresta 71)and
delayed seIf_pollination via stylar movements(Ruan
et a1．，2005b)．If newly—opened flowers are polli—
308 广 西 植 物 3O卷
nated，stigma lobes remain erect；but if un—pollina～
ted，the stigma lobes begin curving and will toucb
the anthers when completing the flowering life—
span．
This study addresses two main questi0ns
．
First，if the curvature of un—pollinated stigma lobes
within flowers responds to seed set and different
number polen loads on the adj acent stigmas?We
applied different numbers of pollen grains to the
stigmas from one to four to observe curvature re—
sponses of the adjacent un—pollinated stigma lobes
and pollinated lobes within flowers
． This was used
to test the effects of pollen loads and seed set 0n
the curvature of un-pollinated lobes within flow—
ers．Second，what is mating system the curvature
of un—pollinated lobes within flowers cOmbining
with other floral traits exhibit?W e tested Dolien—o
vule ratio，stigma receptivity，pollen viability and
pollination limitation to analyze the mating system
influenced by multiple floral traits
．
2 Materials and．Methods
Seeds of K．w；rginica from the Halophyte Bio—
technology Center(University of Delaware
，USA)
were sOwn at the Dafeng site for controlled experi—
ments and assessment in l993．The Dafeng site is
located on the tideland of Yancheng City(1ongitude
119。27 一 120。54 E，latitude 32。34 一 34。28 N)
，
Jiangsu，China．The continuous field exDeriments
from 1994 to 2004 demonstrated that K
． irg cn
is a good halophytic species with a great potentia1
industrially and ecologically for new marshlands in
China．The Dafeng naturalized populations of K
．
virginica grow more than 100 000 individua1s in
2004．Seeds，collected from the Dafeng naturalized
population in 2004，were sown in Dalian site in the
spring of 2005． Individuals growing in Dalian site
were over 2O 000 in the autumn of 2005
． The
Dalian site is located on the tideland of Dalian citv
(1ongitude 121。39 E，latitude 38。55 N)，Liaoning
，
China
2．1 Effects of pollen loads and seed set on curvature
of un-pollinated lobes
In spring 2 0 0 7，two thousands mixed seeds
from Dalian naturalized population were germinated
on a compost substrate(humus：pearlstone--3 ：1)
in polystyrene plates flowerpots(1 00 seeds per
plate)in the culture room ，with 45 ／~mo1．s～ ．n1_
light，a 14-h photoperiod，70 9／5～80 relative hu—
midity and at 28 ℃ ．After 10— 15 days，when grew
to about 2— 3 cm in length．seedlings were trans—
planted into the compost substate(ash and slay with
sand，or fluvial heavy sand)in an insect—free green—
house(Dalian Nationalities University)
． Individuals
were kept in a controlled growing condition at 20．——
25℃(25—30℃ for flowering period)with a 16h／
8 h light／dark regime and under a relative humiditv of
7O ±5％．We gained 1493 individuals that n0rma1
grew to over 1 m tal in．M ay，2007．
To determine the effects of different amounr
of pollen loads and seed set on the eurvature of un-
pollinated stigma lobes，eight flowers were ran—
domly selected from eight different individua1s(one
flower per individua1)randomly selected daily，and
randomly variable amounts of pollen grainS from
the un—selected individuals were applied to one to
four stigmas within a flower(two flowers per treat—
ment)at 7：00，at that time all stigma lobes are iust
curving． After pollination，the flowers were con—
tinuously observed(4-5 h)to determine if Dol1ina—
tlon prevented the curvature of pollinated stigmas
and if pollination was sufficient to inhibit the cur—
vature of the adjacent un—pollinated lobes within
flowers． Once un—pollinated lobes continued to
curve and their location relative to the monade1一
phous column was clearly different from the lobes
that ceased curvature after pollination
， these was
cut using sharp，sterile scissors to prevent delayed
autonomous selling． This phenomenon is easv to
observe because it is the same as the angle between
each of un—pollinated stigma lobe within a flower
and the line along the entire length of the monade1一
phous column． The angle of the stigma lobe wil
not change once It ceases curving after pollination
3期 阮成江等：海滨锦葵花内五柱头裂片弯曲的独自调节 309
takes place(see Fig．1)；in contrast，since the curva—
ture of un—pollinated lobes will 1ast untiI OCCur_
rence of pollination，the angle continues to decrease
gradually(see Fig．1)． The top of hand—pollinated
lobes(2 mm length)in each treated flower were re-
moved and collected six hours after pollination at
about 1 3：00．This cut did not influence the flowers
to set fruit，because pollen tubes required 1
． 0— 2．5
h to pass style and reach ovules and only required
less than 0．5 h to pass the lobes(Ruan et a1． ，u n—
published data)．The number of pollen grains per
treated flower was estimated using a dissecting mi—
croscope(Olympus SZ2-ILST)． Mature capsules
were harvested and the number of seeds in each
capsule was counted． For each treatment(one to
four stigmas)，fifty flowers from fifty different in—
dividuals(one flower per individua1)were treated in
a total of 25 days．
W e determined whether cessation of the cu卜
vature of un—pollinated lobes corresponded to the a—
mount of pollen grains deposited on the stigmas
from one to four and seed set from pollination of
stigmas from one to four or not．W e also examined
the effect of the number of pollen grains deposited
on a single stigma on itself curvature by the data of
one stigma treatment．
Our previous experiments on a deleterious
effect of cut stigma lobes(about 2 ram)on seed set
by physical damage to the flowers indicated that it
did not produce negative effects on seed set，this
was because there was no significant difference in
the mean number of seeds per capsule between the
hand—pollinated flowers that four stigmas was cut
and the intact flowers that only one stigma was
hand—pollinated(mean± s．e．，4．29± 0．08 versus
4．3l_-+-0．07，a two—sample t—test：t：一0．541，d厂一
98，P= 0．589)．
2．2 Pollination limitation and pollen-ovule ratio
W e randomly selected and tagged 20 pairs of
individuals at the Dalian field population in 2005，
with the members of each pair 1oeated within 3 m
of each other and different pairs 1ocated over 10 m
apart from each other．In each day across the flow—
ering season，for each individuai each pair．alj flow—
ers exposed to open pollination as the control
group(Contro1)；where all flowers on the second
individual were also exposed to natural pollinators，
but we supplemented the number of pollen grains ar—
riving to the stigmas by hand—polinating al flowers
with pollen from at least 5 donor plants(Pollen-sup—
plementation treatment)． Artificial pollination was
repeated in each flower at least twice(one at 9：O0，
and another at 1 4：0 0)to ensure successful pollina—
tion．Mature fruits were harvested for COUnting fruit
sets and seeds per capsule，and the comparison be—
tween pollen-supplementation treatment and control
flowers was used to assess pollen limitation in K
．
virginica by the method of Rathcke(2003)． W e
tested the differences in fruit set and in the number
of seeds per capsule between control and pollen—-sup—．
plemented individuals，using a two—sample t-test test
(SPSS l1．0)．
In addition，we measured pollen—ovule ratio(P／
O)by the method of Cruden(1 977)．Five flowers
were randomly selected from five different individ—
uals(one flower per individua1)randomly selected
from the Dalian field site． All stamens on the mon—
adelphous column per flower before anthers dehisce
were collected，and the number of pollen grains per
stamen was estimated using the dissecting micro—
scope(Olympus SZ2一ILST)．The number of ovules
per flower was counted under the dissecting micro—
scope(Olympus SZ2一ILST)． A total of 1 20 flow—
ers were treated in 2 4 days．Breeding system of K．
virginica was estimated by the data of Cruden
(1 9 7 7)：cleistogamous with pollen—ovule ration
ranging from 2．7 to 6．7，l8．1— 38．9 for autoga—
mous—obligate，3 1．9— 3 9 6．0 for autogamous—fae—
ultative，245．0— 1 894．0 for xenogamous—faculta—
tive and 2 108．0～ 1 95 525．0 for xenogamous—ob—
ligate．
2．3 Stigma receptivity and pollen viability
To test receptivity in the Dalian field popula—
tion in 2005。we used an indirect assay modified
from Kalisz et aZ．(1999)and Pu et a1．(2008)that
detects the presence of stigma peroxidase． W hen
31O 广 西 植 物 3O卷
receptive stigmas are placed in a 3％ solution of
hydrogen peroxide，vigorous bubbling OCCURS on
the stigmatic surface．Non receptive stigmas do not
produce bubbles． In the field，gynoecia were dis—
sected from flowers on plants loosely covered with
muslin bags before flowers open，at intervals of 2 h
from 6：O0 to 18：O0 on the day of anthesis．Individ—
ual stigma lobes were immediately sandwiched be—
tween two cover slips with a drop of hydrogen per—
oxide． The stigmatic area was examined under a
dissecting microscope in the field． Stigmas were
scored as positive for peroxidase activity only if we
observed vigorous bubbling across the entire sur-
face of the stigma． For each time interval，three—
hundred fifty(350)lobes of 70 flowers from 70 dif～
ferent individuals(one flower per individua1)were
tested for peroxidase activity．
W e applied TTC(2，3，5一triphenyl tetrazolium
chloride)to examine pollen viability during a flow～
er’s lifetime in the field populations by the method
of Huang et a1．(2004)．At interval of 2 h from 6：
00 to 18：00 on the day of anthesis，30 flowers were
collected from thirty flowering plants，and pollen
viability was assessed by the percentage of pollen
stained red． Freshly harvested pollen w3s dusted
onto a microscope slide with a brush to which four
or five drops of stain were added． Then the slide
was immediately covered with a cover slip and the
edges sealed with nail varnish． The percentage of
pollen of ZO0— 300 grains per slide(one to three
slides for each treated flower)that exhibited the
appropriate staining reaction was determined using
an Olympus I× 7 1 microscope at× 100 magnifica—
tion．
3 Resuhs
3．1 Effects of pollen loads and seed set on curvature
of an—pollinated lobes
The un～pollinated stigma lobes within flowers
still curved after 1——4 stigmas received different a—
mount of pollen grains(Table 1，Plate I)，and this
curvat ure was independent of the number of seeds
per capsule(Table 1)． For one stigma，pollination
of different amount of pollen grains ranging from 1
to 162 immediately and irreversibly prevented itself
curvature．
3．2 Pollen—ovule ratio and pollination limitation
The number of pollen grains per flower was
2831．56±43．14(mean_4-s．e．，n一 120)．The hum—
ber of ovules per flower was five．The pollen—ovule
ratio was 566．31± 18．63，ranging from 296．64 to
939．36． This indicated that breeding system be—
longs to facuhative xenogamy by the criterion of
Cruden(1977)．
Table 1 Effects of polen loads and seed set
on curvature of un-polinated stigma lobes
Note：a：For treatments of stigmas from two to four。mean was the av—
erage of total numbers of pollen grains deposited on the all stigmas per
flower．
Polen 1imitation did not occur in K．virginica，be—
cause there were no significant differences in fruit set
and in the number of seeds per capsule between control
and pollen—supplemented individuals(fruit set：mean±
s．e．，81．32 0A4-_2．64 v5．82．41％±1．86 0A，a two-
sample t-test：t一一1．837，dr=38，P一0．134；seeds per
capsule：4．43±0．07 ．4．51±0．07，t=-1．632，df=
28，P一0．079)．
3．3 Stigma receptivity and pollen viability
Stigmatic peroxidase tests indicated that stigma
receptivity was 56．0Z ±10．2 at 06：00，and opti—
mum stigma receptivity was about 10：00(Fig．3A)．At
16：00，82．86 ±6．91 of the stigma lobes collect—
ed in the field had receptive stigmas．Pollen viabili—
ty at 6：00 was 53．79 4-1．97 ，and it was higher
from 8：00 to l2：00(Fig．3B)．
3期 阮成江等：海滨锦葵花内五柱头裂片弯曲的独 自调节 31l
Plate T Independent regulation of curvature of the five stigma lobes within Kosteletzkya virginica flowers a：one
stigma lobe(solid arrowhead)ceased curving after pollination took place，the remaining four un—pollinated lobes(dashed arrowheads) continued
to curve．b：two stigma lobes(solid arrowheads) ceased curving after polination took place，the remaining three un—polinated lobes(dashed ar—
rowheads)continued to curve；c：two stigma lobes(solid arrowheads)ceased curving after polination took place，the remaining three un—polli—
nated lobes continued to curve．0ne of them(triangle)ceased curving after polination occurred in the progress of curvature，the last two stig—
ma lobes(dashed arrowheads)continued to curve until contacting the anthers to self-polinate；d：two stigma lobes(solid arrowheads)ceased
curving after pollination occurred at the curved primary stage．Other two stigma lobes(triangles)ceased curving after pollination occurred at
the curved middle stage．The last stigma lobe(dashed arrowhead)was not hand—polinated，and continued to curve until contacting the anthers
to self-pollinate．
4 Discussion
The present work provided a first empirical
demonstration that independent regulation of cur—
vature of un—pollinated stigma lobes within K．vir—
ginica flowers does not appear tO be a response to
the amount of pollen grains deposited on the adja—
cent stigmas or seed set，depending on if each re—
ceived pollen．In other words，each of the five lobes
with a flower acts independently，where pollinated
lobes do not curve back，while adjacent lobes will
still curve if un—pollinated．Even if the pollen loads
0f 1— 4 stigmas is enough tO set 5 seeds in a single
capsule。un—pollinated stigma lobes still curve．
W hen an un—pollinated stigma lobe continues to
curve until it enters into contact with the anthers。
it takes longer(about 5— 7 h)tO do SO than it takes
the pollen tube of other pollinated stigmas tO reach
ovules(< 3 h，Ruan et a1．，unpublished data)，pro—
viding addition support for the suggestion that in—
dependent curvature is not a response tO the num—
ber of fertilized ovules．This mechanism iS not con—
sistent with from reported bi—-lobed stigma behav—
ior：reopening of some stigmas after pollination in
M imulus aurantiacus appears tO be fl response to
lOW seed set(fewer than one—third of the ovules are
fertilized)rather than to JOW pollen load(Fetscher
3l2 广 西 植 物 3O卷
＆ Kohn，1999)，and the pollinated stigma seconda—
ry open in Campsis radicans may be due to the
limited pollen grains deposited on the stigmas
(Yang et a1．，2004)．Style curvature in H．trionum
var．vesicarius is prevented when 50 or more grains
are deposited per stigma(Seed et a1．，2006)．Some
reopen stigmas in C．radicans close permanently if
a sufficient amount of pollen(> 350 grains)is re—
ceived(Yang et a1．，2004)．In contrast，in K．vir一
壬 A
壬 壬 壬 壬 士
壬
6：00 8：0O 10：00 1 2：00 14：00 16：00 18：00
Time of day
ginica if the curvature of un—pollinated lobe ceases
only responds to if stigma is pollinated but not to
pollen grains number deposited on the stigmatic
surface．Even one pollen grain is sufficient to halt
curvature if it germinated．This is consistent with
the results of(Fetscher& Kohn，1999)，upon re—
ceipt of pollen，most stigmata in M．aurantiacus re—
main closed for the remaining lifetime of the flow—
er．even if less pollen is received than is needed for
6：O0 8：OO 10：O0 12：O0 14：O0 16：O0 18：O0
Time of day
Fig．2 Stigmatic receptivity(A)and polen viability(B)in different times
on the day of anthesis in Kosteletzkya virginica
±ull seed set．
Independent regulation of the curvature of un—
pollinated lobes within K．virginica flowers，com—
bining with other floral traits，harmoniously exhib—
ited a mixed mating system within flowers． This
could be supported by the following three condi—
tions：(1)high stigma receptivity and pollen viabili—
ty when stigmas are in contacting with the anther
by the curvature；(2)floral traits adapting for out—
crossing and selfing resulting from un—pollinated
lobes curvature；(3)delayed self—pollination via cur—
vature of un—pollinated stigma lobes desponding on
pollinator environments．
First，stigma lobes，if polen is not received，
curve to ensure contact with anthers at the end of
the day(about l4：0O一 16：00)(Ruan et a Z．，
2005a)，providing a mechanism for delayed selfing
in K．virginica．This is because at that time stig—
ma receptivity and pollen viability were over 82
and 56 ，respectively(Fig．3)．If pollen is proxi—
mal to the stigma and the stigma is receptive when
the pollen is viable，autonomous selfing maY occur
frequently(L1oyd＆ Schoen，1992)．Flowers of K．
virginica fully open at about 05：O0，un—pollinated
stigma lobes begin to curve at about 7：O0．Corolla
close before l 6：00，at that time delayed autono—
mous selfing via un—pollinated stigma curvature oc—
curs．M ost pollinators of K．virginica(observation
by CJ Ruan)act at 7：00— 16：00，at this times the
flowers have good stigma receptivity and pollen vi—
ability(Fig．2)．
Second，K．virginica exhibits floral traits a—
dapting to outcrossing and selfing． On the one
hand，flowers of K．virginica display some charac—
ters of outcrossed species
(5．27_+0．70 cm)，showy
，such as large corolla size
color(common pink，pur—
ple and white)attractive to pollinators，nectar and
pollen rewards．Pollen-ovule ratio provides a use—
ful insight into the breeding system of species(Cru—
den，1977；Jfirgens et a1．，2002；Jacquemart，2003)．
There is a substantial decrease in pollen grain hum—
bers and in pollen—ovule ratio from xenogamy to
autogamy(Cruden，1977)． The value of pollen—O—
yule ratio indicated breeding system of K．virginica
∞ ∞ ∞ ∞ ∞∞ 加伯 0
^一 一一一 叮一，、c0一一。
∞ ∞ ∞ 加 ∞ ∞ ∞ ∞ ∞ m 0
一，6一 l^ Q∞ooL西E ∞
3期 阮成江等：海滨锦葵花内五柱头裂片弯曲的独自调节 313
belongs to facultative xenogamy． On the other
hand，K．virginica flowers appear special traits
that change as the flower matures and may be asso—
ciated with selfing． W hen pollinators are scarce or
absent，stigma lobes curve to overcome herkogamy
and achieve self—pollinate． If the opportunity of
outcrossing has passed，styiar movements as a
means to effect selfing has also been reported in
some plants，as in many Campanula species(Faegri
& van der Pijl，1979)，H．1aevis(Klips& Snow，
1997)，Viola pubescens(Culley，2002)and many or—
chids etc．
Finally，pollination modes in K．virginica in—
eluding insect pollination and context—dependent
autonomous self—pollination can occur in different
stigma lobes within a flower，providing a mixed
mating system within flowers and optimal mating
system plasticity． Pollinators’activity of K．vir—
ginica directly responds to pollination conditions。
There are more kinds and amounts of pollinators
and higher pollination frequency in sunny days
than in rainy f cloudy days，especially in late flow—
ering lower pollinator abundance and／or activity
under the low temperature conditions display a var—
iable pollinator environment(Ruan et a1．，2005 b)．
Percentage of stigma lobes that curve to contact
the anthers on sunny days was significant lower
than on cloudy／rainy days(Ruan＆ Jin，2007)．
The potential value of curvature of un 。pollina 。
ted stigma lobes within flowers as a mechanism for
reproductive assurance depends highly on the fit—
ness of selfed relative to outcrossed seeds．Though
inbreeding depression of K．virginica was 0．54
(Shan Pt nZ．，2007)，delayed autonomous selfing in
K．tJirginica directly responds to the abiotic envi—
ronment，as the percentage of flowers displaying
delayed selfing was significantly lower on sunny
davs than on inclement days(Ruan et a1．，2005b，
Ruan＆ J in，2007)．This indicates that delayed an—
tonomous selfing in K．virginica is adaptive for
promoting reproductive success under unpredicta—
ble Dollinator environments，because emasculated
flowers(there is no seed set from delayed selfing
via the curvature of un—pollinated stigma lobes)set
fewer seeds than intact flowers(there is seed set
from delayed selfing via stigma curvature)when
open—pollinated(Ruan et a Z．，2008)and there are no
Do11en limitation．Fisher(1 94 1)showed that selling
is advantageous from the“automatic selection hy—
p0thesis”；an allele for selling will spread if selfed
progeny are at least half as fit as outcrossed proge—
ny(Lloyd，1979；Nagylaki，1976)．In contrast，Dar—
win(1 87 6)proposed that an adaptive benefit of self-
pollination is to provide reproductive assurance
when the opportunity of outcrossing has passed
(Baker，1955；Lloyd，1992；Kalisz et a1．，2004)．In
addition，self-pollination under variable pollinator
environments can be advantageous(Kalisz Vo—
gler，2003)despite strong inbreeding depression
(0．64 for H．trionum)(Goodwillie et a1．，2005，
Seed et a1．，2006)．
Independent regulation of curvature of un-pol—
linated stigma lobes within flowers of K．virginica
ensures that if outcross—pollination was insufficient
to fertilize all ovules and the opportunity of out—
cross_po1lination has passed，delayed autonomous
selfing from this curvature would augment overall
seed set．This provides incentive for testing the re—
productive assurance hypothesis at the flower level
in future studies．Reproductive assurance is one of
the most longstanding and widely accepted expla—
nations for the evolution of selfing． This theory
predicts that self—pollination is advantageous in en—
vironments or conditions where mates or pollina—
tors are rare．Cases of selfing leading to reproduc—
tive assurance have been reported in several species
(Donnelly et a1．，1998；Nagy et a1．，1999；Ander—
son et n￡．，2003；Elte＆ Carney，2003)．More corn—
Drehensive analyses were conducted by Kalisz et
口f．(2004)and Herlihy& Eckert(2002)to test
pop．ulation and time effects，and alternatives to the
reproductive assurance hypothesis，such as pollen
discounting(Holsinger，1996)，seed discounting
(Lloyd，1 992)，selfing rate，and inbreeding depres—
sion．Kalisz et a1．(2004)showed that delayed sel—
fing of Collinsia r Ⅱ，responding to unpredictable
314 广 西 植 物 3O卷
polinator environments，supports reproductive as—
surance．Herlihy and Eckert(2002)indicated that
auto-。pollination in Aquilegia canadensis increases
seed production；however，this benefit was out-
weighed by the loss of high quality seed as a result
of seed discounting and inbreeding depression,．The
above cases were conducted at the level of popula—
tions，but this has not yet been addressed at the
level of individuMs or flowers(Zhang，2004)．
References：
Anderson B．Midgley JJ，Stewart B八 2003．Fadlitated selfing of-
fers reproductive assurance：a mutulism between a hermipteran
and carnivorous plantU]．AmJ Bot，90：l 009一l 015
Baker HG 1 955．Self-compatibility and establishment after‘long-
distance’dispersal[J~．Evolution，9：347—348
Blanchard OJ．1976．A revision of spedes segregated from Hibis
CUS sect．Trionum(Ⅳ[edicus)de Candole sensu lato(Malvaceae)
[M]．Doctor Dissertation，USA：Cornel University．
Butrose M ，Grant WjR，Lot JNIA．1977．Reversible curvature
of style branches of Hibiscus trionum L．，a polination meeha．
nism[J]．Aust J Bet，25：567～570
Bynum MR。Smith WK．2001．Horal movements in response to
thunderstorms improve reproductive effort in the alpine species
Gentianaatgid．(Gentianaceae)[J]．AmJ Bet，88：1 O88—1 095
Cruden RW．1977．Polen-ovule ratios：a conservative indicator of
breeding systems in flowering plantsFJ]．Evolution，31：32—46
Culey TM．2002．Reproductive biology and delayed selfing in
Viola pubescens(Vi]daceae)，an understory herb with chas-
mogamous and deistogamous flowers[J]．Int J Plant Sci，
163：113— 122
Darwin C 1876．TIhe effects of CROSS and self-fertilizafion in vege-．
table Kingdom Ch．9[M]．London：John Murray
Dole JA．1990．Role of corola abscission in delayed self-polhna-
tion of Mimulus gutatus(Scrophula riaceae)[J]．Am J Bot，
77：1 505— 1 507
Donnely SE，L0rtie CJ，Aarssen UW．1998．PoIfinafion j[1 Verbas．-
CUTI~th,zpsus(Scrophulariaceae)：the advantage of being tall,J3．
Am J Bet ，85：1 618—1 625
Eckert CG，Schaefer A．1998．Does stir-pollination provide repro—
ductive assurance in A掣 ￡ canadensis(Ranunculaceae)?
[J]．Am J Bet，85：919—924
Edwards J，Whitaker D，Klionsky S，et a1．2005．A record-break—
ing polen catapultU~．Nature，435：1 64
Hle E，Carney R．2003．Reproductive assurance varies with flower
size in Collinsia parviflora(Scrophulariaceae)[J]．Am J Bot，
90：888— 896
Etcheverry AV。Protomastro J，Westerkamp C．2003．De~ ed au-
tonomous self-polination in the colonizer Crotalaria micans
(Fabaeeae：Papilionoideae)：structural and functional aspeetsU]．
PlantSyst Evol，239：1-28
Faegri K，van der Pij lL．1 979．The principles of polination ecolo-·
gY[M]．Pergamon Press，Oxford
Fetseher AE．Kohn JR．1999．Stigma behavior in ：mulus auran-
tiacus(Scrophulariaceae)l,J3．Am J Bet，86：1 13O一1 135
Hsher R八 1841．Average excess and average efect Of a gene sub—
stitutionE．I]．Ann Eug，I1：53—63
Gallagher JL．1985．Halop],by,tic crops for cultivation at seawater
sannitXJ]．Plant Soil，89：323—336
Goodwilie C．Kalisz S．Eckert CG．2005．The evolutionary e—
nigma of mixed mating systems in plants：occurrence．theo—
retical explanations，and empirical evidence[J]．Ann Rev
Ecol Syst，36：47— 79
Herlihy CR。Eckert CG．2002．Genetic cost of repmduetive assur—
ance in a self-fertihzing plant[J]．Nature，416：320-323
Holsinger KE．1 996．Pol~nation biology and the evolution of mat—
ing systems in flowering plants[J]．Evol Biol，29：107—149
Huang SQ。Takahashi Y，Dafni A． 2002． Why does the flower
stalk of PUlsatila ceFtlua(R．anunculaceae)bend during anthesis?
[J]．Am J Bot，89：1 599—1 603
Huang ZH，Zhu JM ，Mu XJ，“a1．2004。Polen dispersion，polen
viability and pistil receptivity in Leymus chinensis[J]．Ann Bet，
93：295— 301
Jaequemart A L．2003．Horal traits of Belgian Ericaceae species：
are they good indicators to assess the breeding systems?l,J 3．
Belg J Bet ，136：154—164
Jtirgens A，Witt T，Gottsberger G．2002．Pollen grain numbers．o-
yule numbers and polen-ovule ratios in Caryophyloideae：corre—
lation with breeding system，po[finafion，life f0ITn，style number,，
and sexual system[J]．Sex Plant Reprod，14：279—289
Kalisz S，Vogler DW，Hanley KM．2004．Context-dependent au-
tonomous seIf-fertilization yields reproductive assurance and
mixed mating[J]．Nature，430：884-887
Kalisz S．Vog|er Dw．200&．Benefits of autonomous selfing under un-
predictable polinator environments[J]．Ecology，84：2 928—2 942
K~alisz S，Vogler D，Fails B．eta1．1999．The mechanism of delayed
salting in Collinsia mⅡ(Scrophulariaeeae)[J]．Am J Boe，86：
1 239— 1 247
Klips RA，Snow AA．1997．Delayed autonomous self-polination in
Hibiscus laevis(Malvaceae)FJ-]．Am J Bet，84：48—53
Li ，Xu ZF，Kress WJ，et a1．2001．Hexihie~tyle that encoura～
ges outcrossing[J]．Nature，410：432
Hu KW ，Hu zj，Huang D ，et a1．2006．SeIf-fertilization strategy
in an orchid[J]．Nature，441：945
Lloyd DG Schoen DJ．1992．Self-and cross—fertilization in
plants．I．Functional dimensions[J]．1nt J Plant Sci，153：
358——369
Lloyd DG．1992．Sdf-and CROSS-ferfihzation in plants．I．The
selection of self-fertilization[J]．Int J Plant S(i。153：37O一380
Lloyd 13(3．1979．Some reproductive factors affecting the selection
of sd／-fertilization in plants[J]．Am Nat，113：67—79
Nagy ES，Strong L，Galloway LF。1999．Co ntribution of delayed
autonomous string to reprod~luctive Success in Mountain Laure1．
Kalmialatifolia(Ericaceae)[J]．AmMidl Nat，142：39—46
Nagylaki T．1976．A modeI for the evolution of self-fertilization
and vegetative reproduction[J]．J Theor Biol，58：55-58
Peter CI，johnson SD．2006．Do ing the twist：a test of Darwins
3期 阮成江等 ：海滨锦葵花 内五柱头裂片弯曲的独 自调节 3l5
cross-pollination hypothesis for polinium reconfiguration[J]
BiolLett，2：65— 68
Pu GZ，PanYM，Lin CR，eta1．2008．Comparison on floral dynam-
ic，pollen viability and stigma receptivity between Chirita gui—
linensis and Chirita baishouensis[J]．Guihaia，28(3)：32O一323
Rathcke BJ．2003．Floral longevity and reproductive assurance：
seasonal patters and an experimental test with Kalmia lati／blia
(Ericaceae)[J]．Am J Bot，90：l 328—1 332
Rathcke B，Real L．1993．Autogamy and inbreeding depression in
mountain laurel，Kalmia latifolia(Ericaceae)EJ]．Am J Bot，
80：143— 146
RUaD CJ。Jin H．2007．Efects of climate conditions on delayed
self-polination in Kosteletzkya virginica[J]．Acta[col Sin，
27：2 259— 2 264
Ruan CJ，Qin P，Han RM．2005a．Strategies of delayed self-pol-
lination in Kosteletzkya virginica[J]．Chinese Sci Bull，50：
94— 96
Ruan CJ，Qin P，He ZX．2004．Delayed autonomous selling in
Kosteletzkya virginica(Malvaceae)[J]．S Afr J Bot，70：
640— 645
Ruan CJ．Qin P．Xi YG．2005b．Floral traits and polination
modes in Kosteletzkya virginica(Malvaceae)[J]．Belg J
Bot，138：39— 46
Ruan CJ，Zhou LJ，Zeng FY，el a1．2008．Contribution of delayed
autonomous selfing to reproductive success in Kosteletzkya vir‘
ginica[J]．Belg J Bot．141：3一l13
Seed L．Vaughton G，Ramsey M．2006．Delayed autonomous
selling and inbreeding depression in the Australian annual
Hibiscus trionum var．vesicarius(Malvaeeae)[J]．Aust J
Bot，54：27— 34
Shan Y，Zhang Lj，Zheng X，et a1．2007．Breeding system and in—
breeding depression in halophyte Kosteletzkya virginica[J]．J
Na ing U 而：N“t Sc ，43：284—289
Stephens we 1948．Kansas wild flowers[M]．Lawrence：Univer-
sity of Kansas Press，KS．
Sun SG，Guo YH ，GitUl~l RW ，et a1．2005．Co rola wilting facili—
tates delayed autonomous self-pollination in Pedicularis dunni—
ana(Orobanchaceae)I-J3．Plant Syst Ew ，251：229—237
Verma S，Magotra R，Koul AK．2004．Stylar movement avoids
self-pollination and promotes cross-polination in Er~nurus hi—
malaicus[J]．Cur,Sci，87：872—873
Wang Y，Zhang D，Renner SS，et a1．2004．A new self-polination
mechanism[J]．Nature，431：39—4O
Yang SX，Yang CF，Zhang T，eta1．2004．A mechanism facilitates
pollination due to stigma behavior in Campsis radicans(Bignoni—
aceae)[J]．J Integr Plant Biol，46：1 071—1 074
Zhang DY．2004．Plant Life History Evolution and Reproductive
Ecology[M]．Beijing：Science Press
(上接第 361页 Continue from page 361)
R6cas G，Barros CF，Scarano FR．1997．Leaf anatomy plasticity of
Alchornea triplinervia(Euphorbiaceae)under distinct light re—
gimes in a Brazilian montane Atlantic rain forest[J]．Trees，
(11)：469—473
Sift GR(史刚荣)．2005．Study on developmental plasticity of leaf
blades structure of Hibiscus syriacus(木槿叶片结构的发育可
塑性研究)[J]．Guihaia(广西植物)，25(1)：48—52
Strauss—Debenedeti S。Berlyn GP． 1 994．Leaf anatomical respon—
ses to light in five tropical Moraceae of different successional sta-
tus[J]．American J Bot，81：1 582—1 591
Su ZM(苏宗明)，Mo LX(莫新 礼)．1988．Geographic distribu-
tioin of Camellia section Chrysantha from China(我国金花茶组
植物的地理分布)[J]．Guihaia(广西植物)，8(1)：75—81
Tang JM(汤景明)，Zhai MP(翟 明普)，Cui HX(崔鸿侠)，et a1．
2008． Morphological responses and adaptation of seedlings of
three tree species of Fagaceae family to diferent light environ—
ments(壳斗科三树种幼苗对不同光环境 的形态响应 与适应)
[J]．Sci Silv Sin(林业科学)，44(9)：41-47
Weatoby M，Falster DS，Moles AT，et a1．2002．Plant ecological
strategies：some leading dimensions of variation between species
[J]．AnnualReview[col Systematics，33：125—159
Wei XL(韦 小 丽)．2003．Ecological adaption of Cinnamomum
camphora and Cinnamomum bodinieri seedlings in different light
environment(不同光环境下香樟猴樟苗木的生态适应)[J]．J
Mountain Agric Biol(山地农业生物学报)，22(3)：2O8—213
Wei x(韦霄)，Jiang YS(蒋运生)，Tang H(唐辉)，et a1．2008．
Phytocoenologieal feature of the rare and andangered plant Ca—
mellia nitidissima(珍稀濒危植物金花茶 的群落学特征)[J]．
Guihaia(广西植物)，28(2)：183一l9O
Wei x(韦霄)，Jiang YS(蒋运生)，Wei JQ(韦记青)，et a1．2007．
Investigation on the geographical distribution and habitat of Ca —
mellia nitidissima(珍稀濒危植物金花茶地理分布与生境调查
研究)[J]．Ecol Environ(生态环境)，16(3)：895-899
Weijschede J，Martinkova J，Kroom H，eta1．2006．Shade avoidance
in Tri[blium repens：costs and benefits of plasticity in petiole
length and leafsize[J]．NewPhytologist，172：655-666
Wu QZ(吴庆锥)．2005．Study of Mytilaria laosensis plantation
biomass(米老排人工林生物量研究)[J]．J Fujian Fore Sci
f (福建林业科技)，32(3)：125—129
Xiao CW(肖春旺)，Liu YC(刘玉成)．1999．Ecological adaptation
of Gordonia“(uminata seedlings in different light environments
(不同光环境的四川太头茶幼苗的生态适应)[J]．Acta[col
Sin(生态学报)，19(3)：422—426
Xu CY(徐程扬)．2001．Response of structural plasticity of Tilia
arauremis sapling crowns to diferent light condition(不同光环
境下紫椴幼树树冠结构的可塑性响应)[J]．Chin J Appl
＆D (应用生态学报 )，12(3)：339—343
Zhao YH(赵永华)，LEI RD(雷瑞德 )，JIA X(贾夏)，eta1．2003．
Quantitative analysis on sharp-toth oak stands in Qinling
Mountain(秦岭锐齿栎群落数量特征的研究)[J]．Chin J Ap—
pl[col(应用生态学报)，14(12)：2 123～2 128