全 文 ：Pattern and Significance of Seedling Development in
Titanotrichum oldhamii (Gesneriaceae)
WANG Yin_Zheng1 , LI Zhen_Yu1 , PAN Kai_Yu1 , ZOU Xiu_Hong2
(1.Laboratory of Systematic and Evolutionary Botany and Herbarium , Institute of Botany , The Chinese Academy of Sciences , Beijing 100093 , China;
2.Forestry Bureau of Yongchun County , Fujian Province 362600 , China)
Abstract:　The seedling development of Titanotrichum oldhamii(Hemsl.)Solereder has been investigated to
provide insight into the evolution and systematic position of Titanotrichum .In T.oldhamii , the size differen-
tiation of the two cotyledons begins while the hypocotyl grows.However , both of the two cotyledons develop
normally and locate at the same level.Finally , the two cotyledons are almost equal in size.The aerial shoot(including stem and leaves)is produced from the permanent activity of the apical meristem in the plumular
bud.Even though the seedling development in Titanotrichum basically conforms to the general growth pattern
of the seedling in the Cyrtandroideae , it is remarkably different from that of other Cyrtandroideae.Based on the
revealed evidence in seedling development in Titanotrichum and other comparative data , the authors have eval-
uated the possible evolutionary pathway of Titanotrichum and further discussed the familial placement of this
genus.
Key words:　seedling development;cotyledon;functional evolution;familial placement;Titanotrichum;
Gesneriaceae
　　The species Titanotrichum oldhamii (Hemsl.)Sol-
ereder is distributed from Fujian and Taiwan of China to
southern Japan.It was first described by Forbes and
Hemsley[ 1] as a member of the Scrophulariaceae with the
name Rehmannia oldhamii Hemsl.Meanwhile , Forbes
and Hemsley suggested that the unilocular ovary of this
species pointed to Gesneriaceae , which was the reason
why he noted with a question mark after Rehmannia
[ 1 , 2] .
Solereder[ 3] lifted this species to a generic rank , Titan-
otrichum Solereder , as a distinct and monospecific genus
transferred to the subfamily Cyrtandroideae in the Gesneri-
aceae.But , Sealy[ 4] suggested that Titanotrichum was
closer to another subfamily Gesnerioideae than to the Cyr-
tandroideae.Kvist and Pedersen[ 5] still considered Titan-
otrichum closer to Scrophulariaceae than to Gesneriaceae.
However , Wang et al[ 6] raised Titanotrichum to the sta-
tus of a monogeneric tribe Titanotricheae in the Cyrtan-
droideae (Gesneriaceae), also recognized by Burtt and
Wiehler
[ 7] .The position of Titanotrichum as the sister to
the remainder of the Cyrtandroideae was weakly supported
by the ndhF sequence data[ 8 , 9] .Nevertheless , the
cladistic analysis of ndhF sequence data also showed that
placing Titanotrichum in the Scrophulariaceae and the
subfamily Gesnerioideae (Gesneriaceae)only required 39
-50 and 24 additional steps , respectively[ 9] .The
molecular systematic studies indicated that the exact posi-
tion of Titanotrichum was still somewhat in doubt[ 9] .
One morphogenetic pattern was discovered in the
subfamily Cyrtandroideae , which was the unequal growth
of the cotyledons during seedling development
[ 10-12] .Al-
most all species in the Cyrtandroideae that have been cul-
tivated have had cotyledons which become unequal in size
soon after germination[ 13-16] .This seedling developmen-
tal pattern serves to differentiate the subfamily Cyrtan-
droideae in the Gesneriaceae.Therefore , Burtt[ 13] pro-
posed a system of the Gesneriaceae , in which two subfam-
ilies Cyrtandroideae and Gesnerioideae were delimited
mainly by the situation of the cotyledons.He treated Ti-
tanotrichum as an anomalous group in his system due to no
knowledge of the seedling structure of this plant
[ 13] .This
system has been widely accepted by other re-
searchers[ 6 , 16 ,17] .However , because its distribution is
very limited and flowers are often replaced by clusters of
bulbuls serving as vegetative propagules in the upper part
of inflorescences , the seeds of T .oldhamii are difficult
to collect in the field.Therefore , there has been no infor-
mation of the seedling development and structure in Ti-
tanotrichum up to date.Two local populations with harvest
of seeds recently found in Fujian Province allow us to con-
duct a detailed study on the seedling development , which
is necessary for the understanding of Titanotrichum and
its familial placement.
The goal of this study is first to explore the develop-
mental pattern and structure of the seedling in Titan-
otrichum.Secondly , comparing the developmental features
of the seedling in Titanotrichum with those found in relat-
ed groups , we also evaluate the possible evolutionary
pathway of Titanotrichum .Finally , the familial placement
of Titanotrichum is discussed , based on the revealed evi-
dence in seedling development together with comparative
data.
1　Materials and Methods
The seeds of Titanotrichum oldhamii (Hemsl.)
Received:2002-04-08　Accepted:2002-06-30
Supported by the National Natural Science Foundation of China(39870056).
植　物　学　报　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　
Acta Botanica Sinica　2002 , 44(8):903-907
904　 植物学报　Acta Botanica Sinica　Vol.44　No.8　2002
Solereder were collected from two natural populations in
Fujian Province.Voucher specimens (LI Zhen_Yu and
ZOU Xiu_Hong , 11327 , 11328)are deposited in the
Herbarium (PE), Institute of Botany , the Chinese A-
cademy of Sciences.Seeds for germination were spread on
the surface of the filter paper saturated with sterile water
in the petri dish.Seedlings at different developmental
stages were fixed in formalin_acetic acid_alcohol (FAA)
and then dehydrated in a series of ethanol(75%, 85%
and 95%).Materials were checked in 95% ethanol and
further dehydrated in three changes of 100% ethanol ,
100% ethanol_amylacetate (1∶1)and amylacetate , and
then critical point dried with a Hitachi HCP_2 Critical
Point Dryer and Au/Pd sputter coated with a SPI_MOD-
ULE Sputter Coater.The prepared samples were examined
with a Hitachi S_800 scanning electron microscope(SEM)
and recorded on Shanghai Panchromatic film.
2　Results
In a seed before germination , the two cotyledons are
equal in size.Seeds germinated about one week after sow-
ing.At first , the growing hypocotyl pushed the radicle to
burst the seed coat , and several fibrous roots were pro-
duced from the surface of the radicle (Fig.1).After-
ward , the elongating hypocotyl got out of the seed coat
and only the two cotyledons remained in the coat (Fig.
2).As the two cotyledons enlarged , the young seedling
gradually escaped from the seed coat (Figs.3 -5).
Meanwhile , the main taproot gradually developed and
continued to grow downward(Figs.3 ,4).However , the
initiation of both fibrous roots and taproot was delayed
sometimes during seedling development (Fig.5).The
seed coat was usually carried up attached to the tip of the
two cotyledons or to one of the cotyledons which were
greenish in color (Fig.5).While the growing hypocotyl
got out of the seed coat , the primary growth of the two
cotyledons showed an initial differentiation in size (Fig.
6).One cotyledon was slightly larger than the other , in
which the larger one was 250-260μm long and the small
one 245 -255 μm long (Fig.6).While the young
seedling escaped from the seed coat , two cotyledons be-
came more rapid in wide growth(Figs.7 ,8).Meanwhile ,
The different sizes of the two cotyledons became slightly
more manifested , in which the larger one was about 320μm long and the small one about 260 μm long (Fig.8).
Accompanying the enlargement both in width and length ,
two cotyledons opened up and gradually spread out flat(Figs.9 ,11 , 12).When the two cotyledons became ex-
planate , the large cotyledon was about 350 μm while the
small one was about 290 μm in length(Fig.13).During
the development of the two cotyledons , a plumular bud
became visible between two cotyledons(Figs.10-13).
The true leaves were initiated in pair(Fig.14).The first
opposite pair of the true leaves was alternate to the two
cotyledons(Fig.16).The second pair of the true leaves
arose at the position alternate to the first pair of the true
leaves(Fig.17).At this stage , the two cotyledons were
still greenish in color and almost equal in size (Figs.16 ,
17).While sequentially detaching the true leaves , the
apical meristem was continuously active for apical growth ,
which formed a simple stem(Figs.16 ,17).In an excep-
tional case , there were three cotyledons developed in the
seedling in which one was remarkably larger than the oth-
er two (Fig.15).
3　Discussion
3.1　Comparison between Titanotrichum and others
in the Cyrtandroideae
Since Crocker[ 10] first described the unequal growth
of the cotyledons in Streptocarpus , more and more species
have been recognized with more or less unequal develop-
ment of the cotyledons in the subfamily Cyrtan-
droideae[ 11-16] .In an extreme case , even though the ini-
tial seedling possesses two equal cotyledons , one of the
cotyledons increases considerably in size after a few days ,
and continues to grow until the plant has reached the ma-
ture size , while the other remains stationary and eventual-
ly dies off[ 10 ,11 ,14 ,15] .No other leaf is produced in the life
of the plant.This growth pattern of seedling has been
→
Figs.1-15.　SEM micrographs of the seedling development.1-5.Lateral view , showing the process of the seedling bursting the seed coat.
1.Radicle(arrow)just bursting the seed coat with several fibrous roots.Scale bar=209 μm.2.Showing that the elongating hypocotyl(ar-
row)has got out of the seed coat and only two cotyledons remain in the coat.Scale bar=165μm.3 , 4.Showing that two cotyledons gradually
escape from the seed coat as enlarged in size , meanwhile the main taproot(arrow)gradually develops and continues to grow downward.Scale
bar=209 μm.5.Showing the seed coat being carried up attached to the tip of the two cotyledons and the case that both fibrous roots and tap-
root are delayed in growth during seedling development.Scale bar =149 μm.6.Similar stage to Fig.2 with the seed coat removed , showing
that the primary growth of the two cotyledons demonstrates an initial differentiation in size in which one cotyledon is barely larger than the other.
Scale bar=95μm.7 , 8.Similar stages to Fig.3 and Fig.4 respectively with the seed coat removed , showing that the different sizes of the two
cotyledons become slightly more manifested while they become more rapid in wide growth.Scale bar=120 μm in Fig.7 and 195 μm in Fig.8.
9.Oblique_lateral view , showing two cotyledons being opening up accompanying their enlargement both in width and length.Scale bar=127μm.10.Similar stage to Fig.9 , showing the initial plumular bud.Scale bar=209μm.11 , 12.Oblique_polar view of young seedlings , show-
ing that two cotyledons gradually spread out flat with primary development of the plumular bud(arrow).Scale bar=127μm in Fig.11 and 83
μm in Fig.12.13.Polar view of two explanated cotyledons with a plumular bud(arrow)between them , note almost equal size of two cotyle-
dons.Scale bar=110μm.14.A close_up of the micrograph of Fig.13 , showing the initial true leaves(arrow).Scale bar=41 μm.15.
Oblique_lateral view , showing three cotyledons as an exceptional case developed in the seedling in which one is remarkably larger than the other
two.Scale bar=138 μm.
Abbreviations:C , cotyledon;Hy , hypocotyl;LC , large cotyledon;P , plumular bud;SC , small cotyledon;Te , seed coat;Tr , main taproot.
WANG Yin_Zheng et al:Pattern and Significance of Seedling Development in Titanotrichum oldhamii(Gesneriaceae) 905　
Figs.16 , 17.　Photograph of young seedlings , note that the two
cotyledons are still greenish in color and almost equal in size.16.A
young seedling with the first opposite pair of the true leaves alternate
to the two cotyledons.Scale bar=1 612μm.17.A young seedling
with the second pair of the true leaves becoming visible at the posi-
tion alternate to the first pair of the true leaves.Scale bar=3 891μm.
C , cotyledon;Hy , hypocotyl;Tf , first pair of the true leaves;Tr ,
main taproot;Ts , second pair of the true leaves.
found in several species with unifoliate habit , in which
there is only one enlarged and foliaceous cotyledon in the
plant life.These species respectively belong to two tribes
and several genera , such as Streptocarpus and Monophyl-
laea [ 15 , 18] .In these species , a typical plumule or apical
meristem is lacking[ 15 ,18 ,19] .However , it is more frequent
that the two cotyledons differentiate remarkably in size af-
ter germination due to accrescence of one cotyledon in the
pair as a result of continued activity of basal intercalary
anisocotyly meristems[ 15 ,19] .In these species , the apical
meristem is still maintained even though delayed in activi-
ty , which develops a rosette or caulescent plants[ 19] .This
unequal growth pattern of seedling is widely distributed in
the Cyrtandroideae.
In T .oldhamii , the size differentiation of the two
cotyledons begins while the hypocotyl grows.The size dif-
ference in the two cotyledons is manifested during the
seedling development.However , both of the two cotyle-
dons develop normally and locate at the same level.The
two cotyledons are almost equal finally in size.The aerial
shoot(including stem and leaves)is produced from the
permanent activity of the apical meristem.Apparently ,
even though the seedling development in Titanotrichum
basically conforms to the unequal growth pattern of the
seedling in the Cyrtandroideae , it is remarkably different
from that of other Cyrtandroideae.
3.2　Phylogenetic consideration
The subfamily Cyrtandroideae is characteristic of un-
equal growth of the cotyledons accompanied with the de-
layed formation of a plumule.Burtt[ 20] suggested that
small and numerous seeds and an accrescent cotyledon
were interlocked during functional evolution in the Cyrtan-
droideae.As compensating for the lack of food reserves in
the seed , the continued growth of one cotyledon may act
as an additional photosynthetic tissue before energy is re-
quired to mobilize a plumular growing point and new
leaves
[ 20] .The delayed development of the shoot apical
meristem and the continued growth of one cotyledon re-
quire an uncoupling of the strict developmental linkage
between meristematic activity and the shoot apex[ 21] .This
uncoupling may result from gain_of_function dominant mu-
tations in plant_meristem_determining genes such as stm
(shootmeristemless)[ 21 , 22] .The unifoliate form , as an ex-
treme case , would be derived from indefinite delay of the
activity of the shoot apical meristem and extreme develop-
ment of the meristematic cotyledon[ 21] .The interplay of
the reproductive strategy and photosynthetic economy
gains a more advantage for adapting to different environ-
mental conditions.
In Titanotrichum , seeds are averagely 900 μm long
and 200μm wide.The seed size of Titanotrichum is larg-
er than the average size of seeds in the Cyrtandroideae.It
seems that Titanotrichum is not strongly involved in the
functional evolution toward remarkable accrescence of one
cotyledon and mini_size of seeds.This suggestion is fur-
ther confirmed by the normal development of the plumule
in Titanotrichum , which indicates that there exists no un-
coupling of the developmental linkage between meristem-
atic activity and the shoot apex in Titanotrichum .In ad-
dition , a very curious reproductive strategy is developed
in Titanotrichum oldhamii.In the upper part of the race-
mose inflorescence , clusters of bulbuls serving as vegeta-
tive propagules are often produced in the axils of bract ,
instead of a flower.The bulbuls are very deciduous ,
which gives rise to a new plant.The development of the
bulbuls shows that each cluster of bulbuls is developed
from the same floral primordium as a flower (personal
communication with Dr.WANG Cun_Neng).This curious
vegetative propagation developed in Titanotrichum is par-
allel to the generative propagation with numerous mini_
seeds and continued accrescence of one cotyledon as in
other Cyrtandroideae in the reproductive strategy.There-
fore , it is likely that Titanotrichum , representing a lin-
eage , diverges early in the evolution of this subfamily ,
and does not strongly evolve toward remarkable accres-
cence of one cotyledon , but toward additional vegetative
propagation in the functional evolution.This evolutionary
change may be related to lack of convenient pollinators.
In addition to the seedling development and vegetative
propagation , some morphological features are also exclu-
sive with Titanotrichum in the Cyrtandroideae , such as
the terminal racemes , dimorphic flowers and scaly
stolons.The above facts are coincident with the cladistic
analysis of ndhF sequence data which show an isolated
position of Titanotrichum within the Cyrtandroideae
[ 8 ,9] .
Putting the evidence of seedling development togetherwith
comparative data into consideration , we suggest that Ti-
tanotrichum should be considered as the sister to the re-
mainder of the subfamily Cyrtandroideae within the Gesne-
riaceae.The final judgement should be waited for further
information from more detailed sampling of DNA se-
quences , pollination and reproductive ecology.
906　 植物学报　Acta Botanica Sinica　Vol.44　No.8　2002
Acknowledgements:We are indebted to Mr.XIAO Yin_
Hou , YANG Xue_Jian for their technical assistance for
SEM and photographs.We also say thanks to Prof.
CHEN Wei_Lun for seed germination.
References:
[ 1] 　Forbes F B , Hemsley W B.Enumeration of all the plants
known from China Proper , Formosa , Hainan , the Corea ,
the Luchu Archipelago and the Island of Hongkong , together
with their distribution and synonymy.J Linn Soc Bot ,
1890 , 26:194.[ 2 ] 　Hemsley W B.Descriptions of some new plants from eastern
Asia , chiefly from the Island of Formosa, presented by Dr.
Augustine Henry , F.L.S., to the Herbarium , Royal Gar-
dens , Kew.Ann Bot , 1895 , 9:154-155.[ 3 ] 　SolerederH.Uber die Gattung Rehmannia.Ber Deutsch Bot
Gesellsch , 1909 , 27:390-404.[ 4 ] 　Sealy J R.Titanotrichum oldhamii.Curtiss Bot Mag ,
1949 , (166):tab.78.[ 5 ] 　Kvist L P , Pedersen J A.Distribution and taxonomic impli-
cations of some phenolics in the family Gesneriaceae deter-
mined by EPR spectroscopy.Biochem Syst Ecol , 1986 , 14:
385-405.
[ 6 ] 　Wang W T , Pan K Y , Li Z Y.Key to the Gesneriaceae of
China.Edinb J Bot , 1992 , 49:5-74.[ 7 ] 　Burtt B L , Wiehler H.Classification of the family Gesneri-
aceae.Gesneriana , 1995 , 1:1-4.
[ 8 ] 　Smith J F , Wolfram J C , Brown K D , Carroll C L, Denton
D S.Tribal relationships in the Gesneriaceae:evidence
from DNA sequences of the chloroplast gene ndhF.Ann
Missouri Bot Gard , 1997 , 84:50-66.[ 9 ] 　Smith J F , Wolfram J C , Brown K D , Carroll C L, Denton
D S.Familial placement of Cyrtandromoea , Titanotrichum
and Sanango , three problematic genera of the Lamiales.
Taxon , 1997 , 46:65-74.[ 10] 　Crocker C W.Notes on the germination of certain species of
Cyrtandreae.J Linn Soc Bot , 1860 , 5:65-66.[ 11] 　Hill A W.The monocotylous seedlings of certain dicotyle-
dons with special reference to the Gesneriaceae.Ann Bot ,
1938 , 2:127-144.[ 12] 　Burtt B L.Studies in the Gesneriaceae of the Old WorldⅠ .— General introduction.Notes Roy Bot Gard Edinb ,
1954 , 21:185-218.[ 13] 　Burtt B L.Studies in the Gesneriaceae of the Old WorldⅩⅩⅣ.Tentative keys to the tribes and genera.Notes Roy
Bot Gard Edinb , 1962 , 24:205-220.[ 14] 　Hilliard O M , Burtt B L.Streptocarpus:an African Plant
Study.Pietermaritzburg:Natal University Press , 1971.[ 15] 　Jong K , Burrt B L.The evolution of morphological novelty
exemplified in the growth patterns of some Gesneriaceae.
New Phyt , 1975 , 75:297-311.[ 16] 　Weber A.Beiträge zur Morphologie und Systematik der
Klugieae und Loxonieae (Gesneriaceae):Ⅶ .Sproβ_ , In-
floreszenz_ und Blǜtenbau von Rhynchoglossum.Bot Jahrb
Syst , 1978 , 99:1-47.[ 17] 　Ivanina I I.Application of the carpological method to the
taxonomy of Gesneriaceae.Notes Roy Bot Gard Edinb ,
1965 , 26:383-402.[ 18] 　Rosenblum I M , Basile D V.Hormonal regulation of mor-
phogenesis in Streptocarpus and its relevance to evolutionary
history of the Gesneriaceae.Amer J Bot , 1984 , 71:52-
64.[ 19] 　Jong K.Phyllomorphic organisation in rosulate Streptocar-
pus.Notes Roy Bot Gard Edinb , 1978 , 36:369-396.[ 20] 　Burtt B L.Studies in the Gesneriaceae of the Old WorldⅩⅩⅩⅠ .Some aspects of functional evolution.Notes Roy
Bot Gard Edinb , 1970 , 30:1-10.[ 21] 　Mö ller M , Cronk Q C B.Evolution of morphological novel-
ty:a phylogenetic analysis of growth patterns in Streptocar-
pus(Gesneriaceae).Evolution , 2001 , 55:918-929.
[ 22] 　Long J A , Moan E I , Medford J J , Barton M K.A member
of the knotted class of homeodomain proteins encoded by the
STM gene of Arabidopsis.Nature , 1996 , 379:66-69.
台闽苣苔(苦苣苔科)幼苗的发育式样及其意义
王印政1 　李振宇1 　潘开玉1 　邹秀红2
(1.中国科学院植物研究所系统与进化植物学重点实验室及植物标本馆 , 北京 100093;2.福建省永春县林业局 , 362600)
摘要:　为解决台闽苣苔族(Titanotricheae)这一单种族的科级系统位置 , 通过扫描电镜观察了台闽苣苔(Titan-
otrichum oldhamii(Hemsl.)Solereder)植物的种子发芽和幼苗发育过程。随着下胚轴的向下伸长 , 两个子叶开始不明
显的异率生长 ,其中一片子叶略大于另一片子叶。但两片子叶均正常发育并位于同一高度。当真叶发出后 ,两片
子叶几乎等大 ,并且两个子叶柄等长。在幼苗生长期间 , 随着子叶的生长 ,胚芽也正常萌发出茎的顶芽。顶芽持续
进行顶端生长产生交互对生的真叶。这一幼苗生长式样和苦苣苔亚科其他类群的仅一片子叶发育与胚芽被抑制
的幼苗生长式样有明显区别。考虑到台闽苣苔植物在总状花序的上部大量簇生无性珠芽 , 并落地迅速生长出新的
植株这一在苦苣苔科中独特的无性繁殖方式及相关性状 ,台闽苣苔族可能较早地从苦苣苔亚科中分化出来 ,并在
繁殖体的功能进化方面和其他类群发生歧化进而获得独特的无性繁殖方式。台闽苣苔族在系统发育上应该被认
为是其他苦苣苔亚科类群的姊妹群 ,应当提升为亚科等级。
关键词:　幼苗发育;子叶;功能进化;科级系统位置;台闽苣苔属;苦苣苔科
中图分类号:Q949　　　文献标识码:A　　　文章编号:0577-7496(2002)08-0903-05
收稿日期:2002-04-08　接收日期:2002-06-30
基金项目:国家自然科学基金(39870056)。
(责任编辑:王　葳)
WANG Yin_Zheng et al:Pattern and Significance of Seedling Development in Titanotrichum oldhamii(Gesneriaceae) 907