全 文 ：白车轴草 ( Trif oli um repens)植株抗病性和生长
与植物病史的关系 3
刘登义1 3 3 　王友保1 　Lars Ericson2
(1 安徽师范大学生物多样性研究中心 ,芜湖 241000 ;2Department of Ecology and Environmental Science ,
University of Umea , S2901 87 Umea , Sweden)
【摘要】　从白车轴草 ( Trif olium repens) 自然种群中采集无白车轴草单孢锈菌病史的无性系 (clones) 17
个 ,有白车轴草单孢锈菌病史的无性系 14 个 ,分别作为抗病型和感受型植物实验材料 ;采集白车轴草单孢
锈菌 ( U romyces t rif olii2repentis)菌系 (strains) 10 个 ,作为病菌实验材料. 分别设置并进行了两个温室实验、
一个田间盆栽实验和一个原生长地移栽实验 ,实验处理上分对照、单菌系接种和 10 个菌系接种等 3 种. 实
验结果表明 ,无论是用单菌系接种还是 10 个菌系接种 ,植株发病的概率和程度均与其抗病性有关 ,抗病型
植株 (无病史)发病的概率和程度显著低于感受型 (有病史)植株. 在相同处理的实验中 (无论是田间实验还
是温室实验) ,无病史植株和有病史植株的生长无显著差异 ;不同处理田间实验植株的生长有显著差异 ,病
情愈重 ,生长愈差. 无病史植株的抗病性明显强于有病史植株. 但是 , 原生长地的移栽实验结果表明 ,在无
病原菌存在的情况下 ,有病史植株的 (叶)生长显著好于无病史植株. 可以认为 ,研究生物个体对环境因子
反应性差异的实验应当在自然条件下和自然梯度范围内进行.
关键词 　抗病性 　生长 　白车轴草 　白车轴草单孢锈菌
文章编号 　1001 - 9332 (2002) 01 - 0035 - 08 　中图分类号 　S432 　文献标识码 　A
Relationship between resistance and growth of Trif olium repens plants and their disease history. L IU Dengyi1 ,
WAN G Youbao1 ,Lars Ericson2 (1 Biodiversity Research Center , A nhui Norm al U niversity , W uhu 241000 ,
China;2 Depart ment of Ecology and Envi ronmental Science , U niversity of U mea , S2901 87 U mea , S we2
den) . 2Chin. J . A ppl . Ecol . ,2003 ,14 (1) :35～42.
17 healthy host clones and 14 clones originally diseased with the annual rust , U romyces t rif olii2repentis , were
sampled from the natural population of Trif olium repens , which were used as the experimental plant materials.
10 rust strains of U romyces t rif olii2repentis were collected for the experimental inoculation. Two greenhouse
and two field experiments were done , with three treatments as inoculated with one rust strain , inoculated with
ten rust strains and controls , respectively. The results showed that originally healthy plants were more resistant ,
and originally diseased ones were more susceptible to the rust . However , there was a considerable overlap among
some of the clones. Pot experiments , both in the greenhouse and in the field , revealed no significant differences
in growth between originally healthy and originally diseased plants within the treatments (controls , inoculations
with one strain and with ten strains) . However , there were significant differences in growth between treat2
ments. This lack of difference in growth within treatments contrasted to the result of a transplantation experi2
ment in a natural meadow where the rust was absent . The transplantation experiment showed that originally dis2
eased clones produced significantly more leaves than originally healthy ones (mean + S. E. : + 0. 77 and 17. 6 +
0. 49 , respectively) . Further , originally diseased clones showed more variation in leaf production , compared
with the originally healthy ones. The difference between the pot experiments and the transplatation experiments
tells that to study individual differentiation in responses to environmental factors , such experiments ought to be
performed under a range of natural conditions , where various biotic and abiotic factors are involved.
Key words 　Disease resistance , Growth , Trif olium repens , U romyces t rif olii2repentis .3 Supported by the Swedish Natural Science Research Council.3 3 Corresponding author.
Received 14 January 2002 ,Accepted 2 May 2002.
1 　INTROD UCTION
　The pathogens play an important role in non2agricultural
plant communities[7] . They can influence the recruitment
and the fecundity of host plants[2 ,26 ,30 ,40 ,41 ,46 ,47]. Host
plants exhibit active interactions with their pathogens ,
physiologically or/ and genetically[7] . A number of studies
have shown large variation in disease susceptibility and re2 sistance[6 ,7 ,9 , ,12 ,19 ,52 ,54]. Under various experimental condi2tions it has been frequently found that diseased plants arethose most susceptible[15] . However , standard inoculationexperiments under greenhouse conditions may not tell muchabout whether more susceptible individuals actually face
应 用 生 态 学 报 　2003 年 1 月 　第 14 卷 　第 1 期 　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　
CHIN ESE JOURNAL OF APPL IED ECOLO GY ,Jan. 2003 ,14 (1)∶35～42
more risk of becoming diseased in natural communities[1] .
Many environmental factors affect the interactions be2
tween pathogens and host plants[7 ,20 ,21 ,23～25 ,53 ] . As a
consequence , differences in disease susceptibility will
not necessarily explain whether or not a specific plant
will be come diseased. Indeed , in some cases , envi2
ronmental conditions may be decisive , and highest
disease incidence may occur where the selection to2
wards increased resistance has been strongest [33 ,50 ] .
　In this study , we conducted several comparative
experiments to determine whether the occurrence of
rust on individual T rif oli um repens plants in one
small natural population could be explained by differ2
ences in host susceptibility. The second question we
addressed was whether healthy and diseased plants
differed in their response in growth to the presence
and absence of the rust .
2 　MATERIALS AND METHODS
211 　Studied species
　The host , T rif oli um repens , is a stoloniferous
clonal herb[6 ] . It is both frequent and abundant in
various types of meadows throughout the Nordic
countries. Two rusts were present at the study area ,
which both could infect T . repens plants. The rust ,
U rom yces t rif olii2repentis Liro , is macrocyclic , au2
toecious , and causes annual infections[39 ] . Aecidia are
usually formed on petioles and along nerves , causing
swelling and distortions ; uredinia and telia occur as
small pustules on leaves. Teleutospores germinate
continuously throughout the season , thus aecidia may
be found also late in the season , and uredinia are
sometimes very small and poorly developed.
　T . repens also has a second rust , U rom yces f lectens
Lagerh , which is microcyclic and causes systemic ,
perennial infections. The mycelium is able to over2
winter in the buds[36 ] . The telia are more elongated
and occur mainly on vascular tissue ; that is the usual
location of the aecidia of U . t rif olii2repentis . The
two rusts may occur on the same plant , even on the
same leaflet , which has caused much confusion[36 ] .
　These two rusts are not always treated as different
taxa. As they differ with regard of their life2cycles ,
occurrences and effect on the host , it is important to
keep in mind which of the taxa that is referred to , or
whether they are clumped together ( see further be2
low) .
212 　Plant materials
　In early November 1997 , 31 different T . repens
clones were identified in a heavily rusted lawn , 30 ×
40m2 , situated in Obbola , 15 km south of Umea ,
Sweden (63°42N , 29°19E) . The sampling procedure
was as follow. Clones that were possible to identify
based on leaf characters , were searched for. When
one such clone was found a second clone of the oppo2
site disease status was searched for. This resulted in
18 pairs. However , four of the diseased clones were
later excluded as the disease was too poorly devel2
oped , and one of the healthy due to bad condition.
The clones were numbered with 14 of them being dis2
eased , in the following ,“originally diseased”, with
U rom yces t rif olii2repentis . The remaining 17 clones
were uninfected , in the following ,“originally
healthy”.
　The sampled stolons were immediately placed in
separate plastic bags that were sealed. The plants
were brought to the greenhouse at the university cam2
pus and stored outdoors in the plastic bags. These
plants were subsequently used for all the experiments
reported here. During the experiments , the healthy
clones and the diseased clones were kept separately.
　At the above2mentioned sampling , all diseased
leaves ( telia) were removed and placed in separate
plastic bags , which were sealed. In order to avoid
contamination , the sampling equipment was sterilized
with alcohol and flamed after the sampling of each
clone. This inoculum was used to build up strains on
the original T . repens clones in the greenhouse. The
rust st rains were later used for inoculation experi2
ments.
213 　Greenhouse experiment 1 : Does history of in2
fection influence plant growth in the greenhouse ?
　In late November 1997 , three 5cm2long stolon cut2
tings were taken from each of the 31 T . repens
clones. Each cutting had two disease2f ree leaves and
two axillary buds. Each of the cuttings was planted in
a separate plastic pot (diameter 11 cm , 1 dm3) , filled
with plant compost . The pots were randomly placed
63 应 　用 　生 　态 　学 　报 　　　　　　　　　　　　　　　　　　14 卷
on a table in the greenhouse and kept under constant
conditions (18 h light , 18～25 ℃) . The experiment
was terminated on 13 May 1998 , when we counted
the total number of leaves , floral shoots and stolons
and measured the total stolon length for each pot . No
disease was observed throughout this experiment .
214 　Field experiment 1 : Does history of infection
influence plant growth in an experimental garden ?
　On 27 May 1998 , f rom each of the pots , we took
ten 5 cm long stolon cuttings , each with two leaves
and two axillary buds. Cuttings were separately
planted in plastic pots (as in the above experiment) .
There were a total 30 separate cuttings for each of the
31 clones. The pots were placed in shade (30 % light
reduction) in an experimental garden. The experi2
mental area was shaded for two reasons. First , in or2
der to diminish leaf mortality during prolonged
drought periods. Second , as excessive watering will
result in a too rapid increase of the rust . Further , the
lawn from which the plants originated was partly
shaded by trees. The soil was covered with black
plastic and holes dug up for each pot , prior to plant2
ing. Plants were watered when necessary. A total 30
different blocks were established , each of which con2
tained 17 originally healthy and 14 originally diseased
clones. The blocks were assigned to one of three fol2
lowing treatments (for each n = 10) : control ; inocu2
lation with one strain of U . t rif olii2repentis ; inocula2
tion with a composite sample of ten strains of U . t ri2
f olii2repentis . The 10 strains occurred on the original2
ly diseased clones no 2 , 3 , 4 , 5 , 7 , 10 , 11 , 12 , 13
and 14. Those strains were selected because abundant
inoculum material was present .
　The rust st rains used for inoculations were collected
in early Nov. 1997 ( see‘Plant material’) f rom the
originally diseased plants and since were kept in culti2
vation on their original T . repens clones in the green2
house. The inoculations took place on 14 J uly 1998.
The disease (aecidia) started to develop two weeks
after the inoculation. The responses of the different
host clones to the rust st rain ( s) was estimated ac2
cording to a scale 0～5 , where 0 implies complete re2
sistance and 5 the strongest susceptible reaction. The
total number of leaves per plant was scored on 25 Au2
gust. In October 1998 , ten weeks after inoculation ,
ten leaves per plant were randomly selected and
scored for the number of rust pustules present .
215 　Greenhouse experiment 2 : Does single or multi2
ple inoculation history influence plant growth in the
greenhouse ?
　 This experiment was designed to investigate
whether the previous treatment , single or multiple in2
oculation , had any long2term effects. After the termi2
nation of field experiment 1 , we took one cutting of
T . repens (5 cm in length , with two leaves and two
axillary buds) f rom each of the pots in the previous
field experiment . These cuttings were all disease
f ree. Each cutting was separately planted in a new
pot (diameter 11 cm , 1 dm3 , filled with plant com2
post) . Thirty blocks were established , each including
31 individuals ( 14 originally diseased , 17 originally
healthy) . All pots were placed in the greenhouse (18
～25 ℃; 16 h light , 8 h dark) .
　From 1 November 1998 until 30 January 1999 , we
marked the emerging leaves with different2colored
plastic ribbons every ten days. Leaf mortality was
recorded. At the termination of this experiment , we
randomly chose one healthy and one diseased individ2
ual f rom each block of three treatments for determina2
tion of total biomass (f resh weight ) . Before weigh2
ing , each plant was carefully rinsed with water , dried
between filter paper , and excess water was absorbed.
216 　Field experiment 2 :Does history of infection in2
fluence plant growth under natural field conditions ?
　This experiment was conducted to study whether
the treatments in field experiment 1 had any long2
term effects under natural conditions. In early J une
1998 , a meadow , corresponding to the A nthox an2
thum odorat um type[29 ] , was selected on Obbola Is2
land , 16 km south of Umea , Sweden. Ten cuttings
(5 cm in length , with two leaves and two axillary
buds each) were taken from each clone of field experi2
ment 1. All these cuttings were disease f ree. Prior to
planting , the grass sward was removed with a knife ,
and holes were 5cm diameter , dug for the plants , and
filled with a mixture of mull and coarse sand. Cut2
tings were transplanted during 12～ 14 J une 1998.
Ten experimental stations were established ( 7～8m
apart) . One cutting of each clone , in total 14 origi2
nally diseased and 17 originally healthy clones , was
planted at each station. Survival of the cuttings were
731 期 　　　　　　　　刘登义等 :白车轴草 ( Trif olium repens)植株抗病性和生长与植物病史的关系 　　　　　　　　　　　
checked throughout J une and 14 cuttings were re2
placed. Cuttings that died after 13 J uly were not re2
placed. The number of leaves and inflorescences pre2
sent were scored during 17～23 September 1998.
3 　RESULTS
311 　Effects of disease history on plant growth and
resistance
　An ANOVA revealed no differences between origi2
nally healthy and originally diseased clones for the
number of leaves , while originally diseased clones
produced significantly more inflorescences and fewer
stolons than did originally healthy clones( Table 1 and
2) . Total stolon length was also significantly higher
for the originally healthy clones.
　The results of the inoculation experiment ( Fig. 1)
showed a marked variation both between T . repens
clones with regard to their susceptibility and between
U . t rif olii2repentis st rains with regard to their viru2
lence . In general , the originally healthy clones were
Fig. 1 Variation in susceptibility shown by originally healthy and origi2
nally diseased T . repens clones in an inoculation experiment with ten dif2
ferent strains of U . t rif olii2repentis . The strain number refers to the
diseased T . repens clones on which the strain occurred in the field. Sus2
ceptibility is represented by numbers of pustules which were produced by
different U . t rif olii2repentis st rains .
less susceptible when compared with the originally
diseased ones. However , some of the originally
healthy clones (eg. 5 , 6 , 13 and 14) were quite sus2
ceptible , while some of the originally diseased clones
(eg. 5 , 6 , 8 and 9) exhibited low susceptible reac2
tions. This shows a considerable overlap in suscepti2
bility between the originally healthy and originally
diseased clones.
　There was a significant effect of t reatment for
numbers of leaves produced both for originally healthy
and originally diseased plants ( Table 3 and 4) . For
both plant types , inoculation with either one or ten
strains resulted in a significant decrease in numbers of
leaves ( Table 3 ) . However , a comparison within
each of the three treatments showed no significant
difference between originally healthy and originally
Table 1 Mean number of leaves , inflorescences and stolons as well as
total stolon length for Trif olium repens clones which were originally
healthy ( n = 17) or originally diseased ( n = 14) in the f ield. The clones
have been kept in cultivation for 22 weeks under disease free condi2
tions. Data from greenhouse experiment 1. Mean + S. E. given
Variables Healthy clones Diseased clones
Nos of leaves 177. 9 ±11. 5 183. 7 ±7. 9
Nos of inflorescences 0. 35 ±0. 15 1. 47 ±0. 42
Nos of stolons 23. 2 ±0. 66 18. 0 ±0. 54
Total length of stolons 3469. 6 ±143. 15 2712. 6 ±127. 13
Table 2 ANOVA2analyses for a) number of leaves , b) number of inflo2
rescences , c) number of stolons and d) total length of stolons( cm) for
healthy and diseased T. repens clones. Tests performed on the mean
values for each of the 17 healthy and 14 diseased clones. DF = 1 , 29
( a) , 1 , 29 ( b) , 1 , 29 ( c) and 1 , 29 ( d)
Variables MS F2ratio P
a) Leaves 4268. 6 2. 23 0. 1466
Error 1918. 2
b) Flower 10. 4 6. 53 0. 0161
Error 1. 6
c) Stolons 128. 6 8. 94 0. 0056
Error 14. 4
d) Stolon length 38948. 70 6. 72 0. 0148
Error 5797. 14
Table 3 Number of leaves and pustules ( scored on a subset of 10 leaves)
for T. repens clones , which were originally healthy ( n = 17) or origi2
nally diseased ( n = 14) at their natural locality , under three different
treatments : control , inoculated with one and ten strains of U. t rif olii2
repentis , respectively. Data from f ield experiment 1. Mean + S. E.
given
Treatments Nos of leaves
Originally
healthy
Originally
diseased
Nos of pustules
Originally
healthy
Originally
diseased
Control 63. 5 ±0. 56 65. 7 ±0. 66 0 0
Inoculated with 47. 4 ±0. 56 46. 2 ±0. 66 9. 2 ±1. 56 23. 2 ±1. 72
one strain
Inoculated with 38. 5 ±0. 56 38. 5 ±0. 66 10. 5 ±1. 93 45. 0 ±2. 12
ten strains
83 应 　用 　生 　态 　学 　报 　　　　　　　　　　　　　　　　　　14 卷
diseased plants for the production of leaves ( Table 3
and 4) .
　The originally healthy and originally diseased plants
differed with regard to pustule production for both of
the two inoculation treatments ( Table 3 and 4) . The
originally diseased plants developed significantly more
pustules that the originally healthy plants , i . e. a
twofold increase when inoculated with one strain and
a fourfold increase when inoculated with ten strains.
For originally healthy plants , there was no difference
between the two treatments , while for originally dis2
eased plants inoculation with ten strains resulted in
twice as many pustules as inoculation with one strain
( Table 3 and 4) .
Table 4 ANOVA2analyses for a) total leaf production ( DF = 2 , 507 ; 2 ,
507 ; 1 , 308 ; 1 , 308 and 1 , 308 , respectively) , and b) total number
of pustules ( counted on 10 randomly chosen leaves for each individual)
( DF = 1 , 338 ; 1 , 338 and 1 , 338 , respectively) for originally healthy
( n = 17) and originally diseased ( n = 14) T. repens clones under three
different experimental treatments. Data from f ield experiment 1
a) Leaf production
MS F2ratio P b) PustulesMS F2ratio P
Disease history
Originally healthy 27332. 6 501. 9 0. 0000 147. 6 0. 71 0. 4010
Error 54. 5 209. 2
Originally diseased 34282. 7 555. 5 0. 0000 33441. 4 37. 1 0. 0000
Error 61. 7 903. 6
Treatments
Control 1049. 6 1. 67 0. 2350
Error 46. 7
12strain 3 126. 8 2. 48 0. 1150 15014. 3 36. 2 0. 0000
Error 51. 5 414. 7
102strains 3 3 0. 96 0. 02 0. 8845 91507. 4 145. 1 0. 0000
Error 45. 41 630. 53 Inoculated with one rust strain , 3 3 Inoculated with ten rust strains. The same
below.
312 　Effects of previous inoculation on plant growth
and biomass in greenhouse
　The results obtained show that the effects of the
treatments in field experiment 1 were carried over to
this experiment . There was a significant effect of pre2
vious treatment for both originally healthy and origi2
nally diseased plants ( Table 5 and 6) . Control plants
showed the highest leaf production , while those
which had been inoculated with ten strains suffered
most ( Table 5 and 6) . However , no significant dif2
ferences were found with regard to original disease
history for comparison within any of the three previ2
ous treatments ( Table 5 and 6) .
　The pattern observed for leaf production was also
repeated for total biomass ( Table 5) . Control plants
showed the highest biomass ( mean + S. E. :172. 2 +
25. 6) , those inoculated with one strain were inter2
mediate ( 132. 4 + 25. 6) and those inoculated with
ten strains showed the lowest biomass ( 102. 4 +
25. 6) . However , the differences were not signifi2
cant . Neither did the disease history show any signifi2
cant difference in the data set , although the originally
diseased plants in all comparison showed a lower
biomass than the originally healthy ones.
313 　Effects of previous inoculation on plant growth
in field
　Plants that were originally diseased produced al2
most twice as many leaves as originally healthy ones
(mean + S. E. : 29. 2 + 1. 34 and 17. 6 + 0. 90 , re2
spectively , based on means of the mean value of each
station) . An analysis of variance showed significant
effects of both disease history and blocks ( Table 7) .
Table 5 Number of leaves and the biomass for T. repens plants which
had been originally healthy ( n = 17) or originally diseased ( n = 14) in
their original locality , after exposure to the three different treatments.
Data from greenhouse experiment 2. Mean + S. E. given
Treatments Nos of leaves
Originally
healthy
Originally
diseased
Biomass(mg FW)
Originally
healthy
Originally
diseased
Control 26. 3 + 0. 93 25. 7 ±1. 02 186. 5 ±45. 5 157. 8 ±24. 4
12strain 3 20. 5 ±0. 82 17. 8 ±0. 91 154. 7 ±45. 5 110. 2 ±24. 4
102strains 3 3 15. 2 ±0. 65 14. 4 ±0. 71 127. 4 ±45. 5 77. 3 ±24. 4
Table 6 ANOVA2analyses for a) number of leaves ( DF = 1 , 308 ; 1 ,
308 ; 1 , 308 ; 2 , 507 and 2 , 416 , respectively) and b) biomass ( DF =
1 , 18 ; 1 , 18 ; 1 , 18 ; 2 , 27 and 2 , 27 , respectively) for the T. repens
clones which were originally healthy ( n = 17) or originally diseased ( n
= 14) in their original locality , after exposure to the three different
treatments. Data from greenhouse experiment 2
a) Number of leaves
MS F2ratio P b) BiomassMS F2ratio P
Treatments
Control 31. 5 0. 21 0. 6431 174845 0. 12 0. 7292
Error 146. 5 1413712
12strain 3 557. 5 4. 85 0. 7270 18605 0. 05 0. 8274
Error 115. 1 380336
102strains 3 3 59. 0 0. 83 0. 3625 2610754 1. 19 0. 2891
Error 71. 2 2188157
Diseased history
Originally healthy 5212. 8 45. 98 0. 0000 1301327 0. 67 0. 5197
Error 113. 3 1940389
Originally diseased 9404. 1 43. 54 0. 0000 1664083 2. 33 0. 1166
Error 44925. 0 714414
Table 7 Analyses of variance for number of leaves for originally healthy
( n = 17) and originally diseased ( n = 14) T. repens clones after trans2
plantation to a natural meadow. Data from f ield experiment 2
Number of leaves
MS DF F2ratio P
Disease history 10405. 0 1 191. 9 0. 0000
Block 237. 3 9 4. 4 0. 0000
Error 54. 2 299
931 期 　　　　　　　　刘登义等 :白车轴草 ( Trif olium repens)植株抗病性和生长与植物病史的关系 　　　　　　　　　　　
Fig. 2 Leaf production for the 17 originally healthy and 14 originally diseased Trif oli um repens clones after transplantation to a meadow. Data from
field experiment 2. Mean + S. E. given.
　There was also a st riking difference when compar2
ing the leaf production of the originally healthy and o2
riginally diseased clones ( Fig. 2 ) . The originally
healthy clones only showed a minor variation in leaf
production , while the originally diseased ones showed
a large variation in leaf production.
4 　DISCUSSION
　We found a great variation among the 31 T rif oli2
um repens clones in their susceptibility to the ten dif2
ferent U rom yces t rif olii2repentis st rains used. There
was also a marked difference between these st rains in
their virulence. This great variation in resistance and
virulence accords with results of other inoculation ex2
periments using material f rom local natural popula2
tions[6 ,8 ,9 ,19 ]. It was also observed that there was a
marked difference in susceptibility between clones
that were originally healthy and those originally dis2
eased at the sampling occasion. However , four of the
originally healthy clones showed more susceptible re2
sponses against some of the fungal st rains and four of
the originally diseased clones did not show any
stronger susceptibility. Thus the disease pattern
found in the natural population of T . repens to some
extent reflects variations in susceptibility. This sug2
gests that variations in susceptibility may be of , at
least , some importance for explaining disease inci2
dence also in natural plant populations[11 ,14 ,27 ,45 ,48 ] ,
at least for annual pathogens.
　The different cultivation experiments , in which
T . repens cuttings were raised in pots , and the trans2
plantation experiment in the meadow , revealed some
striking differences. We found no differences between
the originally healthy and originally diseased plants in
the controls of the cultivation experiments for num2
bers of leaves produced. In the 1st greenhouse experi2
ment , however , there was a significant difference in
other measured characters , and the originally diseased
plants flowered more , while the originally healthy
plants produced more stolons. As the plants were cul2
tivated in shade in the 1st field experiment , flowering
was inhibited[55 ] . However , this did not increase leaf
production among the originally diseased plants as
could have been expected as a trade2off following
abortion of flower primordia. Neither , was any dif2
ference observed between originally healthy and origi2
nally diseased plants when inoculated with one or ten
fungal st rains , respectively , which could have been
expected considering differences in disease severity.
This lack of difference between the originally healthy
and originally diseased clones in the cultivation exper2
iments contrasts st rikingly to their marked differences
in the transplantation experiment , where the suscep2
tible clones were much more successful. The above
findings suggest that cultivation experiments in pots
may result in st rikingly different results compared to
natural conditions. This result is also in accordance
with data f rom other experiments. For example , Du2
dash[22 ] found that outcrossed and inbred progeny dif2
fered much more in response when grown under natu2
ral conditions compared to greenhouse conditions. E2
vans found that the species2specific patterns of differ2
entiation shown in T . repens against different grass
04 应 　用 　生 　态 　学 　报 　　　　　　　　　　　　　　　　　　14 卷
species were lost after the plants had been grown in
pots for two years in the absence of neighbouring
grasses[51 ] . Turkington[51 ] suggested that the micro2
bial population in the pots became similar , in the ab2
sence of grasses , thus resulting in quite similar
growth responses among the T . repens plants regard2
less of origin. Thus , uniform conditions in our pots
may explain the lack of difference between the origi2
nally healthy and originally diseased clones. Howev2
er , when transplanted back to a meadow , either the
interspecific competition , the presence of soil micro2
bials , or something else resulted in a marked different
response. The only thing we can state for sure is that
this difference in response was not due to the rust , as
that did not appear on the transplants during the
meadow experiment .
　We also observed in the meadow experiment that
the originally healthy clones responded in a uniform
way as they only showed a small variation in growth ,
compared to the originally diseased clones. This ob2
servation is a little astonishing if we consider that
some of the clones which were originally healthy
showed to be rather susceptible in the inoculation ex2
periments and did not differ f rom some of the origi2
nally diseased clones. This suggests that not only sus2
ceptibility is involved , but also other t raits must be
looked for [1 ] . Unfortunately , we did not measure
various stolon characters. However , we observed
during the experiments that originally healthy clones
had rather short , stout stolons which did not differ so
much in length , while originally diseased clones had
more slender stolons showing a large variation in
length. No doubt it would be worthwhile to investi2
gate whether differences in stolon characters may be
one important t rait that differ between the two
groups. This suggests that increased resistance may
be correlated to inferior growth , or rather an effective
exploitation of space (‘phalanx strategy’[43 ] ) . On
the other hand , susceptible genotypes should be
favoured in situations where suitable gaps are more
unpredictable in space (‘guerilla st rategy ’[43 ] ) .
However , it was a tendency in the 2nd greenhouse
experiment that susceptible clones were more severely
affected by earlier rust disease. This suggests that
they may be more severely affected in competitive sit2
uations in the presence of rust [10 ,47 ] which is also sup2
ported by the fact that they are likely to suffer an in2
creased mortality as an effect of disease[27 ,33 ,46 ] . Bur2
don , working with 50 T . repens clones f rom a single
field , found great variation in disease resistance for
two common foliar pathogens[4 ] . He also concluded
that the resistance to the two pathogens was inherited
independently. Our data point towards a great varia2
tion in resistance to U rom yces t rif olii2repentis , too.
As T . repens is f requently diseased by at least some
other pathogens too , it would be worth to study
whether variations in resistance towards different
pathogens , even if inherited independently , may be
correlated to some traits. The complexity of such
host2pathogen interactions may be illust rated by the
observation of Dirzo and Harper [20 ] . They found , in
one field , that cyanogenic T . repens morphs were of2
ten diseased by the systemic rust U rom yces f lectens
(s. n. U . t rif olii ) , while this was only occasionally
the case for acyanogenic morphs. However , they also
found cyanogenic morphs to be more sensitive towards
f rosts and that they suffered an increased mortality
during the winter. Whether this was due to the sys2
temic pathogen , or that cyanogenic morphs are more
sensitive to low temperature , or a combination of
both factors remains to be clarified.
　Systemic parasites are able to survive in the host
once they have established. Thus they ought to be
less dependent upon host population density and size
than are non2systemic parasites[13 ] . This pattern is
evident when comparing the occurrence of the two
U rom yces species , at least in northern Sweden (own
observations) . This suggests , in the case of T .
repens , that selection for resistance against systemic
fungi may increase under more competitive condi2
tions , when host population size is lower. No doubt ,
the different selection pressures offered by different
pathogens and their significance for the great differ2
entiation found in T . repens , even in a single popula2
tion , is worth more detailed experimental work.
REFERENCES
1 　Alexander HM. 1989. An experimental field study of anther2smut
disease of Silene alba caused by Ustilago violacea : Genotypic vari2
ation and disease incidence. Evolution ,43 :835～847
2 　Alexander HM ,Antonovics J . 1988. Disease spread and population
dynamics of anther2smut infection of Silene alba caused by the fun2
gus Ustilago violacea . J Ecol ,76 :91～104
3 　Bull JJ . 1994. Virulence. Evolution ,48 :1423～1437
4 　Burdon JJ . 1980a. Intra2specific diversity in a natural population of
Trif oli um repens. J Ecol . 68 :717～735
5 　Burdon JJ . 1980b. Variation in disease resistance within a popula2
tion of Trif oli um repens . J Ecol , 68 :737～744
141 期 　　　　　　　　刘登义等 :白车轴草 ( Trif olium repens)植株抗病性和生长与植物病史的关系 　　　　　　　　　　　
6 　Burdon JJ . 1983. Biological flora of the British Isles , Trif oli um
repens L . J Ecol ,71 :307～330
7 　Burdon JJ . 1987a. Diseases and Plant Population Biology. Cam2
bridge : Cambridge University Press.
8 　Burdon JJ . 1987b. Phenotypic and genotypic patterns of resistance
to the pathogen Phakosora pachyrhizi in populations of Glycine
canescens . Oecologia , 73 :257～267
9 　Burdon JJ , Brown AHD , Jarosz AM. 1990. The spatial scale of ge2
netic interactions in host2pathogen coevolved systems. In : Burdon
JJ and Leather SR eds. Pests , Pathogen and Plant Communities.
Oxford : Blackwell Scientific Publications. 233～247
10 　Burdon JJ , Groves RH , Kaye PE , Speer SS. 1984. Competition in
mixtures of susceptible and resistant genotypes of Chondrilla
juncea differentially infected with rust . Oecologia ,64 :99～203
11 Burdon JJ , Jarosz AM. 1988. The ecological genetics of plant2
pathogen interactions in natural communities. Philosophical Trans2
actions of the Royal Society of L ondon , Series B ,321 :349～363
12 　Burdon JJ , Oates JD , Marshall DR. 1983. Interactions between
A vena and Puccinia species I. The wild hosts : A vena barbata
Pott ex Link , A . f at ua L . , A . ludovicana Durieu. J A ppl Ecol ,
20 :571～584
13 　Carlsson U , Elmqvist T , Wennstrom A , Ericson L. 1990. Infection by
pathogens and population age of host plants. J Ecol ,78 :1094～1105
14 　Carlsson2Granér U. 1997. Anther smut disease in Silene dioica :
variation in susceptibility among genotypes and populations , and
patterns of disease within populations. Evolution , 51 :1416～1426
15 　Chin KM , Wolfe MS. 1984. The spread of Erysiphe graminis f . sp.
hordei in mixtures of barley varieties. Plant Pathol ,33 :89～100
16 　Clay K. 1991. Parasitic castration of plants by fungi. Trends in
Evolution and Ecology , 6 :162～166
17 　Davis WE. 1970. White clover breeding2review. Occasional Sy m2
posium of the B ritish Grassland Society ,6 :99～122
18 　Dinoor A. 1970. Sources of oat crown rust resistance in hexaploid
and tetraploid wild oats in Israel. Can J Bot ,48 :153～161
19 　Dinoor A. 1977. Oat crown rust resistance in Israel. A nnals of the
New York Academy of Sciences , 287 :357～366
20 　Dinoor A ,Eshed N. 1984. The role and importance of pathogens in
natural plant communities. A nn Rev Phytopathol ,22 :443～466
21 　Dirzo R , Harper JL . 1982. Experimental studies on slug2plant in2
teractions III. Differences in the acceptability of individual plants of
Trif oli um repens to slugs and snails. J Ecol ,70 :101～117
22 　Dudash MR. 1990. Relative fitness of selfed and outcrossed progeny
in a self2compatible , protandrous species , S abatia angularis L .
( Gentianaceae) : A comparison in three environments. Evolution ,
44 :1129～1139
23 　Ebert D , Mangin KL . 1997. The influence of host demography on
the evolution of virulence of a microsporidian gut parasite. Evolu2
tion ,51 :1828～1837
24 　Ennos RA. 1985. Measuring the effects of genetic variation upon plant fit2
ness. In :Haeck J & Woldentorp J W eds. Structure and Functioning of
Plant Populations. Vol. 2. Phenotypic and Genotypic Variation in Plant
Populations. Amsterdam: North2Holland. 53～160
25 　Garcia2Guzman G, Burdon JJ , Ash J E , Cunningham RB. 1996. Regional
and local patterns in the spatial distribution of the flower2infecting smut
fungus Sporisorium amphilophis in natural populations of its host Bothri2
ochloa macra. New Phytologist ,132 :459～469
26 　Garcia2Guzman G,Burdon JJ . 1997. Impact of the flower smut Ustilago
cynodontis (Ustilagiaceae) on the performance of the clonal grass Cynodon
dactylon ( Gramineae) . A mer J Bot ,84 :1565～1571
27 　Garren KH. 1995. Disease development and seasonal succession of
pathogens of white clover. Part II. Stolon disease and the damage2
growth2cycle. Plant Disease Reporter ,39 :339～341
28 　Giles BE , Goudet J . 1997. Genetic differentiation in Silene dioica
metapopulations : estimation of spatiotemporal effects in succession2
al plant species. The A merican Nat uralist ,149 :507～526
29 　Hokkanen H , Raatikainen M. 1977. Yield , vegetation and succes2
sion in reserved fields on Central Finland. Journal of the Scientif ic
A gricult ural Society of Finland ,49 :221～238
30 　Jarosz AM , Burdon JJ , Muller WJ . 1989. Long2term effects of
disease epidemics. J A ppl Ecol ,26 :725～733
31 　Jarosz AM , Burdon JJ . 1992. Host2pathogen interactions in natural
populations of L inum marginale and Melam psora li ni . Oecolo2
gia ,89 :53～61
32 　Jarosz AM , Davelos AL . 1995. Effects of disease in wild popula2
tions and the evolution of pathogen aggressiveness. New Phytolo2
gist ,129 :371～387
33 　Jarosz AM , Levy M. 1988. Effects of habitat and population struc2
ture on powdery mildew epidemics in experimental Phlox popula2
tions. Phytopathology ,78 :358～362
34 　Jφrstad I1964. Observations on life2cycles , spore2forms and alpine
occurrence of the Norwegian Uredinales. Nytt M agasin f or
Botanikk ,11 :27～45
35 　Jψrstad I. 1967. U romyces on Trif oli um repens . Nytt M agasin f or
Botanikk , 14 :19～30
36 　Kobel F. 1920. Zur Biologie der Trif olien2bewohnenden Uromycesarten.
Centralblatt f ur Bakteriologie , Abt . II . ,52 :215～235
37 　Lipsitch L , Herre E A , Nowak MA. 1995. Host population struc2
ture and the evolution of virulence : a‘law of diminishing returns’.
Evolution ,49 :743～748.
38 　Lipsitch M , Siller S , Nowak MA. 1996. The evolution of virulence
in pathogens with vertical and horizontal transmission. Evolution ,
50 :1729～1741
39 　Liro J I. 1906. Kulturversuche mit Finnischen Rostpilzen I. Acta
Societatis Pro Fauna et Flora Fennica ,29 (6) :1～25
40 　Liu D Y. 1994. Pathogenic fungi and natural plant populations I.
Effects of pathogenic fungi on natural plant population. Acta Ecol
S in ,14 (supp . ) :125～130 (in Chinese)
41 　Liu D Y. 1996. Pathogenic fungi and natural plant populations II.
Pathogenic fungi and plant population biology. Acta Ecol S in , 16
(6) :660～663 (in Chinese)
42 　Liu DY. 1997. Pathogenic fungi and natural plant populations III.
Pathogenic fungi2host plant coevolution and population models for vector
dispersed diseases. Acta Ecol Sin ,17(1) :105～108(in Chinese)
43 　Lovett Doust L , Lovett Doust J . 1982. The battle strategies of
plants. New Scientist ,95 :81～84
44 　May RM , Anderson RM. 1983. Epidemiology and genetics in the
coevolution of parasites and hosts. Proceedings of the Royal Society
of L ondon , B . ,219 :281～313
45 　Parker MA. 1988. Individual variation in pathogen attack and dif2
ferential reproductive success in the annual legume , A m phicarpaea
bracteata. Oecologia ,69 :253～259
46 　Paul ND , Ayres PG. 1986a. The impact of a pathogen ( Puccinia
lagenophorae) on populations of groundsel ( Senecio vulgaris) overwinter2
ing in the field I. Mortality , vegetation growth and the development of
size hierarchies. J Ecol ,74 :1069～1084
47 　Paul ND , Ayres PG. 1986b. Interference between healthy and rust2infect2
ed groundsel ( Senecio vulgaris L. ) within mixed populations of different
densities and populations. New Phytologist ,104 :257～269
48 　Roy BA. 1993. Patterns of rust infection as a function of host genet2
ic diversity and host density in natural populations of the apomictic
crucifer , A rabis holboellii . Evolution ,47 :111～124
49 　Scheiner SM . 1993. MANOVA : multiple response variables and
multispecies interactions. In Scheiner SM and Gurevitch J eds. De2
sign and Analysis of Ecological Experiments. New York : Chapman
& Hall. 94～112
50 　Snaydon RW , Davies MS. 1972. Rapid population differentiation in
a mosaic environment II. Morphological variation in A nthoxan2
thum odorat um . Evolution ,26 :390～405
51 　Turkington R , Holl FB , Chanway CP , Thompson JD. 1988. The
influence of microorganisms , particularly Rhizobium , on plant
competition in grass2legume communities. In : Davy AJ , Hutchings
MJ and Watkinson AR eds. Plant Population Ecology. Oxford :
Blackwell Scientific Publications. 343～366
52 　Wennstrom A , Ericson L. 1990. The interaction between the clonal herb
Trientalis europaea and the host specific smut fungus Urocystis triental2
is. Oecologia ,85 :238～240
53 　Wennstrom A , Ericson L. 1992. Environmental heterogeneity and disease
transmission within clones of Lactuca sibirica. J Ecol ,80 :71～77
54 　Wennstrom A , Ericson L. 1994. The effect and transmission of one isolate
of the rust Puccinia minussensis on five clones of Lactuca sibirica. Oe2
cologia ,97 :407～411
55 　Zaleski A. 1970. White clover in seed production. Occasional Symposium
of the British Grassland Society ,6 :147～154
作者简介 　刘登义 ,男 ,1958 年 9 月生 ,博士 ,教授 ,博导 ,主
要从事植物种群生态学、生物多样性与保护生物学和环境生
态学研究 ,发表论文 50 余篇 ,出版专著 3 部 , E2mail : ldy @
mail. ahnu. edu. cn
24 应 　用 　生 　态 　学 　报 　　　　　　　　　　　　　　　　　　14 卷