Content uploaded by Ai-Rong Li
Author content
All content in this area was uploaded by Ai-Rong Li
Content may be subject to copyright.
Host Dependence and Preference of the Root
Hemiparasite, Pedicularis cephalantha Franch.
(Orobanchaceae)
Yong-Quan Ren &Kai-Yun Guan &Ai-Rong Li &
Xiao-Jian Hu &Le Zhang
#Institute of Botany, Academy of Sciences of the Czech Republic 2010
Abstract The interaction between parasitic plants and their hosts is an important
topic in both agriculture and ecology. Little, however, is known about that of the
hemiparasite Pedicularis. It is essential to understand the host dependence and
preference of Pedicularis for successful cultivation of plants in this genus and for
understanding the roles they play in natural communities. We tested the effects of
two herbaceous host species (Trifolium repens and Polypogon monspeliensis) on the
survival and growth performance of Pedicularis cephalantha. Five P. cephalantha
seedlings and two host plants were planted in each pot. In the control, no host plants
were planted (treatment 1). Host plants were planted in three combinations: only
T. repens (treatment 2) or P. monspeliensis (treatment 3) or a mixture of both
(treatment 4). The results showed that P. cephalantha performed better in the
presence of host plants, and host plants are more essential to P. cephalantha for
proper development than for survival. The grass host P. monspeliensis proved to be a
better host plant for P. cephalantha than the legume host T. repens. The high
dependence of P. cephalantha on host plants and its host preference were
demonstrated in this study. This is the first report of the performance of Chinese
Pedicularis species in cultivation throughout all life stages (from seeds to seeds).
Keywords Alpine plant .Haustorium .Host-parasite interaction .Host specificity .
Propagation
Folia Geobot (2010) 45:443–455
DOI 10.1007/s12224-010-9081-6
Y.-Q. Ren :K.-Y. Guan (*):A.-R. Li :X.-J. Hu :L. Zhang
Kunming Institute of Botany, Chinese Academy of Sciences, Kunming 650204(Yunnan,
People’s Republic of China
e-mail: guanky@mail.kib.ac.cn
Y.-Q. Ren :X.-J. Hu :L. Zhang
Graduate School, Chinese Academy of Sciences, Beijing 100039, People’s Republic of China
Introduction
Parasitic plants are common and important members in many plant communities,
and 3,000–5,000 plants can parasitize other plants (Marvier and Smith 1997; Qasem
2006). Parasitic plants can be divided into holoparasites and hemiparasites based on
their nutrient dependence on hosts. Hemiparasites contain chlorophyll and can
photosynthesize, but still partially rely on their host plants as a source of water and
nutrients through nodule-like structures called haustoria (Pate 2001; Wesselingh and
van Groenendael 2005). The best-studied parasitic plant species are generally
economically important parasites. However, compared with economically important
parasitic plants, our understanding of most economically less important parasites is
lacking, which is particularly the case with the knowledge of interaction between
these parasites and their host plants. Accumulating evidence has shown that
economically less important parasitic plants may play significant ecological roles
that have been overlooked before (Joshi et al. 2000; Quested et al. 2003; Press and
Phoenix 2005; Bardgett et al. 2006; Watson 2009), such as determining the structure
of natural communities (Gibson and Watkinson 1992; Joshi et al. 2000; Pywell et al.
2004; Press and Phoenix 2005), enhancing the amelioration of physical stress
conditions (Grewell 2008), and facilitating grassland restoration (Bullock and Pywell
2005; Watson 2009).
Pedicularis L. (Orobanchaceae) is a genus consisting of about 600 root
hemiparasitic species worldwide, primarily distributed in mountainous areas of
southwestern China, and widespread in frigid, alpine belts in the northern
hemisphere (Yang et al. 1998). Plants of this genus have high pharmacological
value (Guan et al. 2006) as well as potential ornamental value. Although Pedicularis
species have long been known as root hemiparasitic plants, only a very limited
number have been experimentally examined for their parasitic habit (Piehl 1963;
Weber 1976,1987; Lackney 1981). Little is known about the host requirements and
the relative importance of particular host species for the hemiparasitic Pedicularis.
The lack of such information represents a gap in our knowledge of the interaction
between these parasitic plants and their hosts.
Of a few existing studies regarding the host range of Pedicularis species,
conclusions about the host range of some Pedicularis species were made based on
field observations, in which cases host plants were determined merely by the
existence of haustoria (Piehl 1963; Weber 1976). Some root hemiparasites, however,
have been found to be able to form haustorium-like structures on non-host plants
(Cameron et al. 2006) or even on inorganic objects such as tiny pebbles (Piehl
1963). In most cases, such haustoria are poorly differentiated and have no actual
function in nutrient acquisition (Cameron et al. 2006). Nevertheless, it is extremely
difficult to tell the non-functional haustoria from functional ones without anatomic
analysis or other more sophisticated examinations. Consequently, it is impossible to
ascertain whether a host species upon which more haustoria are formed functions
better than a host with less haustoria without further examination. In this regard, a
host list compiled merely according to the existence of haustoria may be misleading
as a basis for attempts at the cultivation of these hemiparasites (Marvier and Smith
1997). A relatively reliable method to test whether a species can function as host is
to experimentally grow hemiparasites with it. Previous cultivation studies focusing
444 Y.-Q. Ren et al.
on the interactions between Pedicularis and its host plants have not proven
successful (Li and Guan 2008), which may account, at least partly, for the scarcity of
documented reports on cultivation of Pedicularis species. As far as we know, the
limited published investigations on cultivation of Pedicularis focus mainly on the
performance of Pedicularis in its early stages, such as seed germination and seedling
development (Lackney 1981; Li et al. 2007; Ren and Guan 2008). However, little is
known about the host dependence and preference of Pedicularis species based on
their whole life performance.
We designed the present study to understand host dependence and preference of
one Chinese Pedicularis species based on the whole life performance of the parasitic
species growing with different combinations of host species. We address the
following specific questions: i) Does performance of this hemiparasite depend on
host availability? ii) Does performance of this hemiparasite vary with different hosts
or host combinations? A better understanding of host effects is essential for
developing effective measures to cultivate Pedicularis species successfully. Answers
to these questions are necessary to understand the ecological roles the genus
Pedicularis plays in natural communities.
Material and Methods
Pedicularis cephalantha Franch., the target parasitic species in the present study, is a
herbaceous perennial endemic to Northwestern Yunnan, China, and widespread in
alpine meadows, 12–20 cm in height and with subcapitate inflorescences bearing red
flowers. Hemiparasites with large host plants suffer consequences of competition
with their own host (Calladine et al. 2000). As P. cephalantha are low herbaceous
plants in open habitats, we preferred to select low herbaceous hosts to avoid the
potential competitive influence. Herbaceous hosts are generally categorized into
three functional groups: grasses, legumes and forbs (Jiang et al. 2008). Forbs
generally represent the worst hosts for parasitic plants because they can employ
some defense mechanisms to prevent the abstraction of solutes in these host plants
(Cameron et al. 2006; Cameron and Seel 2007). Grasses are generally considered to
be good hosts because the finely branched architecture of grass roots may greatly
increase the likelihood of contact between the parasite and the host (Gibson and
Watkinson 1989; Marvier and Smith 1997). N-fixing legumes serve as ideal host
plants for some root hemiparasites (Tennakoon and Pate 1996; Radomiljac et al.
1999; Cameron et al. 2006). Furthermore, field surveys of natural habitats of some
Chinese Pedicularis species indicate that these hemiparasites may have close
associations with grasses and legumes (Ren et al., unpubl. data). Therefore, in the
present study, we selected one representative of each of these two groups as hosts.
As a host in cultivation, the plant species must adapt to the local climate and be easy
to grow. Thus, we selected Polypogon monspeliensis (L.) Desf. and Trifolium repens
L., both of which are widespread in Kunming, China, to test their effects on growth
performance of P. cephalantha.
The experiment was conducted in a nursery bed equipped with an auto-irrigation
spray appliance in the Kunming Botanical Garden (25°01′N, 102°41′E, altitude:
1,990 m). The nursery bed was 50 cm above ground to avoid root contamination
Host dependence and preference of the root hemiparasite 445
between treatments. P. cephalantha were planted in 9-inch earthenware pots filled
with humus soil. Five P. cephalantha seedlings and two host plants were planted in
each pot. In the control, no host plants were planted (treatment 1). Host plants were
planted in three combinations: only T. repens (treatment 2) or P. monspeliensis
(treatment 3) or a mixture of both (treatment 4). For each treatment, there were five
replicates with one exception of six replicates in treatment 2. For each replicate,
there were eight pots. Four pots were arranged in a line on the nursery bed, and a
replicate consisted of eight pots (two lines). All replicates were arranged randomly.
Altogether 168 pots were used in this study.
Seeds of P. cephalantha were collected in October 2007 from Shangri-La, Yunnan
(27°45′N, 99°46′E, altitude: 3,370 m). Seeds were dried under ambient conditions
and then stored in paper envelopes at 4°C until the initiation of experiments.
T. repens and P. monspeliensis seeds were collected in the Kunming Botanical
Garden in September 2007 and late March 2008, respectively. These seeds were stored
in paper envelopes under ambient conditions until the experiment commenced. Based
on the results of Ren and Guan (2008), the seeds of P. cephalantha were imbibed for
24 h with 500 ppm GA
3
at 25°C in darkness on 1st April 2008. After treatment, seeds
were cultured in 9-cm Petri dishes with two layers of filter paper saturated with
distilled water. Whenever needed to keep the filter paper moist, additional water was
added. Petri dishes were placed in a growth chamber at 10/20°C alternating
temperature (12 h : 12 h), with fluorescent lamp light during the period at a higher
temperature.
On 1 April 2008, seeds of hosts were sown in the pots, and enough host seedlings
were available 10 days later. Before transplanting P. cephalantha seedlings,
redundant host seedlings were removed, leaving only two according to the treatment
arrangement. On 20 April, five germinated seeds of P. cephalantha were trans-
planted into each pot (radicle was approximately 5–10 mm long). On 27 April, most
P. cephalantha seedlings emerged. The quantity of seedlings in each pot was
recorded. From then on, the survival quantity of P. cephalantha seedlings was
recorded weekly. To reduce competition from hosts, most leaves and stems of both
host plants in all host treatments were cut down to about 4 cm above soil level on 10
June for the first time. From then on, the practice of host defoliation was repeated
every two weeks. To protect the plants from pests, we sprayed pesticide weekly. All
pots were watered whenever needed to avoid water stress during the whole growing
season.
Harvesting commenced on 19 August, about 4 months after the hemiparasite and
hosts were planted together. At this time, some P. cephalantha plants were fruiting.
The growth parameters of P. cephalantha in each pot measured were as follows:
plant height, length of the longest leaf, and quantity of flowering plants. Four pots
(one line) from each replicate were selected to measure root and shoot dry weight.
Roots of P. cephalantha were washed carefully to reduce the damage to rootlets, and
then bagged and oven-dried at 80°C for 48 h before weighing with an electronic
balance. The remaining four pots in the other line from each replicate were left for
examinations of haustorial quantity and size. In each treatment, we selected pots
with only one P. cephalantha, to test the quantity and size of haustoria attached to
hosts. The soil was loosened by soaking for 24 h in water, to reduce the risk of
damaging haustorial connections (Gibson and Watkinson 1989; Kenji et al. 2008).
446 Y.-Q. Ren et al.
Successive washing progressively removed soil from the entangled root systems of
parasite and hosts, with haustorial connections to the host root mostly intact. The
value of haustorial diameter was used in haustorial size measurement, and we used
vernier caliper to measure diameter.
The survival rates of P. cephalantha were calculated from the quantities of living
plants at the time of harvest. The survival and flowering rates of P. cephalantha used
in analysis were obtained from the total of eight pots in each replicate. For length of
the longest leaf and height of plant, mean values of eight pots were used in analysis,
while for dry weight data, values of four pots from each replicate were used. Rate
data were arcsine-transformed, dry weight data were log-transformed, and data of
height and length were square root-transformed before analysis. The data were
analyzed by ANOVA and SNK test. No self-attached haustorium was observed from
the control (treatment 1), and haustoria were obtained only from eight pots in the
other three host treatments (2–3 pots per treatment). Flowering rate of
P. cephalantha was zero in the control (treatment 1). To avoid violating the
homoscedasticity assumption, differences in flowering rate, haustorial quantity and
size among three host treatments were analyzed using one-way ANOVA and SNK
test. Data for haustorial quantity were log-transformed before analysis. All statistical
analyses were performed using SPSS statistical software (SPSS 13.0 for Windows).
Results
At the time of harvest, the life stages of Pedicularis cephalantha plants varied in
different treatments and even within the same treatment. According to the criteria of
Petrů(2005), some Pedicularis plants grown with hosts matured, whereas all
Pedicularis in the control treatment (grown without any host) were still in the
juvenile stage (Fig. 1a). At harvest, most matured Pedicularis plants were at fruiting
stage and some mature seeds were obtained. The developmental stage of
P. cephalantha varied even in the same treatment, and some plants grown with
hosts were still in the juvenile stage. Figure 1bshows the different life stages of
P. cephalantha at the time of harvest.
Except for flowering rate, survival rate and root-shoot ratio did not depend on
host presence and host combinations. In all treatments, survival rates decreased over
time, the highest survival curve was obtained from P. cephalantha grown with a
mixture of both hosts and the lowest one was obtained from that grown without any
host (Fig. 2). However, there was no significant difference on survival rates between
treatments (F=1.267, P=0.317) (Fig. 3). In the control treatment, P. cephalantha
grown independently, the lowest value of root-shoot ratio was obtained, while the
difference between treatments was also not significant (F= 2.025, P= 0.149) (Fig. 3).
Flowering rate depended strongly on host presence, and no Pedicularis flowered in
the control treatment, but the difference was not significant between the three host
combinations (F=3.680, P=0.054) (Fig. 3).
All growth parameters of P. cephalantha depended strongly on host presence and
host combinations (Fig. 3). All growth parameters of P. cephalantha grown with
hosts were better than that of the control treatment, and all growth parameters varied
significantly among host combinations. Growth parameters of P. cephalantha grown
Host dependence and preference of the root hemiparasite 447
Fig. 2 Survival rate curves
of Pedicularis cephalantha
from different treatments. The
hemiparasite was grown
without host (No host), with
two individuals of the legume
Trifolium repens (2T), with
two individuals of the grass
Polypogon monspeliensis
(2P), and with one individual
each of T. repens and P.
monspeliensis (1T + 1P).
Vertical bars denote 1SE
Fig. 1 Growth performance of Pedicularis cephalantha and haustoria formed. aHost dependence of P.
cephalantha. The bottom left pot shows a flowering plant grown with host plants, the upper pots show P.
cephalantha grown in the absence of host being still in the juvenile stage or dead. bDifferent life stages of
P. cephalantha four months after transplanting; rule –30 cm. cHaustoria formed on a rootlet of
Polypogon monspeliensis;H–haustorium, HR –host root, PR –Pedicularis root. dA haustorium-like
structure formed on a tiny dead branch; H –haustorium, DB –dead branch, PR –Pedicularis root
448 Y.-Q. Ren et al.
with Polypogon monspeliensis were better than that grown with Trifolium repens.In
particular, for all growth parameters of P. cephalantha, the highest values were
obtained from plants grown with grass hosts of P. monspeliensis or the mixture of
both hosts. The highest values of total dry weight (DW), total shoot DW and total
root DW, length of the longest leaf, height of plants, and flowering rate were all
obtained from P. cephalantha grown with the mixture of both hosts. The highest
values of mean plant DW, mean plant shoot DW and mean plant root DW were all
obtained from P. cephalantha grown with grass hosts of P. monspeliensis. All these
parameters were significantly different (P<0.01) between treatments. Although
almost all the best growth parameters of P. cephalantha were obtained from the
treatment with a mixture of both hosts, all these parameters were not significantly
different from that of the treatment with grass hosts of P. monspeliensis alone (Fig. 3).
Neither P. monspeliensis nor T. repens had a significant effect on both survival
rate and root-shoot ratio of P. cephalantha by two-factor ANOVAs (Table 1).
However, two-factor ANOVAs indicated that host plants had significantly positive
effects on most other growth parameters. Of the two host species studied,
P. monspeliensis had much greater positive effects on growth of P. cephalantha
than T. repens (Table 1), with extremely significant effects (P<0.001) on flowering
rate, height, longest leaf length and all dry weight parameters four months after
Fig. 3 Growth parameters of Pedicularis cephalantha in different treatments. The hemiparasite was
grown without host (No host), with two individuals of the legume Trifolium repens (2T), with two
individuals of the grass Polypogon monspeliensis (2P), and with one individual each of T. repens and
P. monspeliensis (1T + 1P). Vertical bars denote 1SE, and different letters represent significant different
treatment means by SNK test at 5% level of significance
Host dependence and preference of the root hemiparasite 449
P. cephalantha and hosts were planted together, while T. repens had significant effects
only on height, total shoot DW (P<0.05), total DW and total root DW (P<0.01).
In all host treatments, Pedicularis plants formed haustoria on the root of host
plants (Fig. 1c). In the control, virtually no haustorium was observed except a
haustorium-like structure detected attached to a tiny dead branch in the substrate of
humus soil (Fig. 1d). The highest value of haustorium quantity was obtained from
P. cephalantha grown with two plants of P. monspeliensis, and the lowest value was
obtained from those grown with two plants of T. repens. The difference in
haustorium quantities among different parasite-host combinations was not significant
(P=0.093). Nevertheless, size of haustorium was significantly different in different
treatments (P=0.034), and the largest size was obtained from P. cephalantha grown
with T. repens (Table 2).
Discussion
The performance of Pedicularis cephalantha in the absence/presence of potential
hosts was examined in pot cultivation. In agreement with several previous studies on
root hemiparasitic species (Lackney 1981; Radomiljac et al. 1998; Matthies and Egli
1999; Puustinen and Salonen 1999; Loveys et al. 2002), we found that the presence
of host plants favored the performance of P. cephalantha. Studies on potted
Table 1 Analysis of variance of final Pedicularis cephalantha survival and growth parameters four
months after P. cephalantha plants were planted together with hosts Trifolium repens (T) and Polypogon
monspeliensis (P)
Source of
variation
T P T*P Error
d.f. MS Fd.f. MS Fd.f. MS Fd.f. MS
Flowering rate 1 58.067 0.850 1 1494.801 21.876*** 1 85.859 1.257 17 68.331
Survival rate 1 215.220 1.236 1 464.490 2.668 1 10.033 0.058 17 174.113
Height of plants 1 0.792 4.994* 1 6.295 39.699*** 1 0.159 1.001 17 0.159
Length of the
longest leaf
1 0.243 1.709 1 5.290 37.148*** 1 0.216 1.520 17 0.142
Total dry
weight (DW)
1 0.225 8.719** 1 2.251 87.263*** 1 0.158 6.122* 17 0.026
Total shoot DW 1 0.139 5.949* 1 1.826 78.211*** 1 0.103 4.422 17 0.023
Total root DW 1 0.130 9.995** 1 1.208 92.773*** 1 0.060 4.578* 17 0.013
Mean total DW 1 0.007 1.877 1 0.296 84.018*** 1 0.017 4.872* 17 0.004
Mean shoot
DW
1 0.003 1.226 1 0.168 69.429*** 1 0.012 4.968* 17 0.002
Mean root DW 1 0.003 4.119 1 0.054 68.882*** 1 0.004 5.008* 17 0.001
Root-shoot
ratio
1 1329.306 3.358 1 251.033 0.634 1 814.005 2.056 17 395.849
*–P<0.05, ** –P< 0.01, *** –P< 0.001. Rate/ratio data were arcsine-transformed, dry weight data were
log-transformed, and data of height and length were square root-transformed before analysis.
450 Y.-Q. Ren et al.
seedlings showed that P. cephalantha can survive for up to four months without any
host, indicating no obligate requirement for host species to survive as juveniles.
However, in the absence of hosts, all P. cephalantha plants were trapped in the
juvenile stage even after four months’growth, when they otherwise would be in the
reproductive phase, which means that host plants are essential to P. cephalantha not
for survival but for proper development. With the presence of hosts, we successfully
obtained seeds from seeds in the cultivation of Pedicularis species.
Different host species have different effects on the growth performance of
parasites and parasites achieve optimal performance only when attached to a suitable
host (Lackney 1981; Calladine et al. 2000; Cameron et al. 2006; Jiang et al. 2007).
Polypogon monspeliensis showed much greater effects on promoting the growth of
P. cephalantha than Trifolium repens in the present study. According to previous
research, root hemiparasitic plants generally have less host specificity and may
exploit multiple hosts simultaneously to benefit from different nutritional contributions
(Marvier and Smith 1997;Marvier1998). The treatment with the mixture of two hosts
showed the highest survival rate and the best overall growth performance.
The present study shows that P. monspeliensis is a good host. A similar observation
has been made by a prior field study (Gawler et al. 1987), which found that Pedicularis
furbishiae seedlings survived better under graminoid-dominated vegetation than under
forb-dominated. In a cultivation experiment, Matthies and Egli (1999) demonstrated
that the grass Lolium perenne was a far better host than the legume Medicago sativa for
the hemiparasite Rhinanthus alectorolophus. The probability of contacting and
successfully attacking a host is a main factor involved in determining host suitability
(Marvier and Smith 1997), and grass hosts have a high contact rate and thus likelihood
of a successful parasite attack (Gibson and Watkinson 1989). P. monspeliensis has
highly branched and thin roots, and can therefore be a particularly good host in this
study. The highest quantity of haustoria was obtained from P. cephalantha grown with
P. monspeliensis, and the performance of P. cephalantha grown with P. monspeliensis
was almost as good as that grown with the mixture of both hosts.
Being a legume species, T. repens was expected to be the best host for
P. cephalantha. However, T. repens in treatment 2 was proved to be the “worst host”.
Unlike P. monspeliensis,T. repens has less fibrous rootlets and less surface area and
thus a reduced likelihood of being attacked by haustoria of P. cephalantha. Poor
growth of Cuscuta subinclusa on a legume host Lotus scoparius was previously
suggested to be due to fewer haustorial connections (Kelly 1990). Similarly, in this
study, fewer haustorial connections were formed on T. re p e n s than on
Table 2 Quantity and size of haustoria between one Pedicularis cephalantha individual and two hosts
Hosts 2T 2P 1T + 1P d.f. MS FP
Haustorial quantity 35.25± 22.65a 139.50 ± 79.90a 79.90±31.11a 2 0.297 3.972 0.093
Haustorial size (mm) 0.88 ±0.08b 0.73± 0.07ab 0.62± 0.10a 2 0.046 7.173 0.034
Data are means and standard deviations, different letters represent significant different treatment means by
SNK test at 5% level of significance. The hemiparasite was grown with two individuals of Trifolium
repens (2T), with two individuals of Polypogon monspeliensis (2P), and with one individual each of
T. repens and P. monspeliensis (1T + 1P). Data for haustorial quantity were log-transformed before analysis.
Host dependence and preference of the root hemiparasite 451
P. monspeliensis.T. repens grew quickly and often shaded seedlings of
P. cephalantha. Thus, P. cephalantha had to compete with its host instead of
deriving benefits from it, or the benefit was cancelled out by competition. According
to Matthies (1995), host-parasite relationships may also include competitive
interaction. The bad performance of P. cephalantha in this treatment may be due
in part to competition from hosts. Well developed haustoria and a less defensive
response of the host, rather than the ability to fix nitrogen, may be the primary
factors influencing the performance of hemiparasite (Jiang et al. 2008). Taking these
factors of fewer haustorial connections and strong competition into account,
T. repens is not necessarily as good a host as expected. This result is consistent
with the findings of Radomiljac (1998), who reported that legumes are not
necessarily better hosts than non legumes for Santalum album.
Although the simple comparison between two hosts in the present study does not
allow a general conclusion concerning host quality, it seems that the root architecture
of hosts is an important factor involved in determining host suitability. Host
preference provides a mechanism by which the parasites can selectively parasitize
components of the natural community in which they occur and then can affect the
community structure (Gibson and Watkinson 1989,1992; Pywell et al. 2004). It this
study, we found that P. cephalantha benefit more from P. monspeliensis than
T. repens. Although the effects of P. cephalantha on host performance are not
addressed in this study, the host preference of P. cephalantha can be an important
guideline for the cultivation of this genus as well as other research in ecology.
Parasites have been reported to suffer reduced survival rates as a result of
competition with their host (Keith et al. 2004; Ahonen et al. 2006). To reduce
competition from hosts, host defoliation was conducted in the present study.
However, relative host suitability to a hemiparasite can change with defoliation
(Puustinen and Salonen 1999). Effects of defoliation may vary among hosts, because
the manner in which hosts suffer from defoliation may be significantly different.
Thus, results of this study should be interpreted carefully in comparing host quality.
Further experiments including a gradient of defoliation would be necessary to
determine the relative importance of different host species.
In this experiment, P. cephalantha plants had root-shoot ratios between 0.26 and
0.50, which are similar to the ratios (between 0.29 and 0.34) of the hemiparasite
Santalum album (Radomiljac et al. 1998). Jiang et al. (2007) obtained lower root-
shoot ratios (0.043–0.23) with the hemiparasite Rhinanthus minor. These are low
ratios compared to the overall result (6.3 for mean value and 5.7 for median value)
of Yang et al. (2010) based on a large-scale survey on China’s grasslands plants in
265 sites. Seedlings grown independently in treatment 1 were the smallest plants and
had the lowest root-shoot ratios. In the early life stage, P. cephalantha establishes a
large shoot at the expense of its own root system. This may be an adaptation to
increase its ability of photosynthesis, which maximizes plant growth by giving
priority to shoot growth over root growth. The relatively small root reduces nutrient
and water uptake, and thus a lower root-shoot ratio may lead to a higher mortality
rate of P. cephalantha. However, maybe this represents a strategy for the
hemiparasite that does not depend so much on its own root system. Hemiparasite
rely on its host to avoid water and mineral deficiencies (Glatzel and Geils 2009), and
to obtain part of its carbon nutrition (Logan et al. 2002).
452 Y.-Q. Ren et al.
This study clearly demonstrates that the presence of suitable hosts improves the
growth performance of P. cephalantha. The results of this study support the
hypothesis that hemiparasites depend on hosts and may respond differently to
different hosts. The host preference of Pedicularis species indicates that they may
influence the structure of natural communities by modifying the competitive
relationships between the component host species. The large differences in
P. cephalantha growth performance when grown with different host arrangements
in this study highlight the importance of identifying suitable hosts or host
combinations to promote the cultivation of the genus of Pedicularis.
Acknowledgements We are grateful to Jing-Xiu Li, Xue-Qing Yin, Nan Jiang and Li-Hua Yang from
KIB for their assistance with experiments. We thank Ms. Margaret Cargill and Dr. Patrick O’Connor of the
University of Adelaide, Ms. Juliet Nadeau Lu, Mr. Sailesh Ranjitkar and Mr. Anup Sharma from KIB for
their critical reading of our early draft. We also thank the anonymous reviewers whose comments
significantly improved the manuscript. This research was financially supported by the National Natural
Science Foundation of China (Grant No. 30670207 and 30970288), the National Natural Science
Foundation of Yunnan Province (Grant No. 2009CD114), and a “Western Light”Talent Training Program
of the Chinese Academy of Sciences for Ai-Rong Li.
References
Ahonen R, Puustinen S, Mutikainen P (2006) Host use of a hemiparasitic plant: no trade-offs in
performance on different hosts. J Evol Biol 19:513–521
Bardgett RD, Smith RS, Shiel RS, Peacock S, Simkin JM, Quirk H, Hobbs PJ (2006) Parasitic
plants indirectly regulate below-ground properties in grassland ecosystems. Nature 439:969–
972
Bullock JM, Pywell RF (2005) Rhinanthus: a tool for restoring diverse grassland? Folia Geobot 40:273–
288
Calladine A, Pate JS, Dixon KW (2000) Haustorial development and growth benefit to seedlings of the
root hemiparasitic tree Nuytsia floribunda (Labill.) R. Br. in association with various hosts. Ann Bot
(Oxford) 85:733–740
Cameron DD, Seel WE (2007) Functional anatomy of haustoria formed by Rhinanthus minor: linking
evidence from histology and isotope tracing. New Phytol 174:412–419
Cameron DD, Coats AM, Seel WE (2006) Differential resistance among host and non-host resistance
underlie variable success of the hemi-parasitic plant Rhinanthus minor.Ann Bot (Oxford) 98:1289–
1299
Gawler SC, Waller DM, Menges ES (1987) Environmental factors affecting establishment and growth of
Pedicularis furbishiae, a rare endemic of the St. John river valley, Maine. Bull Torrey Bot Club
114:280–292
Gibson CC, Watkinson AR (1989) The host range and selectivity of a parasitic plant: Rhinanthus minor L.
Oecologia 78:401–406
Gibson CC, Watkinson AR (1992) The role of the hemiparasitic annual Rhinanthus minor in determining
grassland community structure. Oecologia 89:62–68
Glatzel G, Geils BW (2009) Mistletoe ecophysiology: host-parasite interactions. Botany 87:10–15
Grewell BJ (2008) Hemiparasites generate environmental heterogeneity and enhance species coexistence
in salt marshes. Ecol Applications 18:1297–1306
Guan F, Yu L, Yang Y (2006) Research progress of Pedicularis L. Northw Pharm J 21:142–144
Jiang F, Timergalina L, Kudoyarova G, Jeschke WD, Hartung W (2007) Growth and development of
the facultative root hemiparasite Rhinanthus minor after removal of its host. FunctPlBiol34:237–
245
Jiang F, Jeschke WD, Hartung W, Cameron DD (2008) Does legume nitrogen fixation underpin host
quality for the hemiparasitic plant Rhinanthus minor?J Exp Bot 59:917–925
Joshi J, Matthies D, Schmid B (2000) Root hemiparasites and plant diversity in experimental grassland
communities. J Ecol 88:634–644
Host dependence and preference of the root hemiparasite 453
Keith AM, Cameron DD, Seel WE (2004) Spatial interactions between the hemiparasitic angiosperm
Rhinanthus minor and its host are species-specific. Funct Ecol 18:435–442
Kelly CK (1990) Plant foraging: a marginal value model and coiling response in Cuscuta subinclusa.
Ecology 71:1916–1925
Kenji S, Atsushi K, Makoto K (2008) Host rang and selectivity of the hemiparasitic plant Thesium
chinense (Santalaceae). Ann Bot (Oxford) 102:49–55
Lackney VK (1981) The parasitism of Pedicularis lanceolata Michx., a root hemiparasite. Bull Torrey Bot
Club 108:422–429
Li AR, Guan KY (2008) Arbuscular mycorrhizal fungi may serve as another nutrient strategy for some
hemiparasitic species of Pedicularis (Orobanchaceae). Mycorrhiza 18:429–436
Li AR, Guan KY, Probert RJ (2007) Effects of light, scarification and gibberellic acid on seed germination
of eight Pedicularis species from Yunnan, China. HortScience 42:1259–1262
Logan BA, Huhn ER, Tissue DT (2002) Photosynthetic characteristics of eastern dwarf mistletoe
(Arceuthobium pusillum Peck) and its effects on the needles of host white spruce (Picea glauca
(Moench) Voss). Pl Biol 4:740–745
Loveys BR, Tyerman SD, Loveys BR (2002) Effect of different host plants on the growth of the root
hemiparasite Santalum acuminatum (quandong). Austral J Exp Agric 42:97–102
Marvier M (1998) A mixed diet improves performance and herbivore resistance of a parasitic plant.
Ecology 79:1272–1280
Marvier MA, Smith DL (1997) Conservation implications of host use for rare parasitic plants.
Conservation Biol 11:839–848
Matthies D (1995) Parasitic and competitive interactions between the hemiparasites Rhinanthus serotinus
and Odontites rubra and their host Medicago sativa.J Ecol 83:245–251
Matthies D, Egli P (1999) Response of a root hemiparasite to elevated CO
2
depends on host type and soil
nutrients. Oecologia 120:156–161
Pate JS (2001) Haustoria in action: case studies of nitrogen acquisition by woody xylem-tapping
hemiparasites from their hosts. Protoplasma 215:204–217
PetrůM (2005) Year-to-year oscillations in demography of the strictly biennial Pedicularis sylvatica and
effects of experimental disturbances. Pl Ecol 181:289–298
Piehl MA (1963) Mode of attachment, haustorium structure, and hosts of Pedicularis canadensis.Amer J
Bot 50:978–985
Press MC, Phoenix GK (2005) Impacts of parasitic plants on natural communities. New Phytol 166:737–
751
Puustinen S, Salonen V (1999). The effect of host defoliation on hemiparasitic-host interaction between
Rhinanthus serotinus and two Poa species. Canad J Bot 77:523–530
Pywell RF, Bullock JM, Walker KJ, Coulson SJ, Gregory SJ, Stevenson MJ (2004) Facilitating grassland
diversification using the hemiparasitic plant Rhinanthus minor. J Appl Ecol 41:880–887
Qasem JR (2006) Parasitic weeds and allelopathy: from hypothesis to the proof. In Reigosa MJ, Pedrol N,
González L (eds) Allelopathy: A physiological process with ecological implications. Springer Verlag,
Dordrecht, pp 565–637
Quested HM, Press MC, Callaghan TV (2003) Litter of the hemiparasite Bartsia alpina enhances plant
growth: evidence for a functional role in nutrient cycling. Oecologia 135:606–614
Radomiljac AM (1998) The influence of pot host species, seedling age and supplementary nursery
nutrition on Santalum album Linn. (Indian sandalwood) plantation establishment within the Ord
Irrigation Area, Western Australia. Forest Ecol Managem 102:193–201
Radomiljac AM, McComb JA, Shea SR (1998) Field establishment of Santalum album L. –the effect
of the time of introduction of a pot host (Alternanthera nana R. Br.). Forest Ecol Managem
111:107–118
Radomiljac AM, McComb JA, McGrath JF (1999) Intermediate host influences on the root hemi-parasite
Santalum album L. biomass partitioning. Forest Ecol Managem 113:143–153
Ren YQ, Guan KY (2008) Effects of moist-chilling and GA
3
applications on seed germination of three
Pedicularis species from Yunnan, China. Seed Sci Technol 36:225–229
Tennakoon KU, Pate JS (1996) Heterotropic gain of carbon from hosts by the xylem-tapping root
hemiparasite Olax phyllanthi.Oecologia 105:369–376
Watson DM (2009) Parasitic plants as facilitators: more Dryad than Dracula? J Ecol 97:1151–1159
Weber HC (1976) Studies on new types of haustoria in some central European Rhinanthoideae
(Scrophulariaceae). Pl Syst Evol 125:223–232
Weber HC (1987) Studies in cultivated parasitic Scrophulariaceae. II. Host parasite interactions. Flora
179:35–44
454 Y.-Q. Ren et al.
Wesselingh RA, van Groenendael JM (2005) The biology of non-weedy hemiparasitic Orobanchaceae:
synthesis and perspectives. Folia Geobot 40:311–318
Yang HP, Holmgren NH, Mill RR (1998) Pedicularis L. In Wu ZY, Raven PH (eds) Flora of China 18.
Science Press, Beijing and Missouri Botanical Garden Press, St. Louis, pp 97–209
Yang YH, Fang JY, Ma WH, Guo DL, Mohammat A (2010) Large-scale pattern of biomass partitioning
across China’s grasslands. Global Ecol Biogeogr 19:268–277
Received: 29 January 2009 / Revised: 3 June 2010 / Accepted: 15 June 2010 /
Published online: 8 September 2010
Host dependence and preference of the root hemiparasite 455
