• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Impatiens Capensis (Balsaminaceae) Meerb. Is a Necessary Nurse Host for the Parasitic Plant Cuscuta Gronovii (Cuscutaceae) Willd. Ex J.A. Schultes in Southeastern Michigan Wetlands

Posted on: Saturday, 5 February 2005, 03:00 CST

ABSTRACT.-

Generalist parasites have been shown to have the potential to substantially affect the structure of the communities they inhabit. In order to predict the potential effects of these parasites, understanding the relationships they have with their host species is critical. In this study, the host range of Cuscuta gronovii was determined at two different times during the growing season, which corresponded with two life stages of parasitic individuals, seedling and adult. Field observations suggest that most of the successful infections by seedlings occurred upon only Impatiens capensis, one of the many species that it infects as an adult. This paper reports three tests of the hypothesis that I. capensis is a necessary nurse host for the parasite C. gronovii. In early summer survey plots, I. capensis was found to be infected by C. gronovii seedlings significantly more than predicted by the null hypothesis that all species in the plot were equally likely to be infected. In a late winter survey, in which all plants in a wetland infected with C. gronovii were identified and their positions mapped, I. capensis hosts were significantly closer to non-I. capensis hosts than predicted by the null hypothesis that infected I. capensis were distributed independently of non-I. capensis hosts. Finally, in a field experiment, newly germinated C. gronovii seedlings that were tied onto I. capensis and two other common hosts succeeded in infecting only on I. capensis. These results suggest that I. capensis acts as a necessary nurse host for C. gronovii in spring flooded southeastern Michigan wetlands.

INTRODUCTION

Some of the fundamental determinants of a species' distribution and abundance are its interactions with its resources. Due to the nature of the distribution of the resources necessary for plant growth, it is often difficult and time consuming to determine exactly which available resources a species is utilizing and which may be limiting its growth. However, this is less a problem with parasitic plants (Gibson and Watkinson, 1989). The resource environment of parasitic plants, especially those that are holoparasitic, can be characterized as a series of more or less discrete packages in the form of host plants. For generalist parasites, both the size of the resource reward in each package and the how hard it is to access this reward can vary greatly with the identity of the host species (Atsatt and Strong, 1970; Gibson and Watkinson, 1989; Kuiper et al., 1998; Pate and Bell, 2000). Knowledge of the host-range of a given parasite, and the spatial distribution and relative abundance of the various host and non- hosts in its local community, can potentially afford us great insight into demography and dynamics of the parasite population of that community. This paper reports complementary field techniques to determine the host-range of seedlings of the holoparasitic plant Cuscuta gronovii Willd. in southeastern Michigan wetlands.

Cuscuta gronovii is an annual parasitic vine that occurs in wet habitats throughout the eastern United States. Many species of Cuscuta are significant pests of agricultural crops such as mints, carrots, cranberries and flax (Kuijt, 1969). Despite this, until recently, there has been relatively little research directed at how these organisms complete their unique and interesting life cycle as native members in natural communities.

Details concerning the ecology of Cuscuta in general and C. gronovii, the most common member of the genus, are few and have been mostly provided as lists of species on which C. gronovii has been collected. The information that such lists provide about the host range of C. gronovii may be misleading. As vines, Cuscutas spp. are known to twine around many species from which they are unable to extract resources. Thus, further field observation and experimentation is necessary to determine the true host range.

Increasing evidence suggests that relationships between Cuscuata spp. and their hosts can be complex. Kelly and Horning (1999) provide evidence from laboratory experiments with C. attenuata that individuals are not able to exploit all host species with equal effectiveness and that the order in which an individual parasite interacts with hosts of different species may have fitness consequences for the parasite. Specifically, Kelly and Horning (1999) found that an excised piece of C. attenuata was able to attain greatest biomass when it first infected an annual, then a perennial. This suggests that a certain level of vigor may need to be attained by the parasite, by exploiting an easily infected but unproductive 'nurse host' before it can effectively infect other plants, which offer greater reward. This phenomenon may not be very important for an established plant, but could be very important for the establishment of excised stems as in Kelly and Homing's (1999) experiment, or for establishment of Cuscuta seedlings in the field.

Field observation suggests that a similar phenomenon occurs in the life-cycle of Cuscuta gronovii growing in Michigan wetlands. Early in the growing season, most of the infections by newly germinated seedlings appear on an annual host, Impatiens capensis Meerb., while later in the season the infections are concentrated on perennial hosts. In this paper, I describe observations and experiments designed to test the hypothesis that, in Michigan wetlands, I. capensis is used as a preferred 'nurse host' by C. gronovii seedlings, before it can infect perennial hosts preferred by established plants.

METHODS

Field studies were conducted in the Pinckney State Recreation Area (4224.350'N, 8358.560'W) and at the Matthaei Botanical Gardens (4218.280'N, 8339.496'W) in Washtenaw County, Michigan, where Cuscula gronovii has been reported growing on at least 40 different species from 25 different families (Voss, 1996). Nomenclature follows Voss (1972, 1985, 1996).

Surveys.-Surveys were conducted in late February (19-26 February 2001) after snow had melted but before the ground thawed and in early summer (6-24 June 2003), just after Cuscuta gronovii seeds have begun to germinate but each individual had infected only one or two host plants.

EARLY SUMMER SURVEY

Early summer surveys were conducted in two sites, which will be referred to as 'Hankerd Road' and 'Pickerel Lake'. Five circular plots of 0.5 m diameter were located randomly within each site. The identity and number of each species growing in these plots were recorded, as was the identity of infected individuals. The Pickerel Lake site was surveyed 2 wk later than the Hankerd Road site. This allowed more time for Cuscuta gronovii seedlings to germinate, and for early germinated individuals to make secondary infections.

The null-hypothesis that Cuscula gronovii seedlings are equally effective at infecting each species was simulated to test if C. gronovii seedlings were preferentially infecting some species more than others. For each observed C. gronovii seedling, the identity of its host is drawn randomly from all the species present in the plots with a probability equal to relative abundance of all that host's stems in the plot. The number of simulated infections of each species was recorded. This procedure was repeated 10,000 times to generate, for each potential host species, a frequency distribution to estimate the predicted probability distribution of infections. The observed number of infections for each species was compared to the predicted distribution to determine a P-value.

WINTER SURVEY

The winter survey was conducted at the Pickerel Lake site to identify and map the location of the hosts of all infections from the previous growing season. This survey was done while the ground was frozen to minimize the disturbance caused by walking through the wetland site and to facilitate the location and identity of plant stems in the absence of the lush vegetation of spring and summer. A circular plot with a radius of 15 m was established within which the locations and identity of all infected individuals were recorded. The hypothesis that Impatiens capensis is acting as a nurse host to Cuscuta gronovii and subsequently vectoring the parasite to other species predicts that there be a significantly shorter average distance from non-Impatiens hosts to the nearest infected Impatiens than if the infected Impatiens were randomly distributed in the plot. The null-hypothesis that infections of hosts other than Impatiens occur independently of infections of Impatiens was simulated to test this hypothesis. The test statistic used was the average distance of each non-Impatiens host to the nearest infected Impatiens individual. The probability distribution of this statistic predicted by the null hypothesis was estimated by simulation. Simulated locations of each infected Impatiens were chosen randomly by selecting x and y coordinates from a uniform distribution that ranged from negative 15 m to positive 15 m. If the resulting location did not fall in the circular plot, the random selection process was repeated. When all infected Impatiens had been placed in simulated locations among the observed \locations of the infected non-Impatiens hosts, the average distance from each infected non- Impatiens to the nearest infected Impatiens was calculated. This was repeated 10,000 times to generate a distribution of average distance from infected non-Impatiens host to nearest infected Impatiens. This frequency distribution approximates closely the probability distribution predicted by the null hypothesis. The probability of a distance less than the observed value was determined by this predicted probability distribution.

SEEDLING INTRODUCTION EXPERIMENT

In spring 2002, I conducted an experiment at Matthaei Botanical Gardens that consisted of introducing Cuscuta gronovii seedlings onto Impatiens capensis, Solidago patula Muhl. ex Willd. and Aster puniceus L., three hosts that are among the most commonly observed to be infected in end-of-season surveys.

Recently germinated Cuscuta gronovii seedlings, which had not yet made any infections, were collected from the site in which the experiment was to be conducted. These seedlings were gently tied with cotton twine onto 60 Impatiens capensis seedlings, 30 Solidago patula and 28 Aster puniceus stems. After 2 wk, the plants were observed and any successful infections recorded. Any infections resulting from seedlings other than those tied on were not counted. The hypothesis that I. capensis is a preferred nurse host for C. gronovii predicts that there would be a larger number of successful infections on I. capensis than on A. puniceus or S. patula. A chi square test was used to test if the observed distribution of successful infections differs from what would be predicted by the hypothesis that C. gronovii is equally capable of infecting each species.

TABLE 1.-Name, stem counts and corresponding percent of stems in summer surveys and infected stems in winter survey for each species identified. Percentages are calculated as column totals

RESULTS

EARLY SUMMER SURVEY

The early summer surveys resulted in the identification of 91 individuals from 8 species at the Hankerd Road site and 346 individuals from 10 species at the Pickerel Lake site. Of the 91 individuals identified at Hankerd Road, Cuscuta gronovii seedlings infected 11. At Pickerel Lake, C. gronovii seedlings infected 42 of the 346 individuals. At both sites, the most frequently infected species was Impatiens capensis: all 11 of the infections at Hankerd Road and 26 of the 42 infections at Pickerel Lake occurred upon I. capensis. This host species was also one of the most abundant species (24.9%) at both sites, ranking first in relative abundance at 65.9% at Hankerd Road and second at Pickerel Lake. Table> 1 lists the species identified and corresponding relative abundances at each site. For the Hankerd Road site, simulations of the null hypothesis resulted in as many as or more infections on I. capensis as observed in 94 of the 10,000 simulations (P = 0.0094) indicating that C. gronovii infections occur on I. capensis significantly more than predicted by the null hypothesis (Fig. 1). The number of infections on all other species did not differ significantly from the number predicted by the null hypothesis.

Simulations of the data from the Pickerel Lake site indicate that the number of Impatiens capensis stems infected is significantly greater than predicted by the null hypothesis (P < 0.001) (Fig. 2). Additionally, two species at Pickerel Lake, Pilea pumila (Urticaceae) (P = 0.0096) and Onoclea sensibilis (Dryopteridaceae) (P = 0.0054) were observed to be infected significantly fewer times than predicted by the null hypothesis. The number of infections on all other species did not differ significantly from the number predicted by the null hypothesis.

WINTER SURVEY

The winter survey and mapping of the infections in the Pickerel Lake site resulted in the identification of 705 stems from 7 species infected by Cuscuta gronovii (Table 1). The most commonly infected plant was Impatiens capensis (34%), followed by Eupatorium maculatum (Asteraceae) (20%); Solidago patula (18%) and Aster puniceus (14%). The average distance from infected non-I. capensis to the nearest infected I. capensis was 0.736 m, which is significantly less than that predicted by the hypothesis that infected I. capensis were distributed independently of other infected species (Mean = 0.872 m, SD = 0.0135, P = 0.018).

FIG. 1.-Observed (circles) and median predicted (squares) number of infections for each species occurring in plots in the early summer survey at Hankerd Road site. Error bars are 95% CI of predicted probability distributions. Species names are abbreviated as the first two letters of genus and specific epithet. See Table 1 for unabbreviated names

SEEDLING INTRODUCTION EXPERIMENT

Tying Cuscuta gronavii seedlings onto stems of three potential host species resulted in 17 successful infections. All of these infections occurred on stems of Impatiens capensis and none occurred on Aster puniceus or Solidago patula. The observed distribution of successful infections is significantly different than predicted by the null hypothesis that all three species are equally susceptible to infection by C. gronovii seedlings (χ^sup 2^ = 18.9, df = 2, P > 0.0001).

DISCUSSION

The 'nurse plant syndrome' refers to positive spatial correlation between species of plants, if one species, often called a benefactor, has a positive effect on the establishment or growth of another, the beneficiary (Callaway, 1995). There are various mechanisms through which such a positive effect can occur including the alteration of environmental conditions, accumulation of nutrient resources and protection from predators or herbivores. In many cases, nurse plants allow the persistence of beneficiaries in areas where they would otherwise be unable to persist.

The relationship between the benefactor species and the beneficiary species is much more direct and explicit in parasitic plants, which rely on other plant species directly for water, nutrients and/or carbon. In both cases, one species can potentially benefit from a large number of benefactor species. An active topic of research has been to discover the extent to which the species of the benefactor affects the magnitude of the positive effect to the beneficiary (Callaway, 1995).

FIG. 2.-Observed (circles) and median predicted (squares) number of infections for each species occurring in plots in the early summer survey at Pickerel Lake site. Error bars are 95% CI of predicted probability distributions. Species names are abbreviated as the first two letters of genus and specific epithet. See Table 1 for unabbreviated names

This study provides evidence that, in the communities studied, the host range of Cuscuta gronovii changes through ontogeny in such a way that only one of its hosts is available to it as a seedling. Successful establishment of the parasite on subsequent hosts depends establishment on this one species, which acts as a 'nurse host'. This phenomenon is predicted to result in a positive spatial correlation not only between the nurse species, Impatiens capensis, and the beneficiary, C. gronovii, but also between the nurse species and infected hosts of other species. The latter of which has been shown here to occur.

In other plant communities, Cuscuta gronovii seedlings are capable of infecting various perennial plant species. In addition to the communities described in this study, C. gronovii also inhabits other community types in southeastern Michigan. In one of these community types, C. gronovii is hosted predominantly by the perennial Decodon verticillatus (Lythraceae), upon which it establishes as a seedling. It is unknown whether the ability of C. gronovii seedling to infect perennials in some community types, but not others, is due to differences in the traits of the perennial hosts, or to differences in the populations of C. gronovii that occur in each of these community types.

The fact that annual Cuscuta gronovii establishment in some communities is dependent upon a single host in a much larger host range has important implications for understanding the population biology of C. gronovii in these systems. One of the central concepts in the population of infectious diseases is the threshold host population size, which is defined as the population size of host necessary for the long-term persistence of the parasite in the host population (Holt et al., 2003). If C. gronovii is dependent upon interaction with Impatiens capensis for establishment, then the local presence and abundance of the parasite might be much smaller than if its could establish as a seedling on all species in its adult host range. In fact, the presence and density of Impatiens capensis alone may be an important predictor of the presence or absence of C. gronovii in local sites; if I. capensis were absent, or occurred at a density below the threshold at which C. gronovii could sustain a population at a site deemed otherwise suitable, we might expect C. gronovii to be absent. Of course, there are many reasons why we might not find populations of G. gronovii in communities in which I. capensis is present and abundant.

In communities where discuta gronovii depends on Impatiens capensis for establishment, the long-term persistence of C. gronovii will depend upon the long-term persistence of I. capensis. This is a tenuous proposition because annual I. capensis is an inferior competitor for space in communities dominated by clonally spreading perennials. However, C. gronovii may contribute to the maintenance of I. capensis, and consequently its own population, in the community. This could occur, for example, if C. gronovii has a larger negative direct impact on the competitively dominant perennials in the system than it has upon I. capensis. If this were the case, the relationship between C. gronovii and I. capensis could be described as an indirect mutualism (Vandermeer, 1980; Wootton, 1994). Under these co\nditions, C. gronovii would benefit from the presence of I. capensis by using it as a nurse host, and I. capensis would benefit from the presence of C. gronovii through the reduction in asexual reproduction it causes in the perennials in the community, thus increasing the amount of open space available for the establishment of I. capensis seedlings.

I am unaware of published studies reporting interactions between hosts and parasitic plants that may be the result of the evolution of reduced impact to nurse hosts compared with non-nurse hosts. However, there is relevant literature available for parasites that have intermediate hosts. It suggests that if a parasite needs to infect multiple host species in the course of its life cycle, it is often less virulent to the intermediate host than to the host on which it reproduces (Davies et al., 2001; Parker et al., 2003). Similarly, vectored parasites are often observed to be less virulent to vectors than to hosts (Elliot et al., 2003).

The relationship between Cuscuta gronovii and Impatiens capensis is only an imperfect analogy with animal parasites that have obligate intermediate hosts or vectors, because C. gronovii is capable of completing its life cycle on I. capensis hosts. However, C. gronovii is able to produce almost ten-fold more fruits on some perennial hosts than on I. capensis (D. Schoolmaster, pers. obs.), which suggests that reproduction on I. capensis hosts contributes little to its overall fitness. If the amount of reproduction on I. capensis hosts is trivial, the analogy between I. capensis and an intermediate host or vector is more apt.

Cuscuta gronovii is usually regarded as a generalist parasite. However, this study demonstrates that this is not the case for all life stages in all habitats. In some habitats, C. gronovii not only prefers Impatiens capensis, but is dependent upon it to become established each spring. This phenomenon has important implications for the local maintenance of both species. The impact that the parasite has on the fitness of I. capensis and its competitively dominant, perennial hosts may determine whether or not both species are maintained in the community. Discovering these relationships and how they interact to determine the fate of both C. gronovii and I. capensis provides the basis for future study of this system.

Acknowledgments.-I thank the Matthaei Botanical Gardens for access to field sites, G. Estabrook and L. Eideitis for assistance with data collection and R. Burnham, M. Kummel and G. Estabrook for comments on the manuscript.

LITERATURE CITED

ATSATT, P. R. AND D. R. STRONG. 1970. The population biology of annual grassland hemiparasites. I. The host environment. Evolution, 24:278-291.

CALLAWAY, R. M. 1995. Positive interactions among plants. Botanical Reviao, 61:306-349.

DAVIES, C. M., J. P. WEBSTER AND M. E. J. WOOLHOUSE. 2001. Progenesis and reduced virulence as an alternative transmission strategy in a parasitic trematode. Pamsitology, 123:623-630.

ELLIOT, S. L., F. R. ADLER AND M. W. SABELIS. 2003. How virulent should a parasite be to its vector? Ecology, 84:2568-2574.

GIBSON, C. C. AND A. R. WATKINSON. 1989. The host range and selectivity of a parasitic plant: Rhinanthus minor L. Oecologia, 78:401-406.

HOLT, R. D., A. P. DOBSON, M. BEGON, R. G. BOWERS AND E. M. SCHAUBUR. 2003. Parasite establishment in host communities. Ecology Letters, 6:837-842.

KELLY, C. K. AND K. HORNING. 1999. Acquisition order and resource value in Cuscula allenuala. Proc. Natl. Acad. Sci., 96:13219-13222.

KUIJT, J. 1969. The biology of parasitic flowering plants. University of California Press, Berkley, California.

KUIPER, E., A. GROOT, E. C. M. NOORDOVER, A. H. PIETERSE AND J. A. C. VERKLEIJ. 1998. Tropical grasses vary in their resistance to Striga aspera, Striga Hermonthica and their hyprids. Can. J. Bot., 76:2131-2144.

PARKER, G. A., J. C. CHUBB, G. N. ROBERTS, M. MICHAUD AND M. MILINSKI. 2003. Optimal growth strategies of larval helminthes in their inermediate hosts. J Evol. Biol, 16:47-54.

PATE, J. S. AND T. L. BELL. 2000. Host associations of the introduced annual root hemiparasite Parentucellia viscosa in agricultural and bushland settings in western Australia. Annals of Botany, 85:203-313.

WOOTTON, T. J. 1994. The nature and consequences of indirect effects in ecological communities. Ann. Rev. Ecol. Syst., 25:443- 466.

VANDERMEER, J. 1980. Indirect mutualism: varitation on a theme by Stephen Levin. Am. Nat., 116:441-448.

Voss, E. G. 1972. Michigan flora part I: gymnosperms and inonocots. Granbrook Institute of Science Bulletin 61, Bloomfield Hills, Michigan.

_____. 1985. Michigan flora part II: dicots. Cranbrook Institute of Science Bulletin 61, Bloomfield Hills, Michigan.

_____. 1996. Michigan flora part III: dicots concluded. Cranbrook Institute of Science Bulletin 61, Bloomfield Hills, Michigan.

CONFERENCE PAPER

DONALD R. SCHOOLMASTER, JR.1

Department of Ecology and Evolutionary Biology, University of Michigan, Ann Arbor 48109

1 e-mail: dschoolm@umich.edu

Copyright American Midland Naturalist Jan 2005

Source: American Midland Naturalist, The

More News in this Category