Volume 101, Issue 3 p. 607-613
Essay Review
Free Access

Pathogen accumulation and long-term dynamics of plant invasions

S. Luke Flory

Corresponding Author

S. Luke Flory

Agronomy Department, University of Florida, PO Box 110500, Gainesville, FL, 32611 USA

Correspondence author. E-mail: [email protected]Search for more papers by this author
Keith Clay

Keith Clay

Department of Biology, Indiana University, 1001 East 3rd Street, Bloomington, IN, 47405 USA

Search for more papers by this author
First published: 06 March 2013
Citations: 165

Summary

  1. The diversity of pathogens on highly abundant introduced hosts has been positively correlated with time since introduction, geographical range of the introduced species and diversity of invaded habitats. However, little is known about the ecological effects of pathogen accumulation on non-native invasive plants.
  2. Pathogen accumulation on invasive plant species may result from ecological processes such as high plant densities, expanding geographical ranges and pathogen dispersal from the native range, or evolutionary mechanisms such as host range shifts and adaptation of native pathogens to invasive species.
  3. Over time pathogen accumulation may cause decline in the density and distribution of invasive plants and facilitate recovery of native species. Alternatively, pathogens might build up on invasive species and then spill back onto co-occurring native species, further exacerbating the effects of invasions.
  4. Synthesis. Research efforts should focus on determining the long-term outcomes of pathogen accumulation on invasive species. Such research will require multifaceted approaches including comparative studies of diverse invasive species and habitats, experimental manipulations of hosts and pathogens in nature and controlled environments, and predictive models of host-pathogen interactions within an invasion framework. Results of this research will improve our understanding and ability to predict the outcomes of biological invasions.

Introduction

A primary goal of ecological research has been to identify mechanisms underlying biological invasions, which are key drivers of global environmental change and require substantial economic resources for management. Dozens of hypotheses have been proposed to explain why some species become invasive (Catford, Jansson & Nilsson 2009) and why some habitats are vulnerable to invasion (Stohlgren et al. 2002). However, abiotic and biotic factors affecting invasive species are likely to change over time due to varying environmental conditions, habitat disturbance, repeated introductions of distinct genotypes, expanding distributions leading to novel interactions and evolution of invaders or co-occurring pathogens (e.g. Bossdorf et al. 2005; Gilbert & Parker 2010). Research on post-introduction evolution (Maron et al. 2004) and the introduction of biological controls (reviewed by Culliney 2005) suggests that interactions of invasive species with enemies, competitors and mutualists can co-evolve and feedback to affect the dynamics of invasive species over time.

Of the many hypotheses to explain biological invasions, there is perhaps the greatest empirical support for the release of invasive species from natural enemies when they are introduced into a new range, which then gain a competitive advantage over resident species (but see van Kleunen & Fischer 2009; Chun, van Kleunen & Dawson 2010). For example, experimentally excluding pests from the shrub Clidemia hirta promoted growth in its native, but not non-native, range (DeWalt, Denslow & Ickes 2004). Similarly, black cherry (Prunus serotina) seedling recruitment is suppressed near adult trees in native U.S. habitats by below-ground Pythium pathogens, but not in Europe where it invades forest understories (Reinhart et al. 2003). Although invasive species may initially benefit from enemy escape, over the longer-term pathogens can accumulate, potentially reducing the effects of invaders on native communities and ecosystems. For example, in an analysis of 124 plant species, those that were introduced 400 years ago had six times more pathogens than those introduced only recently (Mitchell et al. 2010), but the effects of those pathogens on hosts were not known. Alternatively, pathogen accumulation may exacerbate the effects of invasions if pathogens spill back onto co-occurring native species and reduce their performance and competitive inhibition of invasive species (Kelly et al. 2009). The long-term dynamics of plant invasions and their impacts on native communities and ecosystems may be determined by interactions among pathogens, invasive plants and co-occurring native species. Here, we describe the patterns, mechanisms and potential outcomes of pathogen accumulation on invasive species with the goal of generating discussion and research on how pathogens alter the long-term dynamics of plant invasions. The results of this research will improve our understanding and ability to predict the outcomes of biological invasions.

Patterns of pathogen accumulation

The limited research on pathogen accumulation in natural systems has demonstrated that the richness and diversity of pathogens on introduced hosts are positively correlated with time since introduction (Hawkes 2007), the geographical range of the introduced species (Strong & Levin 1975; Clay 1995) and the diversity of invaded habitats (Mitchell et al. 2010). Many invasive plants, including grasses, forbs, trees and ferns become infected by pathogens in their introduced ranges (Fig. 1). For example, the invasive grass Microstegium vimineum (stiltgrass) is infected by Bipolaris fungi (Fig. 1a, Kleczewski & Flory 2010; Flory, Kleczewski & Clay 2011; Kleczewski, Flory & Clay 2012), Alliaria petiolata (garlic mustard) by powdery mildew (Fig. 1c, Enright & Cipollini 2007) and Pueraria lobata (kudzu) by soybean rust (Fig. 1f, Harmon et al. 2006). Accumulation of below-ground pathogens, which are more difficult to identify, may also occur. For example, the invasive tree Sapium sebiferum (Chinese tallow) survived more poorly and produced less biomass in ‘home’ compared with ‘away’ soils (Nijjer, Rogers & Siemann 2007). Similarly, Diez et al. (2010) found that the strength of negative plant–soil feedback increased with time since introduction for 12 species in New Zealand. However, neither study identified biological agents responsible for decreased plant growth or excluded other potential mechanisms such as reduced benefit from soil mutualists or changes in soil chemistry. Nevertheless, demonstrations of accumulation of above-ground pathogens and increasing negative plant–soil feedbacks over time suggest that pathogen accumulation on invasive species is widespread and potentially important for regulating invasive plant populations.

Details are in the caption following the image
Examples of pathogen accumulation on invasive plants: (a) Bipolaris sp. fungus causing leaf blight disease on Microstegium vimineum (stiltgrass), (b) Pseudomonas syringae bacteria symptoms on Cirsium arvense (Canada thistle), (c) powdery mildew Erysiphe cruciferarum on Alliaria petiolata (garlic mustard), (d) Puccinia lygodii rust on Lygodium japonicum (Japanese climbing fern), (e) rose rosette disease symptoms on Rosa multiflora (multiflora rose) caused by an unknown virus and (f) Phakopsora pachyrhizi (soybean rust) on Pueraria lobata (kudzu). Photo credits: a: S. Luke Flory, b: Bob Hartzler, c: Don Cipollini, d: Min Rayamajhi, e: Fred First, f: Albert Tenuta.

Mechanisms of pathogen accumulation

Highly abundant species, such as invasive species, are expected to exhibit greater accumulation of pathogens over time relative to rarer, less-dense species (Bever 1994; Olff et al. 2000; Clay et al. 2008; Mitchell et al. 2010). However, it is less apparent whether pathogen accumulation occurs through ecological or evolutionary mechanisms, or by some combination. Ecological mechanisms include spatiotemporal dynamics, host density or abiotic conditions. For example, as species invade wider geographical areas, there will be greater encounter rates with native pathogens, and the number of pathogens that accumulate may increase in the same or even higher levels than on native species (Strong & Levin 1975; Clay 1995). Furthermore, the longer a species is present in its new range, the more likely a pathogen from its home range will be introduced (Mitchell et al. 2010). Invasive species growing at high densities also have an increased probability of intercepting inoculum, and shorter distances between hosts can promote the spread of disease (Garrett & Mundt 1999). High host density may also alter environmental conditions such as humidity and temperature that affect pathogen transmission (Burdon & Chilvers 1982; Alexander 2010), and intraspecific competition can increase the effects of infection (Lively et al. 1995). Finally, larger, denser host populations can more easily maintain pathogens that require a minimum host population size (Carlsson & Elmqvist 1992; Holt et al. 2003).

Pathogen accumulation may also be mediated by evolutionary mechanisms. Most pathogens exhibit some level of specificity whereby they only infect certain genotypes, species, genera or families (Keen & Staskawicz 1988; Gilbert & Webb 2007). Hosts may vary in resistance to pathogens or in tolerance to disease and pathogens may infect multiple hosts but may have a more detrimental effect on a particular host (Holah & Alexander 1999; Dobson 2004). In general, introduced species are more likely to accumulate pathogens in habitats containing closely related species compared to habitats with more distantly related species (Agrawal & Kotanen 2003; Parker & Gilbert 2007). However, host–pathogen interactions can change over time; invasive plants may evolve reduced resistance when enemies are lacking (Bossdorf et al. 2004), or pathogens may evolve the ability to attack widespread invasive hosts (Parker & Gilbert 2007). For example, when plant species are introduced to new ranges and escape their natural enemies, they may evolve greater competitive ability at the expense of reduced defences and become more susceptible to generalist pathogens (Evolution of Increased Competitive Ability (EICA); Blossey & Notzold 1995; but see van Kleunen & Schmid 2003; Gurevitch et al. 2011). Identifying the ecological or evolutionary mechanisms underlying pathogen accumulation will help to determine the host ranges and impacts of pathogens on invasive and native species.

Possible outcomes of pathogen accumulation

While the process of pathogen accumulation has been of longstanding interest (Strong & Levin 1975; Mitchell et al. 2010), there has been little research on the ecological consequences of pathogen accumulation for introduced hosts. We hypothesize three possible outcomes of pathogen accumulation on invasive species. First, pathogen accumulation may result in decreased density and extent of invasions and recovery of native species. We refer to this outcome as the Pathogen Accumulation and Invasive Decline (PAID) hypothesis. Invasive species initially increase to high levels followed by increasing pathogen density and species richness, which then causes reductions in the density and distribution of the invasive species and potentially the recovery of native species (Fig. 2). Disease epidemics in both wild and domesticated plants (e.g. American chestnut) and control of invaders by biocontrol agents (e.g. Australian Opuntia spp.) demonstrate the potential of natural enemies to reduce or regulate host populations. Recently, we showed that experimental fungicide applications reduced infection of the invasive grass Microstegium vimineum by the fungal pathogen Bipolaris (Fig. 1a), while simultaneously increasing Microstegium biomass by up to 50% and seed production by up to 200% in natural populations (Flory, Kleczewski & Clay 2011). Similarly, Enright & Cipollini (2007) found that infection of Alliaria petiolata (garlic mustard) by a powdery mildew in the glasshouse significantly reduced plant growth and survival. In parallel, experimental removals of Microstegium (Flory & Clay 2009) and garlic mustard (Carlson & Gorchov 2004) in nature led to recovery of co-occurring native species. However, we are unaware of any system where the sequential pattern of pathogen accumulation, decline of invasive species and recovery of native species has been documented.

Details are in the caption following the image
Conceptual model illustrating the pathogen accumulation and invasive decline (PAID) hypothesis.

Second, invasive species hosting abundant pathogens may cause increased disease on co-occurring native hosts and reduce their competitive suppression of invasive species (Fig. 3). Such dynamics are termed ‘spillover’ when the pathogens are non-native and introduced with the invader and ‘spillback’ when an invasive species hosts native pathogens. The general phenomenon has also been called the ‘Enemy of My Enemy Hypothesis’ (Colautti et al. 2004). For example, spillover of barley yellow mosaic virus from a highly susceptible invasive grass decreased the abundance of two native grasses through pathogen-mediated apparent competition (Power & Mitchell 2004). Similarly, Mangla, Inderjit & Callaway (2008) reported spillback where invasions of the forb Chromolaena odorata resulted in a 25-fold increase in Fusarium, a generalist soil pathogen, which strongly inhibited native plant species, and the non-native invasive cheatgrass (Bromus tectorum) serves as a reservoir for the native seed pathogen Pyrenophora semeniperda, which causes significantly greater death of native seed in invaded areas (Beckstead et al. 2010; see also Kotanen 2007 for seed pathogen effects on native tree species). Kelley et al. (2009) reviewed a number of other potential examples of spillback and suggested that it is a widespread, but underappreciated, component of biological invasions (see also Power et al. 2011). However, it is often unclear whether pathogens were introduced with the invasive species or whether native pathogens accumulated on the invasive plant following its introduction. Further research is needed to determine how frequently pathogen accumulation on invasive species results in spillback to native species.

Details are in the caption following the image
Conceptual model illustrating three alternative outcomes of pathogen accumulation over time: decline of invasive species (PAID hypothesis) and recovery of native species, no effect of pathogens through tolerance or compensation, or spillback of pathogens resulting in further native species decline and release of invasive species from competition.

Finally, pathogen accumulation may have little or no effect on invasive species due to tolerance, compensation or phenotypic plasticity (e.g. Gilbert & Parker 2006; Alexander 2010). For example, many invasive species exhibit substantial phenotypic plasticity such that a reduction in population density would have little effect on biomass or seed production per unit area. Given the theoretical and empirical demonstrations of the negative effects of pathogen build-up (e.g. Clay & Kover 1996; Packer & Clay 2004; Mordecai 2011), this outcome seems unlikely, although possible.

Research needs

Although theory predicts that abundant, high-density species will be more heavily attacked by pathogens than less common species (Clay et al. 2008; Mordecai 2011), we are unaware of any documented cases of pathogen accumulation leading to invasive species decline, and there are very few documented cases of spillback. To determine the dynamics of pathogen accumulation on invasive species and the outcomes for invasive and native species, a combination of literature reviews, meta-analyses, comparative field surveys, population modelling and manipulative experiments in laboratories, glasshouses and natural populations is required (Table 1). This information will help provide insights into the patterns and outcomes of biological invasions and will help guide management strategies.

Table 1. Predicted outcomes for native and invasive host species for each of the three proposed hypotheses: Pathogen Accumulation and Invasive Decline (PAID), Spillback and No effect. Outcomes of increasing (+), decreasing (−) or no change (0) in disease and host population levels are provided for three types of studies: observational across a chronosequence of young to old invasions, experimental suppression of disease in invaded field areas and experimental addition of disease in glasshouse, growth chamber, common garden or field setting
Type of study Species PAID Spillback No effect
Disease Host Disease Host Disease Host
Observations across chronosequence of invasions Native 0 + + 0 0
Invasive + + 0 or + + 0
Experimental suppression of pathogens Native 0 + 0 0
Invasive + + 0
Experimental addition of pathogens Native 0 + + 0 0
Invasive + + 0 or + + 0

Demonstrating pathogen accumulation requires evidence that the diversity of pathogen species or prevalence of pathogen infection on invasive hosts increases over time. However, many invasive species and pathogens were introduced accidentally and escaped evaluation for long periods, so there is little information on changes in their historical distributions and dynamics over time. To acquire better information, we should target invasions of known ages, including new invasions, and gather data as the invasion progresses or, alternatively, substitute space for time by evaluating pathogen richness and abundance at sites along a chronosequence of invasion history (Lankau et al. 2009; Diez et al. 2010). Observations of increasing disease and declining invasive host populations and recovery of native hosts across a chronosequence of invasions would support PAID. Alternatively, increasing disease and no effect or increasing invasive host populations over time, along with increasing disease and decline of native host populations, would suggest spillback dynamics (Table 1).

Field surveys across the range of an invasive species can be time-consuming and costly, but collaborations with natural areas managers, invasive species working groups, cooperative weed management areas and private land owners can significantly increase efficiency. For example, online resources (e.g. Great Britain Non-native Species Secretariat, nonnativespecies.org; US Midwest Invasive Plant Network, mipn.org; EPPO Lists of Invasive Alien Plants, eppo.int) can be utilized to obtain information about invasive plant diversity and distributions. Similarly, online data bases such as the USDA Fungus-Host Distributions data base (nt.ars-grin.gov ), the Virus Identification Data Exchange project (agls.uidaho.edu/ebi/vdie ) or the worldwide Distribution Maps of Plant Pests and Diseases (cabi.org) can be extremely helpful for identifying pathogens on plant species. In addition, species-specific reports of plant disease in the literature can be found using tools such as Web of Science. In some cases, herbarium records can help document spatial and temporal patterns of disease (Hood & Antonovics 2003). Some types of pathogens (e.g. foliar fungal pathogens) will be easier to identify and quantify than other groups (e.g. root pathogens), but there is no reason to expect that general patterns of pathogen accumulation and ecological impacts should differ among types of pathogens. Beyond the existing data bases and literature, determining the identity, source and distribution of pathogens on invasive species will require collaborations among ecologists, plant pathologists and taxonomists, and the use of morphological, microbiological and molecular techniques.

Elucidating the host range of pathogens will help determine whether spillback is possible and whether the pathogen likely colonized from co-occurring native species or from in situ evolution. Key questions include: Are pathogens specialists or generalists? Do they only infect the invasive species or are native species also susceptible? Do infection rates of pathogens vary among isolates, plant populations, habitats or co-occurrence of host species? Determining whether pathogens are from the native or introduced range of the invasive species, and whether they infect any other species, may help to clarify their co-evolutionary history, which could explain ongoing ecological and evolutionary dynamics. The first step in evaluating pathogen host range is to collect symptomatic samples during field surveys from functionally or phylogenetically related species that co-occur with the invasive species. Then, following the identification and culture of isolated pathogens, controlled laboratory and glasshouse inoculation studies should be completed on native species found in invaded habitats (Table 1; see also Charudattan & Dinoor 2000; Flory, Kleczewski & Clay 2011; Kleczewski, Flory & Clay 2012). Such studies must be conducted with caution to prevent the unintended spread of pathogens to uninfected populations. Experimental inoculation studies can also determine whether pathogen genotypes attack certain invasive plant genotypes or whether all invasive genotypes are susceptible (Schmid 1994), which may inform the potential for pathogen spread across the invasive host range. Field surveys in the native range of the invasive species could provide further information on pathogen diversity and prevalence. Further, information on natural pathogen levels in the native range would help inform the role of enemy release in the invasion process and predict which pathogens are likely to attack the host in the invasive range.

To determine how pathogen accumulation may affect the survival, growth or fecundity of invasive species and co-occurring native species, field, glasshouse and laboratory experiments manipulating the presence and density of pathogens, and the invasive and native plant species, are needed (Table 1). The addition or removal of pathogens depends on the feasibility of culturing and inoculating pathogens (while minimizing the risk of pathogen escape), and the efficacy of pesticides to reduce pathogen loads in nature. These types of approaches can also help determine how the effects of pathogens on invasive and native species vary with climate, latitude, soils and other environmental factors. Simultaneous infection by multiple pathogens will reveal whether their combined effects on invasive and native plant species are additive, synergistic or antagonistic.

Testing for spillback requires glasshouse or laboratory experiments using native communities with and without the invasive species in the presence and absence of pathogens. Spillback may also be tested in the field by planting uninfected natives into infected and pathogen-removed (via pesticides) plots in invaded and uninvaded communities. If invasive populations perform more poorly when pathogens are experimentally added, or better when pathogens are experimentally decreased (Table 1), we can conclude support for the PAID hypothesis. If increasing invasion density results in greater pathogen loads on invasive species, but greater negative effects of disease on native species than on the invader, then spillback may be occurring (Table 1; Kelly et al. 2009; Diez et al. 2010). However, non-susceptible native species could also be released from competition with other resident species subject to spillback, leading to further indirect changes in community structure and composition (Table 1).

Experimentally altering the presence of pathogens and native species can help determine the immediate consequences of pathogen accumulation on populations and communities, but will not predict the effects of pathogen accumulation over longer-time frames. Matrix population models offer the opportunity to project population dynamics beyond the shorter-term effects of pathogen accumulation (Crone et al. 2011). Such models can compare population growth rates under different environmental conditions or disease conditions and identify life history stages with the largest impact on population growth. Data on each life history stage could be collected from field or glasshouse experiments where pathogens are manipulated and used to parameterize models (Parker 2000; Davelos & Jarosz 2004). Matrix population models would provide important information regarding the longer-term effects of pathogen accumulation on invasive hosts and resident native species, help determine whether spillback or PAID is occurring and inform management decisions for plant invasions. However, we should recognize that these models generally do not incorporate potential evolutionary changes in the pathogen, the invasive species or their interaction.

Conclusions

Given the biological impacts of invasive species, there is an urgent need for studies across diverse ecosystems that identify pathogens on invasive species and quantify their distribution, elucidate abiotic and biotic correlates of infection and evaluate the long-term ecological consequences for both invasive and native species. Pathogen accumulation may lead to the decline of invasive species and recovery of native species or to spillback of pathogens onto native species and increased ecological impacts of invasions. For example, if pathogen accumulation causes invasive decline, the passage of time may result in natural regulation of invasions and recovery of native plant communities. But, if spillback is common, then pathogen accumulation on invasive plants may suppress native species and promote invasions. Pathogens that attack and suppress invasive species (e.g. bioherbicides; Hallett 2005) also may be deliberately spread to help manage invasions once research is conducted to identify pathogens that do not infect co-occurring native species. On the other hand, if biological invasions result in spillback to native species (e.g. Mangla, Inderjit & Callaway 2008), more urgent control efforts may be warranted. Research efforts across diverse plant–pathogen systems and habitats focused on understanding pathogen and host population dynamics, the source of pathogens, and interactions among pathogens and invasive and native plants, will help elucidate the long-term consequences of pathogen accumulation.

Acknowledgements

We thank Wim van der Putten and two anonymous reviewers for comments on an earlier draft of this manuscript and acknowledge the past and current members of the Flory and Clay labs for helpful discussions. Funding was provided in part by the Center for Research in Environmental Sciences at Indiana University and the Agronomy Department and the Institute of Food and Agricultural Sciences at the University of Florida.