Some of these successes are attributed to rusts – plant diseases caused by pathogenic fungi of the order Pucciniales. These include a rust (Puccinia chondrillina) released against skeleton weed (Chondrilla juncea) in Australia and North America, a rust (Puccinia myrsiphylli) released against bridal creeper (Asparagus asparagoides) in Australia, and the balloon vine rust (Puccinia arechavaletae) released against balloon vine (Cardiospermum grandiflorum) in the Cook Islands. Despite these successes, 60% of rust fungi intentionally released for biocontrol are reportedly having only a medium or variable impact on their target weed, and at least 15% of all rusts released have failed to establish at all, have established but had no impact on the target weed, or their impact is unknown or not documented.
Interactions between plant pathogens and fungal antagonists
There are many explanations and over 12 hypotheses to explain why fungal weed biocontrol agents fail to establish or are ineffective when they do. One of these, the ‘endophyte-enemy-release hypothesis’ (E-ERH) is modified from the enemy release hypothesis (ERH), which underpins weed biocontrol. The ERH states that invasive species dominate local species in novel environments because they don’t have their natural enemies that keep them in check in their region of origin. The E-ERH explains why the presence or absence of mutualistic endophytes can, in part, be responsible for the variable outcomes of classical weed biocontrol. Their presence increases plant fitness in the absence of co-evolved natural enemies, or their absence, coupled with the release from co-evolved natural enemies, contributes to increased plant fitness but leaves them highly vulnerable to classical biological control agents.
There is a diverse range of microorganisms likely to be associated with an invasive weed that may significantly affect the pathosystem. These include fungal endophytes that form part of the microbial community (or microbiome) and inhabit above- and below-ground tissues of all plants without causing visible infection or disease. Fungal endophytes affect plant ecology, fitness, and evolution, and shape plant communities. They are able to change the plant community structure and the diversity of associated organisms through increased fitness (by conferring abiotic and biotic stress tolerances, increased plant biomass, or decreased water consumption), or decreased fitness (by altering resource allocation).
It is important to understand not only the interaction between endophytes and host plants but also the interaction between endophytes and plant pathogens to determine their impact on the efficacy of classical biological control. Recent publications highlight how fungal endophytes interact with and affect classical fungal biocontrol agents of invasive weeds.
For example, interactions between endophytic fungi of the invasive weed Japanese knotweed (Fallopia japonica) and the rust fungus (Puccinia polygoni-amphibii var. tovariae) were studied in the native range of Japan to look for potential synergistic interactions. Pre-inoculation of the host plant with five endophytic fungi most frequently associated with Japanese knotweed gave varying results in terms of the number of rust pustules (raised masses of coloured spores that rupture epidermal leaf tissue) produced by the rust fungus. Two of the endophyte species (Alternaria sp. and Phoma sp.) reduced/suppressed the production of rust pustules, while two other species (Colletotrichum sp. and Pestalotiopsis sp.) were neutral, having no effect. The presence of a fifth species of endophytic fungus (Phomopsis sp.) increased the number of pustules produced by the rust, thereby increasing its potential as a biological control agent.
Similarly, variable disease severity of a rust fungus (Sclerotinia sclerotiorum) on Californian thistle (Cirsium arvense) led to the hypothesis that the variability was caused by the presence or absence of key endophytic assemblages. Using both culturing and molecular techniques, the researchers identified which endophytic fungi were present in the plants, and the amount of variation present within a plant and between plants at varying distances. The authors showed that endophytic fungi had a significant impact on the ability of S. sclerotiorum to cause disease on C. arvense and potentially influenced the success/failure of this biocontrol agent.
Clearly endophytic fungi can play a role in the success of fungal weed biocontrol agents, but there is another variable to be factored into the mix: whether other types of fungi could play a role in the success of rust fungi as weed biocontrol agents. Rust fungi have their own natural enemies, called ‘mycoparasites’. Mycoparasites are essentially fungi that parasitise other fungi. These mycoparasitic interactions form part of the microbiome of the plant and are considered a significant contributor to fungus–fungus antagonism. In fact, mycoparasitic interactions have been observed on several fungal biocontrol agents, either in the native range of the weed or in the introduced range, and two of these examples are well known to us.
A rust fungus (Puccinia araujiae) approved for release against moth plant (Araujia hortorum) was found to be heavily parasitised by another fungus (Cladosporium uredinicola) in the field in Argentina. Attempts to obtain a mycoparasite-free culture through superficial disinfection and multiple sequential inoculations (>8) in the laboratory were only partially successful. However, despite the high levels of mycoparasitism, testing for pathogenicity and host range were successfully completed, and the rust was approved for release by the EPA in 2015. The release has not yet been exercised due to delays with obtaining an export permit for the rust from Argentina.
The same mycoparasite (Cladosporium uredinicola) that parasitises the moth plant rust fungus in Argentina is present in New Zealand, associated with other native rust fungi. If we go ahead and release the moth plant rust, the question remains whether the mycoparasite could parasite the biocontrol agent and reduce its impact in the field. However, evidence from Argentina suggests it would be a successful agent since the rust is still damaging to its host plant there, causing heavy defoliation even in the presence of high levels of parasitism.
A rust fungus (Uromyces pencanus) recently imported into New Zealand as a potential biocontrol agent for Chilean needle grass (Nassella neesiana) was found to be associated with a mycoparasite (Simplicillium sp.) during pathogenicity and host range testing in Argentina. The mycoparasite was not obvious in the field but emerged in the glasshouse, which impeded the production of ‘clean’ rust spores to conduct the testing. Fortunately, our Argentinian collaborator, Dr Freda Anderson (CERZOS-CONICET), was able to produce clean rust cultures by storing them in the freezer, which killed off the mycoparasite but not the rust spores.
With all this evidence from weed biocontrol programmes worldwide, three of our researchers, Alana Den Breeyen, Claudia Lange, and Simon Fowler, recently conducted a review of how antagonistic fungi potentially affect fungal weed biocontrol programmes. Because the impact of fungal antagonists on the establishment and effectiveness of intentionally released fungal agents for invasive weed biocontrol is not well studied and often anecdotal, their review focused on how endophytic fungi and mycoparasites potentially reduce the effectiveness of classical biocontrol agents.
Plants, pathogens, and antagonists interact with each other in the environment, and an imbalance of these interactions can lead either to weed invasion or to successful weed control. In the native range the interactions are in balance and the plant is non-invasive. However, in the introduced range the plant is present as an introduced exotic species.
Three scenarios are discussed in the paper:
- The plant pathogen, introduced as a biocontrol agent, successfully suppresses the plant. Its effect is stronger than that of any present endophytes or mycoparasites.
- A protective fungal endophyte inhibits the plant pathogen. Biocontrol fails, and the plant remains an invasive weed.
- A mycoparasite inhibits the plant pathogen. Biocontrol fails, and the plant remains an invasive weed.
Five main challenges were identified from the literature and anecdotal evidence in terms of how the inadvertent introduction of naturally occurring fungal antagonists potentially contributes to the varying establishment and success of intentionally released fungal weed biocontrol agents: reduced infection pressure in the field, potentially affecting agent efficacy and the ability to keep the agent alive; reduced inoculum availability during spore production, affecting the ability to complete laboratory and glasshouse tests; reduced inoculum safety, due to inability to produce mycoparasite-free cultures for testing in the invaded range; reduced efficacy, due to potential accumulation of native mycoparasites in the invaded range; and reduced impact in the field, due to accumulation of the native pathogen.
The authors concluded that these naturally occurring endophytic fungi and mycoparasites may well contribute to the reduced success of intentionally introduced fungal biocontrol agents. A lack of actual evidence highlights the need for the collection and publication of plant-associated and mycoparasitic taxa. Investigations of the how the antagonists infect their fungal hosts, their host range, and their response to abiotic factors will ultimately improve our understanding of the interactions between the target plants, biocontrol pathogens, and potential antagonists that can disrupt successful biocontrol.
Funding
This project is funded by the Ministry of Business, Innovation and Employment as part of Manaaki Whenua – Landcare Research’s Beating Weeds Programme.