Feature

Taming a Wild Fungus

How do you find out if a naturally occurring fungus is a safe biological herbicide?

Graeme Bourdôt

When scientists in AgResearch first discovered that the fungus, Sclerotinia sclerotiorum, a common pathogen of Californian thistle (Cirsium arvense), had potential as a biological herbicide for controlling the weed in pasture, they were also acutely aware of the disease this fungus can cause in a large number of important crops.

Plant pathologists around the world have devoted much research effort to improve the management of this pathogen, and so the prospect of its use as a mycoherbicide was greeted, not surprisingly, with words of caution.

Used as a mycoherbicide, the vegetative (mycelium) phase of the fungus would be applied to the foliage of the thistle at a dose rate that would far exceed natural inoculum levels in thistle populations. Given suitable environmental conditions, the watery softrot disease caused by the fungus would develop and kill the thistle.

The problem is that once the pathogen has killed the thistle, it leaves behind perennating bodies called sclerotia. These fall out of the dead thistle and survive in the soil beneath the pasture for several years.

In the springtime these soilborne sclerotia may produce either mycelia (fungal threads) that directly infect a susceptible pasture plant, or tiny spore-producing structures called apothecia. Ascospores are formed in the apothecia and these are forcibly ejected into the air where wind currents may transport them out of the pasture to neighbouring susceptible crops where they in turn may germinate and cause disease.

So at a biocontrol site, the pathogen can disperse in both space and in time.

Avoiding risk completely by using a strain of the fungus that is effectively contained at the biocontrol site is one possibility. Auxotrophic strains, produced by UV mutagenesis of spores by a collaborating research team at Montana State University, US, have been tested in New Zealand. These strains must be supplied with the essential amino acid that their mutant genome cannot synthesise, to be able to grow and kill the thistle.

While these chemicals are readily supplied in a mycoherbicide formulation, they may be absent on many crops plants, providing immunity against the mutant ascospores. An elegant biotechnological solution? One catch is that these mutants have reduced pathogenicity. A second problem is that some susceptible crops have sufficient of the required amino acid on their foliage to fulfill the mutant's requirement.

Another way to avoid any risk is to use a sterile strain, unable to form sclerotia. The MSU team also produced one of these, but NZ tests have shown it to have only low virulence. Because the concept of sterility is very attractive, the AgResearch team has granted a PhD scholarship to a Lincoln University student to study the molecular biology of sclerotium formation in S. sclerotiorum. It is hoped that a single gene controlling sclerotium formation can be found, making it possible to engineer a sclerotiumless mutant that will be both highly pathogenic and genetically stable.

Because a biotechnological solution to risk may not be realised, the scientists in AgResearch are also assessing the disease risk to crops from the use of highly pathogenic wild strains of the fungus. S. sclerotiorum is a widespread and common pathogen in New Zealand and there is therefore always a natural risk of disease in any susceptible crop from airborne ascospores and soilborne sclerotia. This means that the appropriate measure of risk is the ratio of added inoculum (resulting from the biocontrol) to the naturally occurring inoculum. An acceptable value for this ratio may be 1.0 in regions where susceptible crops are common and where there is, as a result, a high natural concentration of ascospores in the air.

How Long will it Last?

In considering the temporal aspect of risk, the question the AgResearch scientists are seeking an answer to is how many years does it take for the elevated levels of soilborne sclerotia in pasture following thistle biocontrol, to decay back to some nominated fraction of the natural level?

Experimental data is being used to construct simple empirical models to predict future soil population densities of sclerotia following biocontrol of Californian thistle. While the studies are not yet completed, current estimates reveal annual decay rates of 80% per year. This is higher than expected from overseas studies on arable soils, and probably reflects the high soil microbial activity in high fertility sheep pastures.

How Far will it Go?

The spatial element of risk is in many regards much more complicated to quantify. This involves estimating the production of ascospores in the pasture, their escape and aerial dispersal. The question being asked is how far downwind from a thistle biocontrol pasture must you be to find an acceptably low ratio of added to natural ascospore concentration? Empirical studies in Canterbury have been unable to detect ascospores above natural levels beyond 5 metres of the edge of a treated pasture.

Courtesy of Dr Ian Harvey, Plantwise

However, on some occasions ascospores are likely to travel further and to estimate this, simulation modelling is being used. This is where the skills of a collaborating Dutch scientist, Dr Meindert de Jong, from the Department of Theoretical Production Ecology, Wageningen Agricultural University, NL, are crucial to the risk analysis of S. sclerotiorum. Dr de Jong is developing models to simulate ascospore production, escape and dispersal from pasture.

The models show firstly that not all of the spores released at soil level from apothecia escape above the pasture. The proportion that escape is dependent upon several interacting factors including windspeed, pasture height and pasture leaf area. Leaf area is important because S. sclerotiorum ascospores are sticky, and thus are easily trapped by pasture foliage. Simulations suggest that only a small fraction of ascospores ejected from the apothecia will escape from pastures which have the high foliage cover values typical of late spring, when ascospore production is at a peak.

Dispersal of escaping ascospores downwind is modelled by a Gaussian Plume Model, commonly used in the discipline of air pollution theory. This model assumes that the ascospores distribute in a Gaussian (or statistically "normal") manner in both the vertical and horizontal planes. The standard errors of these distributions depend on air turbulence, windspeed and distance downwind allowing calculation of the expected ascospore concentration at any given position downwind from the biocontrol site.

The de Jong escape/dispersal models are being validated by comparison of their output with US data on the spread of fungal spores from crops and with Canterbury data on the concentrations of S. sclerotiorum ascospores downwind of thistle stands treated with a mycoherbicidal preparation of the fungus.

Meaningful simulations of ascospore escape and dispersal from pastures will be possible after further study of key aspects of the life history of S. sclerotiorum have been quantified including:

• the annual progression of apothecial surface area in pasture
• daily spore flux from apothecia under field conditions
• seasonal progression of pasture leaf area on sheep and cattle farms in Canterbury

So by a process of experimentation in New Zealand and simulation in the Netherlands, the collaborating scientists will be able to quantify the risk of using this common disease-causing fungus as a biological herbicide for controlling Californian thistle in New Zealand pastures.

Then a risk evaluation, consisting of further research into risk reduction and risk perception with regard to the quantified risk, becomes an issue for the New Zealand Pesticides Board and the people of New Zealand at large. New Zealand society will decide for which cases and places the intended biocontrol of thistles may get an official approval.

Graeme Bourdôt works for AgResearch based in Lincoln, Canterbury.