How dangerous is the use of fungal biocontrol agents to nontarget organisms?


Biological control of plant pathogens is a method based on the antagonism between microorganisms (Andrews, 1992) – fungi or bacteria known to be antagonistic to a given plant pathogen are artificially multiplied and then released into an agricultural field to control a plant disease. Most biocontrol agents (BCAs) of plant diseases, similar to most plant pathogens they control, are fungi. Their use is considered, in general, as a safe and environmentally friendly alternative for plant disease control compared to the application of conventional pesticides (Whipps & Lumsden, 2001). Recently, Brimner & Boland (2003) published a review of the nontarget effects of fungal BCAs of plant pathogens in which they attempt to demonstrate the way in which many hazards may be associated with the use of fungi as BCAs of plant diseases. However, as the examples highlighted here indicate, their case was based mainly on unsubstantiated statements, which might mislead and be detrimental to the application of BCAs in the future.

Brimner & Boland (2003) use expressions such as ‘significant environmental impacts’, ‘significant threat’ and ‘unforeseen ecological repercussions’ in order to dramatize suggested damaging effects of fungal BCAs. However, none of the data reviewed in the paper support these serious warnings. Similarly, key statements such as ‘released BCAs have the potential to disrupt entire ecosystems through changes in the native soil community’ or ‘likely all fungi, including mycorrhizal fungi, that have cell walls composed of chitin, would be at risk for attack from G. virens’ are not supported by published data. The data cited show only that, in some cases, some fungi used as BCAs are antagonistic to fungi other than the target plant pathogens, or have negative effects on the host plants, but their nontarget activities are not correlated with any ‘significant environmental impact’.

In the case of Ampelomyces quisqualis, Brimner & Boland state that ‘… it may be possible for this BCA to attack nontarget fungal species and until its host range is identified, it is difficult to determine the risk to beneficial fungi and other soil organisms’. However, A. quisqualis, a highly specialized intracellular mycoparasite of powdery mildews, is one of the best known BCAs in terms of its host range (Falk et al., 1995; Kiss, 1998; Kiss et al., 2004) and, in addition, has nothing to do with soil organisms because its entire life cycle takes place on the aerial plant surfaces.

Brimner & Boland also state that ‘cultivated mushrooms are also at risk’ because it is known that some Trichoderma strains have been responsible for a disease of the commercially grown Agaricus while some other Trichoderma strains have long been used as commercial BCAs. However, the authors add that the disease-causing Trichoderma strains, recently described as belonging to a new species, T. aggressivum (Samuels et al., 2002), have never been used for biocontrol purposes. Therefore, the warning concerning the attack of commercially grown mushrooms by BCAs is unsupported.

Another example of an ‘undesired effect’ of a BCA is even more disturbing. Based on a long-term field experiment carried out by Gerlach et al. (1999), Brimner & Boland mention that the application of some Trichoderma strains against S. sclerotiorum result in the increase of the disease instead of achieving control. This data was used to demonstrate that BCAs may contribute to the spread of plant pathogens. However, the authors do not mention that this data came from only a single, small-scale experiment; Trichoderma strains have never been used as BCAs of S. sclerotiorum in an extensive way in plant protection practice; and Gerlach et al. (1999) excluded the treatments with Trichoderma from the 5-yr experiment 3 yr after its start because they showed no promise for practical control.

Some parts of the review raise conceptual problems. For example, Brimner & Boland write that ‘the most likely nontarget effect of BCAs is a reduction in the diversity and/or abundance of other fungi in an ecosystem’. However, fungal BCAs used for plant disease control have usually been applied in agricultural systems or in forestry, where the presence of other fungi has always been controlled to some extent by human activity. These locations should not be confused with natural ecosystems. Another conceptual problem is the statement that ‘several species of fungi commonly found in soils control plant diseases by preying on pathogenic soil microorganisms, such as nematodes (Jansson et al., 2000), pathogenic fungi (Foley & Deacon, 1986; Huang & Kokko, 1987; Falk et al., 1995, …’ Fungi parasitize, and do not ‘prey on’, microorganisms; nematodes are not microorganisms; and the paper written by Falk et al. (1995) and cited here deals exclusively with powdery mildews and A. quisqualis mycoparasites, which are not soil fungi.

Unfortunately, the Brimner & Boland review might be interpreted by decision makers as evidence of the harmful effects of BCAs and this might easily result in a more negative attitude towards their use in the plant protection practice. However, as the examples as selected show, there is good reason to mistrust the conclusions of the review. The ‘precautionary principle’ might be, at least to some extent, acceptable when little is known about the nontarget effects of a novel method. However, it is unreasonable to insist on ideas that are not backed up by well-founded data. It is well known that fungi may produce toxic metabolites and may parasitize each other. However, this is not a reason to stop using them to produce beer, wine, bread and antibiotics, or to control plant diseases, whenever a careful scientific investigation has shown that this is a feasible method and has no major nontarget effects.