Trichoderma in Biological Control: a Taxonomist Reports

Some have estimated that fewer than 10% of the fungi have been described. Isn't it interesting that almost all of the reports of fungi used in the biological control of diseases caused by other fungi refer to one of about three species of Trichoderma, viz. T. harzianum, T. virens and T. viride? Why Trichoderma and why just these three?

Is Trichoderma a Super Hero of comic book proportions ('Mighty T, Super Fungus'), fighting an unceasing battle against the evil parasites that would deprive us of our daily... chocolate, for example? Will fungi in other groups give effective biological control? Will other species of Trichoderma give effective control of fungus-induced plant diseases? Are all reports based on correct identifications? Are species waiting in nature to be discovered? These are questions that I have been investigating as a research scientist in the Systematic Botany and Mycology Laboratory of the United States Department of Agriculture, Agricultural Research Service (USDA-ARS).

What is a Species of Trichoderma?

In a recent brief article, Christian Kubicek and his collaborators compared DNA sequences of biological control strains identified as T. harzianum. Of the eight strains that they studied, half were reidentified as either T. atroviride or T. asperellum. The low success rate in identifying Trichoderma strains to species is not very surprising. This is a genus that presents few morphological highlights: defining and identifying species is difficult. The 'up side' of the lack of diagnostic characters is that few species have been described. Between its first description early in the 19th century and 1984 no more than nine species were included in the genus Trichoderma. In 1939 G. R. Bisby could recognize only one species, T. viride, and for much of its life as a genus, the vast majority of reports of Trichoderma (primarily in ecology) referred to only that one species. The diversity of activities attributed to T. viride in the older literature is certainly extraordinary!

In 1969 Mein Rifai reviewed Trichoderma and proposed a taxonomy with nine 'species aggregates.' He said of most of these aggregates that they most likely comprised more than one morphologically indistinguishable species. While this was an advance over Bisby's 'single species' taxonomy (which made identification very easy!), it still left the observant user unfulfilled. In order to make an identification, species boundaries had to be very elastic indeed and, given that, there was no confidence that a species name used by different researchers actually referred to the same species. Between 1984 and 1991 John Bissett again took up Trichoderma taxonomy. Critical microscopic observations led him to recognize about 35 species, which he distributed among five sections. Obviously there is a great difference between nine species on the one hand, and 35 on the other. Which taxonomy best reflected the species in nature?

In the mid 1990s DNA sequence analysis was first applied to Trichoderma taxonomy. DNA sequences provided the much-needed independently derived data that would enable a better understanding of species of Trichoderma. A series of papers from Katrin Kuhls and Christian Kubicek and their collaborators established that Bissett's view of Trichoderma was a good approximation: there are certainly more than nine species of Trichoderma, although only some of his morphology-based groupings, or sections, are monophyletic.

Today, DNA sequence analysis is absolutely essential for the description and characterization of Trichoderma species. It is possible that Trichoderma is the only genus for which every species is represented in GenBank, the international database of DNA sequences, by at least one partially sequenced gene and many species are represented by sequences of two or more genes. In truth, it turned out that even Bissett's species concepts were not finely enough drawn. Some of the species that he recognized on the basis of their morphology, such as T. viride and T. koningii, have been divided among two or more species following DNA sequence analysis and a re-evaluation of their respective phenotypes.

Trichoderma entered the modern era of taxonomy (i.e. as defined by DNA sequence analysis) with a small number of species, all of which were represented by correctly identified cultures and at least one gene of each species has been sequenced. Because of this, the likelihood of proposing new species that are the same as previously described species is very low. At least Trichoderma taxonomists will not be accused of being 'name changers' by pathologists. This contrasts to genera such as Fusarium in which hundreds of species have been placed in the genus since the early 19th century; the finding of older names to replace familiar Fusarium names in the literature has been upsetting to some, however necessary the changes. DNA sequence analysis has permitted us to uncover misidentified species: many reports in the literature are based on misidentifications (see, for example, the article mentioned above from Kubicek). While this can lead to confusion in the literature, the end result is an accurately defined species. For example, T. aureoviride is often cited in the biological control and ecology literature. However, Elke Lieckfeldt and her collaborators found that despite these many reports, the species is only found in northern Europe (UK and The Netherlands) and is only known in cultures derived from the ascomycete Hypocrea aureoviridis. In another example, T. viride has long been known as THE species of Trichoderma that has globose and warted conidia and, as such, is easily identified. Lieckfeldt and her collaborators examined cultures identified as T. viride and found a correlation between DNA sequences and the type of warts, which were more or less conspicuous depending on the strain. They separated T. asperellum from T. viride. Currently, the soil fungus T. asperellum is being evaluated in various biological control applications from head-blight of wheat in Russia, caused by a Fusarium, to black pod disease of cocoa (Theobroma cacao) in Cameroon, caused by Phytophthora megakarya. Pierre Tondje and his collaborators in Cameroon believe that cellulase enzymes produced by the Trichoderma might be responsible for degradation of P. megakarya cell walls.

Trichoderma stromaticum, which is effective in control of witches' broom disease of cocoa caused by the mushroom Crinipellis perniciosa in South America, is the heart of the commercial biocontrol product TRICHOVAB®. This Trichoderma was reported first in the literature as T. viride and then later as T. polysporum. A combined study of morphology and DNA sequence analysis in our lab revealed it to be a new species. Today this species is being applied in areas affected by witches' broom in eastern Brazil, where it reduces inoculum of the pathogen. Indications are that it is becoming established in the area and that sexual reproduction, with possible genetic recombination, is taking place. Prakash Hebbar and Jorge T. de Souza, at USDA-Beltsville and Masterfoods, are investigating this.

The most commonly reported biocontrol Trichoderma is T. harzianum. However, this species was implicated as the cause of the green mould epidemic of commercially grown mushrooms in North America and Europe. The consequences of T. harzianum being a pathogen of such an economically important crop as mushrooms would have been disastrous to biological control. However we demonstrated here at USDA-Beltsville, through study of the morphology and cultural characters of T. harzianum and the mushroom parasite combined with use of DNA sequence analysis, that the mushroom parasite is a morphologically similar but phylogenetically distinct new species of Trichoderma, T. aggressivum. The mushroom parasite can be distinguished reliably from T. harzianum by its greatly diminished ability to grow at 35°C.

The bottom line is that one must view reports of identified species of Trichoderma with some scepticism. While Trichoderma species can be identified using the microscope and cultural characters, the most secure way for most people to identify a species of Trichoderma is through DNA sequences. Keys to the identification of most species of Trichoderma described up to the year 2000 are available in printed form and an interactive key is available at: http://nt.ars-grin.gov/taxadescriptions/keys/TrichodermaIndex.cfm

Can We Predict Biological Control Ability?

DNA sequence analysis has permitted us to 'see' the interrelationships of species through the formation of phylogenetic trees. Are phylogenetically homogeneous groups predictive of biological ability?

Christian Kubicek and collaborators found that relatives of T. reesei¸ a strain of which is the industry standard for cellulase production, tended to produce cellulase in higher concentrations than did species outside the T. reesei/longibrachiatum group. Members of this group are also able to grow and sporulate at 40°C and have been isolated from humans who have compromised immune systems: the warning is that the use of T. longibrachiatum strains in biological control should be considered very carefully with special regard for the possibility that those who prepare or apply the fungus are at high risk of inhaling it. However, in one case a strain of T. longibrachiatum that was reported to have biocontrol ability in Costa Rica was actually T. asperellum.

One of the first volatile antifungal compounds isolated from a Trichoderma species is 6-pentyl-α-pyrone (6PAP). This non-toxic compound has the distinctive coconut odour that characterizes T. viride and its relatives (Trichoderma sect. Trichoderma). Production of 6PAP has been attributed to T. harzianum, which is not a member of the viride group, but we have never had a culture of true T. harzianum that has the coconut odour. The only isolates in which we have noticed this odour are members of the T. viride group, in confirmation of an observation that was made in 1971 by C. Dennis and J. Webster. Reports from New Zealand that T. harzianum produces 6PAP are in fact based on T. atroviride, a member of Trichoderma sect. Trichoderma. As 6PAP can inhibit oospore formation in isolates of Phytophthora cinnamomi and inhibit conidial germination in Botrytis cinerea, an effort should be made to assay the many members of Trichoderma sect. Trichoderma for enhanced ability to produce 6PAP.

From these preliminary observations, it is possible to predict some biological activity based on phylogenetic relationships. The phylogeny of Trichoderma is being revealed at a great pace. From this phylogenetic framework it is reasonable to ask whether some kinds of biological activity, such as production of certain chitinases or the ability to direct parasitism of another fungus, are phylogenetically based. Unfortunately, there have been as yet no concerted efforts in this direction.

Where Are New Biocontrol Strains and Species Found?

Trichoderma species are found in almost all soils. They have been considered to be at least partially responsible for the control effect of 'suppressive soils', soils on which crops or trees are unaffected by a given pathogen. Control induced by suppressive composts has been attributed in part to elevated levels of some Trichoderma species. But species reported from these habitats are typically identified as one of the usual soil species, viz. T. harzianum or T. hamatum. Whether these species have been correctly identified is, of course, another question but it is 'reasonable' to expect to find these common species in such soils. Several new species of Trichoderma from eastern and southeast Asian soils have recently been described by John Bissett and his collaborators but the biocontrol potential of these new species was not assessed. The high number of new species found in the Asian soils by scientists who were looking for new species was surprising. Reports of Trichoderma species in soils come from all around the world - these can be seen by a quick search of the Internet - and one must wonder how many new (or incorrectly identified) species are found in the normal course of searching for biocontrol fungi in soils.

Another 'source' of new species of Trichoderma is the DNA-based phylogenetic analysis itself. As mentioned above, phylogenetic analysis of the morphological species T. viride revealed that it comprises two or more new species, including the biocontrol species T. asperellum. Similarly, T. koningii is representative of a common morphology in Trichoderma sect. Trichoderma that is shared by at least four species, of which three are undescribed. Were it not for DNA sequence analysis, all would have been made to fit the T. koningii morphological concept. In the light of the sequence analysis, though, T. koningii can be defined in a strict sense by the length/width ratio of its conidia and by its growth rate. One of these T. koningii morphological species is being evaluated for biological control use against frosty pod rot of cocoa, caused by Crinipellis roreri (formerly Moniliophthora roreri and Monilia roreri), in Ecuador but the same species has been found in soil in Brazil, and Germany, and from mushroom compost in Canada (Ontario). It has been isolated as an endophyte from trunks of Theobroma species in Brazil, Ecuador and Brazil and its Hypocrea sexual stage has been found in Cuba, Puerto Rico and the USA (Kentucky).

Phenotypic and genetic diversity in Trichoderma harzianum has led some to question whether it represents a single species. Priscila Chaverri and her collaborators examined a wide range of isolates using DNA sequences of four genes and found T. harzianum to be a species complex. While some consistent lineages had developed within the species, there were no consistent geographic, biological or phenotypic characters associated with any of the lineages. She rejected the hypothesis that T. harzianum comprises more than one species.

A previously unexplored niche is the tissue of healthy cocoa trees where Trichoderma species are found as endophytes. CABI Bioscience scientists Harry Evans and his collaborators are using 'classical biological control' techniques in searching for biocontrol agents in the area of origin of the crop and/or pathogen. The object is to find biocontrol agents that coevolved with the pathogen but that probably did not follow it to new areas. In the case of cocoa, they have isolated endophytic fungi from asymptomatic cocoa and cocoa relatives (Theobroma and Herrania species) in the upper Amazon region, which is where cocoa is thought to have evolved along with one of its major American pathogens Crinipellis perniciosa and in the western Andes where the second major pathogen, C. roreri, evolved on its original forest host T. gileri. Coevolved antagonists of Phytophthora megakarya have been sought in the Korup National Park of western Cameroon, where the pathogen is thought to have evolved in association with Cola, a relative of cocoa. We have also taken soil samples from the Korup forest in the hope of finding more effective strains of Trichoderma asperellum than are currently being used in Cameroon by Pierre Tondje. Keith Holmes and Harry Evans have isolated endophytes from trunks of Cola species in the Korup National Park. The Trichoderma species found as endophytes of cocoa include well-known species, such as T. harzianum, as well as a high proportion of new species. In Ecuador, they have found an endophytic strain of T. stromaticum, the species referred to above that is used in control of witches' broom. However, the endophytic strain from Ecuador is more effective, in in - vitro studies, than the one currently used in TRICHOVAB. The finding of Trichoderma - and other soil fungi such as Clonostachys rosea (formerly Gliocladium roseum, a species often used in biological control) - living endophytically within asymptomatic cocoa trees is surprising. The substantial endophyte literature (exclusive of that on the grass endophytes) emphasizes leaves of dicotyledonous plants and not stems, and the fungal endophytes of leaves tend to be species that, outside of the leaf, are found on decaying leaves or as twig, leaf and fruit inhabitants but not soil fungi. We are in the process of describing several new Trichoderma species that are endophytic within trunks of Theobroma species. Keith Holmes from CABI Bioscience was able to reinfect cocoa seedlings with some of them and he could reisolate many from the shoot meristem of the cocoa seedlings. At least one of them, a new species, inhibited radial growth of Crinipellis roreri in vitro. It also persisted on the surface, and within the tissues, of cocoa pods in the field for at least 10 weeks.

Why Trichoderma?

One thinks of Trichoderma in biological control simply because most of the fungi used in biological control of fungus-induced plant diseases are species of that genus. Trichoderma species are successful in biological control because they have shown variously an ability to parasitize pathogenic fungi, or out-compete pathogenic fungi for nutrients, or produce compounds that are toxic to pathogenic, or enzymes that lyse the cell walls of pathogenic fungi. They may also enhance plant growth and vigour, enabling the host plant to successfully defend itself against attack. Many aspects of Trichoderma biology and biological control have been reviewed in a two-volume publication edited by Christian Kubicek and Gary Harman.

Recently, Betsy Arnold, Allen Herre and their collaborators studied the leaf endophytes of cocoa in Panama. Typical of endophytes of leaves of dicotyledonous plants, they have isolated many fungi in leaves of cocoa. They have found that the endophytes might be adapted to cocoa in preference to other tree species. Some of these endophytes can be reinoculated into, and reisolated from, cocoa seedlings. None of the isolates is a Trichoderma species but some, unnamed species, can protect seedlings from infection by Phytophthora.

Stem and leaf endophytes of cocoa, respectively, are represented by ecologically and phylogenetically different fungi. Nonetheless, each ecological group offers exciting prospects for biological control applications not only in cocoa but other crops. Moreover, the interactions of endophytic fungi with their host plants may offer insights into how plants defend themselves against attack by fungal pathogens.

In Conclusion

Trichoderma species are effective in biological control of fungus-induced plant disease. A search of the Internet will show literally hundreds of examples. New species will be found as different niches are explored and also as the phylogenetic species concept comes into greater use, which is certain to happen as more people utilize DNA sequences and the incredible GenBank database. The study of endophytic fungi in stems and leaves from a biological control perspective, especially when combined with exploration in areas of diversity of hosts and their pathogens, holds the promise of finding new or more effective biocontrol agents and not just in the genus Trichoderma. The study of the interaction between host plants and their endophytes, especially at a molecular level, will certainly give new insights into the resistance of plants to diseases causing fungi. It's all very exciting at this point!

Selected Further Reading

Although this news section does not habitually include references, in this instance many of them are very recent and many readers are likely to clamour for them.

Arnold, A.E., Mejia, L.C., Kyllo, D., Rojas, E.I., Maynard, Z., Robbins, N. & Herre, E.A. (2003) Fungal endophytes limit pathogen damage in a tropical tree. Proceedings of the National Academy of Sciences, USA, 100: 15649-19654.

Bissett, J., Szakacs, G., Nolan, C. & Druzhinina, I. (2003) New species of Trichoderma from Asia. Canadian Journal of Botany 81: 570-586.

Chaverri, P., Castlebury, L.A., Samuels, G.J. & Geiser, D.M. (2003) Multilocus phylogenetic structure of Trichoderma harzianum/Hypocrea lixii complex. Molecular Phylogenetics and Evolution 27: 302-313.

Dennis, C. & Webster, J. (1971) Antagonistic properties of species-groups of Trichoderma. II. Production of volatile antibiotics. Transactions of the British Mycological Society 57: 41-48.

Evans, H.C., Holmes, K.A. & Thomas, S.E. (2003) Endophytes and mycoparasites associated with Theobroma gileri. Mycological Progress 2: 149-160.

Gams, W. & Bissett, J. (1998) Morphology and identification of Trichoderma. In: Trichoderma and Gliocladium Vol. 1. Basic biology, taxonomy and genetics, Kubicek, C.P. & Harman, G.E. (Eds.) Taylor & Francis, London, pp. 3-25.

Harman, G.E. & Kubicek, C.P. (Eds.) (1998) Trichoderma and Gliocladium. Vol. 2. Enzymes, biological control and commercial applications Taylor & Francis, London, 393 pp.

Kindermann, J., El-Ayouti, Y., Samuels, G.J. & Kubicek, C.P. (1998) Phylogeny of the genus Trichoderma based on sequence analysis of the internal transcribed spacer region 1 of the rDNA cluster. Fungal Genetics and Biology 24: 298-309.

Kubicek, C.P. & Harman, G.E. (Eds.) (1998) Trichoderma and Gliocladium. Vol. 1. Basic biology, taxonomy and genetics Taylor & Francis, London, 278 pp.

Kuhls, K., Lieckfeldt, E., Samuels, G.J., Börner, T., Meyer, W. & Kubicek, C.P. (1997) Revision of Trichoderma sect. Longibrachiatum including related teleomorphs based on analysis of ribosomal DNA internal transcribed spacer sequences. Mycologia 89: 442-460.

Kullnig-Gradinger, C.M., Szakacs, G., Kubicek, C.P. (2002) Phylogeny and evolution of the genus Trichoderma : a multigene approach. Mycological Research 106: 757-767.

Kullnig, C.M., Krupica, T., Woo, S.L., Mach, R.L., Rey, M., Lorito, M. & Kubicek, C.P. (2001) Confusion abounds over identities of Trichoderma biocontrol isolates. Mycological Research 105: 773-782.

Lieckfeldt, E., Kullnig, M., Kubicek, C.P., Samuels, G.J. & Börner, T. (2001) Trichoderma aureoviride : phylogenetic position and characterization. Mycological Research 105: 313-322.

Lieckfeldt, E., Samuels, G.J. & Nirenberg, H.I. (1999) A morphological and molecular perspective of Trichoderma viride : is it one or two species? Applied and Environmental Biology 65: 2418-2428.

Samuels, G.J., Dodd, S.L., Gams, W., Castlebury, L.A. & Petrini, O. (2002) Trichoderma species associated with the green mold epidemic of commercially grown Agaricus bisporus. Mycologia 94:146-170.

By: Gary J. Samuels, USDA-ARS, Systematic Botany and Mycology Lab., Rm. 304, B-011A, Beltsville, MD 20705, USA
Email:

❑

Origin of Florida Air Potato Disentangled

One of the most destructive invasive plants in Florida is air potato, Dioscorea bulbifera, a climbing member of the yam family, which ascends rapidly into tree canopies and smothers native vegetation. Air potato invades a variety of natural habitats, including hardwood forests and pine habitats. It is present in at least 23 Florida counties, extending from the northern panhandle to the southern peninsula, and is also a weed in Georgia and Alabama. Current control involves cutting the vine and pulling it clear of the underlying vegetation, and then removing all 'potatoes' (aerial bulbils) from the site - a laborious process - to prevent them from sprouting into new plants.

A more promising strategy for combating this invasive plant is classical biological control - the introduction into Florida of insect herbivores that feed on air potato in its native home range. Such biological control agents would have to be shown to feed only on the target plant and, thus, pose no threat to native or economically important plants. For air potato, this is a particularly important issue because there are two native yams in Florida: D. villosa and D. floridana. The screening for nontarget effects would be carried out in a highly secure quarantine facility.

The best biological control agents of air potato are likely to be found on plants that are genetically similar to those found in Florida. Air potato is not known to reproduce sexually in Florida, and thus has probably undergone little genetic change since arriving in the state. This means that it will still be genetically similar to the original source population in its native range, which would make that the first-choice location to survey for natural enemies. Unfortunately, until very recently this presented a stumbling block, as the source of Florida's air potato was unknown. What was known was that in 1905, the US Department of Agriculture (USDA) sent aerial bulbils of air potato to Henry Nehrling, a nurseryman based in Gotha, Orange Co., Florida, but the origin of these bulbils is unknown. It has been speculated that D. bulbifera was introduced into the USA on slave ships coming from Africa in the early days of the Atlantic slave trade. Yams store well, which made them ideal for long sea voyages. Nevertheless, whether this was the original (or only) pathway for introduction remained unclear.

The area of origin of the Florida material needed to be narrowed down (for both practical and economic reasons) before surveys could begin, as air potato is widely distributed in Asia and tropical Africa, where it occurs in wild populations and is a minor agricultural crop. The centre of origin of the genus Dioscorea is thought to be Asia, although a secondary centre of yam species diversity exists in West Africa. Inconveniently, from the point of view of the scientists trying to determine the provenance of the Florida material, D. bulbifera is the only member of the genus that occurs in the wild in both Asia and Africa, thus it could have been introduced into Florida from anywhere in this range.

Recently, the University of Florida (UF) and the University of Miami (UM) joined forces to solve the mystery. Previous work by Japanese scientists had shown that air potatoes from Asia and Africa are quite different at the molecular level, and can be readily distinguished by examining their chloroplast DNA. Using this technology, and with financial support from the Florida Exotic Pest Plant Council, the universities' scientists showed that air potato in Florida is most likely to be of African origin.

Now that mystery is solved, efforts are underway to explore for natural enemies. A trip to Africa in June 2003 identified the Crops Research Institute in Ghana and Makerere University in Uganda as collaborating institutions. Florida's Department of Environmental Protection recently awarded a grant to UF-IFAS (Institute for Food and Agricultural Science) and UM to support the work in Ghana and Uganda, and to allow further genetic studies to narrow the search in Africa to areas with genotypes similar to those found in Florida.

By: Bill Overholt (Indian River Research and Education Center, University of Florida) and Colin Hughes (Department of Biology, University of Miami)

Contact: Bill Overholt
Email:

---

Biocontrol of Aflatoxin Contamination of Groundnuts

The fruits of more than 20 years' research by US Department of Agriculture - Agricultural Research Service (USDA-ARS) scientists may provide a natural solution to a contamination problem affecting some of the world's major crops, including groundnuts and maize.

Groundnuts are a tasty, highly nutritious food enjoyed by people around the world, and it is one of the major cash crops of the southeastern USA. Nonetheless, aflatoxin contamination of groundnuts is a major concern from both food safety and economic points of view. Aflatoxin is a chemical produced by a fungus, Aspergillus flavus, which is naturally present in soils in which groundnuts are grown. Under certain environmental conditions (drought with high temperature, a not-uncommon combination in the southeastern USA), A. flavus can invade groundnuts as they develop in the soil and contaminate them with aflatoxin. Aflatoxin contamination can also occur during crop storage, while groundnuts are awaiting processing, if the right combination of temperature and moisture enables the fungus to grow.

Aflatoxin can be lethal in large doses, while prolonged exposure can cause liver disease and cancer, so groundnuts are closely monitored for its presence, and contaminated groundnuts must be diverted from the edible food supply. The US federal limit for food for human consumption is 20 p.p.b. (parts per billion), and the groundnut industry sets a limit of 15 p.p.b. It has been conservatively estimated that the cost of aflatoxin to the groundnut industry of the southeastern USA alone averages more than US$25 million per year.

There are many different strains of A. flavus, and not all have the ability to produce aflatoxin. However, while most of the strains that naturally occur in groundnut soils in the southeastern USA are toxigenic (i.e. have the ability to produce aflatoxin), the fact that nontoxigenic strains exist provided the key to developing a biological control solution. The strategy for reducing aflatoxin contamination of groundnuts developed at the USDA-ARS National Peanut Research Laboratory in Dawson, Georgia is based on competitive exclusion. A certain nontoxigenic strain of A. flavus was discovered that is highly competitive against other naturally occurring strains present in soil. When that strain is applied to the soil of the developing groundnut crop and conditions become favourable for A. flavus to invade the groundnuts, the nontoxigenic strain invades preferentially, and the result is a significant reduction in aflatoxin contamination, generally in the range of 70-90%. Studies have shown that not only is contamination reduced in the field, but a carryover effect also reduces contamination during storage.

Once the strategy had been shown to work, the next step was to develop an efficient application method, which led to the development of a unique formulation technique. Spores of the nontoxigenic A. flavus are coated onto the surface of a small grain, such as shelled barley. The barley serves both as a carrier to deliver the fungus to the field, using conventional farming equipment, and as a substrate for it to grow on after application. The coated barley is applied in a band over each planted groundnut row using conventional application equipment for granular materials. When the barley becomes moist in the field, the spores adhering to the surface germinate, grow, and produce more spores that become the competitive inoculum in the soil.

Circle One Global, Inc. (Cuthbert, Georgia) has licensed the ARS-developed technology and has given the trade name Afla-Guard to the product that it hopes to commercialize in the near future. The company is currently seeking registration of Afla-Guard as a biopesticide with the US Environmental Protection Agency.

The technology could have much wider application against other A. flavus -susceptible crops. It has shown promise for reducing aflatoxin contamination of maize already, and studies are planned to refine the methodology for this crop. Since there is no method currently available for preventing aflatoxin contamination of groundnuts and maize (except to control the environment through irrigation), the strategy has a lot to offer in combating a major agricultural problem in the USA. Further afield, aflatoxin contamination is a serious issue in the developing world, where contaminated crops may be eaten either in the absence of any alternative or through lack of awareness. The ease of use of this technology could make it suitable for resource-poor farmers in the developing world.

Contact: Joe W. Dorner,
USDA-ARS, National Peanut Research Laboratory,
P.O. Box 509, 1011 Forrester Dr.,
SE Dawson, GA 39842, USA.
Email:
Fax: +1 229 995 7416

---

Identifying Weak Links for OTA in Cocoa

Ochratoxin A (OTA) is a mycotoxin, a natural toxin produced by a number of fungal species in the genera Aspergillus and Penicillium. It is commonly associated with many food crops grown in semi-tropical and temperate climates, including cereals, vine fruits, coffee and cocoa. Growth of the fungus and production of the mycotoxin are dependent on factors such as temperature and humidity during crop growth, harvesting, drying, storage and shipment. Because it is stable to heat and other physical food processing, OTA can end up in food products made from these crops. However, recognition of the threat to health posed by this genotoxic carcinogen has led to limits being set.

CABI Bioscience (UK Centre), CABI-SEARC (South-East Asia Regional Centre) and ICCRI (Indonesian Coffee and Cocoa Research Institute) have recently been awarded funding by FAO (Food and Agriculture Organization of the UN) for a 2-year project to work on the management and identification of OTA-producing fungi in the cocoa commodity system in Indonesia. The project will apply a HACCP (Hazard Analysis Critical Control Point) approach to the whole production system.

OTA is of particular concern in cocoa, and especially in Indonesia, where there are increasing instances of rejection of cocoa beans by buyers because of unacceptable OTA levels. Data confirm that OTA is carried over from its source to the final retail product (i.e. chocolate). Sulawesi is the main cocoa-producing region of Indonesia, where the crop is predominately produced by smallholder farmers (see BNI 24(3) [September 2003] 55N-58N, Sulawesi setting for cocoa IPM initiatives) and exported to the USA. Generally the cocoa beans do not undergo on-farm fermentation but are sold directly to a middleman. The length of time the cocoa beans are stored before being sold on is highly variable. Thus there are many points in the supply chain where OTA contamination could occur.

Since OTA contamination can occur during any stage of the commodity system and current data are based entirely on examination of retail products, an epidemiological survey of OTA-producing fungi together with an analysis of OTA levels throughout the system is essential. The project will examine when the OTA-producing fungi start to appear, when they start producing OTA and if there is anything in the production system which augments OTA levels. In addition to the survey the project activities will include a technical questionnaire to collect data throughout the whole commodity system, i.e. harvesting, fermentation, drying, local storage, export, shipping and import storage. This information, combined with the results of the survey, will identify the critical sources of fungal and toxin production to determine where they are initiated and where OTA contamination occurs.

Other activities aimed at ensuring that the project has a sustainable impact include:

• Training local staff in sampling, isolation and identification techniques and data collection.
• Establishment and maintenance of a reference collection, an important part of the systematic development of a management process for cocoa bean production.
• Training-of-Trainers (TOT), covering participatory training techniques and communication skills, and how to prevent contamination of OTA at various points in the cocoa chain.

This is a timely project in that the European Union and other countries are likely to introduce regulatory limits for levels of OTA in imported cocoa from the end of 2004.

Contact: Zofia Lawrence,
CABI Bioscience UK Centre, Bakeham Lane,
Egham, Surrey TW20 9TY, UK.
Email: [email protected]
Fax: +44 1491 829100

---

Local Solution to African Armyworm

The African armyworm (Spodoptera exempta) is a major threat to food production in a large area of eastern and southern African. Outbreaks threaten the livelihoods of subsistence farmers, in particular, as the pest can inflict crop losses that represent a major portion of the families' annual food supply. However, the armyworm is now providing the means of its own destruction. A team of Tanzanian and UK scientists is developing cost-effective methods for harvesting diseased caterpillars to obtain naturally occurring S. exempta NPV (nucleopolyhedrovirus), and for multiplying and reapplying the NPV in the field.

The African armyworm appears between December and May as armies of black caterpillars covering many hectares with densities of up to 1100/m. As the crops grow, armyworms swarm across the countryside stripping the fields of grass, wheat, maize and other crops. Overall losses of 30% for crops have been estimated, but in major outbreak years losses of up to 92% in maize, a staple food in the region, are recorded. Outbreaks begin in Tanzania and are serious in 9 out of 10 years; in 2001 they covered 157,000 ha of crops and pasture. In major outbreak years the pest subsequently migrates to produce further extensive outbreaks in Kenya, Uganda, Ethiopia, Somalia, Mozambique, Zambia, Eritrea and may travel as far as Yemen.

Until now, the response to armyworm outbreaks has been to spray chemical insecticides; these are effective but are too expensive. In 2001, Tanzania could afford to control only some 30% of the outbreaks and in large outbreak years local stocks are exhausted early in the season. Poorer subsistence farmers, in particular, often cannot afford to purchase sufficient insecticide. Also weighting the balance against chemical insecticide use are growing concerns about their safety. Although future bans on older, cheaper chemicals on safety grounds may alleviate these concerns, their newer replacements, while less persistent and much less environmentally damaging, are much more expensive. An additional obstacle to chemical control is that many of the early armyworm outbreaks start in national parks, where widespread application of chemical insecticides is not desirable.

It has been recognized since the early 1960s that many armyworm outbreaks collapse late in the season as a result of a disease that kills up to 98% of the caterpillars. The disease-causing organism has been identified as a host-specific NPV. Late-season disease-associated population collapse is a recognized phenomenon in lepidopteran pests in a number of crops. Under natural conditions this does little to help the farmer as the rise in infection levels lags behind the growth of the pest population, and the disease spreads too slowly to prevent crop damage being inflicted. However, inundative application of NPV early in the season can cause sufficient early-season pest mortality to prevent populations reaching damaging levels.

In contrast to chemical insecticides, NPVs are highly selective, attacking only specific pest insects and posing few if any nontarget risks. They have been recommended by FAO (Food and Agriculture Organization of the UN) as pest control agents and a recent OECD (Organisation for Economic Co-operation and Development) report concluded that baculoviruses, which include the NPVs and closely related granuloviruses (GVs), are safe and do not present any health hazard. Several of these baculoviruses have been developed as commercial biological insecticides in the USA, Europe, China, Thailand and India. These are sprayed like chemicals onto pest outbreaks causing epidemics of disease that kill off the pests. They act as a natural insecticide and also the organisms multiply inside infected insects, which then serve as a source of disease inoculum for subsequent generations, thus the effect can be more persistent than that of chemical insecticides even though the persistence of the NPV particles outside the insect is generally shorter than that of many chemical insecticides.

One commonly suggested obstacle to biopesticide development is the cost of production (see BNI 24(4) [December 2003] 81N-82N, Harmonization sounds good for biopesticide business]. Production of NPVs currently relies on in vivo systems, which can make large-scale production of NPV products more expensive than the older, cheaper chemical insecticides, but NPVs are robust organisms and amenable to mass production, provided quality control issues are adequately addressed. Simple systems for producing and processing NPVs from infected insects have been developed in a number of countries. In India and Thailand locally produced NPVs are used to control a number of important pests such as Helicoverpa armigera, Spodoptera exigua and S. litura whose resistance to chemical pesticides has made them a serious problem for agriculture. In Brazil, 40 tonnes of NPV infected insects are harvested annually, largely by farmers. These are subsequently used to control the velvetbean caterpillar ( Anticarsia gemmatalis ) in soyabeans on over two million hectares at an average cost of US$1.26/ha.

The aim of the project in Tanzania funded by the Crop Protection Programme of the UK Department for International Development (DFID) is to develop a similar system for the armyworm NPV. A team led by Wilfred Mushobozi from the Tanzanian Ministry of Agriculture's Pest Control Services together with scientists from the Natural Resources Institute (University of Greenwich, UK), the NERC Centre for Ecology and Hydrology Oxford (UK), Stirling University (UK) and CAB International Africa Regional Centre (Kenya) has reported substantial progress following intensive studies to develop a low-cost, effective and safe armyworm control that can be produced in Tanzania. Field trials over the last 2 years in Tanzania have shown the armyworm NPV to be highly effective in controlling the pest if sprayed early in the outbreaks. The NPV reduced armyworm populations as much as chemical treatment (sumithion) and was more effective than a neem preparation.

The system adopted involves producing NPV in the field from natural armyworm outbreaks. Outbreaks on low-value pastureland are sprayed with NPV. Subsequently, the dying insects, by now full of NPV, are collected and processed to produce more NPV. This system is an adaptation of the field production techniques developed originally at EMBRAPA (Empresa Brasileira de Pesquisa Agropecuária) Brazil, and Dr Moscardi who leads the Brazilian team has helped train Tanzanians in the low-cost NPV production technology. NPV is highly productive and one armyworm infected by a thousand NPV particles will contain some 2 ∞ 109 particles by the time it dies. Recent research has indicated that larvae surviving viral challenge may carry a 'covert' viral infection with them as adults, which they may then pass on to their young. Under some circumstances, these covert viral infections may be triggered into lethal infections, which may infect other larvae and initiate viral epidemics at armyworm outbreaks beyond those that were initially sprayed. Thus, NPV potentially provides biocontrol beyond the originally targeted pest population. On-going studies are attempting to identify which cocktail of viral isolates best yields the most effective method of control.

One important future control strategy would be to strategically target the first outbreaks of the season in Tanzania, so as to minimize the number of insects migrating to initiate outbreaks in other parts of Tanzania and beyond. Predicting where these early outbreaks are going to occur, so NPV can be sprayed before pest populations begin to build up, is clearly key to successful control. Unlike chemicals, there is a time lag before NPVs begin to have an impact on the target pest. The NPV work is therefore progressing hand in hand with the development of a better armyworm forecasting system.

A successful centralized forecasting service has been operating in Tanzania based on a network of pheromone traps and rain gauges supplemented by satellite-derived rain cloud images. The pheromone traps specifically attract the armyworm male moths using an artificial female sex pheromone. However a problem with this centralized system is poor information flow to many intended recipients, especially poor subsistence farmers. To help overcome this, a community-based forecasting system is being developed. Village communities keep and operate their own pheromone trap and rain gauge, and use these to make their own forecast. Although some loss of forecasting accuracy might be expected compared to the national forecast, poor information flow ceases to be a problem: the forecast is generated by the people who need to act on it and this is in line with Tanzanian government policy to devolve responsibility for armyworm control to farmers.

This work builds on many years of armyworm studies by organizations such as the Desert Locust Control Organization for Eastern Africa, FAO and the old Centre for Overseas Pest Research (UK), which have provided a wealth of biological and population information that has been essential underpinning for efforts to develop improved forecasting and control of armyworm.

The first aerial applications of NPV are planned for Tanzania in early 2004 jointly funded by DFID and USDA-ARS (US Department of Agriculture - Agricultural Research Service) to confirm the efficacy of the NPV on a larger scale.

By Dave Grzywacz, Natural Resources Institute, Chatham Maritime, Kent ME4 4TB, UK
Email:
Fax: +44 1634 883379
Web: www.nri.org

Wilfred Mushobozi, Pest Control Services,
Ministry of Agriculture, PO Box 7433,
Arusha, Tanzania
Email:

---

Root Fungi Affect Leafminer Parasitoids

Increasing calls for scientists to predict the likely outcome of biocontrol releases make it important to be able to assess interactions between potential biocontrol species, and how they are likely to interact with organisms already present in the target environment. A recent study* is of interest because it tackled a previously unstudied relationship: parasitoids and soil organisms.

Interactions between fungi colonizing plant roots (arbuscular mycorrhizal (AM) fungi) and parasitoids attacking leafminers have been demonstrated in a common European meadow flower, ox-eye daisy (Leucantheum vulgare). A team of scientists from Royal Holloway College (University of London) and the University of Reading conducted studies in an experimental flower meadow using fungicide treatment to reduce soil fungal populations, and in controlled laboratory conditions using plants infected with one or more of three species of AM fungi and exposed to leafminers and parasitoids.

When experiments began in spring 1999, ox-eye daisy was the dominant plant in the meadow, which had been established 2 years previously, but it declined in density over the next 4 years. The results of field trials over this period showed that AM colonization was reduced by fungicide (iprodione) treatment (8.3% plants colonized compared with 36% for an untreated control), and treated plants were also smaller. Densities of the generalist leafminer Chromatomyia syngenesiae were highest in the first 2 years and then declined, but did not vary with treatment. Leafminer parasitism by the generalist parasitoid Diglyphus isaea fell over the 4 years of the experiment. However, in the first 2 years, when plant and leafminer densities were also highest, parasitism was lower in the untreated control where AM colonization was higher.

Laboratory experiments looked at colonization of ox-eye daisy roots by three AM species common in soil samples at the field site: Glomus caledonium, G. fasciculatum and G. mosseae (Gc, Gf and Gm, respectively). Complex interactions between species were apparent; for example, colonization by either Gc or Gf alone reduced ox-eye daisy leaf number, but dual colonization by both species together increased leaf number over uncolonized control plants. Single-species AM colonization reduced leafminer density, with Gm reducing it the most. However, the effects of multiple-species colonization varied was variable, with some leafminer densities higher and some lower than for the single species colonizations, and Gc and Gf together gave leafminer densities indistinguishable from the control. The impact of AM colonization on parasitism was even more variable: by themselves, Gc and Gf allowed higher and Gm lower rates of parasitism than an uncolonized control. Multiple colonization again gave variable results, but in general Gm decreased parasitism while Gc increased it. An interaction between these species was also evident: Gm 's antagonistic effect on parasitism was only evident in the absence of Gc.

The authors highlight the variation in the study, noting the changes over the course of the field study: parasitism was only significantly lower when the host plant and leafminer densities were highest. They suggest that multitrophic effects are related to the successional age of the community.

Three possible mechanisms (or a combination of them) could be involved in AM-mediated alterations in parasitism levels: plant size, plant chemistry and density-dependence between parasitoid and host insect. The authors suggest that the first two of these are important in the AM-parasitoid relationship:

• Decreased parasitism is correlated with the larger plant size of AM-colonized plants; the parasites have a larger leaf area to search for their leafminer hosts and are therefore less efficient.
• AM fungi have significant effects on the chemistry of host leaves, which may alter their attractiveness to parasitoids. Parasitoids react to plant volatiles emitted by herbivore-damaged plants, but AM-mediated reduced carbon availability may affect their production. Alternatively, they may (though not demonstrated) affect food quality by altering nitrogen availability.

The authors point out that the reasons for the variation found in this study need to be understood, "not just to enhance our knowledge of community structuring forces, but for the application of strategies involving natural enemies in pest control situations."

*Gange, A.C., Brown, V.K. & Aplin, D.M. (2003) Multitrophic links between arbuscular mycorrhizal fungi and insect parasitoids. Ecology Letters 6: 1051-55.

Contact: Alan C. Gange, School of Biological Science, Royal Holloway, University of London, Egham,
Surrey, TW20 0EX, UK
Email:

---

Code of Conduct Revision Proceeding

Several times last year, BNI flagged the imminent revision of International Standard for Phytosanitary Measures No. 3 (ISPM3), the Code of Conduct for the Import and Release of Exotic Biological Control Agents (e.g. see BNI 24(1) [March 2003] 1N-4N and 15N-27N). This process is now formally underway. An IPPC (International Plant Protection Convention) Expert Working Group meeting was held in Rome on 9-12 December 2003 and was comprised of eight biological control experts (scientists and regulators) from around the world.

The revised draft ISPM3 will be reviewed by the IPPC Standards Committee and then be released for country consultation, hopefully in 2004. It is during this country consultation phase that all interested parties can comment, but this needs to take place through the relevant National Plant Protection Organizations (NPPOs). The relevant contact people for the NPPOs can be found at: www.ippc.int/IPP/En/nppo.jsp
ISPMs under consultation will also be posted on this website.

Contact: IPPC Secretariat, FAO-AGPP,
Vialle delle Terme di Caracalla, 00100 Rome, Italy.
Email:
Fax: +39 06 5705 6347