BIOLOGICAL CONTROL OF PLANT PATHOGENS
---- Please CLICK on desired underlined categories [ to search for Subject Matter, depress Ctrl/F ]:
Root Diseases (mycoparasites)
Root Diseases (antagonists)
Root Diseases (Conservation)
The biological control of plant pathogens was detailed by Van Driesche & Bellows (1996). It involves the ecological management of a community of organisms. In the case of plant pathogens, however, there are two distinctions from biological control of organisms such as insects and plants. First, the ecological management occurs at the microbial level, typically in microcosms of the ecosystem such as leaf and root surfaces (Andrews 1992). Second, biological control agents include competitors, as well as parasites. While hyperparasites of plant pathogens and natural enemies of nematodes function in much the same way as do parasitoids, in arthropod systems (by destroying the pest organisms), competitors function by occupying and using resources in a nonpathogenic manner and in so doing exclude pathogenic organisms from colonizing plant tissues. Microbes which negatively affect pathogenic organisms are called antagonists.
Diseases of roots, stems, aerial plant surfaces, flowers, and fruit are caused by a wide variety of pathogens. Because of this diversity, the antagonist species, which negatively affect plant pathogens and the mechanisms by which they accomplish their beneficial action, are also quite varied. Their biological and taxonomic diversity is covered in some detail in several texts and reviews, including Cook and Baker (1983), Fokkema and van den Heuvel (1986), Campbell (1989), Adams (1990), and Stirling (1991). This section briefly introduces the antagonists of some important plant pathogens as representative of the broad taxa which are important in this field, beginning with agents affecting microbial pathogens of roots, and proceeding through pathogens of stems, leaves, flowers, and fruit. Natural enemies of plant parasitic nematodes are treated in the last section.
Root diseases are caused by a wide variety of fungi, and by some bacteria, in many crops and plant systems. Biological control agents recognized as significant in suppression of these diseases are largely antagonists that can occupy niches similar to the pathogens and either naturally or through manipulation out compete the pathogens in these niches. Antibiotic production is also important in a few cases, as are mycoparasitism and induced resistance. Streptomyces scabies, the causative organism of potato scab, is suppressed by naturally occurring populations of Bacillus subtilis, and saprotrophic Streptomyces sp.). Other microorganisms recognized as suppressing fungal diseases include species of Pseudomonas and Bacillus. Saprotrophic Fusarium fungi are able to suppress populations of pathogenic Fusarium spp. through competition for nutrients. There are few well-documented cases of induced resistance for soil-borne pathogens, and these are mostly of wilt diseases. Examples of organisms that induce resistance in plants to pathogens include nonpathogenic strains of Fusarium spp. Verticillium spp. & Gaeumannomyces spp. Mycoparasitic flora such as Anthrobotrys pp, Coniothyrium minitans Campbell and Sporidesmium scerotivorum Uecker et al., can be added to soil against fungal diseases. Bacillus spp. and especially Pseudomonas spp. are among bacteria that have properties particularly suited to effective suppression of root-infecting pathogens in soil, such as antibiotic production and competition for Fe3+ ions. Mycetophagous soil amoebae have also been noted feeding on pathogenic fungi. These amoebae generally require moist conditions in which to function, and may be important in the natural control of some fungi.
Diseases of plant stems produce symptoms which include decay and cankers on forest and orchard trees. and such wilts @ts I)utch elni disease and chestnut blight (caused by the fungus Cryphonectria parasitica (Murrill) Barr of Asian origin infecting the American chestnut. Castanea dentata [Marshaml Borkjauser). Because the etiologies of stem diseases vary, the taxa involved in biological control also vary. In many stem diseases, the pathogen colonizes a part of the host which initially is relatively free of microorganisms, such as a pruning wound. Successful biological control in such circumstances depends on rapidly colonizing this pristine environment with a nonpathogenic antagonistic competitor (VanDriesch & Bellows 1996). Primary among these are competitively antagonistic fungi, including saprotrophic members of the genera Fusarium, Cladosporium, Trichoderma, and Phanerochaete, and such antibiotic-producing bacteria as Bacillus subtilis and Agrobacterium spp. In the case of chestnut blight, hypovirulent strains of the pathogen itself are crucial in bringing about biological control. In this case, hypoviruience is transmitted cytoplasmically to virulent strains already infecting trees, and disease symptoms decline and disappear (VanDriesch & Bellows 1996).
The growth of microorganisms on leaves is normally severely restricted by environmental factors. Nutrient levels generally are low on leaf surfaces, and microclimate variables, especially leaf surface moisture, temperature, and irradiation, are often unfavorable for microbial development. In temperate climates and arid tropical regions, water will be intermittent on leaf surfaces, but may be continually present in humid tropical regions. Temperatures on leaf surfaces exposed to direct radiation may rise to several degrees above ambient, The result of such variation is that microbial floral development on leaf surfaces varies from general scarcity in temperate climates to more extensive microbial films in tropical rain forests (Campbell 1989).
Microbes that most frequently are recorded as saprotrophs on surfaces of crop plants in temperate conditions and, therefore, the species which are candidates as antagonists of pathogens, include the fungi Aureobasid m pullulans (de Bary) Arnaud, Cladosporium spp., and such yeasts as Cr.yptococcus spp. and Sporobolomyces spp. Beneficial bacteria in the phyllosphere include members of such genera as Erwinia, Pseudomonas, Xanthomonas, Chromobacterium and Klebsiella. These lists,based on microbial surveys, usually give no indication of activity of the organisms, but this information can be obtained from experimental studies. For example, early studies on control of botrytis rot in lettuce (Wood 1951) indicated that several organisms were successful in suppressing the disease when sprayed on lettuce (Lactuca sativa L.) plants, among them Pseudomonas sp., Streptomyces sp., Trichoderma viride Persoon: Fries, and Fusarium sp. Similar studies show varying degrees of effectiveness in other cropping systems (Peng & Sutton 1991; Sutton & Peng 1993a,b; Zhang et al. 1994). The microbial composition and biological activity of phylloplane microbes can vary with season, position on the top or bottom of the leaf and on location in the plant canopy, depending on the degree of exposure relative to prevailing winds and rain (Campbell 1989).
Biological control of the black-crust pathogen (PhyIlacbora huberi Hennings) on rubber tree (Hevea brasiliensis Müller Argoviensis) foliage is accomplished by the hyperparasites Cylindrosporium concentricum Greville and Dicyma pulvinata (Berkeley & Curtis) Arx (Junqueira & Gasparotto l991). Botrytis leaf spot in onion (Allium cepa L.) was suppressed by Gliocladium roseum Link: Bainier (Sutton and Peng 1993a). Other examples include control of powdery mildews, other botrytis rots, and turfgrass diseases (Sutton and Peng 1993a).
Nonpathogenic species of the fungal genus Colletot richum 1981; Dean & Kuc 1986) can be used to induce resistance in cucumbers against pathogenic species of the same genus. Inoculation with a nonpathogenic strain of a virus confers protection to plants from pathogenic strains in many diseases. The bacterium Bdellovibrio bacteriovorus Stolp & Starr is aparasite of pathogenic bacteria. Finally, there are numerous parasitic fungi that attack pathogenic fungi (Kranz 1981). Among those that have been studied in detail, principally as agents against leaf rusts and mildews, are Spbaerellopsis filum (Bivona-Bernardi ex Fries) Sutton, Verticillium lecanii (Zimmerman) Viegas, and Ampelomyces quisqualis Cesati ex Schlechtendal (VanDriesch & Bellows 1996).
Flower & Fruit Pathogens
Flowers are ephemeral structures and as such have limited opportunity to become infected. One major disease of flowers which has received attention is fire blight of rosaceous plants, caused by the bacterium Erwinia amylovora (Burril) Winslow et al. Biological suppression of the disease has been achieved through use of the nonpathogenic species Erwinia herbicola (Lohnis) Dye (Beer et al. 1984; Lindow 1985b), sometimes in combination with Pseudomonas syringae van Hall. rwinia herbicola was used successful by spraying aqueous suspensions of it onto the flowers just before the time of potential infection (Campbell 1989). The mode of action is primarily competitive exclusion, with the antagonist competing with the pathogen for a growth limiting resource and possibly other effects such as induced cessation of nectar secretion or accumulation of a host toxin (Wilson and Lindow 1993a).
The diseases attacked through biological control include diseases of fruit on the plant and post-harvest diseases. One of the first systems developed was against Botrytis cinerea Persoon Fries in vineyards, where sprays with spore suspensions of the antagonist Ttrichoderma barzianum Rifai were effective in suppressing disease incidence. Several organisms, including Gliocladium roseum, Penicillium sp., Trichoderma viride, and Colletotrichum gloeosporioides were as effective as fungicides in suppressing B. cinerea on strawberries (Peng and Sutton 1991). A number of other examples also have been reported (Sutton and Peng 1993a).
Post-harvest diseases, which can be responsible for 10-50% loss of produce (Wilson and Wisniewski 1989; Jeffries and Jeger 1990), have received considerable attention. Numerous reports deal with suppression of post-harvest disease in fruit crops (Campbell 1989; Wilson and Wisniewski 1989; Jeffries and Jeger 1990) by such organisms as species of Penicillium, Bacillus, Trichoderma, Debaryomyces, and Pseudomonas. The mode of action of many of these is generally antagonism, often through the production of antibiotics, which reduce the longevity, and germination of spores of pathogens. Others appear to suppress pathogen growth through nutritional competition or induction of host resistance (Wilson and Wisniewski 1989). Postharvest rots include major diseases caused by Botrytis cinerea, Rhizopus spp., and other fungi in several crops. Competitive and parasitic fungi, including Tiichoderma spp., Cladosporium herbarum (Persoon: Fries) Link and Penicillium spp., give control as good as commercial fungicides. Enterobacter cloacae (Jordan) Hormaeche and Eduards reduces rots liy Rhizopus spp., but there are restrictions in its use on uncooked food products (Van Driesche & Bellows 1996).
Plant-parasitic nematodes inhabit many soils and attack the roots of plants. They are affected by a range of natural enemies, including bacteria, nematophagous fungi, and predacious nematodes and arthropods. There is some limited evidence for virus association with nematodes (Loewenberg et al. 1959), but the etiology of these viruses is not well known (Stirling 1991). The biologies of natural enemies of nematodes were reviewed by Sayre and Walter (1991) and Stirling (1991).
Bacteria That Affect Plant-Parasitic Nematodes
A few bacterial diseases of nematodes have been reported (Saxena and Mukerji 1988); other bacteria produce compounds that are detrimental to plant-parasitic nematodes (Stirling 1991). The most widely studied of the bacterial pathogens of nematodes are in the genus Pasteuria. Early work was focused on Pasteuria penetrans (Thorne) Starr and Sayre. Recent evidence indicates that this taxon represents an assemblage of numerous pathotypes and morphotypes, and probably represents several taxa (Starr and Sayre 1988). This bacterium has been found infecting a large number of nematode species (more than 200 in about 100 genera, Sayre and Starr 1988; Stirling 1991), does not attack other soil organisms, and is the most specific obligate parasite of nematodes known. Its spores attach to and penetrate the nematode cuticle. Most attention has been centered on populations (Pasteuria penetrans sensu stncto, Start and Sayre 1988) that attack root-knot nematodes (Meloidogyne spp.). The spores of P. penetrans germinate a few days after a contaminated nematode begins feeding on a root (Sayre and Wergin 1977). The bacterium reproduces throughout the entire female body, and the female may either be killed or may mature but produce no eggs. Bacterial spores (about 2 million from each infected nematode, Mankau 1975) are released when the nematode body decomposes, and they remain free in the soil until contacted by another nematode. They tolerate dry conditions and a wide range of temperatures, and may remain viable in the soil for more than six months. Because it is an obligate parasite, it has not yet been possible to develop in vitro culturing techniques for this bacterium. Different populations of the bacterium show varying degrees of specificity to small numbers of nematode species, but the mechanisms and degree of specificity remain to be elucidated (Stirling 1991). Pasteuria penetrans appears responsible for some cases of natural regulation of nematode populations (Sayre and Walter 1991).
Some strains of Bacillus tburingiensis are also known to have activity against nematodes, including plant-parasitic species. Zuckerman et al. (1993) report efficacy of a strain against Meloidogyne incognita (Kofoid and White) Chitwood, Ratylencbus reniformis Linford and Oliveira, and Pratylenchus penetrans Cobb in field and glasshouse trials. The body openings of these nematodes are too small to permit the ingestion or other ingress of the bacterium, and Zuckerman et al. (1993) suggest that the mode of action is either a beta exotoxin (Prasad et al. 1972; Ignoffo and Dropkin 1977) or a delta endotoxin released following bacterial cell lysis. A strain of B. thuringiensis with a nematotoxic delta endotoxin is the subject of a European Patent Application by Mycogen Corporation of San Diego, California (Zucherman et al. 1993).
Fungi That Affect Plant-Parasitic Nematodes
Many fungi attack nematodes in the soil (Barron 1977; Stirling 1991). Numerous species have been reported from all types of soils. The taxonomy of the group has been subject to revision, and the generic names recognized in Stirling (1991) are used here (Van Driesche & Bellows 1996).
Some nematophagous fungi are endoparasitic in nematodes. Among these are genera which reproduce through motile zoospores (e.g., Catenaria anguillulae Sorokin, Lagenidium caudatum Barron, Aphanom.yces sp.), which generally appear only weakly pathogenic in healthy nematodes (Stirling 1991). Other endoparisitic fungi possess adhesive conidia, and the infection process begins when conidia adhere to a nematode's cuticle (e.g., the genera Vellicillium, Drechmeria,Hirsutella, Nematoctonuss). In Nematoctonus spp.,the germinating spores secrete a nematotoxic compound which causes rapid immobilization and death of nematodes (Giuma et al. 1973). A few species (Catenaria auxila [Kuhn] Tribe, Nematophthora gynophila Kerry and Crump) parasitize adult females or nematode eggs rather than juveniles.
Other fungi capture nematodes through use of special trapping stmctures, and have been termed "predatory." Among the more common of these fungi are species in such genera as ,Monacrosporium, Arthrobotrys, and Nematoctonus. These fungi consist of a sparse mycelium, modified to form organs capable of capturing nematodes. These organs include adhesive structures, such as adhesive hyphae, branches, knobs, or nets (Stirling 1991). There are also nonadhesive rings, the cells of which expand when touched on their inner surface, constricting the interior of the ring and trapping nematodes. Most of these fungi are not specific and attack a wide range of nematode species. They are widely distributed (Gray 1987, 1988) and most are capable of saprotrophic growth, but often appear limited in this phase in the soil. Many soils suppress the growth of these fungi (a condition called soil fungistasis or mycostasis). This is possibly due to two different causes. Mankau (1962) concluded that a water-diffusible substance was responsible for inhibited germination in tests of soil from southern California (U.S.A.). Other studies have indicated increased activity following soil amendments with nutrients (Olthof and Esrey 1966) or organic material (Cooke 1968), which implies fungistasis may be a result of resource limitation. Following saprotrophic growth, formation of trapping structures occurs which is apparently stimulated by nematodes (Nordbring-Hertz 1973; Janssen & Nordbring Hertz 1980). Stirling (1991) suggests that this phase of predacious activity is followed by diversion of resources to reproduction, followed by a relatively dormant phase (Van Driesche & Bellows 1996).
Other fungi are facultatively parasitic on nematodes. Of the few of these fungi that are significant pathogens of root knot and cyst nematodes, Verticillium spp, are among the most important. These fungi can parasitize nematode eggs, and Verticillium chlamydosporium Goddard plays a major role in limiting multiplication of Heterodera avenae Wollenweber in English cereal fields (Kerry et at. 1982a,b). Paecilomyces lilacinus (Thom) Samson parasitizes eggs of Meloidogyne incognita (Jatala et al. 1979) and Heterodera zeae Koshy, Swarup, and Sethi (Dunn 1983; Godoy et al. 1983). Dactylella oviparasitica Stirling and Mankau, a parasite of Meloidogyne eggs, is thought to be at least partially responsible for the natural decline of root-knot nematodes in Californian peach orchards (Stirling et al. 1979).
Predacious Nematodes That Affect Plant-Parasitic Nematodes
Predatory nematodes are found in four main taxonomic groups: Monochilidae, Dorylaimidae, Aphelenchidae and Diplogasteridae. Each possesses a distinct feeding mechanism and food preferences (Stirling 1991). The monochilids have a large buccal cavity that bears a large dorsal tooth; all species are precdacious, feeding on protozoa, nematodes, rotifers, and other prey, which may be swallowed whole, or pierced and the body contents removed. The dorylaimidss are typically larger than their prey and possess a hollow spear which is used either to pierce the body of the prey or to inject enzymes into the food source and suck out the predigested contents. The group is considered omnivorous. but the feeding habits are known only for a few species (Ferris and Ferris 1989). Almost all the predatory aphelenchids are in the genus Seinura. Although small, they can feed on nematodes larger than themselves by injecting the prey with a rapidly paralyzing toxin through their stylet. The diplogasterids, typically a bacteria-feeding group, have a stoma armed with teeth, and the species with large teeth prey on other nematodes. Species in all these groups are generally omnivorous, feeding on free-living as well as plant parasitic nematodes. The role of individual species in the population dynamics of plant parasitic nematodes in the soil has been difficult to quantify, but it is possible that a number of species may act together to produce a significant impact (Stirling 1991).
Insects and Mites
Several microanthropods in the soil, including mites and Collembola, prey on nematodes, and high predation rates have been recorded in vitro (Stirling 1991). A few genera are obligate predators of nematodes, while other genera are more general feeders and consume nematodes as well as other foods (Moore et al. 1988; Walter et al. 1988; Sayre and Walter 1991). The information available suggests that as a group, microanhropods are probably significant predators on nematodes in some soils and habitats. However, limited information about predation rates in soil is available, and more work is required to assess the impact of this group on nematode populations.
VanDriesch & Bellows (1996) concluded that this overview touched briefly on groups of organisms which are antagonistic to plant pathogens and nematodes. These antagonists vary both in their innate ability to suppress plant pathogens and in their ability to thrive and compete in different environments. Consequently the selection of an organism or organisms for any particular biological control program is a compromise among these parameters and abilities. In addition, the selection of organisms depends on the approach taken for their use (inoculative augmentation, inundative augmentation, or natural control through conservation.
Organisms for biological control of plant disease can be used in various ways, but most attention has been given to their conservation and augmentation in a particular environment, rather than to the importation and addition of new species as is often done for insect or weed control. The choice of these approaches is in part because there is usually a diverse set of microbes already associated with plants. These microbes provide substantial opportunity for development of resident species as competitors or antagonists to pathogenic organisms. Both conservation and augmentation have some application in each of the main groups of plant diseases. The use of microbes for control of plant pathogens is covered in more detail in several texts, including Cook and Baker (1983), Parker et al. (1983), Fokkema and van den Heuvel (1986), Lynch (1987), Campbell (1989), and Stirling (1991) and in other review articles (Wilson and Wisniewski 1989; Adams 1990; Jeffries and jeger 1990; Sayre and Walter 1991; Andrews 1 92; Cook 1993; Sutton and Peng 1993a).
Plant pathogens are attacked with biological control through conservation is accomplished either by preserving existing microbes which attack or compete with pathogens or by enhancing conditions for their survival and reproduction at the expense of pathogenic organisms. Conservation is applicable in situations where microorganisms important in limiting disease causing organisms already occur, primarily in the soil and plant residues but in some cases also on leaf surfaces. They may be conserved by avoiding practices which negatively affect them (such as soil treatments with fungicides). The soil environment may be enhanced for some beneficial organisms through adding organic matter, such as soil amendments (Van Driesche & Bellows 1996).
Biological control of plant pathogens through augmentation is based on mass culturing antagonistic species and adding them to the cropping system. In the context of the examples discussed in this text, this is augmentation of natural enemy populations, because the organisms used are usually present in the system, but at lower numbers or in locations different than desired. The purpose of augmentation is to increase the numbers or modify the distribution of the antagonists in the system. In some cases, such organisms are taken from one habitat (for example the soil) and augmented in another (for example the phyllosphere). Tire activity of augmenting microbial agents is sometimes termed "introduction" in the plant pathology literature, in the sense of "adding@ them to the system (Andrews 1992; Cook 1993). However, he organisms introduced are usually found in a local ecosystem and are not introduced from another region of the world.
Augmentation of antagonists naturally involves two approaches. The first is direct augmentation, at potential infection sites or zones, with organisms antagonistic or parasitic to the pathogens themselves. In this approach, the antagonist population is directly responsible for disease suppression. A second approach is to inoculate plants with nonpathogenic organisms that prompt general plant defenses against infection by pathogens (induced resistance). Disease control is then achieved through greater plant resistance to infection.
Substantial work has been done to characterize the role of microorganisms in biological control of plant diseases. The biological mechanisms underlying the success of these antagonists in such settings may include initial competition for occupancy of inoculation sites, competition for limiting nutrients or minerals, antibiotic production, and parasitism (Van Driesche & Bellows 1996).
Understanding the principles that apply to biological control of plant pathogens, the ecology of the system is considered at the level of the pathogens and the agents used for control. Aerial plant surfaces, usually present hostile environments to colonizing microbes, in many cases consisting of surfaces protected by cuticular waxes, with very small amounts of nutrients available on these surfaces. Further, surfaces of the aboveground portions of plants may be dry. Consequently, pathogenic microbes attempting to colonize these surfaces may face a number of difficulties, including competition with other, nonpathogenic, microbes.
The rliizosphere (the roots and the region immediately adjacent) is a somewhat richer environment than the phyllosphere because of simple sugars, amino acids, and other materials exuded by the roots, but in the remainder of the soil the growth of microbes is often carbon limited (Campbell 1989), Moisture in the rhizosphere may be more continuous in time and space than on the above ground surfaces of plants (the phylloplane), but the rhizosphere may be subject to periodic drying.
Some forms of competition in these environments are important to the ability of any particular organism to increase in numbers and consequently to reduce the numbers or activity of other organisms, including plant pathogens (Campbell 1989; Andrews 1992). Microbial competition can be important at two main stages of growth of pathogen populations. First, there may he competition during initial establishment on a fresh resource that was not previously colonized by microorganisms. Second, after initial establishment, there is further competition to secure enough of the limited resources present to permit survival and eventual reproduction. Microorganisms show many traits that may characterize them as particularly adept at either the colonization phase or subsequent phases of competition. Species referred to as r strategists (ruderal species) have a high reproductive capacity. These species produce so many spores or reproductive bodies that there is a high likelihood that some will be found near any newly available resource. These species are effectively dispersed and establish readily in disturbed habitats or in the presence of noncolonized resources. They are found in disturbed settings where easily decomposable organic matter or root exudates are found, and where initial resource capture is crucial for survival. In contrast to these r-strategists, species found in more stable situations face competition for space and limited resources (Begon et al. 1986). These organisms, termed K-strategists, become more dominant as a community matures and becomes more crowded. These concepts form the endpoints of a continuum, and there are varying degrees of r- and K-related characteristics in different microbes in various habitats (see Andrews and Harris [19851 for further discussion on these concepts in microbial ecology).
Plant pathogens are spread across this r-K range of characteristics and vary in other important biological characteristics (Van Driesche & Bellows 1996). There are opportunistic pathogens that are able to attack young, weakened, or predisposed plants, but may be poor competitors (Botrytis, Pythium, Rbizoctonia). There are pathogens that tolerate environmental stresses. These organisms often live in situations with few competitors, because few species are able to exist in such environments. Some pathogens, such as the Penicillium species that cause postharvest rots, produce antibiotics that inhibit competitors. Other species (such as Fusarium culmorum [Smith] Saccardo) have a very high competitive ability. It is important to understand the ecology of a target pathogen before one can effectively consider what biological control strategy might be most effective. Stress-tolerant and competitive species, for example,require different biological control strategies and agents than ruderal ones.
Similar to the way that antagonists of plant pathogens vary in r-K and other characteristics, the properties of an effective biological control agent will depend on the setting in which it is intended to function, in many agricultural settings, disturbance makes new resources available to microbes through crop residue burial, cultivation, or planting. A frequent need, therefore, is a control agent that has the characteristics of an r-strategist (Campbell 1989), which can grow quickly and colonize new resources rapidly, with minimal nutrient and environmental restrictions. It should function well in disturbed environments and have some means (such as spores) of surviving in the soil or on the plant near to the pathogen inoculum or the Source or site of infection. Biological control agents that are r-strategists are an approximate equivalent of a protectant fungicide, being in place before the pathogen infection cycle can begin. In other programs, such as those directed against a pathogen which has already invaded the plant host, a more competitive species will be required. Finally, a biological control agent may have to be tolerant of abiotic stresses, particularly for use in dry climates or on leaves.
Although there is much variation in soil types in different locations, soils are typically rich in microflora, with propagules numbering in the hundreds of thousands per gram of soil (Campbell 1989). In most soils, growth of microorganisms is carbon-limited. either because what carbon is available is not physically accessible or because the microbes do not possess the enzymes necessary to degrade the carbon-containing molecules that are present. An exception to this general limitation is the region immediately surrounding plant roots. This region, the rhizosphere, contains easily metabolized carbon and nitrogen sources such as amino acids, simple sugars, and other compounds exuded by the roots. Consequently, this region is more favorable than surrounding soil for the support of microflora. Root pathogens and plantparasitic nematodes may be found growing on or in roots, but many microbes in the soil will be dormant because of resource limitations. Because there are many dormant organisms in the soil prepared to take advantage of any favorable period or opportunity, competition for resources in the soil may be significant and may limit the ability to augment beneficial organisms and have them flourish, unless soils are first sterilized to eliminate potential competitors. Therefore, much research surrounding biological control of root diseases and nematodes has centered around identifying soils which are naturally suppressive to particular disease organisms and investigating the microbial components of the soil responsible for the suppression. Management of such antagonistic organisms for biological control can range from treatment of soil to favor the desirable organisms (conservation) through inoculation of soils or plants with specific beneficial microorganisms (augmentation) (Van Driesche & Bellows 1996).
The phyllosphere is significantly different from the rhizosphere in its structure, ecology, nutrient availability, and exposure to climatic factors (Andrews 1992). Leaves are relatively hostile to microorganisms. They are generally hydrophobic and covered with chitin and wax, which limits the amount of exudate (and hence nutrients) that reaches the leaf surface. These and other factors impose severe environmental restrictions to microbial growth on leaf surfaces. Fungal pathogens of leaves often enter the leaf tissue very shortly after germination of the pathogen and, consequently, are protected inside the plant for much of their growth. Bacterial pathogens may multiply on the leaf surface before invading leaf tissues. Biological control of disease can take place either through general inhibition and competition on the leaf surface prior to invasion of leaf tissues or through suppression of the disease after the pathogen has invaded. Biological control within leaf tissues can occur through one of several mechanisms, including induced resistance in the plant and hyperparasitism of the pathogen. Woody stems are habitats low in nutrients and often difficult for pathogens to penetrate, Because the wood itself supports very few saprotrophic microorganisms, pathogens colonizing the wood through wounds, dead branches, or roots find very few competitors. Because there are few organisms present to conserve, protection of the wood from these decay organisms can be achieved by protecting the relatively small, well-defined wound or branch stub through inoculation (augmentation) with specific microorganisms. These wounds are initially very low in sugars or other nonstmctural carbohydrates, and antagonists such as Trichoderma spp. can successfully compete for these limited resources. Many of the organisms used in the biological control of stem diseases are employed by applying them directly to stem wounds, where they colonize resources and subsequently exclude pathogenic forms. This initial occupancy by antagonists subsequently limits infection by decay-causing organisms, and hence controls the succession of microorganisms in the wood. Of the successful, commercially-available biological control products for plant diseases, several are for diseases of woody stems (Campbell 1989).
There are several different ways in which a microbial biological control agent can operate against a targeted plant pathogen (Elad 1986). Among these are competition, induction of plant defenses, and parasitism.
Some agents act through competition for limited resources, and through this competition the growth of the pathogen population is suppressed, reducing the incidence or severity of disease. One important component of competition can be competition for Fe-31 ions. Chemicals called siderophores, which are produced by many species of plants and microbes, sequester these ions. Highly efficient siderophores from nonpathogenic microbes can remove Fe-3l ions from the soil, outcompeting siderophores from pathogens and thereby limiting the growth of pathogen populations. Some biological control agents compete through the production of antimicrobial substances such as antibiotics which inhibit the growth of pathogens directly, rather than by preemptive consumption of limiting resources.
An important mechanism limiting infection is the induction of plant defenses against pathogensby nonpathogenic organisms. Cross-protection and induced resistance are mechanisms in which plants are intentionally exposed to certain (nonpathogenic or mildly pathogenic) microbes, thereby conferring in the treated plants some resistance to infection by pathogens. induced plant defenses may include lignification of cell walls through the addition of chemical cross-linkages in cell wall peptides which makes the establishment of infection through lysis more difficult, suberification of tissues (where plant cell walls are infiltrated with the fatty substance suberin, making them more corklike), and other general defenses, including production of chitinases and Beta 1,3-glucanases. These plant defenses then limit later infection by pathogens. The biological control agent employed may be an avirulant strain of the pathogen, a different forma specialis, or even a different species of microorganism.
A third mechanism by which beneficial microorganisms suppress plant pathogens is parasitism. Some species of Tricboderma, for example, attack pathogenic fungi, leading to the lysis of the pathogen. Natural enemies of plant-parasitic nematodes include bacterial diseases and nematophagous and nematopathogenic fungi.
As is the case of conservation of natural enemies of pest arthropods and weedy plants, conservation activities for the suppression of plant pathogens consist of either avoiding practices which reduce desirable antagonists or actively modifying the environment to favor or selectively enhance the growth of such species. in the case of soil microflora, species employed for biological control of plant pathogens are often competitive antagonists. Adding amendments to soil is one way in which soil microorganisms may be managed to enhance populations of these beneficial organisms. Addition of organic matter to soils for control of Streptomyces scabies, the causative organism of potato scab, is one example. Addition of carbon sources to soil increases general microbial activity that leads to reductions in S. scabies. Specifically, Bacillus subtilis and saprotrophic species of Streptomyces were encouraged by barley, alfalfa, or soy meal (Campbell 1989). Soy meal was also a substrate for antibiotic production against S. scabies. A general rise in soil organic matter also gave control of Phytophthora cinnamomi Rands in avocado in Australia (Manajczuk 1979). The addition of more than 10 tons of organic matter per hectare per year led to general increases in numbers of bacteria. Lysis of the hyphae and sporangia of the pathogen were attributed to species of Pseudomonas, Bacillus, and Streptomyces.
Some soils appear to suppress disease naturally and may contain antagonistic or antibiotic flora which flourish without the need for amendments. An example of such suppressive soils is the Fusarium-suppressive soil in the Chateaurenard District of the Rhone Valley in France Here, Fusarium oxysporum f: sp. melonis Snyder and Hansen is present, but no disease develops when susceptible melon varieties are grown. These soils are suppressive for several other types of F. oxysporum, but not to other species or genera of pathogens. The suppressive nature of the soils is clearly biotic, because the soils lose their suppressive ability when steam-sterilized, and the suppressive ability can be transfered to other soils. The antagonists principally responsible for this suppression are nonpathogenic strains of F oxysporum and F. solani (Martius) Saccardo. The suppression appears to be due to fungistasis induced by nutrient limitation. The competing fungi appear to have nearly the same ecological niche as the pathogenic forms, and the saprotrophic forms outcompete the pathogens for limiting resources so that dormant chlamydospores of the pathogen do not germinate in the presence of host root exudates. It may be possible to develop systems for other areas using the antagonists from the Chateaurenard area (Campbell 1989), although additional research may be necessary to permit their effective operation in different soils. Other soils suppressive to Fusarium wilts are known. There are numerous other examples of suppressive soils, although some soils or combinations appear to give somewhat variable results (Van Driesche & Bellows 1996).
The conservation of existing flora may be important in limiting the extent of a number of leaf diseases (Campbell 1989). These effects are often revealed through the use of fungicides, which deplete extant fungi, permitting the development of previously unimportant diseases, Fokkerna and de Nooij (1981), for example. evaluated the effects of various fungicides on leaf surface saprotrophs that have been used in biological control. Wide-spectrum fungicides allowed almost no growth of saprotrophs, while more selective agents permitted some growth of several genera of saprotrophs. in cases where these saprotroph populations play an important role in limiting disease organisms, the application of fungicides would eliminate their contribution to pathogen suppression. A case illustrated by Fokkerrrt and de Nooij (1981), is where plants treated with benomyl (a systemic fungicide) had fewer saprotrophs and developed more necrotic leaf area when inoculated with Cocbliobolus sativus (Ito and Kuribayashi) Drechsler ex Dastur than nontreatect plants (C. sativus is insensitive to benomyl), Another example (Mulinge & Griffiths 1974) is leaf rust of coffee (Coffea arabica L.), caused by Hemileia vastatrix Berkeley and Broome. The disease can be controlled by proper application of fungicide. However, if fungicides are applied in one year and not in the next, the disease is worse on the treated plants than on those which did not receive treatments either year. The elimination of the saprotrophic flora by the fungicide removes their natural suppressing influence on the disease organisms, permitting the disease to worsen , Here, careful use of selective fungicides are crucial to conserving the important antagonistic flora and permitting their beneficial action (Van Driesche & Bellows 1996).
Several reports exist of substantial natural control (control by natural enemies without intentional manipulation) of plant-parasitic nematodes. Stirling (1991) and Sayre and Walter (1991) review several of these; one example is that of the natural suppression of the cereal cyst nematode Heterodera avenae in cereal cultivation in the Great Britain (Gair et al 1969). In this case, populations of the nematode initially increased for the first 2-3 years of cultivations, and then declined continually during 13 years of continuous cultivation of both oats and barley (a more susceptible crop). Four species of nematophagous fungi were present in the soil. The two species principally responsible for nematode suppression were Nematopbtbora gynophila and Verticillium chlamydosporium. Both fungi attacked female nematodes, either destroying them or reducing their fecundity. The activity of both fungi was greatest in wet soils during laboratory trials (Kerry et al. 1980). Although natural suppression of the nematode population required some time to develop in these soils, but once established it maintained the population below the economic threshold (Stirling 1991).
Conserving nematode antagonists in soils (as opposed to directly enhancing their numbers), is a matter that has received relatively little attention (Van Driesche & Bellows 1996). The application of toxins (insecticides, fungicides) to aerial portions of crops or directly to soils often leads to pesticide activity in the soil. All nematicides are nonselective in their action and, hence, will kill predatory nematodes (Stirling 1991). In addition, herbicides have well-documented effects on soil microorganisms (Anderson 1978) and may well exert some influence on microbial antagonists of nematodes, and insecticides may negatively affect soil microarthropods. Many fungicides are known to be detrimental to nematophagous fungi (Mankau 1968; Canto-Saenz and Kaltenbach 1984; Jaffee and McInnis 1990), but at levels higher than would be expected under normal field practice. Among the fumigant nematicides, ethylene dibromide (EDB) and eibromo-chloro-propene (DBCP) appear nontoxic to the nematode-trapping fungi (Mankau 1968), and several herbicides were shown to be unharmful to Arthrobotrys sp. (Cayrol 1983). Despite these potentially significant effects on beneficial microflora and fauna and the possibility of conserving these organisms by appropriate choice of material, little has emerged to integrate these ideas into normal farming practice (Van Driesche & Bellows 1996). Perhaps because there has been no serious emergence of nematode problems associated with the use of these materials, this status quo is justified. Nonetheless, the opportunities for conserving biologically important agents should be considered in the development of future integrated management programs for plant-parasitic nematodes. Similarly, cultivation practices may be selected to favor natural enemies of nematodes, Among these are minimum or conservation tillage, which reduced the number of cysts of Heteroderci avenae on roots and the amount of damage caused by the nematode on wheat in Australia (Roger and Rovira 1987). other practices which may affect populations of natural enemies include normal tillage (which adds crop residue to the soil and thus may favor certain beneficial organisms) and crop rotation sequences (Stirling 1991). The knowledge that some soils are naturally suppressive to nematodes prompts the question of whether or not the features of these soils can be used to improve biological control. in all documented instances where they have been studied, the suppressive properties of these soils appear to result primarily from the action of one or two specific biological control agents (Stirling 1991). The suppressiveness requires substantial time to develop, and considerable crop loss might be incurred during such an initial phase. Some risk is involved also, because the suppressive nature of the soil may not develop to suitable levels. Careful management of crop varieties, particularly using varieties resistant or tolerant to nematode damage during the initial phases of land use for cropping, is an important part of taking advantage of the potential of these resident natural enemies. Agriculturists have large amounts of capital invested in land, equipment, and cropping costs, and consequently require a certain degree of reliability in pest control measures. Because of the variable nature of natural suppressiveness of nematodes, any natural control of nematodes in the foreseeable future is most likely to arise fortuitously rather than result from any deliberate actions by scientists or farmers (Stirling 1991). Where soils are not naturally suppressive to nematode populations, they may be manipulated to enhance what natural control agents are present. Most attention in this arena has been given to the addition of organic matter to the soils. Much of the information regarding the effects of these amendments is circumstantial, but the beneficial effects appear widespread. Many different soil amendments have been considered and evaluated, and the reduction of plant damage from nematodes following such amendments may occur through a variety of mechanisms (Stirling 1991).
One way is through the general improvement of soil structure and fertility. Addition of crop residue or animal manures increases ion exchange capacity of the soil, chelates micronutrients to make them accessible by the plant, and adds available nitrogen. Grown under such improved conditions, plants are better able to tolerate damage from nematodes. Certain amendments may directly improve plant resistance to nematodes (Sitaramaiah and Singh 1974). Others may contain or release compounds which adversely affect nematodes. Among amendments containing such compounds are those of neem (Azidiracbta indica A. jussien) seeds or leaves and of castorbean (Ricinus communis L.) (Stirling 1991 and references therein). Other amendments release nematicidal compounds during decomposition. The most widely studied of these compounds is ammonia. Because nitrogen is a constituent of nearly all soil amendments, ammonia is usually produced during decomposition. A careful lialance must be maintained in the carbon:nitrogen ratio, together with sufficient concentrations of ammonia, to provide optimal effect without phytotoxicity (Stirling 1991).
Finally, there is the direct stimulation of nematophagous or antagonistic organisms. Spores of many nematophagous fungi fail to germinate in otherwise suitable but nonamended soils (Dobbs and Hinson 1953), and this soil mycostasis can affect both spores and mycelia (Duddington et al. 1956ab, 1961; Cooke and Satchuthananthavale 1968). Before predation of nematodes can take place, mycelial growth and trap formation must occur. The addition of organic matter provides a substrate that may stimulate spore germination. Organic amendments stimulate a broad range of soil microorganisms, so the effects of amendments on populations of these organisms are complex. Microbial population growth generally increases immediately following the addition of organic matter and, subsequently, as pan of the community succession, there is an increase in populations of nematode-trapping fungi. The general hypotheses regarding the beneficial effects of organic amendments center around the stimulation of the saprotrophic growth phase of nematophagous fungi, and stimulation of other general microorganisms which may be detrimental to nematodes, such as antibiotic producing bacteria. A general rise in enzymatic levels also occurs following soil amendment, and the enzymes may attack the structural proteins in nematode cuticle or egg shell. Chitin amendments in particular have received attention, and addition of chitin to soil is followed by a relatively long-term (4-10 weeks) rise in chitinase activity in the soil. Chitin is the principal structural component of nematode eggshells, and the increase in chitinase activity may be accompanied by decreased survival of nematode eggs. However, the decomposition of chitin also releases ammonia, which may contribute to its beneficial effects. Speigel et al. (1988, 1989) concluded that the beneficial effects of chitin amendments resulted from the action of specialized microorganisms.
A current limitation of the implementation of amendments for nematode control is that such amendments must be applied in large amounts, between 1-10 tons/ha to be effective. The use of local resources for such amendments will keep transport costs minimal. One product, the chitin-based Clandosant (derived from crab shells), has been marketed commercially. There is some evidence that the effectiveness of certain amendments may be enhanced by inoculating them with degradative microorganisms (Galper et al. 1991), and Stirling (1991) suggests consideration of systems in which amendments can be inoctitated with a specific microorganism as they are applied to the soil.
Augmentation of antagonists of plant disease organisms can generally be of two types, inoculation and inundation. Inoculative releases consist of small amounts of inoculum, with the intention that the organisms in this inoculum will establish populations of the antagonist which will then increase and limit the pathogen population, In inundative releases, where large amount of inoculum is applied, with the expectation that control will result directly from this large initial population with limited reliance on subsequent population growth. Biological control of plant pathogens may also rely on a hybrid of these two concepts. A large amount of inoculum must be applied, both to increase the population of the antagonist and to improve its distribution to favor biological control. Also, antagonism can result from both these applied organisms and the increased population of antagonists resulting from their reproduction. Biological control of blackcrut (Phyllachora huberi) on rubber tree foliage by the hyperparasites Cylindrosporium concentricum and Diicyma pulvinata (Junqueira and Gasparotto 1991) is one example of long-term control of a plant pathogen by a single augmentation in an agricultural system (Cook 1993) In this case, rubber trees were treated with spore suspensions of the antagonists (inundatively), which resulted in control over more than one season. More generally, beneficial microorganisms are added seasonally or more frequently.
Where the beneficial organisms involved are being placed into a habitat or environment other than where they originated, the organisms are often referred to as "introduced" in plant pathology (Andrews 1992; Cook 1993). There are several examples where such organisms, when moved to a new habitat (for instance, from the soil to the above-ground part of a plant) colonize and serve as successful agents of biological control (Andrews 1992; Cook 1993).
Root Diseases. One way in which flora may be manipulated to protect against disease is to intentionally inoculate soils or seeds with microbial antagonists. Such antagonists, to be successful in their task, must be able to colonize plant surfaces and survive in the competitive environment of the soil. Flora with demonstrated ability to achieve this under field conditions include fungi, principally Trichoderma spp., and, among the bacteria, Bacillus spp. and Pseudomonas spp.
Among the bacteria, species of Bacillus are regularly used for biological control of root diseases, Members of the genus have advantages, particularly that they form spores which permit simple storage and long shelf life, and they are relatively easy to inoculate into the soil. However, the consequence of this biology is that although the inoculants may be present in the soil, it may be in dormant or resting stages. Nonetheless, species of Bacillus have provided good control on some occasions. Capper and Campbell (1986) showed a doubling of wheat yield over wheat plants naturally infected with take-all by those also inoculated with Bacillus pumil Meyer and Gottheil. Bacillus pumilus and B. Subtilis were also used to protect wheat from diseases caused by species of Rhizoctonia (Merriman et al. 1974). A major difficulty with the use of Bacillus spp. is that the control provided is often variable, with different results in different locations. or even in different parts of a season in the same location (Campbell 1989). Bacillus subtilus is used as a seed inoculant on cotton and peanut (Arachis hypogaea L.) with nearly 2 million ha. treated in 1994 (Blackman et al. 1994). Treatment promotes increased root mass, modulation, and early emergence, and suppresses diseases caused by species of Rhizoctonia and Fusarium.
Of much more promise as antagonists of root diseases are species of Pseudomonas, particularly the Pseudomanas fluoresce and Pseudomonas putida (Trevisan) Migula groups (Campbell 1989). These bacteria are easy to grow in the laboratory, are normal inhabitants of the soil, and colonize and grow well when inoculated artificially. They produce a number of antibiotics as well as siderophores. Several have received patents and are marketed commercially for control of root rot in cotton (Campbell 1989). An isolate of another species of Pseudomonas has been used as anantagonist of take-all disease of wheat (Welterl983). Isolates of Ps.fluorescens from soils showing some control of take-all can be applied as seed coats and inoculated into fields suffering from the disease. Such treatments give 10-27% yield increases compared with untreated, infected control groups. Evidence points to both siderophore and antibiotic production as important.
Species of the fungal genus Trichoderma can be saprotrophic and mycoparasitic and have been used against wilt diseases of tomato, melon, cotton, wheat, and chrysanthemums. The antagonists were applied to seeds or through a bran mixture incorporated into the planting mix at transplanting. Although disease did develop, it did so much more slowly than in untreated soils, resulting in a 60-83% reduction in disease (Siven and Chet 1986). The mode of action against Verticillium albo-atrum Reinke and Berthier wilt of tomatoes appeared to be antibiosis.
Stem Diseases. The control of Heterobasidion annosum, the causative agent of butt rot in conifer stumps, by Pbanerochaete gigantea was one of the first commercially available agents for biological control of a plant pathogen (Campbell 1989). The disease caused by H. annosum is primarily a disease of managed plantations. The fungus colonizes freshly cut stumps, invades the dying root system and can then infect nearby trees through natural root grafts, causing death of the trees. However, Heterobasidion annosium, is a poor competitor, and when a stump is intentionally inoculated with Ph. gigantea (and usually with chemical nitrogen sources which encourage growth of the antagonist) the antagonist rapidly colonizes the resource, excluding future attack by the pathogen and even eliminating existing pathogen infection (Table 12.1). Very little inoculum is needed on a freshly-cut stump, and the shelf life of the pellet formulation is about two months at 22'C. The antagonist is able to outcompete H. annosum even when the initial inoculum favors the pathogen by as much as 15:1 (Rishbeth 1963).
The ascomycete fungi Eutypa armeniaceae Hansford and Carter and Nectria galligena Bresadola & Strass infect apricots and apples, respectively, and cause stem cankers and eventual death of the trees. Pruning wounds in apricots are treated with Fusarium laterium Nees:Fries through specially adapted pruning cutters. Fusarium laterium produces an antibiotic which inhibits germination and growth of E. armeniaceae. When applied, the concentration of the antagonist must be greater than106 conidia/ml. Integrated application which includes a benzimidazole fungicide gives better control than either fungicide or antagonist alone. Nectria galligena infection can be reduced through sprays of suspensions of Bacillus subtilis or of Cladosporium cladosporioides (Fresenius) de Vries. These antagonists are not in commercial use because apples are treated for Venturia inaequalis (Cooke) G. Winter (apple scab) so frequently that V. galligena is controlled by those sprays.
Crown gall is a stem dlisease caused by the bacterim Agrobacterium tumefaciens (Smith & Townsend) Conn. It affects both woody and herbaceous plants in 93 families. Infection is typically from the soil, rhizosphere, or pruning tools. Control can be effected by treating plants with a suspension of a related saprotrophic bacterium Agrobacterium radiobacter (Beijerink and van Delden) Conn strain K-84. This strain of the bacterium produces an antibiotic that is taken up by a specific transport system in the pathogen bacterium, which is then killed. The commercially available formulations of this agent are effective primarily against pathogen strains which attack stone fruits, but other bacteria are under investigation for use against strains pathogenic in other crops. This agent has been altered by gene-modifying technology to produce a new strain (strain 1024) which lacks the ability to transfer antibiotic resistance to the target bacterium (Van Driesche & Bellows 1996).
The fungus Cbondrostereum purpureum (Persoon: Fries) Pouzar infects stems of fruit trees and produces a toxin which leads to a condition known as silverleaf disease. Stems can be inoculated with a species of Trichoderma grown on wooden dowels or prepared as pellets which are inserted into holes bored in the affected stem. Treated stems recover from the disease more rapidly than untreated stems. The Trichoderma sp. can be applied to pruning wounds to prevent initial establishment of C. purpureum.
Leaf Diseases. Control of leaf diseases at the time of pathogen germination has been demonstrated in the laboratory. This control occurs in the presence of competitive organisms, which may include fungi, yeast, or bacteria. The mode of action in some cases is competition for nutrients that, together with water, are necessary for successful germination and invasion of many pathogens. The germination of Botrytis sp., for example, is inhibited by certain bacteria and yeasts (Blakeman and Brodie 1977). This inhibition is less pronounced if additional nutrients are supplied, indicating that the mechanism is, at least in parrt, resource competition. Studies on control of Botrytis rot in lettuce (Wood 1951) indicated that several organisms were successful in suppressing the disease when sprayed on lettuce plants, among them species of Pseudomonas, Streptomyces, Ttichoderma viride, and Fusarium. Peng and Sutton (1991) evaluated 230 isolates of mycelial fungi, yeasts, and bacteria and tested them as anntagonists of B. cinerea in strawberry in both laboratory and field trials. Several organisms (including members of each taxonomic group tested) were effective, some as effective as captan (a commercial fungicide). Sutton and Peng (1993b) further evaluated Gliocladium roseum and determined that the suppression of B. cinera by this antagonist was probably a result of competition for leaf substrate. The fungi Gliocladium roseum and Myrothecium verrucaria (Albertini and Schweinitz) Ditmar were also effective in suppressing B. cinerea in black spruce (Picea mariana [Miller] Britton Stearns Poggenburg) seedlings (Zhang et al. 1994).
Bacteria may also be used to limit frost damage to leaves and blossoms of plants. Certain bacterial species such as Pseudomonas syringae and Erwinia herbicola serve as nucleation sites on leaves for the formation of ice, and, in their presence, ice forms soon after temperatures fall below freezing. If these ice nucleating bacteria are replaced by competitive antagonists (such as certain strains of Ps. syringae) that lack the protein that causes ice nucleation, frost is prevented even at temperatures from -2 to 5E C.(1 (Lindow 1985b). The protective bacteria, after being applied it) the leaves, colonize them for up to two months, an interval suitable to protect from frost during the limited season that low temperatures are likely. A naturally-occurring, non-ice nucleating strain of Ps. fluorescens is registered in the United States as a commercial product (Frostban B ) for suppression of frost damage (Wilson and Lindow 1994).
Spraying suspensions of propagules, generally at high concentrations, is the principal method for applying biological control agents to foliage (and to flowers), and dusts (such as lyophilized bacterial preparations) are also used. Spray methodology has yet to be refined in terms of sprayer characteristics, droplet size, and pressures, and other methods of application with greater efficiency may be necessary to effectively target certain plant parts (Sutton & Peng 1993a).
Flower Diseases. A principal disease of flowers, which has received attention, is fire blight of rosaceous plants, which is particularly severe on pear (Campbell 1989). The causal bacterium, Erwinia amylovora, also occurs on leaves and may cause stem cankers. Insects transfer the bacterium to flowers in the spring from overwintering sites on stem cankers, and subsequently from flower to flower. Infection enters the pedicel and from there the stem. Infected flowers and small stems die, and cankers form on other stems. Chemical control is difficult and expensive, and sometimes is ineffective because of resistance to copper compounds and streptomycin. Biological control has been effective using E inia berbicola, sometimes in combination with Pseudomonas syringae (Wilson and Lindow 1993). Suspensions of E. berbicola are sprayed onto the flowers just before the period of potential infection. The antagonist occupies the same niche as the pathogen, reducing the numbers of E. amylovora by competition, and there is also evidence for the production of bacteriocins (chemicals which suppress population growth of related bacteria) by some strains. Control can be good, comparable to that achieved by commercial bactericides, though repeated application of the bacterium was necessary (Isenbeck and Schultz 1986). Another approach to control is to reduce secondary infections on leaves, which leads to reductions in the overwintering population of the pathogen. This control is achieved by treatment with the antagonists Ps.syringae and other bacteria (Lindow 1985b). A novel approach to dissemination of the antagonistic bacteria has been evatuated by Thomson et al. (1992). They mixed E. herbicolaancl Ps. fluorescens with pollen in a special apparatus at the entrance to honey bee (Apis mellifera) hives. Bees emerging from these hives through the mixtures transmittecl the antagonists to the flowers efficiently, although disease control was not evaluated because of absence of disease in the test orchards.
Fruit Diseases. Fruits are subject to attack both by general pathogens (Botrytis, Rhizopus, Penicillium) and by a few specialist pathogens such as the coffee berry disease fungus Colletottichum coffeanum Noack and Monilinia spp., which cause brown rots of rosaceous fruits. While many of these are controlled by fungicides, Trichoderma viride has been shown to limit disease from Monilinia spp. Various Bacillus spp. also are antagonistic to these fungi through production of antibiotics and by reducing the longevity and germination of spores. Both the bacteria and culture filtrates have been used with some success against these pathogenic fungi, but there has been no commercial development, probably because fungicides used routinely in orchards for control of other diseases give some control of brown rot (Campbell 1989). Among the most serious diseases of soft fruits are postharvest rots (Dennis 1983), especially that caused by Botrytis cinerea. Potential for biological control of postharvest diseases was reviewed by Wilson and Wisniewski (1989) and Jeffries & Jeger (1990) (also see Wilson and Wisniewski 1994). In strawberries, B. cinerea grows saprotrophically on crop debris and from there infects flowers or fruit. Various species of Trichoderma have been evaluated and gave control as good as standard fungicides (Tronsmo and Dennis 1977). The antagonists Clado,@Cladosporium herbarium and Penicillium sp. gave excellent results in controlling Botrytis rot on tomato (Newhook 1957). Honey bees have been used to distribute Gliocladium roseum to strawberry flowers (Peng et al. 1992) and raspberry flowers (Sutton and Peng 1993a) to suppress Bot?ytis rot.
Root Diseases. Induced resistance is a form of biological control in which the natural defense responses of the plant, which may include production of phytoalexins, additional lignification of cells, and other mechanisms (Horsfall and Cowling 1980; Bailey 1985), are promoted in the plant prior to exposure to the pathogen (Van Driesche & Bellows 1996). Challenging the plant with a nonpathogenic organism induces these resistance mechanisms. The induced plant defenses then limit later infection by the pathogen. The organism employed may be an avirulent strain of the pathogen, or a different specialized form, or even a different species, There are few well-documented cases of induced resistance for soil-borne pathogens, and these are mostly of wilt diseases. Dipping tomato roots in a suspension of Fusarium oxysporum f.sp. dianthi a few days before likely exposure to the pathogen F oxysporum f.sp. lycopersici (Saccardo) Snyder and Hansen conferred protection that lasted a few weeks. Cotton may be protected for three months or longer by spraying the roots at transplanting with a mildly pathogenic strain of the disease causing pathogen Vcrticillium albo-atryum. The role of some fungi against take-all of wheat includes some elements of induced resistance. Gaeumannomyces graminis var. graminis grows on grass roots and also has been found on wheat, where it occupies a niche similar to that of the pathogen G. graminis var. tritici Walker. The antagonist invades the root cortex but not the stele, and is halted by the lignification and suberization of the cortex and stele. Root cells with these chemically-changed walls are less susceptilile to invasion by the pathogen. Although this interaction produced yield increases in Europe, the strains or species present in the United States did not appear to confer resistance, and in Australia there were only slight yield increases (Campbell 1989). These variable results, while somewhat common for biological control of soil borne pathogens, do not reduce the value of the antagonists where they do work, but rather indicates some potential challenges in defining the taxonomy, biology, and host-plant relationships important to biological control in this group of organisms.
Leaf and Stem Diseases. Induced resistance can control anthracnose diseases caused ny Colletotrichum spp. (Kúc 1981; Dean and Kúc 1986). Colletotricbum lindemthianum (Saccardo & Magnus) Lamson Scribner causes anthracnose of beans, Colletotrichum lagenarium (Passerine) Saccardo causes cucumber anthracnose, and Cladosporium cucumerinum Ellis and Arthur causes scab in cucumbers. Inoculation of cucumbers with Colletotrichum lindemuthianum (which does not cause disease in cucumbers) made plants resistant to both Colletotrichum lagenarium and Cladosporium cucumerinum. Treatment applied to an early leaf resulted in protection of later leaves, even when the initially inoculated leaf was removed. The factor causing resistance travels systemically through the plant. Variations on this approach include inoculating an early leaf with a pathogen, inducing resistance throughout the plant, and then removing the infected leaf. Induced resistance also occurs in some virus diseases (Thomson 1958) and may last for years, as in the case of healthy citrus seedlings being inoculated with an avirulent strain of citrus tristeza vinus
Stem rot in carnations, caused by Fusarium roseum Link: inoculating wounds inflicted during propagation with the nonpathogenic F. roseum 'Gibbosum' can prevent Fries 'Avenaceum,'. This inoculation produced a germination inhibitor and also reduced the time needed for the stems to develop resistance to the pathogen. This hastening of resistance was caused by activation of the host's defense mechanisms, and is another example of induced resistance.
Employing hypoviruient strains of the disease pathogen controls chestnut blight, caused by Cryphonectria parasitica. A number of hypovirulent strains are known, and inoculating infected trees with a hypovirulent strain leads to reduced canker size and greater stem survival. In the field, hypovirulent strains are inoculated into infected trees at the rate of 10 inoculated trees/ha. The hypovirulent strain spreads from these locations and, on contacting more virulent strains, fuses with these strains and exchanges a viral element infecting the pathogen (Van Driesche & Bellows 1996). The hypovims, which causes hypovirutence, is transferred to the virulent strains, attenuating their effects. Active cankers are eliminated in 10 years (van Alfen 1982).
Root Diseases. The mycoparasites Tiichoderma spp, have been used successfully against diseases caused by Rbizoctonia and Sclerotium pathogens. One example is the pathogen Sclerotium rolfsiii Saccardo, which attacks many crop plants and survives unfavorable periods by forming sclerotia in the soil. Strains of T barzanium that have Beta 1-3 glucanases, chitinases, and proteases have been isolated. These enzymes permit T barzanium to parasitize the hyphae and sclerotia of the pathogen, invading and causing lysis of the cells. Trichoderma harzanium is grown on autoclaved bran or seed, and this material is then mixed with the surface soil (Chet and Henis 1985). Two other fungi known to parasitize sclerotia are Coniothryium minitans and Sporidesmium sclerotivorum (Ayers and Adams 1981).
Sporidesmium sclerotivorum is a hyphomycete that in nature behaves as an obligate parasite of sclerotia of Botrytis cinerea and several species of Sclerotium (Adams 1990). It has been studied as an agent against botrytis rot in lettuce, where it shows considerable potential. It can be grown in vitro on various carbon sources and is efficient in converting glucose into mycelium. Spores produced in mass culture are collected, processed, and applied to infected soil, and field tests are promising (Adams 1990).
Leaf Diseases. Some plant pathogens, including fungi and some bacteria, are known to be attacked by other pathogens. Bdellovibrio bacterivorus is a bacterium that can attack other bacteria by penetrating the cell wall and lysing the host bacterium, subsequently reproducing inside its host. Different strains of Bd. bacterivorus have been examined for virulence against Pseudomonas syringae pv. glycinae (Coeper) Young, Dye and Wilkie, the cause of soybean blight. By applying Bd. bacteriovorus at sufficiently high rates, disease symptoms were reduced more then 95% (Scherff 1973). Parasites of fungi pathogenic on leaves are numerous (Kranz 1981), but only a few have been studied in much detail, such as Sphaerellopis filum, Verticillium lecanii and Ampelomyces quisqualis. The mycoparasite typically penetrates the host hypha or spore and kills it. Some of the control may be from the pathogen overgrowing the sporulating pustules of the pathogen and preventing spore release and thus reducing inoculum in the environment, even if the spores are not killed. A typical problem with implementation of these mycoparasitic fungi is that they often do not affect a large proportion of the pathogens unless humidity and temperature are high. Consequently, although much reduction of spore production may take place, there is still sufficient inoculum of the pathogen remaining to cause disease. These mycoparasites often are seen only at high incidences of disease, which is unsuitable for general control of the target pathogens. They may have some use in particular systems, either in the tropics or in greenhouses, where environmental conditions are more favorable.
Plant-Parasitic Nematodes. The bacterial pathogen of nematodes most studied is Pasteuria penetrans sensti stricto Starr and Sayre (Starr and Sayre 1988), which is an obligate parasite of root-knot nematodes (Meloidogyne spp.) and has not been successfully cultured in vitro. This restriction in mass culturing has limited attempts to test the bacterium's effectiveness (Stirling 1991). In experimental trials, it has shown potential for controlling root-knot nematodes (Meloidogyne spp.) (Mankau 1972; Stirling et at. 1990), infesting a high proportion of nematodes in soil to which bacterial spores had been added, and in other trials (U. S. Department of Agriculture 1978) reducing damage to plants in plots containing the bacterium. Observations by Mankau (1975) indicated that populations of the bacterium did not increase rapidly in field soil. The development of a mass production method in which roots containing large numbers of infected Meloidogyne spp. females were air-dried and finely ground to produce an easily handled powder enabled more extensive testing (Stirling and Watchel 1980). When dried root preparations laden with bacterial spores were incorporated into field soil at rates of 212-600 mg/kg of soil, the number of juvenile Meloidogyne. javanica (Treub) Chitwood in the soil and the degree of galling was substantially reduced (Stirling 1984); other authors have reported similar results (Stirling 1991). Effective use of this bacterium through such inundative release would require concentrations on the order of 105 spores/g soil (Stirling et al. 1990). Such quantities could only be produced on a large scale with an efficient in vitro culturing method, a problem which has received attention but has not yet yielded a solution (Stirling 1991). Use in inoculative releases, where smaller numbers of spores are applied and a crop tolerant of nematode damage is grown to permit the increase of both nematode and bacterial populations, has been suggested (Stirling 1991). Conserving the bacterium in the presence of nematicides appears possible. Of seven tested nematicides, only one showed slight toxicity to the bacterium (U.S. Department of Agriculture 1978). The use of Bacillus thuringiensis strains with activity against nematodes is also possible. As these bacteria may be cultured in fermentation media. their mass culture is simpler than for P penetrans. Suppression of nematodes was possible through drench applications and through incorporating the bacterium into a methyl cellulose seed coat (Zuckerman et al. 1993).
Considerable attention has been given to the nematode-trapping fungi as possible augmentative agents, Mass culture on nutrient media is possible for these fungi. Two cultures of nematophagous Arthrobotrys fungi have been developed and tested for addition to soil for specific target environments. Cayrol et al. (1978) reported the successful use of Arthrobotrys robusta Cooke and Ellis var. antipolis, commercially formulated as Royal 300 against the mycetophagous nematode Ditylenchus myceliophagus Goodey in commercial production of the mushroom Agagaricus bisporus (Lange) Singer. The nematophagous fungus was seeded simultaneously with A. bisporus mushroom compost, which led to 280/o increases in harvest and reduced nematode populations by 40%. The results justified the commercial use of the fungus for nematode control in mushroom culture, Cayrol and Frankowski (1979) reported the use of Arthrobotrys superba Corda (Royal 3509) in tomato fields, applied to the soil at a rate of 140 g/M2, resulting in protection of the tomatoes and colonization of the soil by the fungus. Other reports have indicated little efficacy of fungal preparations when added alone to soil (Barron 1977; Sayre 1980; Rhoades 1985). In general, there has been limited success in the use of these agents (see Stirling  for a summary). The fungistatic nature of soil (Mankau 1962; Cooke and Satchuthananthavale1968) may limit the ability of these fungi to grow even when added in substantial numbers to soil. Additional work is needed, perhaps in the areas of colonization and soil amendments together for the use of nematophagous fungi to become suitably reliable for general use as a control method (van Driesche & Bellows 1996).
Many predacious fungi may be unsuited for control of root-knot nematodes, Meloidogyne spp. Stirling (1991) suggested that Monacrosporium lysipagum (Drechsler) Subramanian and Monacrosporium ellipsosporum (Grove) Cooke and Dickinson, which can invade egg masses, may warrant further investigation. The nematode-trapping fungi are likely to be more effective against ectoparasitic nematodes and such species as Tylenchus semipenetrans Cobb, where juvenile stages migrate through the rhizosphere. Little attention has been given to testing predacious fungi against such nematodes (van Driesche & Bellows 1996).
Fungi which are internal parasites of nematodes are difficult culture on nutrient media, and consequently there have been few attempts to use them for augmentative control of nematodes. Alternative mass-culturing techniques may hold promise (Stirling 1991). In the few experiments reported, the fungistatic effects of soil often limited fungal growth and the effectiveness of the antagonists. Lackey et al. (1993) report the production and formulation of Hirsutella rhossiliensis Minter et Brady on alginate pellets (see also Fravel et al. 1985) which, when added to soil, led to transmission of the fungus to the nematode Heterodera schachtii Schmidt and suppressed nematode invasion of roots.
Among the facultatively parasitic fungi which attack nematodes, Paecilomyces lilacinus and Verticillium cblamydosporium have received the most attention as possible augmentative agents. The results of studies on P lilacinus have been variable, with some studies showing some positive effect of the fungus, while others show little or no effect (Stirling 1991). The mechanisms leading to the beneficial effect have not been clearly elucidated, but may be from metabolic products or effects other than direct parasitism of eggs. Studies have generally involved the addition of fungal preparations to the soil at the rate of 1-20 tons/ha, which is really too great for widespread commercial use. Additions at lower rates (0.4 tons/ha) in a variety of carriers (alginate pellets, diatomaceous earth, wheat granules) have also shown limited beneficial effects (Cabanillas et al. 1989; Stirling 1991). Tribe (1980) suggested the direct addition of V. Chlamydosproium to the soil. Kerry (1988) added hyphae and conidia, formulated in sodium alginate pellets or in wheat bran, to soil, and the fungus proliferated in the soil only from granules containing bran. When chlamydospores were used as inoculum, the fungus was able to establish without a food base (De Leij and Kerry 1991). Of three isolates studied, only one successfully colonized tomato root surfaces. This species apparently has considerable promise, but screening programs will be necessary to identify isolates with characteristics suitable for biological control (Stirling 1991).
Predacious microanhropods and nematodes have evoked considerable interest. Most work has been done in simple microcosms, and there have been no attempts to evaluate augmentative release of these organisms in a field setting. In one experiment, Sharma (1971) found nematode numbers reduced by 50% or more in glass jars inoculated with mites and springtails compared with similar jars containing no predators, but the author pointed out possible causes for the reduction other than simple predation. Experiments with predacious nematodes have in general failed to demonstrate a measurable impact of the predator (Stirling 1991). One exception was the reduction of galling by Meloidogyne incognita on tomato by predacious nematodes (Small 1979). The general suitability of these groups of organisms for inundative release is questionable, because of the potential difficulties in developing technologies for their rearing, packaging, transport, and delivery beneath the soil in a viable state (Stirling 1991).
Mycorrhizae are nonpathogenic fungi associated with roots in some temperate forest trees. Ectomycorrhizae are mostly basidiomycetes which form a sheath over the root, and hyphae spread out into the soil. These fungi have been studied in relation to nutrient uptake, but they also affect root disease. Because they completely enclose the root, they change the quantity and quality of exudates reaching the soil; consequently, roots with mycorrhizae have a different rhizosphere flora than uninfectect roots (Campbell 1985). In at least one case, the mycorrhizal fungus Pisolitbus tinctorius (Persoon) Coker and Couch, the thick symbiont sheath forms a barrier to infection by such pathogens as Phytophthora cinnamomi attacking eucatyptus trees. Other mycorrhizal fungi produce antibiotics effective against P cinnamomi in plate tests. The intentional manipulation of mycorrhizal fungi for disease control has not been widely implemented, but opportunities for selected uses may be possible (Campbell 1989).
Another group of fungi are the vesicular arbuscular mycorrhizae (VAM), which are phycomycete fungi associated with the roots of many plant species including many crops. These fungi do not form a sheath surrounding the root, and their effects on disease are complicated, but are in general beneficial (Campbell 1989). Some of these effects may involve changes in host plant physiology in the presence of the symbiont, as there is no direct evidence of pathogen inhibition by these fungi.
Growth of our knowledge about biological control of plant diseases has been extensive since the first experimental reports (Hartley 1921; Sanford 1926; Millard,,ind Taylor 1927; Henry 1931), and substantial potential for microbial control of pathogens has been demonstrated. A number of products or programs have reached the stage of commercial development or availability (van Driesche & Bellows 1996). Products in current use include both those aimed at specialty markets for control of certain stem or flower diseases (for which chemical control is either unavailable or expensive) and those aimed at larger scale markets such as seed treatments for widely planted crops.
The cycle for research, development, and implementation of antagonists of plant pathogens is composed of several steps. These include initial discovery of candidate agents, refinement of knowledge of their biology, ecology, and mode of action, microcosm and field trials of their efficacy, and large-scale development for commercial production.
The first challenge in the development of a biological control program is the discovery process. Many microorganisms show potential as antagonists of particular pathogens. Protocols have been proposed to make the process of screening these candidates more efficient (Andrews 1992; Cook 1993). The principal difficulties are screening out candidates that are effective only during in vitro (agar plate) trials but are not effective in natural settings, and in selecting candidates that can be successfully cultured in large quantities. Following discovery of suitable candidates, research focuses on their mode of action and on factors which may enhance or limit their efficacy in targeted settings (glasshouses, field plots). In addition, experimental fermentation and formulations must be developed for production of materials suitable for use in agricultural settings. Finally, issues of large-scale production and delivery must be addressed. Products for use must be effective on an economical basis, and economies of scale may play an important role in the eventual availability of any organism or product. Each must have a satisfactory shelf life, and safe and effective methods for application must be discovered or developed (Cook 1993; Sutton and Peng 1993a). Such application methods might include sprays of suspensions or dusts, contact application, honeybee and other bee vectoring, and production of antagonists in a crop environment (Sutton and Peng 1993a).
Adopting any biological control agent in commercial agriculture is dependent on its reliability and availability. Limitations to the process of eventual adoption, therefore, include cost of development and size of potential market (van Driesche & Bellows 1996). Many pesticides for control of plant diseases have a broad spectrum of activity, are applicable in a variety of crops and settings, and may act either prophylactically, therapeutically, or both. Biological controls, in contrast, often have narrow ranges of activity and may work in only a few crops or soil types, and while they can often act both as a barrier and therapeutically, their action may take some time to develop. Therefore, they may have a narrower market than a chemical pesticide and be unattractive for development by major corporations (Andrews 1992). In this context, it may be appropriate for public institutions such as government experiment stations to undertake the development of such biological controls, in the same way that they take the responsibility for development of new plant varieties (Cook 1993).
Those microorganisms intended for use as biological control agents must be viewed in a biological rather than a chemical control (Cook 1993). Where an effective pesticide may work in many places, each place may have unique edaphic, and biological features which limit or enhance the effectiveness of microbial antagonists of pathogens. Consequently, each microbial biological control system may have to make use of locally adapted strains, taking advantage of resident antagonistic flora and fauna and augmenting their effectiveness with additional species or strains, or enhancing resident populations through soil amendments. Although the different strains may use common mechanisms to achieve biological control (such as production of antibiotics), competitive abilities adapted to local conditions may be vital to permit the organisms to compete for resources and effectively control pathogens (van Driesche & Bellows 1996)