This review discusses the mechanisms underlying synergistic interactions between phytophagous nematodes and soilborne pathogens, and identifies biotic and abiotic factors affecting these interactions. Approaches towards the resolution and management of nematode–pathogen complexes are considered and discussed.
The development of disease in cultivated crops has long been known to depend on the complex interrelationship between host, pathogen and prevailing environmental conditions. In the case of soilborne pathogens, further opportunities exist for interactions with other microorganisms occupying the same ecological niche. The significant role of nematodes in the development of diseases caused by soilborne pathogens has been demonstrated in many crops throughout the world (Table 1). In many cases, such nematode–fungus disease complexes involve root-knot nematodes (Meloidogyne spp.), although several other endoparasitic (Globodera spp., Heterodera spp., Rotylenchulus spp., Pratylenchus spp.) and ectoparasitic (Xiphinema spp., Longidorus spp.) nematodes have been associated with diseases caused by soilborne fungal pathogens. While nematode–fungus complexes have been reviewed previously (Pitcher, 1965; Powell, 1971; Mai & Abawi, 1987; Evans & Haydock, 1993), this review (i) discusses the mechanisms underlying synergistic interactions; (ii) identifies biotic and abiotic factors affecting their progress; and (iii) outlines potential approaches for the resolution and management of nematode–soilborne pathogen complexes.
Table 1. Examples of nematode–pathogen disease complexes reported in crops between 1997 and 2001
The natural soil environment harbours a multitude of microorganisms. As many as 106–108 bacterial cells, 106–107 actinomycete cells, 5 × 104–106 fungal colony-forming units (CFU), 105–106 protozoa and 104–5 × 105 algae were estimated to be present in a gram of field soil taken from the surface (Gottlieb, 1976), while Richards (1976) found c. 1 × 107 nematodes in an area of 1 m2 of fertile soil. Although many of these organisms are saprophytic, having little, if any effect on cultivated crops, the moist soil environment is favourable for the activities of plant-parasitic nematodes (PPN) and for the growth and multiplication of pathogenic fungi. It is of no surprise, therefore, that a variety of interrelationships between these organisms have been demonstrated.
It has long been understood that the development of disease symptoms is not solely determined by the pathogen responsible, but is dependent on the complex interrelationship between host, pathogen and prevailing environmental conditions. In addition, in nature plants are rarely, if ever, subject to the influence of only one potential pathogen. This is especially true of soilborne pathogens, where there is tremendous scope for interaction with other microorganisms occupying the same ecological niche. Disease aetiology and reasons for the multifactorial nature of disease causation are described by Wallace (1978).
Examples of interactions between soil microbes influencing disease development can be seen in PPN–pathogen complexes. A disease complex is produced through a synergistic interaction between two organisms. Synergistic interactions can be summarized as being positive where an association between nematode and pathogen results in plant damage exceeding the sum of individual damage by pest and pathogen (1 + 1 > 2). Conversely, where an association between nematode and fungus results in plant damage less than that expected from the sum of the individual organisms, the interaction may be described as antagonistic (1 + 1 < 2). Where nematodes and fungi are known to interact and are shown to cause plant damage that equates to the sum of individual damage by pest and pathogen, the association may be described as neutral (1 + 1 = 2). Although the former two associations can be readily demonstrated experimentally, the latter can prove difficult to identify, as neutral associations can result in similar plant damage to that seen in additive associations, where nematode and pathogen are known not to interact with one another.
The first recorded case of a nematode–fungus interaction was made by Atkinson (1892), who observed that fusarium wilt of cotton (caused by Fusarium oxysporum f.sp. vasinfectum) was more severe in the presence of root-knot nematodes (Meloidogyne spp.). Further evidence for the interaction between Fusarium spp. and root-knot nematodes in cotton was later provided during field experiments in which ethylene dibromide or 1,3-dichloropropene was used to sterilize soil (Smith, 1948; Newson & Martin, 1953). Where a soil sterilant was used, the incidence of wilted cotton plants was significantly reduced. As the chemicals are regarded as having little fungicidal activity, it was assumed that they indirectly reduced pathogen infection by reducing the population densities of nematodes with which they interact.
Ectoparasitic nematodes such as Belonolaimus and Trichodorus spp. are rarely recorded to have a role in synergistic interactions with fungi, probably because their feeding behaviour causes only minor tissue damage to plant roots (Hussey & Grundler, 1998). In comparison, ectoparasites such as Xiphinema (dagger nematode) and Longidorus (needle nematode) have longer stylets for feeding in the vicinity of the vascular cylinder, and are recognized as important vectors of plant viruses (Taylor, 1990; Brown et al., 1998; Ipach et al., 2000).
Compared with those of ectoparasites, the life cycles of endoparasitic nematodes are far more complex and involve closer associations with their plant hosts. This means that plants infested with endoparasites are usually subject to various nematode-induced modifications. These can vary from localized forms of damage caused during invasion and feeding to overall systemic effects such as retarded plant growth. It is these changes which influence infections by soilborne pathogens. The endoparasites Globodera, Heterodera, Meloidogyne, Rotylenchulus and Pratylenchus are the genera most commonly reported to be involved in disease complexes with fungal pathogens. These typically interact with the wilt fungi Fusarium and Verticillium and the root-rot pathogens Pythium, Phytophthora and Rhizoctonia. Root-knot nematodes (Meloidogyne spp.) are perhaps the most recurrently recorded nematodes found in disease complexes with fungi. This is illustrated well by the interaction between wilt disease caused by F. oxysporum f.sp. vasinfectum and Meloidogyne incognita, which has long been a problem in cotton crops (Gossypium hirsutum) and has frequently been documented in the literature (Atkinson, 1892; Garber et al., 1979; Mai & Abawi, 1987; De Vay et al., 1997; Abd-El-Alim et al., 1999). More recently, M. incognita has been found in association with the pathogen Thielaviopsis basicola, which causes black root-rot of cotton (Walker et al., 1998). In cotton, neither of these organisms is considered to cause acute effects individually, and plant mortality rarely occurs in their presence (Walker et al., 1998), yet in combination they have consistently been found to increase seedling mortality, increase root necrosis, suppress early seedling growth and subsequently reduce the percentage of bolls (Walker et al., 1998; Walker et al., 1999; Walker et al., 2000; Wheeler et al., 2000).
On potatoes, the Globodera–Verticillium dahliae and Pratylenchus–Verticillium dahliae disease complexes have become particularly notorious. Early senescence or ‘early dying’ caused by V. dahliae and V. albo-atrum is accentuated by populations of Pratylenchus spp. (Martin et al., 1981; Wheeler et al., 1992; Bowers et al., 1996; Hafez et al., 1999), G. rostochiensis (Evans, 1987), and G. pallida (Hide et al., 1984; Storey & Evans, 1987). In terms of yield, Martin et al. (1982) calculated that 15, 50 and 150 P. penetrans per 100 cm3 soil in combination with V. dahliae would result in 36, 60 and 75% reductions in potato tuber weight, respectively. However, tuber weights were unaffected by the presence of the individual pathogens, except where nematode populations were high (150 P. penetrans per 100 cm3), when a 12% reduction was found. Yield reduction from the Pratylenchus–V. dahliae complex has also been reported elsewhere (Botseas & Rowe, 1994), as have other damaging effects such as the disruption of photosynthesis, stomatal conduction and transpiration (Saeed et al., 1997a; Saeed et al., 1997b). However, the fundamental importance of P. penetrans in potato early dying is its ability to activate low populations of V. dahliae that would otherwise be inconsequential in disease development (Bowers et al., 1996).
Another disease complex involves the soyabean cyst nematode Heterodera glycines and the fungus Fusarium solani. Sudden death syndrome (SDS) caused by F. solani is a major disease of soyabean which, among other symptoms, induces root rot, crown necrosis, interveinal chlorosis, defoliation and abortion of pods (Rupe, 1989; Nakajima et al., 1996). The aetiology of SDS is complicated, and changing abiotic factors such as temperature and moisture (McLean & Lawrence, 1993) are influential in its development. Furthermore, H. glycines is considered by many to be an important player in the incitement of SDS (McLean & Lawrence, 1993; Rupe et al., 1993). In two years of microplot experiments, McLean & Lawrence (1993) found that the incidences of SDS symptoms in plots containing both H. glycines and F. solani were 35 and 18% higher than in plots where the fungus was inoculated alone. Recent research on SDS has focused on identifying genes for dual resistance against both nematode and fungus (Chang et al., 1997; Meksem et al., 1999; Prabhu et al., 1999).
Mechanisms underlying synergistic interactions
Utilization of nematode-induced wounds by soilborne pathogens
Depending on specific life cycles, PPN are able to cause a variety of types of wound on host plant roots while entering or feeding. For example, ectoparasitic nematodes such as Trichodorus spp. and Tylenchorhynchus spp. feed on root epidermal cells, leaving behind simple micropuncture-type wounds. In contrast, endoparasitic nematodes are far more disruptive to their hosts’ roots. The root-lesion nematode Pratylenchus spp. is a migratory endoparasite that travels intracellularly through the cortex of roots by cutting through cell walls with its stylet to create a path. The sedentary endoparasites Meloidogyne spp., Globodera spp. and Heterodera spp. have highly specialized feeding strategies together with elaborate life cycles. Vermiform juvenile nematodes (J2) select penetration sites behind growing root tips (Doncaster & Seymour, 1973) and migrate either intracellularly (Globodera and Heterodera spp.) or intercellularly (Meloidogyne spp.) (Von Mende et al., 1998) to the vascular cylinder, where specialized ‘nurse cell systems’ are initiated (Jones & Northcote, 1972).
Some authors (Bergeson, 1972; Taylor, 1990) have regarded nematode invasion sites and tracts as inconsequential in the aetiology of fungal diseases. However, there are a number of reports which clearly illustrate that nematode damage has a role in the establishment and development of disease caused by soilborne pathogens. Inagaki & Powell (1969) adopted a number of methods to investigate the importance of mechanical wounding by the root-lesion nematode Pratylenchus brachyurus on the development of black shank symptoms (Phytophthora parasitica) in flue-cured tobacco (Nicotiana tabacum cv. Hicks). Plants subjected to an artificial root-wounding treatment produced significantly more severe disease symptoms of black shank. Additionally, plants receiving either (i) simultaneous inoculation of the nematode P. brachyurus and the oomycete P. parasitica, or (ii) introduction of P. brachyurus 1 week prior to P. parasitica, also exhibited elevated disease development in comparison to plants inoculated with the oomycete alone. However, plants inoculated with nematodes 2 or 3 weeks before the introduction of P. parasitica did not favour the rapid development of disease. In a further greenhouse experiment, root samples taken from plants inoculated with P. parasitica alone, P. brachyurus alone, or a combination of both were sectioned, stained and examined. While the development of P. parasitica was not found to differ when in close proximity to the feeding sites of P. brachyurus, colonization by the oomycete was reduced in the necrotic lesions caused by the nematode. This might offer some explanation as to why plants inoculated with P. parasitica 2–3 weeks before inoculation with the fungus developed few black-shank symptoms. Inagaki & Powell (1969) suggested that the simultaneous introduction of nematode and P. parasitica allowed the latter to utilize minute openings created by the migratory action of P. brachyurus in the roots. It seems unlikely, however, that the artificial wounding technique employed would be able to simulate the type of damage caused by invading PPN. In truth, such a treatment could have triggered any number of different processes to produce the disease symptoms seen, processes that are discussed below.
Histological studies appear to be the key to unravelling the association between fungal pathogens and the injuries caused to plants by PPN. This is particularly highlighted in the work of Polychronopoulos et al. (1969), where the invasion process of Heterodera schachtii (beet-cyst nematode) was found to facilitate the infection of sugar beet (Beta vulgaris) by the damping-off fungus Rhizoctonia solani. During their investigation, sugar beet seedlings grown in either nematode-infested or nematode-free soil were exposed to R. solani before being examined microscopically over a series of 12 h intervals for 3 days. On inspection of the seedlings 36 h after inoculation, distinct differences could be seen between the two treatments. When in combination with H. schachtii, the hyphae of R. solani were found to grow vigorously through the epidermis and cortex. Closer examination showed that hyphal colonization frequently followed tracts made by invading nematode juveniles. On the epidermal surfaces of the seedlings, the pathogen was found to produce fewer infection cushions in the presence of nematodes than when it was present alone. The authors suggested that infection cushion synthesis could have been hindered in some way by the invading nematodes. However, nematode invasion sites may provide R. solani with the necessary portals for penetration and entry, consequently reducing the need for developing more sophisticated infection structures such as infection cushions. In a similar way, R. solani is known to exploit natural openings on the outer surfaces of plants such as stomata (Chand et al., 1985) and lenticels on potato tubers (Ramsey, 1917) to invade underlying tissue.
Nematode wounding damage has also been found to be fundamental to several other disease complexes. For example, a recent scanning electron microscope (SEM) study on banana roots (Orion et al., 1999) revealed that the mycelium of an unidentified soilborne fungus was closely and frequently associated with the invasion tracts and lesions created by the spiral nematode Helicotylenchus multicinctus. On potatoes, Storey & Evans (1987) have found that the pathology of the wilt fungus V. dahliae is dependent on the timing of invasion by the cyst nematode G. pallida, and also on the potato cultivar. The fungus was found to enter and use the invasion channels of juvenile G. pallida if it was introduced at the same time as the nematode on cvs Pentland Javelin, Maris Peer and Maris Anchor. On both Maris Peer and Maris Anchor, lignified tissue developed following the invasion of the nematodes. If the fungus was introduced 8 days after the nematode, wilt symptoms caused by V. dahliae were less severe than on plants treated with the fungus alone. When sections of root tissue were examined, it could be seen that the lignified tissue had sealed the nematode penetration sites. In fact, the living tissue was partially protected by areas of woody tissue where this response had taken place. This response was not produced by cv. Pentland Javelin after nematode invasion, and V. dahliae was still able to colonize roots to a larger extent than on plants where it was introduced alone.
In addition to the cavities caused during PPN invasion, nematodes produce other forms of mechanical damage to plant roots that are open to exploitation by soilborne fungi. Fagbenle & Inskeep (1987) used SEM to study concomitant infections of Meloidogyne hapla and R. solani on peanut (Arachis hypogaea). Eleven weeks after inoculation, the root galls were often found to be split, leaving a rough surface comprised of cortical cells which rapidly became colonized by R. solani. However, it was unclear whether the cracks of the galled roots aided the penetration of R. solani.
In order for female cyst and root-knot nematodes to reproduce, the females/cysts must rupture through the root cortex to allow the vermiform males to fertilize them (Fig. 1). This event often produces a number of cracks and crevices where the swollen female has emerged. Several authors have suggested that these openings might be used by opportunistic pathogens to reach the underlying tissue of roots more easily (Bergeson, 1972; Golden & Van Gundy, 1975; Evans & Haydock, 1993).
The hypothesis that PPN-induced wounds facilitate the invasion process of some fungal pathogens seems the most likely explanation behind a synergistic interaction, although there are relatively few reports that demonstrate this mechanism, and positive quantitative data coupled with convincing histological evidence are required to validate this hypothesis. Future work in this area might benefit from using equipment such as time-lapse and video-enhanced light microscopy together with some form of image analysis system. Video-enhanced light microscopy has previously been used for studying nematodes (Wyss & Zunke, 1992) and fungi (McCabe et al., 1999). The application of image analysis systems could help overcome the problems of quantifying the density of fungal pathogens in regions of nematode damage.
Nematode-induced physiological changes to the host plant
The feeding sites of sedentary endoparasitic nematodes (giant cells or syncytia) are zones of high metabolic activity, having a large number of Golgi apparatus and mitochondria, while the cytoplasm is dense and contains many ribosomes (Jones, 1981). It is therefore no surprise that these nutrient-rich cells should become the substrate for fungal colonization (Meléndez & Powell, 1970; Wajid Khan & Muller, 1982; McLean & Lawrence, 1993; Abdel-Momen & Starr, 1998). Mayol & Bergeson (1970) made preliminary observations of this effect when tomato plants in gnotobiotic culture were treated with M. incognita and a soil suspension taken from around the roots of healthy tomato plants. Disease assessments of these plants 7 and 12 weeks after planting showed extensive necrosis of their root systems. Fungi isolated from the galls of M. incognita were identified as Trichoderma sp., Fusarium sp. and R. solani. Similarly, Negrón & Acosta (1989) found that F. oxysporum f.sp. coffeae caused increased root necrosis and chlorosis on the foliage of coffee plants (cv. Bourbón) if the plants had been inoculated with M. incognita 2 or 4 weeks previously. Sections taken from the roots of plants preinoculated with M. incognita were found to be colonized by F. oxysporum f.sp. coffeae in a uniquely different way from those where the fungus and nematode were either inoculated simultaneously, or where the fungus was inoculated alone. In the former case, the hyphae of F. oxysporum f.sp. coffeae were found to be abundant in the xylem vessels, giant cells and female nematodes. Giant cells, colonized by the fungus were in varying states of disrepair, with depleted or partially depleted contents. In comparison, plants that had been simultaneously inoculated with M. incognita and F. oxysporum f.sp. coffeae had fewer giant cells colonized by the fungus and no hyphae within the xylem. Taylor (1990) suggested that the 3–4-week nematode preinoculation, found to be critical in investigations of nematode–fungus disease complexes (Golden & Van Gundy, 1975; Wajid Khan & Muller, 1982; Negrón & Acosta, 1989), could be linked to syncytial development which, under optimal conditions in a susceptible host, will take 3–4 weeks to reach peak activity.
Perhaps the most comprehensive studies into such localized nematode-induced modifications are those of Golden & Van Gundy (1975). In their field studies with okra and tomato, fumigants were applied to field plots to reduce M. incognita and R. solani (ethylene dibromide and methyl bromide, respectively) and create independent and combined treatments. In untreated plots (plots with both M. incognita and R. solani), R. solani was isolated from the galls of M. incognita a week after gall formation. Two weeks later, numerous black sclerotia were found encrusted to the galls. In contrast, sclerotia were absent from ungalled regions of the roots. Four weeks after gall formation, substantial root decay occurred. Furthermore, histological sections revealed that R. solani had penetrated cells from the sclerotia attached to the gall surfaces. Rhizoctonia solani appeared to have a marked trophic intercellular pattern through the cortex of galled roots towards nematode-induced giant cells. Wajid Khan & Muller (1982) reported similar observations with M. hapla and R. solani on radish. Galls infected by R. solani had abundant sclerotia on their surfaces, while giant cells were colonized extensively by hyphae. Similarly, Abdel-Momen & Starr (1998) found that a reduction in the pod yield of peanut was significantly greater in coinfections of Meloidogyne javanica and R. solani where a concentration of fungal growth was found around the galled regions.
Cyst nematodes also form nutrient-rich syncytia for the purpose of development. Histological studies of H. schachtii-infested sugar beet seedlings exposed to R. solani indicated that syncytia were a more favourable substrate to the fungus than normal cells (Polychronopoulos et al., 1969). The authors also describe how the syncytium appeared to be a suitable ‘food base’ for colonization of other tissues by the fungus. Hyphae were seen to spread from the syncytia to the corticovascular tissue, which had not been invaded by nematodes.
These studies indicate that nematode-infected plant tissue may be actively selected by certain plant pathogens. According to Taylor (1979); Taylor (1990) and Abawi & Chen (1998), syncytia or giant cells contain higher levels of total protein, amino acids, lipids, DNA and sugars which would be beneficial to many fungi. This would support the suggestion that nematode infection enhances the nutritional composition of portions of plants to fungi, but the relationship remains unproven.
As well as these localized effects, some authors (Bowman & Bloom, 1966; Batten & Powell, 1971; Hillocks, 1986) have supported the notion that nematode-induced physiological changes can be systemic. In such cases, the nematode-induced factors or substances beneficial to fungi are hypothesized to be translocatable within the plant (Wajid Khan, 1993). While investigating this process, Bowman & Bloom (1966) and Hillocks (1986) employed a ‘split-root’ technique whereby the root system of the plant of interest was bisected into two separate containers, and one half of the root system was infested with the interacting nematode species while the other was inoculated with the interacting fungal pathogen. Bowman & Bloom (1966) infested one half of a tomato root system with nematodes (M. incognita) and the other half with F. oxysporum f.sp. lycopersici. Their results revealed that disease development on plants was dependent on being exposed to both M. incognita and the fungus. Regardless of the results, these split-root experiments do not identify systemic physiological modifications or any other mode of interaction. They simply indicate that an interaction might exist. Indeed, it could equally be concluded that the effects seen were a result of nematode occupation causing a loss of resistance or plant stress. Further studies involving critical biochemical analysis of plant material taken from plants either infested or uninfested with nematodes would significantly increase understanding of nematode-induced systemic change. For example, the nutritional quality of plants infested with PPN may prove more favourable to fungal pathogens. By determining the nutritional requirements of the fungus and quantifying these metabolites in nematode-infested and uninfested plants, a clearer picture could be obtained. Equally, PPN infestations may reduce the levels of fungitoxic compounds. Perhaps the sequence of events that leads to systemic induced changes is similar to, or the reverse of, that described in plants with systemic acquired resistance (SAR). Unfortunately, there has been little, if any work to address this question.
Modifications within the rhizosphere
The release of plant root exudates is considered an important factor in the attraction of both soilborne fungi (Flentje, 1957; Reddy, 1980; Grayston et al., 1997) and PPN (Klinger, 1965; Clarke & Hennessy, 1987). There are a number of ways in which PPN might influence the release of root exudates and thus alter the subsequent response of soilborne pathogens. First, the damage inflicted on plant roots during the process of PPN invasion could result in greater volumes of root exudates attractive to fungal invaders. Second, certain potato cultivars have been shown to produce increased numbers of lateral roots in response to invasion by potato cyst nematodes (Evans & Stone, 1977). Such an increase in root surface area may give rise to increased production of root exudates. Finally, PPN infestation may influence the chemical profile of the root exudates released, making them more favourable to fungal pathogens (Bergeson, 1972).
Perhaps the classic examples of this process are those involving root-knot nematodes. The aggregation of fungi, primarily R. solani, around root-knot galls of many plants has drawn attention to changes occurring within the rhizosphere. Golden & Van Gundy (1972) made preliminary observations of this effect during their studies of the M. incognita–R. solani complex of tomato. Tomato roots infested with M. incognita were seen to become more susceptible to fungal attack by R. solani with increasing age. By adopting the cellophane membrane techniques of Kerr (1956) and Flentje (1957), they observed that R. solani would aggregate on cellophane which was directly opposite the galled regions of the roots. In contrast, ungalled areas received only sparse mycelial coverage. In a later publication, Golden & Van Gundy (1975) undertook further studies with semipermeable membranes (cellophane) on tomato and okra infested with M. incognita. Introduction of R. solani (via mycelial plugs) to the external surfaces of the cellophane once again produced sclerotia opposite the galls of M. incognita. Microscopic examination of the sclerotia showed that their formation consisted of irregular branching and interwinding to form loosely constructed, undifferentiated structures. From these studies, the authors concluded that metabolic leakage from the galls of M. incognita could explain the elevated attraction of R. solani.
Van Gundy et al. (1977) undertook an extensive investigation using tomato to evaluate the hypothesis regarding metabolic leakage. First, a technique known as ‘double-root’ was utilized, whereby a secondary root system was induced to allow experimentation on the natural or primary root system. The attraction of R. solani to nematode-infested plants was facilitated by the use of a hydroponic system to remove root leachates. Tomato plants either infested with M. incognita; exposed to R. solani; infested with M. incognita and exposed to R. solani; or left untreated were found to be free of root necrosis after 5 weeks under the hydroponic regime. However, when leachates taken from M. incognita-infested roots were applied to plants exposed to R. solani alone, necrosis developed. Conversely, treatment of R. solani-treated plants with leachate from untreated plants did not result in root necrosis. Furthermore, it was found that if the experiment was repeated in the absence of the hydroponic system, plants would develop root rot only when exposed to both organisms. The results of these studies implied that M. incognita-infested plants were producing some form of attractant for R. solani. When the properties of exudates emanating from the nematode-infested roots were examined, they were found to have elevated levels of 14C metabolites. During the time of sclerotial development, 14–21 days following nematode invasion, the major constituents of the 14C-labelled metabolites were nitrogenous compounds such as amino acids and proteins; such nitrogenous compounds are important in the virulence of R. solani (Weinhold et al., 1972).
Bergeson et al. (1970) recorded a marked increase in the number of propagules of F. oxysporum f.sp. lycopersici found around the galls of M. javanica. Galls produced by M. javanica on tomato roots stimulated the number of spores of F. oxysporum f.sp. lycopersici while reducing the number of cells of actinomycetes antagonistic to Fusarium (Bergeson, 1972).
Typically, loss of resistance has been tested with the application of split-root methods as described earlier. Investigators have adopted this type of approach to determine whether the loss of pathogen resistance induced by nematode infestation occurs as a result of the breakdown of a systemic chemical defence system within the host plant. Bowman & Bloom (1966) found that the tomato cvs Rutgers and Homestead, previously resistant to F. oxysporum f.sp. lycopersici, developed symptoms of wilt during split-root experiments with M. incognita. Further studies (Sidhu & Webster, 1977) using root layering and grafting techniques confirmed that a nematode-induced factor could be passed through a resistant scion (a graft from a resistant tomato cultivar) and render it susceptible to F. oxysporum f.sp. lycopersici. In contrast, resistant scions in tomato plants free from M. incognita infestation could block infection by F. oxysporum f.sp. lycopersici. Vargas et al. (1996) observed a similar effect on chilli (Capsicum annuum), where the nematode Nacobbus aberrans caused a loss of resistance to Phytophthora capsici even if the nematode and oomycete were physically separated on split roots. While few reports have addressed how this might occur, Marley & Hillocks (1994) demonstrated that nematode-induced loss of resistance to Fusarium udum in pigeonpea (Cajanus cajun) was associated with reduced levels of the isoflavanoid phytoalexin cajanol. Previously, Marley & Hillocks (1993) determined that the rapid accumulation of cajanol in some pigeonpea cultivars was responsible for conferring resistance to the pathogen. However, cajanol content was 62% lower and resistance was lost during combined infections of F. udum, M. incognita and M. javanica where wilt disease incidence and severity were significantly higher than in plants inoculated with F. udum alone. Although this study clearly shows that nematode infestation reduced a chemical defence mechanism to fusarium wilt in pigeonpea, it is still not known how nematode activity modified the plant. Marley & Hillocks (1994) suggested that either the overall metabolic rate of the plants was reduced, or specific changes were made to the synthesis of isoflavonoids during nematode attack.
There is general agreement among some authors (France & Abawi, 1994; Sugawara et al., 1997) that polygenic resistance is comparatively less stable than monogenic resistance. Francl & Wheeler (1993) state that plants with polygenic resistance to fungal pathogens are frequently found to become susceptible to fungal attack during nematode infestations, whereas plants with a single dominant gene for resistance are rarely affected. This was observed by Abawi & Barker (1984) on tomatoes, where resistance to F. oxysporum f.sp. lycopersici was disrupted by infestations of M. incognita on cultivars with polygenic resistance, but not on those where resistance was expressed by a dominant single I-gene. Transgenic plants involving quantitative trait loci may have a greater capacity for providing durable resistance in the presence of interacting fungal pathogens and PPN.
Wajid Khan (1993) postulated that resistant plants are rendered vulnerable to pathogens via physiological alterations made by the nematode, which have no effect on the gene(s) responsible for encoding resistance. For example, the process of invasion by PPN may provide soilborne pathogens with portals (Powell & Nusbaum, 1960) through a previously impenetrable physical barrier selected for in a plant breeding programme.
Just as nematode activity can increase the severity of diseases caused by fungal pathogens, so nematode populations can be elevated during concomitant infections with root-infecting pathogens (Vrain, 1987; Taheri et al., 1994). While there are far fewer reports of such phenomena, several hypotheses have been proposed. For example, Zahid et al. (2002) investigated a large and complex set of interactions between 12 species of root- and stolon-infecting fungi and three species of root-colonizing nematodes on white clover (Trifolium repens), an important forage crop for dairy herds in eastern Australia. Although a good number of interactions were found between the various combinations of nematodes and fungi, the root-knot symptoms of Meloidogyne trifoliophila were commonly found to be increased in treatments containing the stolon-infecting fungus Drechslera halodes. Moreover, the final densities of M. trifoliophila and two other nematode species (Helicotylenchus dihystera and Heterodera trifolii) were significantly increased in plants infected with D. halodes. This case exemplifies the problems of defining disease aetiology where a large number of pests and diseases are present. Previously, Faulkner & Skotland (1965) observed that Pratylenchus minyus reached its reproductive peak at the same time as the maximum expression of wilt disease (V. dahliae f.sp. menthae) on peppermint plants (Mentha piperita). The authors suggested that V. dahliae may produce root growth-promoting substances such as indole-3-acetic acid (as previously recorded for V. albo-atrum; Pegg & Selman, 1959), resulting in an enlarged root system, releasing greater volumes of root exudate and thereby attracting more PPN (Clarke & Hennessy, 1987; Rolfe et al., 2000). Some studies illustrate how plant roots infected with fungal pathogens can be more attractive to PPNs. For example, Nordmeyer & Sikora (1983) considered how the attraction of Heterodera daverti might be affected by Fusarium avenaceum infection in clover (Trifolium subterraneum) seedlings when compared to uninfected plants. In vitro experiments showed that a significantly greater proportion of H. daverti migrated towards diffusates from F. avenaceum-infected clover roots than towards diffusates from healthy plants.
Fungal modifications to PPN host-finding has also been investigated by Edmunds & Mai (1967), who showed that P. penetrans would conglomerate around a CO2 source under in vitro conditions, in agreement with the previous findings of Bird (1959) and Klinger (1965). CO2 measurements taken from alfalfa plants (Medicago sativa) infected with Trichoderma viride and, particularly, F. oxysporum were considerably higher than those found in healthy roots. Elevated levels of CO2 from infected plants may have contributed to the increased attraction of P. penetrans towards alfalfa roots previously seen in earlier experiments (Edmunds & Mai, 1966a; Edmunds & Mai, 1966b). Whether or not CO2 emissions would produce a similar nematode response within the natural soil environment remains open to debate, and subject to further experimentation. Moreover, the subject of nematode orientation is still largely unknown in the field of plant nematology.
Several authors have speculated that nematode penetration is increased in plant roots previously subjected to the enzymes of fungal pathogens (Edmunds & Mai, 1966a; Edmunds & Mai, 1966b; Edmunds & Mai, 1967; Nordmeyer & Sikora, 1983). While the results of Edmunds & Mai (1966b) were largely inconsistent, Nordmeyer & Sikora (1983) reported more convincing findings in their study of H. daverti and F. avenaceum in the roots of clover. By treating clover roots with filtrate from F. avanaceum cultures, a significantly higher number of H. daverti juveniles were recovered than from untreated roots. They also observed that the duration of exposure of clover roots to F. avenaceum was critical to the invasion success of H. daverti, with an optimal exposure time of 450 s. As the fungal enzyme pectinmethylesterase was detected in the filtrates of F. avenaceum, the authors suggested that such a cell wall-degrading enzyme may have enabled H. daverti to penetrate more easily.
Some research has suggested that fungal infections cause a deterioration or breakdown of plant resistance to nematode attack. Hasan (1985) encountered this effect during routine field screening of chilli pepper cultivars and lines which, under controlled greenhouse conditions, had shown promising resistance to M. incognita. In this case, resistance was lost in two out of five previously fully resistant and eight out of 16 previously moderately resistant lines. Furthermore, individual plants subject to resistance loss expressed symptoms of collar rot and damping-off diseases. The disease-inducing soilborne pathogens were isolated and positively identified as R. solani and Pythium aphanidermatum. Subsequently, a greenhouse experiment was devised to test the effect of R. solani and P. aphanidermatum on the resistance of cvs Jawala (resistant) and Longthin Faizibadi (moderately resistant) to M. incognita. On both cultivars, the presence of R. solani or P. aphanidermatum caused a significant increase in the reproductive capacity (number of egg masses and eggs produced) of M. incognita. In terms of resistance, the ratings of Jawala and Longthin Faizibadi were demoted to moderately resistant and susceptible, respectively. The exact mechanism of this phenomenon was not determined, but the activities of pathogen-produced enzymes may have compromised the physical barrier conferred on the resistant chilli lines, or chemical defences, such as the antifeedant proteinase inhibitors described by Lilley et al. (1999), may have been disrupted during nematode infections. New technologies such as proteomics may provide a better understanding of such interactions, but in the meantime the interactions between PPN and soilborne pathogens need to be considered in future plant breeding programmes.
Factors affecting synergistic interactions
As with nearly all investigations in science, many reports on disease complexes contradict one another. While some of these disparities might be explained by experimental procedure and accuracy, there are findings that highlight the specificity of certain disease complexes and the influence of biotic and abiotic factors on them. This is exemplified by studies on the V. dahliae–Pratylenchus complex of potato, where it has been found that the interaction between these organisms varies among different nematode species (Riedel et al., 1985) and populations, as well as fungal genotypes (Botseas & Rowe, 1994). For example, Riedel et al. (1985) and Bowers et al. (1996) observed that potato early dying disease (V. dahliae) was enhanced by populations of P. penetrans but not by P. crenatus or P. scribneri. Furthermore, greenhouse experiments undertaken by Hafez et al. (1999) demonstrated that populations of P. neglectus collected from Ontario, Canada would interact synergistically with V. dahliae, while populations of P. neglectus from Parma, Idaho did not increase disease or yield loss any more than treatment with V. dahliae alone. Restriction analysis of the ITS1 region on rDNA gene from nematodes from the Canada and Idaho populations revealed unique fragments for each population, implying variation within this species. The differences between nematode species and populations/pathotypes in their ability to accentuate V. dahliae wilt are likely to be related to their proficiency as parasites on potato. This is certainly evident from the work of Hafez et al. (1999), where the fecundity of the Canadian population was approximately 50% higher than that of the Idaho population, in which initial populations (Pi) were 5000 and 10 000 nematodes per 5000 cm3 soil. It is also possible that species and pathotypes of Pratylenchus spp. may favour different physical environmental conditions. Indeed, Kimpinski & Willis (1981) recorded differential effects of soil pH and temperature between populations of P. penetrans and P. crenatus.
Fungal genotype can also affect potato early dying complex, as shown by Botseas & Rowe (1994), who found that two pathotypes of V. dahliae vegetative compatibility group 4 (VCG 4) formed different relationships with P. penetrans. In greenhouse and field microplot experiments, the two pathotypes of V. dahliae (VCG 4A and B) exhibited no differences in aggressiveness when inoculated alone. However, plants inoculated with V. dahliae VCG 4A and grown in soil infested with P. penetrans had higher levels of disease severity, lower tuber yields and earlier senescence than plants inoculated with V. dahliae VCG 4B in the presence of P. penetrans. This type of specificity has also been reported by Johnson & Santo (2001), who found that P. penetrans would interact synergistically with V. dahliae VCG 2B but not with VCG 4A on cultivated peppermint and Scotch spearmint. In these findings, the governing factor was host-specific fungal aggressiveness. Although Botseas & Rowe (1994) were unable to detect differences in aggressiveness between V. dahliae VCG 4A and B, a number of investigations have reported that VCG 4A is the more aggressive of the two pathotypes on potato (Joaquim & Rowe, 1991; Strausbaugh, 1993; Omer et al., 2000). This is also relevant to the work on mint, where previously Douhan & Johnson (2001) described VCG 2B as the most aggressive pathotype. These types of specific interactions are likely to be applicable to other disease complexes, such as those involving R. solani, where 12 or more anastomosis groups exist (Carling, 1996; Carling et al., 1999).
As mentioned above, fluctuating environmental parameters are often found to affect one or more of the interacting organisms in a disease complex. Temperature has been found to be critical in some nematode–fungus interactions (France & Abawi, 1994; Walker et al., 2000), but not in others (Griffin et al., 1993; Uma Maheswari et al., 1997; Walker et al., 1999). Interestingly, France & Abawi (1994) observed that a bean genotype with dual resistance against F. oxysporum f.sp. phaseoli and M. incognita exhibited visible wilt symptoms if both organisms were inoculated together at 27°C. It was suggested that the high temperature caused a breakdown of resistance to M. incognita and, in turn, the nematodes broke the resistance to the wilt-inducing pathogen.
Soil type has been shown to have no influence over disease complexes involving M. hapla and Phytophthora megasperma f.sp. medicaginis on alfalfa and R. solani, and M. javanica on soyabean (Griffin et al., 1993; Agu & Ogbuji, 2000). Conversely, Uma Maheswari et al. (1997) state that soil type can affect interactions between F. oxysporum f.sp. ciceri and M. javanica on chickpeas. According to their findings, the development of wilt disease by F. oxysporum f.sp. ciceri was higher in a clay soil (48% clay) while M. javanica caused greater plant damage in a loamy sand. On the basis of these results, heavier textured soils appear to be unsuitable for root-knot nematodes, and therefore the establishment of disease complexes with F. oxysporum f.sp. ciceri would be improbable. From the many studies on nematode and fungal epidemiology, it seems highly plausible that additional environmental factors such as soil pH (Rupe et al., 1999), soil moisture and meteorological conditions (Robinson et al., 1987) will also be linked to the development of a disease complex, depending on the organisms involved. The impact of abiotic factors on nematodes and soilborne pathogens emphasizes the importance of undertaking field experiments to test for synergistic interactions. Evans & Haydock (1993) discussed the validity of pot-based and laboratory experiments and argued that the ‘acid test’ for determining the agricultural significance of an interaction is field experimentation.
Indirect effects of disease complexes on interacting organisms
Thus far in this review, nematode–pathogen complexes have been discussed only with reference to the type and level of damage caused to the plant host. However, this is not the only outcome of these synergistic relationships, as either nematode or fungus can be indirectly affected during their cohabitation on a mutual host. Direct antagonistic interactions, which involve the parasitism of nematodes by soilborne fungi, have been extensively studied and reviewed (Kerry, 2000). The indirect effects that fungi exert on nematodes in disease complexes are less well known, yet remain important in terms of future nematode multiplication.
Earlier, work was described in which feeding sites of sedentary endoparasites were shown to be preferable substrates for plant pathogenic fungi (Polychronopoulos et al., 1969; Negrón & Acosta, 1989; Abdel-Momen & Starr, 1998). In all these investigations, the nematode syncytia or giant cells were disrupted or damaged to some extent during fungal colonization. Fattah & Webster (1983) monitored changes within M. incognita giant cells during coinfection of tomato roots with F. oxysporum f.sp. lycopersici and M. incognita. Transmission electron microscopy of root tissue, 3 weeks after infection with the fungus, revealed that the giant cells of M. incognita had become smaller and spherical. Chromatin within the nuclei of these cells had condensed along the nuclear membrane, which was partially fragmented and swollen. In contrast, the giant cells of plants inoculated with M. incognita alone were much larger and had well defined membranes, large nuclei and dense cytoplasm. The authors stated that successful assemblage and maintenance of giant cells is vital for the growth and reproduction of root-knot nematodes, and that destruction of large proportions of giant cells will result in the premature death of female nematodes. Consequently, these types of disturbance are likely to affect the development of subsequent nematode populations.
In addition to fungal disruption of nematode feeding sites, plants affected by disease complexes may be more prone to early senescence and death (Griffin et al., 1993; Walker et al., 1998) which, in turn, may prevent nematodes from completing their life cycles. Competition for nutrients (Jorgenson, 1970) or root space (Ketudat, 1969) may be responsible for the decline of nematode populations, although these concepts appear to be difficult to demonstrate. However, soyabean plants affected by the H. glycines–F. solani complex (sudden death syndrome) have been found to have stunted root systems (Rupe et al., 1999), which might explain the lower final populations of H. glycines found.
As previously indicated, PPN can induce systemic changes within their host plants to render them susceptible to fungal attack. The reverse of this effect is that nematode invasion activates some form of hypersensitive response (Plowright et al., 1996) or triggers a mechanism of SAR (Sticher et al., 1997). These types of plant response are widely recognized for the protection they provide plants against subsequent attacks (Sticher et al., 1997; Van Loon et al., 1998), but to date no studies have been undertaken that demonstrate SAR to pathogen infection induced by nematode invasion.
Synergistic interactions between PPN and fungal pathogens are undoubtedly complex affairs. The research reviewed here underlines the necessity of understanding individual disease complexes before deciding on appropriate control methods. Sampling soils for known interacting organisms may assist in forecasting potential disease problems. Applications involving remote sensing and digital image analysis are currently being refined and used for the determination of spatial distributions of both plant pathogens (Nilsson, 1996) and PPN (Heath et al., 2000) in crops. Further development of this type of technology is likely to be invaluable for the prediction of disease complexes. It would also be interesting to know whether there are any correlating patterns in spatial population densities between interacting organisms.
The specificity of some of the disease complexes mentioned indicates the need to use appropriate diagnostic measures to determine whether management strategies are suitable. For example, the findings of Botseas & Rowe (1994) showed that P. penetrans would form a disease complex with V. dahliae VCG 4A and not V. dahliae VCG 4B. Dobinson et al. (2000) demonstrated that VCG 4A and 4B could be characterized with the use of molecular markers. The absence of a subspecies-specific repetitive DNA sequence (E18) and differential restriction fragment length polymorphisms (RFLPs) in the nuclear rDNA and Trp1 loci allowed VCG 4A to be distinguished from 4B with relative success. In addition, these authors claim that marker analysis could be employed to identify V. dahliae VCG 4A from either soil or plant tissues, and there is potential for the development of a PCR assay. Numerous other complexes would undoubtedly benefit from adopting modern molecular-based diagnostic techniques.
Management of disease complexes appears to be less straightforward than one might anticipate. The most obvious solution is to use chemical methods to control one of the interacting organisms and thus prevent the disease complex from occurring. However, it is fundamental to have prior knowledge of the interaction concerned, as even low densities of fungi or nematodes can result in a disease complex of significant importance (Bowers et al., 1996; Saeed et al., 1998). Consequently, the reduction of one pathogen may not resolve the problem of the interaction. Wheeler et al. (2000) conducted a survey on the management choices made by cotton producers in Texas, USA and found that in fields affected by both T. basicola and M. incognita, growers generally applied higher rates of the nematicide aldicarb (c. 30–50% higher) than where the organisms were found alone. Although these statistics suggest that the growers were aware of the damaging effects of this disease complex, according to the authors, there was no rationale behind the decisions made. This study illustrates the need for further research into the development of targeted approaches to disease complexes. Furthermore, with current public awareness and education regarding the nontarget effects of agrochemicals, there is an increasing demand to reduce their usage.
Encouragingly, a number of alternative strategies have been investigated for the management of disease complexes. In plant breeding, several studies (Chang et al., 1997; Meksem et al., 1999; Prabhu et al., 1999) have concentrated on characterizing the loci responsible for dual resistance in soyabean against F. solani and H. glycines involved in the sudden death syndrome. Cultural practices such as multiyear cropping regimes (Chen et al., 1995) and soil solarization (Lazarovits et al., 1991) have had variable success in reducing population densities of V. dahliae and P. penetrans. Finally, the detrimental effects of a disease complex on pigeonpea involving the sedentary endoparasite Heterodera cajani and the fungus F. udum were reduced following application of the fungi Paecilomyces lilacinus and Verticillium chlamydosporium together with the vesicular arbuscular fungus Gigaspora margarita (Siddiqui & Mahmood, 1995). This type of integrated approach, where both interacting organisms are tackled, appears to be the most promising way of defeating disease complexes involving nematodes and fungi.