SEARCH

SEARCH BY CITATION

Keywords:

  •  biological control;
  • consecutive epidemics;
  • disease dynamics;
  • continuous cropping;
  • Rhizoctonia solani;
  • inoculum dynamics;
  • saprotrophic

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • • 
    A combination of experimentation and modelling is used to examine the role of epidemiological dynamics on the production and infectivity of inoculum and the efficiency of biocontrol by Trichoderma viride during consecutive epidemics of damping-off disease caused by the pathogen Rhizoctonia solani in crops of radish.
  • • 
    Changes in the net infectivity of inoculum at the beginning of first and second crops caused a switch in epidemiological dynamics. Epidemics of first crops were dominated by secondary infection leading to amplification of inoculum so that epidemics of second crops were overwhelmingly determined by primary infection.
  • • 
    The biocontrol agent reduced primary infection and hence parasitic amplification of inoculum in both first and second crops but the efficiency of control dropped from 91.7% in first crops to 64.8% in second crops, with sudden outbreaks of disease in second crops which had previously been disease-free.
  • • 
    We conclude that parasitic amplification can cause a rapid build-up of disease and inoculum over consecutive crops, leading to loss in the efficiency of biocontrol. This form of inoculum production is supplemented by saprotrophic infestation which can result in sudden outbreaks of disease in protected crops where control of disease had previously been fully successful.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Many arable and, in particular, horticultural systems involve growing crop monocultures in quick succession. The risk of a soil-borne epidemic when crops are grown under such intensive regimes is initially high, and often results in the very rapid build-up of disease from one season to the next (Hide & Read, 1991; Gilligan et al., 1996). Biological control provides an environmentally appealing method of reducing disease yet, even when disease is relatively well controlled in one season, severe disease patches can develop in the next (Schneider et al., 2001). Under field conditions, variation in the severity of epidemics (MacNish, 1996; Schneider et al., 2001) and in the efficiency of biocontrol (Weller, 1988) is commonly observed under a regime of continuous cropping. Much of the variability is undoubtedly caused by environmental heterogeneity in the form of differences in soil type, soil temperature and soil moisture. But underlying this are other stochastic influences, driven by the demographic interactions of the pathogen, root and microbial populations (White & Gilligan, 1998). These demographic factors also determine the course of an epidemic and the outcome of disease control. Consequently, a major challenge in devising a coherent theory for the performance of biological control in botanical epidemics is first, to understand the underlying demographic mechanisms driven by the host–pathogen–biological control complex that together influence the dynamics of disease during successive seasons. Here we initiate this by analysing replicated epidemics in microcosms in which we minimize environmental variation within and between epidemics on consecutive crops.

The production and survival of inoculum in a preceding crop and during the intercropping period is central to the development of disease across seasons (Gilligan et al., 1996; Schneider et al., 2001). The production of inoculum depends on the dynamics of disease in a previous crop. We have shown elsewhere that this depends, in turn, on the balance between primary and secondary infection (Bailey & Gilligan, 1999), as well as changes over time in the susceptibility of the host (Kleczkowski et al., 1996; Gibson et al., 1999). This suggests a simple sequence of events, whereby the amount of initial inoculum determines the levels of primary infection in the first crop. Together with changes in the susceptibility of the host this, in turn, influences the rate of secondary infection and the final level of disease at the end of the first season. Following harvest, diseased tissue is converted into inoculum from which the residual, after decay between crops, initiates the epidemic in the succeeding crop. It follows then that the balance between primary and secondary infection in a first crop and the dynamics of inoculum between crops affects the relative amounts and infectivity of inoculum, and hence the balance between primary and secondary infection in a second crop.

In this paper we ask two broad questions: (1) How do the dynamics of disease differ between successive crops? and (2) How is this influenced by repeated applications of a biocontrol agent? We address these questions using a combination of modelling and experimentation in microcosms filled with sand to provide replicated epidemics of damping-off disease caused by the soil-borne plant pathogen Rhizoctonia solani Kühn on radish (Raphanus sativus L.) in the presence or absence of the biological control agent Trichoderma viride Pers. Ex. Gray. The simplicity of sand omits many of the biological processes, interactions and variability associated with field soils, but improves repeatability and allows us to identify and quantify the main epidemiological features responsible for the spread of disease. For this system we ask the following sequence of questions: (i) What is the balance between primary and secondary infection within the first and second crops? (ii) How does the addition of the biocontrol agent affect disease in the first crop? (iii) Does the presence of the biocontrol agent in the first crop affect disease dynamics in the second crop? (iv) Does this in turn affect the likelihood of successful control in the second crop?

We use the results of these analyses to construct a schematic framework to account for the evolution of epidemiological dynamics across successive crops, and the consequences for biological control.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Microcosm experiments

Population experiments were conducted using 10 replicate microcosms without and 10 microcosms with the addition of T. viride, in which 50 radish plants were exposed to known amounts of inoculum of R. solani. Throughout the experiments considerable care was taken to maintain identical conditions among the replicates. This ensured that differences in disease dynamics between replicate epidemics were attributable to the location of initial inoculum, the inherent stochasticity of the transmission of infection, and the survival of inoculum between seasons in the presence or absence of a biological control agent. Two additional uninfested microcosms were planted to monitor for contaminant R. solani or any other organisms.

Microcosms comprised clear plastic boxes measuring 100 mm wide, 200 mm diameter and 100 mm depth to which 1 kg of white, acid-washed sand (grade 16/30; Hepworth Minerals and Chemicals Ltd, Redhill, Surrey, UK) was added with a moisture content of 10% by weight. Fifty radish seeds (cv. Cherry Belle) were sown 20 mm apart and 10 mm below the surface. Ten mycelial discs (1 mm in diameter) cut from the margin of a 5-d-old colony of R. solani (isolate R5, anastomosis group 2-1), growing on a Millipore filter over potato dextrose agar (Gilligan & Bailey, 1997), were introduced at randomly selected locations, 5 mm below the surface. Inoculum of T. viride, comprising previously colonized poppy seeds (Bailey & Gilligan, 1997), was added to half the replicates. A single unit of T. viride inoculum was introduced next to (within a distance of 1 mm of) each radish seed at the time of sowing. The boxes were sealed with clear plastic lids and incubated in a growth chamber at 23°C with a daylength of 16 h light. The cumulative numbers of seedlings damped-off were counted daily until day 20, when seedlings were cut at the sand surface and the tops removed. After a further 21 d, 50 seedlings were introduced at identical locations to the first crop. No additional inoculum of R. solani was added, but T. viride was introduced, as before, to the protected replicates. The cumulative numbers of seedlings damped-off were again counted daily for 20 d following the second sowing. There were 10 replicates of each treatment in the first season and eight in the second.

Distinction between primary and secondary infection

Plants can only become infected in one of two ways, by primary or secondary infection. We used the cumulative number of damped-off plants, I(t), over time as a measure of the change in the numbers of diseased plants that had been infected by primary infection, Ip(t), and by secondary infection, Is(t), thus I(t) = Ip(t) + Is(t). We used the spatio-temporal maps of the daily locations of newly diseased plants to distinguish primary from secondary infections using the methods of Otten et al. (2003). The time taken to transmit infection and express disease between an infected and a susceptible plant has previously been shown to be approximately 2 d (Bailey & Gilligan, 1997; Kleczkowski et al., 1997). It is also known that transmission occurs between nearest neighbouring plants (Kleczkowski et al., 1997; Otten et al., 2003), and that the probability of infection by primary inoculum decays to zero after 9 d in these microcosms (Bailey & Gilligan, 1997). Accordingly, primary infection was ascribed to those plants that damped-off during the first 9 d after sowing in locations not neighbouring a plant previously diseased for more than 1 d. All other infection is ascribed to secondary infection.

Model derivation

A mechanistic model was used to compare epidemics in consecutive crops and between treatments and to estimate differences in the density and infectivity of inoculum responsible for the initiation of disease, in first and second crops. Within each season primary infection is triggered by an initial number of inoculum units, X, in a single cohort of radish plants introduced into a system at time t = 0. Plants damped-off by primary infections become a source of infection as the fungus spreads to susceptible plants. For the infection of radish by R. solani, the spread of disease does not continue ad infinitum but is subject to two forms of decay as the fungus exhausts the nutrient supply, and host plants become increasingly resistant with age. The decay is reflected in the rate of primary infection, rp(t), from particulate inoculum and the rate of secondary infection, rs(t), decreasing with time with decay rates dp and ds, respectively. We define an initial delay, t0 d, between sowing and the first appearance of disease. This reflects the relative rates of germination of plants and pathogen. It may also differ in the first and second crops because of differences in the nature of the inoculum between the two crops. We assume that pre-emergence damping-off is responsible for the reduction in emergence of the second crop.

Changes in the total number of plants damped-off, I, over time, t, were described by:

  • image(Eqn 1a)
  • image(Eqn 1b)
  • image(Eqn 1c)

where N is the number of plants, and the remaining parameters and variables are defined above. As we could distinguish primary and secondary infections from the disease maps, we decoupled equation 1a to allow for the availability of separate variables in fitting the model to data. In practice, as Is is obtained by difference, I(t) − Ip(t), we fitted the model to data for the total number of infected plants and the number of primary infections:

  • image(Eqn 2a)
  • image(Eqn 2b)
  • image(Eqn 2c)
  • image(Eqn 2d)

The model solutions were used to estimate the contribution of primary and secondary infection in first and second crops. Differences in disease progress (primary and secondary infection) were interpreted according to change in the density and infectivity of inoculum preceding the first and second crops. For the first crop, the number of initial inoculum units is known, X = 10, therefore we can interpret the initial value of the rate of primary infections, rp(0) as the infectivity of the initial inoculum. For the second crop it is not possible to distinguish between the rate of primary infection, rp(0), and the density of particulate inoculum, X, as the latter is unknown and we cannot assume the new inoculum is of the same strength as the original one. However, combined differences in infectivity and density of inoculum can be compared across treatments and seasons, using the product αp = rp(0)X, representing the net infectivity of inoculum at the beginning of an epidemic. Differences in the efficiency of disease control on primary and secondary infection in consecutive epidemics were therefore assessed by comparing the relative values of αp.

Model fitting

Estimation of rp(0), dp, rs(0), ds and t0 for epidemics of first crops was performed by fitting model 2 simultaneously to both change in total numbers of damped-off plants and plants damped-off by primary infection over time. Maximum-likelihood estimations as well as 95% approximate confidence intervals were calculated using an MCMC technique, under an assumption of normal errors with the variance estimated from the replicate data. This method is essentially equivalent to weighted non-linear least squares. For the second crop we found that the rates of decay dp and ds were strongly correlated with the initial delay, t0, and the fit was substantially improved by fixing the rates at their values from the first crop.

The results shown for model 2 represent the best combination in terms of quality of fit as assessed by the Akaike information criterion (Williams, 2001).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

First crops

The averaged disease progress curves were sigmoidal in the first season, with final levels of disease affecting an average of 26.3 plants in unprotected radish crops and only 2.44 plants in protected crops (Fig. 1a). Model 2 was fitted to disease data to estimate the balance between primary and secondary infection and the net infectivity of inoculum, αp, at the time of first disease. The latter was estimated as αp = 0.097 for unprotected crops (Table 1) and primary infection was responsible, on average, for 6.1 (Fig. 1c) and secondary infection for 20.2 (Fig. 1e) damped-off plants. Trichoderma viride reduced the net infectivity of inoculum by more than an order of magnitude (92.6%) to αp = 0.0072 (Table 1), hence primary infection was responsible, on average, for 0.86 (Fig. 1c) and secondary infection for 1.4 (Fig. 1e) damped-off plants in the presence of T. viride. Rates of decay for primary and secondary infection, dp and ds, together with rs, were not significantly affected by the biocontrol treatment.

image

Figure 1. Curves describing change in the total (a,b), primary (c,d) and secondary (e,f) infection over time for first (a,c,e) and second (b,d,f) crops of radish unprotected (solid lines, open circles) or protected (broken lines, closed circles) by the biocontrol agent Trichoderma viride. Results of fitting models 1 and 2 for change in total, primary and (by subtraction) secondary infected plants are shown (bold lines). Because curves are shown for all replicates some overlap of data occurs, particularly for replicates with no disease.

Download figure to PowerPoint

Table 1.  Summary of fitting model 2 to data describing change in numbers of plants diseased (damped-off) by R. solani over time for consecutive epidemics of radish either unprotected or protected by the biocontrol agent T. viride
ParameterFirst cropSecond crop
–T. viride+T. viride–T. viride+T. viride
  1. Approximate upper and lower 95% confidence intervals of parameters estimates given in brackets.

Primary infectivity0.0970.00720.740.26
αp(0.073–0.123)(0.0017–0.0147)(0.70–0.78)(0.24–0.29)
Secondary rate0.00870.00650.00440.0076
rs(0.0064–0.0114)(0.0000–0.0174)(0.0033–0.0061)(0.0045–0.011)
Primary decay0.710.410.710.41
dp(0.54–0.98)(0.07–1.00)(fixed)(fixed)
Secondary decay0.160.200.160.2
ds(0.09–0.24)(0.0–1.12)(fixed)(fixed)
Starting time5.75.72.650.97
t0(5.5–5.8)(4.9–6.1)(1.17–3.34)(0–2.72)

Second crops

The onset of disease was markedly earlier in second crops, and there was more disease. Average disease progress curves were monomolecular is shape with, on average, 40.5 and 32.9 damped-off plants in unprotected and protected crops, respectively (Fig. 1b). Notably, replicates that were disease-free in the presence of T. viride in the first season produced between 21 and 43 damped-off plants in the second season.

Model 2 was fitted to disease data in the second crops to estimate the balance between primary and secondary infection in replicate epidemics, and net infectivity of inoculum. Epidemics in both unprotected and protected crops were dominated by primary infection. In unprotected crops, the net infectivity of inoculum was estimated as 0.74. Primary infection was responsible, on average, for 32.0 (Fig. 1d) and secondary infection for 8.5 (Fig. 1f) damped-off plants. Trichoderma viride reduced the infectivity of inoculum by 64.8% to 0.26 (Table 1) where primary infection was responsible, on average, for 23.1 (Fig. 1d) and secondary infection for 9.8 (Fig. 1f) damped-off plants. Moreover, six of the replicates in which disease was controlled completely in first crops developed high levels of disease in second crops, despite reintroduction of T. viride.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We have shown how a change in epidemiological dynamics affects the efficiency of biological control of R. solani during consecutive crops. This allows us to explain why control by T. viride may succeed in a first crop but fail in the second, even though the biocontrol agent is reintroduced with each sowing. The change in epidemiological dynamics is caused by differences in the net infectivity of inoculum preceding each crop. This leads to a switch in the balance between primary and secondary infection in the succeeding crop.

In unprotected first crops there was a rapid build-up of disease from a low initial density of inoculum. First crop epidemics were characterized by low levels of primary infection and high levels of secondary infection (Fig. 1). By contrast, second crops were characterized by a significant increase in primary infection, leaving relatively fewer plants available for secondary infection. This switch from secondary to primary infection is consistent with the large increase in net infectivity of inoculum detected before second crops were planted (Table 1).

Application of the biocontrol agent, T. viride, produced an estimated 10-fold reduction in the net infectivity of inoculum (Table 1), as well as a substantial reduction in average levels of disease in first crops. This included six out of 10 replicates for which no damping-off was detected. Where disease was initiated, epidemics were dominated by secondary infection (Fig. 1e) and control of disease was attributed to a reduction in primary infection (Table 1). This is consistent with a previous analysis by Gibson et al. (1999) for biocontrol of R. solani on radish in a single crop, based on fitting a stochastic model for primary and secondary infection. However, when the crop was resown and the control agent reapplied, the efficiency of control was reduced significantly in all replicates, and epidemics were dominated by primary infection (Fig. 1; Table 1). Moreover, in replicates in which disease was completely controlled in first crops, significant outbreaks of disease were detected in second crops, suggesting considerable amplification of inoculum in the absence of disease (damping-off). These results show that inoculum produced from diseased, damped-off plants alone was not the only reservoir of inoculum for infection of a second crop.

Two mechanisms contribute to the amplification of inoculum. The first occurs by parasitic growth during the cropping period and produces inoculum in the form of damped-off plants. That both primary and secondary infection can contribute to the parasitic amplification of inoculum is now evident for a wide range of soil-borne pathogens including, for example, infection of onions by Sclerotium cepivorum (Entwisle, 1990); wheat by Gaeumannomyces graminis (Bailey & Gilligan, 1999; Schoeny & Lucas, 1999); lettuce by Sclerotinia minor (Gubbins & Gilligan, 1997); and tomato by Fusarium oxysporum (Rekah et al., 2001). However, as we have demonstrated here, the precise contribution of primary and secondary infection to final levels of disease depends ultimately on the interaction between initial inoculum density, changes in host susceptibility, and the rates of disease transmission. The second mechanism for amplification of inoculum involves saprotrophic growth during the intercropping period, and produces inoculum in the form of colonized root debris. The competitive, saprotrophic ability of R. solani and other soil-borne plant pathogens has long been recognized (Garrett, 1970), and the contribution of saprotrophic growth to epidemics of plant disease is well documented (Papavizas, 1970), but the consequences for disease control have received little epidemiological analysis. These facultative parasites are capable of extensive saprotrophic multiplication on root residue remaining during the intercropping period, which means that production of inoculum for a second crop is not likely to depend solely on final levels of disease in the first crop. The consequences of these two forms of inoculum production for the spread and biological control of disease in consecutive crops are described schematically in Fig. 2. Disease is initiated by primary infection from particulate inoculum (Fig. 2ai) and spreads by secondary, plant-to-plant infection to create disease patches (Fig. 2aii). During the intercrop period diseased plants are converted to inoculum and, at the same time, the pathogen continues to spread on decaying roots by saprotrophic growth (Fig. 2aiii). This generates large amounts of inoculum (Fig. 2aiv) and high levels of primary infection in the next crop (Fig. 2av). In this study, biocontrol reduced primary infection and disease of the first crop (Fig. 2bii) and the production of inoculum from infected plants, but did not control colonization of roots (Fig. 2biii). This probably reflects the relatively low competitive saprotrophic ability (sensuGarrett, 1970) of Trichoderma vs Rhizoctonia in this system. From colonized roots, large quantities of inoculum developed (Fig. 2iv) to cause high levels of primary infection in a second crop when disease was absent in the first crop (Fig. 2v). It is striking that the inferred amplification of inoculum in the absence of damping-off or other symptoms of disease can result in severe and unexpected outbreaks of disease in the second crop.

image

Figure 2. A schematic interpretation of the spread of disease and inoculum over consecutive seasons for (a) unprotected and (b) protected radish crops. (i) Inoculum (open circles) initiates disease by primary infection of plants (dots) in the first crop. (ii) Disease in the first crop spreads by secondary infection to create disease patches (dark shading). (iii) After the crop is harvested the pathogen continues to spread by saprotrophic growth. (iv) The combination of disease dynamics and saprotrophic growth determine the density and distribution of inoculum for (v) disease dynamics in a second crop. Biological control may reduce or, as shown here, completely control disease in a first crop, but saprotrophic amplification of inoculum can still compensate for parasitic growth leading to severe and unexpected outbreaks of disease in a second crop.

Download figure to PowerPoint

The analysis of epidemics using mechanistic models offers a powerful but as yet poorly exploited tool with which to study biological control. Moreover, it offers the opportunity to develop a coherent strategy for the long-term management of disease over seasons (Gubbins & Gilligan, 1997) whereby the consequences of control in one season change the dynamics in the following season, necessitating a change in control strategy. It is likely that different forms of control, whether biological, chemical or physical, will affect different components of the parasitic (primary infection and secondary infection) and saprotrophic amplification of inoculum. The modelling and analysis introduced here provide a theoretical and experimental framework to investigate these systems.

While we have proposed a mechanism for the bulking-up of inoculum and disease of R. solani, and for the failure of biological control for crops grown in quick succession, long-term decay in infectivity of inoculum (disease suppression) attributed to microbial competition (Schneider et al., 2001) or the development of native antagonists (Whipps, 1997) is also commonly observed under a regime of continuous cropping (Henis et al., 1978; MacNish, 1988; Weller et al., 2002). The sand system used here was experimentally appealing because of its biological simplicity and homogeneity. A natural progression is to examine the influence of environmental factors such as soil type, soil temperature and soil moisture on the parameters for epidemic development of both average levels of disease and the variability between replicate epidemics. The consequences of such an increase in biological complexity and heterogeneity for the performance of biocontrol in soil await further analysis. The current work shows that epidemiological analysis of disease data can be used to develop a coherent understanding of inoculum dynamics and that, in the absence of this, the performance of biocontrol from one crop to the next will remain unpredictable.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This work was funded by the award of a Research Grant from the Biotechnology and Biological Sciences Research Council which we gratefully acknowledge.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Bailey DJ, Gilligan CA. 1997. Biological control of pathozone behaviour and disease dynamics of Rhizoctonia solani by Trichoderma viride. New Phytologist 136: 359367.
  • Bailey DJ, Gilligan CA. 1999. Dynamics of primary and secondary infection in take-all epidemics. Phytopathology 89: 8490.
  • Entwisle AR. 1990. Root diseases. In: RabinowichHD, BrewsterJL, eds. Onions and allied crops. Boca Raton, FL, USA: CRC Press, 105154.
  • Garrett SD. 1970. Pathogenic root infecting fungi. Cambridge, UK: Cambridge University Press.
  • Gibson GJ, Gilligan CA, Kleczkowski A. 1999. Predicting variability in biological control of a plant–pathogen system using stochastic models. Proceedings of the Royal Society Series B 266: 111.
  • Gilligan CA, Bailey DJ. 1997. Components of pathozone behaviour. New Phytologist 136: 343358.
  • Gilligan CA, Simons SA, Hide GA. 1996. Inoculum density and spatial pattern of Rhizoctonia solani in field plots of Solanum tuberosum: effects of cropping frequency. Plant Pathology 45: 232244.
  • Gubbins S, Gilligan CA. 1997. Persistence of host–parasite interactions in a disturbed environment. Journal of Theoretical Biology 188: 241258.
  • Henis Y, Ghaffar A, Baker R. 1978. Integrated control of Rhizoctonia solani damping-off of radish: effect of successive plantings, PCNB, and Trichoderma harzianum on pathogen and disease. Phytopathology 68: 900907.
  • Hide GA, Read J. 1991. Effects of rotation length, fungicide treatment of tubers and nematocide on diseases and the quality of potato tubers. Annals of Applied Biology 119: 7787.
  • Kleczkowski A, Bailey DJ, Gilligan CA. 1996. Dynamically generated variability in a plant pathogen system with biological control. Proceeding of the Royal Society London Series B 263: 777783.
  • Kleczkowski A, Gilligan CA, Bailey DJ. 1997. Scaling and spatial dynamics in plant pathogen systems: from individuals to populations. Proceeding of the Royal Society London Series B 264: 979984.
  • MacNish GC. 1988. Changes in take-all (Gaeumannomyces graminis var. tritici), rhizoctonia root rot (Rhizoctonia solani) and soil pH in continuous wheat with annual applications of nitrogenous fertilizer in Western Australia. Australian Journal of Experimental Agriculture 28: 33341.
  • MacNish GC. 1996. Patch dynamics and bare patch. In: SnehB, Jabaji-HareS, NeateS, DijstG, eds. Rhizoctonia species: taxonomy, molecular biology, ecology, pathology and disease control. Dordrecht, the Netherlands: Kluwer Academic, 215226.
  • Otten W, Filipe JAN, Bailey DJ, Gilligan CA. 2003. Quantification and analysis of transmission rates for soilborne epidemics. Ecology (in press.)
  • Papavizas GC. 1970. Colonisation and growth of Rhizoctonia solani in soil. In: ParmeterJR, ed. Rhizoctonia solani, Biology and Ecology. Los Angeles, CA, USA: University of California Press, 108122.
  • Rekah Y, Shtienberg D, Katan J. 2001. Population dynamics of Fusarium oxysporum f. sp radicis-lycopersici in relation to the onset of fusarium crown and root rot of tomato. European Journal of Plant Pathology 107: 367375.
  • Schneider JHM, Kocks CG, Schilder MT. 2001. Possible mechanisms influencing the dynamics of rhizoctonia disease of tulips. European Journal of Plant Pathology 107: 723738.
  • Schoeny A, Lucas P. 1999. Modelling of take-all epidemics to evaluate the efficacy of a new seed treatment on wheat. Phytopathology 89: 954961.
  • Weller DM. 1988. Biological control of soil-borne pathogens in the rhizosphere with bacteria. Annual Review of Phytopathology 26: 379407.
  • Weller DM, Raaijmakers JM, Gardener BBM, Thomashow LS. 2002. Microbial populations responsible for specific soil suppressiveness to plant pathogens. Annual Review of Phytopathology 40: 309348.
  • Whipps JM. 1997. Developments in the biological control of soil-borne plant pathogens. Advances in Botanical Research 26: 1134.
  • White KAJ, Gilligan CA. 1998. Spatial heterogeneity in three-species, plant–parasite–hyperparasite, systems. Philosophical Transactions of the Royal Society of London, Series B – Biological Sciences 353: 543557.
  • Williams D. 2001. Weighing the odds. Cambridge, UK: Cambridge University Press, 236239.