Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Pesticide Resistance: Strategies and Tactics for Management. 1986. National Academy Press, Washington, D.C. Response of Plant Pathogens to Fungicides M. S. WOLFE and I. A. BARRETT Genetic variationforfungicide resistance must occur if a pathogen is to respond to fungicide use. The rate of pathogen response depends on a complex interaction between the exposure of the pathogen to the fungicide, the biology of the pathogen, and the environment. An example of this interaction is the response of the barley mildew pathogen Erysiphe graminis f. sp. hordei to the widespread use of triazole fungicides in the United Kingdom, which also illustrates the interaction of fungicide resistance and host pathogenicity. The current strategies of fungicide use tend to exacerbate the problem of restraining pathogen response. Other strategies, based on different forms of diversification, may be helpful in practice, at least under western European conditions. Experiments were con- ducted with fungicide treatments of the seed of single components of mixtures of host varieties having different resistance genes. On the farm this system can give good disease control and predictably high yields at low cost. Durability is not predictable, except that it is likely to be better than with current strategies, with the additional benefit of restricting the response of the pathogen to resistant hosts. INTRODUCTION This paper is an amalgam of first principles and practical experience gleaned largely from research on the control of Erysiphe graminis f. sp. hordei on barley. The use of fungicides changes the environment of the pathogen, and to understand its response requires a knowledge of how such changes affect selective differences between different genotypes in the population. Only 245
246 POPULATION BIOLOGY OF PESTICIDE RESISTANCE then can a way that is acceptable biologically and for practical crop production be developed to modify the response. FUNGICIDE USE The Attraction of Fungicides Why are fungicides used? Broadly, there are three reasons. The first is to control disease during crop development. Among field crops the view is encouraged that a particular species or variety is susceptible and thus losing yield to a disease, that the plant breeders have failed to deal with the problem, and that fungicides will provide the answer. The perception of susceptibility in commercial production, however, is based on an assessment relative to complete absence of disease. Truly susceptible host lines are eliminated during the breeding process and are rarely seen in agriculture; those that are deemed susceptible but remain in cultivation often have yields of only 20 percent (or less) below their potential maximum. Fungicides are used ex- tensively to remove this limitation so as to achieve the "ideal" of a disease- free crop. Initially at least, fungicides remove these restraints consistently and reliably because the recommended dose rates are determined from field trials with adequate pathogen inoculum applied to the currently most susceptible com- mercial varieties. For the farmer the fungicide controls the disease perfectly because his varieties, on average, will be less susceptible than those used in manufacturers' trials, and his farm conditions will tend to be less favorable for disease development. For these same reasons many fungicide applications expose the pathogen to a fungicide for no economic return, but the psychological impact of the clean crop more than offsets this hidden factor. A similar psychological problem arises from using fungicides to eliminate blemishes completely from produce for direct consumption. Perfect produce has become the norm for the marketplace even though it may not be essential, productivity is not improved, and exposure of pathogens to fungicides is maximized. The demands for clean crops and perfect produce mean that fungicides are used increasingly as prophylactic treatments known to cereal farmers in eastern England as the sleep-easy factor despite the consequences. The second reason for the use of fungicides is to improve the storage of produce. Perfect control of storage diseases increases the size and duration of the market available for the product. Thus, the marketplace again encourages widespread use of fungicides, particularly since plant breeders do little or nothing directly to breed for resistance to storage diseases. Third, with fungicides growers can increase production of a particular crop
RESPONSE OF PLANT PATHOGENS TO FUNGICIDES 247 and reduce their dependence on conventional controls of crop rotation and sanitation. Moving away from the costs and constraints of conventional controls is double-edged: the fungicide usage per unit area is increased, as is the total area of the crop and the size of the potential medium for the target pathogen. The increased potential for the crop provided by the fungicide is often so dramatic initially that some manufacturers suggest that breeders need no longer breed for host resistance. Any decrease in attention to inherent host resistance, however, is almost certain to exacerbate and accelerate se- lection of fungicide resistance, simply because pathogen survival is made easier. Fungicide Application and Type The area treated with a fungicide contains the effective treated area, defined as the proportion of the crop at any one time in which the fungicide level is higher than the threshold of control of the common fungicide-sensitive gen- otypes of the pathogen. For example, if equal amounts of two different fungicides are applied to a crop but one is more systemic and persistent than the other, the effective treated area of the first will be greater. Disease control will be greater, but so will the advantage accruing to resistant genotypes of the pathogen. Fungicides may be formulated for use as seed treatments, or as foliar sprays, or both. Seed treatments are potentially more effective because they may control the pathogen when the population is at its smallest and thus delay epidemic development, particularly if the compound is systemic and persistent. The corollary is that the pathogen population has a longer exposure to the treatment. If a fungicide is formulated both as a seed treatment and as a foliar spray and the compound is used widely and sequentially in the two forms, the effective treated area and the advantage to resistant genotypes are greatly increased. Broad-spectrum fungicides, as opposed to selective fungicides, may com- pound the problem if they remove competitors or hyperparasites that would assist the activity of a selective fungicide. Thus, the greatest potential for fungicide resistance comes from the large-scale prophylactic use of a broad- spectrum, systemic, and persistent material applied to the seed and then to the foliage. The fungicide initially controls the disease dramatically, and it is easily sold to farmers who are mostly risk-averse. The alternative of a nonpersistent, selective foliar spray, applied only when the disease level passes a defined threshold, is risky and demands accurate monitoring, fore- casting, and assessment of yield loss, but it reduces the time over which the pathogen is exposed to the fungicide and thus reduces the probability of resistance evolving.
248 POPULATION BIOLOGY OF PESTICIDE RESISTANCE PATHOGEN RESPONSE A Priori Considerations Any response to fungicide use depends, first, on whether genetic mech- anisms exist to reduce or eliminate the effects of the fungicide. The mech- anisms may occur at low frequencies before the fungicide is introduced, they may occur as mutations, or both. The rate at which the pathogen responds then depends on the interaction between the mechanisms available and their genetic control, the use of the fungicide, the biology of the organism, and the environment. One major factor is whether the organism is diploid or haploid in the asexual stage. If haploid then any mutation to fungicide resistance is im- mediately expressed, and the frequency of the mutant will be influenced by its effect on fitness. With a diploid organism the situation is more complex; there may be a cryptically high frequency of resistance, depending on the fitness of the heterozygotes and resistant homozygotes relative to the wild type, in the presence and absence of the fungicide (Barrett, in press). The rate of response of a pathogen also depends on its breeding system, principally on whether there is an obligate sexual or parasexual sequence in the life cycle. An effective sexual stage allows for more rapid formation of novel combinations of appropriate characters through recombination, which may increase the fitness of the resistant pathogen genotypes. With no sexual stage, linkage disequilibrium between resistance and other characters is likely to persist, which may limit or delay adaptation of the pathogen to the treated host population. The spread of fungicide resistance depends on the distribution of propa- gules: populations of foliar pathogens with airborne spores will respond more rapidly than soil-borne pathogens. Finally, the ability of a pathogen to respond to fungicidal control depends on its ability to cope with other environmental stresses. An organism at the limits of its ability to survive in a particular environment will be less able to respond to an extra stress. For example, the greater the level of disease resistance and diversity in the host crop the less likely it will be for a pathogen to develop and spread resistance to a fungicide. Dynamics Wolfe (1982) summarized the interaction of selection for resistance and for other characters. Whether fungicide resistance increases in a population is determined by the size of the effective treated and untreated areas and the fitness of the forms of the pathogen with different sensitivities to the fungicide on each of these areas. There will tend to be large differences in fitness on the treated crop and smaller differences on the untreated. If the differences
RESPONSE OF PLANT PATHOGENS TO FUNGICIDES 249 on the untreated crop area are small, then a small area of treated crop may allow resistant forms of the pathogen to predominate in the population as a whole. If the fitness differences on the untreated crop are large, then fun- gicide-resistant forms of the pathogen may not become apparent until there is a large treated area. The overall fitness of sensitive and resistant forms of the pathogen, therefore, depend on the area of fungicide treatment. Growth rate differences between isolates measured in the laboratory may have little relevance to the fate of those isolates in the field. Monitoring the range of forms of a pathogen with reduced sensitivity to a fungicide is difficult. The phenotypes isolated first may not be the ones that eventually become common, because recombination and selection may change the expression of resistance during its spread. Indeed, if selection is maintained it is never possible to predict when the response will cease. In the example of barley mildew adapting to the use of ethirimol, Brent et al. (1982) noted a shift to an apparent equilibrium between sensitivity and re- sistance in the pathogen population. In this case, however, selection for resistance declined when ethirimol was replaced by other fungicides and more resistant varieties: the apparent equilibrium may have been a temporary peak associated with maximum use of the fungicide. AN EXAMPLE The worst case in terms of selection for resistance is where a systemic, persistent, and broad-spectrum fungicide is applied sequentially on the major part of the crop area to control a well-adapted foliar pathogen that is efficiently dispersed by airborne spores and has an effective sexual stage. Among field crops this combination of characters is exemplified by the use of triazole fungicides to control barley mildew in western Europe. Shortly after introduction of these fungicides into commercial use in the United Kingdom, the first isolates with some resistance were identified in small populations surviving on treated crops (Fletcher and Wolfe, 1981~. From 1981 the air spore was monitored continuously (Wolfe et al., 1984a) by means of a simple spore trap mounted on a car roof (Wolfe et al., 1981; Limpert and Schwarzbach, 19811. The numbers of colonies that incubated on seedlings with different doses of the fungicide increased annually relative to the numbers on untreated seedlings. The early surveys could not always detect isolates with fungicide resistance in the small populations on treated crops; by 1984, however, such isolates were detected easily on untreated crops. The increase in frequency of the less-sensitive phenotypes showed two interesting characteristics. The first was that the rate of increase varied during the year. This variation was repeated between years, which suggested that during the spring, following seed treatment and early foliar sprays, there
250 POPULATION BIOLOGY OF PESTICIDE RESISTANCE TABLE 1 Mean Pathogenicity (Pathog.) on Six Differential Barley Hosts of Powdery Mildew Isolates with Different Levels of Sensitivity to Tr~adimenol Obtained from Untreated and Treated Seedlings in a Car Spore Trap in East Anglia, 1981-1983 Seedling 1981 1982 1983 Source EDso Pathog. ED50 Pathog. ED50 Pathog. Untreated 0.028 32 0.060 40 0.080 35 0.025a 0.045 27 0.080 35 0.093 35 0.125a 0.085 7 0.093 25 0.108 35 aGrown from seed treated at 0.025 or 0.125 g a.i./kg. SOURCE: Wolfe (in press [a]). was rapid selection toward resistance. During the summer the response slack- ened or reversed, presumably following dissipation of the fungicide. At the beginning of autumn, however, frequency sharply increased, probably due partly to release of ascospores from cleistothecia formed at the time of relatively high frequencies of resistance at the end of spring and partly to the influence of emerging crops of treated winter barley. During autumn and winter the frequency of resistant forms again declined. In pathogen populations on individual field crops of treated winter barley, the frequency of the most resistant forms was high on seedlings in the autumn because of the selection imposed by the high concentration of fungicide in the seedling leaf tissue (Wolfe et al., 1984a). As the plants grew and the concentration decreased, the frequency of these forms decreased and forms with intermediate resistance became predominant. On the untreated crops sensitive forms were initially predominant, but, again, forms with interme- diate resistance eventually became more common, presumably due to spores migrating from other crops, most of which would have been treated at some stage. The second major feature of interest was the relationship between resistance and pathogenicity. During the early stages of the overall increase in resis- tance, the more resistant forms of the pathogen were less pathogenic on the range of host varieties in common use at the time (Table 11. In subsequent seasons, however, pathogenicity of the sensitive fraction remained constant, but the resistant fraction gradually increased to the same level. The increase in pathogenicity in the resistant part of the population occurred earlier for some characters than for others. For example, resistance increased first in Scotland and northern England in populations having a high frequency of pathogenicity for varieties with the Mla6 resistance gene. This created linkage disequilibrium, and isolates having these characters rapidly became common throughout the United Kingdom. The potential value of Mla6 was thus diminished in areas where it was not in current use. Simultaneous with
RESPONSE OF PLANT PATHOGENS TO FUNGICIDES 251 these changes the resistant variety Triumph became extensively cultivated and increasingly susceptible. Triazole fungicides thus became widely used on Triumph; isolates resistant to triazoles are now commonly pathogenic on Mla6 or Triumph or both. As fungicide resistance in the pathogen population increases, there may be loss of disease control and a reduction in the yield advantage expected from treatment. Initially such effects have a patchy distribution. Not all resistant isolates will be associated with poor fungicide performance and, conversely, not all poor fungicide performance will result from the oc- currence of fungicide-resistant isolates. Inevitably, during the first seasons of using a new fungicide, there will be some instances of poor control due to incorrect application and other environmental problems. This small proportion will fluctuate from season to season; a real deterioration in fungicide performance will be signalled by a continuing increase in in- stances of poor control. For example, with triazoles and the control of barley mildew, following the increase in frequency of resistant forms in eastern England, performance of triazoles both in disease control and in yield benefit rapidly declined (Table 21. The effect was most marked in varieties with the Mlal2 resistance gene; the yield increase following treatment declined from 25 percent in 1982 (P < 0.001) to 3 percent in 1984 (not significant), during which time ethirimol- a different seed treatment that was less widely used gave a consistent yield advantage of around 10 percent (P < 0.051. A similar yield advantage during this period was obtained with ethirimol applied to Carnival (Mla6), but there was no advantage with triazole treatment, probably because of the higher frequency of resistant isolates carrying pathogenicity for Mla6 compared ~ ~ 1 _ ~ 1 ~ with those pathogenic against Blab;. A more complex ~n~erac~on warn Anise fungicides was obtained with Triumph and Tasman because of the declining resistance of the varieties during this same period. Nevertheless, the perfor- mance of the triazoles declined relative to that of ethirimol. CONTROLLING THE PATHOGEN RESPONSE Reducing exposure of the pathogen to the fungicide is the most obvious way to deter resistance, and this can be helped by making disease forecasting more precise and educating growers to the problems. Commercial pressures against such actions, however, may be strong. Reducing the fungicide dose may or may not delay resistance development. If the dose is reduced to a level at which some sensitive genotypes survive, there may be some delay; however, the pathogen may cause unacceptable yield loss. On the other hand any delay caused by an increased dose is likely to be followed by emergence of highly resistant strains of the pathogen. Other changes in the formulation of the compound or inefficiency of application may also alter the fitness
252 POPULATION BIOLOGY OF PESTICIDE RESISTANCE TABLE 2 Yield (t/ha) of Spring Barley Varieties with Different Mildew Resistance Genes, Untreated or Treated with Ethir~mol or Tr~adimenol, 1982-1984 Variety Year Untreated Ethir~mol-trt. Tr~adimenol-trt. Ml al2 Egmont 1982 5.01 5.49 6.25 ref. 100 110 125 Patty 1983 3.51 3.90 4.12 ref. 100 111 114 Patty 1984 6.90 7.46 7.13 ref. 100 108 103 Ml ad Carnival 1982 5.38 5.87 5.64 ref. 100 109 105 Carnival 1983 3.83 4.11 3.84 ref. 100 107 100 Carnival 1984 6.60 7.07 6.53 ref. 100 107 99 Ml flay Triumph 1982 5.40 5.81 ref. 100 108 Tasman 1983 3.57 3.85 3.70 ref. 100 108 104 Tasman 1984 5.66 6.43 6.05 ref. 100 114 107 NOTE: Standard error for 1982, + 0.11; 1983, + 0.23; 1984, + 0.14. SOURCE: Wolfe (in press [a]). differences between sensitive and resistant genotypes and make prediction difficult. Reducing the use of a particular compound may need to be accompanied by other means of limiting pathogen increase, such as diversifying between fungicides with different modes of action known or thought to be matched by different pathogen mechanisms. For commercial and technical reasons, there are considerable constraints to the kinds of action that can be recom- mended. The current system is the use of mixtures, usually a tank mix of a systemic and a nonsystemic compound. The data to support this approach are inconclusive. Adding a nonsystemic material may only temporarily reduce the absolute population size of the pathogen, while the systemic material will be more persistent so that after the initial combined action of the fun- gicides, the pathogen population will be exposed uniformly to the systemic compound on all plants and thus selected for resistance. A more effective system, analogous to the use of variety mixtures (Wolfe, 1981), may be to ensure that the compounds eliciting different responses are applied to adjacent plants. The pathogen must then either adapt to a single
RESPONSE OF PLANT PATHOGENS TO FUNGICIDES 253 plant or become versatile between plants. Compared with a uniformly treated stand, there is a greater space between plants receiving the same treatment, so that increase of the population resistant to that treatment is delayed. Further, any genotypes with combined resistance to all of the fungicides used are likely to be less fit on any one plant than the genotype specifically adapted to the treatment on that plant. Currently, this approach can be contemplated only for fungicides applied to seed. Even here treatment on one seed may spread to other seeds treated differently, and different treatments may vary in their effects on the flow rate of seed either in a mixing process or in a seed drill. Recent developments in film coating of seeds may eliminate such problems. Fungicides can be applied to seed in a carrier material, improving the precision of individual seed treatment. The material is fixed firmly to the seed, and the flow char- acteristics of the seed are similar to those of seed treated with other fungicides (M. D. Tebbit, Nickersons Seed Specialists Ltd., personal communication, 19841. Seeds can also be simultaneously color coded so that intimacy of mixing can be confirmed. Future developments in application technology may allow a similar ap- proach with foliar sprays. For example, ultra-low-volume equipment such as the electrostatic sprayer raises the possibility of using a square matrix of containers holding different fungicides, mounted on a frame with a system of rapid on-off switching so that a fine mosaic of different materials can be applied. INTEGRATED DISEASE CONTROL Unfortunately, much of the discussion on controlling pathogen response to fungicides makes no reference to the host crop. In the simplest case, with partially resistant host varieties, the number of treatments and the dose can be reduced, thereby reducing selection on the pathogen for resistance to the fungicide and indeed for pathogenicity to the host (Wolfe, 19811. Sometimes it is more effective to use intimate mixtures of host varieties with different resistance genes (Jensen, 1952; Wolfe and Barrett, 1980; Wolfe, 1985~. Particularly if diversity between mixtures is maintained in space and in time, disease control is more consistent and durable than if the components are used in monoculture. By changing the composition of mixtures as new varieties become available, both the yield potential and the diversity are maximized, which suits both the farmer and the plant pathologist. From 1980 through 1984 four barley varieties with different resistance genes and the four mixtures of three varieties that can be made from them were grown in field trials at the Plant Breeding Institute, Cambridge, England (Wolfe et al., 1984b; Wolfe et al., 19851. Over the trial series the mixtures outyielded the pure stands by 7 percent (P < 0.0011. The best strategy found
254 POPULATION BIOLOGY OF PESTICIDE RESISTANCE TABLE 3 Average Yields (t/ha) and Infection (total percent leaf cover) for 1983 and 1984 of the Three Spring Barley Varieties Carnival, Patty, and Tasman, Grown as Pure Stands or Mixtures, Untreated or Treated with a Triazole or Ethirimol at the Normal Field Rate Yield (t/ha) Infection (total % leaf cover) . Pure Rel. Mixed Rel. Pure Rel. Mixed Rel. Untreated 5.192 100 5.61 ~108 25.7 100 19.3 75 Tr~azole 1/3 S.253a 101 5.622 108 22.62 88 15.2 59 N 5.372 103 5.44 ~105 16.4 64 13.6 53 Ethir~mol 1/3 5.3S3a 103 s.682 109 20.4a 79 10.2 40 N 5.672 109 5.65 ~109 9.7 38 6.3 25 NOTE: The 1/3 treatment of the mixtures is the mean of three mixtures in each, of which only one component is treated with tr~azole or ethir~mol. The 1/3 treatments of pure stands are calculated values obtained from the sum of the three pure varieties treated, plus twice their sum untreated, divided by nine. aCalculated values. ~SE= +0.16. 2SE = + 0.09. 3SE = + 0.07. SOURCE: Wolfe (in press [b]). for the farmer, given the choice of only those four varieties each year, would have been to grow any one or more of the mixtures. Based on this research variety mixtures are now grown commercially in the United Kingdom and Denmark, with generally favorable reports from the farmers involved. A much larger scale of development is being undertaken in the German Dem- ocratic Republic, particularly because of the high cost of fungicides in eastern Europe. Despite the obvious advantages of the variety mixtures, disease control is sometimes considered to be inadequate, and some mixtures are treated with fungicides even though the benefit may be uneconomic. For this reason and to provide long-term protection for the varieties and the fungicides, exper- iments have been conducted with fungicide-integrated mixtures (Wolfe, 1981; Wolfe and Riggs, 19831. The seed of one component of a three-variety mixture is treated with a fungicide and then mixed with the two untreated components. Data for two field experiments in 1983 and 1984 are summarized in Table 3. In these experiments Carnival (Mla6 resistance), Patty (Mlal2), and Tasman or Triumph (both Mla7 plus MlAb) were grown alone, un- treated, or treated either with ethirimol or a triazole fungicide. They were . also grown as a mixture and in plots where only one component was treated. All plots were surrounded by guards to reduce interplot interference.
RESPONSE OF PC4NT PATHOGENS TO FUNGICIDES 255 Although it reduced infection, treating pure stands with triazole did not increase yields significantly, probably because fungicide resistance increased during the period. The effect of ethirimol treatment on yield, however, was highly significant (P < 0.001) and was associated with greater disease con- trol. Mixing varieties without a fungicide treatment increased yield signifi- cantly (P < 0.05) and reduced infection, although fungicide treatments of the mixture had no significant effect. An interesting but not significant result was that the highest absolute yields were obtained with the mixtures in which single components had been treated. For both fungicides the yields of these 1/3 treatments were significantly higher (P < 0.01) than the equivalent calculated treatment of pure stands. Moreover, there was considerably less infection on these mixtures than on untreated mixtures; they were only slightly more infected than the mixtures that received the conventional fungicide treatment. Comparing the 1/3 treatments of the mixture with the conventional treatment of the pure stands, the mixture yields were higher, significantly so for the triazole treatments, and infection levels were the same. Thus, for the farmer, using the 1/3 treatment of a variety mixture would produce a yield as high and a crop as clean as from conventionally treated pure stands, but at a lower cost. Epidemiologically the fungicide seed treat- ment protects the crop at the beginning of the epidemic, when variety mixing is least effective. Later in the growth cycle the crop is protected more by the varietal heterogeneity, after the fungicide concentration has declined below the threshold for disease control. Biologically the pathogen is less able to overcome each variety and fungicide component, and less fungicide is delivered into the environment. We may also expect to maintain higher yields with the partly treated mixtures than with the conventionally treated pure varieties. CONCLUSION The response of a pathogen population to fungicide use depends on genetic variation for resistance being present in the population. When such variation is present and can be demonstrated, the rate and form of the response will depend on a complex interaction of the genetic and breeding system and general biology of the target organism, the range of host varieties in use, cultivation practices, and the physical environment. The example of powdery mildew of barley shows how responses can be manipulated using different forms of crop husbandry. The ability to modify the pathogen response requires at least an understanding of the genetics and population dynamics of the pathogen so that the consequences of changes in cultivation practices can be predicted. Without a reasonable understanding of the population biology of the pathogen and of the consequences of crop husbandry methods, it is not
256 POPULATION BIOLOGY OF PESTICIDE RESISTANCE possible either to understand the responses or to suggest changes in agr~- cultural practices that might modify the response. The only certain conclusion is that if variation for resistance exists, and the fungicide is used extensively and homogeneously, then its effectiveness will soon decline. Unfortunately, the pathogen may ultimately find a way around any strategy designed to control it. ACKNOWLEDGMENT We wish to acknowledge financial help from ICI Plant Protection Ltd. for part of the experimental work. REFERENCES Barrett, J. A. In press. In Populations of Plant Pathogens: Their Dynamics and Genetics, M. S. Wolfe and C. E. Caten, eds. Oxford: Blackwell. Brent, K. J. 1982. Case study 4: Powdery mildews of barley and cucumber. Pp. 219-230 in Fungicide Resistance in Crop Protection, J. Dekker and S. G. Georgopoulos, eds. Wageningen, Netherlands: Centre for Agricultural Publishing and Documentation. Fletcher, J. T., and M. S. Wolfe. 1981. Insensitivity of Erysiphe graminis f. sp. hordei to triadimefon, triadimenol and other fungicides. Pp. 633-640 in Proc. Br. Crop Prot. Conf. Fungic. Insectic. Vol. 2. Lavenham, Suffolk: Lavenham. Jensen, N. E. 1952. Intra-varietal diversification in oat breeding. Agron. J. 44:30-34. Limpert, E., and E. Schwarzbach. 1981. Virulence analysis of powdery mildew of barley in different European regions in 1979 and 1980. Pp. 458-465 in Proc. 4th Int. Barley Genet. Symp. Edinburgh: Edinburgh Univ. Press. Wolfe, M. S. 1981. Integrated use of fungicides and host resistance for stable disease control. Philos. Trans. R. Soc. London, Ser. B 295:175-184. Wolfe, M. S. 1982. Dynamics of the pathogen population in relation to fungicide resistance. Pp. 139-148 in Fungicide Resistance in Crop Protection, J. Dekker and S. G. Georgopoulos, eds. Wageningen, Netherlands: Centre for Agricultural Publishing and Documentation. Wolfe, M. S. 1985. Current status and prospects of multiline cultivars and variety mixtures for disease resistance. Annul Rev. Phytopathol. 23:251-253. Wolfe, M. S. In press [a]. Dynamics of the response of barley mildew to the use of sterol synthesis inhibitors. EPPO Bull., Vol. 15. Wolfe, M. S. In press [b]. Integration of host resistance and fungicide use. EPPO Bull., Vol. 15. Wolfe, M. S., and J. A. Barrett. 1980. Can we lead the pathogen astray? Plant Dis. 64:148-155. Wolfe, M. S., and T. J. Riggs. 1983. Fungicide integrated into host mixtures for disease control. P. 834 in Proc. 10th Int. Congr. Plant Prot. Brighton, 1983. Vol. 2. Wolfe, M. S., P. N. Minchin, and S. E. Slater. 1981. Powdery mildew of barley. Annul Rep. Plant Breed. Inst. 1980:88-92. Wolfe, M. S., P. N. Minchin, and S. E. Slater. 1984a. Dynamics of triazole sensitivity in barley mildew, nationally and locally. Pp. 465-470 in Proc. 1984 Br. Crop Prot. Conf., Pests and Dis. Washington, D.C. College Park, Md.: Entomological Society of America. Wolfe, M. S., P. N. Minchin, and S. E. Slater. 1984b. Annul Rep. Plant Breed. Inst. 1983:87- 91. Wolfe, M. S., P. N. Minchin, and S. E. Slater. 1985. Powdery mildew of barley. Annul Rep. Plant Breed. Inst. 1984:91-95.