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. Detection and Monitonng of Resistant Fonns: An Overview K. J. BRENT Detection and monitoring are major components of pesticide re- sistance management, for several reasons. The different steps that should be taken in any detection and monitoring program, as well as examples of successful programs, are described. It is important to monitor for sensitivity and to establish a resistance management strategy early in the life of a new product. The need to distinguish clearly between detecting less-sensitive forms and concluding that practical resistance problems have arisen is also stressed. The most effective programs can be developed and carried out only with the collaboration of private and public organizations. INTRODUCTION What precisely is meant by "the detection and monitoring of resistance"? This basic question must be considered at the outset of any discussion on this topic, because much vagueness and misunderstanding exist about the terms involved and their meanings. "Detection" indicates simply the obtaining of initial evidence for the presence of resistant forms in one or more field populations of the target organism. Consideration of the degree of resistance, the proportion of resis- tant variants in a population, or the effect on practical field performance of the pesticide is not involved. "Monitoring" needs more consideration. To many people it denotes a routine, continuous, and random "watch dog" program, analogous to the official monitoring for levels of pesticide residues in foodstuffs. Such year- in, year-out surveillance aims to detect and then follow the spread of any 298
DETECTION AND MONITORING OF RESISTANT FORMS 299 markedly abnormal forms should they arise, or with sufficiently sensitive and quantitative methods, to reveal any gradual erosion of response, as has occurred with certain plant pathogens. Campaigns of this kind can be pro- tracted and unrewarding, although sometimes they may be justified for certain very important pesticide uses when the risk of resistance is already known to be considerable. More specific, shorter term investigations are also (less aptly) referred to as monitoring. These are done either to gain initial or "baseline" sensitivity data before the widespread commercial use of a new pesticide or, more commonly, to examine individual cases of suspected resistance indicated by obvious loss of field efficacy of the product. Thus, monitoring can be used to indicate either continuous surveillance or ad hoc testing programs; this double use is acceptable, providing the meaning of the term is made clear in any particular context. "Resistance" and "resistant" have many different shades of meaning. For precision either a particular usage must be specified as the correct one or resistance must be defined clearly whenever it is used. The first of these options is unattractive, because new, narrow definitions of commonly used and fairly general terms are seldom adopted universally or even remembered, and they force us to define a whole range of other narrow terms. Hence, "resistance," "tolerance," "insensitivity," and "adaptation" should not, as some suggest, be given separate, precise meanings. The second option, however, is both feasible and sensible and should be encouraged. Resistance can be used in a general way and interchangeably with the other terms to mean any heritable decrease in sensitivity to a chemical within a pest pop- ulation. This can be slight, marked, or complete and may be homogeneous, patchy, or rare within a population. It can cause complete loss of action of an agrochemical or may have little practical significance. Thus, resistance and similar terms must, like monitoring, be defined carefully within each particular context. In reports on monitoring, the absolute use of resistance (as in "the pop- ulation was resistant") causes more problems of misinterpretation than rel- ative use ("population A was more resistant than B"), and a quantitative definition of how resistance was categorized and measured should always be given. A "resistance index" or "resistance factor" (the ratio of the doses, commonly EDso, required to act against resistant and sensitive forms, re- spectively) is often used, but the basis of its calculation needs careful con- sideration. The choice of sensitive reference strains (sometimes merely a single one is used) and any shift in their response with time can affect greatly the value of the index and inferences made, at least with regard to fungicide resistance. If a reference strain has been kept away from all chemical treat- ments for years in a laboratory culture, it may be abnormally sensitive. The Fungicide Resistance Action Committee (FRAC) has recommended that the term "laboratory resistance" should be used to indicate strains of
300 DETECTION, MONITORING, AND RISK ASSESSMENT fungi with significantly reduced sensitivity as demonstrated by laboratory studies, whereas "field resistance" should be used to indicate a causal re- lationship between the presence of pathogenic strains with reduced sensitivity and a significant loss in disease control. The intention is to avoid false alarms such as have occurred when certain authors, having found some specimens to be more resistant than others in laboratory cultures or field samples, implied without evidence that these variants were causing or were about to cause problems in practical pest control. The use of the above terms as suggested by FRAC, however, can also be misleading: resistant forms found in the field in low numbers or with a low degree of resistance or fitness are certainly field and not laboratory resistant, yet such forms may not be affecting practical control. Whatever terms are selected there is no substitute for defining clearly the implications and limits of their use in all publications. THE AIMS OF DETECTION AND MONITORING There are at least seven distinct motives for resistance, detection, and monitoring, and whichever of them predominates will affect the scope and design of the surveys that are done. The aims, which are discussed in turn below, are as follows: · Check for the presence and frequency of occurrence of the basic genetic potential for resistance (expressed resistance genes) in target organism pop- ulations. · Gain early warning that the frequency of resistance is rising and/or that practical resistance problems are starting to develop. · Determine the effectiveness of management strategies introduced to avoid or delay resistance problems. · Diagnose whether rumored or observed fluctuations or losses in the field efficacy of an agrochemical are associated with resistance rather than with other factors. · If resistance has been confirmed, determine subsequent changes in its incidence, distribution, and severity. · Give practical guidance on pesticide selection in local areas. ~ Gain scientific knowledge of the behavior of resistant forms in the field relation to genetic, epidemiological, and management factors. Potential for Resistance To obtain an initial indication of possible sources of future loss of effec- tiveness, we would need to be able to isolate and characterize rare mutants at, say, 1 in 10~° frequency. This is not feasible, however, without vast expense and effort. Resistant forms can be detected only after reaching much
DETECTION AND MONITORING OF RESISTANT FORMS 301 higher frequencies of 1 in 100 or perhaps 1 in 1,000 units (individual disease lesions, spores, pests, weeds), depending on the number of samples taken and the degree of statistical significance required. For example, if 1 in 100 units is resistant, 298 samples must be examined to achieve 95 percent probability of detection of 1 resistant unit; 2,994 samples must be checked if the frequency is 1 in 1,000. If a particular pesticide application normally allows 10 percent survivors (i.e., pest control is 90 percent effective), such detectable frequencies will occur only one or two applications prior to serious and obvious loss of practical control. With some pests and diseases this may be too late to allow any avoidance action to be introduced in the area con- cerned. The relatively late first indication of the occurrence of resistance forms, however, can still give a valuable alert for certain purposes or situations. For example, it can indicate to other regions or countries that the potential for resistance exists. Or there may be time to introduce or modify avoidance strategies in cases where the rate of reproduction of target organisms is low (one or two generations per year), where lack of fitness in resistant mutants leads to an interrupted or fluctuating buildup (as with resistance of Botrytis cinerea to dicarboximide fungicides), or where a range of variants with different degrees of resistance arise and resistance tends to build up in a stepwise manner (as in the resistance of powdery mildews to 2-amino- pyrimidine and triazole fungicides). In such situations loss of efficacy is still a gradual process, even after relatively high frequency levels are first de- tected. Shifts in Frequency or Severity of Resistance After initial detection systematic monitoring can reveal subsequent changes (if any) in the frequency and degree of resistance and in its geographic distribution. For this reason repeated surveys have been done by public- sector organizations such as the Food and Agriculture Organization of the United Nations (FAG), the World Health Organization (WHO), and national agricultural and health research authorities. Surveys are also increasingly done by agrochemical companies, sometimes in cooperation with Resistance Action Committees. Examples are considered in the later section on achieve- ments in resistance monitoring. Shifts in resistance can be very rapid. Sen- sitive populations have been known to be replaced completely by resistant ones over large areas within a year of first detection, particularly when the variants are highly resistant and retain normal or near normal fecundity and the ability to invade a host crop or animal. Shifts may be much more gradual, however, as mentioned above. It is essential to obtain information at each sampling site on the efficacy of field performance of the chemical following the latest and earlier applications, on the numbers and types of chemical
302 DETECTION, MONITORING, AND RISK ASSESSMENT treatments applied, and on management factors (e.g., cultivar grown, method of cultivation), in order to permit assessment of the practical impact of resistant forms at different stages of their buildup and to aid identification of factors that encourage or suppress resistance. Checking Resistance Management Strategies It is sometimes said that monitoring for resistance is a waste of time and money, because if positive results are obtained it is then too late to take effective action. This point of view may be valid under circumstances where the first variants detected are sufficiently resistant to cause loss of control and sufficiently fecund and competitive to accumulate rapidly and persist and where selection pressures are sufficiently heavy and widespread to induce large-scale shifts. Such has been the case with certain combinations of fun- gicides and plant pathogens, for example, the use of dimethirimol against cucumber powdery mildew (Sphaerotheca juliginea) in Holland (Brent, 1982) or of benomyl against sugar beet leaf spot (Cercospora beticola) in Greece (Georgopoulos, 1982b). Insecticide resistance commonly arises in this way (Keiding, this volume). There is now, however, an increasing and very welcome trend toward establishing, in the light of risk assessments, some kind of strategy of resistance management at the very outset of the commercial life of a new chemical. Monitoring then is done not to warn of the need to initiate action but with the much better aim of checking whether an established strategy is working adequately or needs to be modified or intensified. This type of approach is indicated in Table 1. investigation of Suspected Resistance Problems When observed losses of field efficacy are reported, they may be so dramatic that testing a few samples under controlled conditions against high doses of the chemical is sufficient to confirm resistance as the cause. The situation is sometimes less clear-cut: farmers may be using higher and higher rates of a chemical to achieve the same degree of control, or the period of persistence of protection may be gradually shortening. In such situations studies that are more extensive in area and time can reveal a great deal about the cause of these problems, and if there are correlations of reduced sensitivity of the target organism with loss of field performance, then the need for a change in the strategy of chemical use is indicated. Subsequent Changes in Resistance Later surveys, following a demonstration that resistant populations exist, can indicate whether shifts toward resistance are spreading or contracting in
DETECTION AND MONITORING OF RESISTANT FORMS TABLE 1 Phases of Monitoring and Resistance Management for a New Pesticide 303 Timing 1-2 years before start of sales Resistance Monitoring Activities - Establish sampling and testing methods Survey for initial sensitivity data (include treated trial plots) Other Management Activities During years of use As soon as signs of resistance are seen visually or through monitoring Monitor randomly in treated areas for resistance, only if justified by risk assessment or special importance Monitor to determine extent and practical . . ,~ ~ slgnlrlcance or resistance Subsequently Check rate of spread or decline of resistance Assess risk Decide strategy of use Work the decided use strategy Watch practical performance closely If resistance problem is confirmed, review strategies and modify Study cross-resistance, hltness of variants and other factors affecting impact of resistance Watch performance, review strategies SOURCE: Brent (unpublished). geographic distribution, whether they are increasing or decreasing in fre- quency or severity, or whether an equilibrium is reached. Attempts should be made to correlate any such changes in resistance with either initial or modified strategies of chemical use or crop management. Guidance in Pesticide Selection Immediate practical guidance to individual growers, based on resistance monitoring on the farm, may be feasible in some situations. The only example known to the author is in the control of Sigatoka disease of bananas (caused by Mycosphaerella spp.) in Central America, where the United Fruit Company and du Pont have recommended that growers use a simple agar-plate test every month and postpone the use of benomyl if they find that the proportion of resistant ascospores exceeds 5 percent (du Pont, 19821.
304 DETECTION, MONITORING, AND RISK ASSESSMENT Scientific Knowledge The use of monitoring to aid our understanding of the nature of the resis- tance phenomenon is important because of our present limited state of knowl- edge of the population dynamics of resistant forms in relation to biological, agronomic, and environmental factors. For example, are different races of target organisms or cultivars of host plants more prone to resistance problems than others? There is evidence of this in the resistance of barley powdery mildew to fungicides (Wolfe et al., 1984~. How far are theoretical models borne out in practice? Surprisingly few attempts have been made to validate the various proposed mathematical models of the progress of resistance in insects, plant pathogens, and weeds. How do factors such as dose applied, spray coverage, and timing affect the rate and severity of resistance devel- opment? The few studies that have been made for fungicides (Skylakakis, 1984; Hunter et al., 1984) have depended greatly on the development of precise and reproducible detection and monitoring procedures. TIMING AND PLANNING OF SURVEYS A new pesticide should work well initially on the target organisms against which it is recommended. If not, it would have failed in the large number of field trials that generally are done before marketing. Surveys should be started early, however, by testing field samples of each major target pest for degrees of sensitivity under controlled conditions before the chemical is used extensively (Table 11. Such testing provides valuable initial sensitivity (or baseline) data against which the results of any subsequent tests or surveys can be compared. These data could indicate the initial incidence of forms with resistance genes if their frequency and the number of samples tested were sufficiently high. Normally, however, testing will reveal the range of initial sensitivities of different populations of the pest; it also will provide an early opportunity to gain experience with and to check the precision of test methods that may be required at short notice if problems arise later. Some degree of variation in the results of initial sensitivity tests will occur, and it is necessary by replication or repetition of tests to separate experimental variation from real differences in response between populations. As part of the baseline exercise, it is very useful to check the sensitivity of surviving target populations shortly after successful use of the chemical in field trials: the less-sensitive elements of heterogeneous populations tend to predominate after treatment. Although these might persist and create problems later, often they lack fitness or are unstable and decline as the effects of the chemical wear off (Shepherd et al., 1975~. Once initial data are obtained a decision must be made as to whether further surveys are needed. Unless there is a special reason such as the
DETECTION AND MONITORING OF RESISTANT FORMS 305 critical importance of the particular target-chemical combination, an indi- cation of high risk from a risk-assessment exercise, considerable variation between samples in the initial survey, or evidence from other regions for resistance phenomena the effort and expense of further sampling will not be justified until signs of practical loss or erosion of efficacy are seen. A close watch should always be maintained, however, on the efficacy of treat- ment in practical use ("performance monitoring"), in comparison with initial field trial results and with the performance of other kinds of chemicals. If either an obvious major loss of effect or a gradual decline of performance are observed, all possible alternative causes of the difficulty (e.g., poor application, misidentification of target organism, increased pest or disease pressure) should be investigated, in addition to resistance. If possible, re- sistance sampling should be done at sites of poor and good control and at sites where the particular chemical has and has not been used. Positive correlations of degree of resistance with practical performance and with amount of use at the sampling sites must be sought. Sometimes highly resistant strains of fungi or insects have been detected readily at sites where the effectiveness of the product has been retained (Carter et al., 1982; Den- holm et al., 1984~. If tests indicate an appreciable shift in sensitivity from the baseline position, then further monitoring, preferably at the same sites, may well be justified to reveal whether resistance is spreading, worsening, declining, fluctuating, or showing little change and how far it is associated with losses of control. METHODS OF SAMPLING AND TESTING In an extensive survey many sites (e.g., farms, fields, or glasshouses) containing the target organism throughout a region or country are examined, and one or a few representative samples of the population are taken at each site. At the extreme, area populations of insects or spores can be trapped by using suction traps for aerial populations of insects or by mounting test plants on a car top and driving through a cropping area to sample the powdery mildew spore population (Fletcher and Wolfe, 1981~. In an intensive survey one or a few sites are visited, and many smaller samples perhaps comprising single disease lesions or even spores, single insects, or single weed seeds- are collected on several occasions. Often, it is best that an extensive survey be done first, followed by a more detailed study if necessary. These two approaches are complementary, however, and it may be advantageous to use both concurrently or to adopt an intermediate method. Information gathered at each sampling site should include the types, tim- ing, and effectiveness of past chemical treatments and the amounts of target pests, disease, or weeds present. Differences in these factors should be compared with differences in sensitivity.
306 DETECTION, MONITORING, AND RISK ASSESSMENT Sample size should relate to the circumstances. If searching for first signs of resistance in a largely sensitive population, a large bull; sample is more likely to find the "needle in a haystack." To determine the proportion of resistant forms in a population or the differences in degree of resistance, a number of small, specific samples should be tested. Samples should be as fresh as possible, and repeated culture in the absence or presence of chemical should be avoided or minimized. One way to achieve this, which is particularly useful for obligate parasitic fungi, is to place treated test plants in pots in the field crop, allow them to collect inoculum, and then remove them for incubation in a controlled-environment facility or glasshouse to determine response. Conversely, it is valuable to retest samples after repeated subculture in vivo or in vitro to check for genetic stability of response. For increased accuracy and to check degree of resistance, it is generally best to use a range of concentrations during initial testing rather than a single, arbitrary, discriminating dose. The response can be scored in various ways. The EDso value is often used; it is a good "general purpose" value that is widely understood and can be measured relatively accurately, compared with an ED's value. For large-scale surveys, however, and particularly where responses of sensitive and resistant forms are well separated (as with some fungicide and herbicide resistance and most insecticide and rodent resistance), the use of a single discriminating dose permits quick and adequate testing. When resistance is clear-cut, different methods tend to reveal similar trends; only in marginal cases does the method of testing or scoring affect the picture. It is advantageous where possible, however, for one agreed method to be used by different workers nationally or internationally. The WHO standard tests for insecticide resistance in a range of insects of public health importance (WHO, 1970, 1980) have been used internationally since the first test, on anopheline mosquitoes, was introduced about 27 years ago. Test kits, based on diagnostic test dosages for susceptible, fully resistant, and sometimes intermediate populations, are available at cost for about a dozen pest species, including rodents. FAO-recommended methods to measure pest resistance in crop and livestock production and in crop storage have also been adopted widely: Busvine (1980) has drawn together details of tests against 20 im- portant pests, published at intervals since 1969 in the FAO Plant Protection Bulletin; more recent issues of the bulletin contain new or updated procedures. Recommended methods for testing fungicide resistance in crop pathogens have also been published by FAO (1982), and general reviews of procedures are given by Georgopoulos (1982a) and Ogawa et al. (19831. During testing it is important to investigate differences in pathogenicity, growth rate, reproductive rate, and other properties that contribute to the fitness of an organism. Often the more highly resistant forms are less fit or competitive than normal forms in the absence of chemical treatment, and knowledge of this can help to explain and predict their behavior.
DETECTION AND MONITORING OF RESISTANT FORMS 307 Biochemical methods for detecting and monitoring resistant forms have been developed for insecticides and are increasingly used in surveys (Miyata, 1983; Devonshire and Moores, 19841. In some situations they can detect resistance at lower frequencies than do bioassays. They can also be more convenient and permit the degree of resistance to be measured quantitatively without the need to test several samples at different doses. Inhibition of photosystem II, as revealed by loss of chlorophyll fluorescence of herbicide- treated leaves, leaf discs, or isolated chloroplasts irradiated with short wave- length light, has proved a convenient method for monitoring atrazine-resistant weeds (Gasquez and Barralis, 1978, 1979~. Another rapid method for testing response to photosynthesis inhibitors is the sinking-leaf disc technique. The buoyancy of discs floated on surfactant solutions appears to depend on the O2/CO2 ratio in the air spaces, which is decreased by the action of herbicides (Hensley, 1981~. Biochemical monitoring is not yet used for fungicide re- sistance because mechanisms of resistance for field isolates are not well characterized and appear to involve changes at biosynthetic or genetic sites that are not easily detected. More research on this aspect seems justified. Specific diagnostic agents, such as cDNA probes or monoclonal antibodies, may offer new possibilities for future biochemical tests for all types of target organisms (Hardy, this volume). As pointed out by Truelove and Hensley (1982), however, biochemical methods should be used with caution, since resistance that depends on alternative mechanisms to the method under test could be missed; in this respect, bioassay tests on whole organisms remain the most reliable indicators of resistance. ACHIEVEMENTS IN RESISTANCE MONITORING Only a few examples of the many monitoring projects done in different countries and on different target organisms can possibly be considered here. Since the first case of insecticide resistance was reported by Melander in 1914 (Melander, 1914), response to insecticides has been monitored exten- sively in many countries (Georghiou and Mellon, 19831. Global programs have been organized by WHO to survey insecticide resistance in anopheline mosquitoes (WHO, 1976, 1980) and by FAO to survey insecticide resistance in pests of stored grain (Champ and Dyte, 1976) and acaricide resistance in ticks (FAO, 1979~. These very large projects have provided valuable infor- mation on the geographic distribution and intensity of resistance, on its relationships to the successful use of chemicals, and to failures in control. The coordination and interpretation of results have benefited greatly from the general use of recommended methods of testing and reporting mentioned earlier. Many national surveys have been conducted. An outstanding example is the study of resistance in house flies on farms in Denmark, discussed in this volume by Keiding, which has been sustained since 1948 and has shown
308 DETECTION, MONITORING, AND RISK ASSESSMENT clearly the large-scale shifts in response to successive introductions of dif- ferent types of insecticide (organochlorines, organophosphorus compounds, and pyrethroids). Other notable programs have included studies of rice leaf- hoppers and planthoppers in Japan (Hama, 1980), cotton leaf worm in Egypt (El-Guindy et al., 1975), and the aphid Myzuspersicae in the United Kingdom (Sawicki et al., 1978~. In the last study biochemical (esterase-4) tests as well as bioassays were used; both approaches gave rapid and satisfactory results and to some extent were complementary in distinguishing different types of resistance. International surveys comparable with those undertaken with pests have not been done for fungi. Although some recommended methods have been published by FAG, in practice a variety of test methods have been used by different workers. National or regional programs have included surveys of resistance of cucumber powdery mildew to dimethirimol in glasshouses in Holland (Bent et al., 1971) and later to other systemic fungicides (Schepers, 1984), the response of barley powdery mildew to ethir~mo} in the United Kingdom (Shepherd et al., 1975; Heaney et al., 1984) and to triazole fun- gicides (Fletcher and Wolfe, 1981; Heaney et al, 1984; Wolfe et al., 1984), of metalaxy} resistance in Phytophthora infestans on potatoes in Holland (Davidse et al., 1981) and in the United Kingdom (Carter et al., 1982), benomyl resistance in sugar beet leaf spot in Greece (Georgopoulos, 1982b), and dicarboximide resistance in Botrytis on grape vines in West Germany (Lorenz et al., 19811. Each of these studies, as well as others not mentioned here, to some extent tells an individual story. Two main patterns can perhaps be distinguished: a rapid, widespread, and persistent upsurge of resistance and loss of disease control (as with dimethirimol and cucumber powdery mildew, metalaxy} and P. infestans in Holland, benomy! and sugar beet leaf spot) and a slower, fluctuating increase in resistance, with either partial or undetected loss of disease control (as in the cases of ethirimo! or triazoles and barley powdery mildew, metalaxyl and P. infestans in the United King- dom, and dicarboximides and Botrytis). The intensity and exclusivity of fungicide use and the degrees of resistance and fitness of the resistant forms are important factors in determining these patterns. In the former cases mon- itoring tended to follow reports of loss of control and results were obtained too late to permit any management strategy other than withdrawal of the product, but in the latter, where monitoring preceded any major breakdown in performance, avoidance strategies either were already operating or were introduced following the results of monitoring. Since the early 1970s the incidence of triazine-resistant biotypes of various weeds in different crops has been monitored extensively in different parts of the United States, mainly by collecting seeds and growing progeny for glass- house tests. The initial indications of resistance, obtained after 10 years of widespread use of these herbicides, came from farmer observations of obvious
DETECTION AND MONITORING OF RESISTANT FORMS 309 lack of control; the monitoring has served primarily to confirm resistance and to follow the problem in time and space (Bandeen et al., 19821. Atrazine resistance has also been observed in monitoring studies in several countries of continental Europe (Gressel et al., 19821. The rate of development of resistance appears to have varied between different parts of the United States and has been relatively slow in the United Kingdom (Putwain et al., 1982~. Forms resistant to other herbicides, for example, phenoxy compounds and bipyridyls, have been detected in different countries, but their incidence has been sporadic, their resistance less marked, and little monitoring has been done. COOPERATION AND COMMUNICATION Detection of and monitoring for resistance call for close cooperation be- tween scientists as individuals and as representatives of industrial and public- sector organizations. Although coordination does take place, such as in the work of the Fungicide Resistance Action Committee (FRAC) and Insecticide Resistance Action Committee (IRAC), much of the research is still too frag- mented and haphazard. Industry has felt it has been excluded from some collaborative schemes and planning meetings organized by the public sector, but, equally, the RAC system does not fully involve the public sector, since it is primarily an intercompany concern. There is much that scientists in industry and the public sector can do to increase contact, review progress and priorities, and plan collaborative research. Such collaboration would be best focused on particular resistance problems and should be in work groups rather than in conferences, with one person or organization as the focal point for each topic. At this time of retrenchment of national research expenditures in many countries, the selection of priorities in resistance monitoring which despite its importance is an expensive and essentially defensive area of re- search-is especially important. The results of monitoring programs should be reported in the open scientific literature, not retained in confidential reports or computer files. The storage of information from many sources in a data bank from which it can be retrieved and disseminated readily is valuable, however; the data bank for insecticide resistance at the University of California (Riverside) is a good example (Georghiou, 1981~. Education in resistance monitoring is improving. Conferences are helpful, but the international courses on fungicide resistance organized by Professor Dekker and colleagues and held at Wageningen and more recently in Ma- laysia- have proved particularly useful, since they included laboratory ses- sions and a tactical exercise in addition to lectures and group discussions. Perhaps similar courses could be organized on insecticide and herbicide resistance.
310 DETECTION, MONITORING, AND RISK ASSESSMENT CONCLUSION Detection and monitoring form an integral part of pesticide resistance management. To avoid misunderstanding and waste of effort, very careful definition, planning, and interpretation of these activities are required. Mon- itoring denotes different operations, ranging from global surveillance pro- grams to much smaller investigations of cases of suspected resistance. Distinction must be made between detecting resistant forms and establishing that resistance has reached levels of severity and frequency sufficient to cause practical loss of pesticide performance. Cr~tena for defining resistance and sensitivity have differed greatly, especially when several different degrees of resistance occurred, and must always be made clear. Test methods should be developed and initial sensitivity data sought before new compounds are brought into widespread use; avoidance strategies should also be established prior to widespread use, since monitoring cannot be relied on to give sufficient early warning of the need for such strategies. Subsequent monitoring should be done if risks are considered high, if the particular pest-control system is especially important, or when visible signs of resistance problems arise. Selection of test procedures will depend on the nature of the pest and of the pesticide treatment, but the adoption of inter- nationally recommended methods aids the comparison and coordination of results. Biochemical methods have already proved useful and have a prom- ising future. Further collaboration between and within the industrial and public sectors in planning and conducting monitoring programs must be fostered. ACKNOWLEDGMENTS The author is grateful to a number of persons for providing information, and especially to Drs. A. Devonshire, G. P. Georghiou, H. LeBaron, and L. R. Wardlow. REFERENCES Bandeen, J. D., G. R. Stephenson, and E. R. Coweet. 1982. Discovery and distribution of herbicide resistant weeds in North America. Pp. 9-19 in Herbicide Resistance in Plants, H. M. LeBaron and J. Gressel, eds. New York: John Wiley and Sons. Bent, K. J., A. M. Cole, J. A. W. Turner, and M. Woolner. 1971. Resistance of cucumber powdery mildew to dimethirimol. Pp. 274-282 in Proc. 6th Br. Insectic. Fungic. Conf., Vol. 1, Brighton, England, 1971. 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. Busvine, J. R. 1980. Recommended methods for measurement of pest resistance to pesticides. FAG Plant Prod. and Prot. Paper No. 21.
DETECTION AND MONITORING OF RESISTANT FORMS 311 Carter, G. A., R. M. Smith, and K. J. Brent. 1982. Sensitivity to metalaxyl of Phytophthora infestans populations in potato crops in southwest England in 1980 and 1981. Ann. Appl. Biol. 100:433-441. Champ, B. R., and C. E. Dyte. 1976. Report of the FAO global survey of pesticide susceptibility of stored grain pests. FAO Plant Production and Protection Paper No. 5. Davidse, L. C., D. Looigen, L. J. Turkensteen, and D. van der Wall 1981. Occurrence of metalaxyl- resistant strains of Phytophthora infestans in Dutch potato fields. Neth. J. Plant Pathol. 87:65- 68. Denholm, I., R. M. Sawicki, and A. W. Farnham. 1984. The relationship between insecticide resistance and control failure. Pp. 527-534 in Proc. Br. Crop Prot. Conf. Pests and Dis., Vol. 2, Croydon, England: British Crop Protection Council. Devonshire, A. L., and G. D. Moores. 1984. Immunoassay and carboxylesterase activity for iden- tifying insecticide resistant Myzas persicae. Pp. 515-520 in Proc. Br. Crop Prot. Conf. Pests and Dis., Vol. 2, Croydon, England: British Crop Protection Council. du Pont. 1982. Black and Yellow Sigatoka, Improved Identification and Management Techniques. Coral Gables, Fla.: du Pont Latin America. El-Guindy, M. A., G. N. El-Sayed, and S. M. Madi. 1975. Distribution of insecticide resistant strains of the cotton leafworm Spodoptera littoralis in two governorates of Egypt. Bull. Entomol. Soc. Egypt 9:191-199. Fletcher, J. T., andM. S. Wolfe. 1981. Insensitivity of Erysiphegraminis f. sp. hordei totriadimefon, triadimenol and other fungicides. Pp. 633-640 in Proc. Br. Crop Prot. Conf. Pests and Diseases, Vol. 2, Croydon, England: British Crop Protection Council. Food and Agriculture Organization. 1979. Pest resistance to pesticides and crop loss assessment. FAO Plant Production and Protection Paper No. 6/2. Food and Agriculture Organization. 1982. Recommended methods for the detection and measurement of resistance of agricultural pests to pesticides. Plant Prot. Bull. 30:36-71, 141-143. Gasquez, J., andG. Barralis. 1978. Observation et selection chezChenopodium album L. dtindividus resistants aux triazines. Chemosphere 11:911-916. Gasquez, J., and G. Barralis. 1979. Mise en evidence de la resistance aux triazines chez Solanum nigrum L. et Polygonum lapathifolium L. par observation de la fluorescence de feuilles isolees. C. R. Acad. Sci. (Paris) Ser. D 288:1391-1396. Georghiou, G. P. 1981. The occurrence of resistance to pesticides in arthropods: An index of cases reported through 1980. Rome: FAO. Georghiou, G. P., and R. B. Mellon. 1983. Pesticide resistance in time and space. Pp. 1-46 in Pest Resistance to Pesticides, G. P. Georghiou and T. Saito, eds. New York: Plenum. Georgopoulos, S. G. 1982a. Detection and measurement of fungicide resistance. Pp. 24-31 in Fungicide Resistance in Crop Protection, J. Dekker and S. G. Georgopoulos, eds. Wageningen, Netherlands: Centre for Agricultural Publishing and Documentation. Georgopoulos, S. G. 1982b. Case study I: Cercospora beticola of sugar beet. Pp. 187-194 in Fungicide Resistance in Crop Protection, J. Dekker and S. G. Georgopoulos, eds. Wageningen, Netherlands: Centre for Agricultural Publishing and Documentation. Gressel, J., H. U. Ammon, H. Fogelfors, J. Gasquez, Q. O. N. Kay, and H. Kees. 1982. Discovery and distribution of herbicide-resistant weeds outside North America. Pp. 32-55 in Herbicide Resistance in Plants, H. M. LeBaron and J. Gressel, eds. New York: John Wiley and Sons. Hama, H. 1980. Mechanism of insecticide resistance in green rice leafhopper and small brown planthopper. Rev. Plant Prot. Res. (Japan) 13:54-73. Heaney, S. P., G. J. Humphreys, R. Hutt, P. Montiel, and P. M. F. E. Jegerings. 1984. Sensitivity of barley powdery mildew to systemic fungicides in the UK. Pp. 459-464 in Proc. Br. Crop Prot. Conf. Pests and Diseases, Vol. 2, Croydon, England: British Crop Protection Council. Hensley, J. R. 1981. A method for identification of triazine resistant and susceptible biotypes of several weeds. Weed Sci. 29:70-78.
312 DETECTION, MONITORING, AND RISK ASSESSMENT Hunter, T., K. J. Brent, and G. A. Carter. 1984. Effects of fungicide regimes on sensitivity and control of barley mildew. Pp. 471-476 in Proc. Br. Crop Prot. Conf. Pests and Diseases, Vol. 2, Croydon, England: British Crop Protection Council. Lorenz, G., K. J. Beetz, and R. Heimes. 1981. Resistenzentwicklung van Botrytis cinerea gegenuber Fungiziden auf Dicarboximid-Basis. Mitt. Biol. Bundesanst. Land- Forstwirtsch., Berlin-Dahlem 203:278-285. Melander, A. L. 1914. Can insects become resistant to sprays? J. Econ. Entomol. 7:167-174. Miyata, T. 1983. Detection and monitoring methods for resistance in arthropods based on biochemical characteristics. Pp. 99-116 in Pest Resistance to Pesticides, G. P. Georghiou and T. Saito, eds. New York: Plenum. Ogawa, J. M., B. T. Manji, C. R. Heaton, J. Petrie, and R. M. Sonada. 1983. Methods for detecting and monitoring the resistance of plant pathogens to chemicals. Pp. 117-162 in Pest Resistance to Pesticides, G. P. Georghiou and T. Saito, eds. New York: Plenum. Putwain, P. D., K. R. Scott, and R. J. Holliday. 1982. The nature of the resistance to triazine herbicides: Case histories of phonology and population studies. Pp. 99-116 in Herbicide Resistance in Plants, H. M. LeBaron and J. Gressel, eds. New York: John Wiley arid Sons. Sawicki, R. M., A. L. Devonshire, A. D. Rice, G. D. Moores, S. M. Petting, and A. Cameron. 1978. The detection and distribution of organophosphorus and carbamate insecticide-resistant Myzas persicae (Sulz.) in Britain in 1976. Pestic. Sci. 9:189-201. Schepers, H. T. A. M. 1984. Resistance to inhibitors of sterol biosynthesis in cucumber powdery mildew. Pp. 495-496 in Proc. Br. Crop Prot. Conf. Pests and Diseases, Vol. 2, Croydon, England: British Crop Protection Council. Shephard, M. C., K. J. Brent, M. Woolner, and A. M. Cole. 1975. Sensitivity to ethirimol of powdery mildew from UK barley crops. Pp. 59-66 in Proc. 8th Br. Insectic. Fungic. Conf., Brighton, 1975. Skylakakis, G. 1984. Quantitative evaluation of strategies to delay fungicide resistance. Pp. 565- 572 in Proc. Br. Crop Prot. Conf. Pests and Diseases, Vol. 2, Croydon, England: British Crop Protection Council. Truelove, B., and J. R. Hensley. 1982. Methods of testing for herbicide resistance. Pp. 117-131 in Herbicide Resistance in Plants, H. M. LeBaron and J. Gressel, eds. New York: John Wiley and Sons. World Health Organization. 1970. Insecticide resistance and vector control. 17th Rep. WHO Exp. Comm. on Insectic. WHO Tech. Rep. Ser. No. 443. World Health Organization. 1976. Resistance of vectors and reservoirs of disease to pesticides. 22nd Rep. WHO Exp. Comm. on Insectic. WHO Tech. Rep. Ser. No. 585. World Health Organization. 1980. Resistance of vectors of disease to pesticides. 5th Rep. WHO Exp. Comm. on Vector Biol. Contr. WHO Tech. Rep. Ser. No. 655. Wolfe, M. S., P. M. Minchin, and S. E. Slater. 1984. Dynamics of triazole sensitivity in barley mildew nationally and locally. Pp. 465-470 in Proc. Br. Crop Prot. Conf. Pests and Diseases, Vol. 2, Croydon, England: British Crop Protection Council.