.. no. A .m. vnc:n.wc 'Tmi‘ L133. THESiS SITY Ll IBRAF“ |SE it'll” \\\\\\\\\\\\\\\\l\\\\\\l\\lllllllllllll 3129 This is to certify that the thesis entitled Biological Control of Tetranychus Urticae Koch with Phytoseiulus Persimilis Athias-Henriot in a Rose Production System presented by Anthony J. D'Angelo has been accepted towards fulfillment of the requirements for Master's degree in Entomology [QM/v Major pro [hue L/ Aung_‘77/ \J l 0-7639 MSU is an Affirmative Action/Equal Opportunity Institution 3 UBERAR‘Y_ Michigan State LUniversity PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE MSU Is An Affirmative Action/Equal Opportunity Institution czwImema-ot BIOLOGICAL CONTROL OF TETRANYCHLJS URTICAE KOCH WI'I'H PHYTOSETULUS PERSIMILIS ATHIAS-HENRIOT [N A ROSE PRODUCTION SYSTEM By Anthony John D'Angelo A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Entomology 1991 éayvsag/ ABSTRACT BIOLOGICAL CONTROL OF TETRANYCHUS URTICAE KOCH WITH PHYTOSEIULUS PERSIMILIS ATHIAS-HENRIOT IN A ROSE PRODUCTION SYSTEM By Anthony John D'Angelo Twospotted spider mites have been a major pest of greenhouse grown crops for at least 40 years with heavy infestations resulting in serious economic losses. Fluctuations in the intensity of spider mite infestations can be attributed to the development of mite resistance to previously effective miticides. Standard slide-dip and 36 h residue tests were used to evaluate selectivity of commonly used miticides, fungicides and insecticides against twospotted spider mite and the predator mite, Phytoseiulus persimilis. Longer term efficacy studies and predator feeding studies were found to be useful in determining indirect toxicity of miticides to the predator. Absolute population densities and distributions of spider mites on 14 rose plants were determined. For research purposes, random samples of 20 leaflets per plant will provide an acceptable estimate of population density. Sampling schemes designed to detect mites at low density should be aimed at the base of rose plants where the oldest leaves are found. To my family iii ACKNOWLEDGEMENTS I thank my major professor, Dr. David Smitley, for his patience, support and encouragement throughout my graduate work. I thank the members of my committee, Dr. George Ayers, Dr. Gary Simmons and Dr. Dean Krauskopf for their constructive criticism and technical expertise. A special thanks to George Ayers for allowing me to exercise and enhance my teaching skills by assisting him in ENT 302 and to Dr. Mark Scriber and Dr. Fred Stehr for giving me the opportunity to do so. I am grateful to all the employees of the "Smitley Lab" for all their help in completion of my project. My special thanks to Terry and Maria Davis and Kim Kearns, not only for their all their help on this project but mostly for their friendship. Thanks to all the members of the fourth floor family and all the members of the department that I have been associated with over the last 8 years. There are far too many to acknowledge individually. However, I would like to personally thank my office mates Joan Davis, Adam Peters, John (The Skink) Wilterding, Kaja Brix and James Jasinski, and 4th floor members Beth Bishop, Walter Boylan-Pett and Matt Hohmann (zoologist) for their support, friendship and for putting up with my eccentricities. I would also like to thank Dr. Robert Ruppel for giving me my first oppurtunity to work and develop my skills as an entomologist. Finally I would like to thank my parents, John and Verna D'Angelo, and the rest of my family for making the completion of this degree possible. Without their support, encouragement and love this degree could not have been completed. iv TABLE OF CONTENTS List of Figures ........................................................................... vii List of Tables ....................... viii Introduction ................................................................................ 1 Article 1: Determining the selectivity of the miticides avermectin, dienochlor and propargite to Tetranychus urticae and Phy. toseiulus persimilis ................................................................................... 12 Abstract ............................................................................ 13 Introduction ....................................................................... 14 Materials and Methods ........................................................ 15 Slide-dip test ............................................................. 16 36 hour residue test ................................................... l7 Predator feeding study ............................................... 18 Long-term population studies ..................................... 19 Statistical analysis ...................................................... 20 Results .............................................................................. Slide-dip test ............................................................. 20 36 hour residue test ................................................... 21 Predator feeding study ............................................... 22 Long-term population studies ..................................... 23 Discussion ......................................................................... 23 Literature cited .................................................................. 29 Article 2: Selectivity of benomyl, dodemorph acetate, piperalin,- triadimefon, triforine, acephate and endosulfan for Phytoseiulus persimilis or Tetranychus urticae ................................................................. 37 Abstract ............................................................................ 38 Introduction ....................................................................... 39 Materials and Methods ........................................................ 4O Slide-dip test ............................................................. 4O 36 hour residue test ................................................... 42 Statistical analysis ...................................................... 43 Results .............................................................................. Slide-dip test ............................................................. 44 36 hour residue test ................................................... 44 Discussion ......................................................................... 45 Literature cited .................................................................. 47 Article 3: Estimating population densities of twospotted spider mites, Tetranychus urticae, on greenhouse grown roses .......................................................................................... 53 Abstract ............................................................................ 54 Introduction ....................................................................... 55 Material and Methods ......................................................... 56 Numbering system ..................................................... 57 Spider mite counts ..................................................... 57 Sampling .................................................................. 57 Data analysis ............................................................. 58 Results and Discussion ........................................................ 59 Literature cited .................................................................. 62 Summary and Conclusion ............................................................. 73 vi LIST OF FIGURES ARTICLE 1: Figure 1: Slide-dip evaluation of three miticides commonly used in greenhouse rose production .......................................................... 33 Figure 2: Mortality of I. urticae and _E. persimilis on bean leaves treated with three miticides ..................................................................... 34 Figure 3: Mortality of P. persimilis fed spider mites surviving applications of averrnectin, dienochlor or propargite ........................................ 35 Figure 4: Responses of populations of I. urticae over a 15 day period following treatment with dienochlor .............................................. 36 ARTICLE 3: Figure l: A leaf numbering system for rose plants ......................... 67 Figure 2: Mean number of spider mites per leaf in relation to leaf position ...................................................................................... 68 Figure 3: Actual and sample estimates of mean mites per leaflet for rose plants with different population densities ....................................... 69 Figure 4: Actual and sample estimates of mean mites per leaflet for rose plants with different population. densities. Samples of leaflets 3 and 4 from leaves 1 and 2. Sample means multiplied by conversion factor 0.548 ......................................................................................... 70 Figure 5: Actual and sample estimates of mean mites per leaflet for rose plants with different population densities. Samples of leaflets 3 and 4 from leaves 1 and 2 ............................................................................. 71 Figure 6: Average variance/sample mean for rose plants with different population densities ..................................................................... 72 vii LIST OF TABLES ARTICLE 1: Table 1: Number of _'I_‘. urticae and _12. persimilis per sample on potted rose plants treated with abamectin ................................................ 32 ARTICLE 2: Table l: Contact activity of five fungicides to I. urticae and B. persimilis ................................................................................... 49 Table 2: Residual activity of five fungicides to I. urticae and P. persimilis ................................................................................... 50 Table 3: Regression statistics for contact and residual activity of five fungicides to I. urticae and _P_.persimilis ........................................ 51 Table 4: Regression statistics for contact and residual activity of acephate and endosulfan to I. urticae and _P_. persimilis ................... 52 ARTICLE 3: Table 1: Sample means of selected leaflet groups compared to mean mites per leaflet of entire plant ..................................................... 63 Table 2: Accuracy and variance of the mean number of mites per leaflet for 10, 20, and 30 leaflet samples from rose plants with various mite densities ..................................................................................... 64 Table 3: Accuracy and variance of the mean number of mites per leaflet for 10, 20, and 30 leaflet samples from rose plants with various mite densities, sampling leaflets 3 and 4 from leaves 1 and 2 only ............ 65 viii Table 4: Accuracy and variance of the mean number of mites per leaflet for 10, 20, and 30 leaflet samples from rose plants with various mite densities, sampling leaflets 3 and 4 from leaves 1 and 2 only. Sample means are multiplied by 0.548 to reflect populations per plant ......... 66 INTRODUCTION Predaceous Phytoseiid mites are probably the most important biological control agents of spider mites in deciduous orchards, vineyards, and greenhouses around the world (Hoy, 1982). Control programs utilizing natural occurring Phytoseiids have been successful on a number of orchard and field crops. Typhlodromus occidentalis Nesbitt was first used successfully on apples where mite control costs have been significantly reduced (Hoyt & Caltagirone 1971). Important factors contributing to the success of this program were; the predator developed resistance to azinphosmethyl used for codling moth, Laspegresia pomonella L., and an alternate food source, the apple rust mite, Aculus schlechtendali (Nalepa), was present and could maintain I. occidentalis when Tetranychus macdanieli McGregor was scarce (McMurtry 1982). By applying the same principles used on apples, similar programs utilizing I. occidentalis have been successful on peaches (Hoyt & Caltagirone 1971), pears (Westigard 1971), grapes (Flaherty & Huffaker 1970, Kinn & Dout 1972), and walnuts (McMurtry & Flaherty 1977). Amblyseius fallagis (Garman) has been used successfully to control European red mite, Panonyghus m (Koch) in apples when ground cover maintained for overwintering sites provides an alternate Tetranychus host. A. Lam moves into the trees in spring where it is effective in controlling _B. um as well as Tetranychus species (Croft & McGroarty 1973, Meyer 1974) . As with I. occidentalis an important factor in its success is its resistance to certain pesticides (McMurtry 1982). Typhlodromus pyri Scheuten is another predator which occurs on apple trees in various parts of the world (Schuster & Pritchard 1963, Specht 1968). Although I. pygi is recognized as an effective predator of P. u_lr_n_i_ in some situations (Downing & Moilliet 1972), its use in integrated mite control programs began only after developement of resistance to various pesticides (Croft 1982). A number of other naturally occurring Phytoseiid mites have shown potential for spider mite control on a variety of crops (Oatman 1973). Over the last two decades interest in biological and integrated pest management programs utilizing mass production and release of Phytoseiid mites has increased significantly. (van Lenteren & Woets 1988). Twenty years ago there were few commercial producers of natural enemies. As of 1988 there were about 15 large producers supplying beneficial arthropods (van Lenteren & Woets 1988). Large scale production and transportation techniques have been described for a variety of Phytoseiid mites (Kamborov 1966, Scriven & McMurtry 1971, Hoy et al. 1982, Hussey & Scopes 1985). Biological control of spider mites in field crops by mass release of predators has been largely developed on strawberries (McMurtry 1982). Good control has been achieved in several studies using predator mites. On strawberries, Phytoseiulus persimilis Athias-Henn'ot provided good control of spider mites at a rate of 5 to 10 mites per plant (Oatman & McMurtry 1966, Oatrnan et a1. 1968, 1976, 1977). Phytoseiulus persimilis has also shown good control of spider mites on field grown ivy and dahlias (Gould & Light 1971). Interest in biological control in greenhouses did not begin until 1926 when a tomato grower noticed blackened pupae among the normal white pupae of the greenhouse whitefly, Trialeurodes vapgrariorum Westwood. The parasites that emerged were identified as Encarsia formosa (Speyer 1927). Within a few years a research station in England was supplying 1.5 million of these parasites annually to 800 nurseries in Britain. Use of E. formosa spread to other countries until distribution was discontinued when more convenient and efficient insecticides were introduced after World War II (van Lenteren & Woets 1988). However, the first signs of pesticide resistance in spider mites brought about renewed interest in biocontrol. Since then considerable advances have been made using Phytoseiid mites to control spider mites on greenhouse grown crops (van Lenteren & Woets .1988). Phytoseiulus persimilis is the most widely used natural enemy for spider mite control in greenhouses (van Lenteren & Woets 1988). Interest in the potential of B. persimilis as a biocontrol began when Dosse (1958) reported on its high reproductive rate and efficiency in reducing spider mite populations (McMurtry 1982, van Lenteren & Woets 1988). This potential was again demonstrated by a number of experiments using P. persimilis to control spider mites on greenhouse grown crops (Chant 1961, Bravenboer & Dosse 1962, Bravenboer 1963). Small-scale application of spider mite biological control in commercial greenhouses started in 1968. (van Lenteren & Woets, 1988). By 1980 use of B. persimilis in Europe on greenhouse grown cucumbers had increased to over 60% of the acreage in the Netherlands, 75% in England and 75 % in Sweden, Denmark (McMurtry 1982, van Lenteren & Woets 1988) and Finland (Markkula & Tiittanen 1980). Programs have also developed in British Columbia, Canada (Tonks & Everson 1977, Costello et al. 1984) and the USSR (Beglyarov & Smetnik 1977). Commercially B. persimilis has also been used on greenhouse grown peppers (van Lenteren & Woets, 1988), and tomatoes (French et a1. 1976, Gould 1977, Stenseth 1980). In research tests B. persimilis has been effective on Strawberries (Simmonds 1971, Gould & Vernon 1978, Port & Scopes 1981) and Chrysanthemums (Hamlen & Lindquist 1981). In the United States, grower interest in greenhouse biocontrol has increased substantially in the last 10 years (Osborne et al., 1985), although very few commercial greenhouses have actually tried biocontrol as a pest control approach. In greenhouse rose production the major pests include the twospotted spider mite, Tetranychus urticae Koch, rose aphids, Macrosiphum rcLae (L.), western flower thrips, _Flamfiiniella _o_c_c_i_denta_l_i_s_ Pergande, and powdery mildew, Sphaerotheca pannosa. Twospotted spider mites have been a major pest of greenhouse grown crops for many years, with heavy infestations resulting in serious economic loss (Anonymous 1961, Hamlen & Lindquist 1981, Field & Hoy 1984). On greenhouse grown roses, as many as 19 miticide applications per year are applied by some growers (Field & Hoy 1984). The development of mite resistance to previously effective miticides has caused fluctuations in the intensity of spider mite problems on roses and other crops (McMurtry et a1. 1970, Brador 1977, Georghiou & Saito 1983). Increased interest in biocontrol has resulted not only from miticide control failures, but also from pesticide toxicity to workers, phytotoxicity to plants, and increased government regulations. In order to implement a successful integrated pest management program in a greenhouse a grower must have a detailed plan. He must know what predators to use, when to release them, and how many to release. Previous studies have determined the predator/prey ratio necessary for successful biocontrol (Oatman et a1. 1976). A grower must know what selective pesticides can be used and when they should be used. The primary focus of this study was to evaluate pesticides commonly used in commercial rose production greenhouses for their relative toxicity to B. persimilis and I. urti_ca_e. Secondly, biocontrol of spider mites requires more intensive monitoring of mite populations than chemical control. Accurate sampling methods are needed for researchers studying biocontrol of mites on roses, and for greenhouse growers trying to detect spider mites at low densities to properly time release of predator mites. A quantitative approach to biological control of spider mites using predaceous mites requires estimating population densities (Nachman 1984). To be useful, a sampling program should provide estimates having the highest accuracy commensurate with the amount of work expended (Southwood 1978). Intelligent decisions concerning pest control strategies are directly dependent upon knowing the status of all important species GBeardsley et a1. 1979). In order to determine if a successful integrated pest management program for I. urticae can be developed it is necessary to know if the release of predators is compatible with natural controls and pesticides used for the other major pest problems. Boys and Burbutis (1972) suggested that with the use of a short-lived selective miticide P. persimilis may have potential value in an integrated pest management program on commercially grown greenhouse roses. The probability of successful initiation of a spider mite biocontrol program could be increased by the use of a selective miticide to suppress spider mite populations. In conclusion, potential problems in initiating predator mite biocontrol programs in rose production greenhouses are: (l) adequate monitoring of spider mite populations; (2) use of fungicides; (3) pesticide residues on foliage; and (4) outbreaks of aphids and thrips. Literature cited Anonymous, 1961. Predators versus spider mites. Rose Incorporated Bulletin, July, P. 20-21. Beardsley, J. W., M. T. AliNiazee, & T. F. Watson. 1979. Sampling and monitoring, pp. 11-22. I_n Biological Control and Insect Pest Management. eds. D. W. Davis, S. C. Hoyt, J. A. McMurtry, and M. T. AliNiazee, Division of Agric. Sc., Univ. of Calif. Boys, F. E. & P. P. Burbitis, 1972. Influence of Phytoseiulus persimilis on populations of Tetranychus turkestani at the economic threshold of roses. J. Econ. Entomol. 65: 114-117. Brador, L. 1977. Resistance in mites and insects effecting orchard crops. I_n Pesticide and Insecticide Resistance. eds. D. L. Watson and A. W. A. Brown, Academic Press, New York. Bravenboer, L. & G. Dosse, 1962. Phytoseiulus riegeli Dosse as a predator of red spider mites of the Tetranychus urticae group. Ent. exp. & appl.,-5: 305-312. Burgess, E. P. J., 1984. Integrated control of two-spotted mite on glasshouse roses. Proc. N. Z. Weed Pest Control Conf. Palmerston North, p. 257-261 Chant, D. A., 1961. An experiment in biological control of Tetranychus telarius (L.) (Acarina: Tetranychidae) in a greenhouse using the predacious mite Phytoseiulus persimilis Athias-Henriot (Phytoseiidae). Can. But, 93: 437-443. Costello, R. A., D. P. Elliot, & N. V. Tonks, 1984. Integrated control of mites and whiteflies in greenhouses. Prov. of B. C. Min. of Agric. and Fd. Croft, B. A., 1976. Establishing insecticide-resistant Phytoseiid mite predators in deciduous tree fruit orchards. Entomophaga, 21(4): 383-399. Croft, B. A., 1982. Arthropod resistance to insecticides: A key to pest control failures and successes in North American apple orchards. Ent. Exp. Appl. 31: 88-110. Croft, B. A. & D. L. McGroarty, 1973. A model study of acaricide resistance, spider mite outbreaks, and biological control patterns in Michigan apple orchards. Environ. Entomol. 2: 633-638. Downing, R. S. & T. K. Moilliet, 1972. Replacement of Typhlodromus occidentalis by I. ELri (AcarinazPhytoseiidae) after cessation of sprays on apple trees. Can. Entomol. 104: 937-940. Field, R. P. & M. A. Hoy, 1984. Biological control of spider mites on greenhouse grown roses. Calif. Agr. March-April, pp. 29-32. Flaherty, D. L. & C. B. Huffaker, 1970. Biological control of Pacific mites and Willamette mites in San Joaquin Valley vineyards. Part 1. Role of Metaseiulus occidentalis. Part H. Influence of dispersion patterns of Metaseiulus occidentalis. Hilgardia 40(10): 267-330. French, N., W. J. Parr, H. J. Gould, I. J. Williams & S. P. Simmonds, 1976. Development of biological methods for the control of Tetranychus urticae on tomatoes using Phytoseiulus persimilis. Ann. appl. Biol. 83: 177-189. Georghiou, P. & T. Saito, 1983. Pest Resistance to pesticides. Plemun Press. New York 809p. Gould, H. J., 1977. Biological control of glasshouse whitefly and red spider mite on tomatoes and cucumbers in England and Wales. Pl. Path. 26: 57-60. Gould, H. J. & W. I. St. G. Light, 1971. Biological control of Tetranychus urticae on stock plants of ornamental ivy. Pl. Path., 20: 18-20. Gould, H. J. & J. D. R. Vernon, 1978. Biological control of Tetranychus urticae (Koch) on protected strawberries using Phytoseiulus persimilis Athias-Henriot. Pl. Path. 27: 136-139. ‘ Hamlen, R. A. & R. K. Lindquist, 1981. Comparison of two Phytosieulus species as predators of twospotted spider mites on greenhouse omamentals. Environ. Entomol. 10: 524-527. Hoy, M. A., 1982. Introduction. h Recent advances in knowledge of the Phytoseiidae. Div. Agr. Sci. Unv. Calif, Publ. 3284. Hoy, M. A., D. Castro, & D. Cahn, 1982. Two methods for large scale production of pesticide-resistant strains of the spider mite predator Metaseiulus occidentalis (Nesbitt) (Acarina: Phytoseiidae). Z. ang. Ent. 94: 1-9. Hoy, M. A., R. T. Roush, K. B. Smith, & L. W. Barclay, 1979. Spider mites and predators in San Joaquin Valley almond orchards. Calif. Agric., 33(10): 11-13. Hoyt, S. C. & L. E. Caltagirone, 1971. The developing programs of integrated control of pests in Washington and peaches in California. In Biological Control. C. B. Huffaker, ed., Plenum Press, N. Y., 395-421. Hussey, M. W. & N. E. A. Scopes, eds. 1985. Biological pest control: The glasshouse experience. Poole, Dorset: Blandford. 240 pp. Kamburov, S. S., 1966. Methods of rearing and tranporting predacious mites. J. Econ. Entomol. 59(4): 875-877. Kinn , D. N. & R. L. Doutt, 1972. Natural control of spider mites on wine grape varieties in Northern California. Environ. Entomol. 1(4): 513-518. Markula, M. & K. Tiittanen, 1976. "Pest in first" and "natural infestation" methods in the control of Tetranychus urticae Koch with Phytoseiulus persimilis A.-H. on glasshouse cucumbers. Ann. Agric. Fenn. 15: 81-85. Martin, N. A., 1987. Progress towards integrated pest management for greenhouse crops in New Zealand. Bull. SROP. 10(2): 111- 115. McMurtry, J. A., 1982. The use of Phytoseiids for biological control. pp. 23-48. In Recent advances in knowledge of the Phytoseiidae. eds. M. A. Hoy. Div. Agr. Sci., Unv. Calif., Publ. 3284. McMurtry, J. A. & D. L. Flaherty, 1977. An ecological study of phytoseiid and tetranychid mites on walnut in Tulare County, California. Environ. Entomol. 6: 287-292. 10 McMurtry, J. A., C. B. Huffaker & M. van de Vrie, 1970. Ecology of tetranychid mites and their natural enemies: I. Tetranychid enemies: Their biological characters and the impact of spray practices. Hilgardia, 40: 331-390. McMurtry, J. A. & G. T. Scriven, 1965. Insectary production of phytoseiid mites. J. Econ. Entomol. 58: 282-284. Meyer, R. H., 1974. Management of phytophagous and predatory mites in Illinois orchards. Environ. Entomol. 3: 333-340. Nachman, G. 1984. Estimates of mean population density and spatial distribution of Tetranychus urticae (Acarina: Tetranychidae) and Phytoseiulus persimilis (Acarina: Phytoseiidae) based upon the proportion of empty sampling units. Journal of Applied Ecology. 21, 903-913. Oatman, E. R., 1973. An ecological study of arthropod populations on apple in northeastern Wisconsin: Population dynamics of mite species on the foliage. Ann. Entomol. Soc. Am. 66: 122-131. Oatman, E. R. & J. A. McMurtry, 1966. Biological control of the two- spotted spider mite on strawberry in southern California. J. Econ. Entomol., 59: 433-439. Oatman, E. R., F. E. Gilstrap & V. Voth, 1976. Effect of different release rates on Phytoseiulus persimilis (Acarina: Phytoseiidae) on the two-spotted spider mite on strawberry in southern California. Entomophaga, 21: 269-273. Oatman, E. R., J. A. McMurtry & V. Voth, 1968. Suppression of the two-spotted spider mite on strawberry with mass releases of Phytoseiulus persimilis. J. Econ. Entomol., 61: 1517-1521. Oatman, E. R., J. A. McMurtry, F. E. Gilstrap & V. Voth, 1977. Effect of releases of Amblyseius califomicus, Phytoseiulus persimilis, and Typhlodromus occidentalis on the twospotted spider mite on strawberry in southern California. J. Econ. Entomol., 70: 45-47. Osborne, L. S., L. E. Ehler, & J. R. Nechols. 1985. Biological control of the twospotted spider mite in greenhouses. Agr. Exp. Sta., Ins. of Food and Agr. Sci., Unv. of F1. GV. 11 Port, C. M. & N. E. A. Scopes, 1981. Biological control by the predatory mites (Phfloseiulus persimilis Athias-Henriot) of red spider mite (Tetranychus urticae Koch) infesting strawberries grown in "walk- in" plastic tunnels. Pl. Path. 30: 95-99. Schuster, R. O. & A. E. Pritchard, 1963. Phytoseiid mites of California. Hilgardia 34(7): 191-285. Simmonds, S. P., 1972. Observations on the control of Tetranychus urticae on roses by Phytoseiulus persimilis. Pl. Path. 21: 163-165. Simmonds, S. P., 1971. Observations on the possible control of Tetranychus urticae on strawberries by Phytoseiulus persimilis. Pl. Path. 20: 117-119. Southwood, T. R. E. 1978. Ecological Methods. Chapman and Hall, New York, N. Y. Specht, H. B., 1968. Phytoseiidae (Acarina: Mesostigmata) in the New Jersey apple orchard environment with descriptions of spermathecae ‘and three new species. Can. Entomol. 100: 673-692. Speyer, E. R., 1927. An important parasite of the greenhouse white-fly (Trialeurodes vaporariorum Westwood). Bull. Entomol. Res. 17: 301-308. ' Stenseth, C., 1979. Effect of temperature and humidity on the development of Phy_toseiulus persimilis and its ability to regulate populations of Tetranyghus urticae (Acarina: Phytoseiidae, Tetranychidae). Entomophaga, 24(3): 311-317. Tonks, N. V. & P. Everson, 1977. Phytoseiulus persimilis (Acarina: Phytoseiidae) for control of two-spotted mites in a commercial greenhouse. J. Entomol. Soc. Brit. Columbia, 74: 7-8. van Lenteren, J. C. & J. Woets, 1988. Biological and integrated pest control in greenhouses. Ann. Rev. Entomol. 33: 239-269. Westigard, P. H., 1971. Integrated control of spider mites on pear. J. Econ. Entomol., 64(2): 496-501. Determining the selectivity of the miticides abamectin, dienochlor and propargite to Tetranyghus urticae or WM’M ' A. J. D'Angelo Department of Entomology Michigan State University East Lansing, Michigan 48824, USA 12 13 Key Words: Acari, Twospotted spider mite, Tetranychus urticae, Phytoseiulus persimilis, miticides, selectivity. ABSTRACT Standard slide-dip and 36 h residue tests were used to evaluate activity of abamectin, propargite and dienochlor against twospotted spider mites, Tetranychus urticae Koch and the predaceous mite, Phytoseiulus persimilis Athias-Henriot. Results showed that abamectin was more toxic to T urticae than _13. persimilis, while dienochlor and propargite had little or no effect on either mite. Dienochlor was shown to be toxic to spider mites in a long term experiment. A predator feeding study, where spider mites from populations initially recovering from a miticide application were fed to predator mites, was found to be useful in determining indirect toxicity of miticides to E. persimilis. 14 INTRODUCTION The twospotted spider mite, Tetranychus urticae has been a major pest of greenhouse crops for at least 40 years. Heavy infestations frequently result in serious economic losses (Anonymous, 1961; Field and Hoy, 1984; Hamlen and Lindquist, 1981). As many as 19 miticide applications per year are applied by some rose growers to control spider mites (Field and Hoy, 1984). Fluctuations in the intensity of spider mite problems on roses and other crops in the last 40 years can be attributed to the development of mite resistance to previously effective miticides (Brador, 1977; Georghiou and Mellon, 1983; McMurtry et al., 1970). In some cases resistance is present in greenhouse and field populations of mites several generations removed from those originally treated (Overmeer et a1. 1975, van Zon & Overmeer 1975). Studies have shown that I. urticae populations in a rose production greenhouse were still resistant to tetradifon and dicofol up to 7 years after the miticides had last been used (Overmeer et al. 1975, Helle and van de Vrie, 1975). In addition to control failure, the issues of labor costs, pesticide toxicity to workers, phytotoxicity to rose plants, and increased government regulations have resulted in an increased interest in biocontrol. The most widely adopted greenhouse predator release strategy is the use of Phytoseiulus persimilis for control of I. urticae. In the United States, grower interest in greenhouse biocontrol has increased substantially in the last 10 years (Osborne et a1. 1985), although very few commercial greenhouses have actually relied on biocontrol as their major pest control approach. 15 Investigators have reported successful control of I. m on roses in research greenhouses by release of Metaseiulus occidentalis (Nesbitt) (Field & Hoy 1984, 1986) or B. persimilis (Simmonds 1972, Hamlen & Lindquist 1981, Burgess 1984, Rasmy & Ellaithy 1988). However, successful biocontrol of spider mites becomes less certain when release programs are scaled up for use in commercial rose greenhouses. Some projects have successfully held spider mite damage to a tolerable level (Burgess, 1984; Martin, 1987), while others have failed (Boys & Burbutis, 1972; Lindquist et al., 1980). Potential problems in initiating predator mite biocontrol programs in rose production greenhouses are: (l) adequate monitoring of spider mite populations; (2) use of fungicides; (3) insecticide residues on foliage; and (4) outbreaks of aphids and thrips. The probability of successful initiation of a spider mite biocontrol program could be increased by the use of a selective miticide to suppress spider mite populations until undesirable pesticide residues are no longer toxic to 12. persimilis. The objective of this study was to evaluate three miticides for relative toxicity to '_I‘_. urticae and B. persimilis. MATERIALS AND METHODS The toxicity of abamectin, dienochlor and propargite to _'I_:. m and E. persimilis was evaluated in three different tests: (1) a standard slide-dip; (2) a 36 h residual activity study on bean leaves; and (3) a predator feeding study. Tetranychus urticae cultures were established from mites collected from the pesticide research center greenhouses at Michigan State 16 University in 1986. Twice each year I. urticae were collected at the same place and added to the culture. Tetranychus urticae were maintained in the green-house and laboratory on 'Henderson Bush' and 'Eastland Bush' lima beans (Phaseolus vulgaris ). Phytoseiulus persimilis was originally obtained from 'Nature Control', Medford, Oregon. The population was maintained under laboratory conditions on excised bean leaves infested with spider mites. The relative humidity (RH) was kept at 95% and the temperature at 27°C. Spider mites were tapped from infested bean leaves to provide a daily food source. Slide-dip test. In slide-dip tests with spider mites, a 2 cm piece of double-sided Scotch® tape (666 07340-3 3M Commercial Office Supply Division, St. Paul, MN 55133-3053) was placed across a standard microscope slide. Scotch® nylon ribbed packaging tape (34-7023-2006—9, 3M Home products Division, St. Paul, MN 55133-3053) was placed on top of the double-sided tape with its sticky side facing upwards. Adult female mites on bean leaves were placed in a C02 filled sealable plastic bag. After 15 minutes the mites were tapped off the leaves onto a Tyler equivalent 32 mesh screen (500um pore size). Shaking the screen allowed immature and male mites to fall through, leaving only adult females. Adult females were removed with a fine paint brush and placed on slides so that their legs could move freely. After each slide was fitted with twenty mites it was placed in a covered plastic holding chamber at 95 % relative humidity (RH) and 27°C. Miticide solutions were prepared at concentrations of 1x, 1/5x, l/25x, l/650x and 0x, where x is the recommended concentration on the USA. federal product label for greenhouse application. The 1x rates for abamectin, dienochlor and propargite were 5.6, 150, and 360 ppm active 17 ingredient, respectively. Three hundred ml of each solution was prepared in a 400ml beaker. Triton AG-98 was added to each solution at a concentration of 0.12%. All treatments were replicated twice and randomized with respect to time of mite placement on slides. Slides with mites were agitated in test solutions for 5 seconds, and placed on edge in a drying rack lined with absorbent paper towel to remove excess moisture from the slides. After 15 minutes the slides were placed back into the holding chamber. After 36 h the mites were examined to determine survival. Each mite was touched with a fine paint brush. If it responded with movement it was considered alive. A similar procedure was followed for the B. persimilis slide-dip tests. However, the nylon ribbed packaging tape did not hold the predators in place. They were able to turn themselves over and crawl off the tape. Duct tape (Nashua 357, Watervleit, N.Y. 12189) has a stronger adhesive than packaging tape and holds the predator mites in place better. This tape was used for all predator mite slide dip tests. Also, 15 _P_. persimilis were used per slide compared with 20 I. urticae. Treatment rates and all other details were the same as discussed for the I. urticae dip tests. 36 hour residue test. Residual activity of abamectin, dienochlor, and propargite were tested by spraying bean leaves at 5 P.M., allowing them to dry overnight, and placing mites on the leaves between 9 and 11 AM. the next day. In each of these tests miticides were applied- at 1x, .5x, .25x, .125x, .0625x and 0x, where 1x is the recommended concentration on the USA. federal product label. As in the slide dip tests, the 1x rates for abamectin, dienochlor and propargite were 5.6, 150, and 360 ppm active ingredient, respectively. The treatments were replicated four times. Treatments were applied in a spray tower equipped with a conveyor belt 18 that moves at a constant rate of 9.lcm/sec. Miticide solutions were applied through a TeeJet 8001-E even flat spray nozzle. Excised bean leaves in petri dishes layered with absorbent cotton were sprayed uniformly by running the petri dish through the spray tower. Miticides were applied beginning with the lowest concentration first. Petri dishes were arranged in numerical order on a large cafeteria tray after treatments were applied. After the spray had dried on the leaves, one rubber hose washer was glued to each leaf using Elmer's Glue-all® (Borden's Inc., Dept. CP, Colombus, Ohio 43215). Thewashers served as arenas to hold the mites in place. Twenty adult female I. urticae and six 2. persimilis were placed in each arena. A small circle of nylon mesh screening was glued over the arenas after the mites were in place. The cotton base was kept moist to prevent desiccation of the leaves. Phytoseiulus persimilis adults were transferred to arenas by working in a 5°C room with a fine paint brush. In all tests the mites were kept in arenas on treated bean leaves for 36 h. After this time the arena covers were removed and mortality data taken. Predator feeding study. Predaceous mites were allowed to feed on spider mites surviving spray treatments to determine if the predators received a toxic dose of miticide from ingesting mite prey. Four greenhouse grown potted rose plants were sprayed to run-off with the miticide to be tested. After three days each plant was inoculated with four spider mite infested bean leaves. The introduced spider mite population rapidly declined at first due to the miticide residue. A small proportion of the introduced mites survived on the treated plants and the mite population began to increase some 4-6 weeks after the initial miticideapplication. At that time spider mites that had been feeding on the miticide treated plants 19 were fed to the predator mites over a five day period. Small (60ml) clear plasic diet cups were used as arenas. Six B. persimilis were placed in each arena along with enough spider mite food to last five days. The diet cups were then sealed with plastic covers. To allow for oxygen exchange a hole had been made in the center of each cover. Fine mesh screening was placed over the hole to keep the mites from escaping. All treatments were replicated six times. Spider mites from untreated plants were added to another set of diet cups for a control treatment. Long-term population studies. Long-term studies of the effect of miticides on I. m and B. persimilis were conducted because the results of slide-dip tests and 36 h residue tests suggested that dienochlor did not have any effect on spider mites. This miticide is known to be slow acting. Therefore, efficacy of dienochlor was evaluated over a longer (15 day) test period. Abamectin was also tested in a long-term mixed population study to see how results of slide-dip and 36 h residue tests compared with survival of I. urticae and E. persimilis on rose plants treated with abamectin. In the dienochlor study, each of eight greenhouse grown potted rose plants were inoculated with four spider mite infested bean leaves. Spider mites were allowed to establish and increase for two weeks. After this time a fifteen leaflet precount sample was taken from each plant. Four plants were then sprayed to run-off with a 150 ppm concentration of dienochlor. Four were left untreated as controls. Fifteen leaflets were sampled from each plant every five days and the number of live I. urticae recorded. The long term residual effect of abamectin on roses was studied by spraying 20 potted rose plants previously infested with I. urticae. Three weeks later 1000 _12. persimilis were released. Tetranychus um and _B. 20 persimilis were sampled approximately every seven days for a 45 day period starting one day before application of abamectin on 23 July. Each sample consisted of 60 leaflets from five rose plants (12 per plant). Each five plant unit was replicated four times. The temperature was maintained at 25 i 2°C. Statistical analysis. An arcsin square root transformation was performed on all percent mortality data before analysis. Dose effect of abamectin, dienochlor and propargite on I. urticae and E. persimilis was analyzed by regressing transformed percent mortality of mites against the concentration of the test solution (Wilkinson 1987). If the slope of this regression was greater than zero (p < 0.05), the miticide was determined to have a dosage effect. In 36 h residue tests, mortality of I. urticae was compared with that of B. persimilis at each miticide concentration with the Student's T-test (Steel and Torrie, 1980). Mite mortality was also regressed against miticide concentration as described for the slide-dip tests. In the predator feeding study the effect of each miticide on R. persimilis was evaluated by testing for differences (Student's T-test) in survival of E. persimilis fed spider mites from miticide treated plants with survival of E. persimilis fed spider mites from untreated plants. In the long term population studies with dienochlor, the number of spider mites per sample from miticide treated plants was compared with~that on control plants at each sample date with an analysis of variance (Steel and Torrie, 1980). RESULTS Slide-dip tests. When mites were dipped in test solutions, abamectin was more toxic to T. urticae than B. persimilis. At a 21 concentration of 0.224 ppm abamectin killed 100% of I. urticae and only 28% of P. persimilis tested (Figure la). When I. urticae mortality was regressed against concentrations of abamectin, the slope of the regression line was greater than zero (Figure 1A, P: 115.94, p < 0.01). Regression of B. persimilis percent mortality against concentration of abamectin gave a slope that was not different from zero (Figure 1, F: 3.92, p= 0.076). Therefore, a dosage effect of abamectin on mite mortality was observed for I. u_rtica;e_ but not for E. persimilis. Dienochlor as evaluated in our slide dip tests had no 36 hour contact activity against I. urticae or E. persimilis. Regression of I. urticae or _12. persimilis against test concentrations of dienochlor yielded slopes that were not different from zero (Figure 1B, F= 0.146, p= 0.710 and F: 0.276, p: 0.611 respectively). * Propargite also had little effect on I. um; or R. persimilis at any concentration in the slide-dip tests. Regression of I. urticae or _P_. persimilis mortality against concentration of propargite gave slopes that were not different from zero (Figure 1C, F= 0.297, p= 0.598 and F: 1.338, p= 0.274 respectively). 36 hour residue tests. Abamectin was more toxic to I. m than to B. persimilis at all concentrations in 36 hour residue tests on bean leaves. When I. urticae mortality was regressed against concentration of abamectin, the slope of the regression line was greater than zero (Figure 2A, F= 30.750, p < 0.001). Regression of _13. persimilis mortality against abamectin concentration gave slopes that were greater than zero at p = 0.10 but not at p = 0.05 (Figure 4, F: 4.227, p=0.052). Although 50% mortality of P. persimilis was observed at 5.6 ppm abamectin, this was not different from the control treatment (T(.05) = 1.445, d.f.= 6). Mortality 22 of I. urticae was significantly higher than that of _R. persimilis at a concentration of 5.6 ppm (T(.05)= 7.659, d.f.= 6). In 36 h residue tests dienochlor had little effect on I. 1m; or B. persimilis, and no selectivity for either species. Regression of I. m mortality against concentrations of dienochlor was not significant. Phytoseiulus persimilis mortality increased slightly as the concentration of dienochlor increased in test solutions (Figure 2B). Although the maximum mortality to _R.persimi1is was only 20%, regression analysis indicates a positive relationship to concentrations of dienochlor (F = 5.5, p = 0.029). Propargite also had no effect on I. m or B. persimilis mortality at any concentration in 36 hour residue tests. Regression of I. m or _I_’_. persimilis mortality against concentrations of propargite gave slopes that were not different from 0 (Figure 2C, P: 3.714, p= 0.067 and F: 0.368, p= 0.550, respectively). Tetranychus urticae mortality was not different from that of B. persimilis at the standard concentration of 360 ppm propargite (T(.05)= 0.659, d.f.= 6). Predator feeding study. When 12. persimilis were fed I. 1_1gi_ca_e_ from populations initially recovering from a miticide application, significant mortality to the predaceous mites was observed for two of three miticides tested (Figure 3). Tetranychus urticae were collected from rose plants 6 weeks after application of 5.6 ppm abamectin, 150 ppm dienochlor, 360 ppm propargite, or water. When 13. persimilis were fed I. urticae from abamectin and dienochlor treated plants, _E. persimilis mortality was greater than when they were fed I. urticae from untreated plants (T(.05)= 4.27, and T(.05)= 4.38, respectively). When _P_. persimilis were fed I. urticae from propargite treated plants, 13. persimilis survival 23 was the same as when they were fed I. urticae from untreated plants (Figure 3). Long term population study. When dienochlor was applied to potted rose plants it effectively suppressed I. urticae populations for the 15 days of the study (Figure 4). Tetranychus urticae populations at 5, 10 and 15 days after treatment were reduced by 95, 80 and 85%, respectively, compared with populations on control plants (Figure 4). Abamectin applied to 20 potted rose plants infested with I. urticae steadily reduced I. urticae populations from 189 per sample prior to application to 1.8 per sample 4 weeks after application (Table 1). Tetranychus urticae populations slowly increased during the following three weeks to 9.3 per sample at 7 weeks after application. Phytoseiulus persimilis adults released onto the same rose plants at 3 weeks after application of abamectin disappeared to undetectable levels within 2 weeks following release (Table 1). DISCUSSION The results of standard slide-dip and 36 h residue tests showed that dienochlor and propargite had little or no effect on T. urticae or B. persimilis (Figs. 1,2). ' Dienochlor was shown to be toxic to spider mites in a long term experiment (Figure 4). Efficacy testing of dienochlor and propargite in previous studies demonstrate that they are highly effective against I. urticae (Shultz & Coffelt 1987, Horsburgh & Cobb 1988). Apparently, propargite and dienochlor work slowly and require more than 48 hours before mortality is observed in spider mite populations. Abamectin was much more toxic to I. urticae than to B. persimilis in slide- 24 dip and 36 h residue tests (Figs. 1,2). Zhang and Sanderson (1990) had similar results when I. u_rt_i_cae and _P_. persimilis were placed on abamectin treated leaf discs. Our tests with abamectin verified its effectiveness against spider mites but raised the question of a food-chain toxicity problem. Predator mites disappeared two weeks after they were released on abamectin treated plants. In further tests, where spider mites collected from populations recovering from miticide treatments were fed to predator mites, abamectin and dienochlor were toxic to predator mites feeding on spider mites from treated plants, while propargite still showed no effect on B. persimilis (Figure 3). Previous studies have demonstrated that organophosphate compounds remain in resistant spider mites for several weeks in sufficient amounts to kill predators feeding on them (Binns 1971, Hussey & Parr 1965, Hussey 1968). In similar studies Lindquist and Wolgamott (1980) found that acephate at concentrations of 75, 150 and 300 ppm applied as a soil drench were toxic to _13_.persimilis. In that study nearly all P. persimilis feeding on I. urticae on treated plants died 21 days after acephate application. In this study food chain effects of miticides were critical in evaluating their probable impact on mixed populations of spider mites and predaceous mites. The likely impact of a dienochlor or abamectin application is a 4-8 week suppression of I. urticae followed by an outbreak period where the miticide residues are too small to effect I. urticae but are still lethal to B. persimilis feeding on the spider mites. In reviewing the responses of arthropod natural enemies to insecticides, Croft and Brown (1975) cited 13 studies where insecticides were determined to be relatively non-toxic to Amblyseius fallacis. The methods used to determine selectivity were divided approximately equally among slide-dip, residue, and field tests. Little attention was paid to food- 25 chain toxicity; and most of the food-chain studies that were reviewed were for systemic insecticides (Croft & Brown 1975). Some systemic insecticides and acaracides, initially proclaimed as non-toxic to predators, were found to have a food-chain toxicity effect on predators as early as 1967 (McClanahan 1967). Zhang and Sanderson (1990) found only a slight decrease in survival of P. persimilis when predators were fed spider mites "intoxicated"(immobilized) from feeding on abamectin treated leaves. However, In the same study Zhang and Sanderson (1990) found that the reproductive rate of _P_. persimilis was significantly reduced. Grafton- Cardwell and Hoy (1983) showed a similar decrease in fecundity in M. occidentalis when the predators were exposed to leaves treated with 4, 8 and 16 ppm abamectin. The survival of adult M. occidentalis was also greatly reduced at these rates. One difference between our study and that of Zhang and Sanderson (1990) is that in our study the mites fed to predators were those that survived miticide application and were active at the time of our test. These mites were from an actively reproducing population on abamectin treated plants. In the Zhang and Sanderson (1990) study immobile spider mites were collected 24 hours after treatment with abamectin. These differences in test methods may account for the greater mortality of predator mites in our tests compared with Sanderson and Zhang (1990). In non-replicated greenhouse tests, abamectin applied to cucumbers at 7 and 14 day intervals reduced the number of predator mites after the first week, and eliminated them completely after 2 weeks (Heyler & Ledieu 1986). Predator mites may have been eliminated because of a direct effect of abamectin on survival and reproduction of predator mites or a lack of prey as a result of 26 quick elimination of spider mites, or both (Zhang & Sanderson 1990). Further testing is needed before any definite conclusions can be made. A selective miticide is needed for adjusting predator/prey ratios in integrated mite programs (Grafton-Cardwell & Hoy 1983, Hoy & Cave 1985, Zhang & Sanderson 1990). Of the three miticides tested, propargite appears to be the most compatible with P. p_er_sim_ilis. Further testing is needed to determine the impact of averrnectin on predator mites. 27 Résumé On a utilise trois sortes d'essai biologique pour determiner la sélectivité des miticides contre Tetranychus urticae et Phytoseiulus persimilis: 1) le comportement d'alimentation des prédateurs sur la proie, 2) un essai dont des acariens a deux points étaient adheres a une lame de microscope et immergés dans du miticide et 3) un essai dont on a etudié l'effet du miticide résiduel, sur des acariens a deux points, 51 24h apres l'arrosage des feuilles. Utilisant ces deux demiers essais normalisés on a évalué l'efficacité d'abamectin, de proparjite et de dienochlor contre I. urticae et_P_. persimilis. Les résultats de ces deux essais-Ci étaient fallacieux et ne correspondaient pas aux études a long terme de la population. Pour determiner 1a toxicité indirecte des miticides contre B. persimilis on a expose aux prédateurs des acariens a deux points qui venaient d'une population déja traitée par une dose de miticide. Des études du comportement d'alimentation des prédateurs sur la proie et des études a long terme d'une population qui consiste de proie et de prédateur sont les plus utiles pour prévoir l'activité selective des miticides. A partir des résultats de cette experience il serait possible de créer un essai normalise dont les acariens a deux points, traités par une dose sous-létale de pesticide, seraient exposés aux acariens prédateurs afin d'évaluer l'effet, dans la chaine alirnentaire, du miticide sur B. persimilis. 28 ACKNOWLEDGEMENTS Terrance Davis helped analyze the data and prepare the figures. This research was partially supported by Roses, Inc. and the Michigan Agricultural Experiment Station. Thanks to Jan Eschbach for preparing the manuscript and to Kaja Brix for preparing the French translation of the abstract. LITERATURE CITED Anonymous, 1961. Predators versus spider mites. Roses Incorporated Bulletin, July, p. 20-21. Binns, E. S., 1971. The toxicity of some soil-applied systemic insecticides to Aphis gossypii (Homoptera, Aphididae) and Phytoseiulus persimilis (Acarina, Phytoseiidae) on cucumbers. Ann. Appl. Biol. 67: 211-222. Boys, F.E. & P.P. Burbutis, 1972. Influence of Phytoseiulus persimilis on populations of Tetranychus turkestani at the economic threshold of roses. J. Econ. Entomol. 65: 114-117. Brador, L., 1977. Resistance in mites and insects effecting orchard crops. In Pesticide and Insecticide Resistance. eds. D. L. Watson and A. W. A. Brown, Academic Press, New York, 648p. Burgess, E.P.J., 1984. Integrated control of two-spotted mite on glasshouse roses. Proc. NZ. Weed Pest Control Conf. Palmerston North, p. 257-261. Croft, B.A. & A.W.A. Brown, 1975. Response of arthropod natural enemies to insecticides. Ann. Rev. Entomol. 20: 285-336. Field, R. P. & M. A. Hoy, 1984. Biological control of spider mites on greenhouse grown roses. Calif. Agr. March-April, p. 29-32. Field, R. P. & M. A. Hoy, 1986. Evaluation of genetically improved strains of Metaseiulus occidentalis (Nesbitt) (Acarina: Phytoseiidae) for integrated control of spider mites on roses in greenhouses. Hilgardia 54(2): 1-31. Georghiou, P. & R. B. Mellon, 1983. Pesticide resistance in time and space, pp. 1-22. In Pest resistance to pesticides. eds. P. Georghiou and T. Saito, Plenum Press, New York. 29 30 Grafton-Cardwell, E. E. & M. A. Hoy, 1983. Comparative toxicity of avermectin B1 to the predator Metaseiulus occidentalis (Nesbitt) (Acari: Phytoseiidae) and the spider mites Tetranychus urticae Koch and Panonychus u__lmi (Koch) (Acari: Tetranychidae). J. Econ. Entomol. 76. 1216- 1220. Hamlen, R. A. & R. K. Lindquist, 1981. Comparison of two Phytoseiulus species as predator of twospotted spider mites on greenhouse omamentals. Environ. Entomol. 10: 524-527. Helle, W. & van de Vrie, 1975 . Problems with spider mites. Outlook Agric. 8: 119-125. Helyer, N. L. & M. S. Ledieu, 1986. The potential of heptenophos and MK-936 pesticides for control of minor pests in integrated pest control programmes under glass. Agric. Ecosystems Environ. 17: 287-292. Hoy, M. A. & F. E. Cave, 1985. Laboratory evaluation of avermectin for use with Metaseiulus occidentalis (Nesbitt) (Acari: Phytoseiidae). Exp. Appl. Aca. 1: 139-152. Hussey, N. W., 1968. Prospects for integrated control in protected cultivation. Chemy Ind. 498-502. Hussey, N. W. & W. J. Parr, 1965. Glasshouse red spider mite. Rep. Glasshouse Crops Res. Inst. p.78-79. Lindquist, R.K., C. Frost, & M. Wolgamott, 1980. Integrated control of insects and mites in Ohio greenhouse crops. Bull. SROP 3: 119-126. Lindquist, R. K., & M. L. Wolgamott, 1980. Toxicity of acephate to Phytoseiulus persimilis and Tetranychus urticae. Environ. Entomol. 9: 389-392. Martin, NA, 1987. Progress towards integrated pest management for greenhouse crops in New Zealand. Bull. SROP 10: 111-115. McClanahan, RI. 1967. Food-chain toxicity of systemic acaracides to predaceous mites. Nature 215: 1001. 31 McMurtry, J. A., C. B. Huffacker, & M. Van DeVrie, 1970. Ecology of tetranychid mites and their natural enemies: a review. I. Tetranychid enemies, Their biological character and the impact of spray practices. Hilgardia 40: 331-390. Osborne, L.S., Ehler, L.E. & J.R. Nechols, 1985. Biological control of the two spotted spider mite in greenhouses. Univ. Florida Tech. Bull. 853, 40 pp. Overmeer, W. P. J., A. Q. Van Zon, & W. Helle, 1975. The stability of acaricide resistance in spider mite (Tetranychus urticae) populations from rose houses. Ent. exp. and appl. 18: 68-74. Rasmy, A. H. & A. Y. M. Ellaithy, 1988. Introduction of Phytoseiulus persimilis for twospotted spider mite control in greenhouses in qupt (Acari: Phytoseiidae, Tetranychidae). Entomophaga 33(4): 435-438. Simmonds, SP, 1972. Observations on the control of Tetranychus urticae on roses by Phytoseiulus persimilis. Pl. Path. 21: 163-165. Steel, R. G. D. and J. H. Torrie, 1980. Principles and Procedures of Statistics: A Biometrical Approach. McGraw-Hill, Inc. 633p. van Zon, A. Q. & W. P. J. Overmeer, 1975. The occurrence of pesticide resistance in red spider mite populations (Tetranychus urticae Koch), collected from wild plants in the glasshouse district of Aalsmeer, the Netherlands. Z. ang. Ent. 79: 213-222. Wilkinson, L. 1986. Systat: the system for statistics. Systat, Evanston, IL. Zhang, Z.-Q. & J. P. Sanderson, 1990. Relative toxicity of abamectin to the predatory mite Phytoseiulus persimilis (Acari: Phytoseiidae) and twospotted spider mite (Acari: Tetranychidae). J. Econ. Entomol. 83(5): 1783-1790. 32 Table 1. Number of I. urticae and P. persimilis per sample on potted rose plants. Samples were collected approximately every seven days for a 45 day period starting one day before application of abamectin on 23 July. Each sample consisted of 60 leaflets from five rose plants (12 leaflets per plant). Each five plant unit was replicated four times. One thousand Phytoseiulus persimilis were released evenly throughout the 20 plants on 8 August. I. m P_. persimilis Sampling date per sample per sample 22 July ” 189.1 3.: 130.0 0 29 July 27.5 i: 17.9 0 5 August 2.6 i 9.9 0 10 August » R 6.4 i 8.0 1.0 i 0.8 15 August 1.8 i 2.6 0 19 August 3.2 i 5.0 0 25 August 7.3 i 21.1 0 6 September 9.1 i 14.2 0 33 Ionuurv (as) 0.009 0.049 0.224 couc ENTIAW (ppm) 60- a tune» I P.pemmls uonuurv (as) coucamumou (m) MORTALITY PA) Figure 1. Slide-dip evaluation of three miticides commonly used in greenhouse rose production to Tetranychus um and Phytoseiulus persimilis (A = Abamectin, B = dienochlor, C = propargite.) The concentration listed on the U.S.A. federal product label for greenhouse application is denoted by an asterisk (*) above those bars on the graph. 34 MORTALITY (*) D tulcu °°‘ l Rum-I- IORTALITV (in) ccucan‘numu (ppm) uric“ w ‘ l r. painll nonuurv (as) couc autumn (ppm) Figure 2. Mortality of 1m and g persimilis placed on bean leaves treated with abamectin (A), dienochlor (B), or propargite (C) in a 36 h residue test. The concentration listed on the U.S.A. federal product label for greenhouse application is denoted by an asterisk (*) above those bars on the graph. The standard error of each treatment mean is indicated by the brackets on each bar. 35 120— @ 100 - I Treated >_ 80 - Untreated 1'- _l E 60 I: O E °\o 40 - 20‘ Abamectin Dienochlor Propargite TREATMENTS Figure 3. Mortality of Lpersimilis fed spider mites surviving applications of abamectin, dienochlor or propargite compared to the mortality of Lpersimilis fed spider mites from untreated plants. (@ indicates that treated and untreated means are different, ANOVA P < .05). 36 2000 - NUMBER OF TWOSPOTTED SPIDER MITES DAYS AFTER TREATMENT Figure 4. Efficacy of a single treatment of 150 ppm dienochlor in suppressing populations of L urticae 5, 10 and 15 days after treatment as compared with an untreated population. (@ indicates that treated and untreated means are different, ANOVA P < .05). Selectivity of Benomyl, Dodemorph Acetate, Piperalin, Triadimefon, Triforine, Acephate and Endosulfan for Phytoseiulus persimilis or Tetranyghus m Anthony J. D'Angelo Michigan State University East Lansing, MI 48824 37 38 Key words: Acari, Twospotted spider mite, Tetranychus urticae, Phytoseiulus persimilis, fungicides, insecticides, selectivity, testing ABSTRACT Standard slide-dip and short residue tests were used to evaluate activity of the fungicides benomyl, dodemorph acetate, piperalin, triadimefon, and triforine; and the insecticides acephate and endosulfan against twospotted spider mites, Tetranychus urticae Koch and the predaceous mite Phytoseiulus persimilis Athias-Henriot. None of the five fungicides tested were toxic to I. urticae or _P_. persimilis. Both insecticides were more toxic to P. persimilis than to I. m. Therefore, benomyl, dodemorph acetate, piperalin, and triadimefon have potential for use in an integrated mite management system for greenhouse grown roses, while acephate and endosulfan should be avoided because of their toxicity to _13. persimilis. 39 INTRODUCTION Release of the predator mite, Phytoseiulus persimilis Athias-Henriot, has been effective for control of Tetranychus urticae Koch on a number of greenhouse grown vegetable and ornamental plants ( Scopes 1970, Gould 1971, Parr & Scopes 1971, Simmonds 1971, Simmonds 1972). Pesticides used for control of other pests may be toxic to predators used in biological control programs (Croft & Brown 1975, Schulten et a1. 1976, Ledieu & Stacey 1978). Predators may be affected by pesticides through direct contact with sprays, contact with spray residues, or via the food chain (Lindquist & Wolgamott 1980). In greenhouse rose production the major pest problems are twospotted spider mites, Tetranychus urticae, rose aphid, Macrosiphum we (L.) western flower thrips, Frankliniella occidentalis Pergande, and powdery mildew, Sphaerotheca pannosa. In order to determine if a successful biological control system for I. m using P. persimilis can be developed it is necessary to know if the release of predators is compatible with pesticides used for the other major pest problems. This study was initiated to evaluate five fungicides used to control powdery mildew and two insecticides frequently used in aphid control, for compatibility with the use of P. persimilis in integrated control of I. urticae in rose production greenhouses. 40 MATERIALS AND METHODS Toxicity of the fungicides benomyl, dodemorph acetate, piperalin, triadimefon, and triforine; and the insecticides acephate and endosulfan to Tetranychus urticae and Phytoseiulus persimilis was evaluated using a slide-dip test and a 36 hour residual activity test with bean leaves. Spider mites used for these experiments were maintained in the greenhouse and laboratory on 'Henderson' and 'Eastland Bush' lima beans (Phaseolus vulgaris). The culture was initiated in 1986 from spider mites collected at pesticide research center greenhouses, at Michigan State University. Twice every year, additional mites are collected from the same greenhouse range and added to the culture. Phytoseiulus persimilis was originally obtained from "Nature's Control" in Medford, Oregon. The population was maintained under laboratory conditions with periodic additions to the population supplied from N ature's Control. The predators were fed spider mites from our lima bean culture. Slide-dip test. In slide-dip tests with I. m, a 2.5 cm-long piece of double sided Scotch® tape (666 07340-3 3M Commercial Office Supply Division, St. Paul, MN 55133-3053) was placed across a standard microscope slide. A piece of Scotch® nylon packaging tape'(34-7023- 2006-9, 3M Home products Division, St. Paul, MN 55133-3053) was placed on top of the double-sided tape with its sticky side facing upwards. Twenty I. m were placed on each slide. Spider mites on each slide were dipped into test solutions for 5 seconds. Fungicide solutions were prepared at concentrations of 1x, and solutions of insecticides at 41 concentrations of 1x, l/5x, l/25x, 1/125x, and 1/625x, where x is the recommended concentration on the U.S.A. federal product label for greenhouse application. For the fungicides benomyl, dodemorph acetate, piperalin, triadimefon, and triforine the 1x rates were 600, 900, 150, 150, and 269 ppm, respectively. The 1x rates for the insecticides acephate and endosulfan were 600 and 412 ppm respectively. The treatments for the fungicide tests were replicated four times in a randomized complete block design with respect to time of mite placement on slides and treatment. The treatments for the. insecticide tests were replicated twice in the same manner. Adult female mites were collected from spider mite infested bean leaves which were placed for 15 minutes in a C02 filled sealable plastic bag. The mites were tapped off the leaves onto a Tyler equivalent 32 mesh screen (500nm pore size). Unwanted mite stages fell through the screen leaving only adult females. The collected adults were then observed under a microscope where the needed numbers were removed with a fine paint brush and placed on slides. The mites were placed on the tape so that their legs could move freely. After each slide was fitted with mites it was placed in a holding chamber at approximately 95% relative humidity and 27°C. Three hundred ml of each solution was prepared in a 400ml beaker. Triton AG-98 was added to each solution at a concentration of .12%. Slides with mites were agitated in test solutions for 5 seconds, and placed on edge in a drying rack lined with absorbent paper towel to remove excess moisture from the slides. After 15 minutes the slides were placed back into the holding chamber and allowed to sit for 24 hours. Each mite was touched with a fine paint brush. If a mite showed any movement it was considered alive. 42 A similar procedure was followed for the P. persimilis slide-dip tests. However, the nylon ribbed packaging tape did not hold the predators in place. Duct tape (Nashua 357, Watervleit, N .Y. 12189) was stickier and proved to be a suitable replacement. Also, 15 mites were used per slide instead of 20. Treatment rates and all other details were kept the same. 36 hour residue test. Residual activity of benomyl, dodemorph acetate, piperalin, triadimefon, triforine, acephate, and endosulfan were tested by spraying bean leaves at 5 P.M., allowing them to dry overnight, and placing mites on leaves between 9 and 11 AM. the next day. In each of these tests the fungicides were applied at concentrations of 4x, 2x, 1x, 1/2x, 1/4x and 0x. The insecticides were applied at concentrations of 1x, 1/2x, 1/4x, 1/8x, l/l6x and 0x, where x is the recommended concentration on the federal product label for greenhouse application. The 1x rates for all pesticides remained the same as in the slide-dip tests. The treatments were replicated four times in a complete randomized block design. Treatments were applied in a spray tower equipped with a conveyor belt that moves at a constant rate of 9.1 cm/second. Miticide solutions were applied through a TeeJet 8001-E even flat spray nozzle. Excised bean leaves in petri dishes layered with absorbent cotton were sprayed uniformly by running the petri dish through the spray tower. Miticides were applied beginning with the lowest concentration first. Petri dishes were arranged in numerical order on a large cafeteria tray after-treatments were applied. After the spray had dried on the leaves, one rubber hose washer was glued to each leaf using " Elmer's® Glue-all" (Borden's Inc., Dept. CP, Colombus, Ohio 43215). The washers served as arenas to hold the mites in place. Twenty adult female I. urticae and six E. persimilis were used in each arena for their respective tests. Adult female spider 43 mites were collected by the C02 method described in the previous section. A fine paint brush was used to place mites in arenas. A small circle of nylon mesh screening was glued over the arenas after the mites were in place. The cotton base in the petri dishes was kept moist to prevent desiccation of the leaves. In residue tests with B. persimilis, the mites were added to leaves in a 5°C cold room to prevent them from leaving before the screening was glued in place. In all tests the mites were kept in arenas on treated bean leaves for 24 hours. After this time the arena covers were removed and mortality data taken. Statistical analysis. An arcsin square root transformation was performed on all percent mortality data before analysis. Dose effect of the fungicides benomyl, dodemorph acetate, piperalin, triadimefon, and triforine; and the insecticides acephate and endosulfan was analyzed by regressing transformed percent mortality of mites against the concentration of test solution (Wilkinson 1987). If the slope of this regression was greater than zero (p< 0.05), the pesticide was determined to have a dosage effect. In 36 h residue tests, mortality of I. urticae was compared with that of E. persimilis at each of pesticide concentraion with the Student's T- test (Steel & Torrie 1980). Mite mortality was also regressed against pesticide concentration as described for the slide—dip tests. 44 Results Slide-dip tests. The powdery mildew fungicides benomyl, piperalin, and triadimefon had no contact activity against I. urticae or P. persimilis in slide-dip tests (Table 1). Dodemorph acetate was toxic to spider mites but not to predator mites, while triforine was toxic to spider mites and predator mites. Dodemorph acetate treatment caused 29% mortality of I. urticae adults compared with only 9% mortality of mites treated with a water control (T(.OS) = 3.312, df = 6). Dodemorph acetate did not appear to have any effect on B. persimilis adults ( T(.OS) =0.873, df = 6). T riforine had the greatest effect on I. Lime and P. persimilis. Spider mites dipped in triforine suffered 47% mortality compared with 24% in the control, while 46% of the predator mites died compared with 22% in the control (T (.05) = 3.243, df = 6 and T(.OS) = 2.561, df = 6, respectively, Table 1). Regression of I. urticae or B. persimilis mortality against concentrations of benomyl, piperalin and triadimefon solutions indicate no relationship between variables (Table 3). The concentration of dodemorph acetate was related to spider mite mortality but not to _13. persimilis mortality (r2 = 0.67, F = 12.4; r2 = 0.11, F = 0.7, respectively). Also, I. urticae and P. persimilis mortality increased as the concentration of triforine increased (r2 = 0.53, F = 6,9; r2 = 0.57, F = 8.0, respectively, Table 3). The concentrations of acephate and endosulfan solutions were not related to mortality of I. m or _P_. persimilis mortality (Table 4). 36 hour residue test. The powdery mildew fungicides benomyl, piperalin, triadimefon, and triforine had no activity on I. u_r_ti_ca_e_ or 1_’_. persimilis in the residue tests (Table 2). Dodemorph acetate was toxic to 45 spider mites at rates above the 1x rate of 900 ppm (22% mortality compared with 11% mortality in control), but had no effect on spider mites at 1x (11% mortality). Dodemorph acetate was not toxic to P. persimilis. Regression of _P_. persimilis mortality against concentrations of benomyl and dodemorph acetate indicates no relationship between variables. The same regression analysis for I. m indicated a decrease in mortality with an increase in concentration of benomyl (Table 3). The insecticides acetate and endosulfan had greater activity against predator mites than spider mites. Predator mite mortality increased with an increase in the concentration of acephate or endosulfan applied to bean leaves (Table 4). The same regression of spider mite mortality against concentrations of acephate and endosulfan indicates no relationship between variables (Table 4). DISCUSSION None of the five powdery mildew fungicides tested had a major impact on spider mite or predator mite survival in our slide-dip or short residue studies. Triforine and dodemorph acetate caused mortality in slide dip tests but not in residue tests. In the residue tests the activity of dodemorph acetate only affected spider mites when applied at rates above the 1x rate of 900 ppm and therefore cannot be considered useful in helping to control spider mites in an integrated control program. Although these tests suggest that the fungicides tested are not harmful to P. persimilis, a food-chain toxicity test is needed to determine if spider mites feeding on fungicide treated plants have any effect on predators feeding on them (Lindquist & Wolgamott 1980). Field and Hoy (1986) also saw no 46 impact of benomyl, piperalin, and triforine on female twospotted spider mites in residue tests. However, other studies have shown that use of benomyl may decrease the fecundity of twospotted spider mites (Spadafora & Lindquist 1972, Field & Hoy 1986). In one study benomyl applied as a soil drench controlled powdery mildew with no adverse effects to B. persimilis (Parr and Binns 1971). The insecticides acephate and endosulfan were more toxic to B. persimilis than I. urticae in short term residue tests. Therefore, application of acephate or endosulfan may induce outbreaks of spider mites by removing predator mites. Other studies have also shown that acephate (Lindquist et a1. 1979, Lindquist & Wolgamott 1980) and endosulfan (Smith et a1. 1963) are toxic to l_’_. persimilis. In conclusion, the fungicides tested seem to be compatible with use of P. persimilis. Benomyl may have an effect on mite fecundity and should be investigated further in population studies before determining its role in integrated control programs. LITERATURE CITED Croft, B. A. & A. W. A. Brown. 1975. Responses of arthropod natural enemies to insecticides. Annu. Rev. Entomol. 20: 285-335. Field, R. P. & M. A. Hoy. 1986. Evaluation of genetically improved strains of Metaseiulus occidentalis (Nesbitt) (Acarina:Phytoseiidae) for integrated control of spider mites on roses in greenhouses. Hilgardia 54: 1-31. Gould, H. J. 1971. Large-scale trials of an integrated control programme for cucumber pests on commercial nurseries. P1. Path. 20: 149-156. Ledieu, M. S. & D. L. Stacey. 1978. Integrating pesticides and biological control. The Grower, 2 Feb.,1978: 245-247. Lindquist, R. K., C. Frost, & M. L. Wolgamott. 1979. Integrated control of insects and mites on greenhouse crops. Ohio Agr. Res. Dev. Cent. Res. Circ. No. 245. 18pp. Lindquist, R. K., & M. L. Wolgamott. 1980. Toxicity of acephate to Phytoseiulus'persimilis and Tetranychus urticae. Environ. Entomol. 9: 389-392. Parr, W. J. & E. S. Binns. 1971. Integrated control of red spider mite. Rep. Greenhouse Crops Res. Inst. 1970: 119-121. Parr, W. J. & N. E. A. Scopes. 1971. Recent advances in the integrated control of glasshouse pests. A.D.A.S.q. Rev. No. 3: 101-108. Schulten, G. G. M., G. Van de Klashorst, & V. M. Russell. 1976. Resistance of Phytoseiulus persimilis Athias-Henriot (Acari: Phytoseiidae) to same insecticides. A. Angew. Entomol. 80(4): 337-341 Scopes, N. E. A. 1970. Biological control of Chrysanthemum pests in commercial production. Rep. Glasshouse Crops Res. Inst. 1969: 107-108. Simmonds, S. P. 1971. Observations on the possible control of Tetranychus urticae on strawberries by Phytoseiulus persimilis. Pl. Path. 20: 117-119 47 48 Simmonds, S. P. 1972. Observations on the control of Tetranychus urticae on roses by Phytoseiulus persimilis. P1. Path. 21: 163-165. Smith, F. F., T. J. Menneberry, & A. L. Boswell. 1963. The pesticide tolerance of Typhlodromus fallacis (Garman) and Phytoseiulus persimilis A-H with some observations on the predatory efficiency of E. persimilis. J. Econ. Enotomol. 56: 274-278. Spadafora, R. R. & R. K. Lindquist. 1972. Ovicidal action of benomyl on eggs of the twospotted spider mite. J. Econ. Entomol. 65 : 1718- 1720. Wilkinson, L. 1987. SYSTAT: the system for statistics. SYSTAT, Evanston, IL 49 .62. u a .8: 2839 amass: .228 e: s a5 :9: 5806 was 8:28 62295.. E bongo more. B bzmtoE 85 $8065 xtflmm c< .. 5. mo. mo. mo. cm. #4 6360 .3! mo. 3. NF. mm. v pogo: ___E_ a .m NP. mo. NF. mo. mm. v .928 .me. 3. NF. .3... E. v 33on 33H ES 89 ES 83 ES 8: ES 8: 898a ES 89 c octet: coaoEBmE 5:833 cacoEmuon 3:5ch EoEEofi 02295“. jOE “Emmy—0Q CNQE OF: M“ Dmmwmh XQ mhm .m. Pa 3.93 .H 2 8229.2 8% 28558 a bases 8980 ”F 28» 50 mm. ov. mi NF. mm. v xooé 9. mm. 9. 5. mo. v xood C. we. mm. 4.0. t. v xoo; 9. am. 5.. NF. mm. v xomd .30. mm. am. mo. of v med mm. cm. 5. mo. mt v 3:80 29:60: .M 3. mt 9. mm. Po. v 30% no. mm. C. 3. NF. v xood 3. of mf Z. vo. v xooé .3. mm. 3. mo. mo. v xomd om. 9. mo. 3. I. v xmmd 3. 3. E. I. or. v .8200 _ €62.80: 0&3 .H €53 €58 353 353 29QO E53 c octet: coamEEmE 5::an 5908968 380ch E6686; 62295“. stator: 38.8 :38 mm commoaxo 2m Ema .85. 882$ C F 2 3229:: wow: >_coEEoo Co >=>=om _msnfimm ”m 632. .mod n a 2m EmoECQw m_ _oUoE 2.28650. 02:. .. .82: .22. 8:223... 92.. Ucw co_.mE.2mcm.. 80. 6.909535 9.6: 2.2.62.8 mm; 3:95.: 22.8.3 mcozmscm cemmmam. __m .0”. ._. 225.362.2222 Rod 82.2. mt... 22.23 + EN u 2 828. . .m $.62. 84... 2.3.2. 2890 - NS u 2 0:28. 25.2. .802. 28.2 228.2. + me. u .2 2.6-2:; 4.8.2. .392. at... 228.. + mg n 2 2.6-8:... 2.5.2. m8... .8... 29.2. - 48 u 2 $28. :EGB .m. 82.... Sad 48... $5.2. - 4:. u .2 828. Seam .H «82. m5... 92.2. 28:. + mam u 2 2.6-823 : . .m 82. 2.8.2. 52... .89.... + mam u 2 2.6-2:; am... .H 2.3.2. 2.2.02. 32... 2&2. - 5.. u 2 2.28. 2:56... in... Sod ~82. mom... 22.8.2. + 2.5 u .2 0228. .H 1 mm: 03.2. mm: .53. + mam n 2 2.6626 . .m 5 .82. 3.3. wood .88... - 22.. u .2 2.6-8:... 2... _. .H 50... 2.8.0 El. x232. + 2N u 2 $28. _ . .m $3 .48... mm; .68... + mom a .2 0:28. .H «I... 2:3. 32.2. 28:. + mam n .2 2.6-8:... .m 420.2. .392. 84.3 seam... + mam u 2 2.6.8.; .H 48.2. «3.2. 82.2. 38.2. + m8. u 2 2.28. d 2.22. .392. m8... :8... - mom u 2 @228. .H .3... 82.2. 52. 8.3... - 8.. u .2 2.6-2:; .. . .m 2.2.2. 2.8.2. Bail 63.2. - mm... n 2 2.6-8:... v.83: .H a. a u. | .5822 wozmzfim cemwmaom 68 9:2 2:... fig; . .H 2 $2295. 2.2. B 2:228 .323. 2:... 85:8 .2 8:236 cofiwoaom 2Eq3co.mE_qu 3.223538%“ 282.3998... 2308662. 28213380ch 0229.3”. 5:: .. . .m. ”m 233. 52 .mod n n. 8 EmozEgm m_ _muoE c2399.... 0.: .. door .3 66:92:22. :9: 96 5sz.239. woo. 053092.62... 9.6: 2.9.62.8 was 259.2: 22.8.3 mcozmacm 2.06350. __m .0”. + mm: .890 Ema. .622... + 8. u .2 $28. , of... 2.3.0 $4.2” 29.3 + 84 u 2 328. H 08.2. «32. 2.0.0 .682. + 48 u .2 2.6-2.27. .m Sod :2... 9.2. 33.2. + KN n 2 2.6-3:... .H 2333.385 micro .ooo.o «mm. a .685 + So u .2 328. d mood 32.2. 8:. 5.3.2. - mom u .2 26.8. .H $2 $2 $2 $2 2.6-82; . . . .m 82.2. 2.9.2. 2.8.2. .88... + man u .2 2.6.2.2; mama... .H 252.38.22.86. :80: .._. 8:285 cemwmamm _ 5...; oEozoowE ___E_ . .m. new idulul: _. 2 cmzamooco new 3288 .0 3328 _szmm. Ucm 89:8 .2 8228?. 2.93663. x. 03w... Estimating Population Densities of Twospotted Spider Mites, Tgtranyghgs rtic e, on Greenhouse Grown Roses. Anthony J. D'Angelo Department of Entomology Michigan State University East Lansing, Michigan 48824 53 54 Key Words: Acarina, Twospotted spider mite, Tetranychus urticae, sampling, bias, accuracy, precision, roses ABSTRACT Absolute population densities and distributions of spider mites on 14 rose plants were determined by counting the number of mites on every leaflet (450 -900 leaflets per plant) on each plant. Leaf, leaflet and stem positions were also noted and entered into a data file. The number of mites found on leaves 1 and 2, the oldest leaves, was 1-8 fold greater than the number of mites found on randomly selected leaves. The actual mean number of mites per leaflet was most closely estimated by the means of leaflets 3 and 4. Eight replications of simulated samples of 10, 20, and 30 randomly selected leaflets were taken from each plant. This process was repeated using only leaflets 3 and 4 from leaves 1 and 2. Because samples from leaves 1 and 2 are consistently higher than random samples from the entire plant, the bias was determined and used for a sample adjustment computation that more accurately reflected the actual mite density. The precision and accuracy of all sample methods were determined. When mite populations are low (< 2.5 mites per leaflet) samples of leaflets 3 and 4 from leaves 1 and 2, adjusted for known bias (multiplied by conversion factor 0.548), and random samples from the entire plant are equally accurate and precise. When mite populations exceed 2.5 mites per leaflet random samples from the entire plant are more accurate but not more precise than leaflet samples from leaves 1 and 2 only. For research purposes, random samples of 20 leaflets per plant will provide an acceptable estimate of spider mite population density (average s2/‘x_ = 0.79). Because spider mites are most heavily distributed among the oldest leaves, sampling schemes designed to detect mites at low density should be aimed at the base of rose plants where the oldest leaves are found. 55 INTRODUCTION Frequent outbreaks of twospotted spider mite, Tetranychus urticae Koch, on roses and strict guidelines proposed by the United States Environmental Protection Agency for re-entry of greenhouses after pesticide applications has led to increased interest in developing an integrated pest management (IPM) program using the predator mite Phytoseiulus persimilis Athias-Henriot. Biocontrol of spider mites requires more intensive monitoring of mite populations than chemical control. A quantitative approach to biological control of spider mites using predaceous mites requires estimating population densities (Nachman 1984). To be useful, a sampling program should provide estimates having the highest accuracy commensurate with the amount of work expended (Southwood 1978). Accurate sampling methods are needed for researchers studying biocontrol of mites on roses, and for greenhouse growers trying to detect spider mites at low densities to properly time release of predator mites. Intelligent decisions concerning pest control strategies are directly dependent upon knowing the status of all important species (Beardsley et al. 19791 ‘ The purpose of this project was to determine the distribution of spider mites on greenhouse grown rose plants, and to use this information to develop sampling procedures with known precision and accuracy limitations that can be adapted by other researchers and rose growers. 56 MATERIALS AND METHODS Rose plants of Livia and Darling varieties were received in good condition from Devor Nurseries on 26 June, 1987. Two hundred roses were planted in 8 liter plastic nursery pots filled with Baccto® Professional Plant Mix potting soil. Rose plants were maintained with biweekly irrigation and weekly fertilization. An average of 300 g of 20—10-20 of Peter‘s® professional fertilizer (W. R. Grace and co., Fogelsville, PA 18061) applied as a soil drench using a 10-1 proportioner and a 5 gallon stock tank. Mite cultures were established in 1986 from mites collected from the Pesticide Research Center greenhouses at Michigan State University. Once per year additional twospotted spider mites are collected at the same place and added to the culture. Twospotted spider mites were maintained in the laboratory and in the greenhouse on 'Henderson Bush' and 'Eastland Bush' lima beans (Phaseolus vulgaris). Mites were allowed to infest the rose plants naturally by movement from other parts of the greenhouse. This prevented any bias in mite positioning on the plants that might occur from hand infesting of plants. In order to develop and determine the most efficient sampling method for twospotted spider mites on greenhouse grown roses it was decided to determine the actual population and distribution of spider mites on selected rose plants, enter the data into a file and simulate sampling procedures with a computer. When the potted rose plants reached a size of approximately 75 to 95 cm, individual plants were chosen for sampling by 57 visual inspection for mite injury. Selected plants had a range of 450 to 900 leaflets, supporting from 50 to 24,000 mites. Spider mite numbers on every leaflet were recorded for fourteen rose plants. Each plant was counted within a 3-day time period to avoid significant shifts in mite distribution. Numbering System: A five digit identification number was assigned to each leaflet based on its position on the plant. Digits 1 to 5 of the identification number represent the stem, stemlet, branch, leaf or leaflet, respectively (Figure 1). Leaves can arise from either the stern, stemlet or branches. Any plant part not present was assigned a 0. When leaflets were missing the lowest leaflet present was labeled as #1. Spider Mite Counts. Counts were made starting with the lowest leaf on stem #1. This pattern was repeated for all stems. Each leaf was taken from the plant for counting. Leaflets were assigned the appropriate identification number and all motile mites were counted using a binocular dissecting scope. Data were recorded as the number of motile mites per leaflet. Eggs were not counted. A data file for all sampled plants was created in Systat® (Wilkinson 1987) to allow computer sampling. Sampling. Data collected were used to determine: (1) total number of mites/plant; (2) mean number of mites/leaf; (3) mean number of mites/leaflet; and (4) distribution of mites on the plant. For each plant, eight simulated random samples of 10, 20, and 30 leaflets were taken from the entire plant and from leaves 1 and 2, leaflets 3 and 4 only. Using Systat®, random numbers were generated and assigned to each leaflet. For each plant the random numbers were sorted and arranged from lowest to highest numerical value. Groups of 10, 20 and 30 random leaflets were pulled from the data file for statistical analysis in Systat® (Wilkinson 58 1987). The means of all samples were calculated and compared to the actual plant means for their accuracy and precision. Data Analysis. Accuracy of a sample estimate is the difference between the sample estimate and the true population parameter being estimated (Fowler & Witter 1982). Therefore, in our study accurancy can be defined as a measure of the deviation of the sample mean (x) from the true population mean (u). Precision of a sample is the difference between the sample estimate and the mean of the estimates of all possible samples that can be taken from the population (Fowler & Witter 1982). Therefore it can be said that precision is a measure of the variance among the sample means. In order to determine the accuracy and precision of our sampling method the variance of sample means (precision) and the accuracy of random leaflet samples were calculated for all fourteen plants. The mean accuracy was calculated using the formula: Accuracy: 2',(x - 1;) where; n = 8 11 Accuracy was calculated in this manner for all sampling variations. Because the sample means were consistently higher for leaves 1 and 2 than the actual mean, a conversion factor was calculated to adjust samples from leave 1 and 2 to more accurately reflect actual mite density. First, the percent accuracy for each plant was obtained by the formula: % Accuracy = ccuracy u The mean percent accuracy for all fourteen plants was used to calculate the conversion factor by the formula: Conversion factor =1.0 / (1.0 + mean % Accuracy) 59 By this method an average conversion factor of 0.548 was determined for all 14 plants. Sample means were multiplied by this value for each plant and the accuracy, variance and sample means were calculated with the adjusted values. For research purposes, random samples of 20 leaflets per plant will provide an acceptable estimate of spider mite population density (average SZ/T = 0.79, Figure 6). All other data were analyzed with Systat® and Mystat® statistical programs from Macintosh® (Wilkinson 1987). Regression of twospotted spider mites distribution against leaf position was performed for high, medium and low populations on greenhouse grown roses. RESULTS AND DISCUSSION When the mean number of mites per leaf for high (5.0 -36.1), medium (1.5 - 5.0), and low (0 - 1.5 mites per leaflet) populations were regressed against leaf position, the slopes of the regression lines were 0.045x, -0.52x and -1.34x, respectively (Figure 2). Leaf positions 1 and 2, the oldest on any given stem, contained the highest density of mites (Figure 1B). Mite density decreased as leaf position increased. On a potted rose plant leaf positions 1 and 2 are mostly located in the center of the plant closest to the ground. The actual mean number of mites per leaflet were compared with the means for various leaflet sample patterns. In most cases, the actual mean was most closely estimated by the mean of leaflets 3 and 4 (Table 1). Therefore, it was theorized that the most sensitive estimate of mite populations on experimental rose plants could be obtained by sampling only leaves 1 and 2, and counting the live mites on leaflets 3 and 4. In 60 order to test this assumption randomly selected samples of 10, 20, and 30 leaflets were taken from each rose plant using the data file created from the mite population counts. Two sets of samples were taken. One set from leaves 1 and 2 with only mites on leaflets 3 and 4 being counted. The other set of samples were taken randomly from the entire plant. Samples from leaves 1 and 2, leaflets 3 and 4 only, were more precise and less accurate than random samples taken from the entire plant (Table 2; Figure 3). Samples from only leaves 1 and 2 generally gave estimates that were too high (Table 3; Figure 5). When the data were adjusted by multipling the sample means by the conversion factor 0.548, the accuracy of the estimates were increased (Table 4; Figure 4). When mite populations were below 2.5 mites per leaflet random samples from all leaflets and samples from leaflets 3 and 4 on leaves 1 and 2 were equally accurate and precise. When mite populations exceed 2.5 mites per leaflet random samples taken from the entire plant were found to be more accurate and less precise than the adjusted leaflet samples from leaves 1 and 2. In summary, the most accurate but not the most precise estimate of mite populations are determined from samples taken from the entire plant. Samples taken from areas of highest mite concentrations only consistently give estimates that are too high. When this data is adjusted using a known accuracy conversion factor the accuracy and precision of the estimate is increased. The adjusted data were the most precise of the three samples. An important step in biological control programs is the early detection of pest populations followed by release of the control organism (French et a1 1976, Hamlen & Lindquist 1981). Selective sampling of leaflets 3 and 4 from leaves 1 and 2 is a good method for detecting low 61 populations of spider mites on potted rose plants. Further research is needed to confirm that spider mite populations on the larger rose plants grown in production greenhouses have the same spatial distribution as we found for mites on potted rose plants. LITERATURE CITED Beardsley, J. W., M. T. AliNiazee, & T. F. Watson. 1979. Sampling and monitoring, pp. 11-22. In Biological Control and Insect Pest Management. eds. D. W. Davis, S. C. Hoyt, J. A. McMurtry, and M. T. AliNiazee, Division of Agric. Sc., Univ. of Calif. Fowler, G. W. & J. A. Witter. 1982. Accuracy and precision of insect density and impact estimates. Great Lakes Entomol. 15: 103-117. French, N ., W. J. Parr, H. J. Gould, J. J. Williams, & S. P. Simmonds. 1976. Development of biological methods for the control of Tetranychus urticae on tomatoes using Phytoseiulus persimilis. Ann. appl. Biol. 83: 177-189. Hamlen, R. A. & R. K. Lindquist. 1981. Comparison of two Phytoseiulus species as predator of twospotted spider mites on greenhouse omamentals. Environ. Entomol. 10: 524-527. N achman, G. 1984. Estimates of mean population density and spatial distribution of Tetranychus urticae (Acarina: Tetranychidae) and Phytoseiulus persimilis (Acarina: Phytoseiidae) based upon the proportion of empty sampling units. Journal of Applied Ecology. 21: 903-913. Southwood, T. R. E. 1978. Ecological Methods. Chapman and Hall, New York, N. Y. Wilkinson, L. 1987. SYSTAT: the system for statistics. SYSTAT, Evanston, IL 62 63 Table 1. Spider mites per leaflet for computer generated samples of selected leaflet groups compared to spider mites per leaflet for the entire lant. Spider mites per leaflet Rose All Leaflets Leaflets Leaflets Plant Leaflets 1 & 2 3 & 4 5 - 9 A 0.102 0.046 0.067* 0.241 8 0.244 0.167 0.321 0254* C 0.341 0.188 0.301* 0.795 D 2.012 1.780 2.328 1.936* E 2.638 1.489 2503* 4.093 F 2.677 2.131 2613* 3.580 G 2.725 2.455 2.969 2905* H 3.162 3.176* 3.830 2.439 I 6.265 3.558 6.480* 9.523 J 6.368 5.051 8.174 5946* K 8.395 5.442 8.732* 11.868 L 11.513 8.138 13.141* 15.608 M 14.822 8.724 15.594* 24.008 N 36.105 29.428 40.955* 42.200 * Selected leaflet samples with means closest to the mean number of spider mites per leaflet for entire plant are indicated by an asterisk. 64 .mEmEzmm some :36 .o 82.9.? new 2023092.. .2 omém ofim omom ooov ooo- oo.mo oft. om.m- ooom Zoo 2 mNN moo- oi: omNF mod 8.3 3.8 R; 3.3 No.3 .2 8.: moo- moor mod omé- 8.3 no.3 tum our: 5.: ._ omo :o- omo ovo woo woo ooom ooé ooow ovo x moor woo- mod and owo moo ono moo- omo moo o and So who 8.3 moo omo ooom :4 woo 3o _ oo; mfio Foo 3.2 woo. woo No.4 ooo ovum ofm I om; moo mud on; mfo om.w ow; to- co; and 0 mod ooo- mod mo; mfo oo.m 2.2.... ofio mod oo.m u. mo; «No- Ned No.2. woo woo mod 5o- mo.m vod m or; vmo mm.m no; two. vm; moo who- ow; rod 0 omoo oofo- oomo oooo onto oFmo omfio oooo Fovo End 0 mnoo nmoo- two mmfo oo_..o ommo wwfio owoo momo ego m mooo oroo omfio oroo eooo oofio oooo mmoo- omoo 3. Noto < 25 25 . 25 .8sz 5:282 .282 Em> fioo< .8 93:2 +..m> food. .62. $52 +..m> .469... .ma 8:: 23 , 26:82 EmE mmEEmw 8:83 on mmaEmm 8:86.. om 838% 8.83 or .8 $52 $2. 2.2.82 Om .o .om .3 30 838$. 88% 388.8% 5:82 .522 3:8 28 .388 65 8 CC 888.8 :82 .3388 828% 88.2 8882828 83 A3 888 :89. BE. .8889. 28 2588 8.3 memo. 80¢ :83 Ba 32.2.2 Om can .om .2: E 8388. .8 388.9% 803 888:3 838 we 882.2; o8. 2.8228288 22288. Ewmo n. 2. 80:? 53. M VNV 5898.22 .8388 Home“: 8288 28 888228 838 Mo 88.8..» 88 .8858“ 2:. .m 03mg. 65 .m2mEz8 88:. £06 3 88:8 :8. 88:02:. ,P 3.3 3.8 mmdm mmdm 8.3 when 3.8 m :2 omém= 3.8 2 5.: Ed 3.8 m2: :5: mmdm stm 88 vodw 8.3 .2 88 m 8 8.8 88 :3 8.8 8.2 mg. 8.8 __ E. I ._ an; 3: mmd 3.3 mmé no.3 mi: 8; 8.3 3.: x 3.: mm...“ 8.: mo; no; ~53 modr 3.: 8.3 3.0 _. .38 mm; 8.5 om.m om; mfim 3.? mad mfim 8.: _ 58 mm...» mmk and mm.m mad no.“ afin mad :2.” I and and mod 3.: mud mad 3.? 5.: 88 mud 0 we; Pod- 88 0:6 mud mad 3.x. and mud mod n. 2.: mud and mfg mvd mfim mmé mud 88 88 m mod mm; $8 um; 8.: mud 2: mod 88 58 a mad mnmd mid mfio vmvd mmmd end 88: mmmd i 2%.: o vmd «8.0 05.: mod 3?: 50.: mod 38.: mmmd _ 38.0 m mood Pmod- Bod mood mmod- wood 3.: Rod- mnod Nerd < E E A5 8.82 8:8. $.82 88> hoo< 5a 822 +.._m> +.8< .8 92:2 +8> +86. 5: 852 :3 . 8:8. 82: 838mm 8:84 on 838mm 8:84 cm 8383 8:84 or .5: 8:5. 80m 82.82 cm 8 .om .3 .8 8388 88: 888.88: “282 8: 828 .8 898:: 2: 2 CS 888:8 :82 .8388 2288 8:88 8:888: 83 A3 :88 2888 BE. .8888: 28 82:; 8:5 88: 8:8 82: 8: 3282 cm :8 .ON .02 Ho 8388 :8 88888: 825 2888 :88 “o 8:28; :8 A.m:o:8::8 2:88 EwB u : 88>» 83-56 8880< .N :8 2 882 88: v :8 m 32.82 :0 82:88 80¢ 888:8 :88 we 8:88? :8 8888 2: .m 28,—. 66 000000000 0000. £90 05 00 00:0:0> 0:0 50500.0. n 0, 3.9 2.0- 00.00 00.: 00.0- 00.00 00.00 00.0- 00.00 :00 2 00.0 00.7 00.0? mvh 00.7 00.0; 00.: aim- 00.0.? 00.3 _2 00.0 3.0- 00.: 0?; 00.0- 00.: 00.3 00.0- 00.: 2.0.: ._ 0m.0 00.0- 00.0 5.0 00.0- 00.0 00.? 00.0- 00.0 0.0.0 v. 00.0 00.0- 00.0 00.0 00.0- 00.0 05m. 000. 00.0 00.0 0 0F; 00.? 00.0 on; 00.7 00.0 00.0 mmé- 00.0 00.0 _ 05.0 00.0 070 :0 00.0 00.0 04.0 00.0 5.0 00.0 I 00.0 00.0 00.0 mm; 00.0 00.0 00.0 00.0 00.0 00.0 0 5.0 00%. 00.0 00.0 00...- F0; 00.0 00.0- 00; 00.0 n. 00.0 00.0- 00.? 00.0 00.0- mm; 004 00.7 00.0 00.0 m 00.0 00.0- mm; 3.0 00.0- 00.? 3.0 00.0- 3: 00.0 o 05.0 000.0 000.0 000.0 _,:.0 000.0 00.20 000.0 000.0 30.0 0 P30 000.0 000.0 05.0 03.0 000.0 000.0 000.0 000.0 3&0 m 50.0 $0.0- 0.00.0 000.0 000.0- 000.0 000.0 50.0- 30.0 .. 03.0 < E E E 00:00. 00:00. 00:02 30> +00< .00 00:2 30> +00< .00 00:2 30> +000. 000 00:5. :0 , 00:00. 80.0 00.080m 00:00.. 00 00_0E0w 00:00.. 00 003003 00:00.. 0— .00 00:5. 000m 0000002 0m 00 .om A: ¢0 003800 80¢ 008800000 00:02 000 0008 ¢0 000800 08 00 AM 0 00008000 0002 003800 0020000 80¢ 000800000 003 A3 0008 00800 00.0 00000000 008 000000, 003 00000 80¢ 0008 000 000E000 Gm 000 .00. .00 ¢0 003800 00¢ 008000000 0003 0008000 0008 ¢0 0000¢0> 000 00000002000 03800 0030 u 0 00003 05.70.05 0000000< 000008200 00000000 00800 wvmd 83¢ 000000800 08 3 00:00:00 N 000 fl 000,000 80¢ 0 000 m 32.002 ¢0 003800 80¢ 00008000 0008 ¢0 0000¢0> 000 00000000 000. .0 030,—. I” B Figure 1. Rose plant numbering system. (A) Generalized plant showing the numbers assigned to stems and branches. (B) The numbering system used to identify leaf and leaflet positions. 68 5:" m A _1 Y = 16.690 - 1.335X El 40 " MSE=10.184 a d F 3 52.815 L013] 30 _I R‘Z = .56 t o E 0 0 IL 20 '4 O 6 z z 4 LU 2 LEAF POSITION 12 "' 0 . B I.“ .1 1o _. 0 Y =- 5.340 - 0.521X E . MSE = i 0.091 a. 8 _ ° F = 33.04 3 R02 8 .44 .— o s 6 - 3 . . IL 0 4 _. o g . 5 2" i m 0 5 0 v v ' I ll 0 2 4 6 8 1 O 1 2 LEAF POSITION 08 ‘ o . ‘ C 0 Y a 0.434 - 0.045X 0.6 " MSE :- i 0.009 ‘ F a 23.33 R02 8 .43 1M. MEAN NO. OF MITES PER LEAF LEAF POSITION Figure 2. Mean number of spider mites per leaf in relation to leaf position. (A) High populations (plants with 4000 - 7000 mites per plant); (B) medium populations (plants with 1000 - 3000 mites per plant); (C) low populations (plants with less than 1000 mites per plant). 69 12- g A : 10 ‘ D - Census Data I“ + a Sample Data .1 1 ll: 8 p E . 3 t 6 ‘ E + u. + 1 I O 4 " + d + + i Z + z 2 " g 2 g < + + m + + + 5 a as E 3 0 l g l l l r r A B C D E F G ROSEPLANT 80- l- B m .1 E2 5’ 60 - u - Census Data + a Sample Data a: m a. 4- $ + t 40- a 2 * ‘5' 1 ' + O .. z 20 < II in a 0 j E a l l l I I I r H l J K L M N ROSE PLANT Figure 3. Actual (from absolute samples) and sample estimates (eight random samples of twenty leaflets/plant) of mean mites per leaflet for rose plants with different population densities. (A) Plant with lowest mean number of mites per leaflet; (B) plants with highest mean number of mites per leaflet. 70 12- l- . A In : 10 a u - Census Data _, + - Sample Data 5 8 - a (D E + u. o 4 -I I o' z ‘ a 9 ‘2: 2 - E i i I.” 4. q- : a i + + 0 [it T I T I I I A B C D E l: G ROSEPLANT 80- E .3 :- U - Census Data 3 60 _ + . Sample Data I: In a. a) g 40 - 2 i LL 0 g 20 - + b 2 < a 1 L“ l'- = E i 0 I l T l T l l I J K L M Pl ROSE PLANT Figure 4. Actual (from absolute samples) and sample estimates (eight random samples of twenty leaflets/plant) of mean mites per leaflet rose plants. Samples of leaflets 3 and 4 from leaves 1 and 2. Sample means multiplied by conversion factor. (A) Plant with lowest mean number of mites per leaflet; (B) plants with highest mean number of mites per leaflet. 71 12- I- . A #- "J + -| + l; 10 - D = Census Data I_l‘l + a Sample Data + E .- e 8 + m + E 4. IL 3: * + o 4 - 4: 0' ll! t a Z :I + I: Z 2 ‘ U + 3 i i + + 2 . 0 rIl| I g I I I I A B C D E F G ROSE PLANT 80 - l- B In _I “' + E, H - Census Data + .1 6° ‘ + -- Sample Data 3: E + a. 1- fl I- 40 ‘ + 2 a ll. 0 t= g 20 - i E z + i I: < 3 Lu 2 3 5 * 0 r I I I I I I H I J K L M N ROSE PLANT Figure 5. Actual (from absolute samples) and sample estimates (eight random samples of twenty leaflets/plant) of mean mites per leaflet for rose plants with different population densities. Samples of leaflets 3 and 4 from leaves 1 and 2. (A) Plant with lowest mean number of mites per leaflet; (B) plants with highest mean number of mites per leaflet. 72 5 - .. ‘I' z 4 - < I.” 5 .. ILI a‘ 2 3 ' < a) + m + I.“ a 2- + m + 3 l: + s + I + a: 1 _ § " + + * + t i + 1‘ A o . , r . r o- 1 o 2 o 3 o LEAFLETS PER SAMPLE Figure 6. Average variance/sample mean for rose plants with different population densities. Eight random samples of 10, 20 and 30 leaflets per plant. SUMMARY AND CONCLUSION It has been suggested that with the use of a short-lived selective miticide B. persimilis may have potential value in an integrated pest management program on commercially grown greenhouse roses. The probability of successful initiation of a spider mite biocontrol program could be increased by the use of selective pesticides to suppress spider mites and other pest populations. Standard slide-dip and 36 h residue tests were used to evaluate activity of the miticides abamectin, propargite, and dienochlor: the fungicides benomyl, dodemorph acetate, piperalin, triadimefon, and triforine; and the insecticides acephate and endosulfan against twospotted spider mites, Tetranychus urticae and the predaceous mite Phytoseiulus persimilis. Results showed that abamectin was more toxic to I. urticae than B. persimilis, while dienochlor and propargite had little or no effect on either mite. Dienochlor was shown to be toxic to spider mites in a long term experiment. In the predator feeding study, spider mites from populations initially recovering from miticide applications were fed to predator mites. This test was found to be useful in determining indirect toxicity of miticides to E. persimilis, None of the five fungicides tested were toxic to T . m or B. persimilis. Both insecticides were more toxic to _13. persimilis than to I. were. Therefore, benomyl, dodemorph acetate, piperalin, and triadimefon have potential for use in an integrated mite management system for greenhouse grown roses. Predator feeding studies need to be performed before recommendations for use can be made. Acephate and endosulfan 73 74 should be avoided because of their toxicity to B. persimilis. A number of biological control and cultural control methods need to be looked at as control of aphids and thrips. Biocontrol of spider mites requires more intensive monitoring of mite populations than chemical control. A quantitative approach to biological control of spider mites using predaceous mites requires estimating population densities. Our findings show that spider mites are most heavily distributed among the oldest leaves therefore, sampling schemes designed to detect mites at low density should be aimed at the base of rose plants where the oldest leaves are found. For research purposes, random samples of 20 leaflets per plant will provide an acceptable estimate of spider mite population density. HICHIGQN STATE UNIV III II II IIIIIIIIII _ :fj" A .. ... |.. C ‘L v- ‘ ‘ ' ‘u - -..'. ,_ -- w... ; \ .,. _ . .. ... ~ __ _ .. , h .. _ ”_m' ~- a ., .. 7,. . . ~ t, . . a ‘ “‘1", 2‘ . .--.... .. ‘~I-'............ ."‘ .WH_‘....:..I. " ' ;~-_-.r.m' . ........‘.' .. . I'I-v y- ”a...” ~-~"v.u:..:.. “A...” ,4 ,