INFORMATION TO USERS This reproduction was made from a copy of a document sent to us for microfilming. While the most advanced technology has been used to photograph and reproduce this document, the quality of the reproduction is heavily dependent upon the quality of the material submitted. The following explanation of techniques is provided to help clarify markings or notations which may appear on this reproduction. 1. The sign or “target” for pages apparently lacking from the document photographed is “Missing Page(s)”. If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting through an image and duplicating adjacent pages to assure complete continuity. 2. When an image on the film is obliterated with a round black mark, it is an indication of either blurred copy because of movement during exposure, duplicate copy, or copyrighted materials that should not have been filmed. For blurred pages, a good image of the page can be found in the adjacent frame. If copyrighted materials were deleted, a target note will appear listing the pages in the adjacent frame. 3. When a map, drawing or chart, etc., is part of the material being photographed, a definite method of “sectioning” the material has been followed. It is customary to begin filming at the upper left hand comer of a large sheet and to continue from left to right in equal sections with small overlaps. If necessary, sectioning is continued again—beginning below the first row and continuing on until complete. 4. For illustrations that cannot be satisfactorily reproduced by xerographic means, photographic prints can be purchased at additional cost and inserted into your xerographic copy. These prints are available upon request from the Dissertations Customer Services Department. 5. Some pages in any document may have indistinct print. In all cases the best available copy has been filmed. University M icrofilms International 300 N .Zeeb Road Ann Arbor, Ml 48106 8308997 Quartey, Solomon Quatekwei POPULATION DYNAMICS OF THE ONION THRIPS, THRIPS TABACI LIND., ON ONIONS Michigan Suae University University Microfilms International PHD. 1982 300 N. Zeeb Road, Ann Arbor, MI 48106 POPULATION DYNAMICS OF THE ONION THRIPS, THRIPS TABACI LIND.f ON ONIONS By Solomon Quatekwei Quartey A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Entomology 1982 ABSTRACT POPULATION DYNAMICS OP THE ONION THRIPS, THRIPS TABACI ON ONION By Solomon Quatekwei Quartey Onion thrips, Thrips tabaci Lindeman, a cosmopolitan species, is a major pest of onions in Michigan. The objec­ tives were to obtain more detailed information on the ecology and biology of the thrips in a near pristine environment, identify and understand the factors affecting field popula­ tion densities and to understand the pest-crop interaction and the role of the onion thrips in the onion agro-ecosystem. Adult thrips colonized the onions when the latter were 3-5 weeks old. The thrips distribution was random at the beginning of the season, but became clumped later. The population rose exponentially and peaked at or soon before harvest. Infestations reached as high as 121 thrips/plant. Adults and second instar thrips fed on onion leaf tissue at the rate of 4.93 and 4.51 mm pectively. 2 x thrips -1 x day -1 , res­ The lower temperature developmental threshold was established as 7.4°C. At high densities, thrips killed young onion plants, but no loss in bulb weight occurred if feeding occurred in the late season. Heavy rainfall was a major mortality factor. ellids were the most abundant natural enemies. Coccin- Although the coccinellid Coleomegilla maculata DeGeer consumed 3-400 thrips per day in the laboratory, the temporal and spatial Solomon Quatekwei Quartey occurrence of this predator and the thrips were not properly synchronized in the field. Adult C. maculata density peaked in early season and before the prey. The coccinellids stayed in grasses and cereals and showed a bimodal diel flight into the onion field to feed. By manipulating the environment, this predator can be made more effective. The results of this study provide pertinent information for developing pest management strategies for onion thrips in the onion agro-ecosystem. To my parents, Emmanuel and Mary, my wife, Ama, and children, Nii Kwate, Naa Kwale and Nana Ama Aboagyawa Kwakor. ACKNOWLEDGMENTS I would like to extend my deepest appreciation to all the people who contributed in various ways to the completion of my Ph.D program. Special thanks go to Dr. Ed Grafius, my Major Professor and Chairman of my Guidance Committee, whose friendship, direction and support at very crucial times made this work possible. I wish to express my appre­ ciation to the members of my Guidance Committee - Dr. James E. Bath, Dr. James R. Miller and Dr. Thomas C. Edens, for their contribution in forging the direction of my program. Special thanks should go to Dr. James E, Bath, the Chairman of the Department, for providing excellent facil­ ities and a congenial atmosphere for pursuing graduate work. My thanks go to Sally Phelps, Wendi Greer and Sherry Cooper for their assistance in collecting the data and also to Dr. S. G. Wellso for placing the facilities of his green­ house and equipment at my disposal. It was rewarding associating with the following persons in the "Ed Grafius Group" - Liz Morrow, Mark Otto, Bob Collins, Resham Thapa, Dave Prokrym and Fred Warner. The author is greatly indebted to the U.S. Agency of International Development, and the University of Cape Coast, Ghana, for their financial support that made my studies here possible. iii Lastly, but in no way the least, I wish to extend my warmest thanks to my wife, Ama for her emotional support throughout this program and my children Nii Kwate, Naa Kwale and Nana Ama Aboagyewa Kwakor, for enduring a sometimes trying yet rewarding experience with me. iv TABLE OP CONTENTS List of Tables.............. ........... . List of Figures................ ............................ 1. Introduction.......................... 1 2. Literature Review....................................... 3 3. 2.1 Economic Impact of Onion Thrips................. 3 2.2 Sex Ratios and Reproductive Biology............. 5 2.3 Developmental Biology................. 10 2.4 Natural Enemies.......... 13 2.5 Population Dynamics.............. 14 Methods and Materials.............................. 3.1 3.2 3.3 18 Sampling Methods for Onion Thrips............... 18 3.1.1 Passive or Liquid Extraction.............. 18 3.1.2 Dynamic Extraction................... 19 3.1.3 Knockdown Method........................... 20 3.1.4 Direct Counting............................ 20 Preliminary Field Survey............. 21 3.2.1 Occurrence and Importance of Onion Thrips. 21 3.2.2 Natural Enemies Survey.................... Population Dynamics of Onion Thrips..... 3.3.1 24 24 Field Data, 1978........................... 25 v 3.4 3.5 3.3.2 Field Data, 1979.................. 26 3.3.3 Field Data, 1980................. 28 Developmental Biology.................... 3.4.1 Design of Rearing Cells for Onion Thrips.. 31 3.4.2 Developmental Biology of Onion Thrips 35 3.4.3 Developmental Biology of Coleomegilla maculata........ ........... *............. 40 Population Dynamics of the Coccinellid Complex... 40 3.5.1 Coccinellid Population Estimates, 1979.... 40 3.5.2 Coccinellid Population Estimates, 1980.... 45 3.5.2.1 Visual Count on Ragweed......... 45 3.5.2.2 Quadrat Sampling................. 46 3.5.2.3 Stickyboard Traps................ 46 3.5.3 3.5.4 3.5.5 3.6 4. 31 Arthropod Complex Associated with Coccinellids.............................. 47 Diel Activity Patterns of Coccinellids.... 47 3.5.4.1 Visual Counts on Onion Plants.... 47 3.5.4.2 Flight Interception Method...... 48 Predation by Coleomegilla macula ta....... 48 3.5.5.1 Laboratory Predation Experiments. 48 3.5.5.2 Field Predation Experiments 51 Onion Thrips - Onion Plant Interaction.......... 52 3.6.1 Feeding Rate on Onion Leaf Segments....... 53 3.6.2 Feeding Rate on Whole Onion Plants........ 54 Results and Discussion................................. 57 4.1 Preliminary Field Study.......................... 57 4.2 Population Dynamics of Onion Thrips.............. 59 4.2.1 Relationship between Thrips Infestations of Onions and Distance from Adjacent Field that is Chemically Protected.............. vi 66 4.2.2 4.3 4.3.1 Relationship between Temperature and Rate of Development of Thrips tabaci...... 72 4.3.2 Theoretical Lower Temperature Develop­ mental Threshold of Thrips tabaci......... 87 4.4 Developmental Biology of Coleomegilla maculata... 87 4.5 Population Studies on Coccinellids............... 4.5.2 4.5.3 4.5.4 4.5.5 91 Coccinellid Complex at Eaton Rapids and Laingsburg, 1979......................... 91 Vertical Distribution of Adult Coccinellids ...... 95 Population Dynamics of Coccinellids....... 114 4.5.3.1 Population Index using Behavioral Traps, 1979........... 114 4.5.3.2 Population Densities of Coccinellids at Eaton Rapids..... 119 Arthropod Complex Associated with Coccinellids................. Diel Activity Pattern of Coccinellids 126 127 4 .6 Predation by Coleomegilla maculata............... 136 4.7 Onion Thrips - Onion Plant Interaction.... 146 General Discussion.............. ...... ................ 153 6 . Summary................... 7. 68 Developmental Biology of Thrips tabaci............ 72 4.5.1 5. Dispersion Pattern of Onion Thrips in Field............. *.. *... * , , , , , , , , , . 165 Bibliography................ ........................... 167 8. Appendix Tables ................... vii 175 LIST OF TABLES 1 Fecundity of Thrips tabaci...................... 8 2 Developmental durations of the different stadia of Thrips tabaci........................ 9 3 Questionnaire for Onion Thrips Survey.......... 23 4 Ratio of variance to mean number of thrips/plant and required sample size (1978 Data)................................ 27 Summary of preliminary survey of some onion fields in Michigan, 1978................. 58 Chemical control of onion thrips Thrips tabaci................................... 65 Correlation coefficients of distance versus number of thrips........................ 67 Mean developmental time and percent development per day (100/y) for the different stages of onion thrips at various constant temperatures.................. 73 Linear regression statistics for fitting temperature-velocity curve to straight line and K-values calculated from three equally spaced temperatures.................... 74 Straight line equations for In [(K-P)/P] and the logistic equations..................... 81 Theoretical lower temperature developmental thresholds, t, for various stages for Thrips tabaci................................... 88 Developmental duration in days (mean + S.D.) of Coleomegilla maculata when reared under different conditions........................... 89 Numbers of adult coccinellids caught by three different trapping methods at Eaton Rapids and the Muck Farm, 1979................. 92 5 6 7 8 9 10 11 12 13 viii 14 15 16 Comparison of the total stickyboard trap . catch «j>f coccinellids (catch x 100 traps x day" ) for similar periods and trapping durations at Eaton Rapids and the Muck Farm, 1979............................ 94 Summary of vertical distribution of coccinellids caught on stickyboard traps located in onions at the Muck Farm, Laingsburg, 1979............................... 98 Summary of vertical distribution of coccinellids on 15 stickyboard traps located in spring sorghum at the Muck Farm, Laingsburg, 1980................. 99 17 Vertical distribution of C. maculata in sorghum during the periods 7-06-80 to 7-25-80 and 7-31-80 to 8-07-80 at the Muck Farm................................... 104 18 Summary of vertical distribution of coccinellids caught on 10 stickyboard traps located in winter wheat at Eaton Rapids, 1980............................. 109 19 Summary of vertical distribution of coccinellids caught on 11 stickyboard traps located in or adjacent to spring oat at Eaton Rapids, 1980...................... 110 20 Occurrence of adult coccinellids on giant ragweed, Ambrosia trifida, growing among onions and in an adjacent oat plot, Eaton Rapids, May 24, 1980..................... 120 21 Visual count of coccinellids occurring on 150m length rows of onions at different parts of the field.............................. 130 22 Occurrence of coccinellids on exposed parts of oat plants in early morning (6:30 a.m., 13.5°C) and mid-morning (10:00 a.m., 18°C), July 24, 1980, Eaton Rapids.................... 131 23 Rate of consumption (#/day) of 2nd instar onion thrips by different stages of C. maculata..................................... 139 24 Linear regression data for age of predator (hours) versus number of prey consumed......... 143 ix 25 Linear regression of age (hours) of C. maculata versus number of thrips consumed for three C. maculata larvae together........... v ....... 144 26 Mean thrips count at end of season and mean bulb weight of onions grown adjacent to different cereals (n = 10plants)............. 147 27 Feeding rate of Thrips tabaci on onion leaf segments............................ 148 28 Feeding duration of adult Thrips tabaci for destroying whole plant surface area....... 150 29 Hypothetical thrips densities through the season that will cause certain percentage onion leaf area loss............................159 x LIST OF FIGURES 1 The relationship between egg mortality and temperature for Thrips tabaci (after Ghabn, 1948)............................ 11 Michigan, Localities sampled for pre­ liminary field survey.......................... 22 3 • Layout of field at Eaton Rapids, 1980.......... 29 4 Felt rearing cage (after Lewis, 1973).......... 32 5 Plastic clay rearing cell...................... 34 6 Constant humidity chamber...................... 36 7 Set-up for collecting freshly hatched first instar thrips larvae..................... 38 8 Glass globe for rearing coccinellids........... 41 9 Flight interception trap....................... 42 10 Stickyboard trap.............. 43 11 Set-up of coccinellid predation cell........... sc 12 Set-up for measuring thrips feeding rate on whole onion plant........... 55 Population trend of onion thrips, Eaton Rapids, 1978................................... 60 Population trend of onion thrips, Eaton Rapids, 1979................................... 61 Population trend of onion thrips, Eaton Rapids, 1980................................... 62 Population trend of onion thrips, M.S.U. Muck Farm, 1980............................... 64 Relationship between variance and mean (number of thrips per plant) on log/log scale........................................... 70 2 13 14 15 16 17 xi 18 Temperature-velocity curve for first instar Thrips tabaci......................... 75 19 Temperature-velocity curve for second instar Thrips tabaci.. .................... , . ,. .... 20 Temperature-velocity curve for prepupal stage of Thrips tabaci........................... 77 Temperature-velocity curve for pupal stage of Thrips tabaci .............................. 78 Temperature-velocity curve for four immature stages (1st and 2nd instar, prepupa and pupa) of Thrips tabaci......................... 79 Lower temperature developmental threshold for first instar Thrips tabaci................ 82 Lower temperature developmental threshold for second instar Thrips tabaci.......... 83 21 22 23 24 76 25 Lower temperature developmental threshold for prepupal stage of Thrips tabaci............ 84 26 Lower temperature developmental threshold for pupal stage of Thrips tabaci............... 85 Lower temperature developmental threshold for immature stages of Thrips tabaci....... 86 Vertical distribution of coccinellids on stickyboard traps, in onions through the season, at the M.S.U. Muck Farm, 1979 (a) Coleomegilla maculata plus Hippodamia tredecimpunctata, (b) Adalia bipunctata....... 96 27 28 29 Vertical distribution of coccinellids on stickyboard traps, in sorghum through the season, at the M.S.U. Muck Farm, 1980 (a) Coleomegilla maculata, (b) Hippodamia tredecimpunctata, (c) Adalia bipunctata........ 101 30 Vertical distribution of coccinellids on stickyboard traps, in sorghum, at the M.S.U. Muck Farm, 1980 (all trapping dates together)................................. 102 Vertical distribution of coccinellids on stickyboard traps, in winter wheat (July 2 through July 23, 1980) at Eaton Rapids (a) C. maculata, (b) H. tredecimpunctata, (c) A. bipunctata.............. 106 31 xii 32 33 Vertical distribution of coccinellids on stickyboard traps, in winter wheat, at Eaton Rapids, 1980 (all trapping dates together)....... 107 Vertical distribution of coccinellids on stickyboard traps, in spring oat (July 2 through July 23, 1980) at Eaton Rapids, (a) C. maculata, (b) H. tredecimpunctata, (c) bipunctata......................... 112 34 Vertical distribution of coccinellids on stickyboard traps, in spring oat, at Eaton Rapids, 1980 (all trapping dates together)..... 113 35 Population trend of coccinellids at Eaton Rapids as indexed by different trapping methods, 1979 (a) stickyboard trap, (b) flight interception trap, (c) pitfall trap....................................... 116 Population trend of coccinellids at the M.S.U. Muck Farm using stickyboard traps, July, 1979................................ 117 Occurrence of coccinellids on ragweed in spring oat and onion fields (a) number per ragweed, (b) density (number per unit ground area)........................................ 123 Density of coccinellids in onions, spring oat, winter wheat and alfalfa.......... 125 36 37 38 39 Diel activity pattern of Coleomegilla maculata.132 40 Diel activity pattern of H. tredecimpunctata..,133 41 Diel flight pattern of C. maculata, (a) flight into onion field, (b) flight out of onion field 135 42 Rate of consumption of second instar Thrips tabaci larvae by Coleomegilla maculata larvae..142 43 Onion leaf production through the season........156 44 Comparison of field recorded thrips density levels (broken lines) with hypothetical thrips densities that assume constant 5%, 10%, and 15%. leaf area loss (LAL) through the season........ 161 xiii KEY FOR FIGURES 28 TO 34 HEIGHT (m) □ 0-.3 .3-.6 .6 - 9 .9-1.2 ACTUAL NUMBERS TRAPPED ARE SHOWN xiv 1. INTRODUCTION The onion thrips, Thrips tabaci Lindeman, is cosmo­ politan; it occurrs in almost every part of the world, in temperate, tropical and subtropical areas, and at altitudes ranging from sea level to 2,000 m (Lewis, 1973). It is polyphagous and is reported as a pest of many crops (Lewis, 1973; Stannard, 1968; Metcalf et al, 1962). is not known for certain. Its origin However, through the study of the mode of reprodution and the sex ratio in different parts of the world, O'Neill (1960) and Mound (1973) concluded that it originated from the eastern Mediterranean. Damage done by the onion thrips to onions, cotton and other crops makes it economically important (Sakimura, 1932; Harris et al, 1936; Richardson, 1957; Metcalf et al, 1962; Abdel-Gawaad and Shazli, 1970; etc.), and hence there is an accumulation of literature on this insect. However, most of these works have been directed towards the use of in­ secticides for control and not much information on the de­ tailed biology and ecology is available. Where the biology and ecology have been treated, they lack important informa­ tion needed for an onion pest management program. In Michigan, onions are grown on 7,500 acres and the crop is worth over $18.5 million (Anon., 1981). The crop is attacked by a complex of diseases and insect which separa­ tely and together affect the growth and yield of the crop. As a first phase in designing a pest management program, it is necessary to understand how each pest component 1 interacts with and affects the crop. Having obtained this information, it then becomes pertinent to put together all the pieces into one whole that will simulate the real world situation. On its own, however, the onion thrips are estimated to cause an annual yield loss of 11% in Texas (Richardson, 1953). Being parthenogenetic and with a life cycle of under 20 days, onion thrips are capable of developing very high population densities in a short period when the weather is favorable, leading to severe crop losses. The choice of this topic for the dissertation research afforded me the chance to work on an insect pest that is also a problem of my home country of Ghana, West Africa. In Ghana, onion and shallots form an important part of the agri-business in the north-eastern and south-eastern regions, respectively. Thrips tabaci have been reported to be an important pest of these crops, particularly in the Keta (south-eastern) region (Halteren, 1970). An understanding of the biology and ecology would be a useful basis in designing management programs, even though the environmental conditions in Ghana are different. The aim of this study was to obtain information on the * biology and ecology of the onion thrips, to identify and under­ stand the factors that are responsible for changes in the field populations, to understand the pest-crop interaction and the role of the onion thrips in the onion agro-ecosystem. Information from these studies will hopefully form the basis for developing a management program of onion thrips in an ecologically compatible and economically feasible onion pest management system. 2. .REVIEW OF RELEVANT LITERATURE The taxonomy of the onion thrips, Thrips tabaci,Lindeman (1888) (Thysanoptera: Thripidae), has been changed many times, but all previous names have been synonymized by Priesner (1925). Stannard (1968) revised the taxonomy of the order Thysanoptera and provided keys to identifying the various species that occur in Illinois, Thrips tabaci has been referred to by the following common names: onion or potato thrips (Lewis, 1973), and the tobacco thrips (Federov, 1930), but onion thrips is the most popular in the literature. 2.1 Economic Impact of Onion Thrips Thrips tabaci is polyphagous (Lewis, 1973; Stannard, 1968; Metcalf et al, 1962), the latter stating that it attacks nearly all garden plants, including cauliflower, bean, melons, tomato, cucumber, and also many weeds and some field crops. The onion thrips is a pest of many crops, the most important being onion (Harding, 1961; Lall and Singh, 1968; Richardson, 1953), cotton (El-Saadany et al, 1975; Shoeib and Hosny, 1973; Hosny, 1964) and tobacco (Tutel, 1963; Federov, 1930). The preferred host plant is onion (Stannard, 1968). Reports of the effect of onion thrips on onion yields are very conflicting and seem to indicate that there is a host of other factors involved in the yield obtained. These factors include onion varieties, cultural practices, soil types and environmental conditions. Horsfall (1921) reported 75% yield reduction compared to previous years 3 that had practically no thrips infestation. The loss was directly attributed to onion thrips since surrounding fields which had no thrips produced 400 bushels/acre compared to 60 bushels/acre for the thrips infested field. Maughan (1933, 1934), Hibbs and Ewart (1946) and Sleesman (1946) all showed increased onion yield by suppressing thrips infestations with insecticides. Richardson (1953, 1954) reported 11% and 21% yield reduction respectively when com­ paring the best yield from insecticide-treated plots with untreated plots. Wilcox and Howland (1948) reported yield increases from thrips control, but not in every test as reported by the preceeding authors. Sleesman (1943) and Douglas and Shirck (1949) showed little yield differences using thrips-resistant varieties. Harding (1961) used a Parathion and Dieldrin mixture and obtained drastically reduced levels of thrips infestation, but yields were not significantly different: the best insecticide treated plants averaged about 3 thrips/plant over the entire season compared to 80 thrips/plant for the untreated plants. Shirck and Douglas (1956) showed that onion thrips did not reduce yield unless the population was very high, particularly if the severe infestations occurred early during the growing season. There is no information in the literature on the economic threshold and what thrips numbers mean in terms of their quantitative effects on onion plant growth and yield, although high numbers will kill young plants. Apart from feeding on and interfering with the normal physiological functioning of the leaves, Thrips tabaci have also been reported to cause sterility in onions (Pearson, 1930). They have been reported to be vectors of the tomato spotted wilt virus (Sakimura, 1963), but there is no report of their serving as vectors of any disease in onions. The wounds created by its feeding could become secondarily infected. 2.2 Sex Ratios and Reproductive Biology The field populations of most species of thrips are bisexual but often the females predominate and in some species the males are rare or reproduction is wholly parthenogenetic (Lewis, 1973). In some cosmopolitan species, the sex ratios differ in different regions. Thrips tabaci have a sex ratio of about 1:1 in the eastern Mediterranean and Iran (Mound, 1973), but in most other areas the males are rare. ratio is 1:1000 females in Hawaii The sex (Sakimura, 1932), 0:3000 in Sudan (MacGill, 1927), and no males found in Illinois (Stannard, 1968). Shull (1914) collected 2 males and 226 females in Michigan giving a sex ratio of 1:113. for the vast disparity in sex ratio is not clear. The reason In the latter case, since the 2 males were collected in September, it was concluded that the species is wholly parthenogenetic but that the appearance of males might have been caused by cooler climate conditions. Morrison (1957) also suggested that the difference in sex ratios from place-to-place may be dependent on the temperature and observed that in England, males of Thrips tabaci were found in the fields where it was cooler but not in the artificially warm sub-tropical climate of the greenhouse where reproduction was always parthenogenetic. O'Neill (1960) suggested a different reason for this difference in sex ratios over the range of distributions of the species. He believes that the scarcity of males, and hence parthenogenesis, is common in introduced species because the parthenogenetic forms are more easily spread than the sexual forms. Thus in the area in which the species originated, the sex ratio may be equal or close to 1 :1 , but males will be rare or absent in other areas. Thrips tabaci lays its eggs indiscriminately in leaves, cotyledons, petals, sepals and glumes. The eggs are kidney­ shaped and measure 0.26 x 0.12 mm (Ghabn, 1948). The eggs may be laid singly into the underlying tissue (Ghabn, 1948), or in clusters (Lall and Singh, 1968) or may sometimes be laid in rows (Lewis, 1973). The total number of eggs laid by most plant thrips range from 30 to 300 depending on the species, the individual, and the amount and quality of food available (Lewis, 1973). For some species (e.g., Thrips imaginis) the availability of protein is very critical. Practically no eggs are laid if the females are reared on a protein deficient diet (an average of 20 eggs/female when fed on stamens with anthers removed), but when this diet is supplemented with protein they lay an average of 209 eggs (Andrewartha, 1935). Loan and Holdaway (1955) reported that 7 although pollen is the preferred diet of Haplothrips leucanthemi the females are able to lay fertile eggs without it. There is no mention of the requirement of protein for oviposition of Thrips tabaci in the literature and the species reproduces effectively without any observable protein source. Evidence of the possible effect of food constitution on oviposition is given by Abdel-Gawaad and Shazli (1970) who reported that more eggs were laid and the duration in the various stages were shorter if onion thrips were fed on green leaves of stored onion bulbs, and on caster oil seedlings compared to a food source of stored onion bulbs, new onion bulbs or garlic bulbs. Lall and Singh (1960) also noticed a change in the developmental durations with respect to the age of the onion plant and attributed this to changes in nutritive constitution of the plant. Reported fecundity of the onion thrips ranges from 4.5 to 80 eggs per female (Table 1). It is not surprising that this range occurs since the data reported are for different host plant species. Even on the same host plant, fecundity changes with host nutrient constitution. Ghabn (1948) reported that only 73% of females that were observed laid any eggs. Temperature probably does not affect total egg production of Thrips imaginis once the threshold for laying has been exceeded (Andrewartha, 1935), although the rate of oviposition depends on it. Lewis (1973) reported that Thrips tabaci lay more eggs at higher temperatures. The lower threshold temperature for oviposition has been determined as 8.5°C for Table 1. Temp.°C Egg production of Thrips tabaci. Average # egg/female Range Host Reference 18-33 4.5 2-22 cotton Ghabn (1948) --- 14.5 ?-46 cotton Eddy & Clarke (1930) 29 37.4 0-109 Emelia sagittata Sakimura (1932) 18 80 --- Onion? Sakimura (1937) --- 28.7+3.2 13-54 cotton Abdel-Gawaad & Shazli 15.8 49.8 --- Onion Lall & Singh (1968) 18.0 51.7 --- Onion Lall & Singh (1968) 23.4 55.0 --- Onion Lall & Singh (1968) 30.8 28.2 --- Onion Lall & Singh (1968) Table 2 . Duration of Thrlpa tabaci In various stages In days. L1 L2 P P P PreA Ov.A Post.A Z A Host Reference 4.7 2.3 2.8 1.4 3.2 3.8 (1-7) 8.6 (1-26) 2.7 (0-6) 14.5 (4-28) Cotton Eddy & Clarke (1930) - 4.6 (2-9) 2.0 (1-4) 2.8 (1-6) 1.0 (1-4) 2.7 (1-5) 3.1 (1-5) - - 18.3 Cotton Watts (1934) 25 - 6.0 6.1 1.2 2.8 - - - - Onion Harris et al (1936) 30 - 4.0 4.2 1.0 2.0 - - - 19.9 Onion Harris et al (1936) 21 - 3 50 6 Onion? Sakimura (1937) 18-33 - - Cotton Ghabn (1948) 22 65-70 4.9 + 0.7 30.8 Kean 47.6 4.8 23.4 54.4 18.0 15.8 T°C RH° - - - E 1.5 (1-2) 2.1 (1-3) 6.9 - - 14.5 (1-30) - - 2.1+0.3 12.9+5.5 3.7+0.4 18.7+3.6 Cotton Abdel-Gawaad & Shazll (1971) 5.9 1.4 2.4 - - - 20.2 Onion Lall 6 Singh (1968) 6.0 5.5 1.7 2.8 - - - 20.1 Onion ibid 61.8 7.9 6.2 2.0 3.5 - - - 19.6 Onion ibid 78.5 8.5 2.0 4.0 - - - 18.8 Onion Ibid E = Egg L ” Larva P P*= Prepupa P = Pupa 2.3 (1-4) - 6.5 PreA ■ Pre-oviposition Adult Ov.A * Ovipositing Adult Post.A “ Post-ovipositing Adult E A » Sum for Adult 10 Thrips imaginis by Andrewartha (1935) and 12.5°C for Heliothrips haemorthoidalis by Rivnay (1935). Thrips imaginis laid similar numbers of eggs at 12.5°C, 15°C, 20°C and 23°C (Andrewartha, 1935). No threshold temperature for oviposition has been found in the literature for Thrips tabaci. 2.3 Developmental Biology The life cycle of the onion thrips involves the following stages: egg, two larval instars, prepupa, pupa and adult. The duration of development from egg to adult is dependent on temperature but generally falls within 12-16 days. Egg. Incubation lasts 4.8 days at 30.8°C (Lall and Singh, 1968), but ranges from 4.0-8.5 days depending on the temperature (Table 2). The rate of successful hatching is also affected by temperature. At 24°C, the hatchability of eggs is about 41%, it drops sharply at lower temperatures, and averages 38% for 18-33°C (Ghabn, 1948) (Figure 1). These data compare with 30-40% hatchability for Anaphothrips obscurus (Muell.) Larvae. (Hinds, 1903). The two larval instars are similar in appearance and, except for the period soon after the emergence of the first instar, the use of size to distinguish the two stages is unreliable. Both instars are pale yellow in color. The main distinguishing feature in these two stages is the shape of the third antennal segment. The first instar has a short, top-shaped third antennal segment which is as long as it is wide and a pointed terminal (6th) segment. The second instar has a slender third antennal segment that is longer than 11 h a tc h 45 egg 33 29 25 21 23 25 Temperature 27 °C Figure 1. The relationship between egg mortality and temperature for Thrips tabaci {after Ghabn, 1948). 12 it is wide and a more rounded terminal segment (Ghabn, 1948). Each stage lasts between 2-3 days (Table 2). The develop­ mental duration varies with the age of the onion plant (Lall and Singh, 1968) and with the food source (AbdelGawaad and Shazli, 1970). No larval developmental temperature threshold has been determined for Thrips tabaci, but Ewald and Bust (1959) reported the lower threshold temperature of 8°C for Taenothrips laricivorus Kratochvil. The mature second instar onion thrips stops feeding and migrates 3-5 cm or more into the soil (Ghabn, 1948) where it molts into the prepupa. Prepupa and Pupa. The prepupa and the pupa are both pale yellow but can be easily distinguished from the larvae by the presence of wing buds. The prepupa has short wing buds which hardly exceed the length of the head and the pro­ thorax together. The antennae are folded back in the head, as in the pupa, but are shorter and scarcely reach the anterior margin of the prothorax. The wing buds of the pupa are much longer than the head and thorax together. They usually extend beyond the anterior half of the abdomen. The antennae are also long and cover about half the length of the prothorax. The prepupa and the pupa last a little over 1 or 2 days, respectively (Lall and Singh, 1968) Adult. (Table 2). The adults vary in color from pale brown when they first emerge to dark grey brown and measure 0.8 mm (segments contracted) to 1-2 mm (segments extended). Antennal segment I is light brown but the other segments are light 13 brown except the bases of segments III-V which are somewhat paler (Stannard, 1968). Occellar pigment is grey to yellowish grey as opposed to red pigments found in most other species. The forewings usually have 4 or more apical bristles on the fore vein instead of 3 or fewer as in other species. The longevity of the adult has been recorded as 30 days (Shepard, 1925). Table 2 gives the mean durations reported by various authors. 2.4 Natural Enemies Lewis (1973, Appendix 3a, 3b) recorded the parasitoids and predators of various species of thrips. The onion thrips, Thrips tabaci, is attacked by 6 different parasitoid species of which 2 are found in the U.S.A. — Dasyscapus parvipennis Gah.(Eulophidae: Hym.) and Thripoctenus russelli Crwf. (Eulophidae: Hym.). dators in the U.S. He also listed 8 species of pre­ Sakimura (1937b) found a density dependent relationship between Thrips tabaci and Thripoctenus brui Vuil in Japan and recorded 20-80% parasitism. Saxena (1971) found another eulophid, Ceranisus sp., parasitizing the onion thrips. The females of this parasitoid select and ovipost on the second instar onion thrips and the adult parasitoid emerges from either the prepupa or the pupa. Bourne and Shaw (1934) recorded a fungus, Entomophthora sphaerosperma Fres., attacking onion thrips in Massachusetts. Carl (1975) found another Entomophthora sp. which attacks and kills both adults and immatures in 3-6 days, but doubted if this was the same species reported in Massachusetts. Incidence 14 of the latter Entomophthora sp. was found to be densitydependent. It causes high host mortality but is not an effective control agent, since it does not become abundant in the field until late in the season when most of the damage has already been done by the pest. 2.5 Population Dynamics To understand how the numbers of a particular pest species change within and between seasons, one needs to know how the initial population becomes established, how the numbers build up and change through the season, and whether there is migration during the season or at the end of the season. Overwintering and sources of infestation. Thrips tabaci overwinter principally in the adult stage (Boyce and Miller, 1954; Ghabn, 1948). Vinson (1929) reported that this insect overwinters as pupae inside onion pulps. Ghabn (1948) also reported seeing larval onion thrips in late December and early March but noted that such larvae were found in periods preceded by warm weather and that the larvae died when the temperature fell again. Dimitrov (1975) reported that adults overwintered in the soil. This was contrary to the findings of Boyce and Miller (1954) that the adults overwintered in fields of clover and alfalfa and that grass sod bordering onion fields, onion culls and muck soils did not appear as suitable overwintering sites. Irrespective of the stage and site of overwintering, it is the adult onion thrips that colonize available suitable plants at the onset of warm weather. The overwintered adults start reproducing and building up their numbers on weeds and forage crops including lucerne (Banham, 1968; Boyce and Miller, 1954), alfalfa and set onions (Horsfall, 1921) and then later migrate into fields of seeded onions. Boyce and Miller (1953) reported that the initial infestation was by adults that dispersed from nearby crops and that further population increases in onions could be associated with the cutting of adjacent hay crops. Horsfall (1921) observed the spread of Thrips tabaci from an adjacent alfalfa field into seeded onion fields. The spread was initiated by cutting of the alfalfa and by prevailing winds that blew from the alfalfa to the onion field. Banham (1968) found a similar increase of thrip populations on asparagus spears following the cutting of bordering forage crops. Developmental rates. A close relationship exists between the rate of development of insects and other animals and the temperature. This relationship has been reviewed by various authors (Crozier, 1926; Shelford, 1929; Belhradek, 1930; Uvarov, 1931; Janisch, 1932; Hoskins and Craig, 1935; Huffaker, 1944; Davidson, 1944; Pry, 1947; Wigglesworth, 1953; Andrewartha and Birch, 1954; Messenger and Flitters, 1958). The rate of development at different constant temper­ atures does not increase proportionately with increasing temperature throughout the range suitable for development. The rate is slower at the lower temperatures, faster in the median range and slower at the upper temperatures. This rela­ tionship is best described by a logistic curve, the equation 16 for which was first formulated by Pearl and Reed (1920) using the Pearl and Reed equation: where: Y = the time required for complete development of a particular stage at a given constant temperat X, and K, a, b = constants. When 1/Y is plotted on the ordinate against temperature, X, on the abscissa, K is the upper asymptote of the resulting velocity curve which is typically sigmoid or S-shaped. a is the parameter which indicated the relative position of the origin of the curve on the abscissa and b represents the degree of acceleration of development of the stage in relation to temperature and hence determines the slope and course of the curve. Davidson (1942) gave reasons why the data on the rate of development at temperatures above the peak should not be included in calculating the formula for the temperaturevelocity curve. Commonly the reciprocal of the time required for development of each stage, 1/Y, is multiplied by 100 so it becomes 100/Y, the percent development per unit time (Davidson, 1944; Andrewartha and Birch, 1954). Davidson (1944) claims the temperature-velocity curve represents the trend of the speed of development of insects for 85-90% of the complete range of temperature at which development can occur. According to Matteson and Decker (1965) the point where the velocity line crosses, or is extrapolated to cross, 17 the temperature axis: is theoretically the threshold of development of that particular stage. The number of day degrees it takes to complete development can be calculated using the developmental threshold temperature: Baskerville and Emin (1969) have developed another method for calculating degree days using a modified sine curve. Day degrees is one of the simplest thermal heat units used in measuring development and gives a more useful measure of the duration of development than simple time units. 3. METHODS AND MATERIALS 3.1 Sampling Methods for Onion Thrips The various sampling methods used by different authors to estimate relative population densities of thrips on vegetation have been discussed by Lewis (1973). The choice of a particular method depends on the sampling accuracy desired, the physical characteristics of the plant and the behavior of the thrips. Sampling techniques that have been used in this study and also cited in the literature for the onion thrips, have been restricted to the larvae and adults only. The eggs of the onion thrips are laid embedded in the plant tissue and since the onion leaves are thick, they need to be chemically treated to see the eggs. The prepupal and pupal stages are spent in the soil and are consequently not sampled, since soil extractions could be cumbersome. The larvae and adults of the onion thrips tend to be cryptic and usually hide in the narrow spaces between the bases of the onion leaves. They do not ordinarily crawl down deep into these spaces and hence can be easily seen when the leaves are slightly pulled apart. However, both these stages will crawl deeper down the leaf-bases if the plant is grossly disturbed. The adults may fly away with such disturbance. 3.1.1 Passive or Liquid Extraction The type of leaf surface is particularly critical for this method since the thrips are killed by the liquid and must 18 19 be washed out. Seventy percent ethyl alcohol (Le Pelley, 1942; Ota, 1968), petrol (Bullock, 1963) or detergent (Taylor and Smith, 1955; Ota, 1968) have been used with good recovery for flat leaves, e.g., coffee and rose leaves. This method was not found suitable in this study for the reasons given below: 1) whole onion plants need to be uprooted for the extraction, 2) during the process of uprooting the plant, adult thrips fly away, 3) alcohol-killed thrips get caught in their hiding places, and 4) the process is cumbersome, requiring one to carry cylinders of alcohol from plant to plant, and spend long periods of washing and counting the thrips. At very high densities of thrips, however, this method was used in some parts of this study. The plants were cut immediately above the soil surface with a pair of scissors if they were not too thick, dipped in 70% alcohol and then each leaf was pulled off and washed in the alcohol. Thrips were sieved through a fine nylon mesh and counted through a binocular microscope. 3.1.2 Dynamic Extraction In this method, the thrips are made to move away from the plant surface to a collection point by expulsion with heat through the Berlese or Tullgren funnel (Schirck, 1948); or expulsion with turpentine (Taylor and Smith, 1955). 20 Drawbacks are: 1) and 2) as above, and 3) since extractions cannot be done immediately and in the field, thrips crawl away and are lost during transportation, or if they are held in plastic containers, they get trapped and die in moisture that condenses in the bags. 3.1.3 Knockdown Method Sakimura (1937a) tapped onion plants over a black card and counted the dislodged thrips. andcotton lint cloth instead Other authors have used felt of card so the thrips are temp­ orarily entangled in the fibres (Powell and Landis, 1965; Henderson and MacBurnie, 1943). This method is quick but has the following drawbacks: 1) and 2) as above, and 3) the method underestimates the numbers since the thrips, particularly the larvae, are not easily dislodged from their hiding places. 3.1.4 Direct Counting This method is most suited for plants on which the thrips can be easily seen and for thrips species which remain still long enough to be counted. on the field. Estimates of density can be obtained No damage is done to the plants and no crop loss through uprooting occurs and hence larger samples can be taken, particularly at the low thrips densities, without cost to the grower. However, plants need to be handled with minimum agi­ tation of leaves. For sampling thrips on onion plants, one has to get very close to the plant to count the larvae.‘ The method 21 is laborious at high thrips densities. This was the favored sampling method. Since the thrips are more active and crawl faster in warmer weather, all sampling was done before mid-day, if possible. Counting was done sys­ tematically, starting with the outer leaves and working towards the center. Since the adult thrips move away faster than the larvae when the plant is distributed, these were counted first. The leaves were pulled apart very gently to expose individuals hiding between the bases of the leaves. A small hand lens was occasionally used to identify the first instar larvae. The following environmental factors were measured through the growing season: temperature, relative humidity and rainfall. 3.2 Preliminary Field Survey 3.2.1 Occurrence and Importance of Onion Thrips A survey for the distribution and determination of pest status of the onion thrips in Michigan was carried out in the summer of 1978 at the following locations: East Lansing, Grant, Eaton Rapids, Decatur, Stockbridge and Newago (Fig. 2). The survey was conducted at the early part of the season and again towards the end of the season. At each location, the opinion of the grower was sought on the pest status of the onion thrips in his field using a questionnaire (Table 3). Where the information desired was not yet available for the season being surveyed, the grower provided data from previous years. 22 1. East Lansing 2. Hudsonville 3. Grant 4. Stockbridgs 5. Nswago 6. Eaton Rapids Figure 2. Michigan, Localities sampled for preliminary field survey. 23 Table 3. Questionnaire for Onion Thrips Survey. Michigan State University Department of Entomology East Lansing, MI THRIPS SURVEY Location Date Onion Variety Date Planted ______________________ Plant Age Insecticide(s) Used ___________________________ Rate of Application ___________________________ Frequency 24 The direct counting method was used for the early season survey, but towards the end of the season, the extraction method was used in order to reduce the time spent at each location. One-hundred plants were sampled along each diagonal of the field in the direct counting method, but this size was halved for the extraction method. 3.2.2 Natural Enemies Survey At each location mentioned above, a visual search was made for any predators that occurred in the fields. A search was made in the surrounding bushes for any predators. Large thrips larvae, presumed to be second instars, were brought back to the laboratory and reared on onion leaf sections in petri dishes to see if any were parasitized. Five-hundred larvae were reared from each location. 3.3 Population Dynamics of Onion Thrips The survey of the onion-growing areas of Michigan showed there was only one grower that grows onions on a commercial scale that had a thrips problem most of the time. This grower, Dale Kunkel, grows onions and other crops (potatoes, carrots, soybeans, and some small grain) in the organic fashion and does not use any pesticides. The Dale Kunkel farm, located near Eaton Rapids (35 km SW of the M.S.U. campus), was thus chosen as the site for most of the field research through 1978, 1979 and 1980 growing seasons. Data were also collected from the M.S.U. Muck Farm at Laingsburg during the 1980 season. 25 Lewis (1973) reported that thrips are more abundant near the edge of the field than in the center and hence it was necessary to sample the end rows. With aphids, Sylvester and Cox (1961) reported that the distribution in the field during the initial phase of infestation is random, but later becomes contagious as each aphid reproduces. There is no detailed study of the field distribution of onion thrips in the literature, but it is doubtful if a large onion field will show a random distribution of thrips even at the early phase of infestation. The initial colonizers will most probably land near the edges of the field so that although the distribution in the early phase of infestation may approach random, there may be higher densities near the edges of the field at this time. The distribution may become further clumped as the thrips reproduce. 3.3.1 Field Data, 1978 Sampling done in 1978 at Eaton Rapids was considered as a preliminary study. Sampling was done bi-weekly from the time the plants were 5-7 weeks old, through harvest. The onion field was about 300 x 50 m and was bordered on two sides by an edge row of trees and shrubs and potatoes and onions on the other sides. were numbered. All the rows in the field Starting from the middle row, a bi-weekly sampling scheme was designed so that one whole row was sampled on each visit, i.e. 2 rows were sampled each week. For each row, all plants within the first 1 m length were counted and carefully searched for larvae and adult thrips. 26 Several 1 m length samples were taken along the same row, separated by about 30-40 m (30 long strides), to provide about 500 plants/row and including a sampling of the last 1 m length of the row. On the next sampling date, one row was skipped and the next row sampled, thus reducing the chances of disturbing a row before sampling it. This order of row sampling was continued to the edge of the field and the direction reversed and worked to the other side of the field using the rows that were previously skipped. 3.3.2 Field Data, 1979 A different sampling scheme was used for the 1979 season. Along each row, small plots of 10 m in length were marked off and staked. Each plot was separated by 5 m of length of unmonitored onion plants. Only alternate rows of onions were used so as to reduce the chances of disturbing insects in adjacent plots while sampling another. Data from the 1978 season for single plant entries of thrips were used to estimate the number of samples needed. Southwood (1978) gives the number of samples N as: N = (s/Ex)2 where s = standard deviation, E = predetermined standard error expressed as a decimal, x = mean Table 4 shows the ratio of the variance to the mean during 1978 (which give the measure of the degree of clumpedness) and also the number of samples needed for different times in the season. With the direct counting method, the Table 4. Ratio of variance to mean number of thrips/plant and required sample size (1978 data) Time Mean Thrips/plant Variance 2 s s2/x Required sample size oo s-ss-e i— 1CM 1.8 1.4 106 (E 26 (E Early season 1.7 4.9 2.9 169 (E 42 (E Early season 1.4 5.6 4.0 289 (E = 10%) 71 (E = 20%) Early season 3.3 23.1 7.0 212 (E 53 (E Mid-season 14.7 522.0 35.5 242 (E 60 (E Late season 34.8 1534.8 44.1 127 (E 32 (E II II N>l-* O O S'S S'S W'w' 1.3 II II Early season 8*8 s-2 o o II II pH CM o o II II pH CM N B'SS'S O O CM II II pH 28 chances of seeing and counting all larvae are low and further become decreased at high thrips densities, thus, the error margin was set high, 1 0 % and increased towards the end of the season as the variance and degree of clumpness increased. If a larger error was not permitted, very large samples had to be taken. An error margin of 10% or better was decided on, so that about 200 plants could be sampled at the begin­ ning of the season. An error margin of 20 % or better, in the middle to the end of the season, allowed a sample size of about 100 plants mid-season, and 50 plants late season.. Other authors, Hoerner (1947), Shirck (1948) and Faulkner (1954) have used a sample size of 10 plants and did not alter this size through the season. 3.3.3 Field Data, 1980 Onion sets, which were left in the field at Eaton Rapids at the end of the 1979 season, were monitored regularly from the first week of May, to determine when onion thrips began migrating into the field. Monitoring was done weekly until the numbers of thrips began to rise and the onions from the fields to be used for the regular sampling program were 6 weeks old. The layout of the field at Eaton Rapids was such that the onions were adjacent to spring oat, winter wheat and alfalfa. (Figure 3). All onions and spring oat were planted between April 28 and May 4, 1980. The winter wheat and alfalfa were planted at the end of the 1979 season. The onion cultivar used for the study, 'Southport white globe', was the same cultivar that was used in the preceeding seasons H* IQ C n> u t- i Pi Ol O 3 CO 3 OAT CAGED BORDER EXPTS, 1< O P WEEKLY rt O Hi fl> Qj PI (+ W P) (+ CO ALFALFA *. oo VO CD O o 3 28m o z o z Hi H- o 3 » Pi 'O H0. SAMPLING WMTER WHEAT ---------------------- •

.05). 59 No parasitoids were recovered from any of the larvae that were collected from all the surveyed fields (total n = 3000). 4.2 Population Dynamics of Onion Thrips The population trend of onion thrips on onions through the growing season is summarized in Appendix Tables l a , b, c, d. Data were collected at Eaton Rapids in 1978, 1979 and 1980. In 1980, data were also successfully collected from the M.S.U. Muck Farm at Laingsburg for the first time during this study. In the two previous years, weeds had completely overrun the experimental field. No pesticides were used in this field and it was impossible to keep up manual weed control. The 1980 Muck Farm data were collected from another section of the farm (within the fence) where cultivation has been done yearly and also the weeds were under chemical control. Figure 13, 1978, showed the mean number of adults per plant was either equal to or higher than that of the larvae at the beginning of the season but soon the predominant stage recorded was larvae. the onions This is to be expected since were planted from seeds which cannot harbour any larvalthrips. The adults are the colonizers and they migrate into the field when the plant is only a few weeks old. As the adults reproduce, the numbers of larvae soon increased and these became the predominant stage. This phenomenon was shown even more clearly in 1979 (Fig. 14), where the thrips counted at the first date of sample was (L o g (N + 1 )) # THRIPS/PLANT 0.8 MEAN 1.4 ★ TOTAL ■ LARVAE • ADULTS 1.2 0.6 0 .4 0.2 600 8 0 0 1 0 0 0 120 0 1 4 0 0 ) T I0 JUNE D tf JULY 20 AUGUST Figure 13. Population trend of onion thrips, Eaton Rapids, 1978. RAIN (in) 61 1.0 111 \ .8 1.6 ■ LARVAE • ADULTS .4 # THRIPS/PLANT .2 1.0 MEAN (Log(N *1)) * TOTAL 0 .4 0.8 0.6 0.2 600 8 0 0 1 0 0 0 1 2 0 0 1 4 0 0 1 6 0 0 DD (>7.4°C) I JUNE1 10 JULY 20 I 10 20 ■ AUGUST I 10 I SEPT Figure 14. Population trend of onion thrips, Eaton Rapids, 1979. 62 2.68 2.00 21.3 10S .7 ★ TOTAL ■ LARVAE 60 • ADULTS H Z < -i Q. 40 oc z 10 300 — 5 00 700 900 1 1 0 0 1 3 0 0 DD° ----,--- ,---- .----.---- .— JUNE JULY .----.<* 7r-4°c) AUGUST Figure 15. Population trend of onion thrips, Eaton Rapids, 1980. 63 comprised entirely of adults. It was necessary to transform the number of thrips per plant for some years to a log (N+l), since the numbers were extremely low at the beginning of the season compared to the number at the latter part of the season. This transformation spreads out the lower numbers relative to the larger numbers. The general pattern of the population buildup for the two stages combined, approached an exponential progression, particularly the 1979 data (Fig. 14), although this was not uniformly so. It is interesting to note that in both 1978 and 1979, the graphs for the adults show more perturbation than the larvae. suggested: a) The following reasons are a sudden rise in the adult numbers could be due to recent immigration into the field from adjacent onion fields or other crops e.g., alfalfa that has been cultivated, harvested, or disturbed in a manner that will cause resident thrips populations to emigrate. move beyond a few plants at best, Larvae so disturbed cannot b) since the adults occur more on the exposed parts of the leaves, they are more likely to be disturbed. The larvae, on the other hand, hide between the bases of the leaves and are thus protected until their numbers become high, in the middle to late season, when they can also be found on the exposed parts of the leaves. The graph of the 1980 season (Fig. 15) showed a similar exponential rise in the numbers of thrips per plant (both stages together) until the end of July. The population crashed severely after this time due to heavy rainfall of 64 z < oc ★ TOTAL ■ LARVAE • ADULTS h- z < CO Q. OC X * z < 111 2 900 20 JULY 1100 I 10 1300 1500 DD (> 7 .4 °C) 20 AUGUST SEPT Figure 16. Population trend of onion thrips, M.S.U. Muck Farm, 1980. Table 6. Chemical control of onion thrips, Thrips tabaci. Treatment lb ai/A Harvest Data Yield (lb/lOft. wt/bulb row) Thrips count^ (mean #/plant) 7/25 7/31 8/7 8/14 8/28 Orthene 75WP 0.5 8.8 0.10 0.8b 1.4 1.8a 1.3a 0.7a Orthene 75WP 0.75 9.4 0.10 0.7a 1.1 7.3b 11.4b 2.8ab Orthene 75WP 1.0 9.0 0.11 1.0c 2.3a 1.6a 0.9a Diazinon AG 500 0.5 8.2 0.11 1.3cd 1.6 12.8b 8.8ab 6.8ab Untreated --- 9.1 0.10 1. 5d 1.6 6.5b 12.4a 12.6b ^Means followed by the same letter are not significantly different (P>0.05, StudentNewman-Keuls, test of log transformed counts). 66 6.57 cm (2.68 in.) on August 2nd and 2.33 cm (.95 in.) the following day. At the Muck Farm (1980), the population trend in the untreated onions again assumed an exponential pattern until mid-August and then flattened (Fig. 16). The mean number of thrips per plant in the untreated plots was significantly different from the treated plots on some dates (Table 6). This difference in numbers of thrips was not reflected in the bulb weights and there was no difference between treat­ ments. Thrips infestation at the Muck Farm was generally low compared to Eaton Rapids where infestation at one point was over 120 thrips per plant. 4.2.1 Relationship between Thrips Infestations of Onions and Distance from Adjacent Field that is Chemically Protected By the middle of the 1980 field season, it was noticed that the numbers of thrips per plant varied with the locations within the study area. The sampling design that was being used up to this time involved sampling a given number of ran­ domly chosen plots on a particular date. The assignment of plots was done at the beginning of the season and the sampling scheme was not changed even though a pattern of uneven in­ festations was beginning to be noticed. Lower infestations seemed to occur in plots near an adjacent field where regular pesticide control was practiced. At the end of the season, the study area was divided into six parallel and equal blocks starting from one end of the field. Each block was 50 m long. Table 7. Date Correlation coefficients of distance versus number of thrips. y-intercept n slope r sign, level 6-13-80 4.5 6 6.5 0.90 P > .01 6-19-80 34.7 5 10.1 0.46 P < .05 6-26-80 81.4 5 74.4 0.90 P > .01 -180.0 5 164.3 0.88 P > .01 7-10-80 -75.3 5 197.6 0.94 P > .01 7-16-80 380.0 5 1023.4 0.87 P < .01 7-23-80 1711.3 5 158.7 0.31 P < .05 8-6-80 2114.6 5 445.2 0.56 P < .05 8-13-80 2845.0 5 -349.5 -0.78 P < .01 7-2-80 68 For each date, regression analysis was conducted for distance away from this end versus the number of thrips found. Since the assignment of plots to the sampling dates was purely random, there was not a good control over plot distribution to each date. Fortunately, most dates had plots scattered over at least three blocks except July 30, 1980, so this date was deleted in the analysis. The correlation coefficients obtained showed a strong relationship between distance and infestation (Table 7). The reason for this difference in rate of infestation with location along the length of the field may be due to the fact that to the northeastern side of the study area, just across the ditch and about 10 m away from the "zero end" of the field, is another grower's field where regular insecticide treatment is practiced. It is suggested here that insecticide drift from the adjacent field affords some protections to the onion plants near-by. However, the pre­ vailing wind in the area originate from the south-westerly direction and could not account for the insecticide drift into the organically grown onions. A more thorough study will be needed to confirm and understand this uneven in­ festation rate within the study area. 4.2.2 Dispersion Pattern of Onion Thrips in the Field The distribution of the onion thrips in the field gives an indication of how these insects occur in the habitat and hence how they should be sampled and is also important when considering the method to be used in analysing the data. As stated earlier in the methods, the distribution may not be the same throughout the growing season and thus even though the early adult colonizers of the field may be assumed to be randomly distributed, this will be expected to change as the season progresses and the adults reproduce. To study this dispersion phenomenon, data from the early, middle and 2 late season was analysed. The ratios of variance (s ) to mean (x), (x = number of thrips per plant, in this study), is less than unity for regular (or even) distribution, equals zero for random distributions, and greater than one for contagious distributions (Southwood, 1978). In the early part of the season, the variance/mean ratio was approximately equal to one, indicating a near random dispersion but became greater than one towards the middle and late seasons (Appendix Table 2). harvest. The population became even more clumped towards The non-random distribution of the onion thrips means some transformation of data is necessary for analysis. In general, most statistical tools for analysing data require that the frequency distribution be normal which in turn means the variance should be independent of the mean, or better yet, the variance be homogenous. A normal dis­ tribution also means the individuals are randomly distributed in the field population. The following analysis was conducted on some of the 1980 data, using Taylor's Power Law (Southwood, 1978) to determine if the population was contagious and if so, what transformation was required for statistical analysis (Appendix Table 2). From this law, the relationship between 70 •• (LOG S N a o o a - 13-80 VARIANCE « 6 -1 9 -8 0 a 8-26-80 • 7 -1 0 -8 0 • 7 -3 0 -8 0 A A 1 MEAN (LOG X) Figure 17. Relationship between variance and mean (number of thrips per plant) on log/log scale. 71 the variance, s 2 and the mean, x is given by s2 = a xb --------------------------- (2) where a is largely a sampling factor and b is an index of aggregations characteristic of the species. x and s 2 The values of were plotted on a log/log scale (Fig. 17) and fitted to a straight line using linear regression. The log of equation (2) log s2 = log a + b log x ------ <-— -— (3) is a straight line with "y-intercept" of 0.13, slope b of 1.68 and r of .97. 2 read off on the s The value of a in equation (3) is either axis at the point x = 1or calculated from the linear regression equation for x =1, remembering 2 that the value for s obtained at x = 1 on the graph is the true intercept since log 1=0.0. Southwood (1978) gives the variance stabilizing transformation function, f(x) as f(x) = Q/x - b ^2 d x ---------- (4) The transformation is obtained from the value of z in the equation: z = xp ------------------------------- (5) where x = the original (raw) number, z^ = the transformed value and p = 1 - %b. From the data used, p = 1 - %(1.68) = 0 . 1 6 and the transformed value, z is z = x ‘1 6 ----------------------------- (6) Because of the contagiousness of the distribution of the onion thrips inthe field,it would be necessary form the raw data obtained using equation (6) to trans­ to adjust it to a normal distribution before making any statistical comparison. 4.3 Developmental Biology of Thrips tabaci 4.3.1 Relationship between Temperature and Rate of Development of Thrips tabaci The results of rearing the larval, prepupal, and pupal stages of the onion thrips at various constant temperatures, are shown in Table 8. No complete development of any stage was observed at 5°C for up to 28 days except for one first instar that molted and died almost immediately after 25 days of observation. This suggests that 5°C is close to or below the lower developmental threshold for the immature stages. At 10°C, the development from first instar to the emergence of the adult took 62.9 days. The development for these same stages was much faster at 17°C and above. Thus only about 1.59% of the development occurs in one day at 10°C and 13.51% at 29°C (Table 8). When the percent development per day (100/y) is plotted against temperature, a sigmoid-shaped temperature-velocity curve should be obtained, the middle portion of this curve being close to straight (Davidson, 1944). The plots for the first instar, second instar, prepupal and pupal stages are shown in Figures 18, 19, 20 and 21. Figure 22 is the temper- ature-velocity curve for the four immature stages together. It appears that the temperature range studied, 10°C to 29°C, falls within the middle straight line portion of the sigmoid curve. Addition of temperatures outside this range will provide the typical sigmoid curve which fits the equation: Table 8. Mean developmental time and percent development per day (100/y) for the different stages of I.e. onion thrips at various constant temperatures. Temp. °C x Larva 1 Prepupa Larva 2 y 100 (± S.D) y 5 y <+ S.D) 100 y Pupa y 100 (± S.D) y Total y 100 y 100 (± S.D) y (± S.D) y No complete development after 28 days 10 19.0±1.3 17 3.26 17.6+4.0 5.68 10.8+3.0 9.26 15.5+2.8 6.45 62.9+11.1 1.59 4.040 25.00 4.840.8 20.83 3.040.3 33.33 4.640.5 21.74 16.4+1.6 6.10 20 3.1+0.2 32.26 3.240.6 31.25 2.1+0.4 47.62 4.841.1 20.83 13,2+2.3 7.58 23 3.1+0.2 32.26 3.8+0.1 26.32 1.940.2 52.63 3.140.9 32.26 11.041.4 8.40 26 2.240.3 45.45 3.240.3 31.25 1.640.4 62.50 2.340.4 43.48 9.341.4 10.75 29 2.140.3 47.62 2.340.3 43.48 1.040 2.040 50.00 7.440.6 13.51 100.00 Table 9. Linear regression statistics for fitting temperature-velocity curve to straight line and K-values calculated from three equally spaced temperatures. Stage Regression statistic y-intercept slope r P-value (=100/y) at 20°C 23°C 26°C K 1st instar -15.07 2.23 .98 29.45 36.13 42.81 72.26 2nd instar -10.19 1.76 .95 25.00 30.28 35.36 54.98 Prepupa -38.09 4.27 .96 47.33 60.14 72.96 120.50 Pupa -18.92 2.31 .98 27.20 34.12 41.04 68.24 Total -4.40 0.59 .99 7.49 9.28 11.06 18.30 (1 0 0 /y ) 50 16 40 12 30 8 20 4 5 10 15 20 TEMP. °C 25 30 PER DAY 20 MEAN PERCENT DEVELOPMENT MEAN DEVELOPMENTAL DURATION, y (DAYS) 75 Figure 18. Temperature-velocity curve for first instar Thrips tabaci. (1 0 0 /y ) PER DAY MEAN DEVELOPMENTAL 40 30 MEAN 20 DEVELOPMENT 50 20 PERCENT DURATION, y(D A Y S ) 76 5 10 15 20 25 30 TEMP. °C Figure 19. Temperature-velocity curve for second instar Thrios tabaci. 12 100 10 80 6 60 4 40 2 20 MEAN 8 PERCENT DEVELOPMENT PER DAY (1 0 0 /y ) I MEAN DEVELOPMENTAL DURATION, y (DAYS) 77 5 10 15 20 TEMP. 25 30 °C Figure 20. Temperature-velocity curve for prepupal stage of Thrips tabaci. Figure 21. Temperature-velocity of Thrips tabaci. MEAN DEVELOPMENTAL DURATION, y (DAYS) 00 —1 to «— O) • O) o -J oi 00 to o curve for to CJ1 CO o pupal o to o CO o OI stage o o> o O' o MEAN PERCENT DEVELOPMENT PER DAY (1 0 0 /y ) (1 0 0 /y ) PER DAY 70 DEVELOPMENT 60 50 40 PERCENT 30 20 MEAN MEAN DEVELOPMENTAL DURATION, y (DAYS) 79 5 10 15 20 25 30 TEMP. °C Figure 22. Temperature-velocity curve for four immature stages (1st and 2nd instar, prepupa and pupa) of Thrips tabaci. where £ is the developmental duration at a given temperature x. The constants: K, a and b were explained previously. The parameter K, the upper asymptote could be assessed from the graph by inspection but because the temperature ranges studied did not include this asymptote, the following formula developed by Davidson (1944) was used to calculate K: 2 2P!P2P 3 - P 2 (Pi+P3) K = Z (2) PlP3 " P2 where P x, P 2 and P 3 are values for 100/y on the curve at three equally spaced temperatures. The percent development per day (100/y) for the various temperatures was fitted to a straight line by linear regression (Table 9) and the Pi, P 2 and P 3 for three equally spaced temperatures, 20°, 23° and 26°C were obtained from the regression equation. Using Equation (2), K was calculated (Table 9). To determine the parameters a and b from Equation (1), the value of (K-P)/P for each temperature was calculated for the straight line portions of the sigmoid curve (which for our data included the whole range studied). Fitting Equation (1) to a straight line yielded the parameters a and b as follows: 100 y K 1 + e P a-bx K 1 + e and hence K-P _ a-bx — " e a-bx Table 10. Straight line equations for ln[(K-P)/P] and the Logistic equations. Straight line equations Logistic equation: Stage for ln[(K-P)/P] = a-bx 100 y 1st instar 1.67 -0.07 -0.96 100 = y 72.26 n e3.84-0*16x 2nd instar 1.44 -0.06 -0.91 100 y 54.98 ,, 3.31-0.15x l+e K 1+ea-bx Prepupa 1.89 -0.08 -0.97 100 y 120.50 i+e4.36-0.19x Pupa 1.70 -0.07 -0.99 100 y 68.24 1+e3.92-0.17x Total 1.66 -0.07 -0.98 100 y 18.30 i+e3.82-0.17x ~ 00 82 >* S o ^ w > < o a ui = -15.074 + 2.226x a H Z 40 Ui 2 a. O 30 -i Ui > UI o H Z UI 0 1 z < Ul 2 S 10 16 TEMP. 20 26 30 °C Figure 23. Lower temperature developmental threshold for first instar Thrips tabaci. 83 >• 8 100 10.188 + 1.759x H Z 1 30 ® H- 20 | Z UI 8 8! 5 10 15 TEMP. 20 25 30 °C Figure 24. Lower temperature developmental threshold for second instar Thrios tabaci. 84 = -38.086 + 4.271x a > o o ^ 100 < Q £C £ t- 80 z -j Ui > iz uj O a ui (L 60 40 Z < Ui 3 5 10 15 20 25 30 TEMP. °C Figure 25. Lower temperature developmental threshold for prepupal stage of Thrips tabaci. 85 60 = -18.915 + 2.306x o o > < 50 O DC ® H 40 Z Ui 2 a. 2UJ > UJ a H § o cc Ui 0. 20 z < UJ 2 5 10 20 15 TEMP. 25 30 °C Figure 26. Lower temperature developmental threshold for pupal stage of Thrips tabaci. 86 = -A.401 + 0.594x 14 > s >■ < o 10 CL tz 8 6 111 o s a. 3 UJ 2 4 2 5 10 15 20 25 30 TEMP. Figure 27. Lower temperature developmental threshold for immature stages of Thrips tabaci. 87 This is the form of the straight line, y - a-bx. The natural logarithms of (K-P)/P for the various values were fitted a straight line using linear regression (Table 10), from which the y-intercept and slopes were obtained. 4.3.2 Theoretical Lower Temperature Developmental Threshold of Thrips tabaci The theoretical lower developmental threshold is the minimum temperature at which any development will occur. This is the point of interception of the linearised temperature-velocity curve and the temperature-axis or the extra­ polated point of interception in this case. Table 11 gives the linear regression equation of the temperature-velocity curve which is of the form: 100 _ a_^x The lower developmental threshold is at the point y = 0 in the above equation. Figures 23, 24, 25, 26 and 27 show the graph for the different stages, and Table 11 shows, the calculated lower temperature development thresholds. 4.4 Developmental Biology of Coleomegilla maculata The average developmental period through the four larval stages was found to be 22 days (Table 12). This was higher than what was observed in the continuous feeding predation experiment (see section 4.6) where it was insured that the larvae never ran out of onion thrips. In the predation experiment, the four larval stages together took an average of 13.9 days. The longer developmental period in the former experiment may be due to the fact that the larvae of C. maculata did not get enough thrips to eat. Table 11. Theoretical lower temperature developmental thresholds, t, for various stages of Thrips tabaci. Straight line equations Stage 100 = a - bx y 1st instar 100 - -15.074 + 2.226x .98 6.77° 2nd instar = -10.188 + 1.759x .95 5.79° Prepupa = -38.086 + 4.271x .96 8.92° Pupa = -18.915 + 2.306x .98 8.20° .99 7.40° Total 100 - -4.401 + 0.594x r t°C Table 12. Developmental duration in days (mean + S.D) of Coleomegilla maculata when reared under different conditions. Temp. °C 24° 24° CM CM R.H. 50% 50% -- — Food Thrips tabaci T. tabaci Myzus persicae Therioaphis maculata Egg 4.5 ± 1.2 1st instar 4.8 ± 0.6 2nd instar 3.2 ± 0.1 2.8 3.83 ± 0.2 3.1 ± 0.1 3.0 3.9 ± 0.3 3.03 ± 0.6 2.0 ± 0.2 2.0 3rd instar 6.5 3.63 ± 0.3 2.5 ± 0.1 2.3 4th instar 6.8 ± 1.1 3.42 4.8 3.5 Total larval 22.0 13.91 Pupa 5.2 ± 0.8 — 3.7 ± 0.3 Egg to Adults 31.7 — 19.3 + 0.5 Reference This study (Developmental biology) ± 0.8 -- ± 1.2 This study (Predation experiments) ± 0.3 12.4 Obrycki and Tauber (1978) 10.7 6.7 17.4 Simpson and Burkhardt (1960) 90 Even though enough thrips were always present in the rearing containers, the thrips hid inside the hollow of the cut onion leaves and thus were not always accessible to the predators. The larval developmental period of 13.9 days (thrips-fed) is more comparable to 12.4 days obtained by Obrycki and Tauber (1978) using the green peach aphid, Myzus persicae (Sulzer) as host, and 10.7 days on the alfalfa aphid, Therioaphis maculata (Buck) (Simpson and Burkhardt, 1960) (Table 12). The experimental set up in both cases allowed the predator to easily obtain aphids coupled with the fact that the aphids are not very cryptic. The eggs from adult C. maculata that were fed onion thrips took longer, 4.5 days, to hatch compared to 3.2 days for those fed on green peach aphid (Obrycki and Tauber, 1978), and 2.8 days for those fed on the alfalfa aphid (Simpson and Burkhardt, 1960). No data from eggs of adult coccinellids that had been fed with thrips are available for comparison. Data was also not collected on the fecundity of the adult C. maculata. The observations on the developmental duration of C. maculata that had been fed onion thrips shows that C. maculata can survive and reproduce on onion thrips but the cryptic nature of the onion thrips may affect the amount of food the predator can obtain and thus slow down developmental rate and possibly reduce its fecundity. In the field, Thrips tabaci are cryptic, particularly when they occur at low densities, so that the developmental rate C. maculata under field conditions may be even slower if the predator is to survive on only the onion thrips. 91 4.5 Population Studies on Coccinellids 4.5.1 Coccinellid Complex at Eaton Rapids and Laingsburg, 1979 The following species of coccinellids were trapped at both Eaton Rapids and the MSU Muck Farm at Laingsburg: Coleomegilla maculata DeGeer, Adalia bipunctata (Linnaeus), Coccinella transversoguttata Faldermann, Hippodamia tredecimpunctata tibialis (Say), H. parenthesis (Say), and Cycloneda munda (Say). Species identification was done by Daniel Young of M.S.U. Entomology department. Pitfall, flight interception, and sticky board traps were used at Eaton Rapids. Muck Farm. Only sticky board traps were used at the A summary of the catch of the various species is shown in Table 13. Because the specimens trapped were not identified immediately, all the adults caught at the beginning of the season were grouped together as "unidenti­ fied coccinellids". C. maculata and H. tredecimpunctata were mistaken for two forms of the same species until late in the season when it was observed the proportion of the two "forms" had changed. A closer study of the two then confirmed that they were different species. at Eaton Rapids was: The order of abundance of the coccinellids C. maculata, H, tredecimpunctata, C. transversoguttata, with A. bipunctata, H. parenthesis and Cycloneda munda occurring in very low numbers. At the Muck Farm, C. maculata and H. tredecimpunctata were the most abundant, in that order, followed by A. bipunctata, C. munda, Table 13. Numbers of adult coccinellids caught by three different trapping methods at Eaton Rapids and the Muck Farm, 1979. Location Trapping Method CM+HT Eaton Rapids Pitfall (7/11 - 9/7/79) Flight Interception (6/25 - 9/7/79) Muck Farm CM = HT = AB = CT = HP = CMU= AB CT HP CMU Undeterm. TOTAL 66 0 11 3 0 25 105 41 1 1 0 0 5 48 Stickyboard (6/20 - 9/7/79) 136 4 25 4 3 516 688 Total 243 5 37 7 3 546 841 % Total Det. Sp. 82% 2% 13% 2% 1% Sticky board 653 139 34 12 74 0 912 % Total Det. Sp. 72% 15% 4% 1% 8% Coleomegilla maculata Hippodamia Tredecimpunctata Adalia bipunctata Coccinella transversoguttata Hippodamia parenthesis Cycloneda munda 93 C. transversogut tata and H. parenthesis. Besides occurring in low numbers and together accounting for 18% of the coccinellids trapped at Eaton Rapids and 28% at the Muck Farm, A. bipunctata, C. transversoguttata, H. parenthesis and C. munda appeared to prefer the grass and weed borders rather than the onion plants. Further studies will be needed to determine whether their low occurrence within the onion fields is because of the ab­ sence of the preferred food host or whether this environment is unsuitable for them. With regards to numbers and dis­ tribution, C. maculata and H. tredecimpunctata were considered as the two important coccinellid species in both Eaton Rapids and the Muck Farm. To compare numbers occurring at the two locations studied, only the sticky board trap method was used since this was the only method common to both locations. The plot at the Muck Farm was designed not to receive any pesticide applications and all weeds were controlled mechanically. Unfortunately, data was collected here only up to the end of July 1979, when the plot was overrun by weeds. Even though the sticky board traps were not checked on the same days at both locations, the data from Eaton Rapids was grouped to obtain similar trapping dates and durations (taken from Table 13) as at the Muck Farm. The catch for each period was converted to catch X 100 traps (Table 14). X day Analysis of variance showed no significant difference between the two locations (F = 5.548, p>0.05). Table 14. Comparison of the total stickyboard trap catch of coccinellids (catch x 100 traps-* x day-*) for similar periods and trapping durations at Eaton Rapids and thr Muck Farm, 1979. Date EATON RAPIDS Trapping # of Catch x lOOtrapsDays Traps x day-l Date Trapping Days MOCK FARM # of Catch x lOOtrapsTraps % day-l 7-06-79 7 29 196 7-04-79 9 16 72 7-12-79 6 29 189 7-11-79 7 16 125 7-20-79 8 9 269 7-20-79 9 16 122 7-28-79 8 9 97 7-27-79 7 16 71 95 4.5.2 Vertical Distribution of Adult Coccinellids The vertical flight distribution of the adult coccinellids was studied using sticky board traps. These were the same traps used for the population studies and were marked into 30 cm segments along their heights. Muck Farm, 1979: only at the Muck Farm. In 1979, this study was conducted The traps were located within and along the borders of the onion plot. once a week in the month of July, Trapping was done The study period was short enough to allow the assumption that no appreciable increase in onion plant height occurred during this period. In general, onion plants do not grow tall enough during the season to grossly alter the flight behavior of any adult coccinellids that prefer to fly above the plant tops. This assumption becomes very crucial if the traps are located among plants whose height increase appreciably during the growing season. Only the three abundant species, C. maculata, H. tredecimpunctata and A. bipunctata were used, the first two being grouped together because of misidentification. The number of adults occurring at the four heights were converted to their percentage of the catch (Appendix Table 3), thus emphasizing the proportion of the species at that height even though the procedure overemphasized low catches. This was necessary because the number of traps dropped from 29 to 9 after the second monitoring owing to a shortage of tangle­ foot. These results are shown in Figure 28, A comparison of the percentage occurrence of the adults at the different 96 (a) C. maculata HI + H. tredecimpunctata O 60 z 92 HI CC O O 116 40 22 80 81 o 15 49 20 47 10 31 27 111 11 3_ 12 1 I I 7 -0 6 •*i 7 -1 2 7 -2 0 7 -2 8 -7 9 (b) A. bipunctata Hi O Z Ui 40 QC CC 1 1 36 25 14 14 D 2 2 2 Y/s » i O O fTT o 2010 I 0 0 j 7 -0 6 7 -1 2 7 -2 0 7 -2 8 -7 9 Figure 28 Vertical distribution of coccinellids on stickyboard traps, in onions through the season, at the M.S.II. Muck Farm, 1979 (a) Coleomegilla maculata plus Hippodamia tredecimpunctata, (b) Adalia bipunctata. 97 heights (one-way analysis of variance) considering the four sample dates as replicates, showed there was a significant difference (F = 34.09, p * .01) in the vertical distribution of C. maculata and H. tredecimpunctata but not for A. bipunctata. C. maculata and H. tredec impunc ta ta together show a strong preference for a flight altitude of 0.3 - 0.6 m and 0 - 0.3 m in that order and little preference for altitudes above 0,9 m (Table 15). Muck Farm, 1980: This study was repeated in 1980 at the Muck Farm in a field of sorghum and at Eaton Rapids in fields of winter wheat and spring oat. Records were taken for the same three species as the year before. At the Muck Farm, 15 traps were used and trapping was done for a five week period, July 1 to August 7. In general, the numbers of C. maculata were higher at the beginning of the assessment period and declined thereafter (Appendix Table 4). An analysis of variance of the vertical distribution of C. maculata when the five sampling dates are considered as replicates (i.e., assuming no drastic changes in plant height), yielded no significant difference in the preference to the height segments (Table 16). However, a closer look at the data, showed that a high proportion of adults, 63%, preferred the lower 0 - 0.3 m height early in the season and very few occurred at the higher levels (0.9 - 1.2 m) (Appendix Table 4). This trend was followed through the next three weeks although there was a gradual shift in proportion towards the middle levels (0.3 - 0.9 m). By the final week of July and the Table 15. Summary of vertical distribution of coccinellids caught on stickyboard traps located in onions at the Muck Farm, Laingsburg, 1979. AB Height (m) Mixed Pop. of CM & HT Mean 7„ f-test Mean 0 - .3 29.75a 6.50 .3 - .6 45.50b 27.25 .6 - .9 19.75c 36.25 .9 - 1.2 5. OOd 29.75 34.09** % f-test 3.419 n.s. **f-test significant at p = 0.01. Means followed by the same letter are not significantly different on Student-Newman-Keuls multiple range test, p = 0.05. Table 16. Summary of vertical distribution of coccinellids on 15 stickyboard traps located in spring sorghum at the Muck Farm, Laingsburg, 1980. Height (m) Coleomegilla maculata Mean % F-test 0 - .3 35.40a .3 - .6 29.20a 52.40c 9.60a .6 - .9 20.20a 27.20b 53.20b .9 - 1.2 15.00a 15.80ab 21.20a 1.49 n.s. Hippodamia tredecempunctata Mean 7„ F-test 4.60a 11.831** Adalia bipunctata Mean 7a F-test 12.00a 5.629** *F-test significant at p = 0.05. **F-test significant at p = 0.01. Means followed by the same letter are not significantly different on Student-Newman-Keuls multiple range test, p = .05. 100 Figure 29. Vertical distribution of coccinellids on stickyboard traps, in sorghum through the season, at the M.S.U. Muck Farm, 1980 (a) Coleomegilla maculata, (b) Hippodamia tredecimpunctata^ (c) AaalTa bipunctata. 101 30 (a) C. maculata 60 40 % OCCURRENCE rrri 20 •• 3 :: •• •• •• •• •• • #i (b) H. tradaolmpunctata 60 2 40 rrri •• •• 7 -1 1 7 -0 6 7 -2 5 7 -3 1 8 - 0 7 -8 0 1 100 % OCCURRENCE (e) A. bipunctata 2 i m 60 l 40 § :: i " 20 .0 .0 7 -0 6 Figure 29. 7-11 7-25 L°J , o .o 7-31 I 8 - 0 7 -8 0 102 11 S CM HT AB Figure 30. Vertical distribution of coccinellids on stickyboard traps, in sorghum, at the M.S.U. Muck Farm, 1980 (all trapping dates together). 103 first week of August, the weight of the distribution was heavy at the middle and upper levels and very weak at the lowest level (Fig. 29). It appears then that the assumption about the data being of the "same population" over the 5 week period which allowed these dates to be treated as replicates was not valid. A separate ANOVA for the first 3 weeks showed there was a strong preference for the lowest levels (0 - 0.3 m ) , (F = 30.83, p = 0.01), compared to the other height segments (Table 17). However, for the last 2 sampling dates, C. maculata avoided the lower levels and preferred the upper levels (F = 7.92, p = 0.01) (Table 17). This shows that there was a change in flight height as the season progressed, attri­ butable to the increase in height of the sorghum plants. The sorghum was planted in late June and, by the fourth week of July, was about 0.6 - 0.8 m tall. Adult C. maculata were thus flying just above the plants and altering their elevations as the plants grew taller. Over the five weeks period, larger numbers of C. maculata occurred at 0 - 0.3 m (Fig. 30). Compared to C. maculata, fewer adults of H. tredecimpunctata were trapped at the beginning of the sampling period but the number increased towards the end (Fig. 29). This confirms the observation in 1979 that the numbers of H. tredecimpunctata increased and they became the predominant coccinellid species towards the end of the growing season. Over the 5 weeks sampling period, the proportion of H. tredecimpunctata occur­ ring at the middle level was significantly higher (F =* 11.83, P = .01) Table 17. Vertical distribution of C. maculata in sorghum during the periods 7-06-80 to 7-25-80 and 7-31-80 to 8-07-80 at the Muck Farm. 7-25-80 to 8-07-80 Mean % F-test Height (m) 7-06-80 to 7-25-80 Mean F-test 0 - .3 54.67a .3 - .6 26.00b 34.50b .6 - .9 15.33b 27.50b 4.00b 32.50b .9 - 1.2 30.83** 6.50a 7.92** **Significant at p = 0.01. Means followed by the same letter are not significantly different on Student-Newman-Keuls multiple range test, p = 0.01. 105 Figure 31. Vertical distribution of coccinellids on stickyboard traps, in winter wheat (July 2 through July 23, 1980) at Eaton Rapids (a) C. maculata, (b) H. tredec impuncta ta, (c) A. bipunctata. I 106 (a) C. maculata 30 60 9 9 40 12 12 i 111 O zLU 13 20 2 2 GC CC I Jim, 1£20U. o (b) H. tradacfmpuictata 40 20 I % OCCURRENCE IO O -U 0 O o) m 77-02 -0 2 7 -1 0 2 2 7 - 2 3 -8 0 » (e) • » 7 -0 2 •• •• •m •* •• •• •• •* •• A. bipunctata 1 •9 •* •• •• •• +* 0. 0 Figure 31. 7 -1 7 . • • • 9 • • • • ♦ • • • 00 7- 1 0 : 0 0 0 0 0 0 0 0 7 -1 7 7 - 2 3 -8 0 107 UJ o Z 60 Lil 54 QC CC g 22 40 o o 20 15 CM HT AB Figure 32. Vertical distribution of coccinellids on stickyboard traps, in winter wheat, at Eaton Rapids, 1980 (all trapping dates together). 108 than the lowest and highest levels (Appendix Table 4 and Table 16). The preferred flight height for H. tredecimpunctata then was 0.3 - 0.6 m (Fig. 30) and did not alter much during the five weeks observation period even though the sorghum plants increased in height to 0.6 - 0.8 m. Adalia bipunctata occurred in very low numbers compared to the previous year and, even though they seemed to prefer flying at 0.6 - 0.9 m, the generally low numbers makes this a questionnable conclusion (Appendix Table 4, Figs. 29 and 30). Eaton Rapids, 1980 — Winter Wheat: Adults of both C. maculata and H. tredecimpunctata occurred in fairly good numbers during the period of trapping, July 2, 1980 to July 23, 1980, with C. maculata being almost twice as abundant as H. tredecimpunctata and again A, bipunctata being rare (Figures 31 and 32) (Appendix Table 5). Analysis of variance of the vertical distribution showed C. maculata strongly preferred the lowest level, 0 - 0.3 m, but H. tredecimpunctata and A. bipunctata did not show any preference (Table 18). However, the raw data seems to indicate H. tredecimpunctata preferred the lower two levels whereas the six A. bipunctata trapped preferred the upper level. C. maculata in winter wheat did not show the change in flight height that was ob­ served at the Muck Farm in sorghum. This difference might be explained as follows -- the winter wheat was present in the field before the overwintered adult coccinellids migrated into the field and thus served as the host plant Table 18. Summary of vertical distribution of coccinellids caught on 10 stickyboard traps located in winter wheat at Eaton Rapids, 1980. Height (m) Coleomegilla maculata Mean % F-test 10.302** Hippoclamia tredecempunctata % F-test 31.50a 2.571 n . s . Mean Adalia bipunctata Mean % 0a 0 - .3 46.75 .3 - .6 29.50b 36.75a 0a .6 - .9 17.75ab 17.70a 33.00a 6.00a 14.25a 17.00a .9 - 1.2 F-test 2.186 n.s. **F-test significant at p = 0.01. Means followed by the same letter are not significantly different on Student-Newman-Keuls multiple range test, p = 0.05. Table 19. Summary of vertical distribution of coccinellids caught on 11 stickyboard traps located in or adjacent to spring oat at Eaton Rapids, 1980. Coleomegilla maculata Hippodamia tredecempunctata Mean % F-test Mean 7a F-test Mean % F-test 0 - .3 26.50ab 5.010* 23.50ab 4.027* 2.25a 6.159* .3 - .6 44.75b 42.25b 11.75ab .6 - .9 21.OOab 29.OOab 30.50ab .9 - 1.2 7.75a 5.25a Height (m) Adalia bipunctata 55.50b *F-test significant at p = .05. Mean followed by the same letter are not significantly different on Student-Newman-Keuls multiple range test, p = 0.05. Ill Figure 33. Vertical distribution of coccinellids on stickyboard traps, in spring oat (July 2 through July 23, 1980) at Eaton Rapids, (a) C. maculata, (b) H. tredecimpunctata, (c) A. bipunctata. 112 (a) C. maculata 60 40 UJ O 20 3 UJ OC •• I oc 60 26 40 20 20 2 UJ O (c) A. bipunctata * UJ oc o o o 2 TT 40 20 1 * 1 I u 7 -0 2 Figure 33. 7 -1 0 7 -1 7 7 - 2 3 -8 0 113 CM HT AB Figure 34. Vertical distribution of coccinellids on stickyboard traps, in spring oat, at Eaton Rapids, 1980 (all trapping dates together). 114 on which the various arthropod prey and later the coccinellids became established. By the time the trapping was started the wheat was already about 0.4 - 0.5 m tall and did not increase in height much for the rest of the season. It harboured a rather resident population that did not migrate much but stayed within the height (0 - 0.6 m) of the plants. Most adult coccinellids counted during quadrant sampling (see section 4.5.3.2) were actually found on the soil or near the bases of the wheat plants. Eaton Rapids, 1980 -- Spring Oa t : More H. tredecimpunctata were trapped in the spring oat than any other coccinellid species (Table 19 and Appendix Table 6). C. maculata and H. tredecimpunctata preferred the lower three levels (0 - 0. 9 m ) with the highest proportion occurring at 0.3 - 0.6 m (p = 0.05). A. bipunctata on the other hand preferred the highest level 0.9 m and above. These results are shown in Figures 33 and 34. In general, the flight heights of the adult coccinellids were altered as the height of the surrounding vegetation in­ creased. C. maculata preferred the flight height of up to 0.3 m when the surrounding plants were shorter than this height, while H. tredecimpunctata preferred the flight level of 0.3 - 0.6 m above the ground. The low numbers of A. bipunctata did not permit any firm conclusions to be reached on the flight height of this species. 4.5.3 4.5.3.1 Population Dynamics of Coccinellids Population Index Using Behavioral Traps, 1979 The three trapping methods used to determine the species of coccinellids that occurred in the study areas, Sec. 4.5.1, 115 also served as a means of studying the population dynamics through the season. Since the trapping durations and number of traps were not the same for the three methods, the catch was converted into catch X trap X day which gives a common basis for comparing the methods without implying equivalency of the trapping ability of the traps. Because the numbers thus obtained were very low, they were converted to catch X 100 traps X day The sticky board traps caught the highest numbers of adult coccinellids and appears to be the most efficient of the three methods used. Care however must be used in interpreting the results of trap catches as the ability to trap the adults is dependent on environmental factors such as temperature and rainfall (affecting activity and the chances of moving into the trap vicinity) and by the behavior of the adults of the different species. The pitfall and flight interception traps are passive and are more likely to indicate adult activity. The bright yellow stickboard traps on the other hand can actively attract adults and hence interspecific and intra­ specific (e.g., age and sex) differences in responding to the yellow color become important. The dynamics of the adult populations of all the species of coccinellids as indicated by the three trapping methods at Eaton Rapids, 1979, are shown in Figure 35 (Appendix Tables 7, 8 and 9) for stickyboard, flight interception, and pitfall traps, respectively. The stickyboard trap method showed a distinct major peak in the numbers in the 116 (a) STICKYBOARD TRAPS 150 100 - I CO OL < OC 50 H O O h [ JUn 5 J U L Y 1 AUGUST 1 SEPT 50 O JUNE' 50 n n JD JULY (b) FLIGHT INTERCEPTION TRAPS AUGUST 1 SEPT (c) 20 I JUNE PITFALL TRAPS n~n~n— n~u~TTTij— ^.,rfl I 10 20 I I 20 20 10a 20 JULY 10 20 AUGUST 2 SEPT Figure 35. Population trend of coccinellids at Eaton Rapids as indexed by different trapping methods, 1979 (a) stickyboard trap, (b) flight interception trap, (c) pitfall trap. 117 « 250 < Q X ^ i 200 CO | 150 O o 100 X o H < o 50 tO 20 JULY Figure 36. Population trend of coccinellids at the M.S.U. Muck Farm using stickyboard traps, July, 1979. 118 middle of July and a possible minor peak in August. The flight interception traps recorded extremely low numbers until August, when a weak peak in numbers seemed to occur (Fig. 35b). This weak peak may be confirming the existence of a second population peak as was observed with the sticky­ board traps. It is not clear why the flight interception traps recorded such low numbers at the beginning of the season when many more adult coccinellids were observed in the field at this time. The pitfall traps appeared to catch about the same low numbers through the season (Fig. 35c). At the Muck Farm, only stickyboard traps were used and for only the month of July (Appendix Table 10 and Fig. 36). This trapping period was too short to indicate any trend in the population dynamics. Since the adult coccinellids are not ground-dwellers, although they have been observed to walk across the ground to get to adjacent plants when they fall, very few of them occurred on the ground at one time between the onion plants, explaining the low and constant numbers caught in the pit­ fall traps. The virtual absence of adults in the flight interception traps in warmer periods of the season might be due to the fact that the adults were active enough at that time to fly out again after landing at the open ends of these traps. As the temperature dropped in August, the adults were less active and did not fly away that readily but walked up the nylon mesh of the traps and were caught at the corners. A high activity in the early part of the 119 season placed the adults in the vicinity of the stickyboard traps and increased the likelihood of landing on the boards. Since the surfaces were coated with Tanglefoot R , the adults got trapped and could not get away. 4.5.3.2 Population Densities of Coccinellids at Eaton Rapids Early Season: The coccinellids overwinter as adults and they are present in the fields in the early spring, long before the onion crop is even planted. During the early part of the 1980 season, in mid-May, adult coccinellids were present in the onion fields and surrounding oat and wheat fields at a time when the onions were only 2-3 weeks old, and about 5-8 cm tall. the onion plants but The coccinellids did not occur on rather ongiant ragweed, Ambrosia trifida, that grew particularly in the inter-row spaces. In the spring oat adjacent to the onions, the coccinellids again occurred on the ragweed. The giant ragweed was up to 60 cm tall by May 24, 1980. A visual count made on all ragweed occurring in the first nine rows of onions and ad­ jacent spring oat (Table 20) showed C. maculata was again the predominant species. H. tredecimpunctata occurred in moderate numbers but few C. transversogotata were found. Coccinella novemnotata Herbst, recorded for the first time during this study, occurred in low numbers. The ragweed evidently came from a permanent boarder on the west-side of the field that was almost entirely made up of ragweed. The density of this weed in the field was highest in the spring Table 20. Occurrence of adult coccinellids on giant ragweed, Ambrosia trifida, growing among onions and in an adjacent oat plot. Source N o . of Ragweeds Coccinellid Species CM HT CT CN TOTAL Eaton Rapids, May 24, 1980. No. Coccinellids per Ragweed Coccinellid Density (#/m2) In Oats 235 27 10 0 0 37 0.16 0.026 Rows 1-3 of Onions 170 88 15 7 3 113 0.66 0.116 Rows 4-9 of Onions 125 56 23 11 9 99 0.79 0.051 TOTAL AREA 530 171 48 18 12 249 0.47 0.11 121 oat that was next to the weed border and decreased away from the source. A total of 235 ragweed was counted in the oat and enough onion rows (or rather inter-row spaces) were surveyed to provide a similar number of ragweed for comparison. The first 9 row-area of onions provided 295 ragweed and this occupied an area of about 3000 m 2 compared to 1440 m 2 of oat. Thus the density of ragweed in the spring oat was twice that in the first nine row-area of onions. The density of ragweed decreased sharply for distances further away from the ragweed border. The number of coccinellids per ragweed increased from the spring oat into the field (Figure 37a). contrary to what was expected. outside the oat This was Since the adults overwintered and onion fields, because this area was completely bare over the winter, it would have been expected that the numbers of coccinellids would have been higher near the border - the possible route of immigrating into the field. The adults probably did not come from the ragweed bordering i the field. The density of coccinellids was highest for the first 3 row-area of onions adjacent to the oat, twice the density for the next 6 row area of onions and about 5 times the density in the oat (Figure 37b). occurred in remaining onion area. Very few coccinellids It is not clear why the coccinellids occurred in this distribution pattern but it is understandable that they occurred on the giant ragweed at this time since these plants provided them with shelter and a host of arthropod prey. The early season occurrence of the coccinellids in the oat and onion fields cannot be 122 Figure 37. Occurrence of coccinellids on ragweed in spring oat and onion fields (a) number per ragweed, (b) density (number per unit ground area). -1 * ADULT COCCINELLIDS x RAGWEED j j O o p o ^ N> -U O) O O 00 “T > ° CD Figure 37 * ADULT COCCINELLIDS x M O > H § G> O o 00 T™ p o o M Oft 0.116 “ o o 0-121 P 2 P o ro -2 124 attributed to either of these crops since they were too small at this time to harbor any prey for the coccinellids. Further, when the ragweed were removed from the onion field, the coccinellids virtually disappeared from this area but continued to remain in the oat since the ragweed here were not removed. Within season densities: As stated above, the early season population of coccinellids was present early in the spring oat because there were ragweed present that provided food and shelter. As the oat grew, they became the food host of herbivorous arthropods upon which the coccinellids fed. The density of coccinellids within this area thus increased as shown by quadrat sampling that was started on June 6, 1980 (Appendix Table 11). oat peaked in mid-July at The density in spring 2 2.4coccinellids/m (Figure 38). The densities of coccinellids in winter wheat were much 2 higher than in spring oat and probably peaked (4 coccinellids/m ) in the former in the middle of June (Figure 38). High numbers of coccinellids were observed in the winter wheat at the beginning of the season when only few occurred in the spring oats. It appears that the coccinellids became established in the winter wheat earlyin the provided food and shelter season when these plants but as they matured and dried up towards mid-July they harboured fewer prey and hence fewer coccinellids. The density on onions, based on numbers visually counted per 9 x 150 m row of onions, was extremely low until early August when it started rising. More data is 125 7.55 « Onion • Wheat o Oat □ Alfalfa > t 0.05). Table 27. Feeding rate of Thrips tabaci on onion leaf segments. Thrips Stage Mean Area Fed x Thrips-^ x day"^" 2nd instars 4.51 (S.E. = 0.39) Adults 4.92 (S.E. = 0.36) 149 effects were observed. Thus, providing a larger leaf area per thrips did not consistently alter the feeding rate. In the next experiment, known numbers of adult thrips were enclosed on whole onion plants. The preliminary experi­ ments, using young plants of about 20-30 days old (2-3 leafstage), twenty thrips completely destroyed and killed the plants in 4-6 days, showing the potential damage that can be caused by the thrips. In the actual experiment, using older plants whose senile leaves had been trimmed away, the feeding rate of 4.92 sq. mm x thrips x day was used to estimate the expected number of days a given number of adults will take to consume a known proportion (e.g. 25 or 50%) of the leaf area provided. By stopping the deter­ mination when this proportion of leaf area had been damaged, the experiments would have spanned a shorter period. However, it was not possible to tell with any degree of certainty when 25% or 50% of the leaf area had been damaged owing to the fact that fresh lesion were difficult to observe. The experiments were thus continued to 100% leaf area destruction. In the leaf segment experiments, observing the feeding lesions was not a problem because the leaf segments provided were inspected under binocular microscope and fresh lesions were easily spotted as "watery" green patches. The results of this experiment showed that the adult thrips took signifi­ cantly longer (x = 40% longer) to feed on the leaf area pro­ vided compared to what was predicted (Table 28). for the disparity can be due to the following: The reason Table 28. # of Leaves Feeding duration of adult Thrips tabacl for destroying plant surface area. Total leaf area (sq.mm) // of adult thrips Days to destroy leaf area Exp.1 Obs. , X2 - (°~E) Ed 3 1947.42 81 5 8 1.8 4 2110.64 40 11 9 .36 3 1635.10 64 5 9 3.2 2 603.19 462 3 5 1.3 3 3389.78 78 9 9 0 3 1291.19 116 2 9 24.5 2 D T - 31.2 ^Expected values, to the nearest day, estimated from adult feeding rate studies. 2 About 50 adults died at the start of the experiment. 151 1. Using the surface area measured at the beginning of the experiment, to calculate the expected time for destroying the total surface area, assumes that the plant does not grow or increase in surface area over the period of determination. This assumption is more acceptable if the experiment is done over 1 or 2 days but is certainly wrong for the durations of these experiments since in 5 days or more the plants would have increased in surface area. In fact, new leaves were observed in some plants apart from increase in length of the initial leaves. It was not possible to measure the surface area at the end of the experiment because most leaves had withered by then. The increase in surface area means the thrips will take a longer period than expected. 2. It was assumed that no thrips deaths occurred after the beginning of the experiment. All thrips that survived the anaesthesia and handling at the onset were presumed to survive the whole period. It was not possible to cross­ check the number alive and feeding daily without creating channels for escape since the only method would have required removing the plastic covers and counting the thrips. Also, the thrips were collected from the general thrips culture and hence of different ages and thus some of those used might have died naturally before the termination of the experiments. The design of the whole onion plant experiment could be improved by first estimating the longevity of freshly emerged adults and running the experiments for periods when 152 most of them will still be alive. This will also reduce the variability of rates of feeding of adults of different ages. As stated above, overcrowding of thrips in the leaf segment experiment did not affect the rate of feeding and no overcrowding occurred at all in the whole plant experiments since large leaf surface areas were available. In the field, onion thrips were usually observed to cluster together be­ tween closely touching leaf surfaces particularly the bases of the leaves. The result of such clustering is that larger feeding lesions occur at such sites although feeding lesions can be found all over the general surface of the plant. The cryptic and gregarious nature of the thrips make them prefer plants whose morphology provide them (thrips) with narrow and concealed spaces into which they can hide. This condition is satisfied by onion plants or varieties whose leaf bases are in close contact making a deep "V" rather than a wide "V"; whose newest leaves develop while the two preceeding ones have not yet separated; whose entire leaf surface are not smooth and uniform but develop small troughs along the inner surfaces and finally, those with long flabby leaves that bend over sharply and lie on each other. It is suggested that such plants provide suitable hiding places for thrips where their numbers may increase leading to severe plant damage. These morphological and growth characteristics could be useful for developing thrips-resistant onion varieties. 5. GENERAL DISCUSSION Damage done by the onion thrips is primarily through rasping on the leaf surface and sucking up the contents of the plant tissue. This feeding causes the plant to lose moisture, nutrients and chlorophyll. The loss of chlorophyll reduces the plant's photosynthetic potential but is not obvious until later when the lesions dry up and appear as white blotches. Information on the rate of feeding on onion leaf tissue (or reduction in photosynthetic potential), forms the basis for assessing the physical damage that can be caused by the thrips and establishing the economic threshold. The economic threshold concept is very important in pest management and is critical for making decision on re­ source allocation and utilization for the purpose of controlling the pest. Economic threshold has been defined variously by different authors. Stern et al (1959) defined it as the "pest density at which control measures should be applied to prevent an increasing pest population from reaching the economic injury level". From microeconomic theory, the maximum profit of using an input, e.g. an insecticide, is obtained when "all things being equal", the marginal revenue of using that input equals the marginal cost of that input (Ferguson, 1969). The marginal revenue (MR) and marginal cost (MC) are the incremental changes in the total revenue (TR) and the total cost (TC) respectively. That is, 154 where Q is the output or total yield. as TR - TC. Profit is defined Since profit is maximized at the point where MR = MC, the pest density associated with this point of equality, defines the economic threshold. Higher pest density will decrease the total revenue and hence decrease the profit. Similarly, a high total cost through too many pesticide applications will also decrease the profit margin. The control agent (pesticide input) should only be used when the marginal revenue it produces is equal to or greater than the marginal cost of using that input. Rabb (1972) suggested a method of developing the economic threshold, of which the first important ingredient was studied in his project. The Rabb method involves collecting the following data: 1. Amount of physical damage related to various pest densities. 2. Crop yield and monetary value (TR) and production cost (TC) including cost of control at various levels of physical damage. 3. Amount of physical damage that can be prevented by the control measure. 4. Monetary value of the portion of the crop that can be saved by the control measure. 5. Monetary cost of the control measure. Since thrips damage is done to the leaves and not the onion bulb directly, the damage (or feeding rate) needs to be tied in with knowledge of onion leaf surface area development 155 through the season. The leaf area development, in turn, needs to be related to onion bulb formation through the season, in order to obtain the amount of physical damage caused. Data on he production of leaf tissue (Figure 43 and Appendix Table 20) has been compiled by another project that was to develop an onion plant model (Bolgiano, 1980). The plant model project has accumulated data on leaf tissue production through the season over several years but has not as yet established the all important link with bulb forma­ tion. Typically, the graph for leaf area production over the growing season, has a sigmoid shape. It is hypothesized here that, in the early phase of the growth of the plant, virtually all energy excess of metabolism is channeled into the production of above ground leaf tissue. After a period, above ground leaf production slows down and the plant starts storing the excess energy in the form of below ground leaf tissue i.e. the bulb. It is further hypothesized that the point at which the above ground leaf production stops increasing at an increasing rate, is the critical point at which bulb formation begins or very soon thereafter. The yield or weight of the bulb will, among other things, depend on the total leaf area that the plant is able to produce throughout its life (since this area is proportional to the total photosynthetic potential); the time it takes to reach this maximum potential; the slope at the critical point and the time it takes to attain this slope. Integrating the area under the leaf area/time curve (Figure 43), gives a measure of how much leaf area is produced and maintained 156 TOTAL LEAF SURFACE AREA (C M 400 300 200 100 20 40 60 AGE Figure 43. 80 100 120 140 (D A Y S ) Onion leaf production through the season. 157 through the season, and thus the potential photosynthate that the plant can produce. The area under this curve from the "critical point" will be directly proportional to the size of onion bulb produced. Thrips build-up is essentially exponential in form (baring any sudden drop due to environmental conditions like heavy rainfall) up to a few weeks before harvest. Knowing the amount of leaf area that can be destroyed per thrips per day, the size of the thrips population at any time can be converted to leaf area that can be destroyed per day. To develop the economic threshold, the precise effect of a given number of thrips at the various stages of the plant's growth needs to be estimated. This information is not available at this time but can be obtained from experiments in which onion plants at various stages of growth are subjected to different thrips-days of feeding. For each set of experiment, half the plants are harvested after a known period to determine the mean area damaged, while the remaining plants are relieved of the thrips pressure and allowed to grow to maturity and the weight of bulbs measured. The first half of each experiment needs to be ran for a fairly short period, e.g. two days, during which time it can be assumed the plant never changed in surface area. Even though the precise relationship between the numbers of thrips and reduction in bulb weight is not known, it is quite unlikely that there is only one damage threshold level 158 for thrips throughout the season. Judging from the sigmoid pattern of leaf surface area development, it is likely that the early phase, when leaf area is small and increase is slow, will be more sensitive to loss in photosynthetic area. Thus a lower percentage loss of photosynthetic area will be desirable. For example, if a 5% leaf area loss is acceptable in the middle of the season, the acceptable level of loss at the beginning of the season may be lower. Using hypotheti­ cal levels of damage and the estimated rate of feeding of larval thrips (since this is the predominant stage through the season), and allowing the feeding pressure to last for 7 days a "scenerio" of expected damage and the corresponding density of thrips can be developed. A hypothetical thrips density function of 0.79 thrips/plant at 20 days old, 3.17 thrips at 30 days, 34.9 thrips at 60 days and 60.3 thrips at 90 days old was estimated based on a photosynthetic area loss (Table 29). 5% acceptable A 7-day turn-around time was used because this is a convenient frequency at which fields can be visited although more frequent visits can be made particularly at the more sensitive periods in the plant's development and also under conditions when thrips numbers are building up very fast. prolonged dry and warm periods. The latter occurs under The formula used for develop­ ing the above thrips density function is: Thrips Density = A x D 0.045 x T 2 where A = total onion leaf surface area (cm ) Table 29. Hypothetical Thrips densities through the season that will cause certain percentage onion leaf area loss. Age of Onions (wks) Approx. # of Leaves Hypothetical Thrips Densities (# Thrips/plant) (Percent Acceptable Leaf Area Loss) 1% 5% 107o 15% 20% 2 .14 .70 1.40 2.10 2.79 4 4 .63 3.17 6.35 9.52 5 5 1.11 5.56 11.11 16.67 12.70 22.22 6 6 1.78 8.89 17.78 26.67 35.56 7 7 3.33 16.67 33.33 50.00 66.67 8 8 5.08 25.40 50.79 76.19 101.59 9 7.78 38.89 77.78 116.67 155.56 10 9 12 9.37 93.65 140.48 187.30 11 11 10.95 46.83 54.76 109.52 164.29 12 11 58.73 117.46 176.19 13 11 11.75 12.22 219.50 234.92 61.11 122.22 183.33 244.44 14 9 12.70 63.49 126.98 190.48 253.97 15 9 13.17 65.87 131.75 197.62 263.49 159 3 160 D = acceptable damage as a fraction of total surface area T = turnaround time or duration of thrips feeding pressure in days. 2 The amount of feeding per larval thrips per day is 0.045 cm . If the area used in this calculation process is the initial area, the plant would have increased in surface area by the end of the time period, T. This calculation can be improved by developing a logistic equation (similar to that developed for thrips temperature-velocity curve) and using this equation to predict the surface area by the end of period T, and thus the surface area at the "mid-point" of period T, i.e. one-half of that surface area is used for estimating the thrips density or action threshold. The "scenerio" for a constant 5%, 10% and 15% leaf area loss (LAL) through the season, provided corresponding thrips density, functions (Figure 44). Since the onion leaf area development through the season had a sigmoid shape, the hypothetical constant LAL (of 5, 10 or 15%) yielded sigmoid functions. Field data from Eaton Rapids (1980) and the Muck Farm (1980) were then included for comparison. Both locations ran at 2 - 3% LAL up to day 50, after which the thrips density at Eaton Rapids started to rise faster. Thus by day 70, the Muck Farm plants were about 3% LAL level and those at Eaton Rapids were at 4.5% LAL level. At around days 80-82, the Muck Farm plants were losing about 5% of their leaf area to thrips and those at Eaton Rapids were losing over 10%. No loss in bulb weight was observed at the Muck Farm 161 1981 Muck Farm - 200 n o , Y ie ld lo s s 1980 Eaton R apids- 15% LAL Visid L oss? 180 160 140 10% LAL Z 120 100 80 5% LAL 60 40 20 40 60 AGE OF PLANT (DAYS) Figure 44. Comparison of field recorded thrips density levels (broken lines) with hypothetical thrips densities that assume constant 5%, 105, and 155 leaf area loss (LAL) through the season. 162 (Table 6) although the mean number of thrips from the un­ treated plots (data used here) were at times significantly different from chemically treated plots. No accurate yield data was available from Eaton Rapids, although some loss was suspected due to high thrips infestation. The thrips density at Eaton Rapids may thus be around the damage threshold. The grower at Eaton Rapids reported that his yearly yield was always lower than the State of Michigan average but this may be due to a combination of many factors. Using the data from Eaton Rapids (1980), as a density function that will cause unacceptable damage or at least close to the damage threshold, a sliding hypothetical action threshold can be developed that will fall below this function but above the Muck Farm (1980) function (except for the end of the season when the latter function appeared high). A suggested hypothetical action threshold will be as follows: 0.05 Thrips/plant (or 0.005 LAL) at 5 weeks; 5.00 Thrips/plant (or 0.01 LAL) at8 weeks; 29.00 Thrips/plant (or 0.03 LAL) at 10 weeks, and, 59.00 Thrips/plant (or 0.05 LAL) at 12 weeks. The potential controlling effects of the predator, C. maculata, could be tied in if information on how many thrips. each predator consumes per unit time under field conditions is known. This field predation is not known now but the laboratory predation experiments provided the maximum predation rate. The rates reported here are expected to be much higher than what will occur in the field, since the laboratory 163 experiments provided the predators with an unlimited supply of prey and eliminated searching time and the cryptic nature of the thrips. However, using these high predation rates, and knowing the density of onion plants and hence the maximum allowable number of thrips in an area, the minimum number of predators that will be needed to keep the thrips population below a required level can be estimated. The ability of the onion thrips to use grasses, cereals and a host of other plants as food sources, coupled with their rapid development in warm weather, make them common in all onion fields. Yet the use of chemicals on a regular basis for controlling them may not be justified as found by Shirck and Douglas (1956). Such control may be warranted when the thrips occur as unusually high infestation parti­ cularly if this happens early in the season when the plants are younger. Chemical control methods should be used judi­ ciously since Richardson and Wene (1966) reported thrips resistance to dieldrin and other chlorinated hydrocarbons. Heavy rainfall is a major mortality factor, causing over 70 % mortality, both in this study and another by Harris et al (1936). Rainfall is more destructive to the larval population which is the predominant stage in the field through the season except for the first few weeks. Thus, even in periods of high thrips infestation, it may be necessary to recheck the fields following a heavy rain before instituting other control measures. A good pest management program should include the use of resistant cultivars. Sleesman (1934,1943), 164 Lall and Verma (1959), and Verma (1966) all found the white cultivars to be resistant to the onion thrips. Eawar et al (1975) also reported that cultivars with glossy foliage are resistant to thrips. The onion plant morphology (wrinkled surfaces) and the configuration of the leaves (when touching tightly at the bases) appear to enhance the development of the population by making them and rain action. cultivars. inaccessible to predators These factors could be bred-out of desired 6. SUMMARY The onion thrips, Thrips tabaci is a major pest on onions that are grown in organic fashion, i.e. where no insecticides are used. Adult thrips migrated into the onion field when the onion plants were only a few weeks old and were dispersed nearly randomly through the field at the time. As the season progressed and the adults reproduced, the predominant stage present on the onions were larvae and the distribution became more contagious. The numbers of thrips per plant rose sharply and in exponential fashion to a peak by the end of the season. harvest. Thrips may exceed 100 per plant by the time of Sudden heavy rainfall was a high mortality factor and caused the population of thrips to crash. The theoretical lower temperature developmental threshold was estimated as 7.4°C for the larval and pupal stages to­ gether. Under laboratory conditions at 29°C, the development of the two larval and pupal stages took under 8 days. The rate of feeding on onion leaf tissue was similar for both adults and second instar thrips larvae, being 4.93 and 4.51 mm 2 . -1 -1 x thrips x day respectively. Inspite of differences in thrips infestation rate between chemically created and untreated plots, no yield difference was estab­ lished but preliminary investigation in the greenhouse showed that large numbers of thrips could kill the onion plant. The main predators of the onion thrips in the field were coccinellids, the most important being Coleomegilla maculata. The coccinellid predators overwintered as adults and were present in the field very early in the season on cereals and 165 166 grasses at a time when the onions had not even been planted. At this time they fed on a host of small arthropods, the pre­ dominant prey being various thysanoptera. Coleomegilla maculata appeared to have a preferred flight height of 0 - 0.6 m above the ground for short range flight between plants although it flew high above the plants, parti­ cularly over longer distances. It had a bimodal diel activity pattern with peaks in the late morning and late afternoon and a lull in the heat of the day. C. maculata adults fed on about 250 second instar thrips per day. The first, second, third and fourth instar larvae fed at the rate of 0.11, 0.38, 0.57 and 1.09 compared to the adult. When fed only on onion thrips, C. maculata was able to complete its development and lay fertile eggs, indicating that onion thrips are suitable prey. The developmental rate when fed on onion thrips, is comparable to rates obtained by other authors using aphids as a food source. A sliding hypothetical action threshold was developed for various ages of the onion plant based on the feeding rate of the thrips and leaf area production through the season. BIBLIOGRAPHY Abdel-Gawaad, A.A. and A.Y. Shazli, 1970, Studies on Thrips tabaci Lind. VII Effect of food on the life cycle. Z. Angew Entomol. 67(l):27-30. Abdel-Gawaad, A.A. and A.Y. Shazli, 1971, A new method for rearing Thrips tabaci Lind, and bionomics of its eggs and adult stages. Bull Soc Ihtomol. Rgypte 53:443-447. Andrewartha, H.G., 1935, Thrips investigation. 7. On the effect of temperature and food on egg production and length of adult life of Thrips imaginis Bagnall. J. Coun. Scient. ind. Res. Aust. 8:7il-fZ8Z— ---Andrewartha, H.G. and L.C. Birch, 1954, The distribution and abundance of animals. Univ. of Chicago Press, Chicago 782pp. Anon. 1981, Michigan Agricultural Statistics 1981. Department of Agriculture. 80pp, Michigan Banham, F.L., 1968, Thrips infesting the tips of asparagus spears. J. ent. Soc. Brit. Columb. 65:16-18. Baskerville, G.L. and P. Einin, 1969, A rapid estimation of heat accumulation from maximum and minimum temperatures. Ecology 50(3):514-7. Belehradek, J., 1930, Temperature coefficients in Biology. Cambridge Ihil. Soc. Biol. Rev. 5:30-58. Bishara, J., 1934, The cotton worm, Prodenia littura F. in Egypt. Bull. Soc. Roy. Ehtomol. d'Ejgypt 2^:288-420. Bolgiano, N., 1980, Onion plant model, Michigan State University Botany and Plant Ihthology. (In preparation) Bourne, A.I. and F.R. Shaw, 1934, Notes on fungus attacking onion thrips. J. econ. Entomol. 27:860-61. Boyce, K. E. and L.A. Miller, 1954, Overwintering habitats of the onion thrips, Thrips tabaci Lind. (Thysansptera:-Thripidae) in southwestern Ontario. Ent. Soc. Ontario. Ann. Rept. 84:82-86. Brassier, K . , 1930, 1st Coccinella 7-punctata wirklich nur ein Blattlausfresser? Ztschr. Pflanzenkrank. Pflsch., Stuttgart 40:511-513. 167 168 Bullock, J.A., 1963, Extraction of Thrips from Samples of foliage. J. Econ. Entomol. 56:612. Carl, K.P., 1975, An Entomophthora sp. (Entomophtorales: Entomopthoraceae) Fungi parthogenic to Thrips spp. (Thysanoptera: Thripidae) and its potential as a Biological Control agent in glasshouses. Entomophaga. 20(4):381-8. Crozier, W.J., 1926, On curves of growth, especially in relation to temperatures. J. Gen. Physiol. 10:53-73. Davidson, J., 1942, On the speed of development of insect eggs or constant temperatures. Aust. J. Exptl. Biol. & Med. Sci. 20:233-9. Davidson, J., 1944, On the relationship between temperature and the rate of development of insects at constant temperatures. J. Anim, Ecol. 13:26-38. Dimitrov, A., 1975, Restriction of Density of Tobacco Thrips tabaci with Microlophus costalis Biological control. Rastit Zasht. 22(6):34-7. Douglas, J.R. and F.H. Shirck, 1949, Experiments for control of onion thrips. J. Econ. Entomol. 42:68-72. Eddy, C.O. and W.H. Clarke, 1930, The Onion thrips on Seedling Cotton with a season's record of parthenogenetic developments. J. Econ. Entomol. 23:704-8. El-Saadany, G., M.F. El-Shaarawy and El-Refaei, 1975, The damage and damage threshold assessments of Thrips tabaci to cotton. Zeitschrift Fur Angewandte Rntomologie 79X37"281-284. Ewald, G. and R. Burst, 1959, Untersuchungen uber Larchenblasenfuss und Larchenminiermotte. Ang. Forst-u. Jagdztg. 130:173-181. Faulkner, L.R., 1954, Economic thrips of Southern New Mexico, Bull. New Mex. agric. Expt. Sta. 387. Fedorov, S.M., 1930, Tobacco Thrips (Thrips tabaci Lind.) as a pest of Tobacco plant in Crimea. E o s ,6:229-248. Ferguson, C.L., 1969, MicroEconomic Theory. Irwin Co. Illinois. Fry, F.E., 1947, Effects of environment on animal activity. Univ. of Toronto Stud. Biol. :l-62. . Ghabn, A.A.E., 1948, Contribution to the knowledge of the biology of Thrips tabaci Lind, in Egypt. Bull, Soc. Fouad I Ent'. 32:123-174. 169 Hagen, K.S., 1962, Biology and ecology of predaceous Coccinellidae. Ann. Rev. Entomol. 7:289-326. Halteren, P. Van, 1970, The onion thrips, Thrips tabaci on shallots and onions. Legon Agricultural News 4(l):6-7. Harding, J.A., 1961, Studies on the control of Thrips attacking Onions. J. econ. Entomol. 54:1254-5. Harris, H.M., C.J. Drake and H.D. Tate, 1936, Observations on the onion thrips. Iowa Call. J. Sci. 10:155-172. Henderson, C.F. and H.V. McBumie, 1943, Sampling technique for determining populations of the citrus red mite and its predators. Civc. Bur. Entomol. U.S. Dept. Agric.671. Hibbs, E.T. and W.H. Ewart, 1946, DDT for the control of onion thrips. Texas Agric. Expt. Sta. Progr. Rept. 1042. Hinds, W.E., 1903, Contribution to a monograph of the insects of the order Thysanoptera inhabiting North America. Proc. U . S . Museum, 26. Q11-V55, 26:79-242. Hodek, I., 1966, Food Ecology of Aphidophagous Coccinellidae. in Ecology of Aphidophagous Insects. Sept. 27 - Oct. 1, T964, Proc. of a Symp., Prague, Academic Pub. House 360 pp. Hoemer, J.L., 1947., A separator for onion thrips. J. econ. Entomol. 40:755. Holling, C.S., 1959, Some characteristics of simple types of predations and parasitism. Can. Entomol. 91:385-398. Horsfall, J.L., 1921, Sources of infestation of Thrips tabaci in Iowa. J. econ. Entomol. 14:493-496. Hoskins, W.M. and R. Craig, 1935, Recent progress in insect physiology. Physiol. Rev. 15:525-96. Hosny, M.M., 1964, Testing the validity of a simple method for estimating thrips infestation on cottom seedlings in the field. Agric. Res. Rev., Cairo 42(3):136-140. Huffaker, C.B., 1944, The temperature relations of the immature stages of the malarial mosquito, Anopheles quadrimaculatus, with a comparison of a developmental power of constant and variable temperatures in insect metabolism. Ann. Entomol. Soc. Amer. 37:1-27. 170 Janisch, E., 1932, Influence of temperature on the life history of insects. Trans. Roy. Entomol. Soc. London. 80:137-68. Kamal, M., 1951, The biological control of the cotton-leaf worm, Prodenia litura F. in Egypt. Bull. Soc. Fouad Entomol., Cairo"! 35 :221-270. Lall, B.S. and L.M. Singh, 1968, Biology and control of the onion thrips in India. J. econ. Entomol. 61:676-679. Lall, B.S., and S.K. Verma, 1959, Variability of resistance in the onion varieties against the onion thrips. Sci. Cult. 24:657-658. LePelley, R.H., 1942, A method of sampling thrips populations. Bull. Ent. Res. 33:147-148. Lewis, T . , 1973, Thrips: Their biology, ecology and economic importance. Academic Press London. 349 pp. Loan, C. and J.F.G. Holdaway, 1955, Biology of the red clover thrips, Haplothrips higer (Osborn). Can. Entomol. 87:210-219. MacGill, E.I., 1927, The Biology of Thysanoptera with reference to the cotton plant. II The relation between temperature and life cycle in a saturated atmosphere. Annals Appl. Biol. 14:501-512. MacGill, E.I., 1931, The Biology of Thysanoptera with reference to the cotton plant. VII The relation between temperature and humidity and the life cycle. Annals Appl. Biol. 18: 574-583. Mack, T.P. and J.Z. Smilowitz, 1979, Diel activity of Green Peach Aphid Predators as Indexed by Sticky Traps, Environ. Entomol. 8:799-801. Matteson, J.W. and G.C. Decker, 1965, Development of the European corn borer at controlled constant and variable temperatures. J. Econ. Entomol. 58:344-49. Maughan, F.B., 1933, Naphthalene for the control of the onion thrips. J. Econ. Entomol. 26:143-147. Maughan, F.B., 1934, Further studies on the control of the onion thrips. J. Econ. Entomol. 27:109-112. Messenger, constant of three Entomol. P.S. andN.E. Flitters, 1958, Effect of temperature environments on the egg stage species of Hawaiian fruit flies. Ann. Soc. Amer. 51:109-19. 171 Metcalf, C.L., W.P. Flint and R.L. Metcalf, 1962, Destructive and Useful Insects Their habits and control. 1087 pp. McGraw-Hill Book Co. 4th ed. Morison, G.D., 1957, A review of British glasshouse Thysanoptera, Trans. R. Entomol. Soc. Lond. 109:467-534. Mound, L.A., 1973, In Viruses and Invertebrates Thrips and whitefly p. 230 (Gibbs, A.J. ed.) North-Holland Pub. Co. 673 pp. Nefedov, N.I., 1961, Zonal1 nyi Kharakter raspredelenii Kokcinellid KVASSR. Uch. zap. K a b .-Balkarsk. g o s . Univ. 10:105-127. Obryck, J.J. and M.J. Tauber, 1978, Thermal requirements for the development of Coleomegilla maculata (Coleoptera: Coccinellidae) and its parasite Perlitus coccinellae (Thymeroptere: Braconidae) Can. Entomol. 110:40?-12. O'Neill, K 1960, Identification of the newly introduced phlaeothripid Haplothrips clarisetis Priesner (Thysanoptera) Ann. ent. Soc."Am. 53:507-510. Ota, A.K., 1968, Comparison of three method of extracting the flower thrips from rose flowers. J. Econ. Entomol. 61:1754-1755. Pawar, B.B., A.V. Patil and H.N. Sonone, 1975, A Thrips tabaci resistant glossy foliage selection in white onions. Res. J. Mahalma Phule Agric. Univ. 6 (2):152-3. Pearl, R. and L.J. Reed, 1920, On the rate of growth of the population of the United Stated since 1790 and its mathematical representation. Proc. Nat. Acad. Sci. Washington. 6:275-88. Pearson, H.M., 1930, The Thrips factor in onion sterility. J. econ. Entomol. 23:929-31. Powell, D.M. and B.J. Landis, 1965, A comparison of two sampling methods for estimating population trends of Thrips and mites on Potatoes. J. econ. Entomol. 58(6):1141-1144. Priesner, H., 1925, Die Winterquartiere der Thysanopteren. Ent. Jb. 34:152-162. Rabb, R.L., 1972, Principles and concepts of pest management in Implementing practiced pest management strategies. Proceedings of a national extension pest-management work­ shop. Purdue Univ. Lafayette, Indiana. 172 Richardson, B.H., 1953, Control of onion thrips in the Winter Gargen area of Texas. J. Econ. Bitomol. 46:92-95. Richardson, B.H., 1954, Control of onion thrips in the Winter Garden in 1953, J.. Econ. Ehtomol. 47:616-618. Richardson, B.H., 1957, Control of onion thrips in the Winter Garden area of Texas. J. Econ. Entomol. 50:504-5. Richardson, B.H. and G.R. Wene, 1952, Control of onion thrips and its tolerance to certain chlorinated hydrocar­ bons. J. Econ. Bitomol. 49:333-335. Rivnay, E., 1935, Ecological studies of the greenhouse thrips, Heliothrips haemorrhoidalis„ in Palestine. Bull. Bitomol. Res. 2 6 :267-278. Sakimura, K . , 1932, Life history of Thrips tabaci on Emelia sagittalia and its host plant range in Hawaii. T. Econ. Entomol. 25:884-891. Sakimura, K . , 1937a, The life and seasonal histories of Thrips tabaci Lind, in the vicinity of Tokyo, Japan. Oyo Dobuts. - Zasshi 9:1-24. Sakimura, K . , 1937b, On the bionomics of Thripocterms Brui Vuillet, a parasite of Thrips tabaci Lind, in Japan. (1) Kontyu. 11:370-391T Sakimura, K . , 1963, Frankliniella fusca, an additional vector for the tomato spotted wilt virus, with notes on Thrips tabaci, another vector. Hiythopathology 537512-415:---Saxena, R.F., 1971, Some observations on Ceranisus sp (Hym: Eblophidae) parasitizing Thrips tabaci Lind (Thysanoptera: Thripidae) Indian J. Bitomol. 33(1);91-92. Shelford, V. E., 1929, Laboratory and Field ecology. and Walkins Co., Baltimore. 608 pp. Williams Shepard, H.H., 1925, Notes on Thrips tabaci Lindeman. J. econ. Entomol. 19:549-550. Shirck, F.H. and J.R. Douglass, 1956, Experiments on control of the onion thrips in Idaho. J. econ. Ehtomol. 49:526-529. Shirck, F.M., 1948, Collecting and counting onion thrips from samples of vegetations. J. econ. Bitomol. 41:121-123. Shoeib, A.M. and M.M. Hosny, 1973, The distribution of Thrips tabaci Lind, within cotton fields (Thysanoptera). Bull. Soc. Eht. d 1 Ejgypte 56:267-271. 173 Shull, A. F . , 1914, Biology of the Thysanoptera, 2. Sex and life cycle. Am. Nat. 48:236-247. Simpson, R.G. and C.C. Burkhardt, 1960, Biology and Evaluation of certain predators of Therioaphis maculata (Buckton). J. econ. Bitomol. 5 3 :89-?5~ Sleesman, J. P. , 1943, Variations in thrips populations on onions. Ohio Agric. Expt. Sta. Bimonthly Bull. 221. Sleesman, J. P. , 1946, DDT for the control of onion thrips. Ohio Agric. Expt. Sta. Bimonthly Bull. 239. Solomon, M. E. , 1951, Control of humidity with potassium hydroxide, sulphuric acid or other solutions. Bull. Entomol. Res. 42, 543-554. Southwood, T. R.E., 1978, Ecological Methods with particular reference to the study of insect populations. Wiley & Sons. 524 pp. Stannard, L.J., 1968, The Thrips or Thysanoptera of Illinois. Bull. 111. St. Nat. Hist. Surv. 29:216-552. Stern, V.M., R.F. Smith, R. van den Bosch and K.S. Hagen, 1959, The integrated control concept. Hillgardia 29(2);81. Sylvester, E. S. and E.L. Cox, 1961, Sequential plans for sampling aphids on sugar beets in Kern County, CA. J. Econ. Ehtomol. 54:1080-1085. Szumkowaski, W . , 1955, Observaciones sobre la biologia de algunos Coccinellidae (Coleoptera) (1) Bol. Ent. Venezolana. 11:1-20. Taylor, E.A. and F.F. Smith, 1955, Three methods for extracting thrips and other insects from rose flowers. J. Econ. Bitomol. 48:767-768. Tutel, B . , 1963, Morphological anatomical and physiological studies on the Malatya and Bursa varieties of tobacco in relation to their resistance to thrips. Rev. appl. Ehtomol. (A). 52:467. Uvarov, B. P. , 1931, Insects and climate. Bitomol. Soc. London. 79:1-248. Trans. Roy. Verma, S.K., 1966, Studies on the host preference of the onion thrips, Thrips tabaci Lind., to the varieties of onion. Inidian J. Bit. 28:396-398. Villiers, J.F. De, 1976, Economic importance of Thrips tabaci Lind. (Thysanoptera) on cotton at Vaalhartz, South Africa. J. Bit. Soc. S. Africa 39(l):63-66. 174 Vinson, J., 1929, Le Thrips de L'oignon (Thrips tabaci Lindeman). Leafl. Dept. Agric. Maurice No. 39. 3 pp. Reduit. Watts, J.G., 1934, A comparison of the life cycles of Frankliniel1a tritici (Fitch), F. fusca (Hinds) and Thrips tabaci Lind. (Thysanoptera - Thripidae) in South Carolina. J. econ. Bitomol. 27:1158-9. Wigglesworth, V.B., 1953, The principles of insect physiology. Methuen and Co. Ltd. London. 546 pp. Wilcox, J. and A.F. Howland, 1948, DDT dust for control of onion thrips. J. Econ. Bitomol. 41:694-700. Appendix Table la. Date Summary of onion thrips recorded on onions, Total # plants Faton Rapids, 1978. Total # Thrips # A L Z A Thrips/plant L Z 516 101 116 217 .20 .22 .42 7-12-79 322 341 110 451 1.06 .34 1.40 7-15-79 303 317 558 875 1.05 1.84 2.89 7-19-79 311 184 1216 1400 .59 3.91 4.50 7-24-79 213 143 1122 1265 .67 5.27 5.94 7-26-79 220 258 1253 1511 1.17 5.70 6.87 8-2-79 109 145 865 983 1.33 7.94 9.27 8-10-79 112 5674 2315 7989 50.66 20.67 71.33 A = Adult L = Larvae Z = Total 175 6-30-79 Appendix Table lb. Date Summary of onion thrips recorded on onions, Total # plants Total # Thrip s # A L E A 15 1 16 7-5-79 140 33 8 41 7-13-79 120 70 52 122 7-21-79 120 64 77 140 7-26-79 160 267 150 8-3-79 160 302 8-17-79 120 8-24-79 L .12 E .13 .06 .30 .43 1.58 .53 .64 1.17 417 1.67 .94 2.61 339 641 1.89 2.12 4.01 265 799 1064 2.21 6.66 8.87 120 216 1358 1574 1.80 11.32 13.12 8-28-79 120 98 2641 2739 .82 22.01 22.83 8-31-79 120 151 4130 4281 1.26 34.42 35.68 9-7-79 80 932 2527 3460 11.66 31.59 43.25 9-9-79 60 794 1977 2771 13.23 32.95 46.18 9-10-79 60 970 2479 3449 16.17 41.31 57.48 A = Adult L = Larvae E = Total CM .01 00 120 Thrips/plant 176 6-28-79 Eaton Rapids, 1979. Appendix Table lc. Date Summary of onion thrips recorded on onions, Total # plants Total # A L ThripIS £ Eaton Rapids, 1980. # A Thrips/plant L I 360 139 38 177 .39 .11 .71 6-19-80 300 162 163 325 .54 .54 1.08 6-26-80 300 420 1300 1720 1.40 4.33 5.73 7-2-80 300 633 1570 2203 2.11 5.23 7.34 7-10-80 300 794 2584 3378 2.65 8.61 11.26 7-16-80 300 4255 7324 11579 14.18 24.41 38.59 7-23-80* 150 1098 4525 5623 7.32 30.17 37.49 7-30-80 150 2345 16646 18991 15.63 110.97 126.61 8-6-80 150 1515 8001 9516 10.10 53.34 63.44 8-13-80 150 675 4166 4841 4.50 27.77 32.27 8-22-80 60 417 1384 1801 6.95 23.07 30.02 Cultivation done early that morning. the leaves. A = Adult L = Larvae E = Total No thrips occurred on the upper parts of 177 6-13-80 Appendix Table Id. Date Summary of onion thrips recorded on onions at the Muck Farm, 1980. Total # plants Total # Thrips A L 2 # Thrips/plant A L 2 60 57 96 153 .95 1.60 2.55 7-31-80 60 80 100 180 1.33 1.67 3.00 8-7-80 60 99 294 393 1.65 4.90 6.55 8-14-80 60 45 706 751 .75 11.77 12.52 8-28-80 60 90 667 757 1.50 11.12 12.62 A = Adult L = Larvae 2 = Total 178 7-25-80 Appendix Table 2. Sample means (///plant) and variances for various onion thrips sampling dates transformed to log/log scale, for determining if onion thrips field distribution is normal, (sample size = 10 plants). 2 s , 2 log s Date X 6-13-80 0.9 0.7 0.7 1.1 1.4 1.0 0.5 0.9 -0.05 -0.15 -0.15 0.04 0.15 0.00 -0.30 -0.50 1.88 0.71 0.46 1.43 0.60 1.11 0.50 0.99 0.27 -0.15 -0.34 0.16 -0.22 0.50 -0.30 -0.004 2.09 1.01 0.66 1.30 0.43 1.11 1.10 1.10 6-19-80 0.8 1.1 1.3 -0.10 0.40 0.11 1.51 1.21 1.21 0.81 0.08 0.08 1.89 1.10 0.93 6-26-80 2.8 1.6 1.8 8.1 8.8 8.9 0.45 0.20 0.26. 0.91 0.94 0.95 14.18 1.82 3.73 34.35 51.73 12.32 1.15 0.26 0.58 1.54 1.71 1.09 5.06 1.16 2.07 4.24 5.88 1.38 7-10-80 22.1 21.3 26.0 1.34 1.33 1.41 599.43 247.34 587.33 2.78 2.44 2.77 27.12 11.61 22.59 7-30-80 132.8 61.9 2.12 1.79 6554.18 1206.16 3.82 3.08 49.35 19.49 log X S 2/i 179 180 Appendix Table 3. Vertical distribution of coccinellids caught on Stickyboard traps located in onions at the Muck Farm, Laingsburg, 1979. (A total of 23 traps were used until 7-12-79 and only 9 traps thereafter). Collection Date 7-06-79 llcinht n&xgn l (m) 0 .3 .6 .9 - .3 .6 .9 1.2 TOTAL 7-12-79 0 .3 .6 .9 - .3 .6 .9 1.2 TOTAL 7-20-79 0 .3 .6 .9 - .3 .6 .9 1.2 TOTAL 7-28-79 TOTAL 0 .3 .6 .9 - .3 .6 .9 1.2 CM & HT CM ■& HT■" Total // Percent Total // 75 80 27 11 39 41 14 6 10 36 25 9 13 45 31 11 193 100 80 100 81 116 47 12 32 45 18 5 6 14 14 11 13 31 31 25 256 100 45 100 49 92 31 6 28 52 17 3 0 2 2 2 0 33 33 33 178 100 6 99 10 22 15 3 20 44 30 6 0 0 1 1 0 0 50 50 50 100 2 100 677 133 Percent 181 Appendix Table 4. Vertical distribution of cocclnellids caught on 15 stickyboard traps located In spring sorghum at the Muck Farm, Lalngsburg, 1980. Collection date 7-06-80 Height (m) 0 .3 .6 .9 - .3 .6 .9 1.2 TOTAL 7-11-80 0 .3 .6 .9 - .3 .6 .9 1.2 TOTAL 7-25- 80 0 .3 .6 .9 - .3 .6 .9 1.2 TOTAL 7-31-80 0 .3 .6 .9 - .3 .6 .9 1.2 TOTAL 0 3 6 9 EZ - .3 .6 .9 1.2 Total i CM Percent HT Total if Percent Total t AB Percent 30 11 4 3 63 23 8 6 0 3 0 2 0 60 0 40 1 1 5 3 10 10 50 30 46 100 5 100 10 100 9 4 2 1 56 25 13 6 0 6 4 1 0 55 36 9 2 1 2 3 25 13 25 17 16 100 11 100 8 100 9 6 5 0 45 30 25 0 9 31 8 2 18 62 16 4 1 1 1 1 25 25 25 25 20 100 50 100 4 100 0 4 3 3 0 40 30 30 0 14 13 6 0 43 39 18 0 0 1 0 0 0 100 0 10 100 33 100 1 100 3 7 6 8 13 29 25 33 2 16 17 3 5 42 45 8 0 0 2 1 0 0 66 34 24 100 38 100 3 100 118 137 26 182 Appendix Table 5. Vertical distribution of coccinelllds caught on 10 stickyboard traps located In winter wheat at Eaton Rapids, 1980. Collection date Height (m) 7-02-80 0 3 6 9 - .3 .6 .9 1.2 TOTAL 7-10-80 0 .3 .6 .9 - .3 .6 .9 1.2 TOTAL 7-17-80 0 .3 .6 19 - .3 .6 .9 1.2 TOTAL 7-23-80 TOTAL 0 .3 .6 .9 - .3 .6 .9 1.2 Total CM Percent HT Total tf Percent AB Total t Percent 12 12 9 3 33 33 25 9 5 9 3 2 26 47 16 11 0 0 2 1 0 0 66 34 36 100 19 100 3 100 3 1 2 0 50 17 33 0 3 4 1 5 23 31 8 38 0 0 2 1 0 0 66 34 6 100 13 100 3 100 9 9 2 2 41 41 9 9 7 3 2 1 54 23 15 8 0 0 0 0 0 0 0 0 22 100 13 100 0 0 30 13 2 3 63 27 4 6 3 6 4 0 23 46 31 0 0 0 0 0 0 0 0 0 48 100 13 100 0 0 112 58 6 183 Appendix Table 6. Vertical diBtribution of cocclnellids caught on 11 stickyboard traps located in or adjacent to spring oat at Eaton Rapids, 1980. Collection date 7-02-80 Height (m) 0 .3 .6 .9 - .3 .6 .9 1.2 TOTAL 7-10-80 0 .3 .6 .9 - .3 .6 .9 1.2 TOTAL 7-17-80 0 .3 .6 .9 - .3 .6 .9 1.2 TOTAL 7-23-80 TOTAL 0 .3 .6 .9 - .3 .6 .9 1.2 CM Total // Percent HT Total # Percent AB Total # Percent 5 15 9 3 16 47 28 9 16 26 20 2 25 41 31 3 1 1 4 5 9 9 36 46 32 100 64 100 11 100 1 7 1 2 9 64 9 18 4 12 3 2 19 57 14 10 0 2 5 4 0 18 46 36 11 100 21 100 11 100 12 6 4 1 52 26 18 4 7 3 3 1 50 21 21 8 0 1 2 2 0 20 40 40 23 100 14 100 5 100 2 3 2 0 29 42 29 0 0 1 1 0 0 50 50 0 0 0 0 2 0 0 0 100 7 100 2 100 2 100 73 101 29 Appendix Table 7. Coccinellids caught on 16 stickyboard traps at Eaton Rapids, 1979, transformed to -1 catch x 100 traps -1 x day No. of Traps Total Coccinellids 6-20-79 6-25-79 7-04-79 7-11-79 7-16-79 7-20-79 7-23-79 7-27-79 8-02-79 8-06-79 8-13-79 8-16-79 8-20-79 8-24-79 8-27-79 8-30-79 9-07-79 START 5 9 7 5 4 3 4 6 4 7 3 4 4 3 3 8 16 16 16 16 16 16 16 16 16 16 16 16 16 16 16 12 6 104 140 128 48 56 34 8 16 47 4 31 18 14 18 16 79 252 688 TOTAL Catch ^ traps ^ x Day .075 .722 1.250 1.600 .750 .500 .530 .083 .250 .420 .083 .484 .281 .292 .375 .167 Catch x 100 ^ trap x day 8 72 125 160 75 50 53 8 25 42 8 48 28 29 38 17 184 Date Trapping Days Appendix Table 8. Coccinellids caught in 8 flight interception traps at Eaton Rapids, 1979, transformed to catch x 100 traps ^ x day-"*". Date 6-25-79 6-29-79 7-16-79 7-20-79 7-23-79 7-25-79 7-27-79 7-30-79 8-06-79 8-13-79 8-16-79 8-20-79 8-24-79 8-27-79 8-30-79 9-07-79 TOTAL Trapping Days START 4 17 4 3 2 2 3 7 7 3 4 4 3 3 8 74 No. of Traps Total Coccinellids 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 2 1 0 0 2 0 0 2 0 1 7 10 12 10 1 120 48 Catch x traps ^ Catch x 100 trap ^ x day x day ^ .06 .007 0 0 .125 0 0 .036 0 .042 .219 .313 .500 .417 .016 6 0.7 0 0 13 0 0 4 0 4 22 31 50 42 2 Appendix Table 9. Coccinellids caught in 15 pitfall traps at Eaton Rapids, 1979, transformed to catch x 100 traps Date Trapping Days 7-11-79 7-16-79 7-20-79 7-23-79 7-27-79 7-30-79 8-06-79 8-13-79 8-16-79 8-20-79 8-24-79 8-27-79 8-30-79 9-07-79 START 5 4 3 4 3 7 7 3 4 4 3 3 8 15 15 15 15 15 15 15 15 15 15 15 15 15 58 180 TOTAL ^ x day No. of Traps Total Coccinellids 13 2 10 6 8 5, 21‘ 2 11 10 6 5 6 105 Catch x traps -1 .173 .033 .222 .100 .178 .048 .200 .044 .183 .167 .133 .111 .050 x day -1 Catch x 100 -1 . -1 traps x day 17 3 22 10 18 5 20 4 18 17 13 11 5 Appendix Table 10. Coccinellids caught on stickyboard traps at the Muck Farm, Laingsburg, 1979, transformed to catch x 100 traps ^ x day Date Trapping Days No. of Traps Total Coccinellids Catch x trap ^ x day ^ Catch x 100 trap ^ x day ^ START 7-06-79 7 29 398 1.96 196 7-12-79 6 29 324 1.86 189 7-20-79 8 9 194 2.69 269 7-28-79 8 9 70 .97 97 29 76 986 187 6-30-79 Appendix Table 11. Population density of coccinellids in various crops at Eaton Rapids, 1980. Density includes larval, pupal and adult stages as indicated. Crop Date Sampling Method Spring Oat 5-24-80 Visual count on Ragweed Onions 5-24-80 Onions Spring Oat Sample Size Density (#/m2) Remarks 235 Ragweed 0.026 Adults only Visual count on Ragweed 170 Ragweed 0.116 Adults only 5-24-80 Visual count on Ragweed 125 Ragweed 0.051 Adults only 6-12-80 6-16-80 7-21-80 lm2 Quadrat lm2 Quadrat lm2 Quadrat 20 50 20 1.40 1.80 2.40 8-19-80 lm2 Quadrat 12 1.75 Adults only Adults only Larvae, pupae, Adults Larvae, pupae, Adults Winter wheat 6-16-80 6-17-80 7-21-80 lm2 Quadrat lm2 Quadrat lm2 Quadrat 10 20 20 3.75 4.05 1.25 Alfalfa 8-18-80 lm2 Quadrat 20 7.55 Onions 7-18-80 Visual count on Onions 9x150m row of onions 0.02 Adults only Adults only Larvae, pupae, Adults Larvae, pupae, Adults Larvae & Adults 8-4-80 Visual count on Onions 9x150m row of onions 0.26 Larvae & Adults 189 Appendix Table 12. Arthropods extracted from giant ragweed, Ambrosia trifids, winter wheat and spring oat, at Eaton Rapids, 1980. Source Date Extracted Number Collected ARACHNIDA Giant Ragweed Arthropod 6-02-80 Araneida 2 INSECTA Collembola Thysanoptera Hemiptera (Miridae) Homoptera (Aphidae) Diptera Winter Wheat 13 51 20 3 6 ARACHNIDA 6-05-80 Acarina 2 INSECTA Collenbola Thysanoptera Homoptera (Aphidae) Diptera ARACHNIDA Spring Oat 3 1151 37 2 6-20-80 Acarina 2 INSECTA Collenbola Thysanoptera Homoptera (Aphidae) Coleoptera Diptera 1 280 56 4 6 Appendix Table 13. Diel activity of adult coccinellids in onions as indexed by four flight interception traps at Eaton Rapids, August 4, 1980. Time Temp.°C Number Caught Moving towards onions away from onions CM HT CM HT Total CM HT 13.5 start 900 16.0 2 0 0 0 2 0 1000 21.3 4 2 1 0 5 2 1100 23.0 4 0 5 1 9 1 1200 25.0 2 3 0 1 2 4 1300 26.3 2 1 3 0 5 1 1400 26.8 1 1 3 1 4 2 1500 27.0 5 1 4 0 9 1 1600 25.8 1 0 5 3 6 3 1700 25.4 4 3 3 1 7 4 1800 24.8 3 1 0 0 3 1 1900 23.5 0 1 2 0 2 1 2000 23.0 1 2 0 1 1 3 190 800 191 Appendix Table 14. Number of Thrlps tabacl larvae (second Instar) required for development of