ASPECTS OF TEE FE‘EBING BEHAVIOR OF SCOLYTUS MULTESTREFITUS, THE PREMARY VECTOR 0F WEE ELM MS-EASE, AK?) A. CREYECAL EVAEMTIGI‘! OF ‘ __ PRESENT CHEMLCAL CGK‘FROL HEASURES Thests got- Hm Degree of DE. D. MICHIGEN STATE WWEBSHY Helmut Woifgang Riedi 3973 fimw" _'!_ ‘.' . '. .c T" ’ 1., i - a - Mi-thigvan Fuse iv; v. ' D .W‘W?"14 4‘ r} ' “1-. a... This is to certify that the thesis entitled Aspects of the Feeding Behavior of Scolytus multistriatus, The Primary Vector of Dutch Elm Disease, and a Critical Evaluation of Present Chemical Control Measures presented by Helmut wolfgang Riedl has been accepted towards fulfillment of the requirements for Ph-D. degree in Entomology Major professor Date JXpril 27. 1973 0-7639 Lg , ASPECTS or THE :1; 7. . 06‘“ THF Pawn ~ ‘ AIDA cnmm .. "fl 9" .I' O ‘4 u t". i “V '-r Smpllng 9‘ 3 .i c (_.>? pnhrontlally in ‘l'.‘ H», '1?’ “CINOSCi linear} iy fag“ [hr bi: ~. 1" “Heuristic: sent "was! {, a “7* .: _, , ":‘fix h “it 512‘ Dr. ave {hr 11‘:' . '. ~' on o-v n5?» u no influence 4:. Feeding ' ,. . . t 9 4R... “elation between a were u .. n \ _ ,._h “33.3 a,” 1i, ,, I . ugh.” ”I‘llh re£\jrtadl i'lbU] t" H“ l'.‘ vIJCI'H ‘\l I ‘ J: .55" IU' : ‘l - . ”'5'”? l Wm! feeding. .,'=Mg Staples were :nllerieu from 3 bright strata at u- iv . ‘ri-P. _ . I‘IC blower-sprayed e 7' trees and then x 0“» tar. M b50059!) ‘ I r. “IQ I l _\ ultistrlatus; and m chemical arrays“ naru residue. 0‘ . J ‘ x» ' ; rhncwtcf-Cppllcd :net‘iox/tho: fomula'iom. a ’281 famlotlcu ‘ w J :1; ' . I. has“! as anti-drift material gave LN :44 orotrctlofl. Mar- "'. ,_ x» . .A _ "6m! “than/chic!" carry-Over from min Mum .mticaum ; , your: finvily Interfrrod with an Obje‘llvt valuation o! m r. ifiu) V ‘ ~, '35,“ ’° out" trauma. Corry-over from Mllwptor treatmnts 5;: .0 ”Huh“. A wrlson of ground and aerial apnllcatlon .‘4 vol 6!. all: bio-nor dutch can Inna-I N and 90: pm— -g§nm new; «m guru-u... has mm: i5» v 9" . . ‘1‘ g Wh' 33.nov- ' r .«i ."““ Ir.' . , (r1. ASPECTS OF THE FEEDING BEHAVIOR OF SCOLYTUS MULTISTRIATUS, rfiQVL . THE PRIMARY VECTOR OF DUTCH ELM DISEASE, AND A CRITICAL EVALUATION OF PRESENT CHEMICAL CONTROL MEASURES ABSTRACT ‘u c ‘ . 53"». By ‘ Helmut Wolfgang Riedl Sampling of untreated elms revealed that §, multistriatus feeds 41w. ' preferentially in the whole upper tree region. Feeding attack in- :“0‘zr reroased ilneariiy from the bottom to the top. Physical twig crotch Sharacteristics were investigated with respect to vector feeding. a :ffhES; angle size between the main and lateral member of a twig crotch i M;§é§f;é influence on feeding. There was, however, a significant ? inaociatlon between a more rounded twig crotch base and lateral U‘jf ing which reportedly results in more successful inoculations than ands: blower-sprayed elm trees and then subjected to bioassays with Figlmultistriatus, and to chemical analysis for bark residues. 0f “ Shoiicopter-appiied methoxychior formulations, 3 l2. 5% formulation ?;; Dacagln as anti- -drift material gave the best protection. How- Helmut Wolfgang Riedl liiit'ahiy factor which affected effectiveness of the mist blower. The , If 3¢opter, although favoring tree- top coverage, gave insufficient pro- I§g¢ ffiéctién, never exceeding 30% in any of the sampled height strata. The ~.,%ip responses measured in the bioassays, mortality and 'failure to score ‘ilg'ihlem', were well correlated except in bioassays of twig crotches with , ;“ Xhésh surface deposits. 1; Tree kill in the helicopter and mist blower treatment areas over _ 4‘ f , a 2-year period appeared to be independent of the degree of protectiog as ‘igffifls indicated by the bioassay results. This discrepancy was explained by t Ié‘tlifferential vector pressure and root-graft transmission in the various nireatment areas. To objectively monitor quality and quantity of spray coverage _ "%hroughout the crown region, a sampling system was developed empl0ying 5‘9‘dniensional cube-like sampling devices. The helicopter gave highly Vnscrete spray patterns. Dacagin shifted the draplet size spectrum to fl 43¢ larger sizes, while straight emulsions showed higher dispersion. r I ‘fai~ial application resulted in poor 3-dimensionai coverage of the sample viii;sthICh suggested that less than half the total bark or twig crotch I received spray. Also, dosages applied to each tree (55, 2 quarts '”gh;5%.spray) were found to be too small to provide necessary protection. :Mfi‘Viist blower created finely-dISpersed, almost continuous spray de- it; which often exceeded the levels needed for protectioniand 3-dimen- ;1 coverage was superior, even in the upper tree region. ;%fig correlations of beetle response (mortality) with bark residues a lower Lcso value for the mist blower- and lab-treated-twigs helicopter-treated twigs. This was explained by the difference Helmut Wolfgang Riedl ";fohhtact toxicity of fresh methoxychior emulsion deposits was -,? than solution deposits. The addition of Dacagin to the emulsion i;d contact toxicity signifiCantly. Contact with weathered surface .Fijiy its caused little mortality and had no measurable impact on feeding i‘lig so. Also, the physical structure of methoxychior surface deposits 14V§T§ documented and related to contact toxicity. _;. Hethoxychior deposits on elm bark proved to be quite persistent Lfl;ith a decomposition rate of around 70% after one year. Dacagin delayed gaiaeathering, even when added in small quantities to the formulation. Dis- ;aiecement of methoxychior residue from the point of deposition to other 'fhiites on an elm branch was not observed. '1‘ The Dutch elm disease control program on the campus of Michigan 5;;ri to University was considerably more successful than that of surrounding . {n’ities between l958 and 1972. Elm die-back increased since l968 T“the mist blower-DDT-methoxychlor program was replaced by a helicopter- }31ychlor program. However, increasing vector pressure over the past ears contributed to the growing number of diseased elms on the ‘ l. A l.‘ ‘9". 3". ‘l .4 I V? i.iI‘sI'IIEIrTs OF THE FEEDING BEHAVIOR or scourus 4‘ .J yfi' . MULTISTRIATUS, THE PRIMARY VECTOR OF DUTCH ELM DISEASE, "JZAHD A CRITICAL EVALUATION OF PRESENT CHEMICAL CONTROL MEASURES BY Helmut wolfgang Riedl A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Entomology Hy ,i;l mcwww inml i ' ‘ I L l’n r I ~— ' I, - I I l i fin- \ l s. . i l ’l. _ .081!) err-r ‘ "F, {JI'J'I ." 556cm ' .n .‘. T. vprogress. To Crystal and .~: I 4 my Parents r . tv.lI'i 'I' o a“) .‘F'M‘k' ‘_ “ L! . fl“ '0! P‘IL‘ ‘e ACKNOWLEDGMENTS I would like to express my appreciation to Dr. J. W. Butcher, a AsgAstant Dean for Graduate Programs, College of Natural Science, who "I '( fAvaided the opportunity for me to study in the United States and served as my academic advisor. My sincere thanks also go to Dr. Gordon E. Guyer, a'gbggrman, Department of Entomology, who secured the financial assistance ' A necessary for my studies and displayed great interest in my academic '2' regress. n'.‘ ._ i"; For their suggestions, criticisms, and general contributions in E‘hh$:compietion of this dissertation, my thanks go to my committee members: v ;.u. H. Butcher, G. E. Guyer, F. Stehr, Wm. E. Wallner, L. F. Wilson, . ‘3‘“ Wright, and M. Zabik. Further, I am grateful for the cooperation and help of Burt D. iii TABLE OF CONTENTS 7?” .' ,HILIST or TABLES . . . . . . . . . . . . . . . . . Iuvifl or FIGURES. . . . . . . . . . . . . . . . . -_' INTRODUCTION. . . . . . . . . . . . . . . . . . 7}, LITERATURE REVIEW . . . . . . . . . . . . . . . . “a 7 (A) History and DevelOpment of Dutch Elm Disease in Europe ‘ (I) introduction of Dutch Elm Disease into the United States and Present Status. . . . . . . . . . . (8) Hosts in EurOpe and North America . . . . . . . . }g ”(D) Other Vectors in North America . . ‘1 .(E) Dutch Elm Disease Pathogen: Ceratocxstis_ ulmi (Buisman) , a. C. Moreau. . . . . Biology of Scoixtus multistriatus in Michigan . . Feeding and Breeding Behavior of Scoixtus multistriatus Control Measures . . . . . . . . . . . I. Population control . . . . . . . . 2. Chemotherapy against fungus . . . . . 3. Control of vector feeding . . . . . . .Methoxychior: Properties and Performance. . . . Structure and Toxicity of Insecticidal Deposits. . Spray Technology . . . . . . . . . . . . Bioassay. . . . . . . . . . . . . . . a a o I a I Aspects of the Feeding Behavior ofS Scolytus multistriatus ' 0 Mrs mm D O C I I I I I O O I O I I I Introduction. . . ’. . . . . . . . . . . fiateriais and Methods. . . . . . . . . . . . (A) Regional Distribution of Feeding Injuries along a Vertical Gradient . . . . . . . (B) Regional Distribution of Feeding Injuries along a Horizontal Gradient. . With Feeding! a O I 0 O I a o C a I Results and Discussion . . . Mry and COROlUSiOHS a I e ”-0 a V. I» or a a I? ‘(6) Association of Physical Twig Crotch Characteristics Page viii 1|. Evaluation of Methoxychlor Formulations and Application I Techniques. 1 Introduction . . Materials and Methods . Spray Applications 1969. Spray Applications 1970 . . . . fiwig Sampling . . . Bioassay Technique. . . . . . . . . . . Scolytus Rearing Quantification of Methoxychlor Residues on Elm Bark . . . . . . . . . . . . Bark sampling . . Chemical analysis . Efficiency of extraction procedure Quantitative changes of methoxychlor residues during extended freezer storage. . . . . . . . . .PWN— Results and Discussion. (A) (B) (C) (D) (E) (F) Bioassays 1969 . Bioassays 1970 . 1. Helic0pter 2. Mist blower . . Correlation between Mortality and Feeding ReSponse in the 1969 and 1970 Bioassays . Mortality at 3 Height Levels . . . 1. Bioassays 1969/70: helic0pter. . . . 2. Bioassays May, July,1970: mist blower . Methoxychlor Residues on Treated Elm Trees. . . 1. Carry-over 2. Methoxychlor residues 1968/69. 3. Methoxychlor residues 1970 . Evaluation of Treatment Effects by Comparison of Disease Incidence in Treatment Areas l-Vl: 1968-1969. . . . . . . . . . . Summary and Conclusions . . . . . . 111. Monitoring of Spray Coverage on Elm Trees . Introduction . . . . . . . . . . . . Materials and Methods . . . . . (A) 3- Dimensional Sampling Device . . . . . . . (B) Spray Applications. . . . . . 1. Helicopter application 1971/72.. . . . 2. Mist blower application 1972 . . . . . (C) Analysis of Spray Deposits . . . Analysis of spray patterns on filter disks . . - 2. Quantification of spray deposits on filter disks . . . . . . . . Page 109 113 116 116 117 117 119 119 120 120 120 121 Page 111. (continued) Results and Discussion. . . . . 123 (A) Spray Pattern, Droplet Size and Density, along a Vertical Gradient in Treated Elm Trees . . . . . . . . . 123 l. Helic0pter application . . . . . . . 123 2. Mist blower application . . . . . 128 (B) Horizontal and Vertical Distribution of. Spray Deposits on Treated Elm Trees . . . . 131 1. Helicopter application . . . . . . . 132 2. Mist blower application . . . . . . . 135 Summary and Conclusions . . . . . . . . . . . 136 IV. Insecticidal and Decomposition Pr0perties of Methoxychlor Formulations . . . . . . . . . . . . . . 138 Introduction . . . . . . . . . . . . . 138 Materials and Methods . . . . . . . 140 (A) Bioassay Procedures for 0ra1 LC50 Determinations . . . . . . 140 (B) Fiber Board Units for Contact Toxicity Testing. . . . 143 (C) Application of Methoxychlor Formulations on Elm Branches to Investigate the Behavior of Bark Deposits. . . . . . . 146 1. Structure of surface deposits . . . . . I46 2. Decomposition of bark deposits. . . . . 147 3. Movement of bark deposits . . . . . . 147 Results and Discussion. . . . . 148 (A) Oral Toxicity of Methoxychlor Deposits in Elm Twig Crotches . . . 148 I. Residue-mortality regressions calculated from field samples . . . . . . . 148 a. Helic0pter application . . . . . 148 b. Mist blower application . . . . 150 2. Responses of S. multistriatus on lab- treated twig crotches . . . . 152 3. Comparison of male and female feeding response on untreated twigs . . . . . 155 4. Relationship between mortality and sex in the bioassays. . . . . 155 (B) Contact Toxicity of Methoxychlor Deposits . . . 158 I. Comparison of 3 methoxychlor formulations . 158 2. Contact toxicity and feeding response . . 161 (C) Structure of Surface Deposits of Methoxychlor. . 163 (D) Decomposition of Methoxychlor Formulations. . . 166 (E) Movement of Methoxychlor on Elm Bark. . . . . 169 Summary and Conclusions . . . . _ , . , , , , 17] vi Page $1'l . v. "I. :1! "“jktch Elm Disease Control on the Campus of Michigan '.\« 'State University 1958-71. . . . . . . . . . . . 173 1’.;LT V“nuo CONCLUSIONS . . . . . . . . . . . . . . 180 ,v'IuaE CITED. . . . . . . . . . . . . . . . . I86 Friedmar . I leyc'n HEINOKVu and n.5I tnoxychlo. ~~ pod ndati..ur 1' v}. - x. i f if V. h.55 1 . I!) a)? )3 I ..‘. . “LA-e 5 s 4,»: a u t”S {)3 Mali __ 'x . ‘ . . A. _ - '1 ".‘_ . , a- ‘u Ag'lr'u 5‘ 1,S?t' '_ T. 3:- EL'. Table LIST OF TABLES Page Comparison of bioassay procedures used by investi- gators in evaluating the effectiveness of insecti- cides for controlling S. multistriatus . . . . . . 46 Percent feeding attack at 5 sample points in 3 height levels . . . . . . . . . . . . . 54 Comparison of feeding attack in 4 cardinal quarters and center section of elm trees using Friedman's ANOVA . . . . . . . . . . . . . . . . . 57 T-comparisons of angle means of attacked and un- attacked twig crotches . . . . . . . . . . . . 59 Rate of recovery for methoxychlor from elm bark samples . . . . . . . . . . . . . . . . . 84 Stability of methoxychlor residues during freezer storage . . . . . . . . . . . . . . . . . 86 Responses of S. multistriatus in bioassays of treatments I - V1 (19691 . . . . . . . . . . . 88 Friedman's ANOVA by rank of mortalities at 3 height levels of helicopter and mist blower-treated elms . . . 99 Methoxychlor residues in twig crotches of helic0pter- and mist blower-treated campus elms: 1968, 1969, 1970. . 105 Methoxychlor residues in twig crotches of helic0pter- and mist blower-treated campus elms: 1970 . . . . . 108 Tree losses I968/69 in treatment areas I - VI . . . . . 111 Helicopter application: number of spray droplets per unit area and actual area covered on horizontal filter disks at 2 height levels of treated elm trees . . 124 Average methoxychlor deposits on sample cubes at 2 height levels of elm trees . . . . . . . . . . I33 viii '. ‘.5uv' ‘lh ‘1’ :I.<%1- T1 v ' i D I‘. i s. "P ‘58- dfiin 1":‘19- ‘ . . “4121.). .'. I ‘ ' ,1: .311. Igfi 2%- .; ‘.33. I >, ‘ J u , _..347, V ~. tjyig crotch residues and respective mortalities ”on sprayed campus elms (1970) . . . . . Response of g, multistriatus in bioassays of lab- treated twig crotches. . . . . . . . . . ‘”1;' Results of Finney's probit analysis of the dosage“ mortality data in Table 15 . . . . . . . . -Association of mortality and beetle sex in bio- ‘assays of methoxychlor-treated elm twigs . . . Contact toxicity of methoxychlor deposits on fiber boards to §, multistriatus . . . . . . Beetle mortality and feeding response after ex- pOSure to treated fiber boards. . . . . . . Decomposition of methoxychlor formulations on elm bark over a one year period. . . . . . . . Residue recovery from bark sections A, B, C, and D. Tree losses in DDT and methoxychlor blocks, 1964/65 Summary of DED control and elm losses on the campus [Of MS": ‘958-71 0 o o o o o o a o u o .-DED spraying in Lansing, Michigan . . . . . . Page 142 153 153 I57 159 162 167 170 174 I76 178 LIST OF FIGURES Dutch elm disease cycle: interaction of vector, fungus and host tree (unshaded circle: healthy elm, shaded :1 CirCle: diseased elm). o I o o o o o o o o 2. Sampling scheme for mature elm trees: location of sample points in three height levels . . . . . . 53. Type of twig crotch base. . . . . . . . . . . 1" 4. Type of twig crotch feeding by §, multistriatus . . . f 35. Twig crotch feeding by S. multistriatus on 6 untreated ‘ ' elms. . . . . . . . . . . . . . . . . 1:;16. Dutch elm disease treatment areas 1968/69 . . . . . Dutch elm disease treatment areas 1964-66 . . . . . Aerial spraying of elm trees with Bell G-Z helic0pter . Bell 6-2 helic0pter: part of boom and nozzle assembly. ground Spraying of elm trees with John Bean Rotomist We] 301-6) 0 o I o c o o I o o o o o o ‘: .doho.3ean Rotomist: blower and nozzle assembly . . . 1/Sampiing scheme for treated campus elms: location of "§,sampling with aluminum pole pruner. . . . . . 13.1. n" 'Iééing of brass cage in twig crotch for bioassay . . 9‘1 L§f_§, multistriatus In brass cage with vacuum 0 I t C O I I c I t I o o o o o f_d:twig crotches on plywood tray. . . i. -. . "LU. ,, Page 51 52 52 56 65 65 67 67 67 67 69 69 69 71 Figure Page 18. Classification of feeding response: F3 (weak: xylem not visible), F2 (medium: xylem visible), F1 (strong: xylem deeply penetrated) . . . . . . . . . . . 71 19. Design of emergence drum and ventilation system . . . . 75 20. Emergence drum with elm logs on steel rack. . . . . . 77 21. Ventilation system for emergence drums . . . . . . . 77 22. Bark sampling in twig crotch with cork borer No. 1 . . . 80 23. Bark samples from twig crotches . . . . . . . . . 80 24. Extraction and clean-up of bark samples. . . . . . . 80 25. Analysis of samples with gas chromatograph (Beckman GC-4). 80 26. Sample run on a Beckman GC- -65: DDE, DDT and methoxychlor . . . . . . . . . . . . 83 27. Bioassays I969: mortality in treatments I to V1. . . . 89 28. Bioassays 1969: feeding response in treatments I to V1 . 91 29. Bioassays 1970: mortality and feeding response on helicopter and mist blower treated elms . . . . . . 93 30. Bioassays 1969: correlation of 'failure to score xylem' and mortality for two sample periods . . . . . . . 95 31. Bioassays 1970: correlation of 'failure to score xylem' and mortality for two sample periods . . . . . . . 96 1 32. Bioassays 1970: average mortality at three height levels for helic0pter and mist blower-treated elms (2 and 8 weeks after application) . . . . . . . . 101 33. Sample cube fastened to elm branch . . . . . . . . 118 34. Designation of filter cube sides (I to 6) . . . . . . 118 35. Analysis of spray pattern on filter disk under long- wave UV. . . . . . . . . . . . . . . . 122 36. Fresh methoxychlor deposits on mist blower-treated elm branch. . . . . . . . . . . . . . . . 122 37. Helic0pter application: preportion of droplet sizes per unit area (cm2 ) on horizontal filter disks at two height levels of treated elm trees . . . . . . 125 xi Figure 38. Helicopter spray deposits on filter cubes (fluorescence photographs) . . . . . . . . . . 39. Mist blower spray deposits on filter cubes (fluorescence photographs) . . . . . . . . . . 40. Average methoxychlor deposits on the six sides of the sample cube at two height levels of treated elm trees: helicopter (a) 12.5% emulsion + Dacagin, (b) 12.5% emulsion; mist blower (c) 12.5% emulsion + Dacagin, (d) 12.5% emulsion . . . . . . . . . . . . . 41. Application of methoxychlor formulation on fiber board with pen-like applicator . . . . . . . . . . 42. Disassembled fiber board unit with plastic dish and screen for contact toxicity testing . . . . . . 43. Beetles in gelatine capsules No. 3 after exposure to treated fiber boards . . . . . . . . 44. Decomposition experiment; bark sampling of treated elm branches . . . . . . . . . . . 45. Bark deposit-mortality regression line calculated for helicopter-treated twig crotches: (a) two weeks after application, (b) eight weeks after application. . 46. Bark deposit-mortality regression line calculated for mist blower-treated twig crotches . . . . . . . . 47. Bark deposit-mortality regression line calculated for lab-treated twig crotches (Finney's probit analysis). . 48. Male and female feeding response on untreated twig ' crotches . . . . . . . . . . . . . . . . 49. Time-mortality curves for S. multistriatus after ex- posure to treated fiber boards: (a) deposits aged for 12 hours, (b) deposits aged for 4 days . . . . . 1 50. Structure of methoxychlor deposits on elm bark (a) solution: radial crystal, (b) emulsion: radial and plate-like crystals, (c) emulsion: large plate- like crystals, (d) emulsion plus Dacagin: uneven distribution at lower concentrations (whitish spots). . 51. Translocation experiment: designation of bark sections on twig crotch . . . . . . . . . . . . . . ' 52. Losses to Dutch elm disease on the campus of Michigan State University: 1958-71 . . . . . . . . . . Page 127 130 134 144 149 151 154 156 I60 I64 170 177 INTRODUCTION Dutch elm disease is an insect-borne fungus disease and is characterized by the close interaction of 3 organisms: the vector (or vectors), primarily scolytids; the pathogen (an ascomycete fungus), and almost all the members in the genus ylmgg. When dealing with control strategies of a vector-borne disease one has to understand the whole complex of the problem. Any change in one of the 3 inter- acting components will alter the performance of the whole disease system. Assuming that the amount of inoculum carried by a given pop- ulation is directly related to vector density, the performance of the disease system is determined by (a) the host tree density, and (b) the vector density. Indirectly, the vector provides for new breeding sites by inoculating healthy elms which subsequently become attractive for breeding. If undisturbed (no control measures), the cycle of inoculation and beetle infestation will continue until a threshold host tree den- sity is reached; then the disease will turn from the epidemic to the endemic stage. In other words, a minimum host tree density is a pre- requisite for the perpetuation of Dutch elm disease. Since the Euro- pean elm bark beetle, the most successful vector, has a certain flight range, it determines the threshold density. In other words, this den- sity is reached when the mean distance between host trees will exceed the flight range of the most successful vector. In areas where Dutch elm disease has occurred for the past forty years, such a threshold 1 2 density might be present already. For all practical purposes,if one were to wait until the balanced state of a threshold density were reached in a given area, the loss of the majority of all elms would be the probable result. One can imagine that the elm density in which the disease would become endemic would be very low considering the maximum flight range of 4 miles of Scolytus multistriatus Marsham (Wolfenbarger _£__l., 1939). Therefore, in order to protect and save high density urban elms, the disease transmission cycle has to be interrupted. The problem of disease prevention presents itself in the simple question: How can optimum control be achieved knowing the bio- logical features of all three interacting organisms? A simple answer would be to eliminate (eradicate) one of the three components. For instance, the elimination of the elm from the American scene would certainly solve the problem. This would be, of course, self-defeating since the overall goal is to save the American elm, one of the most ideal and common shade trees in the Northeast of the United States. Thus, besides resistance breeding of elms, control can come only from the manipulation of either the fungus component or the vector popu- lation per se or some combination thereof. Figure 1 schematically shows the interaction of the pathogen, the vector and the host tree in 'the disease cycle. The shaded circles stand for dying elms, the unshaded circle is a healthy elm. A vector emerging from a diseased tree will either go directly to another breeding site or feed in the twig crotches of a healthy tree. Eventually this feeding portion of the population will look for a breeding site. During the feeding pro- cess, the healthy tree becomes inoculated with the fungus, will even- tually die and provide new breeding sites for the vector population. .AE_o commom_u uo_oc_o ooomcm “E_o >Lu_moc uo_ot_o nmumgmcsv owes omoL ucm mamas» .L06um> mo :o_uomtmuc_ u293 ammom_n E_o goose ._ otam_u $3,023.."— ko 013.025."— DQJ‘QUIFOZUIU k0 ZO_FUSN_._.WMQ Q8 UZ_D.wwh+ n_On_ UZK—wWMM k0 JOMFZOU to JONFZOU .UZEUNNU T J x... L, In this way pertetuation of the disease is guaranteed. The small arrows depict the types of control and protective measures which can be applied to a diseased, beetle-infested tree and to a healthy, vigorous tree. Ever since DED has been a problem in the USA, vector control . has been considered the primary way to fight the disease. A great deal of research effort has been devoted also to the chemotherapy of the phytopathological phase. Only recently has a new injection technique in combination with a new fofmulation of the fungicide benomyl given some promise for successful therapy of already infested elms (North- eastern Forest Experiment Station, Upper Darby, Pennsylvania, Photo- story No. 19, 1971). Vector control can be directed either at the population as a whole or at the 'feeding' portion of the vector population which is actually responsible for the perpetuation of the disease by providing indirectly for new breeding sites. Possibilities for control of the vector population as a whole are in reality very limited, and a thor- ough effort would involve costs which are, for all practical purposes, forbidding. During the early stages of the fight against DED in the United States the attempt was made to contain the disease or even to eradicate it totally from this continent. The attempt failed, although it was the most complete control effort ever undertaken against this tree disease. Its goal was to eliminate the introduced primary vector, S, multistriatus, and the pathogen by scouting for diseased or beetle- infested elms and by destroying (burning) them. The eradication pro- gram might have worked in the urban communities but it failed in the rural and forested areas where diseased trees often escaped detection and necessary destruction. The woodlands Surrounding cities were and are, therefore, still a great pool for new vector p0pulations. A 5 iwathy might have an effective tree removal, sanitation and aim tree I atlen program, but the continuous supply of thriving vector pOpu- ffifiijna from neighboring woodlands is always a threat to the success {Wyn-planned control programs in the urban communities. I ‘.,. Vector control can be practiced along the following lines: ~7'Ap, Direct Control: U. 1) Control of breeding population - detection and destruction of u, beetle-infested elms, trapping of beetles. 2) Control of feeding pepulation - prevention of disease trans- mission during feeding process through protective spraying of healthy elms. , B. Indirect Control: 1) Intensive tree care (sanitation) to keep elms in healthy, vigorous condition. , 2) Detection, removal and destruction of possible breeding sites, .fi. such as dying elms, broken limbs, and so on. 1 . c .53) Silvicuitural control: regulation and manipulation of elm tree ',§.r,. density to a point where the probability of disease occurrence fin :4565“ vls greatly reduced. _ j. )3 35“,.“ Integration and combination of all of the mentioned control "?§;jii§§ufesbis necessary in order to provide a well-rounded control system. . ,f;-£§oaearch reported in this thesis dealt with questions concerning ‘c ‘ . k“ .-.. ’ .1 \i71_ éfiei of the "feeding" population of S. multistriatus to prevent g;*§gansmission in healthy elm trees. 6 I. Aspects of the Feeding Behavior of Scolytus multistriatus Marsham. The location of a feeding site (inoculation point) on the tree region determines the probability of infection and also possible re- covery. lnoculations were far more effective when close to the main trunk than at the outer branches (Zentmyer _£ 31., 1946). This means that for adequate elm tree protection one has to take into account not only where the vector feeds on the tree but also possible hazard regions of infection during the period of susceptibility. Therefore, a valid evaluation of elm protection systems must be based solidly on a com- plete understanding of the behavior of both the vector and the pathogen. Not all vectors emerging from a diseased tree make successful inocula- tions. Some go directly to a breeding site and never get involved in direct disease transmission. The current aim of protective spraying is to prevent inoculation by killing the vector at the feeding site, the 1-5 year old twig crotches. Their post-emergence flight carries part of the beetles into the canopy of healthy elm trees where they commence feeding. Wolfenbarger and Buchanan (1939) and Whitten (1958) found most of the feeding scars in the upper and peripheral part of the tree; beyond that no information is available as to regions of pref- erence for feeding within the tree crown. In order to maximize pro- tection against feeding, it is necessary to know where feeding takes place primarily. The chemical aspect of stimulation for feeding has received considerable attention and several phenolics have been identified and verified as such by bioassays (Baker and Norris, 1967). Physical stimuli, such as twig crotch angle and certain bark characteristics, were believed by Meyer and Norris (1964) to influence the feeding attack. The potential of physical characteristics as a basis for elm resistance v—f breeding warranted additional research on factors determining selection 7 of feeding sites by beetles. 11. Evaluation of Methoxychlor.Formulations and Application Technigues Against Feeding by Scolytus multistriatus Marsham. Ever since Scolytus multistriatus and Scolytus scolytus (in Europe) were established as the main vectors of DED, spraying of elm trees has been used extensively to prevent feeding by these vectors. The questionable value of spraying was recognized as early as 1931 by Fransen: materials then available lacked the persistence to be effec- tive during the period of transmission, and application techniques were not available which would provide for complete coverage of all feeding sites on a tree. Initially, arsenicals were recommended; later with the development of the organochlorines, DDT became available and was successfully used for prevention of twig crotch feeding (Becker, 1946; Plumb, 1950). Fall and Spring applications before the leaves appeared were almost equally effective because of the long persistence of DDT (Whitten, 1949, 1960). During recent years, concern over environmental pollution made the continued use of DDT in urban areas increasingly difficult. Trials with many other insecticides such as Lindane, mala- thion, etc., proved that DDT was still superior in all respects. During the 1960's, DDT was phased out of most urban DED spray programs and systemic insecticides received considerable attention for controlling DED vectors (Norris 1959; Peacock,l967). Unfortunately, no systemic proved effective enough to replace DDT. Although earlier tests did not favor methoxychlor as a possible alternative to DDT, investigations by Becker (1956) and Wooten (1962) established methoxychlor as a substitute for DDT. What made methoxychlor so appealing was its low mammalian toxicity and environmental safety. Serious problems arose, however, 7‘~n “A” A I I '| ' 8 with timing of the applications during the dormant season. It was generally assumed that the lower persistence (Norris, I961; Wallner and Leeling, 1968) of methoxychlor would render fall applications in- effective. Consequently only spring applications were recommended. This put great pressure on extensive Spray programs because not enough man- power or equipment were available to complete the spraying in time be- fore bud break. Applications made from the ground, either with the hydraulic sprayer or mist blower, were naturally slow and expensive, so cheaper ways were scught. Wallner and Leeling (1968) pioneered the use of the helicopter for elm tree spraying. Aerial application proved to be cheaper and much f faster once the pilot became familiar with the location of the elm trees in an area. However, the effectiveness of helicopter spraying in com- parison with mist blower application remained in doubt, and further ex- perimental evidence was needed. The Dutch elm disease symposium in Dela- ware, Ohio (Peacock, 1967) placed research priority on methoxychlor for- mulation studies and application techniques. In 1968, Wallner _£ _l. (1969) compared and evaluated several methoxychlor formulations which were applied by helicopter to the campus elms of MSU. This study was continued and expanded during 1969 and the following years. The per- formance of formulations and application techniques was evaluated by bioassays with S. multistriatus and chemical analysis of bark residues. The results are reported in this dissertation. 111. Monitoring of Spray Covgppgg on Elm Ttggg. Two factors made a valid comparison of formulations and appli- cation techniques difficult in 1969 and 1970. One was the great varia- bility in bark residues and the other was carry-over from previous app- lications. A more objective way was sought to study the overall coverage ' ‘l.‘ C a 9 of an elm tree when the material was applied either by helicopter or mist blower. A sampling method was developed which permitted study of spray patterns, actual spray deposition and 3-dimensiona1 coverage. IV. Insecticidal and Decomposition Properties of Methoxychlor Formulations. Extraction and quantification of bark deposits have frequently been used to test the performance (coverage and persistence) of spray applications against DED vectors (Matthysse _p _l., 1954; Thompson, 1965; Wallner and Leeling, 1968). Feeding trials with S. multistriatus ser- ved the same purpose (Doane, 1958, I962; Norris, 1960). Only a few workers attempted to correlate beetle response with residue levels in twig crotches (Miller, 1951; Matthysse pp 21., 1954). Information of this kind was lacking for methoxychlor. In order to establish residue levels at which control is sufficient, beetle response (mortality) on field-collected tWig crotches was correlated with the respective resi- due values. LC50 values were also determined in bioassays with lab- treated twig crotches. During the search for a feeding site on a treated elm tree, S. multistriatus comes in contact with surface deposits of methoxychlor. Although the contact toxicity of methoxychlor is known to be low (Cuth- bert 2; 31., 1970), it was of interest to compare contact toxicity of various methoxychlor formulations. Several experiments were conducted to investigate the effect of contact with methoxychlor residues on feeding behavior. The toxicity and decomposition rates of a compound are largely dependent on its deposit structure. Therefore, crystalli- zation patterns on elm bark were documented for various formulations and an attempt was made to point out possible relationships between con- tact toxicity, weathering and type of surface residue. Decomposition rate is, to a great extent, predetermined by the makeup of the formulation 3% 10 g, .ld‘lm Disease Control on the Cam-us of Michi-.n State Univer- . I. f 1‘ ';{(,£i;h§ history of DED and control measures on the campus was docu- ,fI'for-the period since 1958. The successful elm protection pro- Vrlt Michigan State University is cited as an example of excellent .I‘ ‘ fifiug} gwlon between the Grounds and Maintenance Department and the id;eu;»§ment of Entomology. 8‘ Ts"? "1‘ * .1" 1‘ p‘a'? .‘ . . é" . tcnr‘ w o \ a: . “.(‘LJ " LITERATURE REVIEW (A) History and Development of Dutch Elm Disease in Europe The name Dutch elm disease implies incorrectly that the origin of this fungus disease is Holland. Actually, elm die-back was observed independently in several areas of Europe at about the same time. Spier- enburg (I921) discovered the disease in Holland while Guyot (1921) des- cribed similar disease symptoms on elms from Northern France and a re- port by Georgescu and Orenski (I957) mentions that the disease had been found in Rumania as early as 1910. The question as to the origin of the disease cannot be answered conclusively. The region of origin is believed to be the Far East, in particular, China. The manner in which it was introduced into Europe is not quite certain. During the past century travelling Chinese tradesmen who roamed Europe used baskets made of branches of their own native elm, gimp; parvifolia. Some of these branches might have been infested with the disease agent (Heybroek, personal communication). A further reason to suspect China as the place or origin is the fact that the Chinese elm, S. parvifolia, is rather resistant or tolerant to the disease. This low level of susceptibility can best be explained by a long evolutionary parallel deve10pment of both host tree and disease. The Dutch elm dis- ease problem might be analogous to the chestnut blight caused by Eggp- Ebifl parasitica, a fungus, (definitely of Asiatic origin) in the sense that the Chinese chestnut is also resistant to the blight, whereas other 11 \ 12 chestnut species are very susceptible. The actual cause of the disease was unknown until Schwarz (1922) isolated a fungus from diseased elm wood which she later described as Graphium Eflflfln Ascomycetae. Schwarz speculated that the fungus entered the tree through the leaves and con- siderable controversy developed over the cause of the elm die-back. Some attributed the disease to unfavorable conditions in the soil (Pape, 1924) while others held frequent moisture changes in the soil responsible (Hoestermann, Noack, 1925). The large European elm bark beetle, Scolytus scolytus F. and the smaller European elm bark beetle Scolytus mplgi: striatus Marsham, together with other Scolytids, were also suggested as the primary causes of Dutch elm disease (Malaquin, I923; Knop, 1928; Kaiser, 1931). The question about the cause of the disease was finally resolved by Wollenweber (1927) and Buisman (1928) who both proved Graphium ELEL Schwarz to be the pathogen. Schwarz (1922) described only the imperfect stage of Graphium pypl; but in 1932 Buisman found the perfect stage of the disease agent and named it Ceratostomella ulmi (Schwarz) Buisman. A further name change to Ophiostoma HEEL (Buisman) Nannf. was made by Nann- feldt (Melin and Nannfeldt, 1934). A revision of the group by Moreau (1952) and Hunt (1956) yielded the name, Ceratocystis 31ml (Buisman) C. Moreau, which is now commonly accepted. Even though the causal organism was isolated and identified, the method of the transmission from one host to another still remained un- solved. Wollenweber and Stapp (1928) mentioned birds and beetles as possible vectors. However, the first evidence of a causal vector was obtained by Wollenweber (1929) who observed S. plgi in the galleries of S, scolxtus in diseased elm wood. Subsequent research in Holland (Betrem, I929; Roepke, 1930; Fransen, 1931), in Germany (Prell, 1930) and in -m: 13 other European countries, verified S. scolytus and S. multistriatus as the main vectors of the disease. Using S. colytus, Fransen (1931) obtained experimental proof that the fungus is transmitted during the feeding process shortly after emergence. In fungus-infested elms, S. glml fruits in galleries of S. scolytus and S. multistriatus and will form coremia (fruiting bodies). The sticky spores adhere to the body integument or are carried internally in the digestive tract of emerging beetles (Betrem, 1929). Both the larger and the smaller European elm bark beetle have basically the same feeding behavior. After emerging they feed in the young twig crotches of healthy elms and introduce fungus spores into the vascular system of the tree. (B) Introduction of Dutch Elm Disease into the United States and Present Status Both European elm bark beetles were probably introduced in the United States; however, only S. multistriatus could establish itself. The first record of its occurrence in this country, Massachusetts, dates back to 1909 (Chapman, 1910). Not until 1930 were the first cases of Dutch elm disease reported from Cleveland and Cincinnati, Ohio. These localized and relatively small infestations were soon eradicated. In the early Summer of 1933, the disease appeared in New Jersey. It was soon discovered that all of these recent infestations were caused by diseased elm burl logs shipped from France to New York and from there to several veneer factories in the country. Many of these logs were also infested with the 2 European scolytid vectors, S. scolytus and S, multistriatus. Federal quarantine No. 70 from October, 1933, re- stricted the further importation of elm logs into the United States (Beattie, 1934). This regulatory action came too late to prevent the . iii. , i I4 establishment of the disease and its main vector, S. multistriatus. S. multistriatus was soon joined by the native elm bark beetle, Hylurgopinus rufipes (Eichh.), as an important vector, especially in areas with lower winter temperatures (Collins, 1936, 1941; Finnegan, 1957). The distribution of the introduced vector S. multistriatus and the disease pathogen covers at the present almost the entire range of the American elm, Qlflfli americana. As of 1970 infestations were found in all states except Arizona, Florida and Montana (Barger and Hock, 1971). It is common belief that the disease will eventually Spread to wherever elms grow in North America. (C) Hosts in Europe and North America Practically all of the species in the genus QlEEE are susceptible to varying degrees to S. 31ml. Asiatic Species in general and the Chin- ese elm, gimp; parvifolia, and the Siberian elm, gimp; ppmllg L. show considerable resistance if not immunity (Buisman, 1931, I933, 1934). Similarly, S. multistriatus attacks all the elm species within its range in Europe, in the United States and Canada. In the United States this beetle has never been reported from hosts other than from members in the genus Ulmus (Wallace, 1940), although in Europe it has been recorded from aspen, plum and ash (Escherich, 1923). According to Becker and Mankowsky (1965) there was little or no significant difference between feeding response on gimp; americana and other elm species, including the fungus-resistant Chinese and Siberian elm. (D) Other Vectors in North America The problem of other possible vectors of Dutch elm disease besides S, multistriatus received early attention. Readio (1935) stated that all the insects associated with elm trees must be suspected as vectors. These are the native elm bark beetle, 5. rufipes, the curculionids, 15 Magdalis ssp., the cerambycid, Saperda tridentata, buprestids, cossids, and even leaf feeders, such as the chrysomelid elm leaf beetle, cankerworms, tent caterpillars and several sapsuckers. Significant disease trans- mission was su5pected by the bark and wood borers, particularly by u. rufipes (Clinton and McCormick, 1935). Even certain Acarina were be- lieved to be capable vectors (Jacob, 1934). Collins (1936) performed several transmission experiments on small elm trees with insects which emerged from or fed on diseased elms. Only S. multistriatus and fl. rufipes proved to be successful vectors and no infections occurred with the wood borers Saperda tridentata, Magdalis armicollis and fl. barbita, or with the leaf feeder Galerucella xanthomelaena, Chrysomelidae. Goeden and Norris (1963) failed also to establish Magdalis ssp. as vectors. Collins (1941) isolated S. 31ml from several wood-boring species which developed in diseased wood. Based upon the total number of emer- ging insect Species from several logs the following percentage carried fungus spores: Scolytus multistriatus 6.9, 5.8, 7.7. 5.7% Hylurgopinus rufipes 4.3, 2.4, 3. , 0.7% Xylobiops basilare Xylosandrus germanus } small % with spores Magdalis armicollis and others The same author cautioned against underestimating fl. ufipes as a vector of Dutch elm disease, eSpecially in areas where the European elm bark beetle is absent or cannot establish itself because of low winter temp- eratures. The geographical distribution of the indigenous elm bark beetle follows closely the natural range of its host trees, Slmpg ssp. Finnegan (1957) also believed that S. rufipes deserved more attention considering the fast spread of Dutch elm disease in Quebec where it is the only vector. A detailed description of the biology is given by l6 Kaston and Riggs (1937) and Kaston (1939). Like S. multistriatus, fl. rufipes is a secondary pest and oviposits only in sickly or dying elms. Before constructing galleries, fl. rufipes makes feeding tunnels in healthy elms which rarely reach the xylem. The adults hibernate in bark tunnels which penetrate deeper into the bark than do the feeding tunnels. It is commonly assumed that the disease is transmitted during the con- struction of these tunnels (Kaston and Riggs, 1938). Disease Spread does not depend solely on vectors as Verrall and Graham (1935) pointed out. The pathogen itself can pass from a diseased tree to a healthy tree via root-grafts. This can occur in situations where elms grow close together as in roadside-plantings. The authors mentioned aS evidence for root-graft transmission the fact that dyes passed readily from one tree to the other via the root-grafts, and fur- ther pathogenic discolorations in the vessels of the roots reached into the root system of neighboring trees. (E) Dutch Elm Disease Pathogen: Ceratocystis pull (Buisman) C. Moreau The literature on the phytopathological phase of Dutch elm disease is very extensive and will not be reviewed here since Elgersma (1969) gave a comprehensive review of the subject matter. Instead key references will be given. Vectors introduce fungus spores into the conductive tissue while feeding on the host tree. After inoculation, the infection spreads rapidly as passively moving spores in the sap stream. Invasion of the vessels by S. p191 results in a brownish dis- coloration of the vessel walls. The fungus moves then through pits into adjoining vessels and finally causes complete obstruction of vessels through the formation of gums and tyloses. The water and nutrient supply to leaves is interrupted, and this results in wilting and yellowing of .J‘iiiiiiigi; 17 the leaves (Pomerleau, 1968; Buisman, l932b). It was also suggested that toxin formation alone or in conjunction with tylosis and gummosis is responsible for the pathogenicity of S, plmi (Zentmyer, I942, 1946). A recent study by Landis (1969) supports this conclusion. Wollenweber (1929) found the fungus to be vertically distributed, mostly in the fylem part of the youngest growth ring. Banfield (1968) investigated seasonal difference in movement and distribution of Spores in the con- ductive system of elms and found that they moved only a few centimeters away from the inoculation point between September and April. Suscep- tibility to infection reached a maximum during early July because of the rapid movement of the yeast-like spores in the spring wood vessels (Parker, 1941; Banfield, 1968; Pomerleau, I965). The location of the inoculation point determines the rate of development of disease symptoms. Trunk inoculations were far more effective than inoculations in the upper branches and twigs (Zentmyer _S pl., I946). The number of spores entering the vessel system was definitely correlated with symptom de- ve10pment: the higher the spore load the higher the pr0portion of leaves wilting. Multiple inoculations produced a higher percentage of disease incidence and also a greater percentage of leaves wilting, than did a single inoculation with the same number of spores (Zentmyer _£_gL., 1946). S, plmi was cultured from 6-8% of feeding scars made by naturally infested S. multistriatus. Transmission experiments with artificially infested beetles yielded 40-50% S. Eflll cultures. The fungus became established in many feeding scars without leading to further disease deve10pment, which indicates that many inoculations due to feeding never lead to true infection (Parker SE 31., 1941). The pattern of feeding in the twig crotch seems to influence effective inoculation. W7 18 According to Ouellette (1962), wounds along the lateral part of the twig crotch showed a higher true infection rate than did injuries in the center of the crotch. (F) Biology of Scolytgg MultistriatpS in Michigan Readio (1935) and Wallace (1940) gave a detailed description of the life cycle and breeding habits of S. multistriatus in the United States. The adults, Strongly attracted to recently killed elms or dying elm branches, breed in the trunk portion, the main branches and also the smaller twigs if their diameter exceeds 2.5 cm. The female bores into the bark and excavates a nuptial chamber, after which the , maternal gallery is constructed upward along the grain. The female carves egg niches in both sides of the gallery and lays an average of 68.5 eggs (Wallace, 1940). The males help in the construction and cleaning of the egg gallery and the developing larvae make their own tunnels perpendicular to the maternal gallery. The larvae go through 5 moults. Following the last larval instar, they leave the cambium region and bore into the outer bark region where pupation occurs. De- pending on geographical area and temperature conditions in a given year, S, multistriatus will have 1, 1% or 2 generations per year. Fox (1958) investigated the emergence pattern of the smaller EurOpean elm bark beetle and concluded that there are 2 complete generations per year in northern Michigan. The first peak occurs in late May or early June and the second pOpuIation surge takes place between late July and late August. In the northern part of the lower peninsula Truchan (1970) found that the survival of the majority of the second generation was low. This suggests that there may be 1% to 2 generations per year under Michigan conditions. This knowledge is very crucial for the prOper r I ' no ' , - v 'a » \ fl ‘ .&7 :E. ‘ l9 timing of elm tree protection. After emergence a fair portion of a population will feed in the twig crotches of healthy elm trees before breeding places are found. During this feeding process the healthy elm becomes infested provided the beetle was contaminated with spores of S, ulmi. (G) Feeding and Breeding Behavior of Scolytus MultistriatpS The peculiar feeding behavior of S. multistriatus is the main cause for the effective transmission of Dutch elm disease in the United States. When the disease arrived in the United States, the basic mech- anism of the vector-pathogen-host interaction of this disease problem had been elucidated by European workers. Fransen (1931) was the first to link disease transmission with the twig crotch feeding of elm scolytids. When research on S. multistriatus as a vector of Dutch elm disease star- ted in the United States, the following scientific evidence was available from previous European work (Readio, 1935): l. Beetles are attracted to dying elms for breeding purposes. 2. S, 31ml produces coremia in the galleries of S. multistriatus. 3. Emerging beetles carry the gelatinous spores externally and internally. 4. Freshly emerged beetles exhibit some sort of I'maturation feeding” which can last up to 7-11 days. During this post- emergent feeding in young twig crotches of healthy elms the tree is inoculated with the fungus. Since this feeding is the point where infection occurs, the factors influencing it were of particular interest. Collins (1938) stud- ied feeding injuries on 500 elm trees, but he was unable to find a cor- relation between the number of feeding scars and the diameter or age of 20 the attacked twigs. Wolfenbarger and Buchanan (1939) found that the smaller, lateral member of the injured crotch varied from current sea- son's growth to 3-year old twigs and the age of the larger twigs ranged from 1-5 years. In addition, they gave a detailed description of the actual feeding injury. The adults of both sexes bore into the center or into the somewhat lateral part of the twig crotch where they make holes up to 5 mm in length and about 2 mm in width. The attacked crotch reacts with extensive callus formation around the scar. If the injury was too severe and the scar too deep, then the lateral twig often broke off completely. These two authors character- ized borings in the twig crotches as actual ”feeding“ and not simply a search for a breeding place. They obtained this conclusion from experi- ments where feeding beetles lived 7.6 * 2.6 (SE) days as Opposed to 2.2 i 0.4 days when not feeding. Beetles of each sex of S. multistriatus were confined in glass ball jars together with fresh elm twigs and logs. The average number of feeding injuries per beetle after 7-14 days were 0.91 for males and 1.03 for females. Log injuries were considerably lower with 0.13 for males and 0.33 for females. The question 'if and how' S. multistriatus is attracted to a healthy elm tree for feeding purposes received wide research attention. Upon examination of a rather extensive elm tree population, Collins (1938) was unable to discern a regular pattern of feeding among elm trees. Some trees had numerous feeding wounds, whereas adjacent trees had few, thus indicating that injuries within the elm population were erratic and clumped. Collins (1938) concluded therefore that 'centers of attraction' are responsible for this erratic but clumped feeding pattern. Wolfenbarger and Jones (1943) distinguished between feeding \ o C» 'I '—-——*—-——— —~— 21 near the dispersion point (beetle productive tree) and feeding near the convergence point (beetle attractive tree), but in terms of feeding , they attributed more importance to the diSpersion point. From field observations they concluded that beetles attack twig crotches near dis- persion points regardless of the immediate proximity of breeding material. The feeding injuries decreased as the distance from the dispersion point increased. From these field-collected data, Wolfenbarger and Jones (1943) calculated a maximum flight distance of about 4 miles. However, it has been demonstrated that wind forces are able to aid the dispersal of small insects such as scolytids, to places which lie far outside the natural flight range (Felt, I935). The importance of wind drift for the dispersal of S, multistriatus and local outbreaks of the disease were demonstrated in Connecticut and other states by releasing balloons and following their flight path (Felt, 1937). Wallace (1940) attributed dis- persal beyond the natural flight range to transportation of beetle- infested material. Of interest is his statement that under ordinary conditions, flight range of S, multistriatus is limited to about one- quarter of a mile. Field observations taken in 1939 indicated that only a small percentage of adults feed on elm twigs at any time. Feeding is by no means obligatory for successful breeding Since freshly emerged beetles commenced to construct brood galleries within 24 hours when given the Opportunity (Wallace, 1940). More recent research indicated that the search for a feeding Site is more random and that healthy elms have no olfactory attractiveness to S. multistriatus. Unlike many other scolytids, S. multistriatus exhibits no positive anemotactic response; that is, flight after emergence is initially not directed against the wind (Meyer and Norris, 1964). The emerging beetles fly into the canopy —-—.-— 22 of nearby trees and if they encounter a specific feeding stimulant in the bark, they start to feed. Repellents in the bark of other tree species or lack of the attractive compounds reinitiate flight until a healthy elm is found. These feeding stimulants were extracted from bark of S. americana and characterized as p-Hydr0xybenzaldehyde and other phenolics (Baker and Norris, 1967; Baker _S _l,, 1968). Norris, at the Dutch Elm Disease Symposium in Delaware, Ohio, in 1967 held that the total feeding response of S. multistriatus involves physical as well as chemical characteristics of the bark. Meyer and Norris (1964) pointed out that crotches with more acute angles and rough bark are preferred over those with less acute angles and smooth bark but no data were given to substantiate the above statement. Dixon (1964) placed more emphasis for the induction of a feeding response on thigmotactic stimulation. Attack response on logs of several unrelated species, including elm, seemed to favor bark with many fissures. In cage experiments twig crotches from pear and elm were almost equally attacked. Dixon (1964) concluded that thigmotactic stimuli, bark fissures in the trunk region and twig crotches on the smaller branches, were more conducive to attack than chemical Stimuli. With respect to effective control of feeding beetles it is nec- essary to know about the pattern of feeding and the distribution of feed- ing injuries within an elm tree. Wolfenbarger and Buchanan (1939) found most of the scars in the upper and outer parts of the crown. Whitten (1958) stated also that feeding occurs predominantly at the periphery of the top portion of the tree. The total number of twig crotches in- creases approximately at the same rate as does the diameter of the tree (Wolfenbarger, 1940). HOWever, tree location has an overriding influence 23 on the total number of twig crotches. Therefore, the above author could not give a general formula for computing the number of twig crotches based upon the diameter of a given tree. While no direct attraction seems to be involved during the post- emergent flight to a feeding place, S. multistriatus responds strongly to dying or weakened elm trees as possible breeding places. Chemical changes in the tree during the process of dying, or during extreme stress situations such as drought render the tree very attractive to bark beetle attack. This seems to initiate a strong olfactory reSponse in the beetle (Meyer and Norris, 1967). Even pruning wounds on otherwise healthy elms will attract beetles due to the volatility of degradation products in the wound (Peacock, 1967). Hart _S 21- (1967) reported an increase in elm die-back caused by S. pup: on trees which had been trimmed during the previous summer. The authors explained this as an obviously increased attractiveness which was due to the strong odor exuding from trimming wounds and the weakening of trees by heavy pruning. The beetles infested the trees in their attempt to breed near the pruning wounds. Goeden and Norris (1964) theorized that the “lag between cutting and initial attack by beetles represented the necessary time for the production of the volatile, attractive degradation products.” According to Meyer and Norris (1967), elm logs were attractive to beetles for a period of 18 hours to 16 days after cutting, with a peak attractiveness on the 11th day. They attributed the decline in attractiveness to a decrease in the amount of the olfactory stimuli exuding from the bark. Further attract- ancy studies with logs plus females, versus logs alone, and logs plus males revealed that significantly more beetles were attracted to the logs with females. However, the high correlation between the number of 24 attacking beetles (regardless of sex) and the total number of attracted beetles led them to conclude that the attractancy had its origin in the decaying tree tissue and not in beetles per se. This conclusion is 'n contrast to the secondary beetle-associated attraction demonstrated by several researchers in other scolytids (Anderson, 1948; Wood, 1962). Martin (1935, 1938, I946) inve tigated effects of sunlight, phloem condition and moisture content of phloem on infestation rate by the beetle. Most of the beetles entered during the first 4 weeks after cut- ting and the total infestation period did not last longer than 10 weeks. Phloem older than 4 weeks had lost its attractiveness to the beetle almost entirely. Moisture content did not appear to influence the extent of the infestation unless it was associated with the phloem condition. Also, logs exposed to sunlight were preferred over logs in the shade. (H) Control Measures 1. POpulation Control When the first Dutch elm disease cases were reported from Ohio in 1931, the seriousness of the problem was realized and, since the in- fested area was small and localized, diseased trees were eradicated. May (1931) outlined eradication plans and asked for state, federal and private cooperation to initiate a survey for Dutch elm disease infec- tions. In this paper he gave practical instructions for the removal and burning of all diseased trees. Felt (1934a,b) stated that the control of the disease is “largely an entomological one, since the most feasible method of checking this disease, so far as known at present, is by re- ducing the abundance of the carrier.” But he recognized that locating and destroying diseased trees failed to check Spread because of several difficulties involved. Scouting for infested trees was relatively 25 simple in urban areas, but in rural areas and in woodlands, detection and removal were difficult. Above all, the control costs were too high, and generally state and federal funds were inadequate to support an ef- fective eradication program. The following time table for control meas- ures was advocated by Felt (1934 b) and he urged that a shift of em- phasis be made from tree removal to sanitation and other protective measures: Winter (until April): cut and burn sickly or dead trees or parts 7 thereof (removal and sanitation) May: spray against feeding of Ist beetle generation with lead arsenate or Bordeaux mixture; adjust treatment to control leaf feeders. 1 August (mid): spray against feeding of 2nd beetle generation ‘ April to November: fertilize weak trees to increase vigor ' The primary objective at that time was the elimination of the d i Sease in the United States, although discouraging reports about the ‘3 "<=>gress of the eradication program became more frequent (Worthley, I 935, 1936). Two statements from 1934 expressed the pessimism of many ' lb‘EP‘ZZDPIe who were involved at that time in this large-scale control pro- 9 ram- ”We must learn to live with the disease” (Felt, 1934 b) ”Only a scientific miracle can save the American elm'I (Dr. 0. N. Lining in Felt, 1934 b) ‘r 1 ‘7“‘E= basic avenues for combatting this problem were already laid out Cl 7" "'ing the incipient stages of disease Spread. I Removal and subsequent destruction of diseased and beetle-infested § I h m trees is as important today as it was 40 years ago when the disease 3 hfi rted in Ohio. Usually the infested trees were burned, a practice 26 which is still recommended. Sanitation, that is, keeping the tree in good health by pruning wilting or dead branches, requires active and Well-trained personnel (Felt, 1934 b). After pruning, the wounds Should be painted with coal tar creosote to prevent infection by the beetle (Parks, 1936). Recently, the effectiveness of pruning was questioned by Hart (1970). Although removal of the diseased portion of the elm tree led to complete recovery in 19 out of 84 attempted cases, he concluded frtxn this low success ratio (only 23%) and the high pruning costs that tf1is control method is impractical. Measures such as increasing or maintaining the vigor of the tree t>§l fertilization or Spraying against insect defoliators were considered llsieaful in preventing successful colonization by S. 3121 (Felt, 1934 b). ‘ff1ea possibility of using trap logs of the two Dutch elm disease vectors, 55.- multistriatus and fl. rufipes, was investigated by Whitten and Baker ( I 5339). In 1935 the tests included felled trees, chemically killed t "eaes, girdled trees and elm logs lying and standing on the ground. () '7 these the felled trees were most attractive to Scolytus and Hylur- :asstiiiflfli- Further trap log studies in 1937 and 1938 revealed that the t "”E=es killed with sodium chlorate were most attractive. The authors did r":>’1: draw any conclusions with regard to the control potential of elm- wQQd traps. To prevent emergence of bark beetles from elm wood, Becker (:1. 15’46, 1947) applied DDT and DDD Sprays. A 5% oil solution of DDT gave E3“ ..8% control while the wettable powder and the emulsion sprays were I Applications of DDT and §SS effective. DDD sprays were not effective. '30 . . . .3. sprays to prevent scolytId InfestatIons on elm logs were more Suc- Q Qssful when No. 2 fuel oil solutions were used. Prevention of infes- 1: is. ition ranged from 99.7% to 100% for DDT treatments at concentrations 27 of 0.0625%.to 5.0% Of actual material as compared to the untreated con- trol. No. 2 fuel Oil alone gave 100% prevention while orthodichloro- benzene when mixed 1:8 with NO. 2 fuel Oil also gave complete prevention. Thirty-six chemicals were introduced by Himelick and Neely (1961) into the sapstream of living elms in order to prevent vector breeding. Sodium arsenite (applied in ax frills) gave complete control in healthy trees and 97% prevention when trees were in the early stages of infection. Scudium iodide was somewhat less effective, but because of its lower mam- rnaalian toxicity it should be preferred over sodium arsenite (Himelick and Neely, 1963) . Over recent years the interest in alternatives to chemical con- tir‘c>l has established a firm place for biological control in many control F>l‘<)grams. The deve10pment of new, more selective pesticides made it FWC>sssible to integrate such contrasting concepts as chemical and bio- l‘:>.<_;,lical control. Today, the term integrated control comprises all pos- ‘5 i t>le control measures, used in combination and adjusted to each other, 3 r1 order to bring about Optimal control of the target organism. Bosch and Stern (1962) summarized the literature on this new control concept and gave a critical evaluation of the subject. Chapter 17 in Insect- £325§E£§LSManagement and Control (Subcommittee on Insect Pests, National Acad- fiaujr‘iyc of Sciences, 1969) is a more recent and complete discussion of this t:<:’l':->ic. Dutch elm disease is an example where only the combined use (3"=r several control strategies will lead to the desired goal of reduction Q "=7 disease incidence. Biological control has been neglected for the Q 0 <:)"‘Itrol Of Dutch elm disease vectors for many years. In 1964, a braconId 5‘ ss;p, Dendrosoter protuberans, a native parasite of Scolytus multistriatus W ‘5. as shipped from France to the U. S. Forest Service Laboratory at Dela- “Vie: . . . . . r‘e, Ohio. Further shIpments went to MIchIgan State UnIverSIty at 28 East Lansing, Michigan, in 1965. Valek (1967) and Truchan (1970) reviewed the literature dealing with the natural enemies Of S, multistriatus and discussed the potential of the indigenous and introduced natural enemies in terms Of pOpulation control. Early prominence was given to resistant or immune elm Species or elm varieties. Buisman (1931, 1932, 1934) started a worldwide col- lection of elm Species in Holland. The Asiatic Species, in particular tflie Chinese elm, S, parvifolia, and the Siberian elm, S, pumila L. and i'ts variety Ulmus pumila pinnato-ramosa Henry showed high degrees of resistance. Resistance breeding began in 1928 in the Netherlands (Westerdijk, I 59:28). The first resistant clone was released in 1937, but it was sus- Cltatotible to another fungus, the coral Spot canker, Nectria cinnabarina ('1F<3de x Fr.) Fr. The next clone from 1948, though resistant to Dutch e'll‘ndisease and the coral spot fungus, did not meet requirements in rQSpect to growth rate and form. In 1961 the 'Commelin' elm was re- leaeased. It had less resistance but it had the desired growth and shape characteristics (Heybroek, 1961). The highly resistant 'Groeneveld' elm appeared 2 years later but this clone had a limited growth rate (Hey- b r‘Oek, 1963). pin the United States, resistance breeding programs were not "I'itiated until recently. Selection programs are under way at several .I":><:ations. Neither in EurOpe nor in the U.S.A. did possibilities of re- 53 'i Estance breeding against vector feeding receive much attention. Heybroek (1969) gave the following reasons why breeding for ‘Eabtztor-resistant elms was not further investigated: (a) there are too tries. rny vectors involved, (b) forcing beetle attack is too difficult, and el- 5 ... ‘1‘ .6 AJ an » P\.I 29 (c) disease resistance mechanisms are readily available in elms. The question of elm resistance to S, 2191 was fully reviewed by Bingham _S.gl, (1971). The recent discovery of a new, more aggressive strain of S, ulmi from southern England indicated the possibility that already released resistant elm clones might again be threatened by the disease (Holmes t al., 1972). The newer deve10pments in insect control might find useful apualication in the fight against Dutch elm disease vectors. The poten- t:ial of sex attractants (pheromones) given Off by virgin females when they enter an elm log for breeding is under scrutiny at present at the lJ - S. Forest Service Laboratory in Delaware, Ohio, and at the University (D‘Fr Syracuse, New York. A sex attractant could be used effectively in c3<>Ir1nection with a Sterilant. 2. Chemotherapy against fungus A whole generation of pathologists has worked on the chemotherapy (>"f: DED infections. It is mostly a problem Of getting the active material 3 'W‘tha soluble state and introducing it into the conductive tissue of the tree. The extensive literature will not be reviewed here. Mention ‘“' 3 'Il be made only that a new tree injection system in combination with a: '7‘ effective fungicide showed promise in curing infected elm trees ( l“Icotection was obtained with the higher concentrations, but the dis- advantage was their phytotoxicity to neighboring trees. Feeding trials 3 ndicated that emulsions were more effective than wettable powder for- mu lations. Preliminary experiments with a mist blower (Buffalo turbine ‘3 I ower) resulted in good coverage of the lower portion Of 60-65 ft. €31] rns while the upper part had little coverage. Concentrations Of up to 1'.55c% were used either in the form of Oil solutions or water emulsions and t 31. (1949) ':::’est coverage was Obtained when there was no wind. Dimond a 1 so experimented with DOT formulations (0.5, 1.0 and 2.0 lbs per gallon). l 3‘: field tests in 1947, up to 1 gallon was applied per tree. Feeding t Qsts showed little control in tOp and center portions Of the treated t fees but it was concluded that the dosage per tree was too low. The 7:7‘:>']lowing year an average of 2.6 gallons was used per tree. Good con- t "Ql, also in the upper tree region, was Obtained as a consequence Of the increased amount Of Spray per tree. The Elm Bark Beetle Conference ( Sd':_anlon, 1948) discussed elm protection in detail. Enough evidence 31 was accumulated to prove that only prefoliar Sprays gave satisfactory protection. Further, it was felt that 100% coverage was necessary since one feeding vector might be enough to kill the tree. Dimond _S_gl, (1948) attempted to give exact guidelines for the amount of spray for a given tree and calculated a regression between dosage and diameter Of the tree. Because Of the poor correlation be- tween diameter and volume of the crown region, a regression of dosage (M1 diameter was not very useful. Later recommendations included the sstatement to “spray each elm thoroughly on an individual basis." Commercial DDT formulations were found to be too injurious for tzr-ees. Therefore, special formulations were deveIOped which were not F>flytotoxic and which provided persistent deposits (Whitten, 1949). The 1:C>llowing formulations were recommended: FORMULATION A: 16 pounds Of technical DDT dissolved in 22‘; gallons ben- zene and 1 gallon Velsicol AR-SO (mono- and dimethyl- naphtalenes) plus 1 pint Triton X-100 (arakylpolyether- alcohol) as emulsifier FORMULATION B: 16 pounds of technical DDT dissolved in LI gallons of xylene plus 1 pint Triton X-100 as emulsifier The above emulsifiable concentrates were diluted with water to make I 00 gallons Of spray emulsion in each case. Formula A and B were sug- Sested for hydraulic Sprayers only. About 25-30 gallons spray were '7.‘E=<:ommended for an average 50-foot elm tree. [>0 RMULATION C: 20 pounds of technical DDT dissolved in 5 gallons of xylene and 2% gallons Acme white oil plus 1% pint Triton X-100 as emulsifier It? <:D'r'mula C was recommended only for mist blower application. The con- <:i'E=Intrate was diluted with water to make 20 gallons Of spray. TWO or 32 three gallons was the average amount of spray used for a 50 foot elm tree. The first treatment was applied Shortly before the appearance Of elm flowers and leaves while the second application was usually 2-3 months later with the same formulation but only half as concentrated. The purpose of the second treatment was tO control defoliators and vectors of the phloem necrosis. Mist blower equipment had the following specifications: output c>f 6000+ cubic feet of air per minute at a velocity of 120 miles per l1<>ur. One to four gallons of spray concentrate were injected into the éai r stream at 40 pounds Of pressure. Plumb (1950) used mist blowers Wh ich were modified NO. 12 Bean-Vapo dusters. Output was 8400 cubic feet C>1: air per minute at a speed Of 140 miles per hour. Four nozzles injected "<>IJghly 1 gallon Of spray into the air stream per minute. The formulations LISSo_ +39»; — U m I I O 0 I m m. a. L0“ _ 0< o< om oo 00 0“. 0w om mm 3 mm. mm 00 J u 8. menu}. R .r cl .‘. u a . ...- u 57 was lower at 27.8%. No information could be found in the literature about the importance of feeding in leaf axils with regard to effective disease transmission. Twig crotches were sometimes attacked more than once, even those with old scarred callus tissue, particularly in samples with a great number of feeding injuries. The attempt to determine the influence of branch position on the amount of feeding was unsuccessful. In general the branches in the tOp region pointed upwards, were more or less horizontal in the middle region, and hung straight down in the bottom portion of the tree. Not only is the morphological variation of twig characteristics great, but also location (isolated tree or in a stand) has a lot to do with the morphology of the twigs. Some elms have long stringy twigs which are very flexible and with few crotches (tree E and F) while others had many crotches and are more sturdy in appearance. Location of the feeding *wound, on the upper or on the bottom side of the twig, showed no pattern regardless of the position of a branch. (B) Regional Distribution of Feeding Injuries Along a Horizontal Gradient Friedman's ANOVA by rank gave the results listed in Table 3 below. Table 3. Comparison of feeding attack in 4 cardinal quarters and center section of elm trees using Friedman's ANOVA. Tree XE Significance Highest Feeding in Quarter of Tree 8 7.#7 0.20 N, E C 3.67 n. s. -- D 3.00 n. s. -- E 7.h7 0.20 E F 6 50 ‘ 0.20 -- (0k= 0.]0, df = Li) (ok= 0.20, df = 1+) ><>< 1N1N 0 ll U'I\l .q; \u «‘3‘ v I“ A. . ‘43: l ii! '\-I I t l“ ll 58 Only trees B and E had greater differences in feeding attack between the five sample points N, S, E, W and C (Center) consistently over all 3 height levels. In tree B the percentage of feeding injuries was lowest in the S section, but considerably higher in the N and E sections. Tree E had the least feeding in the w section and the highest number of injuries in the E section. Initially it was suspected that the center section at the lower height levels 2 and 3 would be the region with the fewest number of feeding scars because of a possible vector preference for the periphery of the crown. However, this was not the case as this analysis proved. Uneven distribution of feeding wounds in the 5 vertical sections of the tree (N, S, E, w, C) was probably not the result of preferential feeding in any one of these 5 regions. Rather, the proximity of a beetle-producing elm tree to a particular quarter of the above sample trees might have caused this difference in attack. The question of a possible difference in the overall feeding pattern on elm trees in a closed stand and on road-side elms was also raised. Sample trees A, B, C and D were in closed stands together with other tree Species while trees E and F were part of a road-side planting. The vertical and horizontal distribution of feeding injuries in these 2 groups of trees did not seem to be different; however, both road- side elms showed differences in the number of feeding injuries in the 5 sections of the tree. Whether this is a general trend among road- side trees can only be speculated upon since the sample was too small. (C) Association of Physical Twig Crotch Characteristics With Feeding l. Angle of twig crotch and feeding There was no relation between crotch angle and feeding. None of 59 Table 4. T-comparisons of angle means of attacked and unattacked twig crotches. Tree Height N Mean S SX tcalc. Level Angle B 5 att. 5 85.00 7.07 3.16 unatt. 95 not calc. total 100 87.25 10 att. 32 84.06 14.89 2.63 unatt. 68 82.43 12.38 1.50 0.575 total 100 82.95 15 att. 32 81.72 7.67 3.12 unatt. 68 85.74 16.58 2.01 1.110 total 100 84.45 D 5 att. 18 63.89 11.58 2.72 unatt. 82 64.70 11.56 1.28 0.266 total 100 64.55 10 att. 120 66.53 10.39 0.94 unatt. 30 67.33 11.80 2.15 0.364 total 150 66.82 15 att. 118 68.52 11.25 1.04 unatt. 32 71.41 11.38 2.01 1.277 total 150 69.13 = 1.980 (ct = 0.05, df = 1.960 (at: 0.05. df S n~ln '- fl ;'1 5.": «ii ' a: 3;.‘9 ‘ssb 'n 'U'Jil,‘ \ ‘1 s 1" .\~ n u 1 . u a p 1v 1N 1. 60 the calculated t-values exceeded the table values (Table 4). Of interest is the fact that the angle means of the height level samples for tree B, with 87.250, 82.950, and 84.450 were considerably higher than the means for tree 0, with 64.550, 66.820, and 69.130. The great difference in angle means between these two trees is an example of the morphological variability one can find among elm trees. 2. Association of round or acute crotch base with type of feeding: lateral or central Ouellette (1962) noted that lateral feeding resulted in a higher infection rate than did central feeding. Presumably during lateral feeding the beetle vector with the spores glued to its body establishes better contact with the conductive tissue. Type of Feeding Twig Crotch Base Round Acute Total Lateral 124 60 184 Central 83 72 155 Total 207 132 399 chalc = 6.78 thab = 3.84 (05 = 0.05, df =1) The contingency table suggested that twig crotches with a more rounded base were more likely associated with lateral feeding than with central feeding. The association of these 2 characteristics was signifi- cant at DC = 0.05. These findings suggest that trees with a higher proportion of twig crotches with a rounded base are more likely to be infested be- cause of the preferred lateral type of feeding in these crotches. ‘ a 111'. v ” rm"; .9 4' ,x .g. 1" 61 Summary and Conclusions Extensive sampling of untreated elm trees revealed that S, myljj: striatus feed preferentially in the whole upper tree region. Feeding attack rate increased linearly from the bottom to the top. Differences in attack in vertical sections of a tree were observed but were be- lieved to be related to the close proximity of a beetle-producing tree to a particular quarter of a sample tree. Several physical twig crotch characteristics were investigated in reSpect to their association with vector feeding. The size of the angle between the main and lateral member of a twig crotch had no influence on feeding. There was, however, a significant association between a more rounded twig crotch base and lateral feeding. Lateral feeding reportedly results in more successful inoculations than does central feeding. These findings pointed out a possible resistance mechanism against feeding by S, £2151: striatus. 11. Evaluation of Methoxychlor Formulations and Application Techniques Introduction The protective spraying against the feeding of S, multistriatus has been heralded by many as the main way to prevent Dutch elm disease transmission. But before the availability of organic insecticides, in particular DDT, 'pre-attack' spraying was considered to be of doubtful value; first, because of the nature of the available materials and, second, because of the lack of an adequate spray technology at the time (Fransen, 1931). DDT had to be phased out of all Dutch elm disease programs in the urban communities in the mid-1960's. There was never any doubt about the superior performance of DDT applications against the feeding of §;. multistriatus (Becker, 1946; Whitten, 1958; Doane, 1958, 1962). However, when environmental considerations forced the replacement of DDT with other insecticides, there were few materials which could substitute satis- factorily. DDT's long persistence and its toxicity to S, multistriatus made it very valuable for Dutch elm disease control programs. Methoxychlor, also called methoxy-DDT, emerged in the 1960's as the only promising replacement for DDT (Wooten, 1962; Barger, 1971) although earlier reports did not favor it as a possible DDT substitute (Matthysse _£__l,, 1954). Recent work dealing with the behavior of methoxychlor in ecological sys- tems indicated that this compound is highly biodegradable (Metcalf t al., 1972). This feature, and low mammalian toxicity make methoxychlor a good choice for applications in urban communities. Elm tree protection requires very thorough Spray coverage since 62 63 the feeding sites, or twig crotches, are distributed over the whole crown region. The tree tops deserve the most attention because feeding was observed to be heaviest in the upper crown region (Wolfenbarger and Bu- chanan, 1939; see also section I of this thesis). The problem of ground applications with the mist blower or with hydraulic equipment was often one of insufficient tree t0p coverage, especially on tall trees (more than 18-20 m high). Wallner and Leeling (1968) pioneered the use of helicopter applications for Dutch elm disease control and reported superior coverage of the tree t0ps. The Dutch elm disease conference in 1967 (Peacock, 1967) outlined two important research areas: (a) formulation studies with methoxychlor and (b) research on application techniques. One reason why the switch from DDT to methoxychlor put a stop to many community control programs was the high cost of this insecticide. Increasing labor costs and the expensive, slow ground applications contributed to the decision of commu- nities to discontinue control programs. Aerial spraying of elms had major cost advantages over ground applications which were related to speed of the spraying operation. Also, the amount of insecticide added to the environment by the helic0pter was considerably less than by ground equip- ment. However, the value of aerial spraying for elm protection needed further experimental support. Part of the work reported here followed a formulation study initiated by Wallner g£_§l, (1969) on the elm pOpulation of Michigan State University. In 1969 and 1970 ground and aerial applications were evaluated by feeding trials and residue analysis. 64 Materials and Methods (A) Spray Applications 1969 For study purposes, the campus of Michigan State University, with an elm tree pOpulation of circa 1600, was divided into 5 areas (Figure 6). Each area contained at least 100 elms. The Spray plot arrangement was the same as for 1968 (Wallner t al., 1959)- Treatment 1: Michlin Chemical Co. 'MZ-2' Elm Spray 25% methoxychlor 36.5% xylene 36.5% petroleum distillates 2% inert ingredients and emulsifier Treatment 11: Standard Oil Elm Spray 'M' 25% methoxychlor 65% xylene 8% paraffinic white oil 2% inert ingredients and emulsifier Treatment 111: Michlin Chemical Co. 'MA-2' Elm Spray 25% methoxychlor 67% xylene 6% paraffinic white oil 2% inert ingredients and emulsifier Treatment IV: Michlin Chemical Co. 'MX-2' Elm Spray 25% methoxychlor 73% xylene 2% inert ingredients and emulsifier Two pounds of Dacaginl) (86.9% polysaccharide-gum mixture, 13.1% inert ingredients) were added to 100 gallons of 12.5% emulsion in order to prevent excessive spray drift. The rate was reduced since nozzle clogging occurred during the 1968 application. Treatment V: Amoco Elm Spray 25% methoxychlor 73% xylene 2% inert ingredients and emulsifier All the above concentrates were mixed with water to give 12.5% emulsions. In the middle of April, all 5 areas were Sprayed with a Bell G-Z helic0pter which was equipped with a 30 ft. boom and 59 D-6 nozzles 1) Diamond Alkali Co. ADJ-“m..." \ \x _.. . .\\ s ...\ .x r... .k\_ .. Is KxVN . \~ § .868 9%) 432.1% Duo "43.0.... 65 r \ \\\1|\|\ .mmmmflmMn.“mfimmummmmmmmmfimmmwwmwfi .. 1| 1 1. ...m... .111...“ 1.1 = I l.” m 1- ' 4 I... W — I II. I III .2. _ ....l I 111! = _ :1 fl ‘ I .u OU‘I II- V o v A a.” . .11 . o l 1 1 1.. H 1\1 . 1.1.1 1 1 I. 1 I 11.411 I“ J .I [II I u -..-Hunwhql- @ gang» 2 66 (Figures 8, 9). Flight speed during application was above 20 miles per hour and the flight height above tree tOpS ranged between 15-20 ft. The helic0pter delivered 7 gallons per acre in a 50 ft. swath at a line pressure of 45 psi. The wind conditions during spraying were steady and did not exceed 5 miles per hour. Spraying time was early morning between 5 and 7 a.m. and in the evening between 6 and 8 p.m. Each tree received on the average 2 quarts of 12.5% spray emulsion. Treatment VI: The smaller elms in the University married housing area with a height of 10-12 m were treated from the ground with a John Bean 301-6 roto mist sprayer (Figures 10, 11). The same 12.5% Spray emulsion was employed as for treatment V of about two gallons per tree. (8) Spray Applications 1970 In 1970 most of the elms on campus were treated with the Michlin Chemical Co. 'MX-Z' Elm Spray diluted to give a 12.5% emulsion. Dacagin was added for spray drift control at a rate of 2 pounds per 100 gallons emulsion. Amoco Elm Spray which is similar to Michlin Chemical Co. 'MX-2' in its composition was also used as 12.5% emulsion and with 2 pounds of Dacagin per 100 gallons of spray. The applications were made at the end of April in early morning by helicopter as described for 1969, but the average amount of spray per elm was somewhat less than for the previous year (less than 2 quarts). Wind conditions were favorable and not ex- ceeding 6 miles per hour. Several 15-18 m elms close to dormitories and other buildings, where helicopter application did not seem feasible, were treated with a John Bean 301-0 roto mist. Again, Michlin Chemical Co. 'MX-Z' was used as a 12.5% emulsion but without Dacagin. Figure 8. Figure 9. Figure 10. Figure 11. 67 Aerial spraying of elm trees with Bell G-2 helicopter. Bell G-Z helicopter: part of boom and nozzle assembly. Ground spraying of elm trees with John Bean Rotomist (model 301-0). John Bean Rotomist: blower and nozzle assembly. . ... 9’” 5V. . :~‘ .41‘ .5053 ‘ r l 410 1135 us 68 (C) Twig Sampling lfléfl; From each of the 6 Spray plots 2 elms were chosen at ran- dom and twig samples were taken 6 and 10 weeks after application. The sampling of each tree was designed to take into account the vertical and horizontal variation of spray deposits within the crown. Each tree was subdivided into 3 horizontal sample strata (Figure 12). HEIGHT LEVEL 1: TOP (15 m and above) HEIGHT LEVEL 2: CENTER (10 m) HEIGHT LEVEL 3: BOTTOM (5 m) 0.6 - 0.9 m long branches were taken at 2 points within each height level which gave a total of 6 sample points per tree. About 100 1-5 year old twig crotches were cut from both branches and bulked to one sample. The fresh twig samples were transferred into polyethylene bags, labeled and transported in Styrofoam coolers to the freezer for storage (temperature -10° 0). Samples were stored up to 2 months while awaiting analysis de- pending on availability of beetle material or manpower. The trees were usually sampled with the MSU Grounds Department Stratotower which made the sampling relatively easy (Figure 13). When this equipment was not available, an extendable aluminum pole pruner with a light-weight head was used for sampling trees up to 16 m (Figure 14). 12193 10 helic0pter-sprayed trees were chosen at random from the elm pOpulation on the MSU campus and were sampled 2 and 8 weeks after application. Twig sampling followed the same scheme as in 1969. 2 mist-blower-sprayed trees, both close to dormitories, were sampled as well at the same time. The tree heights of the sampled trees ranged betWeen 15 and 19 m. 5‘4 5‘1 NV. Figure 12. Figure 13. Figure 14. 69 Sampling scheme for treated campus elms: location of sample points at three height levels. Twig sampling with Stratotower. Twig sampling with aluminum pole pruner. W .1 W V. any» AU 5 PF— . .. a... ht ‘II 5 I n o v 1 C . I‘- '5 Al 5v 70 (D) Bioassay Technique The technique used here for the feeding assays was suggested by Lincoln (1968) and described and illustrated by Barger t 21- (1971) in a recent publication. Twenty twig crotches, 5-7.5 cm long, were bioassayed for each height level, totaling 60 twig crotches per tree. Holes (diameter 7/32”) were drilled into 3/4ll plywood boards to hold the twig crotches in an upright position during the feeding trials (Figure 17). Small per- forated brass cylinders were made in 3 diameters, 4.0, 4.5 and 5 mm, to fit various twig crotch sizes. The length of all brass cages was 13 mm and one end was ground to a deep notch to give a tight fit in the twig crotch (Figure 15). The perforated brass was obtained as sheet metal from McMaster-Carr Supply Company, Chicago. The brass cage was pressed into a firm position in the twig crotch to assure a tight fit. This was very important since the twigs dried up and shrank during the 24-hour feeding period which resulted in several cases in loose cages and escaping beetles. Individual unsexed beetles which emerged within 24 hours previous to the bioassay were placed with a 'vacuum-forceps'l) into each cage (Figure 16). The Open end of the cage was closed with a small piece of &“ Permacel masking tape. On several occasions it was observed that beetles gottrapped on the sticky surface of the tape, but healthy, vigorous beetles were usually able to free themselves. Only 'healthy‘ beetles were used for the bioassays. The beetles were considered 'healthy' if they were able to climb the Side wall of a 1 pint 'SEALRIGHT' ice cream carton. The duration of each bioassay was set at 24 hours at 25° C and 80% (f 10%) relative humidity and the light conditions were set at a lZ-hour day. After 24 hours, the bioassay was terminated and the beetles 1) Chemical Rubber, Cleveland, Ohio Figure 15. Fitting of brass cage in twig crotch for bioassay. Figure 16. Placement of S, multistriatus in brass cage with vacuum forceps. Figure 17. Assembled twig crotches on plywood tray. Figure 18. Classification of feeding response: F3 (weak: xylem not visible), F2 (medium: xylem visible), Fl (strong: xylem deeply penetrated). 71 1.1 were removed 72 from their cages. The beetles were then sexed and classified according to their physiological stage as: (a) (b) (C) The (a) (b) (C) (d) alive: full capability of movement, no jerky movements (alive): movements jerky, fall sometimes to the side or on their backs dying or dead: cannot right themselves under intense light, slight spastic movements of antennae or legs, wings often times spread out with rhythmic motions of abdomen or complete immobility feeding reSponse was either indexed as (Figure 18): NF: F1: F2: F3: non-feeding (no feeding activity noticeable) strong (xylem deeply penetrated) medium (bark removed, xylem visible) weak (patches of bark eaten, xylem not visible) or the length of the scar was measured directly with a 20 mm disc micro- meter reticle in a LEITZ stereo microscope. (E) Scolytus Rearing The bioassays required a large number of Scolytus beetles with preferably a high degree of physiological uniformity, same age and equally good health. Therefore, a rearing technique was required which provided beetle material with the desired characteristics. Several rearing con- tainers for bark beetles are described in the literature, ranging from 5-gallon metal pails (Fox, 1958), fiber drums (Valek, 1967), vented 20-gallon metal trash cans (Clark and Osgood, 1964), to collection con- tainers made of 35-gallon lever-lock steel drums (Truchan, 1970). Mold and fungus growth on the logs is a common problem in rearing containers due to insufficient ventilation. Even if a high percentage of bark --' p: ... v ,0. fl - , v 0‘ C .l ‘ 4 (n'a .- - IV . :13 I... . . s 08‘ R )1! 1 . 9", :-~ ‘h 4.. J i .. “ch. FH‘ '51 'VI 73 beetles emerges and is recovered, their physiological condition might render them unsuitable for bioassays. A further requirement was the immediate extraction of newly emerged beetles from the rearing containers to guarantee fresh beetle material of uniform age. S, multistriatus is Strongly phototrophic like many other insects. This behavioral feature allows efficient collection Shortly after emergence. Many of the afore- mentioned emergence drums have collection devices which allow the insect to go back into the emergence drum, or the light conditions in the con- tainer are such that the emerging insects are not drawn immediately into a collection jar. Both of these conditions result in beetle material of different age which is not desirable for a standardized bioassay. There- fore, emergence drums were designed with a ventilation system to prevent fungus and mold growth and with an efficient extraction system which did not allow the insect to move back into the rearing container. Emergence drums (Figure 19: A, B) Three 35-gallon steel drums served as rearing units securely positioned horizontally on a slotted angle steel rack (Figure 20). The drums were covered in front with circular black 'Poly Cover' plastic Sheeting which was conveniently tied down with 5/16” rubber cord. A cir- cular air screen was inserted in the lower half of the plastic cover to aid air circulation. A piece of black plastic sheeting loosely covered the screen vent to prevent light from penetrating into the inside of the drum. The actual extraction device consisted of a 5 cm long piece of PVC pipe (diameter 3”) glued into a circular hole close to the lower rim of the drum after which a 500 ml transparent plastic bottle from which the bottom had been cut out was fitted snugly over the PVC pipe. One end of a short piece of %” tygon tubing was inserted into the mouth 74 Figure 19: Design of emergence drum and ventilation system N .PUU mm 10. 11. 12. 13. I4. 15. 16. Emergence drum plus extraction device A Front View 8 Side View Ventilation system C Front view: blower housing and air pipe plus support 0 View from above: blower housing E Side view: support of air pipe with valve half closed 35 gal. steel drum Black plastic sheeting Tie-down rubber cord Air-screen Black plastic sheet covering air screen PVC pipe (diameter 3“) 500 m1 transparent plastic bottle Tygon tubing 1 gal. ice-cream carton 20 WATT cool white neon light Blower housing 2.5 m long PVC pipe (diameter 3“) Wooden support with adjustable valve Dayton blower (model 2C781) 3-way switch 5 cm c0pper tubing (diameter 5/8”) 75 Figure 19. Design of emergence drum and ventilation system. n n.- n. ‘1’. 1':- 1:? ill ‘1 EV J 76 of the bottle and the other end placed straight down into a l-gallon ice cream carton. The lower end of the tubing reached several centimeters into the carton, and prevented the beetles from crawling back up into the drum. Small collection jars with very little Space for free movement have the disadvantage that the emerging beetles inflict injuries on each other, rendering them unsuitable for bioassay work. The l-gallon cartons were a definite improvement in this regard and with frequent emptying of these collection cartons, beetle-inflicted injuries were very rare. A 20-watt cool white neon bulb provided the necessary light stimulus for drawing the emerging beetles into collection containers. Ventilation system (Figure 19: C, D, E) The ventilation system consisted of the blower-housing with blower, a long PVC pipe (diameter 3”) with several air outlets and an adjustable valve to regulate the airflow (Figure 21). The necessary airflow was generated by a blower (Dayton Model 2C781) housed in a %“ plywood box. It could be set at 3 speeds with a 3-way switch. A 2.5 m long PVC pipe (diameter 2“) was connected to the blower with air outlets every 30 cm. Five cm long pieces of 5/8'' c0pper tubing were glued into the PVC pipe to serve as air outlets. A %“ diameter Tygon tubing guided airflow into the emergence drums. Inlet into the drum was provided by a 5 cm long piece of 5/8'l copper tubing with a fine-mesh screen soldered to one end to prevent insects from entering the air tubing. The first 20 cm of the Tygon tubing next to the air inlet was wrapped with black plastic tape to keep light from entering the drum. The other end of the PVC pipe rested firmly in a wooden support which consisted of &” masonite board sandwiched between 2 pieces of 3/4” plywood. A sliding 77 .mEJLv oocomcmeo LOm Eoum>m co_um__uco> ..N ocam_m .xumL _umum co mmO. E_m ;u_3 Escv oucomLoEm .ON ocsm_u . 1.- 03$ 1 I 0‘ '11 F. 11096 TI 78 piece of masonite board allowed continuous regulation of airflow into the rearing container. With the end of the PVC pipe completely Open and the slide pulled out, airflow into the drums was at a minimum. This setting was used for the whole rearing Operation and guaranteed sufficient air exchange. For comparison purposes, relative humidity (R.H.%) and temper- ature (°C) data were recorded for the rearing room, for an unvented and a vented rearing unit over a 2-week period. The same number of fresh logs from the same tree were placed in each of the 2 units. Temperature in the 2 rearing units did not, as might be expected, differ from the temperature in the rearing room, and was constant at 26° C (a 1.5). How- ever, the relative humidity varied considerably. Relative humidity for the rearing room averaged 54.8%, whereas in the vented rearing drum the average relative humidity was constantly higher at 67.8%, the mean difference being 13%. The unvented rearing unit had constantly maximum humidity and never went below 100% R.H. during the 2-week period. The logs in the vented drum were free of mold with beetle emergence high, and mortality low. The high humidity in the un- vented drum caused extensive fungus growth and an overall unsatisfactory environment for beetle deve10pment. Emergence was very poor due to high mortality inside the drum. NO quantitative comparisons of mortality data in the vented and unvented rearing units were made. The beetles used for the bioassays were reared from field-infested elm logs. A lab culture Of S, multistriatus was also maintained. Fresh elm logs were artificially infested with S, multistriatus in completely closed cardboard boxes and kept there for 2 weeks before they were placed in one of the emergence drums. Parasites or predators (clerids and ostomids) were never a problem in these cultures. 79 (F) Quantification of Methoxychlor Residues on Elm Bark I. Bark sampling For accurate quantification of methoxychlor residues in twig crotches, it was necessary to obtain uniform bark samples. Most commonly, residues on elm bark were expressed in weight per unit of bark surface (micrograms/cmz) (Miller, 1951; Matthysse _£'_l,, 1954; Nielsen, 1962; Parker and Nielsen, 1967). None of the previously suggested methods allowed easy sampling and exact measurement of the twig crotch area. Since feeding scars along the sides of the twig crotch are appar- ently more conducive to true infection (Ouellette, 1962), bark samples had to include this area. With a NO. 1 cork borer (inside diameter 4 mm) one bark circle with an area of 12.56 mm2 was taken from each twig crotch (Figures 22,23). Fifty twig crotches from each height level were sam- pled in this manner, which gave a total of 6.28 cm2 bark surface analyzed per height level. 2. Chemical analysis (Figures 24,25) The analytic procedure basically followed the one described by Wallner, Leeling and Zabik (1969). Extraction Procedure The bark was extracted with a hexane-acetone solvent pair (1:1, 50 ml) in a 500 ml Erlmeyer flask. The sample was allowed to sit overnight and the next day it was shaken on a Burrell-Wrist- Action shaker for 20 minutes at a setting of five. The extract was de- canted intO a 500 m1 separatory funnel. Following this, another 50 m1 Of the solvent-pair was added to the sample and again shaken. The same step was repeated a third time so that the extract amount was finally equal to 150 ml. The acetone fraction of the solvent was then removed Figure Figure Figure Figure 22. 23. 24. 25. Bark sampling in twig crotch with cork borer NO. 1. Bark samples from twig crotches. Extraction and clean-up of bark samples. Analysis of samples with gaS chromatograph (Beckman GC-4). 80 1 0p ‘ .1 Q8. ..1“ 'n . v "I; up! .v ‘4 v" rf ivvu .. a .. K 3 ...-r M 81 with distilled, deionized water. 300 m1 of distilled HOH were added and thoroughly Shaken with the extract. After separation of the 2 layers the HOH layer with the acetone was drained off. If excessive emulsion occurred, a 10% NaCl solution was added to separate the 2 phases. This step was repeated 4-5 times until no more acetone smell could be detected in the aqueous fraction. The hexane extract was transferred quantita- tively to a 500 m1 round bottom flask using a funnel with a glass wool plug, plus 15 g Of anhydrous Sodium Sulfate to remove remaining traces of water. The extract was concentrated down on a rotary vacuum evapor- ator to approximately 10 m1 using a lukewarm water bath. The concentrate was then pipetted onto the prepared clean-up column. The round bottom flask was rinsed 3 times with small aliquots of hexane to insure complete transfer. Preparation of Clean-up Column A 22 x 500 mm clean-up column was used with thin Teflon tubing as the Outlet. The column was prepared with 10 g of Florisil: Celite (5:1) between 1 cm layers Of Sodium Sulfate. The Florisil was 12% de- activated with distilled water and then tested with a known sample of methoxychlor. The total amount of methoxychlor was eluted between 350 and 420 ml, which was considered as adequate. The Florisil was saturated with hexane before the sample was pipetted onto the column. Methoxychlor was eluted with 500 m1 redistilled hexane into a 1000 ml round bottom flask. The sample was then concentrated down to about 15 ml on a vacuum- evaporator as mentioned before, quantitatively transferred to a screw- top vial with a Teflon-lined cap and filled up to exactly 25 ml. The samples were stored at -180 C while awaiting analysis on the gas chrom- atograph. ups-1' ’\ .’C.~ ‘\ . . 1.. v 0'! .r '1: M . v V . , 5 LE 7: let .. "M :. 1.; ;.IA 31 - 'i 82 Quantitative Analysis The methoxychlor samples were quantified on a Beckman GC-4 gas chromatOgraph (electron capture detector- ECO) fitted with a 6 ft. long, 1/8” diameter Stainless steel column (Figure 25). Column packing for the 1969 analysis consisted of Gas Chrom Q 60/80 with 3% SE-30. Temper- ature for the column was 2000 C, for the injection port, 2400 C, and for the detector, 3000 C. The retention time for this column was 4.5 minutes at 200° C. Gas Chrom Q 60/80 with 4% DC-ll in a stainless steel column of the same dimensions was used for the 1970/71 and some 1972 samples. The retention time for this column was 4 minutes at 2100 C. Most of the analysis in 1972 was done on a Beckman GC-65 (electron capture detector) which was equipped with a 6 ft. glass column (diameter l/l6”) packed with Gas Chrom Q 60/80 with 3% SE-30. Figure 26 is a sample run of 3 standards: methoxychlor, pp'DDT and pp'DDE. The peaks of DDT and DOE were super- imposed to show retention time of these 3 compounds and the relative sensitivity of the electron capture detector. Amounts of 2 microliters per sample were injected with a Hamilton microliter syringe. Standards were injected at the beginning of a run, after each 10-12 samples and at the end of the run. Peak height was measured and converted to PPM with the standard graph to give the amount of methoxychlor in the sample. The 2 conversion to micrograms/cm bark surface was calculated with the following formula: (PPM) (DILUTION FACTOR) (SAMPLE VOLUME, ml) = micrograms Meel/cmz bark BARK AREA analyzed (cm‘) The methoxychlor in several samples was confirmed with thin-layer chroma- tography (silica gel-H plates). 3. Efficiency Of extraction procedure The methoxychlor residues in the bark samples were undoubtedly 83 :33 am 5.33 a w... —..._ ..__ _- £83 an. _ w H 5.556. W _ _ a ‘- 0:96: woo . .. ...oo 3 32253505 2 6003 34302 .co_;u>xo;ume pcm Foo .moa b 1 _ __ _,__ , {TIT—T" m ~-—16-"—--~- v "mmuuu cmExuOm m :0 ese o_aEmm m a h m _ a M . . _ .,. . _ _ . _ ,_ _ . . _, r . r « .AHM _ A P . _ . . _ _ c .. _ u ._ _ _ . . n _ W m _ _ _ _ _ _ M m _ m 0 0. L 100. z<_ zomm < _ W L. 20 zamwmaaz 0u _ mucmsummcu c_ >u__mucoe acoczamz.hgso=.~m=: "mom_ m>mmmmo_m .NN 6L:m_u 956.8: 5302.0... ”> 2 _ o. m c— .. _ _ _ w on m x H . . «1 .. e H. 1 8 .1 1 8 - m m. . _> > >_ - 8. w H H H 111H H 111w . . H H11. H «11. 1.: .1 .. 8 .1 .. 8 H. 1 8 .. =. __ _ .... 3:85.: )2 90 treatments alone, rather the response in these assays expressed both carry-over and 1969 treatment effects. This caution in the interpretation of the bioassay data was particularly valid for the helic0pter treatments I-V. In the mist blower treatment VI the carry-over effect was probably insignificant in light of the high residues accumulated after each Spring application. Since this study was initiated after the treatments were applied, it was not possible to differentiate between carry-over and treatment effects. Treatment VI, a mist blower application, resulted in high mor- tality at both sampling dates, 93.8% and 96.2% reSpectively.. The trees in this treatment were smaller than the average helicopter-treated elms (IO-12 m). Therefore, a direct comparison of treatment effects between helicopter and mist blower-sprayed trees at different height levels could not be carried out. The mist blower applied 4-6 times as much spray ( 2 - 2% gallons) as compared to roughly 2 quarts with the helicopter. This is one reason for increased mortality in the mist blower treatment. Further, greater effectiveness of the mist blower spraying is given by the fact that considerably more of the total twig crotch area is reached than is the case with helicOpter spraying. This point was conclusively proven during a later phase of this investigation (see section 111). Feeding Response (Table 7, Figure 28) The degree of protection of an elm tree against vector feeding is perhaps more truly given by the feeding response itself,or rather, the 'failure to score the xylem', than by the percentage mortality. F1 and F2, both strong feeding responses with the xylem visible, were pooled to one value which was termed 'successful feeding'. ReSponses F3 and NF were pooled as well and termed 'failure to score xylem'. It 91 ._> OH _ mucoEumoLu c_ OmcoamoL mc_600m 5.32.08 Eta 9.095 111 2953: .0263 3.8 "S "mmm_ m>mmmmo_m .mN Oc3m_m 3:953: 858:0; ”> 2 _ 2 o 2 o S o v _ H q _ H _ .1 1 8 I H H .. . H H H H . S .. H J. - .. 1 1.3 . H 1% . . «1 _> > >_ - s. .. H H. L. H 1H. 1.2 . H1 H HH1 H j H . .. - s .1 1 8 .. - a - =_ __ _ so? 4 s _ 900m 2 92.3 a 92 was believed that indexing the feeding response was more meaningful than would be an actual measurement of the scar length, since the beetle in close confinement within the brass cylinder could not exhibit 'natural' feeding behavior. Further analysis of feeding and mortality data proved that indexing the feeding response (although only a relative measure of feeding behavior) was very reliable. In general, the figures of percentage 'failure to score xylem' for treatments | to V1 corresponded well to the respective mortality figures. The correlation of these 2 responses received further attention in a separate analysis. Again, the mist blower treatment VI had the highest degree of protection with 88% to 95% in May and July, followed by treatment IV, then V, and with very little protection, the rest of the treatments, 1, 11, and 111. (B) Bioassays 1970 (Figure 29) l. Helicopter Mortality In general, the helic0pter treatment was surprisingly ineffective considering that the first series of bioassays was conducted 2 weeks after application. Mortality figures ranged from 62% to 3.5% with the mean at 23.2% t 12.5 (C.L.)Ikor the 10 trees, C to M. After 8 weeks (2nd sample period) the mortality decreased to 11.3% i 10.9 (C.L.) calculated from the 5 sample trees, C, E, G, H, and K. Feeding Response Failure to score the xylem in the first series of assays was considerably higher than the correSponding mortality figure. The mean value was 41.0% i 8.5 (C.L.), which was almost twice as high as the mean 1) 00= 0.05 93 .....1...... 1’ .mE_o panacea 2030—3 Hm_E ucm Emuaoo__oc co mmcoammc mc_pomm new >u__mucoe "0mm. m>mmmmo_m .mN 0L:m_u 5.823% .28 9.33 H l 1‘1 8— E2? 203 2 95:5 x a N u." H . I L o I o I o I o I o I o 11 ON 5302.6; 1.ov J / .1 8 .0303 EE 1 1 8 8— 3:23:. a 94 mortality. However, the mean feeding response for the July assays of 18.3% 1 10.64 (C.L.) corresponded well with the average mortality figure of that sample period. 2. Mist blower Mortality By comparison, average mortality in the 2 mist-blower-sprayed trees was lying considerably above the mean of the helic0pter-treated elms. Tree A, sprayed before bloom, had 82.7% mortality while tree 8 had only 53.5% due to the fact that the blossoms prevented good coverage of the branches. For the second sample period the mortalities dropped to 66.4% and 47.0% for tree A and B, respectively. Feeding Re5ponse The percent failure to score xylem and the respective percentage mortality corresponded better in the July assays than in the May feeding trials, as was also noted for the helic0pter treatment. (C) Correlation Between Mortality and Feeding Response (Failure to Score Xylem) in the Bioassays 1969 and 1970 (Figure 30, 31) These correlations include the figures for all sample trees, both helic0pter and mist-blower-treated, of a given sample period. The mortality figures in these correlations were not corrected for natural mortality. The feeding response or failure to score the xylem includes all apparently non-feeding individuals plus the cases when only small bark patches were removed (xylem not visible). Both reSponses are highly correlated in the 1969 bioassays with the correlation coefficients r= .98 for the May samples and r= .97 for July. The slope for the regression lines was in both cases 1.09. The intercept was somewhat larger for the July assays, however it was negative in both cases. 95 .muo_coa OHQEmm 03H L0H >H__mucoe 6cm .Eo_>x Ocoom OH Oc:__mH. Ho co_um_oLcoo "mom. m>mmmmo_m .om Oc:m_m 80.? 908 o. 925* 2: 8 8 3 8 8— 8 8 3 cm 1? H H H H H H H H H H H H H b H H H b H 1 o 1 o 8 OO 1 flow 0 f O o I NO. '5 ”a. '5 .36.. + and- ..> .. .33 + I..- 1» re 1 1cm 1 H. l 18 .I T .o 52. o \o 1 12: 3:555 x 96 .mno_gma o_aEmm o3u Low >u__mpLOE ncm .Eo_>x mLOUm Cu mL:__mH_ Ho co_um_mLLoo NoumH m>mmmmo_m Emix 900m 2 23:3 x ..m 2:3... 2: a 8 2 2 2: 8 8 s H H H H H H H H H H H H H H H H H H H O O I O \o\ oo 18 o o 1 ”a. '5 mm. '5 o x»: + and...» .35 + 3.21 1» 12 o I 18 O I 18 r >42. >52 Is. 3:955 x 97 The correlation of the 2 responses, mortality and failure to score xylem, in the May bioassays of l970 is weaker with r= .85; the sl0pe of the regression line was 1.12. In 8 out of 10 data pairs in this correlation, the feeding response (failure to score) was higher than the corresponding mortality figure. A possible reason for this discrepancy between failure to score xylem and mortality could be a repellent action of surface deposits. Cuthbert _£_§l, (I972) suggested that methoxychlor deposits on bark and in the twig crotch might also have some repellent action which keep the beetle from feeding. The May samples were taken 2 weeks after application; that is, enough surface deposits could still have been present in these samples, which prevented a stronger feeding response. It is conceivable that it takes some time (longer than 2 weeks) for the surface deposit to penetrate into the bark (ordecompose on the surface) and a particular twig crotch might become attractive again after this penetration (or decomposition) process is completed. The data for July, l970, were again well correlated with r= .98 and the sl0pe for the regression line was l.l3. Under the conditions of the bioassay, feeding response indexing was very reliable, which was proven by the good correlations between failure to score xylem and mortality. These correlations suggest that beetle mortality could be predicted from observed feeding responses. (D) Mortality at 3 Height Levels The sampling scheme was designed to take into account not only the horizontal but principally the vertical variation in Spray coverage (Figure l2). With a mist blower Operating from underneath the tree and a helicOpter from above the crown region, definite differences in residue deposits at the 3 height levels were expected. It is a well-known fact 98 that with increasing tree height the mist blower loses its effectiveness, that is, less spray is carried into the tree t0p region. The preferential feeding pattern, as elaborated in section I suggests that the greatest protection is needed in the tree t0p region with no particular prefer- ence as to a compass direction. The helic0pter application seemed to offer a valid alternative to the mist blower application since it presumably deposits more spray in the top region. Theoretically the deposits should decrease from the t0p to the bottom due to the interception of spray in the upper tree region. In other words, it was expected that the helicopter would lose efficiency in coverage from the top to the bottom while the mist blower would lose efficiency from bottom to tOp. lf trends of this sort existed, the bioassay data should have reflected differences between height levels with regard to mortality and feeding response. Since mortality was well correlated with the feeding response, only mortality data from the helic0pter applications in 1969 and l970 and from the mist blower treatment in l970 were examined. Because of the nature of the data, a non-parametric test was employed to demonstrate possible differences in mortality between height levels. 1. Bioassays l969/70: Helic0pter The mortalities at the bottom, center and tOp level of all trees of a sample period were compared in Friedman‘s two-way ANOVA by ranks. A separate analysis was performed for each sample period in l969 and l970. All data were corrected for natural mortality in the controls. In two instances the corrected data had to be coded with +l0 because of some negative values after correction with Abbott's formula. The results of the ANOVA are listed in Table 8 and show a contrasting pattern for l969 and I970. It was expected that mortality would be highest in the upper 99 Table 8. Friedman's ANOVA by rank of mortalities at 3 height levels of helic0pter and mist blower-treated elms. 1N Sample Date x No. Trees p 1) Highest mor- tality at height level A. Helicopter Treatments 1969 May 1.25 10 0.569 ---- 1969 July 5.55 10 0.069 3 1970 May 2) 0.66 9 0.814 ---- I970 June 6.30* 5 0.039 I B. Mist Blower Treatments I970: data 7.13* 4 0.042 3 pooled for May and June I) Table N in Siegel, ”Non-parametric Statistics” (I956) 2) One tree omitted from analysis because of incomplete data tree region in all five helic0pter treatments. The Opposite was the case for July, l969, where the mortality was highest at the lowest height level. In I970, 2 weeks after helic0pter spraying, mortality was in general higher at the top level; however, the differences were not significant. The mor- tality in the second series of bioassays, 8 weeks after application, was significantly greater in height level I. This change of the mortality pattern in the helic0pter treatment from l969 tol970 probably reflected the slow, but continuous disappearance of mist blower carry-over in the lower tree regions. The cube sample experiments (section III of this dis- sertation) proved clearly that the helicopter delivers significantly more spray into the top region of the tree than into the lower portions. One must conclude, therefore, that the carry-over from previous mist blower lOO applications (prior to I968) reversed the expected vertical mortality pattern in the l969 helicopter treatments. With further decomposition of the mist blower carry-over the actual distribution pattern character- istic for helicopter-applied residues will become more and more apparent. The vertical mortality pattern of the I970 helic0pter treatment indicates this. The residue analysis of l969 and I970 bark samples gave further support for this conclusion. 2. Bioassays I970: Mist blower Tree A and B were treated by mist blower. Figure 32 shows the average mortality pattern at the 3 height levels for the 2 sample dates. The mortality pattern of the I970 helic0pter treatment (9 tree average at 2 weeks, 5 tree average at 8 weeks after application) is shown for comparison. Typical for the mist blower application was the continuous decrease in mortality as tree height increased. Since only 2 mist blower sprayed trees were sampled, the mortality data from the May and July bio- assays were pooled for the ANOVA. The results are shown in Table 8 and were significant at p:=0.0h2. The high mortality at the lower height levels is a strong contrast to the helic0pter treatment. Since S, multistriatus obviously prefers feeding sites in the t0p region of the tree (see section I of this thesis), particular atten- tion must be given to good tree-tOp coverage. A direct comparison of mortality figures at the upper height level on mist blower and helicopter Sprayed trees is therefore of great interest. The average mortality in height level I of the 9 helicopter-sprayed trees was 27.0% 2 weeks after application (May I970) and l7.3% 8 weeks (July I970) after application. There was a decrease in the mean mortality from height levels I to 3 in the May samples, I970; however, the ANOVA did not indicate significant l0l .Aco_umo__aam Loumm mxomz w 6cm NV mE_m vmummuu1cozo_n um_E ucm Louaoo__mc Low m_o>m_ u;m_m£ mobs“ um >u__mucoe mmmco>m "cum. m>mmmmo_m .Nm meam_u 95 .92 Ego; . 4 9. OH m 0H OH. 0 ..-- II .1111. \ 1cm -..--. \ 9.063 m 10v .33 5:6 9.003 m .--l '00 ."" [om games as mw>>01_m H.25— meaOOSwI 50':al 503w L tree A at the Experi create 5l0wer was de l02 differences between height levels. The mean mortality in the July sam- ples went down from l7.3% in the t0p to 9.9% in the center and to 6.7% at the bottom. The differences in mortality at the 3 height levels were significant as shown by Friedman's ANOVA. 0n the other hand, the mist blower treatment caused the least mortality in the top height level. In the l969 treatments this did not show up because the trees were smaller (l0-l2 m) and tree-tOp coverage was sufficient. The mortality in the t0p region of these smaller trees ranged between 80% and 100% at both sampling dates (6 and lo weeks after application). 0f the mist blower treated trees in I970, tree B was somewhat smaller than Tree A and was in bloom when it was sprayed, which resulted in poorer coverage. Consequently mortality was lower than in tree A, particularly in the t0p height level. Mean mortality was 27.7% at the t0p but fell to 23.2% 8 weeks after spraying. The cube sample experiments indicated that a well-exercised mist blower application will create sufficient deposits on the t0ps of 15 m high trees. The mist blower application for tree A and B was probably not thorough enough and was definitely ill-timed because of tree bloom. (E) Methoxychlor Residues on Treated Elm Trees One aim of the residue analysis of bark samples was to give support to the findings of the biological assays. The amount of residue present on or in the bark is certainly an accurate parameter for the com- parison of spray coverage on elm trees and was suggested by several workers (Thompson, I965; wellner g£_§l,, l969). Although useful for comparison of treatment effects, the amount of spray deposit can only be a relative measure of performance and in itself lacks biol09ical meaning. Therefore, for prOper interpretation, the amount of residue 103 needs to be correlated with the response of the insect in the bioassays. This was done in section IV of this dissertation. l. Carry-over (Table 9) Prior to l968, the elm pOpulation on the campus of MSU was sprayed yearly in the spring with a mist blower. Section V gives a detailed ac- count of DED spray practices on the campus since I958 (see also Figure 6 for treatment areas). In l968, the spray application was performed by helic0pter on the majority of the campus elms. Five methoxychlor formu- lations were applied and evaluated by gas chromatographic analysis (Wall- ner _£__l,, l969). Twig samples from elms in treatment areas IV and V taken before the helic0pter application had very high methoxychlor resi- dues. The above authors attributed the significant carry-over in these treatment areas to heavy mist blower spraying over the previous years, which frequently overloads the lower tree portions. Methoxychlor deposits on bark were believed to have low persistence. Wallner and Leeling (I968) reported originally that methoxychlor deposits had a short life span, not exceeding one year. This could be true if the deposit is primarily laid down on the bark surfaces where it will decompose much faster. However, if a formulation and/or application technique favors bark tissue deposits over surface deposits, carry-over will be higher since the residue is imbedded in the tissue and is protected against weathering. This point was proven in decomposition experiments of several methoxychlor formu- lations (section IV). Because of the heavy carry-over from the earlier mist blower applications (both methoxychlor and DDT were used), it is very difficult to make a valid judgement about the superiority of one or the other l969 helicopter treatment. Carry-over levels were determined each year from 1969 until l97l, shortly before the prefoliar Spray 104 applications. Wallner 33 _13 (I969) found the highest carry-over from I967 to l968 on trees in the treatment areas IV and V (Table 9). The difference in carry-over was significant on these trees which were mist blower-sprayed during the previous year with the same methoxychlor formu- lation. Obviously elm trees in certain locations received greater atten- tion because of their shade tree value. The following year the pre-spray residues were again highest in treatment areas IV and V. Carry-over from l969 to I970 was substantial in treatment IV with the highest residues in the lower height levels. Trees in the only mist blower treatment had high carry-over residues in the upper height level. It was intended to use these carry-over data as the base values for the evaluation by calculating the difference between post-spray residue and carry-over. However, variation in spray deposits within and between trees of the same treatment was great. Only extensive bark sampling and residue analysis would have removed this variability. For these 2 reasons, carry-over and variation of spray deposits, the post-spray residue values from the heli- copter treated elms must be interpreted with considerable caution. The same caution applies for the interpretation of the bioassay data since the response at the feeding site was a composite of carry-over and treatment effect. The high amount of DDT in some of the I970 samples was surprising. DDT was also present in samples from 1969, but it was not quantified. DDT was verified by gas chromatographic analysis with 2 different columns and by thin-layer chromatography on silica-gel H plates. 0nly trace amounts of metabolites, DOE and D00, were present. If these residues were indeed carry-over, it was astonishing to find DDT still in its unmetabolized state. A possible explanation may be that DDT had survived 105 Table 9. Methoxychlor residues in twig crotches of helic0pter- and mist blower-treated campus elms: 1968, I969, I970. 1968 2) 1969 1970 PPM micrograms/cmz micro./cm2 Treat- Height Pre 3 Days 130 Days Pre 8 Weeks 15 Weeks Pre ment Level Post Post Post Post MeCl DDT 1 1 84.0 260.2 72.6 6.1 6.2 4.33) 2 18.8 9.4 3 19.4 3.7 11 I 281.4 328.0 |2#.h 1.2 39.3 2 17.2 3 5#.h III I 20.5 195.2 32.0 4.0 3.9 2.3 2 #. 2.h 3 10.0 2.H 4) IV I 308.8 316.0 257.0 9.5 35.6 16.3 8.7 1.5 2 83.5 17.8 17.5 2.5 3 14.7 27.7 28.4 3.5 V l HOQ.8 432.0 207.6 15.0 23.8 7.5 2 17.3 11.7 3 15.8 18.3 -- VI I 172.1 39.9 25:64)l.67 2 183.5 71.3 3 208.3 80.2 I) I - V: helic0pter; VI: mist blower 2) Data from Wallner g£_§l, (l969) 3) Average of 2 trees per treatment 4) Average of 3 trees per treatment 106 as tissue deposit in the outer bark where neither biological nor envir- onmental factors could induce metabolism. The bark samples were taken from 1-5 year old twig crotches. Therefore, part of the total sample consisted of bark from older twig crotches which apparently carried the DDT residues. On some trees the amount of DDT carry-over was suspiciously high considering that the last mist blower applications with DDT were performed 5 years prior to sampling. 2. Methoxychlor residues l968/69 (Table 9) The results of the residue analysis of twig crotches from treat- ments 1 to V had a similar pattern in both 1968 and 1969. In 1968, for both sample dates, coverage appeared to be best in treatments IV, V and also 11, followed by l and Ill. The sequence from highest to lowest residues for the first sample date in 1969 was: treatment IV, then 11, V, I, and Ill. The last two treatments, I and III, again had the poorest coverage, as was the case in 1968. Results from the second sample date in 1969 established once again that treatments IV and V had the highest residues, while treatment 111 had very poor coverage. The 1969 samples were collected from 3 height levels to demon- strate vertical distribution of residues within the crown region. In general, the highest residues were found at one of the 2 lower height levels while the tOp level had the lowest deposits. The reverse was actually expected in these helic0pter treatments since the upper tree portion presumably intercepts most of the spray. It is possible that the vertical distribution of carry-over from previous mist blower appli- cations still overrode the distribution pattern created by helic0pter spraying. If this assumption was correct, one could expect a continuous transition from a 'mist blower residue pattern' to a 'helic0pter residue 107 pattern'. The mortality patterns at the 3 height levels in 1969 and 1970 reflected this trend and supported the above hypothesis. Also, methoxy- chlor deposits could decompose faster at the t0p region where twigs are more directly exposed to environmental influences. This would create a similar effect whereby residues in the lower tree region would appear to be higher after a longer weathering period. However, this is only speculation. Treatment VI, the mist blower spray plot, had considerably higher residues than did the best helicopter treatments. The typical mist blower deposit pattern, with the lower tree portion having the highest residues, was not very distinct in this treatment. The reason probably was that the low height of the trees in the spray block permitted the entire tree crown, even the tops, to receive equally good coverage. 3. Methoxychlor residues 1970 (Table 10) There were only 2 treatments in 1970, a helicopter and a mist blower treatment. Samples were taken shortly (2 weeks) after application, but residues were considerably lower at all height levels from the year before. Again, the variation in spray deposits between helic0pter-treated trees was great. The average deposits at the 3 height levels showed mini- mal differences 2 weeks after helicopter spraying. Eight weeks after application the deposits were greatest in the t0p height level. As for 1969, the vertical distribution of residues corresponded well with the mortality pattern at the 3 height levels. The strong presence of DDT in many samples from both sample periods was mentioned already. Significant amounts of DDT were detected in the first sample of h helic0pter and 2 mist blower sprayed trees at all height levels. Two helicOpter and 2 mist blower sprayed trees had DDT present in the second samples, but 108 Table 10. Methoxychlor residues in twig crotches of helicopter- and mist blower-treated campus elms: 1970. Treatment') Height 2 Weeks Post 8 Weeks Post Level MeCl DDT MeCl DDT Mist Blowerz) I 5.7 4.9 3.7 3.7 2 17.7 15.1 7.8 5.5 3 27.9 22.7 19.7 15.9 Helicopter I 12.83) 8.hh) 2 12.8 5.8 3 12.5 6.3 1) For description of formulations see page 2) Average of 2 trees, heavy DDT deposits 3) Average of 9 trees, DDT present on 4 trees h) Average of 7 trees, DDT present on 2 trees in smaller amounts. Considerably more DDT was found on the mist blower-treated trees than on the helicopter-sprayed trees. The vertical distribution of de- posits, both methoxychlor and DDT, was typical for the mist blower. The deposit figures for the mist blower in Table leepresent the mean of 2 trees and are almost 10 times lower than the figures for treatment VI in 1969. Data from the cube sample experiments in 1972 suggest that initial n1ist blower deposits can be extremely high (several hundred micrograms/cm2 bark) in the lower tree portions. Either the extraction and the quanti- teat ive analysis was unusually unreliable or the application itself was ;><><>1-. One of the 2 mist blower trees was already in bloom when the spray ”V2355 zapplied. Consequently, on that tree the coverage was particularly ;>c><:1r-- 109 Basically, the bioassays and the residue analysis yielded the same information. Both methods established the mist blower as the most effective application technique, at least for trees with limited height. Also, the same sequence of most effective to least effective helic0pter treatment was indicated by the two methods with the exception of treat- ment II which had high residues but little mortality. Treatment IV, the Dacagin formulation, appeared to be the best choice for the helic0pter; however, the high carry-over of DDT and methoxychlor in this treatment block confounded the results to a large extent. This applied to the other treatments as well. This circumstance made it difficult to draw valid conclusions from this field experiment andtn recommend a superior methoxy- chlor formulation for the helic0pter. (F) Evaluation of Treatment Effects by Comparison of Disease Incidence in Treatment Areas I-VI: 1968-1969 The effectiveness of pre-attack sprays or other control strate- gies has frequently been evaluated on the basis of disease occurrence (Plumb, I950; Matthysse‘_£._L., l95h; Barger, 1971). One has to agree that in the case of DED the true test for insecticidal formulations and application techniques is overall disease prevention rather than the per- formance in a particular bioassay. If the total number of trees lost to DED over a given period of time is used as the sole criterion to compare treatment effectiveness in sometimes widely separated areas, sev- eral assumptions have to be made. The most important one is expressed 5" the following question: Do all elm trees in the treatment areas have 5’15! same chance of becoming infested with the disease? Naturally, the F"'<>t>ability of disease incidence is directly related to vector density (VeCtor pressure), assuming that the amount of inoculum carried by a 110 population is prOportional to its density and rather stable for a given population size. Equal chance of tree infection presupposes, therefore, a uniform distribution of the vector population over all treatment areas. This is, in all probability, an incorrect assumption. Fransen (Heybroek, l969) attempted to select in a field experiment elm trees which were resistant to vector feeding. While S. scolytus was feeding rather uni- formly on the experimental elm trees, the feeding pattern of S, @3131: striatus was very spotty and erratic. Wolfenbarger and Jones (l9h3) made similar observations and found that the feeding pattern among a pOp- ulation of elm trees was influenced by the distribution of beetle-pro- ductive trees. Even if the post-emergent flight to a feeding site was random and not triggered by olfactory stimuli (Meyer and Norris, l96h) the feeding and, subsequently, also breeding will always be higher around beetle-productive trees. Furthermore,wind currents can offset the vec- tor distribution (Felt, 1937) and so contribute to an erratic feeding pattern. The design of field experiments where tree losses in sprayed blocks were compared to the losses in the untreated check blocks fre- quently had inherent flaws. Untended wood lots where the shade tree value of elms approached zero served often as check plots. Naturally, the vector pOpulation was higher there than in the sprayed treatment blocks, which were usually roadside plantings or park trees. Because of this vast difference in vector pressure, disease incidence was under- .standably much higher in the untreated check blocks. In other words, $116! treatment effect was undoubtedly exaggerated in these comparisons. The differences in elm die-back in the six treatment areas of t"€3 MSU campus over the two-year period (Table II) can only partially b‘3 iéittributed to the relative effectiveness of one or the other methoxychlor voEL_Hcou “02 Am me On _:o momma co __muou c_ pon_eomou ocm mucmEumoch AN ucoEuLmamu mucaoLm 3m: ecu mo mucoume EoLH coxmu moL:m_m A. lll H.o H.m_ m J. m 6 mm _> o.m 6.0. :~ mm mm a: 6mm > w.m w.m 1 a a 1 mo. >_ a.mm o.m~ 1 mm m. ON 0:. ___ :.: o.m _ m m N om __ 0.: o.H H ON oH o. owm _ uHmCu ummeu uoom .HUXm .mHOH fi uoox Hmu0H mom. wom— uGHMLQm moc< moock vommom_o Ho x mouth vommom_o Ho .02 momck Ho .02 AuucmEumoLH _> 1 . mmocm acoEumoLu c_ mo\wmm_ mommOH mock . __ o_nmk A. 112 formulation or application technique. Differential vector pressure in the treatment areas was definitely present. In particular, the western section of the campus is close to extensive untended wood lots which were, at the time, still vector productive. In addition, an increasing influx of vectors must have started in 1968 when the city of Lansing, also to the west of the campus, began to reduce DED control practices (see sec- tion V). Other confounding factors which made an objective evaluation of these treatments I to VI difficult were root graft transmissions and possible carry-over from previous DDT treatments (the latter applied also to the feeding trials). Figure 7 shows the campus areas which were con- tinuously mist blower-sprayed with DOT from 1958 to 1967. Disease occur- rence in 1968 in the treatment areas I to V1 did not differ very much from the figures in 1969. The highest percentage of die-back occurred in area 111 in the northwestern part of the campus, followed by area VI, the only mist blower- treated area, then came areas V, I, II, and 1V (see Figure 6 for exact location of treatment areas). Root graft transmission was suspected to be responsible for the substantial tree kill in area VI since most of the infestations were in immediate proximity of each other. Also, it was be- lieved that vector pressure in this area was higher than in any other part of the campus. Treatment IV had the smallest percentage of trees killed overall by the disease. With the exception of treatments III and IV, the results of the 1969 bioassays were a strong contrast to the above Piczture. Treatment VI (mist blower) emerged as the most effective in these feeding trials, giving 90% protection up to 10 weeks after appli- Cat 3 on. From the five helicopter treatments, only IV was satisfactory, fol I Owed by V and, giving minimal protection, I, II, and III. The 113 sequence here was almost the reverse of the disease picture. Again, probably several of the factors mentioned above contributed to confound the results (die-back) in the respective treatments. In order to correct for the varying vector density in the treatment areas, one would have to know the relationship between vector density and disease infection. Therefore, a different experimental design is necessary to safeguard against confounding variables, such as vector density, which can confuse the whole disease picture and can make comparisons of treat- ments to control DED meaningless. A completely randomized design or a randomized block design would certainly be more apprOpriate, but from a practical standpoint, such a design might not be useful since the trees in each treatment would be randomly distributed over a large area. This could make the actual applications complicated and time consuming. Summary and Conclusions In 1969, five helicopter treatments and one mist blower treat- ment were evaluated by feeding trials with S, multistriatus and by anal- ysis of bark residues. A 12.5% MeCl formulation containing Dacagin, a viscous thickening agent, emerged as the most persistent and effective helic0pter treatment. Mortality ranged between 50 and 60%. Consequently, this formulation was recommended for the 1970 helicopter application on the majority of the campus elms. The performance of this treatment was very poor. The mean mortality in the bioassays reached only 23.2% 2 weeks aafter application and fell to 11.3% after 8 weeks. Twig samples taken before the spring applications in 1969 and 1970 suggested that carry-over F7‘<>n11nist blower applications prior to 1968 with DOT or methoxychlor was 5"t>=3tantial. It was concluded, therefore, that the good protection a- c" 3 Iaeved in 1969 in some helicopter treatment areas was the combined result 114 of carry-over and the particular helic0pter treatment. Carry-over in- validated the results of the comparison study of several helicopter- applied methoxychlor formulations to a great extent since it affected the outcome of the bioassays and residue analysis. In 1969, the carry-over in the lower tree regions still exceeded the new helicopter deposits in the tree tops. The feeding trials and residue analysis in 1970 indicated that the carry-over in the lower height levels had diminished to the point where tree top residues began to be higher. The slow transition from a mist blower residue pattern, with higher deposits in the bottom region, to a helic0pter residue pattern, with higher deposits in the tree t0ps, was well expressed by the results of the bioassays and residue analysis in 1969 and 1970. The comparison study of ground application with the mist blower and aerial application with the helic0pter was based on bioassay data and bark analysis and definitely favored the mist blower. Tree height seems to be the only factor which would alter the effectiveness of the mist blower. The investigation concerning the preferred regions for feeding attack within the tree crown established the whole upper tree portion as the primary hazard region. The helicopter application in 1970 gave insufficient protection in tree t0ps, causing only 27.0% mortality, and decreasingly less in the lower height levels (2 weeks after application). The mist blower protected the lower height levels almost completely, vvith mortality figures ranging between 80 and 100% for the whole season. Mist blower protection was poorer in the tree tOps but still better than the helic0pter application. Placing most of the attention during a spraying Operation on t r-€3<3 t0p coverage is inadequate. One has to take into account as well 115 the hazard regions for effective inoculations with the OED pathogen. In- oculations in the lower tree region, particularly close to the main trunk, have been shown to be more effective and lead to faster disease deve10p- ment in the whole tree. Considering both the vector and the pathogen, the whole tree would need equal protection, because of the higher feeding rate in the tree t0p and the higher successful infection rate in the bottom region. Neither mist blower nor helic0pter gave equally good pro- tection throughout the tree. Their obvious draw-backs were pointed out already. Two reSponses, mortality and 'failure to score xylem' (feeding response), were recorded in the bioassays. Both were highly correlated except in one case where twigs with 2-week old deposits were bioassayed. Surface deposits of methoxychlor were still present and they apparently induced a behavioristic type of resistance in S, multistriatus which re- sulted in low mortality due to non-feeding. This researcher believes that the possible repellent action of MeCl surface deposits is of little significance in terms of tree protection since a searching vector will eventually start to feed in a suitable twig crotch. Disease occurrence as the sole criterion for an evaluation of spray treatments against DED was critically discussed, using the helic0pter and mist blower treatment areas in 1968 and 1969 as examples. Tree kill in each treatment area appeared to be independent of the degree of tree Protection (as indicated by the bioassay results). It is su5pected that tfl1e tree kill reflected more the vector pressure in each respective area. 77162refore, a different experimental design was suggested for field experi- ’"eirits where tree kill is used as the criterion of treatment performance. 111. Monitoring of Spray Coverage on Elm Trees Introduction During the course of the field evaluations of application tech- niques and methoxychlor formulations in the spring and summer of 1969 and 1970, it became desirable to determine the degree of coverage of a tree using a more objective method. Bark analysis alone was extremely variable and only extensive sampling would have given the desired accuracy. Besides, carry-over from previous mist blower applications interfered heavily with the evaluation and distorted treatment effects. The objectives were to demonstrate how applications by helicopter and mist blower differed with respect to : (a) variation in spray patterns, drOplet size and density along a vertical gradient, (b) horizontal and vertical distribution of actual spray deposits within the crown region, and (c) coverage in a 3-dimensional sense. A close look at point (c) seemed to have particular relevance in this case because of the 3-dimensional nature of the target sites, the twig crotches. A sampling device was needed, therefore, which could serve the above objectives. Maksymiuk (1971a) successfully used fluor- escent dye-tracers for the assessment of spray deposits and Spray patterns on cards or aluminum plates. Oil-sensitive cards were found to be useful for the immediate appraisal of oil-based spray deposits. Ideally, a sphere would best show a 3-dimensional coverage. However, since actual deposit determinations had to be made, a cube-like sampling device was selected with filter disks fastened to all 6 sides. The filter paper 116 117 trapped the spray in such a way that it could be conveniently extracted and quantified. The total analysis consisted first of a visual inspection of coverage and measurement of Spray patterns, and, second, of an accurate quantification of deposits on these filter disks. Materials and Methods (A) 3-Dimensional Sampling Device Wooden cubes with the dimensions of 4.25 x 4.25 x 4.25 cm were covered with Permacel masking tape to prevent secondary absorption of spray from the filter paper into the wood. Whatman No. 2 filter disks were stapled to all six sides of the cube (Figure 33). The cubes were suspended on wire from branches at the respective sample points. The top and bottom sides of the cube were horizontal but side areas did not necessarily face certain compass directions. Five cubes were placed in each height level at compass points somewhat within the tree region and at the center (Figure 2). For mist blower experiments in 1972, attachment of the sample cubes to the branches had to be redesigned because of the strong air current generated by the blower. For an objective experiment, it was necessary to have the sample cubes securely positioned. One end of a 20 cm long piece of steel wire (0.08” diameter) was inserted into the center of the wooden cube; the other end was wound around a branch and tightly squeezed on with a pair of pliers. Also, the filter disks had to be stapled to the cube with at least two staples, again because of the strong air current. For the helic0pter experiment, one staple in the center sufficed. 118 vm SOHHOD o MIIIIIIil. .Jm 0L:m_u . co_umcmmmoo Am 66 _v maa_m aaau tau__c Co . .mm etam_u .cucmcn 8.0 cu nocoummw onsu 0HQEmm mWMWJWWflJ$;,.n . fl . a. v. 5...... tarp u le The du As WE 1m... 119 (B) Spray Applications 1. Helic0pter application 1971/72 Two elms, one a park tree among a group of shade trees and the other a road-side tree, were randomly selected from the campus elms which were sprayed by helic0pter in April, 1971. One day before application, the sample cubes were positioned in the two trees. Tree A (park tree): 20 m high Cubes were placed at 3 height levels (5, 10, and 15+ meters) and about 1-2 m within the dripline, % meter above ground. Tree B (road-side elm): 15 m high The same setup was used as for tree A, however only two height levels (5 and 15 m) were sampled due to the smaller size of the tree. The: helic0pter used the same type of spray equipment as was employed 30 ft. boom, 59 D-6 nozzles. The weather duridwg the previous years: A Michlin conciitions were ideal with winds less than 5 miles per hour. 'MX-Z' 12.5% methoxychlor emulsion was employed throughout the campus. AS 'F<>r the 1970 application, Dacagin at the rate of 2 pounds/100 gallons was added for spray drift control. The average amount of spray received by each tree ranged above 2 quarts. The Stratotower from the Michigan State University Grounds Department was made available for placing and Samp] ing the filter cubes in the tree crown region. Two elms, both 15 m in height, served as experimental trees in '972- Sample cubes were positioned at two height levels (5 and 15 m) at the 5 sample points as in 1971. JLESE§L_£; (park tree): sprayed with same formulation as in 1971 T‘ .HEEESE_Jl (park tree): sprayed with same formulation as in 1971 but with- out Dacagin 120 The helicopter was equipped with the same spray gear as the year before. Approximately 2 quarts of spray were applied to each tree in the early morning hours. The wind conditions were calm, below 6 miles per hour. 2. Mist blower application 1972 Two days after the helicopter spraying, sample cubes were again placed in the same trees (C and D) at two height levels. These elms were then sprayed from the ground with a John Bean Rotomist 301 G. Approxi- mately 2% gallons of Michlin 'MX-Z' 12.5% emulsion was employed per tree; tree C was sprayed with the Dacagin formulation. (C) Analysis of Spray Deposits The first part of the analysis was concerned with a visual in- Spemztion of the sample cubes and the spray pattern under long-wave UV. Thettsecond portion involved chemical analysis and quantification of the actlmal deposits. The filter cubes were removed from the trees several hours after spray application and they were stored in a cooler (10° C) uP ‘t<> 14 days until the analysis was completed. Each filter disk was ca r'ei:ully taken from the cube by cutting with a sharp dissecting knife along the staple or by pulling out the staples with needle-nose pliers. 1. Analysis of spray patterns on filter disks The xylene-based Spray drOplets which soaked into the filter paper became strongly fluorescent with a light-greenish color under long- Wave UV. A Blak Ray (UVL-22) lamp was employed as the stimulating light soul‘ce, First, the overall Spray deposition on the six sides of the Samp] e cubes was examined visually, then the top disks were analyzed with re . . . . spect to number and slze of spray drOplets per unit area. Since 1t 121 was difficult to count and measure the drOplets over the total area of the filter disk at once, the disk was subdivided into smaller sections with a stencil. This was accomplished by placing an acrylite plate with a circular Opening above the filter disk. Three thin acetate strings glued across the center subdivided the disk into six equal sections. Only three Opposing sections (half the area) were analyzed on each disk (Figure 35). Droplets were classified according to their size into three groups: (1) 2-3 mm, (2) 1-2 mm, and (3) smaller than 1 mm. 2. Quantification of spray deposits on filter disks After the visual analysis of the spray patterns was completed, the methoxychlor residue on each individual filter disk was extracted 1~ith 100 ml of redistilled hexane in a 300 ml Erlmeyer flask. Five grams 0f :SOdium sulfate (anhydrous) was added to each extraction to remove tra<:es of water. Each sample was Shaken lightly and stirred with a glass rDd. After sitting overnight, each sample was analyzed in a Beckman GC-‘+ gas chromatograph or a 10 ml aliquot was stored in a freezer. The total amount Of methoxychlor residue was calculated as micrograms per Squa re centimeter (micrograms/cmz). The extraction of residues trapped on filter disks involved only 5’ s'ifigle step and was, therefore, very efficient. The recovery rate was determined in an experiment where 25 micrograms Of methoxychlor in acetone Were pipetted onto 3 filter disks. The disks were air-dried for 2-3 hours and then extracted with 100 ml hexane. Percentage recovery was 90'00/0, 94.0%, and 95.0%, the average being 93.0%. The residue determin- a ' . . t'ohs from the falter dISkS were not corrected. Figure 35. Analysis of Spray pattern on filter disk under long-wave UV. Figure 36. Fresh methoxychlor deposits on mist blower-treated elm branch. 122 (19%|:- d 611:"; 123 Results and Discussion (A) Spray Pattern, DrOplet Size and Density, Along a Vertical Gradient in Treated Elm Trees 1. HeliCOpter application The mean number of drOplets/cm2 was computed for each size class and height level. These means and the associated Standard errors are tabulated in Table 12. DrOplet density in all size classes definitely changed from the upper height level to the lower tree region. The addi- tion Of Dacagin had a strong influence on drOplet size even at the reduced rate of 2 pounds/100 gallons. The numbers and equivalent proportions of droplets in each size class per unit area (cmz) are graphed in Figure 37 at the levels 1 and 3 for the two formulations. It is apparent that the Dacagin formulation produced a higher prOportion of larger sized droplets at both height levels. About 20% of all droplets exceeded 1 mm in diameter. The straight emulsion resulted in a pattern with only 6-10% of the total number of drOplets being larger than 1 mm in diameter. The prOportion of larger sized drOplets increased somewhat in the lower height level which could have been the result of spray drift. The pro- portions of the 3 size classes remained the same at the 2 height levels in the application with the Dacagin. When the mean density per cm2 in each size class was multiplied by the respective mean diameter (0.5, 1.5, 2.5 mm) the total area covered by residue was Obtained. However, since the drOplets were measured on filter paper, it is doubtful if these figures of the actual area covered apply also to elm bark with different Surface and absorption characteristics. For comparison purposes, they might suffice. It is interesting to note that for both formulations 13%.of 1 cm2 was covered with spray deposit at level 1, while only 6% of ll luil‘ufhu «ll q1riHn1 ...-nu unm- vaufw PV1IHnd 1- ——-d ....Ufln mm Hrd—HHAdH-U )FiLF-wu lu-lu 124 Home“ Hv mxm_u Lmu__w : EOLH team—:OHOO AN AmOOcu my mxm_v cmu__H mH EOLH beam—:OHOO A. ANEEV UOLO>OO ow.m Ho.N mH.N :m.o ou H_.m_ N:.o_ om.m .m._ ou ::.o ow._ om.m ww.o oo mm.H_ om.a am.m ma.m OU AEEV mommOHu ON_m c_ HO>OJ mco_um_: mOc< HmHOH .m .m a mum—note Ho % cmmz unm_oz 1Ecou Hum: co_umo__aa< cmm> .mOOLu 5.0 teammcu HO mHO>OH u;m_os N um mxm_u LOu__H HmucoN_co: co UOLO>OO moLm .mauom ucm mmcm u_c: Lou mum—Qotn Hmcam HO Conan: u:o_umu_Haam cmuaou__oz .N_ O_an 125 height level 1: 15m height level 3: 5m .7 12.5% MeCl emulsion + Dacagin (2 lbs/100 gal) 12.5% MeCl emulsion .4 size class 1 (2-3mm) 2 (1-2mm) A 3(<1mm) Figure 37. Helicopter apEIication: prOportion of drOplet sizes per unit area (cm ) on horizontal filter disks at two height levels of treated elm trees. V—V' (h 126 1 cm2 was covered at level 3. That is, although the spectrum of droplet sizes was quite different for the 2 formulations, there appeared to be little difference in the actual area covered. From the standpoint of better and more even distribution of bark residues, the formulation with- out Dacagin is more desirable with a considerably larger number of drop- lets per unit area. However, the advantages Of Dacagin as anti-drift material and its possible retarding effect on weathering might outweigh the above negative feature. The overall distribution of spray deposits on the sample cubes was very poor in all helic0pter experiments. Figure 38 is an example for the poor coverage of sample cubes at 2 height levels. In all sample cubes, the tOp disk had the highest drOplet density while only one or two side disks had significant deposits. In no case did the bottom disk have any appreciable amount of spray drops. The visual examination Of deposit distribution on the sample cubes showed little difference between formu- lations and height levels. At the lower height levels, however, most of the deposits were concentrated on the top disk, with minimal deposits on side disks. The poor overall coverage Of the sample cubes (at best three disks) demonstrated that helicopter application did not create enough air swirl, which is necessary for good coverage of three-dimensional objects (like branches). Also, it was observed that the Spray, after being ejected through the nozzles, settled down very calmly with little or no apparent turbulences in the air. The revolving blades of the helicopter create a strong vortex moving downwards only when the aircraft is hovering. However, down-draft and air turbulences are minimal as the speed of the helicopter increases. Elm spraying is usually done at speeds around 20 miles per hour, which is too fast for Spray delivery into the downward 127 Figure 38: Helicopter spray deposits on filter cubes (fluorescence photographs) Tree (I Sample point N Height level I (15 m) Height level 3 (S m) b) a) Numbers 1-6 refer to position of filter disk on sample cube (see Figure 34‘ b) Actual spray deposit in micrograms/cmz. (v u- (I) n0 ‘1) 128 vortex. Best results with aerial spraying of apple trees were obtained when the helicopter hovered above an individual tree and delivered the spray into the downward vortex (J. Topliss, Montrose, C01,; pesticide applicator; personal communication). The vortex after hitting the ground was apparently reflected back into the tree region which resulted in ex- cellent coverage. It was speculated that filter disks with no apparent spray de- posits, according to the visual examination under UV light, could still harbor micro-droplets. Therefore, only an accurate chemical analysis could prove if this was the case. Fresh deposits of helic0pter spray examined shortly after appli- cation showed up as dark spots on the bark, especially on smooth bark of younger twigs. The overall distribution of droplets on branches, even in the tOp region of the tree, was poor; only the bark areas which were facing upwards were covered. 2. Mist blower application Ground application with the mist blower resulted in a completely different spray pattern in that it was difficult to distinguish between individual spray droplets. The density of droplets was very high, es- pecially of the smaller size class below 1 mm in diameter. Fluorescence was low in comparison to droplets of similar size of the helicopter pattern; this indicated that the individual droplets might be less con- centrated, due to the fact that the strong air blast from the turbine caused greater spreading upon impact. Droplets were usually more dis- tinct on the tap disk, and the pattern resembled the helicopter treatment. This was understandable since the pattern on the top disk was created by spray droplets with little velocity settling down from above. Tree C 129 was mist blower sprayed with a formulation containing Dacagin (same rate as for the helic0pter) while tree 0 received the same formulation but without Dacagin. Visual inspection of the filter disks on each cube revealed that the overall coverage was very uniform and similar for both formulations. At height level 3, all six disks, even the tap disk, had uniform coverage. The filter cubes in the upper tree region had consistently fewer droplets on the top disk, the other side disks and the bottom disk again having uniform coverage. The superior coverage of the sample cubes (and, therefore, of the branches) with the mist blower application can be explained on the follow- ing grounds: (a) During the Spraying Operation the blower is moved around the tree, thus covering the tree from all Sides. (b) The strong air stream generated by the blower turbine causes great air movement ('swirl') which is necessary for good coverage in a three-dimensional sense. (c) The bark areas facing upwards away from the Spray direction are eventually covered by spray droplets which settle down from above after being carried into the crown region. It was not possible to analyze the mist blower sample cubes visually with UV-Iight in the same manner as the helicopter cubes because of the in- distinct droplets. Fluroescence photographs (Kodak Plus-X, 2 Blak Ray UV-22 lamps, Kodak Wratten filter No. 2A) with the spray patterns at the 4M0 height levels for mist blower are shown in Figure 39. Observations (”1 the coverage of branches were made shortly after application. The ac tual deposits on the bark made the twigs look wet. Later only a Wh‘itish residue remained (Figure 36). Overall coverage Of branches was 130 Figure 39: Mist blower spray deposits on filter cubes (fluorescence photographs) Tree D Sample point E Height level 1 (15 m) Height level 3 (5 m) b) a) Numbers 1-6 refer to position of filter disk on sample cube (see Figure 341. b) Actual spray deposit in micrograms/cmz. I'L'S per 18% C05. $1310: and deg fro, W hen. Gem and 131 complete, particularly at the lower height levels. (B) Horizontal and Vertical Distribution of Spray Deposits on Treated Elm Trees The distribution of spray on the 6 sides of the sample cubes in each height level was interpreted with respect to actual coverage and protection of bark areas. The hypothesis was that best coverage of tar- get sites will be achieved when all 6 Sides of the cube have sufficient spray deposits. 'Sufficient' is defined here with respect to the LC50 values as determined in section IV. The researcher is aware that this anaIOgy cannot be entirely correct due to the fact that an artificial sampling device was employed, rather than actual target sites, i.e. the branches and twig crotches. Also, for this analogy to be correct, it must be assumed that the twig crotches face all directions in equal pro- portion, correSponding to the 6 Sides of the cube. The average methoxychlor deposits on a sample cube per height level were calculated in the following manner. Each tree had 5 sample cubes per height level. The means were computed for each side of the sample cube. First, the average for the two horizontal filter disks, 5 and 6, were calculated. The vertical filter disks on the cube were designated 1, 2, 3, and 4 (Figure 34) according to amount of MeCl deposit, from highest (1) to lowest (4). The means with their associated standard errors are listed in Table 13. The size of the standard error was, in many cases, extremely large, expressing the great variation in spray deposits within each height level. Both spray applications, mist blower and helicopter, gave highly variable deposits at each respective height level. I32 1. Helicopter application On the average, about 2-4 times as much methoxychlor was deposi- ted in the upper height level than in the lower tree region (Table 13). Of the 6 filter disks on a cube, only 3 had appreciable deposits. This was consistent in both height levels. The tOp disk (No. 5) always had the greatest amount of methoxychlor. Only 2 vertical side disks harbored spray deposits, while only small amounts of methoxychlor were detected on the remaining sides Of the cube. Deposits were insignificant on the bottom side (No. 6) of the sample cube (Figure 40a,b). The experimental results suggest that approximately only 50% of the total bark area and, predominantly, those areas exposed to the tOp would be sprayed. Half on the vertical bark area would receive spray, and to a lesser extent. Residue amounts in the lower tree region were much smaller (Figure 40a,b). The LCSO determination on helicopter-sprayed twigs yielded a value of 41.8 micrograms/cm2 (p. 150) which is repre- sented as a dotted line in Figure 40a,b. Only the deposit on one hori- zontal tOp disk approached this line. Since these figures signify fresh deposits, one can imagine that the degree of protection after several weeks of weathering is even less. By putting the deposit figures from the sample cubes in relation to the LCSO one can explain the low beetle mortality in the 1970 helic0pter treatment (27.0% in the tOp tree region 2 weeks after application, and 17.3% after 8 weeks). If all twig crotches in the upper region projected towards the tap, an average residue of roughly 40 micrograms/cm2 could be expected in a twig crotch according to the deposit on disk No. 5 in Figure 40a. A deposit of this magnitude would cause 50% mortality in a given pOpulation. However, since only part of the twig crotches in the upper tree region are exposed to the t0p, and the rest project in 133 Home“ Ocov monau O_aEmm m EOLH ovum—:o_mu Am Ammo.“ 00.56. monau O_aEmm m. EO.H ovum—:o_mo AN 6m o.:m.m mow m1. moo_m Ho co_u_moa .OH H. .m.m m.~6 6.6. 6.6. ...m ..mm m.6H zO4 co_um_:e co_umo onzo O_aEmm Ho 01— mou_m co mu_moaov .o_zo>xO£uoz ucm.OI 1.06 _ooz 1_Haa< emo> .mOO.u E_O Ho m_o>o_ u;m_oc N um monau O_aEmm co mu_mOQOp .o_cu>xo;uOE Ommco>< .m. O_an 134 height level 1: 15m height level 3: 5m MeCl deposit(pg/cm2) 40‘ —————————————————————————— ____mm a m- I H [—1 m l 2 3 4 5 6 l 2 3 4 5 6 ”q ——————————————————————————————— "L050 b 20- l I l L_L FZL—FH J L. I 2 3 4 5 5 I 2 3 4 5 6 ‘ 118 m- Ifil m- C so- r—I w- 20- u-1 I 2 3 4 5 - 214 199 m- 100~ ' ' F— d 00. 1—1 w- FJ ‘0‘ ...l m. _ - ------ _-____-_--_---_-——--wm I 2 3 4 5 G l 2 3 4 5 5 sides on sample cube Figure 40. Average methoxychlor deposits on the six sides of the sample cube at two height levels of treated elm trees: helic0pter, (a) 12.5% emulsion + Dacagin, (b) mist blower, (c) 12.5% emulsion + Dacagin, (d) 12.5% emulsion; 12.5% emulsion. l35 various directions, coverage, and, therefore, mortality, will be less (as indicated by the distribution of deposits on the other sides of the sample cube). This points out that the actual effectiveness of a spray application can not be judged by the amount of spray which is trapped on a horizontal filter disk. Two 12.5% methoxychlor emulsions, one with Dacagin, were tested. Three-dimensional coverage was the same with both formulations (Figure 40 a,b). Deposits on the tree sprayed with the straight l2.5% emulsion were smaller. Because of insufficient replication, no valid conclusions can be drawn concerning the superiority in coverage of either formulation. 2. Mist blower application The horizontal variation in spray deposits at each height level was substantial and comparable to the helicopter application (see the standard error values, Table 13). The actual deposits were up to ID times greater than for the helic0pter application. The vertical distribution pattern of spray deposits was the reverse of the helicopter pattern. The bottom tree region was heavily overloaded, receiving about 3 times as much spray as the t0p region. Amounts of close to 700 micrograms/cm2 were found on single disks in the lower part of the tree. Coverage of the average sample cube in the lower height level was complete. Even the tap disk had sufficient coverage. Towards the tree top, the deposits became weaker, but coverage remained excellent for all sides except the top disk. Making the analogy from the sample cubes to the target sites, coverage of the twig crotch areas was more than sufficient in the lower tree portion and decreased towards the top. The amounts of the initial deposits suggests that, even after weathering for many weeks, residues would be sufficient to protect the feeding sites. Since the I36 dosage-response regression equation for the mist blower-treated twigs was distorted because of DDT residues, the LCSD figure from the lab- treated twig crotches was drawn as a dotted line in Figure 40 c,d. The mortality figures from the mist blower treatments of l969 and 1970 at several height levels correspond well with the vertical distribution pattern of spray deposits as demonstrated here. Mortality in the lower height levels throughout the season never fell below 90%. The percentage drOpped significantly in the t0p on some mist blower-sprayed trees. The researcher intended to compare the performance of a straight 12.5% emulsion to a formulation of the same strength, but with Dacagin. However, after the spraying was completed, it was discovered that the Dacagin formulation was considerably diluted to a spray emulsion of undetermined concentration. The lower deposits in Figure h0c reflect this. Therefore, no conclusions can be drawn as to the effect of Dacagin on tree coverage with the mist blower. Summary and Conclusions Because of the confounding effect of carry-over in the evaluation of formulations and application methods, a sampling system was develoPed to objectively monitor spray coverage of elm trees. Cubes with filter disks on all sides were employed as sampling devices to investigate spray patterns, drOplet density and size spectrum, 3-dimensional cover- age and actual deposition. The helicopter applications gave discrete spray patterns. Less than 20% of l cm2 was covered when the spray drOplet area was summated. The addition of Dacagin to the l2.5% spray emulsion shifted the droplet size spectrum to the larger sizes. Straight emulsions had a more dispersed spray pattern. The poor 3-dimensional coverage of the sample cubes suggested that helicopter spraying will protect less 1:2:- 137 than half of the total bark area. Besides, the dosages applied to each tree were found to be too small to provide the necessary protection. A comparison of fresh deposit data from the tree tOp with the LCSD value for helic0pter-treated twigs indicated that not more than one-third of the potential feeding sites would be protected. After weathering for several weeks, protection would diminish, eSpecially in the lower tree region. The mist blower application created almost continuous deposits and 3-dimensional coverage was excellent, even in the upper tree region. Actual deposits often exceeded several hundred micrograms/cm2 and suggested long-lasting protection. One can deduce from the superior coverage of the sampling cubes that the majority of the potential feeding sites will receive spray protection. This monitoring method nicely showed the dis- tribution pattern of helic0pter- and mist blower- applied spray within the whole tree region. The mist blower gave higher deposits at the bottom while the helic0pter favored tree top deposits. The bioassay results of the helic0pter- and mist blower-treated twigs in l970 were in close agreement with the findings of the cube sample experiment, which demonstrated the reliability of this monitoring method. IV. Insecticidal and Decomposition Properties of Methoxychlor Formulations Introduction The elm bark beetle conference in I948 (Scanlon, I949) laid out certain research objectives for Dutch elm disease control, among which were quantification of bark residues and correlation with bioassay data (feeding response or mortality). Pesticide analysis at that time was in its infancy, and analytical tools were not sufficiently refined to give accurate results at the desired levels. In this study one of the objectives of the residue analysis of field-collected bark samples was to demonstrate differences in coverage which were attributable to certain formulations and application techniques. This was done in section II of this dissertation. Furthermore, an effort was made to determine residue levels which were sufficient to prevent successful feeding to the xylem. Matthysse st 31. (l95h) evaluated DDT formulations applied by hydraulic spray gun or by mist blower to tall American elms (ls-2] m), with chemical analysis of bark residues and with feeding tests. Assays with twigs dipped in l.0, 0.25 and 0.05% DDT emul- sions (ll3, 23 and 5 micrograms/cm2 bark surface respectively) revealed 2 and more gave sufficient control, while that deposits of #0 micrograms/cm deposits below 20 micrograms/cm2 resulted in poor protection (for both native and European elm bark beetle) (see also Miller, l95l). Other investigators made use of chemical analysis of bark residues to measure the performance of application techniques and formulations, but they did not relate actual residue with bioassay data to give it biological meaning (Thompson, I965; Wallner and Leeling, l968; Wallner._£._L., l9b9). It is common knowledge that finely dispersed residue is more toxic than residue of the same amount per unit area but with less subdivision. I38 I39 Subdivision of spray drOplets is primarily influenced by the application technique and formulation (Hoskins, I962). It was therefore expected that the residue-mortality correlations would not be the same for the mist blower and helic0pter treatments because of the different residue patterns on the bark. LCSO values were calculated from bioassays where twigs were dipped in methoxychlor solutions of known concentration. For realistic evaluation of the contact toxicity of insecticidal deposits, it is desirable to use solid support surfaces on which the in- sect dwells in a natural setting (on leaves, on soil, on bark, etc.). For the purpose of uniformity and convenience, researchers have often resorted to supporting surfaces which were easier to handle. Lyon (I965) measured the residual contact toxicity of several insecticides to bark beetles on fiber boards. He argued that natural bark surfaces introduced too much variability in the way deposit patterns developed from the same formu- lation and this prevented good reproducability. Cuthbert _£_§l, (l97l) used the World Health Organization pro- cedure to determine the contact toxicity of DDT and methoxychlor applied on filter paper to S, multistriatus and to its braconid parasite, Dendro- soter protuberans, In these tests DDT was about lOO times more toxic than methoxychlor. Although the contact toxicity of methoxychlor was shown to be low, it was of interest to demonstrate the contact toxicity of various formulations. Further, experiments were carried out to measure if, and to what extent, the feeding response was influenced by prior contact with the surface deposits of methoxychlor. The behavior of methoxychlor deposits on elm bark was also in- vestigated. The type and structure of bark deposits was described and lhO documented for various formulations and concentrations. The possible relationship of deposit structure to contact toxicity and weathering received considerable attention in these experiments. The rate of decomposition of insecticidal deposits can vary greatly depending on the formulation employed (Barlow and Hadaway, I952; Hoskins, I962). Wallner _£_§l, (l969) reported highest methoxychlor residues on trees which were treated with a formulation containing the anti-drift material, Dacagin. This suggestion led to several experiments where the break-down of methoxychlor formulations applied to elm branches was fol- lowed over a whole year. When an elm tree is sprayed, many bark areas are covered which are not potential feeding sites. It was often Speculated that methoxychlor residues might be translocated passively from the original site of depo- sition to another location and possibly be accumulated in the twig crotches. Also, if this assumption was correct, discrete spray patterns where drOplets are far apart would become continuous over time if passive movement of MeCl residues existed. Therefore, it seemed worthwhile to supply information on the possible movement of MeCl deposits and trace their displacement. Materials and Methods (A) Bioassay Procedures for Oral LCSO Determinations In I970, the majority of elms on the campus of Michigan State University received the same helic0pter application. Twigs were sampled from IO helicopter sprayed trees and 2 mist blower treated trees two weeks after application. Only 5 helic0pter treated trees and two mist blower sprayed trees were sampled eight weeks after application. Samples were collected from 3 height levels and subsequently subjected to feeding 14] tests with S, multistriatus and chemical analysis for methoxychlor resi- dues (for details of sampling, bioassay and chemical analysis, see section II). Regression equations were computed from the data pairs, amount of residue (micrograms/cm2 bark surface) and mortality (%) after conversion to a log-probit scale. Regression lines were fitted (a) to the helic0pter data from 2 weeks after application, (b) to the helic0pter data from 8 weeks after application, and (c) to the mist blower data from both sample periods (data were pooled because of the small number of sample trees). Twig samples from helic0pter sprayed trees with considerable DDT resi- dues were not considered for the analysis. DDT residues were present in all mist blower twig samples, in several cases almost equalling the amount of methoxychlor (Table 14 ). For the lab tests, twig crotches were collected from untreated mature elm trees, cut to size and stored in a freezer at -IOo C. Pre-‘ Iiminary feeding assays with twig crotches dipped in methoxychlor solu- tions of contrasting strength were performed to get a starting point for the LC50 determination. Consequently, the following methoxychlor solutions (xylene) were prepared from a commercial stock solution (Michlin 'MX-Z', 25.l%): 0.l%, 0.l5%, 0.2%, 0.3%, 0.4%, 0.5%, 0.6% and 0.8%. Ting crotches were dipped for two seconds in each solution respectively arui then placed upright in a ply wood tray. After a drying period of 3-1llmnns, the treated twig crotches were subjected to the standard biioassay with S, multistriatus. Extended drying periods beyond A hours were avoided since they resulted in a poor feeding response due to desiccation of the twig crotches. Bark samples were taken as described in section II from around 50 twig crotches from the same batch for each concentration. They were then chemically analyzed to determine the I42 Table I4. Twig crotch residues and respective mortalities on sprayed campus elms (I970). Treel) Height 2 Weeks After Application 8 Weeks After Application Level MeCl DDT Mortality MeCl2 DDT 2 Mortality rig/cm rig/cm2 % rig/cm [Ag/cm % A l 6.l 5.6 53.5 4.4 3.l 20.5 2 20.] 14.8 94.8 9.6 8.2 78.8 3 37.9 30.3 I00.0 26.9 18.7 100.0 8 l 5.3 4.2 l.8 2.9 4.2 25.8 2 15.3 15.11 79.3 5.9 2.5 47.0 3 l7.8 l5.0 79.3 l2.4 ll.5 68.0 C I 22.2 3.3 32.8 28.I l3.l 36.4 2 4I.8 l9.7 53.5 15.2 Trace 36.4 3 2l.l 20.4 17.3 3.3 l5.2 D l 1.0 Trace 6.9 6.2 25.8 2 l.8 l.l 4.6 3 l.8 l7.8 5.6 25.8 E I 30.3 0 7 53.5 13.3 25.8 2 2l.8 l 2 63.8 9.8 9.9 3 26.0 1 L1 69.0 11.5 G l 5.0 12.l 0.5 4.6 2 2.2 l2.l 0.9 4.6 3 3.0 27.6 2.5 0.9 H I 23.4 Trace 63.8 l.5 4.6 2 0.5 0.2 1.8 2.] 3 1.9 0.5 17.3 2.0 4.6 J l 9.8 l7.3 4.7 l5.2 2 ll.2 12.l 3.4 4.6 3 ll.9 2.1 32.8 l.2 Trace K I 4.5 38.0 Not Sampled ,2 l6.9 32.8 3 30.3 2.3 32.8 L I 15.4 lZ.l Not Sampled 2 15.9 22.4 3 l.8 M 1 3.7 6.9 4.7 15.2 2 11.7 6.9 7.9 15.2 3 5.6 l.8 5.8 l .2 ]) trees A and B: mist blower-sprayed all others: helicopter-sprayed I43 amount of residue on the bark surface (micrograms/cmz). The control treatment consisted of 50 twig crotches dipped in xylene. Finney's probit analysis was employed to calculate the residue-mortality equation and the LCSO’ the median lethal concentration (Finney, I964). (B) Fiber Board Units for Contact Toxicity Testing Several attempts to test the contact toxicity of methoxychlor deposits on elm bark failed because the bioassay insects invariably be- gan to feed in bark fissures. Therefore, a supporting surface was chosen which was by its nature not stimulating to S, multistriatus and would not induce a feeding response. Untempered l/8” Masonite fiber board had the desired characteristics: easy to handle and absorption qualities similar to bark. The fiber board was cut into smaller pieces measuring 7.5 x 7.5 cm. A special pen-like applicator was constructed to distribute the measured amount of insecticide over a 4 x 4 cm square. The applicator (Figure 4l)was made of a 2 cm wide thin aluminum sheet metal strip which was folded to an acute V on one side and then bent over. An equal amount of each respective formulation was released with a pipette into the V- shaped applicator. Four tiny holes at the bottom of the V permitted the liquid to flow out slowly when the applicator was drawn across the fiber board surface. It took 2 timed strokes with the 2 cm wide applicator to distribute the measured amount evenly over the 16 cm2 area. Another problem was to keep the assay beetles on the treated surface during the duration of the experiment. Lyon (I965) placed a watch glass over the assay insects, but because of possible fumigation effects during the long exposure time (60 minutes), plastic weighing cups (l 5/8” x l 5/8”)l were used instead. Large holes were punched in the bottom of the dish 1) Scientific Products Figure 4i. Application of methoxychlor formulation on fiber board with pen-like applicator. Figure 42. Disassembled fiber board unit with plastic dish and screen for contact toxicity testing. Figure 43. Beetles in gelatine capsules No. 3 after exposure to treatal fiber boards. Figure 44. Decomposition experiment: bark sampling of treated elm branches. mm 13 .1 and 5:15 ml 5': 145 for prOper ventilation. Then the cup was placed upside down above the treated surface and taped down (Figure 42). Unsexed beetles were intro- duced with a vacuum forceps, then a small piece of plastic screen was placed above the hole to prevent beetles from flying out. The inside wall of the cup was rubbed with emulsifier (Emcol 5l00) which kept the assay beetles from climbing up the side wall. Five beetles were tested at a time in one unit over a period of one hour. The choice of exposure time had to be arbitrary since it seemed impossible to determine the average time a beetle will Spend moving around on the bark surface before feeding activity is initiated. After exposure each beetle was placed individually in a gelatine capsule No. 3 to guard against cross contamination (Figure 43). The physiological condition of the beetles was checked and recorded every 6 hours. The same criteria were used here as described for the bioassays in section II: (a) A alive (b) (A) dying (c) + dead For the final analysis (b) and (c) were pooled. The total obser- vation period was 48 hours. The following formulations were tested for contact toxicity: 6.25% MeCl solution (xylene) 6.25% MeCl emulsion (xylenezwater = l:l) 6.25% MeCl emulsion plus Dacagin (2 pounds/I00 gallons) All formulations were diluted from a 25.l% MeCl xylene stock so- lution (Michlin 'MX-2' elm spray). Exactly 0.] ml from each formulation was distributed evenly in a 4 x 4 cm square on the fiber board as des- 2 cibed earlier. The actual deposit per cm amounted to £2, 400 micrograms. Deposits of this magnitude can be found on the heavily overdosed lower I46 portions of mist blower sprayed elms. Two replicates with 25 beetles each were fun for each formulation and the control with untreated fiber boards. Time-mortality curves were established for fresh deposits (l2- hour drying period) and for 4-day old deposits. 8 To determine the influence of contact with surface deposits on their feeding response, beetles were exposed to deposits on fiber boards as described above. Three emulsions were tested: 6.25%, 3.l2%, and l.56%, which corresponded approximately to 400 micrograms/cmz, 200 micro- grams/cmz, and l00 micrograms/cm2 actual MeCl deposit respectively. After a 60-minute exposure to the treated fiber boards, the beetles were each caged in a residue-free twig crotch. Following 24 hours, the feeding test was stOpped and mortality and feeding response determined. The de- posits on the fiber boards were aged for ID days. Twenty-five beetles were tested per concentration in 2 replications. (C) Application of Methoxychlor Formulations on Elm Branches to Investi- gate the Behavior of Bark Deposits l. Structure of surface deposits Small sections of elm branches (l5 cm long, 2-2.5 cm diameter) were sprayed with three concentrations of three formulations. All formu- lations were made from a 25.l% MeCl stock solution (Michlin 'MX-Z' elm spray): Formulation Concentration (a) solution l2.55% MeCl (xylene) 6.25% MeCl 3.l2% MeCl (b) emulsion same as (a) (xylene:water lzl) (c) emulsion same as (a) (as (b) but with Dacagin 2 lbs/IOO gal) l47 An equal volume of each formulation was applied to a branch section with a chromatography spray unit. The l2.55% formulations gave roughly an initial deposit of 500 micrograms/cm2 and the other concentrations corre- spondingly less. The crystallization patterns of the surface deposits generated by the various formulations and concentrations were then obser- ved and documented as they developed. 2. Decomposition of bark deposits Bottom branches of mature elms were sprayed with several methoxy- chlor formulations. The branches had a western exposure, a diameter of l-2 cm and the sprayed portion measured between l.5 m to 2.0 m. It was attempted to make the total bark area for each treatment similar in size. Ten ml of each formulation were applied with a chromatOgraphy Spray gun to the bark area facing south. The spray settled down as a continuous deposit over the whole treated surface. Three days after application, the bark was sampled for the first time to quantify the original deposit. The bark sample was taken with a No. l cork borer (diameter 0.4 cm) by punching out bark circles (Figure 44). A total of 50 bark circles was sampled randomly from a branch amounting to 6.28 cm2 total bark area analyzed per formulation. Table 20 lists the formulations tested, the respective application and sampling dates plus the actual residue data in micrograms per cm2 at the start of each experiment. The break-down of these formu- lations was studied in 2 experiments during l970/7l and l97l/72. All formulations in these experiments had a concentration of l2.55% MeCl and were diluted from 25.l% Michlin 'MX-Z' elm spray. 3. Movement of bark deposits Two twig crotches (2-3 years old) on a young elm at the edge of a woodlot were chosen for this experiment. The position of these twig I48 crotches was close to vertical. The bark area around the twig crotch was divided into I cm sections and marked (Figure 5]). The application was made in section B with 5 microliter of a l2.55% emulsion which was equivalent to 599.7 micrograms of actual methoxychlor. A Hamilton micro- syringe was employed to dispense this small amount in the Specified bark area. After 4 weeks of weathering the two twig crotches were cut and the bark sections A to D were stripped from the branch. Then each bark section was chemically analyzed for amount of residue. Results and Discussion (A) Oral Toxicity of Methoxychlor Deposits in Elm Twig Crotches l. Residue-mortality regressions calculated from field samples a. Helic0pter application The data pairs from the first sampling period showed great vari- ation around the calculated dosage-response regression line (Figure #5). The bioassay beetles came from the same batch. Therefore, the encountered heterogeneity of the data could not be attributed to an inconsistent re- sponse (other than random variation). Non-uniform, discrete distribution of bark deposits, bark sampling and chemical analysis of methoxychlor residues caused this wide spread of points. The researcher believes that, in particular, the non-uniform, discrete distribution of helic0pter spray deposits on elm bark contributed significantly to the variation. Contin- uous carry-over deposits from earlier mist blower treatments (before l968) and their different toxicity could have added a further component to the total variation. The low coefficient of determination of r2 = 0.4l expresses the high unexplained variation. I49 %mortality ”-1 0 ill- log probit regression 00 Y- 341+ -981 X o ”-1 r2 - .41 m_. 0 w_‘ o o o o o O a) 211— o o o 0 O O ”—1 o o LCgo- 41-8 ug/cm2 5-1 p 1 1 r 1 1 I 1 1 1 ' 1 1 2 4 5 l 10 ii :11 fl W bark deposit (HQ/cm?) ”—1 Y- 3-16 + 1053 X o b) zo— l0“ 5— p l 1 1 1 r I 1 T 1 T Ti 2 4 6 I I0 15 20 30 “I Figure 45. Bark deposit-mortality regression line calculated for helicopter-treated twig crotches (a) two weeks after application (b) eight weeks after application. l50 The L050 was calculated from the log-probit regression equation in Figure 45 which yielded a value of 4I.8 micrograms methoxychlor resi- due per l cm2 bark surface. Since this value is extrapolated (just out- side the range of sample points) it must be regarded with caution. The data from the second sampling period were closer to the com- puted regression line which resulted in a higher r2 = 0.67. The log- probit equations from the two sampling periods agreed adequately, con- sidering that the second regression line was constructed from data where the highest mortality was 36.4%. The extrapolated value for the LCSO, as predicted by the equation in Figure 45, would be 55.9 micrograms/cmz. b. Mist blower application The beetle response in the assays of mist blower-treated twigs had the range necessary for the construction of a valid dosage-mortality equation. The data were in good agreement with the regression line (Figure 46) which was indicated by a high r2 = 0.85. The relationship between dosage and mortality, as shown by the log-probit equation in Figure 46 for the mist blower treatment, was quite different. The SIOpe of the regression line was more than twice as steep as for the helic0pter regressions. The higher toxicity of mist blower-treated deposits was attributed partly to better distribution and finer dispersion of residues. The log-probit equation gave a L050 of 6.5 micrograms/cmz, which was roughly one-sixth of the LC50 on helic0pter-treated twigs. However, the regression line was computed with residue data which had high amounts of DDT (see Table I4). The beetle reSponse was, therefore, caused by the combined action of both compounds, methoxychlor and DOT. The calcu- lated log-probit regression is, therefore, not indicative of the rela- tionship between methoxychlor bark deposits and mortality on mist blower lSl .mogob0co m_zu woumocuncozo_n um_E Low boom—:o_mu uc__ co_mmocmoe >u__muL051u_moaon xcmm .w: ocam_u €833 .388 {3 8 8 8 a 2 IF — p — p — - °—— N w W m I In 158 ma 130.. 12 18 18 8. 'N.- I. x35 + Xmé 1> 53852 .395 co. 0 i 18 18 I! 18 T: 3.3.58... 152 sprayed twigs and underestimated, consequently, the L050. 2. Responses of S, multistriatus on lab-treated twig crotches Both reSponses, mortality and feeding, were tested over a wide range of methoxychlor concentrations (Table IS). The observed mortality was not corrected since no natural mortality occurred in the control assays. The results of the probit analysis of the deposit-mortality data are listed in Table 16. This regression equation differed consid- erably from the equation for helicopter treated twigs. The higher beetle response (mortality) on the lab-treated twig crotches was apparently caused by the continuous uniform distribution pattern of bark deposits. The L050 was 12.7 micrograms/cm2 which was about l/3 the value calcu- lated for the helic0pter treated twigs. A comparison with the results of the mist blower assays is difficult to make because of the inter- action with DDT. The substantial difference in response rate on the field-treated and lab-treated twig crotches suggested again that the out- come of lab tests is not necessarily indicative of happenings in the field. The slight heterogeneity of points around the regression line in Figure 47 is partly the result of a longer time lapse between bio- assays. This meant that the physiological response was not completely uniform in all assays since beetle material from different batches had to be used. Further, the inherent variability of the twig crotch ma- terial certainly contributed to the heterogeneity. A X2 test for good- ness of fit was almost significant at cX-= 0.10, therefore the variances were corrected with a 'heterogeneity factor' as suggested by Finney (1964). This correction gave a somewhat wider fiducial belt. The actual feeding reSponse was measured for some bioassays as 153 Table 15. Response of S, multistriatus in bioassays of lab-treated twig crotches. Treatmentl) MeCl No. Feeding ReSponse DeposiE Beetles Mortality Mean Scar Conf. ()Ag/cm ) Tested ‘% Length(cm) Limit(c£ = 0.05) l) 0.1% 4.1 100 4 —- -- 2) 0.15% 5.2 100 10 -- -- 3) 0.2% 7.5 51 20 0.226 0.045 4) 0.3% 11.7 50 38 -- -- 5) 0.4% 15.4 55 73 -- '- 6) 0.4% 16.7 52 50 0.190 0.033 7) 0.5% 21.1 53 77 0.139 0.026 8) 0.6% 23.6 53 87 0.098 0.026 9) 0.8% 31.2 52 96 0.111 0.019 10) Control 1 -- 51 -- -- -- (xylene) 11) Control 2 -- 53 -- 0.327 0.047 (xylene) fl) bioassays l, 2, 4, 5, and 10 performed prior to the others; twig crotches were dipped in MeCl xylene solutions Table 165 Results of Finney's probit analysis of the dosage-mortality data in Table 15. 0.989 + 3.635x log-probit regression equation: y SIOpe t SE = 3.635 t 0.326 7&2 = 11.93 n.s. (oc= 0.05) Lc50 (Hg/cmz) = 12.7 95% C.L. =11.1-1L1.5 LC90 ( Mg/cmz) = 28.6 154 .Am_m>_mcm u_noLa m.>occ_uv monouoLo m_3u puummcuunm_ Lo» coup—:o_mo oc__ co_mmocmoc >u__muLOEIu_moaop xcmm $8063 .688 {on IafivNRo—zfie—u a w h p _ — _ _ _ _ _ _ p _ _ _ \ N «.83; Z? 80.. a l .o ..a. r.e~ r.an 1.9. 1.9m azwn .maoan;._xmxw 1.90 1.e~ ..e- xnngw:mu.+ mzxw. ..er 1.99 86852 .505 no. I...» I..o 3:28.: a .N: otam_u 155 indicated in Table 15. Naturally, the feeding response became weaker as the methoxychlor deposits increased. Accordingly, the lengths of the feeding scars became more uniform for each treatment, which is nicely shown by the narrowing confidence limits associated with the respective means in Table 15. 3. Comparison of male and female feeding response on untreated twigs For an actual comparison of male and female feeding response it was necessary to use a more accurate way of measuring the feeding. In- dexing alone was not refined enough, therefore the feeding scars were measured accurately to 0.1 millimeter (mm) with a disc micrometer reti- cle in a LEITZ stereo microsc0pe. The feeding data were taken from the controls for the oral tox- icity tests. The total sample amounted to 102 feeding scars, 47 of which were made by females, 55 by males. The mean length of a feeding injury for the males was 3.24 mm, for the females, 3.35 mm. The calculated t = 0.337 was smaller than t = 2.01 ( ci.= 0.05); indicating that male and female feeding responses were not Significantly different. For a graphic demonstration (Figure 48) of the distribution of length of feeding scars, the feeding responses were grouped in size classes from .05 - .15 to .55 - .65 cm for both males and females. 4. Relationship between mortality and sex in the bioassays The helicopter bioassays of July, 1969, and the oral toxicity tests (July, 1971) were examined with respect to the association of beetle sex and mortality. The percent mortality was determined for each sex separately in each assay using the total number of males and females as 100% (Table 17). Differences between male and female mortality were then computed and marked with the prOper plus or minus sign. Wilcoxon's 156 .mmsouoLo m_3o woummcuc: co mmcoammc mc_voow o_mEmm 6cm a.mz .m: ocsm_m mo. mo. um mm. mm. x 23.2 9 mm 3.3% 450» wwwwocumwmwmo 9 or m. m o. m. ow _ _ _ _ _ _ . . _ _ _ C _ _ _ p @539 :0: 9. 1 mo. mm. 1 9. - _1 F H as 1 mm. r H— mm. 1 ma. - fi fi 8. 1 mm. - All—1 2.8 mmflo cum 0 , 1 b 157 Table 17. Association of mortality and beetle sex in bioassays of meth- oxychlor-treated elm twigs. MeCl N Treat- (Males + Males Females Mort.]) Sexz) ment Females) Mort.% Mort.% Difference Ratio Bioassays, 1969, July: HelicOpter treatment II A 54 35.5 26.1 + 9.4 1.35 11 B 55 22.2 10.7 +11.5 0.96 IV A 55 55.2 40.0 +15.2 1.16 IV B 55 39.3 33.3 i 6.0 1.04 V A 55 21.4 18.5 + 2.9 1.04 V B 55 62.5 45.2 +17.3 0.77 Bioassays, 1971, July: Oral Toxicity Tests 0.2% 51 23.8 6.7 +17.1 0.70 0.3% 51 37.0 41.7 - 4.6 1.13 0.3% 54 82.6 62.5 +22.6 0.72 0.4% 50 52.4 51.7 + 0.7 0.89 0.4% 50 23.1 16.7 + 6.4 1.08 0.5% 53 82.6 62.5 +22.6 0.72 1) Mean difference in susceptibility: 0': 9.7% 2) Sex ratio (males/females): Mean 5. R. = 0.96 i 0.13 (C.L. for a£=0.05) 158 signed rank test for 2 groups, arranged as paired observations, was per- formed on these data (Sokal, R. R., and F. J. Rohlf, 1969, p. 400). The difference between males and females with reSpect to sus- ceptibility to methoxychlor was significant at 06 = 0.01. These data do not suggest to which degree the males are more susceptible than the females. The fact that in 11 out of 12 observed bioassays the males responded with higher mortality suggests that a difference exists. On an average, the females were 9.7% less susceptible than the males. Al- though this difference in‘susceptibility might be important for precise toxicological investigations, it can probably be disregarded in practical bioassay work. Cuthbert t al. (1971) did not find a difference in sus- ceptibility between sexes when S, multistriatus was exposed to methoxy- chlor-treated filter disks. The mean ratio of males to females in these 12 assays was very close to unity: Sex Ratio = 0.96 i 0.13 (C. L. at d1= 0.05). Since the beetles used for each assay were a random sample from a field pOpu- lation (emerged under lab conditions), the above sex ratio is probably a good estimate for the true ratio of males to females in a natural S, multistriatus population. (B) Contact Toxicity of Methoxychlor Deposits 1. Comparison of 3 methoxychlor formulations The toxicity pattern which emerged from these experiments cor- reSponded well with previous work (Barlow and Hadaway, 1952; Hoskins, 1962). The contact toxicity of the 12-hour old deposits was somewhat higher than for the 4-day old, weathered deposits (Table 18). Barlow and Hadaway (1952) pointed out that the physical state of a surface de- posit determines its toxicity and that a reduction of toxicity usually 159 goes hand in hand with crystallization. The 12-hour old deposits were probably not completely crystallized and consisted of a mixture of cry- stals and supersaturated droplets. The reduced toxicity of the weathered deposits can be attributed to complete crystallization of the mixture since it is unlikely that significant weathering occurred during this relatively short period. Table 18. Contact toxicity of methoxychlor deposits on fiber boards to S, multistriatus. Time After 6.25% 6.25% 6.25% Emulsion Exposure Solution Emulsion + Dacagin Control (hours) Total Mort.% Total Mort.% Total Mort.% Total Mort.% A') 32) A B A B A B 6 9 2 24 19 10 2 2 -- 12 17 8 59 43 17 16 2 -- 18 27 14 84 61 30 24 8 8 24 48 32 89 67 38 32 24 12 30 69 40 93 73 60 44 38 16 36 81 52 98 77 75 60 48 36 42 923) 68 1003) 83 883) 70 603) 56 48 923) 84 1003) 88 923) 78 783) 72 l) deposits aged for 12 hours, mean mortality of 2 replicates (25 beetles 2) Szgggits aged for 4 days, mean mortality of 2 replicates (25 beetles eac 3) mortality of 25 beetles The solution penetrated well into the absorbent fiber board sur- face despite its high volatility, and laid down a thin surface deposit. The emulsion, on the other hand, stayed more on the surface and the quan- titative relationship between surface and tissue deposit favored the 160 91. mortality ' - /' ’ - I . ,-’ l¢emulsion+ solut1on—_/' , Dacagin / c 12 11 24 3': 3'; I, T 1" 7 hours control Figure 49. Time-mortality curves for _S_ multistriatus after exposure to treated fiber boards: (a) deposits aged for 12 hours (b) deposits aged for 4 days. 161 former. The emulsion containing the Dacagin behaved in the same fashion; it penetrated little and the deposit accumulated on the surface. The degree of contact toxicity was directly related to the type of surface deposit which was formed by each formulation. The emulsion caused highest mortality because of richer surface deposits and better pick-up of methoxychlor crystals. When the thickening agent, Dacagin, was added to the emulsion, the toxicity drOpped markedly and was similar to the solution (Figure 49). Obviously, this gel-like, viscous anti-drift material coated the surface with a thin film which prevented direct contact between insect and residue to some extent and made 'pick-up' more difficult. Since the mortality in the controls was rather high in both series of the experiments, the treatment mortalities were not corrected. The time-mortality curve of the control is graphed for comparison and was a little steeper in the experiments with the 12- hour old deposits. The handling of the bioassay beetles after exposure might have increased mortality overa1l, since the gelatine capsules offered little room for movement and were a completely artificial en- vironment for the beetle. The bioassay beetles responded in the same fashion to the fresh and the aged deposits. In both cases, the emulsion was the most toxic, the solution and the emulsion plus Dacagin were comparable in their toxic action. 2. Contact toxicity and feeding response The methoxychlor deposits which the beetle encounters upon emer- gence have experienced several weeks or months of weathering. As sur- face deposits they have undergone changes in their chemical and physical make-up.) The treated fiber boards in this experiment had been exposed to outside conditions for 10 days in order to approximate the natural 162 conditions somewhat. The following Table gives a summary of the beetle response in these assays. Table 19. Beetle mortality and feeding reSponse after exposure to treated fiber boards. MeCl emulsion Mortality % Failure to Mean Scar Score % Length(cm) 1.56% 4 32 22.2 3.12% 6 30 21.5 6.25% 6 36 21.1 Control 4 28 22.7 Fumigation 6 22 25.2 Control The cause of poor feeding response in the controls and in the experi- ment as a whole is not quite clear. The twigs were untreated and were collected from young elm trees with smooth bark in late summer. Meyer and Norris (1964) suggested that feeding is motivated in part by rough bark. Dixon (1964) deduced from his data that feeding in twig crotches is primarily a thigmotactic response. This is one explanation. Elm twigs collected in late summer are not as succulent as during the early season, which could be the cause for the poor feeding reSponse. However, this researcher believes that a comparison to the control is still valid. The mortality and the feeding response in all treatments were not dif- ferent from the control. The time-mortality curves in Figure 49 Show high mortality after an observation time of 24 hours for the fresh and aged 6.25% emulsion deposits. After 10 days weathering, the surface 163 deposits on the fiber boards had apparently lost their contact toxicity and this resulted in insignificant mortality. The mean values for the measured feeding response decreased a little with increasing concentration. The difference to the control proved to be insignificant. To test for possible fumigation effects in these experiments, fiber board units were used where a screen separated the treated surface from the assay beetles. As the low vapor pressure of methoxychlor suggested, no fumigation effect was noticeable. These experiments bore out that the contact toxicity of weathered methoxychlor deposits is extremely low and that the feeding response does not appear to be influenced by prior contact with aged deposits. (C) Structure of Surface Deposits of Methoxychlor Solution The 6.25% and the 12.55% solution produced a fine network of long crystals (Figure 51b) but the bark had still its dark natural appear- ance. Very light crystal formation was observable on bark treated with the 3.12% solution. Crystals, thin and long, sometimes radially bran- ching out from a center, adhered tightly to the bark surface. Emulsion All 3 concentrations produced surface deposits primarily of large-sized crystals and resulted in continuous coats of whitish appear- ance (Figure 50b). Surface deposits were heavier on the branch sections treated with the higher concentration. Crystals ranged from very small to large and some were definitely plate-like (Figure 50c). The radial crystal type was also common, but in contrast to the solution, here the crystals frequently projected away from the surface. Crystallization started uniformly several hours after application over the treated surface. Figure 50. Structure of methoxychlor deposits on elm bark (a) solution: radial crystal (b) emulsion: radial and plate-like crystals (c) emulsion: large plate-like crystals (d) emulsion plus Dacagin: uneven distribution at lower concentrations (whitish Spots) 165 However, a mixture of crystals and supersaturated dr0p1etS persisted for 1-2 days until complete crystallization occurred. Emulsion plus Dacagin The surface deposit patterns deve10ped by the 3 concentrations were distinctly different (FiguneSOd). The pattern formed by the 3.12% formulation showed several irregular whitish spots where deposits seemed to be more concentrated. These Spots became more frequent on the bark which was sprayed with the 6.25%.emulsion and they consisted of ray-like radial crystal structures. The thickener itself showed up in some places as thin transparent film. The area between these centers of intense cry- stallization was covered with a net of ray-like crystals. Their branches were long, sometimes needle-like, and adhered tightly to the bark surface. The Spottiness and obvious uneven distribution of deposits disappeared with the 12.55% emulsion. The crystal coat was thick and continuous. Ray-like crystals branching out from a center were produced by all 3 formulations. The solutions resulted in very delicate radial structures which adhered flat to the bark and did not project away from the surface. The straight emulsion deposits were characterized by many plate-like crystals. The Dacagin emulsions generated both types of crystal structures but plate-like crystals were not as frequent. In general, the solutions produced the thinnest surface deposits since the xylene penetrated well into the bark tissue. The emulsions did not penetrate as well which resulted in rich surface deposits. Although the xylene-based MeCl solution generated better tissue deposits than both types of emulsions, xylene as the sole carrier cannot be recommended because of its high volatility. Heavier oils would be more practical since they favor tissue deposits, but their phytotoxicity makes them 166 oftentimes not usable for tree applications (Whitten, 1949). (D) Decomposition of Methoxychlor Formulations The rate of decomposition on bark depends to a great extent on the quantitative relationship between surface deposits and tissue de- posits (Lyon, 1965). The penetration into the bark is determined by the carrier. Solutions usually penetrate better and create higher tissue deposits than emulsions. At higher concentrations, however, solutions and emulsions were found to penetrate equally well (Lyon, 1965). This ob- servation was confirmed in the first break-down study, 1970-71, where the solution and the emulsion deposits weathered equally fast by 61.1% and 66.5% over a one year period (Table 20). The emulsion containing the thickening agent, Dacagin, produced a very persistent deposit which was reduced by 34.3% after one year. During the first 4 weeks break- down was slow, 15% for the straight emulsion and the solution, and 5.5% for the Dacagin formulation. That thickening agents can have a retarding effect on decomposition of insecticidal deposits was mentioned by Bar- low and Hadaway (1952). Since higher concentrations of Dacagin caused nozzle clogging during helicopter applications, the 1971-72 experiment had the purpose of determining concentration levels for Dacagin at which a break-down retarding effect was still noticeable. The experi- ment was started on September 1, 1971, and was carried through the winter to obtain a measure of the residue degradation during the dormant season. Up to the tenth week only little break-down occurred for all formulations (around 10%). Decomposition was slower during the first 5 weeks than during the first 4 weeks of the previous year. Colder temperatures in the fall might have retarded degradation. The residues of all formulations dropped significantly over the dormant months. The 167 co_umu__aam coumm u_moamp .mc_m_eo co comma Am mo_aEmm N mo come AN xmm.~_ ob a80=__a acm smeam e_o _~-xz. c__;u_z x_.mN sate amum_=etoC we A~:.mo :.m: N... ..m ANm.ON: Amco__mm oo_\mn_ m.ov c_mmomo + co_m_:Em ANN.~Q :.m: m.m m.o Aum.-m Amco__mm oo_\ma_ NV c_mmumc + co_m_:Em Awo.~n m._o ~.o_ m.: Amo.mm: :o_m_:Em Num_ 1 .mm_ m.:m m.m m.~ a.mkm Amco__ma oo_\mn_ av c_mmoma + co_m_:Em m.mm o.m_ 0.0 o.mom A_u_ ucoumzuoco_>xv co_m_38m _._o a.m. 8.0 o._mm Amco_>xv co_b:_om _Nm_ 1 onm— Nm mm o. m d N NEo\mEmcmoco_E A_co_um_:ELOm Amco_umo__aa< coum< mxooz z co_u_moaeoooo x. u_moaoo .mmco AN .po_coa Lmo> oco m Lo>o xcmn 5.0 :0 mco_um_:Ecow Lo_:o>xo;uoE mo co_u_moaeoooo .ON o_nmh c o c .. .......U..._..... .1..— _... n.1_-..- _Hl...::...1~ ...—-_. .....—~IL... . ., ...... _. ...t. ,_....._ ...—... . -11“.....M.. .... . r. ...... :. L 77—) -..._ _> .._..LZ S DCMBENCLECUCO .....A flrummODOQ .mmJC A— A1 . . » o . x3 :_ a t m ,Eo/msm4c040.3 ::__:_:E;:_ u..,.. ..,. . :37“. 1...: .H. .3. 168 straight emulsion was broken down by 61.8% by the following April. The formulation with the higher Dacagin concentration had decomposed by 45.4%. After more than one year, 28% was left over from the original emulsion deposit, 32% and 35% from the high and low Dacagin deposits. Break-down of the Dacagin formulation in the 1970-71 experiment was much slower than in the second experiment. The higher concentration of Dacagin in the formulation of the 1970 experiment was certainly one reason. The quan- titative deposit changes of both Dacagin formulations (2 lbs/100 gal and l lb/100 gal) in the 1971/72 experiment were comparable. These experiments indicated that methoxychlor deposits have a life-Span which definitely exceeds one year. The deposits of formulations without Dacagin were reduced to roughly 1/3 of the original amount within one year. Decomposition was twice as strong for the first 32 weeks (September to April) than it was during the next 20 weeks. This suggested that the weathering of surface deposits was completed after 32 weeks. The break-down of the remaining tissue deposits proceeded, then, at a much slower pace. After 10 weeks, a fine net of methoxychlor crystals was still observed on the bark of all treatments. No surface deposits could be detected in April after 32 weeks of weathering. Dacagin, at the lower concentrations (1 lb/100 gal and 2 lbs/100 gal), must be credited with the retarding action on the decomposition of methoxychlor deposits during the dormant season. This protective action, however, disappeared during the rest of the year. It appeared that this thicken- ing agent can increase persistence at higher concentrations Significantly, as was shown in the 1970-71 experiment. The results of these break-down experiments might give an answer to a key question: can methoxychlor be applied in the fall? For the 169 mist blower, the answer is 'yes' for trees with limited height (up to 15 m) provided the application is done thoroughly. Mist blower deposits frequently reach several hundred micrograms/cm2 in the lower tree region 2 and up to 200 micrograms/cm in the upper height level. Given a decom- 2 would have position rate of 70% over one year, the 200 micrograms/cm decomposed to 60 micrograms/cm2 which is still sufficient for good pro- tection. The helic0pter achieves deposits of 70 micrograms/cm2 in the t0p of elms. Using the same decomposition rate, 25 micrograms/cm2 would remain after one year. This deposit could cause up to 40% mortality according to the residue-mortality regression. In reality, however, this figure would be lower because of the poor Spray distribution on the target sites, the branches. All in all, a helicopter treatment in the fall appears to be ineffective. (E) Movement of Methoxychlor on Elm Bark Table 21 summarizes the residue amounts recovered from the var- ious bark sections. Within the 4-week period, very little displacement of residue had taken place. Section A, the bark area above the appli- cation point, had zero residue. Almost the entire recovery of residue was made in section B where the insecticide was originally applied. A small amount of residue was translocated downwards to section C, which included the crotch area, and section 0. The recoveries for the twig crotches I and II were almost identical. These results clearly indicated that movement of methoxychlor deposits from the original point of application to another location is minimal. The insignificant diSplacement which occurred ducing the 4-week weathering period was probably due to the washing effect of rain water running down the bark. Since the water solubility of methoxychlor is 170 Figure 51. Translocation experiment: designation of bark sections on twig crotch. Table 21. Residue recovery from bark sections A, B, C, and 0.1) Twig Bark Application 4 Weeks After Recoveryz) Crotch Section micrograms MeCl Application % micrograms MeCl I A --- 0.0 0.0 B 599.7 529.4 88.3 c --- 5.9 1-0 0 --- 4.2 0.2 TOTAL 90.0 11 A --- 0.0 0.0 B 599.7 514.7 85.8 c --- 6.2 1.0 D --- 4.6 0.8 TOTAL 37.; 1) for location of bark sections, see Figure 51 above 2) based on total amount applied originally 171 extremely low, it is conceivable that translocation was purely mechanical. Therefore, the degree of coverage is definitely determined by the capa- bility of the application machine and the quality of the application it- self, since bark residues of methoxychlor were not redistributed after initial deposition. Summary_and Conclusions The correlation of beetle response (mortality) with the reSpec- tive amounts of residue on field-treated twig crotches yielded LC50 val- ues for the helicOpter and the mist blower application. The LC50 on helicopter-treated twigs was 41.8 micrograms/cm2 while the LC50 on mist blower-treated twigs was only 6.5 micrograms/cmz. However, DDT residues on mist blower-sprayed twigs interfered with the LCS0 determination. Therefore, the LC underestimated the true deposit-mortality relation- 50 ship. Feeding tests and residue analysis of lab-treated twig crotches 2 with continuous bark deposits yielded an LC of 12.5 micrograms/cm 50 which was also much lower than the L050 for the heliCOpter-sprayed twigs. The difference was explained by the fact that discrete spray patterns are less toxic than spray patterns with great dispersion. These findings pointed out a further advantage of the mist blower application, in addi- tion to better tree coverage. That is, less methoxychlor per unit bark area is required for protection than for the helic0pter. Contact toxicity of several methoxychlor formulations applied to fiber boards was substantial when the deposits were fresh. The emulsion deposit caused greater mortality than a solution deposit. The addition of Dacagin to the emulsion reduced contact toxicity of fresh deposits considerably. Weathered surface deposits had only little contact toxi- city. The contact with aged deposits prior to feeding had no impact on 172 the feeding response. The physical structure of methoxychlor surface deposits was documented and discussed in relation to contact toxicity. These experiments pointed out that methoxychlor deposits are only effective when they are taken in during the feeding process. Since surface deposits are avoided by the vector, the whole toxic action must come from tissue deposits. In order to obtain long-lasting protection, formulations must be employed and deve10ped which favor tissue deposits. Methoxychlor deposits proved to be quite persistent. Deposits from 12.5% emulsions on elm bark broke down to about 30% after one year. When Dacagin was added, even at very low rates, decomposition was defi- nitely delayed. The persistence of methoxychlor deposits suggested that fall applications could be effective with the mist blower provided ex- cellent coverage of the whole tree region is obtained. It was often suspected that methoxychlor residues are partly translocated from the point of application and are accumulated in twig crotches. However, no such displacement was observed in 2 experiments. V. Dutch Elm Disease Control on the Campus of Michigan State University 1958 - 1971. In 1958 the first elm tree succumbed to DED on the campus of Michigan State University. Since then, the elm pOpulation on the campus has been subjected every year to a rigorous and well-planned control pro- gram. This was largely due to the excellent c00peration between the Department of Entomology and the Grounds and Maintenance Department of Michigan State University. Between 1958 and 1963, DDT emulsions were applied by mist blower throughout the campus, but some trees were sprayed by hydraulic equipment. In 1964, the campus was subdivided into 2 treat- ment blocks (Figure 7). One block with about 900 trees was treated with methoxychlor; in the other, DDT was used on approximately 1500 elms. Two woodlots within the campus area were never sprayed and served as check plots. For 3 consecutive years, these blocks were treated in the same way. At the end of 1966, the methoxychlor plot compared favorably to the DDT treated plot in terms of elm losses over the 3 years. However, as Butcher _£_§l, (1967) pointed out, it was difficult to evaluate the effect of DDT carry-over in the methoxychlor plot. Table 22 compares the losses in the 2 treatment plots and the check plots. In 1967, the majority of the campus elms was sprayed with methoxychlor since it appeared that this insecticide was a satisfactory substitute for DDT. With methoxy- chlor, all spraying had to be completed during 2 weeks before bud break whereas DDT applications were made partly in the fall as well as in the spring. Treating the more than 1500 campus elms by mist blower within a 173 174 short period of 2 weeks posed a particular problem to Grounds Department personnel and very often the timely completion of elm spraying could only be accomplished by over-time work. Increasing cost factors (methoxychlor is about twice as expensive as DDT), problems with manpower, and growing concern about pesticide contamination on the campus stimulated the search for more effective and more economical ways to treat elms. Table 22. Tree losses in DDT and methoxychlor blocks, 1964/65. (from Butcher g£_§l., 1966) Year Insecticide Number of Trees Dead Trees No. % DDT 1274 38 2.90 Check 1) 647 238 36.78 1964 Methoxychlor 1163 41 3.53 Check ) 604 398 65.89 DDT 1163 41 3.53 Check ‘1 -- -- -- 1965 Methoxychlor 793 8 1.08 Check ) 480 237 49.37 I) Kalamazoo woodlot - never treated for bark beetle control. 2) Sandord woodlot - never treated for bark beetle control. Woodlot disturbed by sewage plant construction in 1965. Thus, in the spring of 1968, based on previous experience with experimental spraying of city elms in Birmingham, Michigan, all campus elms were treated by helic0pter (Wallner and Leeling, l968; Wallner g£_ 1., 1969). Helic0pter spraying, treatment areas and methoxychlor formulations employed for 1968 and 1969 are described in detail by Wallner t 21° (1969) and on page 64 of this dissertation. In the spring of 1970 and 1971, a 12.5% methoxychlor emulsion with 2 pounds of Dacagin per 100 gallons 175 spray was applied by helicopter to most of the campus elms. Around 200 elms in 2 University housing areas, University Village and Cherry Lane, and several elms close to dormitories have been sprayed consistently by mist blower (Figure 6). Particularly valuable elm trees were sprayed twice, once by helic0pter and again by mist blower. In 1969, Vapam I) was injected into the soil in several places where root graft transmission was suspected. According to the elm tree count at the end of 1967, there were about 1900 elms left from the original elm pOpulation of 2300 trees (estimate). This amounted to a total of roughly 400 trees or 16.6% lost during the first 10 years of DED on the campus from 1958 to 1967. An additional 18.0% was lost between 1968 and 1972. The elm population on the campus is now down to about 1500 elms which represents a loss of 34.6% since 1958. The history of DED control on the MSU campus for the years 1958 to 1971 inclusive is summarized in Table 23 which lists also the yearly tree losses owing to Dutch elm disease. Figure 52 shows a dramatic increase of losses to DED on the MSU campus since 1967. One might conclude that this increase was due to the switch from DDT to the less effective methoxychlor or due to the less thorough aerial application technique. This conclusion is probably correct to some degree and might explain part of the upward trend in elm losses in recent years. However, the most important variable which is definitely correlated with the number of tree losses is the vector density. Unfortunately, no pop- ulation estimates are available for the campus or the greater Lansing area for the period following 1958. The production of Scolytus vectors on the campus itself is probably very low due to the stringent and well- exercised sanitation and tree removal program. An indication of a possible 'p0pu1ation explosion' in the Lansing area and the subsequent increase of 1) product of Stauffer Chemicals 176 'hble 23. Summary of DED control and elm losses on the campus of MSU: 1958-1971 Yau' No. trees No. treesl) % lossz) %loss3) No. trees % trees % trees sprayed yearly yearly Sprayed with Sprayed by total loss DDT MeCl mistbl. helic. 1958 2285 1 0.0 0.0 1157 - 50.6 100 - 1959 2284 8 0.4- 0.4 1641 - ‘71.9 100 - E60 2276 17 0.7 0.8 1960 - 86.1 100 - 1961 2259 45 2.0 2.0 2007 - 88.8 100 - 1962 2214 34 1.5 1.5 2214 - 100.0 100 - W63 2180 33 1.4 1.5 2180 - 100.0 100 - 1961 2147 44 1.9 2.1 1515 898 95.3 100 - 1965 2103 49 2.1 2.3 1274 793 89.8 100 - 1966 2054 69 3.0 3.4 1163 616 89.8 100 - 1967 1985 79 3.5 4.0 208 1320 89.8 100 - 1968 1906 112 4. 5.9 - -u ,4 56) 95 8 “b 1969 1794 126 5 .5 7.0 - =’ g 5 95 u E 1970 1650 90 3 .9 5.5 - 2 g 5 95 > 0) 1971 1560 84 3.7 5.4 - 'g 'g 5 95 1972 1476 D including root graft infestations U % based on tree total of 1958 8 % based on yearly tree total N 18 trees, removed due to construction, not included 9 probably between 95 to 100% Q estimate D % total loss from 1958 to 1971 177 ._m1mmm_ n>ummco>_c: eunum cmm_co_2 mo maaEmo ecu co mmmom_v 5.0 gouge ou mommoa .Nm 0L:m_u .Il 868:0... 1. o .8306 ES. 0 .082 10 o 1.50 L Alllll 33> E 3 3 E S 2 a... x _ _ _ _ _ .1LIH 1N .1! 1 .IIIIL. 11v L1 t 1a meJ R 178 tree infections in surrounding areas such as the campus is the fact that 1967 was the last year where all Lansing city elms received DDT spray protection. In the following years, methoxychlor was used exclusively by the city of Lansing, however, only some of the more important park and street trees were sprayed (Table 24 ). City crews were not able to keep up with timely removal of diseased elms and sanitation of elms was not practiced sufficiently. All these factors led to an increase in the vector pOpulation in the Lansing area and the rising losses of elms on the MSU campus since 1967 are probably a partial reflection of that. Table 24. DED Spraying in Lansing, Michigan. ]) Year Number of elms Sprayed Estimated total No. elms 1968 1200 5000 1969 1400 ? 1970 700 ? 1971 spraying discontinued 2500 1) according to figures supplied by the city forester of Lansing, Michigan The original elm population in the city of Lansing of 8000 trees suffered a loss of 69% and as of 1971 is down to an estimated 2500 trees. The city of East Lansing to the north of the campus lost a similar per- centage of its original number of elms. The loss of elms on private prOperty is not included in these figures. The 34.6% loss to DED since 1958 on the campus compares favorably to the above figures of around 70% in the city of Lansing and East Lansing. Yearly elm losses in some cities in the northeast were as high as 15% (Miller t 1., 1969). Most of the bigger elms in the woodlots surrounding the campus have disappeared 179 because of Dutch elm disease, so that the urban elms in the vicinity of the campus can be considered as the only source of beetle production. The chances that the majority of the well-tended campus elms can escape the disease is enhanced by the fact that the elm density in the surrounding urban communities will eventually reach a level where the disease becomes endemic. With a basically vector-free belt surrounding the campus, a minimal sanitation and spray program will be necessary for adequate pro- tection of the campus elms. By maintaining or even improving the present control measures for Dutch elm disease, a fair chance exists to preserve the elms on the campus of Michigan State University. SUMMARY AND CONCLUSIONS Data on the distribution of feeding zones in elm trees were col- lected because of their obvious importance for elm protection. Extensive sampling of untreated elm trees showed a linear increase in feeding attack by S, multistriatus from the bottom to the tOp. The upper tree region was designated, therefore, as the preferred feeding zone of S, multistriatus. However, inoculations in the lower half of an elm tree reportedly result In faster disease development and tree kill because of the shorter distance between feeding sites (inoculation points) and main trunk. Consequently, twig crotches in the whole tree region require protection. The actual target sites during a spraying operation are the twig crotches which make up only a fraction of the total bark area in an elm tree. Only a small portion of the spray settles in the twig crotches, the rest 'contaminates' other bark areas or drifts to non-target sites in the vicinity of the tree. No spraying method will ever be able to direct the protective spray exclusively to the twig crotch areas. In practice, the whole tree, all bark area, has to be covered to assure protection. What is adequate protection? Theoretically, vector feeding in one unprotected feeding site could result in effective inoculation and disease development. The logical objective of pre-attack Spraying is therefore to prevent inoculation via vector feeding in all potential 180 181 feeding sites during the whole period of susceptibility. The problem is two-fold. First, to achieve the goal of complete protection, suffi- cient amounts of insecticidal spray have to be deposited in every twig crotch. This is basically a Spray-technological problem. Second, the duration of protection is determined by the persistence of residue on and in the bark, which is dependent on the properties of the chosen insecti- cide and formulation. The performance of two spray systems for elm protection, the helic0pter and the mist blower, was critically evaluated in this disser- tation by bioassays, by chemical analysis of insecticidal residues on the bark, and by a monitoring technique for spray coverage. Neither of the two tested application techniques achieved complete coverage or in- sured 100% protection. However, the mist blower was, in general, 3 to 4 times more effective than the helicopter. This was conclusively shown by all three evaluation methods. Reduced tree-top coverage at greater tree heights seemed to be one important short-coming of the mist blower application. The monitoring technique, employing sample cubes which were suspended at several height levels in elm trees, provided an ob- jective evaluation of spray coverage. This technique was unique in the sense that it permitted the study of droplet size spectra, actual de- posits and 3-dimensional coverage. Although the degree of protection provided by a certain application technique should be the main criterion for its selection for elm spraying, other factors, such as environmental and economic considerations, must be taken into account in the decision- making process. The following criteria appear to be crucial to the evaluation of application techniques for elm Spraying: 1) degree of coverage: Coverage depends on the capabilities of the application machine to reach all target sites. The actual amount 182 of insecticide applied per unit bark area and drOplet size Spectrum are critical for the degree of protection. 2) insecticidal input to the environment and drift: Dutch elm disease spraying is practiced primarily in urban areas where the addition of insecticide to the environment should be kept at a minimum. Spray drift should be low because of the obvious health hazards which could arise in the city environment. 3) economics of spraying Operation: The actual spraying costs per tree are certainly a prime consideration. They are composed of the direct cost of the spraying machine, cost of manpower and insecticidal formulation. However, indirect costs could evolve from environmental damage and spray damage to prOperty and human health. 4) speed and accuracy of application: Timing for Dutch elm disease spraying is critical. It is desirable to complete the spraying operation in the weeks before bud break. ln urban areas spraying is usually carried out during early morning and evening hours because of reduced interference with city life. Low temperatures and unfavorable wind conditions can greatly limit the available spraying time before bud break. Speed of the Spraying Operation is then of great importance. Further, with persistent insecticides such as DDT the spraying Operation could be stretched out over a longer time period (fall and spring appli- cations) which is not possible with fast-weathering insecticides. This, too, increases the need for faster application techniques. In terms of accuracy, elm identification is better executed from the ground than from the air, which could mean that, in an aerial spraying Operation, not all elm trees receive spray coverage. Utilizing the criteria established above, the helic0pter and the 183 mist blower application techniques can be compared for a final evaluation. The mist blower was, overall, the more effective application machine. It delivered 22: 4-6 times more spray to each tree (2-3 gallons), reached 22, twice as many feeding Sites and gave between 80 and 90% protection during the first 10 weeks after spraying, as compared to 15 to 30% for the helic0pter. Also, helicopter-applied bark deposits were found to be comparatively less toxic than mist blower deposits, owing to the higher dispersion of mist blower Spray. The helicOpter‘s spray input to the environment is considerably lower (2 quarts per tree) which should be considered when environmentally hazardous insecticides (such as DDT) are used for Dutch elm disease spraying. The question as to which spraying method causes less drift is still Open, however drift control was achieved in helic0pter-applied spray by the addition of an anti-drift material. The economics of both spray methods was not investigated, however, helic0pter application is 2 to 3 times cheaper. The helic0pter saves insecticide, requires less manpower and finishes a spraying job in a shorter time. For instance, on the campus of Michigan State University helicopter spraying of 22: 1200 elms requires a total of £2: 6 flight hours (one pass over each tree) while the mist blower application needs several weeks for completion. The mist blower application can be adjusted better to characteristics of the individual tree and involves no tree identification problem. A pre-requisite for a good spraying job from the air is a thorough know- ledge of the terrain by the pilot applicator. It is certain, however, that always an unknown number of elm trees escape spray coverage from the air. It was suggested recently (Barger, 1971) that 'saturation' 184 spraying (treating all elms in a given area) might not be necessary. This might be a reasonable assumption provided the vector pressure is low. In summary, one can state that the helic0pter application, as it is practiced today, cannot be recommended as a substitute for the mist blower application because it fails to give adequate protection. Its positive features of speed of application and reduced cost make it de- sirable to use if it can be improved as an application machine. Such improvements can be made either in the technical design of the actual spray equipment, adapting it especially for tree spraying, or altering the mode of the spraying Operation. Reducing speed during application above each tree will cause a more favorable down-draft and better cover- age, and will result in higher spray dosages for each tree. Another possibility is to make two passes over each tree ('double application') which will result in twice as much insecticide per tree and possibly better coverage since the spray settles from two different directions. These suggested modifications will, however, increase the expense of helicopter spraying; one would have to weigh the economics of a 'double application'. Methoxychlor is now the exclusive chemical in use for Dutch elm disease vector control. It is surprisingly effective and has long per- sistence, exceeding one year. From an environmental standpoint, it is a safe compound with high biodegradability and low mammalian toxicity. Because of these positive features, no change to another insecticide will be necessary. However, new formulations are needed which favor bark tissue deposits and create persistent deposits SO that fall appli- cations will be feasible. Formulations which are self-spreading after application could improve deposit distribution on the branches. Such 185 a formulation would complement and aid the heliCOpter application with its poor 3-dimensiona1 coverage and discrete spray pattern. Over the long run, the expensive yearly spraying Operations against DED cannot be carried on indefinitely. Many communities, discouraged with the results of many years of intensive DED control, are giving up all protective spraying Operations. No control method to date has been able to halt disease spread. Spraying against vector feeding certainly delayed disease deve10pment and gave city foresters a chance to replace the dying elms with other shade tree Species. This delaying action is probably the biggest merit of elm Spraying, but control of the disease cannot be achieved by it. LITERATURE CITED LITERATURE CITED Al-Azawi, A. F., and J. E. Casida. 1958. The efficiency of systemic insecticides in the control of the smaller EurOpean elm bark beetle. J. Econ. Entomol. 51: 789-90. Al-Azawi, A. F., and D. M. Norris. 1959. Experimental prevention of bark beetle transmission of Ceratocystis ulmi (Buism.) Moreau with the systemic insecticide Chipman R-6l99. J. Econ. Entomol. 52: 902-4. Al-Azawi, A. F., D. M. Norris, and J. E. Casida. 1961. Hazards asso- ciated with the implementation of Tetram into elm trees for Dutch elm disease control. J. Econ. Entomol. 54: 127-9. Anderson, R. F. 1948. Host selection by the pine engraver. J. Econ. Entomol. 41: 596-602. Baker, J. E., and D. M. Norris. 1967. A feeding stimulant for Scolytus multistriatus (Coleoptera:Scolytidae) isolated from bark of Ulmus americana. Ann. Entomol. Soc. Am. 60: 1213-5. Baker, J. E., D. P. Rainey, D. M. Norris, and F. M. Strong. 1968. p- hydroxybenzaldehyde and other phenolics as feeding stimulants for the smaller EurOpean elm bark beetle. Forest Sci. 14: 91-5. Banfield, W. M. 1968. Dutch elm disease recurrence and recovery in American elm. Mass. Agric. Exp. Stn. Bull. NO. 568. Barger, J. H. 1971. Field and laboratory evaluation of methoxychlor for Dutch elm disease vector control. Ph. D. Thesis. Ohio State Univ. Barger, J. H., and W. K. Hock. 1971. Distribution of Dutch elm disease and the smaller EurOpean elm bark beetle in the United States as of 1970. Plant 015. Rep. 55: 271-2. Barger, J. H., R. A. Cuthbert, and N. H. Roberto. 1971. Assessment of three pre-cold storage preparations on the feeding response of Scolytus multistriatus and the degradation of dicrotOphos. J. Econ. Entomol. 64: 621-4. Barlow, F., and A. B. Hadaway. 1952. Some factors affecting the avail- ability of contact insecticides. Bull. Entomol. Res. 43: 91- 100. Beattie, R. K. 1934. The new outbreak of the Dutch elm disease. J. Econ. Entomol. 27: 569-72. 186 187 Becker, W. B. 1947. Sprays to kill Scolytids breeding in elm logs. Pre- liminary spraying experiments to prevent twig feeding by the smaller EurOpean elm bark beetle. Mass. Exp. Stn. Bull. NO. 441: 40-2. Becker, W. B. 1956. Seasonal control of Dutch elm disease and insect pests of elms in Massachusetts. Trees Mag. 16: 8, 9, 18, 19. Becker, W. B., and R. A. Mankowsky. 1965. Twig feeding by the smaller European elm bark beetle on different kinds of trees. J. Econ. Entomol. 58: 132-4. Betrem, J. G. 1929. De eipenziekte en de iepenspintkevers. Tijdschr. PlZiekt. 35: 273-87. Bingham, R. T., R. J. Hoff and G. I. McDonald. 1971. Disease resistance in forest trees. Annu. Rev. of Phytopathology. 9: 433-52. van den Bosch, R., and V. M. Stern. 1962. The integration of chemical and biological control of arthrOpod pests. Annu. Rev. Entomol. 7: 367-85. Buisman, C. J. 1928. De oorzaak van de iepenziekte. Tijdschr. Nederl. Heidemaatsch. 40: 338-45. Buisman, C. J. 1931. Overzicht van de soorten van iepen in verband met het iepenziekte onderzoek. Tijdschr. PlZiekt. 37: 111-6. Buisman, C. J. l932a. Ceratostomella ulmi, de geschlachtelijke vorm van Graphium ulmi Schwarz. Tijdschr. PlZiekt. 38: 1-8. Buisman, C. J. l932b. Verslag van de phytOpathologische onderzoekingen over de iepenziekte, verricht in het laboratorium ”Willie Com- melin Scholten“ te Baarn, gedurende 1931. Tijdschr. PlZiekt. 38: 17-36. Buisman, C. J. 1933. Verslag van de onderzoekingen over de iepenziekte verricht in het phytopathologisch laboratorium ”Willie Commelin Scholten” te Baarn, gedurende 1932. Tijdschr. PlZiekt. 39: 77- 94, 101-13. Buisman, C. J. 1934. Verslag van de onderzoekingen over de iepenziekte verricht in het laboratorium “Willie Commelin Scholten” te Baarn, gedurende 1933. Tijdschr. PlZiekt.~ 40: 65-87. Butcher, J. W., S. Ilnitzky, and J. Shaddy. 1966. Investigations on control of Dutch elm disease vectors in Michigan. Special Re- port to President. Mich. State Univ. 74p. Chapman, J. W. 1910. The introduction of a EurOpean scolytid (smaller elm bark beetle) Scolytus multistriatus (Marsh.) into Massachu- setts. Psyche. 17: 63-8. Clinton, G. P., and F. A. McCormick. 1935. The Dutch elm disease, S527 Ehium ulmi, in Connecticut. Science (Wash. D.C.). 81: 68-70. 188 Collins, C. W., W. D. Buchanan, R. R. Whitten, and C. H. Hoffmann. 1936. Bark beetles and other possible insect vectors of the Dutch elm disease, Ceratostomella ulmi (Schwarz) Buism. J. Econ. Entomol. 29: 169-76. Collins, 0. J. 1938. Feeding habits of Scolytus multistriatus Marsham with reference to Dutch elm disease. J. Econ. Entomol. 31: 196- 200. Collins, C. W. 1941. Studies of elm insects associated with Dutch elm disease fungus. J. Econ. Entomol. 34: 369-72. Courshee, R. J. 1959. Investigations on spray drift. J. Agric. Eng. Research. 4(2): 144-52, 229-42. Cuthbert, R. A., A. C. Lincoln, B. H. Kennedy, and N. H. Roberto. 1970. Laboratory assay of Scolytus multistriatus and Dendrosoter pro- tuberans on DDT and methoxychlor. J. Econ. Entomol. 63: 1889- 9]. Dimond, A. E., G. H. Plumb, E. M. Stoddard, and J. G. Horsfall. 1949. An evaluation of chemotherapy and vector control by insecticides for combating Dutch elm disease. Conn. Agric. Exp. Stn. Bull. NO. 531. Dixon, E. B. 1964. Attack response of the smaller EurOpean elm bark beetle, Scolytus multistriatus, in confinement. J. Econ. Entomol. 57: 170-2. Doane, C. C. 1958. The residual toxicity of insecticides to Scolytus multistriatus. J. Econ. Entomol. 51: 256-7. Doane, C. C. 1962. Evaluation of insecticides for control of the smal- ler EurOpean elm bark beetle. J. Econ. Entomol. 55 : 414-5. Elbs, K. 1393. J. Prakt. Chem. 47: 44. Elgersma, D. M. 1969. Resistance mechanisms of elms to Ceratocystis ulmi. Meded. phytOpath. Lab. ”Willie Commelin Scholten”. No. 77. 84p. (Baarn, Netherlands). English, L. L., and W. L. Hartstirn. 1962. Systemic insecticide con- trol of some pests of trees and shrubs- a preliminary report. 111. Dept. Reg. and Educ. (Nat. Hist. Div.). Biol. Note. NO. 48. Epstein, A. H. 1969. Low temperature sprays for Dutch elm disease con- trol. Plant Dis. Rep. 53: 305-6. Escherich, K. 1923. “Die Forstinsekten MitteleurOpas” Vol. 2: 494p. Felt, E. P. 1934a. Dutch elm disease and its control. Bartlett Tree Res. Lab. Circular. Stamford, Conn. Felt, E. P. 1934b. Dutch elm disease control and the elm bark borer. J. Econ. Entomol. 27: 315-9. 189 Felt, E. P. 1937. Balloons as indicators of insect drift and of Dutch elm disease spread. Bartlett Tree Res. Lab. Bull. NO. 2: 3-10. Felt, E. P., and S. W. Bromley. 1938. Shade tree insects and sprays, 1937. J. Econ. Entomol. 31: 173-6. Finnegan, R. J. 1957. Elm bark beetles in southwestern Ontario. Can. Entomol. 89: 275-80. Finney, D. J. 1964. ”Probit Analysis.II 2nd ed. Cambridge Univ. Press, London. 318p. Fox, R. C. 1958. Seasonal deve10pment of the smaller EurOpean elm bark beetle in southern Michigan. Mich. Agric. Exp. Stn. Bull. NO. 41: 279-87. Fransen, J. J. 1931. Enkele gegevens omtrent de verspreiding van de door S, ulmi Schwarz veroorzaakte iepenziekte door de iepen- Spintkevers, S, (S.) scolytus F. en E, (5.) multistriatus Marsh. in verband met de bestrijding dezer ziekte. Tijdschr. PlZiekt. 37: 49-62. Georgescu, C. C., and S. Orenski. 1957. Morphologische, Kulturelle und physiologische Untersuchungen Uber den Pilz Graphium ulmi Schwarz. Verh. IV. Int. Pfl. Schutz Kongress. Hamburg. 1: 65-9. Goeden, R. D., and D. M. Norris. 1963. The potential of Magdalis ssp. in the transmission of Ceratogystis ulmi (Buis.) C. Moreau. J. Econ. Entomol. 56: 238-9. Goeden, R. D., and D. M. Norris. 1964. Attraction of Scolytus quadri- spinosus (Col.;Scol.) to Carya ssp. for oviposition. Ann. Ento- mol. Soc. Am. 57: 141-6. Guyot, M. 1921. Notes de pathologie vegetale. Bull. Soc. Path. Veg. France. 8: 132-6. Haftstad, G. E., and J. F. Reynolds. 1961. Effectiveness of dormant sprays in the control of Dutch elm disease. Plant 015. Rep. 45: 681. Hart, J. H., W. E. Wallner, M. R. Caris, and G. K. Dennis. 1967. In- crease in Dutch elm disease associated with summer trimming. Plant Dis. Rep. 44: 806-7. Hart, J. H. 1970. Attempts to control Dutch elm disease by pruning. Plant 015. Rep. 54: 985-6. Heybroek, H. M. 1961. De iep 'Commelin'. Ned. Bosb. Tijdschr. 33: 325-8. Heybroek, H. M. 1963. De iep 'Groeneveld'. Ned. Bosb. Tijdschr. 35: 370-4. 190 Heybroek, H. M. 1966. Dutch elm disease abroad. Am. Forests. 72(6): 26-9. Heybroek, H. M. 1969. Host plant resistance in the elms. Proc. N. Centr. Br. Entomol. Soc. Am. 24: 69-74. Himelick, E. B., and D. Neely. 1961. Prevention of bark beetle devel- Opment in undesirable elms for the control of Dutch elm disease. Plant Dis. Rep. 45: 180-4. Himelick, E. B., and D. Neely. 1963. Systemic control of bark beetle colonization of diseased American elms. Plant Dis. Rep. 47: 87-8. Hoestermann, G., and M. Noak. 1925. Uber das Ulmensterben am unteren Rhein. Mitt. dt. dendrol. Ges. 35: 287-9. Holmes, F. W., H. M. Heybroek, and J. N. Gibbs. I972. Aggressiveness in Ceratocystis ulmi. PhytOpathology 62: 939-40. Hoskins, W. M. 1962. Some important prOperties of pesticide deposits on various surfaces. Residue Rev. 2: 66-91. Academic Press, New York. Hunt, J. 1956. Taxonomy of the genus Ceratocystis. Lloydia 19: 1-58. Jacot, A. P. 1934. Acarina as possible vectors of the Dutch elm disease. J. Econ. Entomol. 27: 858-9. Kaiser, P. 1931. Das Ulmensterben. Gartenflora. 80: 369. Kapoor, 1. P., R. L. Metcalf, R. F. Nystrom, and G. K. Dangha. 1970. Comparative metabolism of methoxychlor, methiochlor, and DOT in mouse, insects, and in a model ecosystem. J. Agric. Food Chem. 18: 1145-52. Kaston, B. J., and D. S. Riggs. 1937. Studies on the larvae of the native elm bark beetle. J. Econ. Entomol. 30: 98-108. Kaston, B. J., and D. S. Riggs. 1938. On certain habits of elm bark beetles. J. Econ. Entomol. 31: 467-9. Kaston, B. J. 1939. The native elm bark beetle (HylurgOpinus rufipes Eichhoff) in Connecticut. Conn. Agric. Exp. Stn. Bull. No. 420. Kirch, J. H. 1971. The Microfoil boom system. Proc. 4th Intern. Con- gress Agric. Aviation. 1969: 152-5. KnOp, Th. 1928. Das Absterben der Ulmen und seine Ursache. Entomol. Zeitschr. 42: 68-9. Lamdin, J. M., R. D. Eikenbary, and E. E. Sturgeon. 1969. Biology and feeding habits of the smaller EurOpean elm bark beetle on American elm trees treated with Bidrin. J. Econ. Entomol. 62: 640-3. 191 Landis, Wm R. 1969. Changes in Ulmus americana in response to infection by Ceratocystis ulmi. Ph. D. Thesis. Mich. State Univ. LaUger, P., H. Martin, and P. Muller. 1944. Helv. Chim. Acta. 27: 892. Lincoln, A. Ch. l968. Bioassay of insecticidally treated elm twig crotches with smaller EurOpean elm bark beetles (unpublished manuscript). U.S.D.A. Research Laboratory, Northeastern Forest Experiment Station. Delaware, Ohio. Lyon, R. L. 1965. Structure and toxicity of insecticide deposits for control of bark beetles. U.S.D.A. Tech. Bull. No. 1343. 59p. Maksymiuk, Bohdan. 1971a. Kinetics and physics of pesticidal aerial sprays. Pesticides, pest control and safety on forest range lands. Proc. short course (Corvallis, Oregon): 171-9. Maksymiuk, Bohdan. 1971b. How to minimize drift of pesticidal Sprays. Pesticides, pest control and safety on forest range lands. Proc. short course (Corvallis, Oregon): I80-7. Malaquin, A. 1923. La maladie des ormes. Renaissance Agric. 4: 91-4. Martin, C. H. 1935. Preliminary report of trap-log studies on elm bark beetles. J. Econ. Entomol. 29: 297-306. Martin, C. H. 1938. Field notes on the life history of Hylurggpinus rufipes (Eich.). J. Econ. Entomol. 31: 470-7. Martin, C. H. 1938. Effect of sunlight and location of logs on beetle infestation of elm logs. Bull. Brooklyn Entomol. Soc. 33: I95- 203 a Martin, C. H. 1946. Effect of phloem condition and phloem moisture on the entry of Scolytus multistriatus. J. Econ. Entomol. 39: 481-6. Matthysse, J. G., H. C. Miller, and H. E. Thompson. 1954. Insecticide deposits for control of elm bark beetles. J. Econ. Entomol. 47: 736-46. May, C., and G. F. Gravatt. 1931. The Dutch elm disease. U.S.D.A. Cir- cular. 170: 1-10. Melin, E., and J. A. Nannfeldt. 1934. Researches into the blueing of ground wood-pulp. Svenk. Skogsvardsfoer. Tidskr. 32: 397-616. Metcalf, R. L. 1955. ”Organic Insecticides". Wiley (Interscience), New York. Metcalf, R. L., I. P. Kapoor, and A. S. Hirwe. 1972. DeveIOpment of biodegradable analogues of DDT. Chemtech. : 105-9. Meyer, H. J., and D. M. Norris. 1964. Dispersal of the pathogen and vectors of Dutch elm disease. Proc. N. Cent. Br. Entomol. Soc. Am. 19: 26-9. 192 Meyer, H. J., and D. M. Norris. 1967. Behavioral responses by Scolytus multistriatus (ColeOptera:Scolytidae) to host- (Ulmus) and beetle- associated chemotactic stimuli. Ann. Entomol. Soc. Am. 60: 642-7. Miller, H. C. 1951. Comparative toxicity of some synthetic organic in- secticides to elm bark beetles. Ph. D. Thesis. Cornell Univ. Miller, H. C., S. B. Silverborg, and R. J. Campana. 1969. Dutch elm disease: relation of spread and intensification to control by sanitation in Syracuse, New York. Plant Dis. Rep. 53: 551-5. Moreau, C. 1952. Coexistence des formes ThielaviOQsis et Graphium ulmi chez une souche de Ceratogystis major (Van Beyma) nov. comb. Remarques sur les variations des CeratocyStis. Revue Mycol. 17. Suppl. col. 1: 17-25. Nielsen, G. R. 1959. New additive to old insecticide improves control of Dutch elm disease. Farm Res. 25(3): 15. Norris, D. M. 1959. Potentialities for systemic control of the insect vectors of Dutch elm disease. Proc. N. Centr. Br. Entomol. Soc. Am. 14: 59-60. Norris, D. M. 1960. Systemic insecticidal action in the cortical tissues of elm twigs. J. Econ. Entomol. 53: 1034-6. Norris, D. M. 1961. Insecticides for Dutch elm disease. American City, A-19591. Ouellette, G. B. 1962. Studies on the infection process of Ceratocystis ulmi (Buism.) C. Moreau in American elms. Can. J. Botany 740: 1567-75. Pape, H. 1924. Das Ulmensterben in Deutschland. Mitt. dt. dendrol. Ges. 34: 284-8. Parker, K. G., P. A. Readio, and L. J. Tyler. 1941. Transmission of the Dutch elm disease pathogen by Scolytus multistriatus and the deve10pment of infection. PhytOpathology 31: 657-63. Parks, T. H. 1936. Insects on elms. Ohio State Univ. Agric. Ext. Ser- vice Bull. NO. 172. Peacock, J. W. (editor). 1967. Dutch elm disease conference. Pro- ceedings, U. S. Dept. Agric. Res. Lab., Delaware, Ohio. Plumb, G. H. 1950. Control of Scolytus and Dutch elm disease by con- centrated DDT sprays. J. Econ. Entomol. 43: 110-1. Pomerleau, R. 1965. The period of susceptibility of Ulmus americana to Ceratocystis ulmi under conditions prevailing in Canada (Quebec). Can. J. Botany 43: 787-92. Pomerleau, R. 1968. Progression and localization of infection by Cera- tocystis ulmi in white elm (Ulmus americana L.). PhytOpatholo- gische Zeitschrift. 63: 302-27. 193 Prell, H. 1930. Ulmensterben und Ulmenborken-Kaefer. Die kranke Pflanze. 9: 89'93. 103'5. 124-7. Readio, P. A. 1935. The entomological phases of the Dutch elm disease. J. Econ. Entomol. 28: 341-53. Roepke, W. 1930. Verdere gegevens omtrent de iepenziekte en de iepen- spintkever. Tijdschr. PlZiekt. 36(9): 232-7. Scanlon, E. H. 1949. Elm bark beetle conference. Trees Mag. 8(2): 13, 16, 18, 23. Schmitz, W. R., and M. B. Goette. 1948. Penetration of DDT into wood surfaces. Soap Sanit. Chemicals 24(1): 118-21. Schwarz, M. B. 1922. Das Zweigsterben der Ulmen, Trauerweiden und Pfirsichbaeume. Thesis. Utrecht Univ. Shell Chemical Company. 1964. Implantation of Bidrin insecticide into elm trees. Instruction Manual. 15p. Siegel, S. 1956. Nonparametric Statistics: for the Behavioral Sciences. McGraw-Hill, New York, Toronto, London. 312p. Snedecor, G. W., and W. G. Cochran. 1969. Statistical Methods. 6th ed. Iowa State Univ. Press, Ames, Iowa. 593p. Sokal, R. S., and F. J. Rohlf. 1969. ”Biometry.” W. H. Freeman CO., San Francisco. 776p. Spierenburg, D. 1921. Een onbekende ziekte in de iepen. Tijdschr. PlZiekt. 27: 53-60. Subcommittee on insect pests. 1969. Principles of plant and animal pest control. Volume 3: Insect-pest management and control. National Academy of Sciences (Wash., 0. 0.). Publication 1695. 508p. Sullivan, W. N., and I. Hornstein. 1953. Chlorinated polyphenyls to improve Lindane residues. J. Econ. Entomol. 46: 158-9. Sullivan, W. N., I. Hornstein, A. H. Yeomans, and Tsao Ching-Hsi. 1955. Improved deposits for controlling insects outdoors. J. Econ. Entomol. 48: 153-5- Thompson, H. E. 1965. Residual effectiveness of DDT emulsion sprays with and without horticultural oil against the smaller EurOpean elm bark beetle. J. Econ. Entomol. 58: 165-6. Touhey, J. G., and D. F. Bray. 1961. An evaluation of certain com- pounds as feeding deterrents against the smaller EurOpean elm bark beetle, Scolytus multistriatus. J. Econ. Entomol. 54: 293-6. 194 Truchan, J. G. 1970. Field evaluation of Dendrosoter protuberans as a biological control agent for Scolytus multistriatus, the primary vector of Dutch elm disease. Ph. D. Thesis. Mich. State Univ. Valek, D. A. 1967. A study of the host-parasite relationships of Scolytus multistriatus Marsham and methods of propagation for the introduced parasite Dendrosoter protgberans (Nees) Wesm. M. S. Thesis. Mich. State Univ. Verrall, A. F., and T. W. Graham. 1935. The transmission of Ceratocystis ulmi through root grafts. PhytOpathology 25: 1039-40. Wallace, Ph. P. 1940. Notes on the smaller European elm bark beetle. Conn. Agric. Exp. Stn. Bull. No. 434: 293-311. Wallner, W. E., and J. H. Hart. 1968. Dutch elm disease control. Mich. State Univ. Extension Bull. No. 506. Wallner, W. E., and N. C. Leeling. l968. HeliCOpter application of methoxychlor for elm bark beetle control. Mich. Agric. Exp. Stn. Bull. No. 50: 416-22. Wallner, W. E., N. C. Leeling, and M. J. Zabik. 1969. Fate of methoxy- chlor applied by helic0pter for smaller EurOpean elm bark beetle control. J. Econ. Entomol. 62: 1039-42. Ware, G. W., E. J. Apple, W. P. Cahill, P. D. Gerhardt, and K. R. Frost. 1969. Pesticide drift II. Mist-blower vs. aerial application of sprays. J. Econ. Entomol. 62: 844-66. Westerdijk, J. 1928. 15 de iepenziekte een infectieziekte? Tijdschr. Ned. Heidemaatsch. 40: 333-7. Whitten, R. R., and W. C. Baker. 1939. Tests with various elm-wood traps for bark beetles. J. Econ. Entomol. 32: 630-4. Whitten, R. R. 1949. Control of insect carriers of the Dutch elm disease and elm phloem necrosis. Trees Mag. 9(3): 6, 25. Whitten, R. R. 1958. The Dutch elm disease and its control. U.S.D.A. Agric. Info. Bull. No. 193. Whitten, R. R. 1960. Elm bark beetles. U.S.D.A. Leaflet No. 185. Williams, L. H., and H. E. Brown. 1969. Some biological investigations of the smaller EurOpean elm bark beetle in Missouri with refer- ence to systemic insect control. J. Econ. Entomol. 62: 1381-6. Wolfenbarger, D. 0., and W. D. Buchanan. 1939. Notes on elm twig crotch injuries produced by Scolytus multistriatus Marsham. J. Econ. Entomol. 32: 377-81. Wolfenbarger, D. 0., and T. H. Jones. 1943. Intensity of attacks by Scolytus multistriatus at distances from diSpersion and conver- gence points. J. Econ. Entomol. 36: 399-402. 195 Wolfenbarger, D. O. 1945. Estimations of volumes, surface areas, and numbers of twig crotches in elm trees. Amer. Mid. Nat. 34: 421- 32. Wollenweber, H. W. 1927. Das Ulmensterben und sein Erreger, Graphium ulmi Schwarz. Nachrichtenbl. Deutsch. Pflanzenschutzd. 7(10): 97-100. Wollenweber, H. W., and C. Stapp. 1928. Untersuchungen Uber die als Ulmensterben bekannte Baumkrankheit. Arb. Biol. Reichsanst. Land-und Forst-wirtsch. 16: 283-324. Wollenweber, H. W. 1929. Das Ulmensterben und sein Erreger. Z. Pilzk. N. F. 8: 162-6. Wood, 0. L. 1962. The attraction created by males of a bark beetle 12$ confusus (LeConte) attacking ponderosa pine. Pan-Pacific Ento- mologist 38: 141-2. Wooten, J. F. 1962. Methoxychlor: safe and effective substitute for DDT in controlling Dutch elm disease. U.S.D.A. Central States Forest Exp. Stn. Note No. 156. Worthley, L. H. 1935. The Dutch elm disease eradication project: fed- eral, state and local cooperation. U.S.D.A. Circular No. 353. Worthley, L. H. 1936. Progress in Dutch elm disease eradication. J. Econ. Entomol. 29: 176-81. Yates, W. E. 1960. Minimizing Spray drift hazards. Down to Earth. 16(2): 15-9. Zentmyer, G. A. 1942. Toxin formation by Ceratostomella ulmi. Science (Wash., 0. C.). 95: 512-3. Zentmyer, G. A., J. G. Horsfall, and P. Wallace. 1946. Dutch elm dis- ease and its chemotherapy. Conn. Agric. Exp. Stn. Bull. No. 498.